Pipeline Welding: Processes, Codes, and Field Practice

A girth weld on a 36-inch natural gas line passed the welder’s own visual inspection and the coating crew’s prep check, and still got cut out two days later when the radiograph came back showing a root defect nobody could argue with. That rejection traces back to something that happened maybe forty minutes earlier, on a truck in the dark, on a joint nobody wanted to stop and clean. Pipeline welding is a trade where a chain of small, boring decisions — a bevel wiped down, a joint aligned, an electrode dried — ends up deciding whether a weld stays in the ground for fifty years or comes back out on a truck. Here is how that chain actually works, from the codes that govern it down to the arc.

What Pipeline Welding Actually Is (and Why It Is Nothing Like Structural Welding)

Pipeline welding is joining steel pipe that will hold pressure for fifty years. That single fact explains most of what separates it from other welding. A structural welder connects a beam to a column in a building an inspector can walk up to any time. A pipeline welder connects two lengths of line pipe in a trench, in weather, sometimes over a road, and then the pipe gets buried. That joint is never seen again unless something goes wrong.

A pipeline welder aligns two pipes, lands a stringer or root bead, fills and caps the joint, then moves on. On a modern spread, a handheld stick process still does most of the field work, and one joint may take an hour or more of arc time in any weather.

In regulatory terms, pipeline welding is procedure-driven: every weld traces to a qualified procedure and a qualified welder, and every critical weld is inspected by a qualified, certified NDT technician. Nothing gets welded on a pipeline because it looked like a good idea in the moment. A written procedure states the process, consumable, parameters, and inspection, and it was qualified and recorded.

The Anatomy of a Pipeline

Two very different welds exist in a pipeline system.

• Girth welds join two lengths of line pipe end to end. These are the field welds, and the subject of nearly every pipeline welding procedure you will ever read.
• Longitudinal mill welds join the pipe to itself along its axis as it is formed in the mill, usually by submerged arc welding, under factory conditions and inspected before the pipe ships to the job site.

A complete system is more than a long run of pipe. Compressor and pump stations add piping, headers, and valves. Tie-ins connect a new line to an existing one, meaning cutting into a live system or matching a line in service for decades. Branch connections carry smaller lines off the main. Each is a joint somebody must weld, with its own geometry and paperwork.

Two Joint Geometries, Two Construction Contexts

Two joint geometries dominate. The single-V butt joint is the girth weld: both pipes are bevelled, aligned, and welded in multiple passes. The branch or socket joint connects a smaller pipe into the side of a larger one, as a socket weld or a full penetration branch connection.

Two construction contexts go with them. New-build line pipe welding happens on an empty right-of-way with a spec, a procedure, and a fresh crew. In-service repair happens on an operating asset carrying product right now, sometimes inside a shutdown window you have to hit exactly. In-service work gets its own section later.

Why It Is Unforgiving

A pipeline weld is a pressure-boundary element, not a structural member. Pipe under internal pressure develops hoop stress, a circumferential pull that tries to split it open lengthwise, tied directly to the pipe’s specified minimum yield strength (SMYS). Federal rules use the 20 percent SMYS threshold to help decide how rigorously welds must be tested.

Add decades of cyclic pressure, buried in soil or on the seabed, and the consequence chain follows. A defect can sit there quietly for twenty years, undetectable from the surface, until it becomes a leak or a rupture. Nobody looks at that weld again.

Vocabulary You Will Hear Immediately

Term	What it means in the field
SMYS	Pipe’s rated strength for stress and design calculations
Girth weld	Field weld joining two pipe lengths end to end
Root pass	First bead at the bottom of the bevel
Hot pass	First fill pass over the root; cleans it and builds sound weld metal
Fill and cap pass	Passes that build the joint and form the outer bead
Weld ID number	ID tying a joint to its procedure, welder, and inspection record
Location class	Classes 1 to 4, based on proximity to buildings, roads, railways, and waters

Location classes matter more than they sound. The closer a weld sits to populated areas, railways, or waterways, the more rigorous the testing. Class 1 through Class 4 designations drive the percentage of welds that must be nondestructively tested, so the same joint on a remote gathering line and the same joint in a city crossing carry different inspection burdens.

Where the Standard Comes From

The pipeline-specific welding standard is API Standard 1104. It covers butt, branch, and fillet welds in carbon and low-alloy steel pipe carrying crude petroleum, petroleum products, fuel gases, carbon dioxide, and nitrogen, and applies to distribution systems where applicable. It applies to both new construction and in-service welding, and permits shielded metal arc, submerged arc, gas tungsten, gas metal, flux cored, plasma arc, and oxyacetylene welding, performed manually, semiautomatically, or mechanized. That is the frame for everything that follows: qualified procedures, qualified welders, traceable weld identification, and third-party inspection wrapped around a joint that has to hold pressure for the next fifty years.

The Codes That Govern Pipeline Welding: API 1104, ASME IX, and the CFR

Most pipeline welding confusion starts with the assumption that one rulebook governs the job. There is not. There is a stack, and each layer rests on the one below. Get the stack wrong and you qualify a welder against one code, weld a joint another rejects, and find out in the radiographic film.

The Regulatory Stack

At the top sits federal law: 49 CFR Part 192, Subpart E for interstate gas pipelines, and Part 195 for hazardous liquid pipelines. They are enforceable, and they are what an inspector measures your work against.

The next layer down is the technical standards. 49 CFR 192.225 requires welding by a qualified welder or welding operator using procedures qualified under API 1104 Section 5 (with one exception), Section 12, Annex A, or Annex B, or ASME BPVC Section IX. Only Sections 5, 6, 9, and 12 plus Annexes A and B of API 1104 are referenced by Part 192, so the rest of API 1104 is not federal law. One restriction catches experienced crews off guard: API 1104 Annex A may not be used to accept cracks.

The hierarchy runs federal law, then the API 1104 and ASME Section IX standards it incorporates, then owner and company specifications. That last layer causes the most headache, because operator specs are usually more demanding than the code: a transmission operator may require radiography where the code permits ultrasonic testing. Passing the code is the floor, not the finish line.

The Paperwork Chain: WPS, PQR, and WPQR

Three documents, three different jobs.

• WPS (Welding Procedure Specification) is the instruction: process, filler metal, joint design, amperage range, voltage, travel speed, preheat, shielding gas, essential variables.
• PQR (Procedure Qualification Record) is the evidence: test coupons were welded to that WPS, tested, and passed.
• WPQR (Welder Performance Qualification Record) is proof about a person: this welder passed the test for this process, position, and material.

Here is the rule that catches crews: a WPS without a supporting PQR is not a qualified procedure. If the qualification test was never run, failed, or never recorded, welds made to it are open to rejection. 49 CFR 192.227 requires each welding procedure to be recorded in detail including the results of the qualifying tests, and requires that record to be retained and followed.

Continuity and Requalification

Qualification is not permanent. Under 49 CFR 192.229, a welder may not weld with a given process unless within the preceding six calendar months they were engaged in welding with it, or performed a weld with it that was tested and found acceptable within the preceding seven and a half months.

For pipe operating at 20 percent or more of SMYS hoop stress, the bar is tighter. The welder must have one weld tested and found acceptable within the preceding six calendar months, or maintain qualification with welds tested and accepted at least twice each calendar year, at intervals not exceeding seven and a half months.

Absolute limits apply too. A welder whose qualification is based on nondestructive testing may not weld compressor station pipe and components. Records are not optional either: welder qualification records for steel transmission pipe installed after July 1, 2021 must be kept at least five years following construction.

What Changed in the 21st and 22nd Editions

API 1104 is revised on a regular cycle, and the API announcement page tracks each edition. The 21st edition was incorporated by reference into 49 CFR 192 in April 2024, with the related stay of enforcement ending in January 2025. The 22nd edition was published in July 2021 and is what most current work references.

The 21st edition introduced new SMYS ranges aligned with API 5L pipe grades, eliminated flattening bend test specimens, and tightened the treatment of shielding gas flow rate changes greater than 20 percent, so gas flow drift can now knock a procedure out of compliance. It revised the essential variables for filler metal group changes, and added a practical rule: a welder qualified in the fixed position is also qualified for rolled positions.

That last rule is a quiet time-saver. An all-position electrode guide shows which rods pass 1G, 2G, 5G, and 6G, but the code’s position rules decide which test results translate into which real jobs, which matters more than rod-to-position charts.

Finally, formal requests for interpretation are an official, published channel. If a client argues a technical point, that process exists for exactly that situation. The API 1104 22nd edition preview and the Wisconsin PSC presentation on 21st edition changes show what moved and why.

Learn the stack once and the rest of pipeline welding gets easier. When an inspector asks why your WPS has no PQR, you will know which layer the question belongs to.

Choosing a Welding Process: SMAW, FCAW, GTAW, GMAW, and SAW Compared

On a pipeline spread, process selection is less about theoretical perfection and more about practicality. Joint position, weather, impact-tested requirements, and on-site equipment all play a role. Field research on pipeline welding and coating found shielded metal arc welding to be the most common process, used by about 50 percent of survey respondents in a survey of pipeline welding and coating practice. Its dominance comes from reliability, not raw productivity.

Process	Typical deposition rate	Diffusible hydrogen	Where it fits on a pipeline
SMAW, cellulosic (E6010, E7010)	0.5 to 3 kg/hr	roughly 30 to 60 ml/100 g	Manual girth welds, all positions, standard field work
SMAW, basic (dried)	0.5 to 3 kg/hr	under 5 ml/100 g	Higher-strength or sour service, impact-tested connections
FCAW, gas-shielded (FCAW-G)	3 to 10 kg/hr	under 4 ml/100 g	Fill and cap, mechanized and hybrid root-to-cap
FCAW, self-shielded	3 to 10 kg/hr	typically 8 to 15 ml/100 g	Remote or windy sites, generally not impact-tested joints
GMAW (solid wire)	2 to 6 kg/hr	under 3 ml/100 g	Root or hot pass where automation is available
GTAW (TIG)	0.5 to 2 kg/hr	under 2 ml/100 g	Clean, controlled root pass on thin wall or critical joints
SAW (submerged arc)	5 to 25 kg/hr single wire, up to about 100 kg/hr tandem	Process dependent	Mill longitudinal welds; flat and horizontal fillet positions only

Cellulosic Stick: Built For Girth Welds

For manual field girth welds, E6010 and E7010 cellulosic electrodes remain the backbone for good reason. Their arc digs deep into the joint to handle imperfect bevels, while the fast-freezing slag allows a welder to lay down narrow, vertical-down passes. This combination is what keeps them viable in every position encountered on pipe, including 5G and 6G. These traits are broken down further in our all position welding rod guide and in why SMAW still wins.

The Hydrogen Problem

Cellulosic coatings are hygroscopic, meaning any absorbed moisture releases hydrogen into the arc. From there, it finds its way into the weld metal and heat-affected zone. The spread is stark: cellulosic SMAW runs roughly 30 to 60 ml/100 g, dried basic SMAW stays under 5 ml/100 g, FCAW-G under 4, GMAW under 3, GTAW under 2, and self-shielded FCAW typically between 8 and 15 ml/100 g.

What makes this tricky is timing. Hydrogen can continue to diffuse out of a solidified weld for hours, lingering long enough to combine with susceptible microstructures and cause hydrogen-assisted cracking days later. Low-hydrogen consumables mitigate this, but only if they’re stored and baked to manufacturer requirements, as moisture pickup will raise hydrogen levels regardless of their classification.

Preheat and the Carbon Equivalent

Process choice is often a cracking-control decision once carbon equivalent comes into play. The IIW formula is CE = %C + Mn/6 + (Cr + Mo + V)/5 + (Ni + Cu)/15, and API 5L X65 line pipe typically sits at approximately 0.38 to 0.43 by that formula. Above roughly 0.40 to 0.45, preheat is generally called for to suppress hydrogen-assisted cold cracking in the heat-affected zone. The highest risk scenario is exactly the most common one: cellulosic electrodes on X65 in cold weather with little to no preheat. Low-hydrogen consumables and controlled preheat are the two levers used to pull that risk down. For a deeper look at metallurgy, see our guide to alloy steel welding electrodes.

Why SAW Stays at the Mill

Submerged arc welding is the productivity champion, reaching 5 to 25 kg/hr with a single wire and up to roughly 100 kg/hr in tandem multi-wire configurations. However, its granular flux blanket cannot be retained on vertical or overhead surfaces. For that reason, SAW is restricted to flat (1G) and horizontal fillet (2F) positions, ruling it out for field girth welds entirely. This is exactly why the longitudinal seam of mill-produced API 5L pipe is routinely submerged arc welded, while field joints are not.

Mechanized and Hybrid Approaches

On larger spreads, crews often rely on combinations rather than any single process. Internal root machines, mechanized GTAW root passes followed by FCAW-G fill and cap, or tandem FCAW are all common. Research on high-strength line pipe has demonstrated mechanized GTAW root passes paired with tandem pulsed FCAW-G fill and cap passes, showing considerably lower fill and cap arc time than alternative processes in modern pipeline welding practice. Across thousands of girth welds, that consistency is what justifies the investment.

Those gains come with constraints. Under 49 CFR 192.227, the procedure, equipment, and welders are qualified as a system under API 1104, so even a change in wire size, gas mix, or machine setup can invalidate that qualification. Gas-shielded processes such as GMAW and FCAW-G are also vulnerable to wind disrupting their shielding gas, while SMAW remains the most tolerant of open, draughty conditions. Self-shielded FCAW ignores the wind, but its higher diffusible hydrogen generally keeps it off impact-tested connections.

Joint Preparation, Fit-Up, and Consumable Handling

The arc gets the attention, but the radiograph tells the truth, and it often blames something that happened before the arc lit. A girth weld can look perfect and still fail for porosity if a damp electrode sat in the truck overnight. None of that shows in a photo of the finished joint, which is why preparation gets rushed.

What the Regulations Say Before the Arc Strikes

Federal law is blunt. 49 CFR 192.235 requires that, before welding begins, the welding surfaces be clean and free of any material detrimental to the weld, and that the pipe be aligned to give the most favorable condition for depositing the root bead, with that alignment preserved while the root bead is deposited. Clean it, straighten it, and do not let it move during the hardest pass.

API 1104 Section 7 turns that into the working specification for production welds, covering joint design, alignment, lineup clamps, bevel geometry, weather and clearance, cleaning between beads, position and roll welding, weld identification, and preheat, interpass, postheat, and PWHT requirements. The current edition of API 1104 is worth owning, because the numbers in it are the numbers an inspector uses against your crew.

The Preparation Sequence, in Order

1. Verify material. Confirm pipe grade and wall thickness against the line list before cutting or beveling.
2. Check bevel geometry. Angle, root face, and root gap must sit inside the ranges the qualified procedure specifies. Out-of-spec bevels get ground, not welded around.
3. Clean the bevel. Mill scale, rust, paint, thread compound, and hydrocarbons all come off. They drive porosity, slag inclusions, and hydrogen cracking even when every parameter is right.
4. Fit and align. Set hi-lo and root gap inside procedure tolerance, then lock the joint with internal or external lineup clamps so nothing shifts under the welder.
5. Tack, then clean between passes. Tack welds are inspected and cleaned like any other deposit. Between passes, remove slag, spatter, and grinding burrs.

Misalignment and root gap cause a large share of rejects. When the high side sits outside the qualified hi-lo tolerance, the welder has two bad choices: bridge it with a slow, dragging weave that traps slag and risks cold lap, or burn through the low side to keep the bead moving. A root gap set too wide forces extra filler and heat into the root; too narrow, the root pass fuses poorly and cold laps. The compensation happens in the welder’s body, and the inspector sees it.

Consumable Handling Is a Welding Parameter

Consumables are treated like procedure variables because they behave like them. Electrodes and wire must be stored dry and protected from the air, and cellulosic electrodes need baking and holding in a rod oven at the manufacturer’s set temperature. Moisture in a cellulosic coating raises diffusible hydrogen and degrades arc characteristics: wet, spattery arc, poor fusion, undercut. A rod oven in the weld truck is not a convenience; it is the difference between a qualified procedure and an unqualified one. For electrode selection by position, start with our all-position welding rod guide and our take on the advantages of stick welding in the field.

Wire gets the same treatment: sealed packaging, controlled storage, clean handling of the spool face. A contaminated drive roller or a wet spool is a porosity generator. SAW flux must be kept dry and free of hard clumps, stored per the manufacturer’s instructions. Shielding gas has its own discipline: correct purity and mix, dry regulators and lines, delivery pressure checked before the joint starts, and wind protection over the arc. Cylinder and supply handling details are covered in the welding supplies buyer’s guide.

The Link Between Weld Profile and Coating

A weld that passes inspection can still fail years later, and geometry is one reason. Excessive cap height or a steep reinforcement angle creates tenting: the tape coating or heat-shrink sleeve bridges away from the pipe surface instead of conforming to it, leaving a hollow gap over the weld. Water and chlorides collect there, and corrosion attacks the most protected point on the line. Holding cap profile to your procedure limits is part of the corrosion plan, not extra work.

Pre-Weld Checklist

• Grade and wall thickness verified against the line list
• Bevel angle, root face, and root gap within procedure limits
• Bevel free of mill scale, rust, paint, thread compound, oil, grease, and moisture
• Hi-lo inside tolerance; joint clamped or tacked
• Alignment held steady through the root pass
• Lineup clamps in place, spotters briefed
• Consumables dry, in-date, and from a tracked lot
• Preheat measured on the pipe, interpass limit known
• Shielding gas mix and flow verified
• Weld ID assigned before the first arc

API 1104 requires welds to be identified so each production weld can be traced to its welder, its procedure, and its test results. That number is assigned at the joint before the first arc, because reconstructing identity afterward is guesswork.

Positions, Heat Input, Preheat, and Interpass Control

A pipeline weld that passes radiographic inspection and one that gets cut out are separated by what happens in the few minutes the arc is live. The welder manages four things at once: joint position, heat going into the steel, metal temperature at the start of each pass, and metal temperature at the end of each pass. Get one wrong and the heat-affected zone pays for it years later.

The Four Pipe Positions You Need to Know

Welding positions come from the American Welding Society system, with pipeline shorthand layered on top. Since the pipe does not move once it is in the trench, position is named by where the weld sits relative to the pipe.

Position	What It Looks Like	Where It Shows Up
1G	Pipe rotated, weld runs flat across the top	Shop or spool fabrication; field joints where the pipe can be turned
2G	Horizontal fixed pipe, weld oriented horizontally	Roll-welded field joints on some spreads
5G	Horizontal fixed pipe, weld running vertically	Standard field girth weld position
6G	Pipe inclined at 45 degrees, vertical-inclined weld	Hills, river crossings, joints that cannot be rotated

Field crews live in 5G and 6G. API 1104 lets welders qualify by position and roll welding, or by a combination, and the 21st edition added a simplification: a welder who qualifies in the fixed position also qualifies in the rolled position.

Why Position Changes Everything

Position decides what the molten metal wants to do and how fast it freezes. In 1G, gravity pulls the pool into a flat puddle the welder can shape with the torch. In 5G, the bead runs vertically, so the pool must be held against gravity through the root pass, then filled and capped. In 6G, the joint is inclined and the pool drifts both down and sideways.

• Pool control. Cranking amperage to keep up with a fast vertical-down bead pushes the pool past bridging, producing undercut at the edges and trapped slag in the middle.
• Cooling rate. A vertical-up bead, or one stacked too high, cools slower because the pool is large and exposed. Slower cooling through the transformation range means coarser grains and a harder, less tough heat-affected zone.
• Direction of travel. Gravity encourages downward travel, exactly what cellulosic electrodes are built for.

Why Cellulosic Electrodes Dominate Field Girth Welds

Cellulosic electrodes such as E6010 and E7010 remain the backbone of manual field pipeline welding. Their coating produces a deep-penetration arc and a fast-freezing slag. Deep penetration gets the root bead through a wall that can exceed an inch, and the fast freeze lets the welder lay a short bead and move downward without the pool running out from under the arc. That makes vertical-down progression practical, and vertical-down is the field standard: it gets the welder out of the trench faster, leaves slag where it can be chipped while hot, and produces a bead profile inspectors and automated UT equipment recognize.

Cellulosic consumables demand discipline because the coating pulls a large amount of moisture into the weld. That hydrogen is why the preheat and interpass controls below are not optional. Our all-position welding rod guide covers the electrode families and their storage requirements in more depth.

Heat Input: The Number That Drives Everything Else

Heat input is the energy the arc delivers to the joint, in kilojoules per millimeter of weld. The standard field calculation is:

kJ/mm = (voltage × amperage × 60) ÷ (travel speed in mm/min × 1000)

For a typical SMAW girth weld at 24 V, 130 A, and 180 mm/min travel speed:

(24 × 130 × 60) ÷ (180 × 1000) = 187,200 ÷ 180,000 = 1.04 kJ/mm

Run the same current at 240 mm/min and heat input drops to 0.78 kJ/mm, still inside a 0.60 to 1.10 kJ/mm procedure range but only just. Push the current to 150 A at 180 mm/min and you reach 1.20 kJ/mm, over the limit and in rework territory.

Codes set both a maximum and a minimum because heat input cuts two ways. Too much heat gives a wide, coarse-grained heat-affected zone that is hard and brittle. Too little gives a cold, sluggish fusion zone with incomplete penetration and poor sidewall fusion. The heat-affected zone is the ring of parent metal beside the weld that never melted but was heated and cooled, so its width and grain structure are set by how much heat goes in and how fast it comes back out. Most structural codes set a maximum specifically for impact-tested applications, while API 1104 establishes its own limits through the qualified procedure rather than one blanket number.

Preheat, Interpass, Postheat, and Weather

Preheat is the temperature the parent metal reaches before the first arc is struck. Interpass temperature is the minimum temperature of the weld metal plus adjacent base metal when the next pass begins. Together they are the procedure’s main control over cooling rate and the standard mitigation for hydrogen-assisted cold cracking. Hydrogen from cellulosic coatings, moisture, and wet consumables diffuses into hot, freshly deposited weld metal. If that metal cools quickly through the transformation range while still saturated with hydrogen, it becomes trapped and forms a hard, crack-sensitive phase. Under the restraint of a thick-walled joint, that combination can produce a delayed in-service crack. Preheat slows cooling to buy hydrogen time to escape; interpass keeps that control running pass after pass. For high-strength, high-restraint line pipe, these are often the difference between a sound joint and a crack a federal pipeline operator will find on an in-line inspection tool years later. API 1104 Section 7.10 governs preheat, interpass temperature, postheat, and PWHT requirements, all established by the qualified procedure.

Postheat means holding the completed weld at elevated temperature for a specified period. PWHT is a controlled furnace or induction cycle that relieves residual stress and lets hydrogen diffuse out under a slow cool. Field girth welds rarely use PWHT due to joint geometry and time constraints, so codes often lean on restraint control, hydrogen control, and cooling rate instead. Weather adds another variable: wind strips the shielding of gas or electrode covering and can cause porosity and reduced fusion. Cold and wind also pull heat out faster, raising hydrogen cracking risk and hardening the heat-affected zone, which is why preheat requirements climb in cold weather and why damp or freezing conditions trigger extra precautions.

The Troubleshooting Connection

Look at a rejected joint and the heat input number usually explains it. Too much heat gives undercut, a wide hard heat-affected zone, grain growth at the fusion boundary, and slag trapped from an over-weaved or over-ampered bead. Too little heat gives lack of fusion, incomplete penetration, cold lap, and a sluggish pool that will not wet the sidewall. The same welder on the same joint, on a different day, with different weather and different electrode conditioning, can land on either side of that line. That is why heat input limits live in the qualified procedure rather than in the welder’s memory, and why API 1104 treats it as a controlled variable rather than a guideline.

For anyone working around an operating pipeline, in-service welds that contact the carrier pipe are particularly susceptible to underbead or delayed hydrogen cracking, so the inspection method should be able to detect those cracks at the carrier pipe weld toe. That is a strong argument for radiographic testing, covered in more detail in the inspection and repair sections that follow.

I’ll write section 6 on inspecting pipeline welds.

Inspecting Pipeline Welds: Visual, Radiography, UT, MT, and PT

“It was inspected” is a phrase pipeline crews use loosely, and it hides a lot. A weld can pass a visual look-over by the welder who ran it, get a thumbs-up from the coating crew’s prep check, and still be excavated six months later. Inspection is not a single event. It is a chain of separate tests, performed by different people with different tools, answering different questions about the same joint. Some of those tools are looking for one specific flaw. Some are looking for a different one. Knowing which is which is the difference between a repair schedule and a surprise shutdown.

Visual Inspection: The Baseline Nobody Skips

Before any instrument touches a girth weld, it gets a visual examination. Under 49 CFR 192 Subpart E, visual inspection of girth welds is required to be performed by an individual qualified by appropriate training and experience. That last phrase carries real weight. The welder who laid the bead is not qualified to self-certify their own work, and neither is a coating applicator whose prep check is about surface condition, not weld integrity.

The visual check looks at geometry and surface condition: weld reinforcement profile, cap height, undercut, overlap, burn-through, arc burns, excessive spatter, misalignment, and evidence of cracking visible to the naked eye. It also checks for arc strikes outside the weld, which are stress concentrators and, on some carriers, grounds for removing the joint outright.

The NDT Toolbox: What Each Method Sees

API 1104 covers four nondestructive methods: radiographic testing, magnetic particle testing, liquid penetrant testing, and ultrasonic testing. Each one has a specific job.

Method	Detects well	Misses or struggles with
Radiography (film or digital)	Volumetric defects: porosity, slag, lack of fusion, incomplete penetration, cracks in a favorably oriented condition. The only method that shows the full thickness of the weld in one image.	Defects masked by geometry, very fine indications, and it cannot quantify depth reliably. Also position-dependent for cracks.
Ultrasonic testing (manual and automated)	Planar and reflectors in the weld metal and HAZ: cracks, lack of fusion, lack of penetration, and deep slag. Excellent near-surface and toe sensitivity.	Porosity scatters the beam and masks other indications. UT acceptance criteria do not readily evaluate porosity.
Magnetic particle testing (MT)	Surface and near-surface discontinuities in ferromagnetic material: cracks, toe cracks, and sharp surface defects. Fast and inexpensive on welds and on the carrier pipe itself.	Only works on ferromagnetic materials. Cannot see subsurface defects, and cannot evaluate non-magnetic alloys.
Liquid penetrant testing (PT)	Fine surface-breaking cracks and flaws on clean, non-porous, non-textured surfaces. Very sensitive to tight surface cracks.	Surface only. Requires a clean surface, which is often the limiting factor in the field.

That UT-versus-radiography mismatch is worth dwelling on, because it is a real and recurring source of production rework. API 1104 acceptance criteria for ultrasonic testing do not readily evaluate porosity, so a weld accepted by UT can still be rejected under radiography for excessive porosity. On a job where the owner specifies UT for speed and the code allows radiography as an alternative for verification, crews have learned the hard way that a joint can clear an automated UT gate and fail a radiograph a day later. Magnetic particle, UT, or a combination of both have also been shown effective at detecting hydrogen cracks at the toe of sleeve, saddle, and branch-to-carrier pipe welds, which is why MT and UT are the tools of choice around branch connections.

How Much Testing Is Actually Required

The federal rule sets a floor, not a target. Welds on a pipeline operated at 20 percent or more of SMYS hoop stress must be nondestructively tested. There are two practical exceptions: pipe under 6 inches nominal diameter, when approved by a qualified welding inspector, and low-stress lines where there are too few welds for testing to be practicable.

Beyond that baseline, 49 CFR 192.227 sets required NDT coverage by location class:

• At least 10 percent of each day’s field butt welds in Class 1 locations, except offshore.
• At least 15 percent in Class 2 locations.
• 100 percent in Class 3 and Class 4 locations at crossings of major or navigable rivers, offshore, and within railroad or public highway rights-of-way. Where 100 percent proves impracticable, at least 90 percent.
• 100 percent of all pipeline tie-ins, including tie-ins of replacement sections.

On top of the sampling percentage, there is a per-person rule that often gets overlooked. Except for a welder whose work is isolated from the principal welding activity, a sample of each welder or welding operator’s work for each day must be nondestructively tested when NDT is required. This means a welder on a 10 percent sampling job still gets tested on their own joints, and a new welder on day one gets tested on day one.

Records matter just as much as the testing. Each operator must retain, for the life of the pipeline, a record showing, by milepost, engineering station, or geographic feature, the number of girth welds made, the number nondestructively tested, the number rejected, and the disposition of the rejects. That last item is the reason the record exists. It is an audit trail back to a specific joint in a specific trench, decades later.

Reading the Results: Defect Vocabulary and Disposition

Acceptance is determined per API 1104 Section 9 or Annex A. Defect type drives disposition, because each defect carries a different risk.

• Porosity shows on a radiograph as small round or irregular light-toned spots, clustered or scattered through the bead. Caused by moisture, oil, rust, contaminated gas, drafts, or excessive arc length.
• Slag inclusions appear as dark, irregular, sometimes elongated areas oriented along the weld axis. Slag trapped between passes.
• Lack of fusion shows as an unfused, unbonded edge along the sidewall or between passes, a dark line or irregular area at the fusion boundary.
• Incomplete penetration appears as a continuous dark gap along the center of the weld where the root bead failed to fuse with the root or the opposite wall.
• Undercut shows as a groove or notch at the weld toe. On a radiograph it can be subtle or invisible, which is a good reason to pair radiography with MT at the toe.
• Cracks appear as sharp, linear indications with a dark, tapered profile. Cracks are the defects that get the least latitude.

Disposition follows one of two paths. Either the joint is rejected and repaired, or it is accepted under an alternate criteria path. API 1104 Annex A provides alternative acceptance standards for girth welds based on a fitness-for-purpose assessment, including CTOD-based imperfection limits. Annex A is not a loophole for cracks, though. It may not be used to accept cracks.

Understanding this section keeps the whole job in perspective. If you want a feel for how the process choices feeding into inspection are made, our guides to all-position welding rod selection and why SMAW still wins in the field connect directly to what the radiographer sees on the other end. And for the repair side of the picture, air arc welding and gouging is how defective weld metal actually comes back out before the repair deposit goes in.

The Bottom Line

Inspection does not prove a weld is sound. It proves the weld meets a defined criteria set, using methods that each have known blind spots. The right method for the job, an honest record of what was found, and a disposition that respects the defect type are what actually protect the pipeline.

In-Service Welding, Repairs, and Hot Tapping

Pipeline crews talk most about the welds someone had to cut out. New construction is the clean case: open trench, dry pipe, a qualified procedure. In-service work is the opposite. The pipe carries product, maybe under a road or a river, nobody knows what happened to that steel in 1978, and the schedule says it must be back in service before the next compressor comes online.

That is why the standards treat in-service welding as a separate discipline. API 1104 lays out three regimes. New construction follows the main body; in-service welding falls under Annex B, the normative annex that governs repairs and hot tapping on live assets. It is not optional reading, and it layers requirements on top of the Section 8 inspection and testing rules for new construction. Federal rules match: 49 CFR 192 Subpart E incorporates API 1104 for pipeline welding, and 49 CFR 192.227 requires repaired pipe to be restored to a condition safe for operation before it returns to service.

Three different jobs, three different rulebooks

Paperwork, margins, and acceptable outcomes all differ. Blur the categories and you defend a repair weld against an inspector using the wrong criteria.

Work type	Code path	Condition during welding
New construction	API 1104 main body	Not in service, or isolated and depressurized
In-service repair	Annex B, Section 9 criteria	In service or partly depressurized, approved plan
Hot tapping	Annex B plus operator hot tap procedure	Live, pressurized, flowing system

Two Annex B details matter. In-service welds contacting the carrier pipe are especially vulnerable to underbead or delayed hydrogen cracking, because the weld is deposited into a restraint and hydrogen environment the joint has not been stress-relieved of. That is why magnetic particle testing, ultrasonic testing, or both are recommended for detecting cracks at the carrier pipe weld toe. Separately, Section 9 acceptability criteria apply to in-service welds, while Section 10 requirements for repair and removal of defects apply alongside.

The repair cycle, step by step

API 1104 Section 10 lays out the repair sequence, and the order is where shortcuts cause trouble.

1. Expose and verify the defect. Uncover the area, then confirm the indication with the specified NDT before anyone touches steel.
2. Mark the limits of removal. Grind or gouge out defective weld metal plus a specified amount of sound parent metal. Carbon arc gouging is common because it cuts fast and leaves a surface that welds cleanly.
3. Prepare the excavation. Bevel the cavity for a sound fusion path, clean the surfaces, confirm fit-up.
4. Preheat to the required temperature. Non-negotiable in service work, because the surrounding pipe is a massive heat sink that cools the repair fast.
5. Deposit the repair weld. Usually SMAW with a cellulosic or low-hydrogen electrode, which is why SMAW remains the backbone of field girth welding even on modern projects.
6. Post-weld heat treatment if specified. Field repairs on line pipe rarely call for PWHT, but it happens on heavier or sour-service base metal.
7. Re-inspect. Section 10.6 covers NDT and weld repair acceptance criteria, and the repair weld gets the same scrutiny as the original.

Section 10.4 sets repair welder qualification, and it is stricter than general qualification: qualified for the specific position, process, and joint configuration of the repair. That is why crews carry a range of all-position electrodes, and why repair work often sends a welder back to the test coupon.

The wall thickness constraint

This arithmetic governs every in-service decision. Removing a defect removes sound metal with it, leaving a thinner carrier pipe than the operator has been running. Wall thickness must not drop below what is acceptable for the operating pressure of the carrier pipe. If the remaining wall will not carry design pressure with appropriate margin, the repair is not a weld repair at all. It becomes a pipe replacement or a reinforced sleeve. API 1104 also limits repeated repairs at the same location and requires approval before a third pass, because every excavation takes more parent metal out of the pipe.

Why in-service welds behave differently

• Hydrogen at the carrier pipe weld toe. With restraint and fast-cooling base metal, the classic recipe for delayed cracking, sometimes appearing days after the repair passed inspection.
• Residual stress. A repair bead shrinks the surrounding live pipe toward itself, adding load to the area that was already suspect.
• Unknown material condition. The pipe may be older than the weld procedure, repaired before with a different process, and may carry unknown toughness or sour-service resistance.
• Coating damage. Gouging and grinding strip the coating and adjacent surface preparation. Pipeline coating failures are frequently traced to poor coordination between welding, surface preparation, and coating steps, not the coating material itself.
• Cathodic protection interaction. Welding and stray current disturb the cathodic protection system, and overprotection in some environments drives hydrogen into the steel.

Qualifying a repair procedure under abnormal conditions is harder than it sounds. The joint is a cavity with irregular edges, base metal may not match the plate, and a cramped trench pushes welders toward vertical and overhead positions. API RP 577 is the practical companion reference, covering welding processes, inspection, and metallurgy for in-service pipeline repair and alteration work.

Hot tapping: welding on a live system

Hot tapping means putting an arc on a pressurized, flowing system. It is the highest-consequence operation in pipeline welding and needs its own planning, not a call from the superintendent at 4 p.m.

Before the arc is struck, the operator needs a written plan. Pressure must be reduced or the segment isolated so a leak during welding becomes a controlled event. The line may need plugging, bleeding, or diverting product to reduce the hazard of an ignition source meeting a combustible atmosphere. Gas testing establishes that the work area is safe to enter and weld in. A fire watch with no other duty is assigned and stays until the area is confirmed clear, and an emergency response plan names who shuts the tap, who isolates the segment, and how the job is abandoned if conditions change.

Across all three regimes, in-service work shifts the burden from the weld to the paperwork. The arc is the easy part. What decides whether a repair survives an audit, an inspector, and the next ten winters is whether the procedure was qualified, the welder was qualified, the NDT was performed by a certified technician, the wall thickness math was checked, and the documentation was assembled while the trench was still open.

Safety and Health: Fume, Gas, Heat, and the Confined-Space Problem

Welding fume is the price of the arc, and a technical problem before it is a health problem. A pipeline crew closing a spread in August manages metal oxides, shielding gas, radiant heat, and a dozen non-fume hazards at once. Safety handled as disclaimers taped to a procedure fails; handled as measurable exposures with known controls, it works.

What Is Actually in Welding Fume

The OSHA welding fact sheet lists the metals found in welding fume: aluminum, antimony, arsenic, beryllium, cadmium, chromium, cobalt, copper, iron, lead, manganese, molybdenum, nickel, silver, tin, titanium, vanadium, and zinc. What shows up depends on the base metal and the consumable. Carbon steel girth welds give mostly iron oxide fume with manganese riding along, a stainless repair adds chromium and nickel, and burning off galvanized coating adds zinc oxide.

The gases matter just as much. OSHA names shielding gases, nitrogen oxides, carbon monoxide, ozone, phosgene, hydrogen fluoride, and carbon dioxide. Some arrive with the gas supply, some form in the arc, and phosgene and hydrogen fluoride form when an arc hits a surface carrying solvents, paints, or fluoropolymer coatings. Striking an arc near a just-cleaned, solvent-wiped joint is one of the fastest ways to poison a crew.

As NIOSH notes in its welding fume overview, welding fume is composed of metals and may contain small amounts of manganese, with exposure varying according to the manganese content of the welding wire, rods, flux, and base metal. Internalize that: the consumable on the truck drives much of the exposure, not just the material in the pipe.

The Numbers That Govern Exposure

Hexavalent chromium and manganese drive most of the regulatory attention.

Contaminant	Limit	Authority
Hexavalent chromium	5 µg/m³ as an 8-hour TWA; action level 2.5 µg/m³	OSHA permissible exposure limit
Hexavalent chromium	0.2 µg/m³ as an 8-hour TWA over a 40-hour workweek	NIOSH recommended exposure limit
Manganese	5 mg/m³ ceiling	OSHA permissible exposure limit
Manganese	1 mg/m³ TWA, 3 mg/m³ short-term limit	NIOSH recommended exposure limit
Manganese	0.02 mg/m³ respirable; 0.1 mg/m³ inhalable	ACGIH threshold limit value
Manganese	500 mg/m³ (immediately dangerous to life or health)	NIOSH IDLH

One thing jumps out: the authorities disagree by an order of magnitude on manganese. OSHA’s legacy 5 mg/m³ ceiling is roughly 250 times less protective than ACGIH’s 0.02 mg/m³ respirable TLV. For the neurologic risks chronic manganese exposure brings, design to the modern guidance, not the older ceiling.

Where Hexavalent Chromium Comes From

Chromium metal is not the hazard. Hexavalent chromium is. During welding, chromium in stainless base metal, chrome-bearing consumables, and chromate-containing primers converts to the hexavalent form. Cr(VI) fume damages the eyes, skin, nose, throat, and lungs and causes cancer, which is why NIOSH treats all Cr(VI) compounds as occupational carcinogens and sets its recommended limit 25 times below the OSHA permissible limit. Most pipeline carbon steel carries little chromium, but stainless connectors, repair sleeves, and plant piping show up often enough that crews should know what is in the joint before they start.

The Confined-Space Multiplier

Confined space welding significantly increases exposure to manganese fumes. A welder inside a vessel or small-diameter line has nowhere for the plume to go, so it recirculates and concentrates with every pass, while a welder on an open trench can usually catch a cross-breeze. That is the argument for treating confined space welding as an engineering problem with an attendant, not a permit with a signature on it.

The Hierarchy of Controls

1. Eliminate or substitute. Pick lower-fume consumables where the joint allows it. Lanthanated and other low-fume electrode families cut the fume burden compared with standard cellulosic rods, and a lower-fume process is the cheapest control of all. Our all-position welding rod guide covers the trade-offs between families.
2. Engineering controls. Local exhaust ventilation first. OSHA recommends keeping fume hoods, fume extractor guns, and vacuum nozzles close to the plume source, with respiratory protection added only where controls fall short. This is also where the enduring popularity of stick welding partly comes from: no shielding gas supply means fewer cylinders staged in a trench.
3. Administrative controls. Training, written procedures, and permit-required confined space entry under 29 CFR 1910.146. Know when to stop welding, who monitors the space, and what the rescue plan is.
4. Personal protective equipment. Last on the list, not first. Respiratory protection supplements engineering controls; it does not replace them.

Beyond the Fume

Fume is the hazard everyone discusses, and it is not the only one that puts a welder in an emergency room.

• Arc eye and UV. Looking at the arc brings photochemical conjunctivitis or corneal damage, which is why properly shaded glasses and a hot work permit that shields nearby workers are not optional.
• Noise. Arc noise above 85 dBA is routine; earplugs or earmuffs are required on most pipeline spreads.
• Heat stress. Radiant heat in a trench in July plus heavy PPE is a real thermal load, and dehydration drives mistakes that become weld defects.
• Height and excavation. Scaffolds, ladders, trench walls, spoil piles, and adjacent traffic. A trench collapse or struck-by event ends a career faster than any fume exposure.
• Flammable gas and hydrogen sulfide. Live gas lines, ruptured seals, and low gas concentrations in pits bring ignition risk and asphyxiation. Gas test before entry, every time.
• Zinc oxide fume. Welding on or near galvanized coating produces metal fume fever, often mistaken for flu hours after the exposure.
• Shielding gas cylinders. Storage, separation from oxidizers, secure caps, and upright handling are covered in our welding gas cylinder safety guide.

Respiratory Protection and Its Paperwork

When a respirator is required, the obligations are not optional. Under 29 CFR 1910.134, the employer must provide a medical evaluation before fit testing, conduct an annual fit test for tight-fitting facepiece respirators, and maintain a written respiratory protection program covering training, maintenance, and cartridge change schedules. OSHA’s welding-specific rule, 29 CFR 1910.1026, addresses respirator selection and requires fume controls to be used first. Pipeline work also sits under 1910 Subpart Q, 1926 Subpart J for construction, and 1915 Subpart D for shipyard and marine work.

The NIOSH guidance for welding fumes and manganese makes the practical point cleanly: the welder who controls the plume controls the exposure. Nothing on that list of contaminants is a mystery. They come from the arc, the coating, and the consumable, and every one of them responds to ventilation, distance, and material selection.

Common Pipeline Welding Mistakes and How to Troubleshoot Them

Most pipeline weld rejections come from a small list of preventable mistakes. The key is knowing what each defect looks like on film, what causes it, and how to correct it before it becomes a cutout.

Defect	What it looks like on a radiograph	Common causes	Corrective action
Porosity	Round or slightly elongated dark spots, often clustered near the root or scattered through the cap	Moisture, oil, rust, or mill scale; contaminated shielding gas; wind; too long of an arc length	Clean and dry the joint, verify shielding gas purity and flow, shield from wind, and shorten the arc
Slag inclusions	Irregular dark shapes with jagged edges, between passes or in the heat-affected zone	Poor interpass cleaning, wrong travel angle, excessive weave, too much amperage	Grind or chip each pass, hold a 10 to 15 degree drag angle, and stay within qualified amperage
Tungsten inclusions	Bright white spots, the opposite contrast of porosity and slag	GTAW tungsten dropped into or touchedthe molten pool	Reject and remove. Watch torch-to-work distance and tungsten extension on GTAW roots
Lack of fusion	A straight dark line parallel to the weld bevel	Cold lap, insufficient preheat, travel speed too fast, bevel angle too narrow, joint misalignment	Re-cut the joint, increase preheat, widen the bevel, correct alignment, slow down and hold the puddle against the sidewall
Incomplete penetration	A straight dark line at the root centerline with sharp edges and consistent width	Root gap too large, wrong root pass parameters, wrong process for the joint	Re-establish correct root gap and remake the root with qualified parameters, or requalify the process
Undercut	A dark groove at the cap toe in the parent metal, running parallel to the weld	Too much amperage, excessive travel speed, wrong work angle, excessive weave	Reduce amperage and travel speed, correct the work angle, use a stringer rather than a wide weave
Cracks	Sharp, linear indications, often at the toe or in the heat-affected zone	Hydrogen-assisted cold cracking (missing preheat, high diffusible hydrogen, high restraint, fast cooling) or stress corrosion cracking in sour service	Stop, do not repair blind. Re-examine preheat and electrode choice, then requalify the repair procedure
Excessive reinforcement / high cap	Thick cap profile visible in the image, plus a profile that fails visual or coating checks	Feeding too much metal, holding the puddle too long	Reduce electrode travel and bead size, trim the cap, watch cap height for coating bridging
Hard heat-affected zone	Usually not visible on film. Surfaced by hardness testing or later cracking	Excessive heat input, missing preheat, no control of interpass temperature	Lower heat input, raise preheat and interpass limits, move to a lower-hydrogen electrode

Undercut and internal concavity also have their own visual acceptance limits. API 1104 Section 9 sets visual acceptance standards for undercutting in Section 9.7 and internal concavity in Section 9.8, so a groove that looks cosmetically acceptable can still be a reject if it exceeds the depth limit. Check the standard rather than eyeballing it.

Cracks: the defect that stops the job

Everything else on that list is a repair. A crack in or near the heat-affected zone is a different conversation. When a crack is found, the API 1104 interpretations process confirms the treatment: heat-affected zone cracks found by NDE are not acceptable, referencing the AWS A3.0 definition of a weld crack. There is no marginal case and no interpretation that softens it.

The usual mechanism is hydrogen-assisted cold cracking. Steel picks up hydrogen from cellulosic electrode coatings, and if that hydrogen is still present when the joint cools through the martensite transformation range, it combines with high restraint to open a crack. The levers are preheat, interpass control, electrode selection (a guide to alloy steel welding electrodes and their diffusible hydrogen ratings helps), and holding a realistic inspection hold time before stressing the line. In sour service, add chloride stress corrosion cracking to the list, which is a materials-and-environment problem rather than a welder-parameter problem.

The paperwork failures that never show up on film

Some of the worst pipeline welding failures are invisible to NDE and show up only in an audit. The recurring ones:

• No current WPS at the site. The procedure exists in the office trailer, not on the line. The work stopped.
• Lapsed continuity. The welder has not welded the process for six calendar months or has not had a tested weld within 7.5 months. Federal welder qualification requirements are not a formality, and requalification is straightforward under 49 CFR 192.227.
• An unqualified repair. Repair welds have their own qualification requirements under API 1104 Section 10. They are not a bonus round for whoever was closest.
• Missing weld identification. No weld ID number means the weld cannot be traced, and an untraceable weld is an audit finding even if it is sound.
• A WPS with no supporting PQR. A written procedure that was never tested is a document, not a qualified procedure.

When the schedule becomes the defect

Most defects trace back to a decision made under time pressure. Field research on SMAW girth welding recorded typical qualified settings of roughly 75 to 85 amps and 24 to 28 volts at 3 to 5 inches per minute with a 1/8 inch electrode for the root pass, and about 130 to 140 amps and 25 to 29 volts at 5 to 7 inches per minute for the cap. When a crew is behind, the amperage drifts up and the travel speed drifts up with it, and that is exactly the combination that produces undercut and slag. Why SMAW still wins in the field is a productivity argument, but it only holds when the procedure is actually followed.

Night welding adds its own problems: fatigue, cold pipe that soaks heat out of the joint faster than expected, dew on a joint that has been sitting since dusk, and headlamp shadows that hide undercut until the film comes back. Fatigue is the quietest defect generator of all, and it is the hardest one to argue about with a customer.

Habits of crews with high first-pass rates

• Clean the bevel twice: once when the joint is fit, again immediately before the root pass.
• Keep the electrode oven in the weld truck and count hours, not guess them.
• Hold a short preheat soak and verify with a temperature stick, not a torch.
• Chisel or grind every pass, then inspect the interpass surface before laying the next bead.
• Mark every weld and keep the WPS, PQR and WPQR package physically at the weld site.
• Stop the job when a defect is found instead of welding over it. Air arc gouging is the standard way to remove defective weld metal, and a proper excavation is far cheaper than a cutout.

Conclusion: A Field-Ready Mental Model

If you carry one idea out of this guide, make it this: pipeline welding is a chain, and the finished weld is only as good as its weakest link. Preparation, parameters, position control, cooling rate, and documentation all have to hold at once. A textbook root pass under a rushed cap is not a good weld. Neither is a flawless weld with no qualification record behind it. Both are liabilities waiting on a rainy Tuesday downstream.

Field culture loses that instinct first under pressure. A hot day shortens the preheat interval. A crew changeover puts a new welder on a joint before anyone checks continuity. A bevel that needed a wire wheel pass gets wiped with a rag. Each feels trivial. Stacked together, they are why a girth weld gets cut out and trucked away.

The Pre-Weld Check

Run these before the arc strikes. Most take under a minute, and most rejections begin when one is skipped.

• Procedure on site: the current WPS at the joint with its supporting PQR, matching the process, position, material, wall thickness, and joint configuration in front of you.
• Welder qualified: for that position and process, with continuity current.
• Material verified: grade and wall thickness checked against the line list.
• Bevel clean: mill scale, rust, oil, grease, and moisture removed; root face and gap match the WPS.
• Alignment correct: hi-lo within tolerance, offset corrected, alignment held through the root pass.
• Consumables ready: electrodes dry and in date, rod oven temperature logged, wire and shielding gas in order, gas flow checked. The reasoning behind why SMAW still wins holds up in the field.
• Preheat measured: a pyrometer reading at the joint, logged, with the interpass interval respected.
• Environment controlled: wind shield up, moisture off the joint, cap height planned so the coating crew is not bridging a tented weld.

The Post-Weld Documentation Check

The weld is finished, but the job is not. These records prove the weld was made by a qualified person to a qualified procedure.

• Weld identification: the mark on the pipe matches the map, tally, and welder log.
• Welder qualification on file: 49 CFR 192.227 requires records demonstrating each individual welder qualification at the time of construction, retained a minimum of 5 years for steel transmission pipe installed after July 1, 2021.
• NDT records: welds made, welds tested, welds rejected, and the disposition of each reject, kept for the life of the pipeline, not the project folder.
• Inspection reports: radiographs or UT data with the acceptance criteria cited.
• Repair file: excavation dimensions, weld metal removed, repair WPS used, requalification where required, and the re-inspection result.
• Process data: preheat and interpass readings, consumable lot numbers, and any post-weld heat treatment.

Where the Rules Come From

Start with API 1104, the pipeline welding standard. The 22nd edition expands the scope to cover gas and arc welding used in the construction and in-service repair of pipes and components for the compression, pumping, and pipeline transmission of crude oil, petroleum products, fuel gases, carbon dioxide, and nitrogen. Pair it with 49 CFR 192 Subpart E, which incorporates it, ASME Section IX for welder and procedure qualification, and API RP 577 for in-service inspection and repair. On the health side, read OSHA’s welding safety factsheet, 1910.1026 on hexavalent chromium, and the NIOSH welding resources.

One useful detail when a technical point comes up with a client: API 1104 Annex C sets out the official process for requests for interpretation and revision, and the committee publishes its responses, turning a hallway opinion into a citable reference.

One Weld, One Checklist

Do not try to absorb all of this at once. Pick a single girth weld from the last job you worked or supervised, walk the pre-weld checklist against it, then walk the documentation checklist. You will find at least one gap in the first five minutes.

That is the whole craft: not a perfect bead every time, but a system where the mediocre day still passes inspection and the defects that slip through are caught by records, not by an excavation crew three years later.

Frequently Asked Questions

What makes pipeline welding different from structural or shop welding?

Pipeline welds are permanent pressure-boundary elements that must survive decades of cyclic pressure in buried, underwater, or chemically aggressive service, and they are governed by a specific code — API 1104 — referenced directly into U.S. federal pipeline regulations. The qualification scheme, the nondestructive testing percentages, and the record-keeping requirements all differ from structural work, so a welder qualified in one domain is not automatically qualified in the other.

What is the difference between API 1104 and ASME Section IX?

API 1104 is a complete pipeline welding standard that carries its own procedure and welder qualification rules, while ASME Section IX is a qualification code that other construction codes (such as B31.3) reference but do not apply directly. Procedures and welders qualified under one are not automatically acceptable under the other, which is why a contractor moving between plant and pipeline work must check which code owns each document in the package.

Do pipeline welders have to be qualified in the 6G position?

Qualification position depends on what the welder will actually weld and on the governing code. Under 49 CFR 192.227, welders qualify in accordance with API 1104 Section 6 or ASME Section IX, and API 1104’s 21st edition provides that a welder who qualifies in the fixed position also qualifies in the rolled position, which simplifies many crew assignments. For low-stress service line work, 49 CFR 192.227(b) offers a separate simplified qualification route.

What is the most common cause of a rejected pipeline girth weld?

In practice, incomplete penetration and lack of fusion are among the most frequently rejected indications, and they usually trace back to preparation rather than to the arc — a root gap that is too large, a bevel that is too narrow, misalignment forcing a weave, or a dirty or contaminated bevel. Porosity is also common and is usually a moisture, contamination, or shielding problem. API 1104 acceptance criteria treat planar defects such as lack of fusion and incomplete penetration far more strictly than scattered porosity in normal service.

How much of a new pipeline gets radiographed or ultrasonically tested?

It depends on the location class. Under 49 CFR 192.243, at least 10 percent of each day’s field butt welds must be tested in Class 1 locations except offshore, at least 15 percent in Class 2, and 100 percent in Class 3 and Class 4 at river crossings, offshore, and within railroad or public highway rights-of-way, with a 90 percent floor where 100 percent is impracticable. Pipeline tie-ins require 100 percent testing, and a sample of each welder’s work each day must also be tested.

Can you weld on a pipeline while it is still in service?

Yes, but it is heavily regulated and operationally complex. API 1104 Annex B covers in-service welding, and API RP 577 gives additional guidance on repair and alteration. The work requires a qualified repair procedure, welders qualified for repair, careful control of diffusible hydrogen and preheat, and an inspection method capable of detecting underbead and delayed hydrogen cracking at the carrier pipe weld toe — magnetic particle, ultrasonic, or both.

Which welding process is best for field pipeline girth welds?

There is no single answer, but cellulosic SMAW with E6010 and E7010 electrodes remains the backbone of manual field welding because of its deep penetration, fast freeze, and vertical-down capability in all positions. Mechanized and hybrid approaches — an internal GTAW or GMAW root followed by FCAW-G fill and cap — deliver better consistency and higher deposition rates, but carry significant qualification and engineering cost and are less forgiving of wind.

Why is preheat so important on high-strength pipeline steel?

Preheat slows the cooling rate through the transformation range, which keeps the heat-affected zone softer and less crack-susceptible while hydrogen is still present. Combined with low diffusible hydrogen consumables and controlled interpass temperature, it is the primary defense against hydrogen-assisted cold cracking, the failure mode that has stopped more high-strength line pipe projects than any other weld defect.

How often do pipeline welders have to requalify?

Under 49 CFR 192.229, a welder may not weld with a given process unless within the preceding 6 calendar months they were engaged in welding with that process, or they performed a weld with that process that was tested and found acceptable within the preceding 7.5 months. For pipe at 20 percent or more of SMYS hoop stress, a tested and accepted weld is required every 6 months, or qualification can be maintained by having welds tested and accepted at least twice per calendar year at intervals not exceeding 7.5 months.

What health risks do pipeline welders face from fume?

The main concerns are metal fume constituents such as manganese and hexavalent chromium, which is generated when chromium in stainless base metal, chromate coatings, or some consumables oxidizes during welding. OSHA’s permissible exposure limit for Cr(VI) is 5 micrograms per cubic meter as an 8-hour TWA, while NIOSH recommends a far more protective 0.2 micrograms per cubic meter. Manganese exposure is a particular problem in confined spaces, which is why local exhaust ventilation, confined-space permitting, and respiratory protection matter so much.