The job title suggests someone who looks at finished welds and decides whether they are good. That is a real part of the work and it is a surprisingly small part. Most of an inspector's day is spent on the conditions that produce welds - documents, materials, fit-up, qualifications, equipment, sequence - and on recording what was found in a form that will still make sense to somebody else in two years.
This guide describes the work as it is actually structured rather than as a job advert describes it. It covers what happens before, during and after welding, the paperwork that carries the whole thing, the parts of the role that are about people rather than metal, and the ways shop and field work differ. It reflects general industry practice; specific duties on any job are set by the contract, the governing code, and the employer.
The job in one sentence
An inspector verifies that welding work conforms to a specified set of requirements, and creates the record that demonstrates it. Both halves matter equally. Verification without a record is an opinion nobody can rely on, and a record without honest verification is worse than nothing because it manufactures false confidence.
Notice what is not in that sentence. An inspector does not decide what is acceptable - that was decided when the code and the contract were written. An inspector does not design the joint, choose the process, or tell the welder how to hold the electrode. Staying inside the role is not timidity; it is what makes the verification worth anything. The work is therefore comparative all the way down: every task reduces to putting an observation next to a requirement, and the skill lies in knowing which observations to make and how to make them reliably.
Before the arc: where the value is
The pre-weld phase is the highest-leverage part of the day and the part most often compressed by schedule pressure. It covers whether the drawings and the procedures are the current revision, whether the material is what the drawing calls for and is traceable, whether the consumables match the procedure and have been stored as required, and whether the welders are qualified for the work in front of them.
It also covers the physical joint. Preparation angle, root opening, root face, cleanliness, alignment, tack quality, backing where required, and the condition of the surfaces that will be fused. A joint that is wrong at this stage produces a rejectable weld no matter who welds it, and the correction costs minutes now against a repair later. This is the phase where an inspector prevents work rather than judging it, and turning up at fit-up to ask four questions is worth more than a very careful examination of the finished joint.
- Current revision of drawings, procedures and any project specification
- Material identification and traceability to the documentation supplied with it
- Consumable classification, condition, storage and issue control
- Welder qualification status and its continuity
- Joint preparation, cleanliness, fit-up, alignment and tacking
- Equipment condition and calibration where the procedure depends on a reading
During welding: surveillance rather than supervision
Once welding starts the inspector's job is to observe whether the conditions the procedure requires are actually holding. Are the parameters within the qualified ranges? Is preheat being maintained and is interpass temperature staying inside its band? Is each pass being cleaned before the next? Is the sequence being followed? Is the welding position what the procedure and the qualification permit?
The useful skill here is sampling. Nobody watches every pass on every joint, so the value comes from choosing when to look. After a break, on a second shift, at the far end of a long joint, when the weather changes, when a new welder starts on the work: these are the moments when a requirement quietly stops being met, and they are far more informative than the first pass of the day.
The line between surveillance and supervision is one to keep clear. Pointing out that a parameter is outside the qualified range is inspection. Telling a welder to increase their travel speed is direction, and an inspector who directs the work has compromised their independence in verifying it. The distinction feels pedantic right up until the weld is disputed.
After welding: examination and what follows
The post-weld phase starts with visual examination, which finds more rejectable conditions than every other method combined. Surface condition, profile, weld size and length against the call-out, and the base metal either side of the joint. It then extends into whichever additional methods the contract requires, which the inspector may perform, may witness, or may simply coordinate and review the results of.
Where a condition is found, the work does not end at the finding. Somebody has to define the repair, the repair has to be carried out to a procedure, and the repaired area has to be re-examined. Tracking that loop to closure is a real part of the job and one of the more common places for paperwork to go missing, because attention naturally moves on once the finding has been raised. Post-weld heat treatment, where it applies, adds its own verification of charts, thermocouple placement, heating and cooling rates and hold times, usually by reviewing a record somebody else produced.
The paperwork is the product
It is easy to treat records as the administrative tax on the real work. On most projects the records are the deliverable. What survives handover is the documentation package, and it is what an owner, an auditor or an investigator will actually use if something goes wrong years later. Nobody will remember the walk-around; everybody will read the file.
A usable record names the condition, fixes its location against something the next person can find, gives the governing measurement, identifies the provision it was compared against, and states the disposition separately from the observation. It also records what could not be examined. Access limitations, coated surfaces and areas that could not be lit belong in the record exactly as findings do. The discipline that makes this sustainable is writing at the joint rather than at the end of the day, because reconstructed reports drift toward what the inspector expected to see.
- Record the observation while you are still standing at the joint
- Fix locations against a datum somebody else can find without you
- Record the governing measurement, meaning the worst one present
- Name the provision you compared against
- State the disposition as a separate sentence from the observation
- Record limitations on access, lighting and surface condition explicitly
- Track every finding through repair and re-examination to closure
The part of the job that is about people
An inspector spends a great deal of the day talking to people who would prefer a different answer. Welders, supervisors, project managers and clients all have schedule and cost pressure, and an inspector is frequently the person standing between that pressure and a piece of work. How that is handled determines whether the role functions at all.
The technique that works is separating the finding from the person. A condition measured, located and compared against a stated provision is a fact about steel. A conversation about whether somebody is doing a good job is a different conversation, and mixing the two turns a routine finding into a dispute. Cite the requirement, not your authority.
The harder situations are the ones with a commercial edge: pressure to pass something marginal, to sign for work not witnessed, to date a record conveniently. These are ethics questions rather than technical ones, and they are the reason certification schemes carry conduct requirements at all. Every experienced inspector has faced a version of this, and the only durable answer is a record that reflects what was actually observed.
Shop and field are different jobs
Shop work is repetitive, well lit, weather-proof and close to the paperwork. The same joint types recur, the same welders do them, and the inspector's attention drifts toward efficiency and toward catching the slow drift of a process that is broadly under control. Days are predictable and the commute is the same every morning.
Field work is the opposite. Access is difficult, lighting is whatever you brought, weather affects both the welding and the inspection, and the joints are frequently one-off. There is more travel, more time in a truck, more coordination with other trades, and more situations where the honest answer is that a joint could not be examined properly and that has to be written down.
Neither is better, but they suit different people, and it is worth knowing which one a role actually is before accepting it. The technical knowledge transfers completely; the working life does not. Ask about travel, about shift patterns, and about how much of the week is spent outdoors, because those are the details that decide whether a role fits.
- Shop: repeatable joints, controlled conditions, steady hours, drift-detection focus
- Field: variable access, weather exposure, travel, one-off joints, more coordination
- Both: the same requirement to observe, measure, compare and record
- Field work more often produces legitimate limitations that must be recorded as such