Welder qualification to EN ISO 9606-1

No matter how well a welding procedure specification (WPS) is written, the assurance stays on paper if the person laying the bead is not qualified. Welder qualification closes exactly that gap: the WPQR (EN ISO 15614-1) — the ISO counterpart of the ASME PQR — proves the procedure can be done, while the qualification test proves the welder can do it. On most European and international projects, that proof is called EN ISO 9606.

This article is based on EN ISO 9606-1:2017 (incorporating ISO 9606-1:2012, Cor 1:2012 and Cor 2:2013). The 2017 EN edition superseded EN ISO 9606-1:2013. It does not change the ISO text technically, and the clause and table numbers are identical. The only difference lies in the European annexes: Annex ZA referred to Directive 97/23/EC in the 2013 edition, whereas the 2017 edition links the standard to the Pressure Equipment Directive (2014/68/EU), under which it has been cited in the Official Journal of the European Union (OJEU) since 1 March 2018; Annex ZB was added for 2014/29/EU (simple pressure vessels). ISO is running a revision that merges Parts 1 to 5 into a single standard: the document sits at the ISO/DIS 9606.2 stage (the DIS ballot opened in mid-2026) and, once published, will supersede ISO 9606-1:2012, -2:2004, -3:1999, -4:1999 and -5:2000. At the time of writing, the merged text is not yet published — confirm the current status with your national standards body.

1. Where EN ISO 9606-1 sits and what it is for

EN ISO 9606 is the standard family that qualifies the skill of welders performing manual or partly mechanized fusion welding. Part 1 covers steels, Part 2 aluminium, Part 3 copper, Part 4 nickel, and Part 5 titanium and zirconium. The basic distinction is this: EN ISO 9606 qualifies the welder, so it covers manual and partly mechanized processes that depend on hand skill. For fully mechanized and automatic welding, operator and setter qualification falls under EN ISO 14732.

One boundary deserves a heavy underline: a welder certificate guarantees nothing metallurgical. Preheat, interpass temperature, heat input, post-weld heat treatment (PWHT) and hydrogen-controlled consumable management all sit outside the scope of 9606-1. Clause 6.3 states plainly that PWHT specified in the pWPS need not be applied to the test piece, at the manufacturer's discretion. Cold cracking, hardness, toughness and HAZ control are the job of the WPS/WPQR. When "the welder has a certificate" gets confused with "the joint is metallurgically sound", the bill arrives at hydrotest.

2. The test piece: plate or pipe?

The qualification test is welded on a test piece defined by the standard. On plate, the test piece length shall be at least 200 mm; the test length is 150 mm — a fixed value, not a minimum. If the pipe circumference is less than 150 mm, additional test pieces are required to make up the test length, but the total number of test pieces is three at most — three in total, not three extra.

Under Clause 6.3, the test piece shall be welded in accordance with a pWPS or WPS prepared to EN ISO 15609-1 or EN ISO 15609-2 — this is a requirement, not a recommendation. The same clause also requires that, for a fillet weld test piece, the throat thickness be defined in the pWPS/WPS.

Clause 5.3 defines the coverage relationship between plate and pipe numerically. Read it from the numbers, not from memory:

A pipe test piece with D > 25 mm covers plate.
A plate test piece covers fixed pipe only for D ≥ 500 mm.
A plate test piece covers rotated pipe of D ≥ 75 mm, and only in positions PA, PB, PC and PD.

Branch connections (Clause 5.4 c–d). A butt weld test piece on pipe also covers branch connections with an angle of 60° and above. For a branch, the diameter range is calculated from the outside diameter of the branch. The deposited thickness range, however, depends on the joint type (Table 6, footnote b, and Figure 1): for set-on connections the deposited thickness of the branch governs, while for set-in / set-through the deposited thickness of the main pipe or shell governs. Branches at angles below 60° require a separate, dedicated test piece.

The butt-to-fillet relationship hides the mistake made most often in the field. Clause 5.4 b) is unambiguous: a butt weld test does not qualify the welder for fillet welds, and the reverse is equally false. The old EN 287-1 habit of "if he passed the butt, the fillet is covered anyway" is void under 9606-1. There are three routes to fillet weld coverage:

A separate FW test piece.
The combined BW+FW test piece of Clause 5.4 b): a single-bevel (K) preparation with permanent backing, at least 10 mm thick.
The additional fillet weld test piece of Clause 5.4 e) — separate for each process, each filler material group and each covering/core type; at least 10 mm thick (or the thickness of the butt test piece if that is thinner), completed in a single layer in position PB. A layer is not a pass: a single layer may contain more than one pass.

That additional test gives only positions PA and PB for fillet welding, and it is recorded separately on the certificate. One important exception sits in Table 12, footnote b): a welder qualified with a multi-layer (ml) butt weld who also performs the 5.4 e) additional fillet test is qualified for both single-layer and multi-layer fillet welding.

Another frequent cause of rejection also sits in Clause 6.3: the test piece shall contain at least one stop/restart area in both the root run and the capping run, and these shall be marked. On a multi-process test, that applies to each process separately. Grinding of the capping run is permitted only in that area, and only with the inspector's approval.

3. Essential variables and the range of qualification

Clause 5.1 defines eight essential variables: welding process; product type (plate/pipe); weld type (BW/FW); filler material group; filler material type; dimensions (material thickness / deposited thickness s and pipe outside diameter D); welding position; and weld details (mb/nb, gas backing, flux backing, consumable insert, ss/bs, sl/ml, leftward/rightward welding).

Welding process — identified by EN ISO 4063 numbers: 111 manual metal arc welding (MMA), 131 solid wire with inert gas, 135 solid wire with active gas (MAG), 136 flux-cored wire with active gas, 138 metal-cored wire, 141 TIG. As a rule, each process requires its own qualification, but the exceptions in Clause 5.2 save real money and time on site: 135 ↔ 138 are mutually valid; 121 ↔ 125 are mutually valid; a test made with 141, 143 or 145 covers 141/142/143/145, but 142 covers only 142; and a qualification taken in dip (short-circuiting) transfer with 131, 135 and 138 covers the other transfer modes, though not the reverse — this exception is limited to those three processes. Clause 5.9 adds a special rule for 311 (oxy-acetylene): changing from leftward to rightward welding, or vice versa, requires a new test.
Filler material group and type — this is what really drives coverage. Table 2 gives the groups: FM1 non-alloy and fine-grain steels; FM2 high-strength steels; FM3 creep-resisting steels, Cr < 3.75%; FM4 creep-resisting steels, 3.75% ≤ Cr ≤ 12%; FM5 stainless and heat-resisting steels; FM6 nickel and nickel alloys. Which side of the limit a value falls on changes the range directly: a consumable with exactly 3.75% Cr belongs to FM4.

The direction of coverage in Table 3 runs one way only, and it is worth memorizing:
- FM1 → FM1, FM2
- FM2 → FM1, FM2
- FM3 → FM1, FM2, FM3
- FM4 → FM1, FM2, FM3, FM4
- FM5 → FM5 only
- FM6 → FM5, FM6
In metallurgical terms: on the ferritic side, coverage runs downward (an FM4 test opens FM1, never the reverse); the stainless/nickel side does not intersect the ferritic side in any direction — a stainless test does not cover carbon steel; an FM6 test opens FM5, but an FM5 test does not open nickel alloys.

For filler material type, solid wire/rod (S) and metal-cored (M) cover each other. Basic flux-cored wire (B) covers itself plus the R, P, V, W, Y and Z cores; R/P/V/W/Y/Z cover only themselves; B and S/M do not cover each other in either direction. For process 111, Table 4 sets a separate coverage rule for the electrode covering: a test taken with a basic covering (B: 15, 16, 18, 28, 45, 48) covers both B and the A/RA/RB/RC/RR/R group; the rutile/acid group covers only itself; cellulosic (C: 10, 11) covers only C, and a test taken with C opens nothing else. Hydrogen control is not the subject of 9606-1, but in practice this coverage rule stops a welder qualified on low-hydrogen consumables from moving straight to a cellulosic root. For an unbacked root (ss, nb), the covering or core type used in the test shall be the same in production — and that requirement applies to covered electrodes as well as to flux-cored wire.

Readers often skip the first sentence of Clause 5.6: welding with filler material covers welding without filler material (nm); the reverse does not apply. The note to the same clause sets up an important exception: for processes without filler material such as 142 and 311, the range is governed by the ISO/TR 15608 parent material group used in the test. In autogenous welding the parent material group behaves like an essential variable and takes the place of the filler material group in the designation string (Clause 11 b) 4).
Dimensions — for butt welds, the range is governed not by parent material thickness t but by the deposited weld metal thickness s (Clause 5.7, Table 6): s < 3 → from s to 3, or from s to 2s, whichever is greater; 3 ≤ s < 12 → from 3 to 2s; s ≥ 12 → 3 mm and above, with no upper limit (in that case the test piece shall be welded in at least three layers; the number of layers is a requirement distinct from the number of passes, because of the reheating/tempering effect). The footnotes to Table 6 give a separate range for 311: s < 3 → from s to 1.5s; 3 ≤ s < 12 → from 3 to 1.5s.

On a multi-process joint, s is calculated separately for each process; in addition, Table 1 creates a range for the total s = s₁ + s₂. In a backed/unbacked combination, the root area falls only within the range of the process that welded the root.

For fillet welds, Table 8 works from parent material thickness: t < 3 → from t to 2t or to 3, whichever is greater; t ≥ 3 → 3 mm and above. For pipe diameter, Table 7 applies: D ≤ 25 → from D to 2D; D > 25 → 0.5D (minimum 25 mm) and above, with no upper limit. So a test welded on Ø 60 mm opens Ø 30 mm and above; the diameter range widens downward as well as upward.
Weld type and weld details — butt (BW) / fillet (FW); backed (mb) or unbacked (nb) root; single-sided / double-sided welding. Gas backing, flux backing and consumable inserts are essential variables as weld details; do not confuse them with the shielding gas. The direction of Table 11 is also clear: a test taken unbacked (ss nb) covers mb, bs, gb and fb, but does not cover the consumable insert (ci); a test taken with mb does not cover nb. In terms of root protection and root oxidation, this is the single most useful rule on site.
Welding position — to EN ISO 6947, the main positions are PA, PB, PC, PD, PE, PF, PG on plate, and PA, PB, PC, PD, PH, PJ plus H-L045 / J-L045 on pipe. In Tables 9 and 10, PF and PG belong to the plate rows and PH and PJ to the pipe rows; mixing that distinction up leads straight to a misread range.

Non-essential variables (Clause 10): type of current and polarity, parent material group/sub-group, and shielding gas. All three are recorded on the certificate but do not narrow the range — with the single exception of 142 and 311 noted above, where the absence of filler material means the ISO/TR 15608 parent material group used in the test governs the range. Clause 5.5.1 recommends that the parent material be chosen from groups 1 to 11 of CEN ISO/TR 15608 (the standard says "should", so it is not an absolute prohibition); the group and sub-group used are recorded on the certificate in every case (Clause 5.1).

As a short reminder, the ISO/TR 15608 groups are: 1 non-alloy C-Mn steels; 2 fine-grain normalized/TMCP steels; 3 quenched and tempered high-strength steels; 4–5 low Cr-Mo creep-resisting steels; 6 high Cr-Mo steels; 7 ferritic/martensitic stainless steels; 8 austenitic stainless steels; 9 nickel-alloyed low-temperature steels; 10 duplex stainless steels; 11 high-carbon steels. Cast irons and non-ferrous groups fall outside 9606-1; for those, go to Parts 2, 3, 4 and 5.

This is why the welder test is less sensitive to material than its counterpart in ASME BPVC Section IX: it measures skill, not metallurgy. A change of shielding gas mixture does not invalidate a welder certificate, and neither the shielding gas nor the backing gas enters the designation string.

4. How far a position reaches

Position offers the most productive optimization on site. Per Table 9, a butt weld test welded on pipe in H-L045 covers PA, PC, PE and PF; it does not cover PG. J-L045 covers PA, PC, PE and PG. Clause 5.8 adds two more practical rules: two pipes of the same diameter, one welded in PH and one in PC, are equivalent to the H-L045 range in the upward direction; two pipes of the same diameter, one in PJ and one in PC, are equivalent to the J-L045 range in the downward direction. Furthermore, on pipes of D ≥ 150 mm you may weld two positions on a single test piece (two-thirds of the circumference in PH or PJ, one-third in PC) and cover all positions for that welding direction. That is what the advice "one covering test in the hardest position" actually means in numbers. Even so, read the range tables line by line; not every difficult position automatically covers every easy one.

5. Testing: visual always, the rest depends on the table

Visual testing (EN ISO 17637) is mandatory in every case and on every test piece. Per Clause 7, the acceptance criterion is EN ISO 5817 level B, except for five imperfection types where level C applies: undercut (501), excess weld metal (502), excessive convexity (503), excessive penetration (504) and excessive throat thickness (5214). The level C allowance for undercut starts more arguments in the field than any other detail in the clause.

For butt welds, volumetric and mechanical testing follows Table 13 and its footnotes:

Apply either radiographic testing (EN ISO 17636) or a bend test (EN ISO 5173) / fracture test (EN ISO 9017).
For processes 131, 135, 138 and 311, an additional bend or fracture test is mandatory even when RT is chosen. The reason differs by process: with 131/135/138 it is the characteristic lack of side-wall fusion, which radiography does not reliably catch; with 311 it is the imperfection behavior peculiar to the process.
UT (EN ISO 17640) may replace RT only on ferritic steels and for t ≥ 8 mm; in that case the additional bend/fracture test above is not required.
On pipes of D ≤ 25 mm, a notched tensile test on the whole piece may be used instead of a bend or fracture test.

For fillet welds and branch connections, visual testing is mandatory and RT is not; the bend test is not applicable and the fracture test is mandatory. The fracture test may be replaced by macroscopic examination (EN ISO 17639) provided that at least two macro sections are taken, at least one of them from a stop/restart area (Table 13 footnote e, Clause 6.5.3). On pipes, the fillet weld fracture test may be replaced by RT.

The acceptance criteria for bend test specimens are numerical: no single discontinuity of 3 mm or more (≥ 3 mm) in any direction, and the sum of discontinuities larger than 1 mm but smaller than 3 mm shall not exceed 10 mm on any one test specimen. The standard allows one exception: disregard cracks that open at the edges of the specimen during testing, unless lack of penetration, slag or another discontinuity caused them.

All of this drives the logic of method selection: RT picks up porosity, slag and root penetration imperfections well, but is weak on lack of side-wall fusion (401) and on cracks perpendicular to the beam. Bend and fracture tests expose those imperfections mechanically. Finally, Clause 8: a welder who fails may be given one opportunity to retest without further training — this is a possibility that can be granted, not an automatic right.

6. Validity: six-month confirmation and three revalidation routes

This is the most characteristic — and the most frequently misapplied — part of EN ISO 9606-1.

Clause 9.1 — start and duration. Qualification starts on the date the test piece is welded. The certificate becomes invalid unless it is confirmed every six months. Which of the options a), b) or c) in Clause 9.3 will apply shall be written on the certificate when it is issued. This is the trap auditors find most often.

Clause 9.2 — six-month confirmation. The confirmation is made by the person responsible for welding activities (the welding coordinator under EN ISO 14731) or by the examiner/examining body. It confirms that the welder has been working within the range of qualification and extends validity by another six months.

Clause 9.3 — revalidation. Only the examiner or examining body may perform it; the welding coordinator's signature alone is not enough. There are three routes:

a) Retest every three years.
b) Revalidation every two years from production weld records: two production welds made within the last six months of the validity period are tested by RT, UT or destructive testing and recorded; acceptance criteria are per Clause 7, and the test shall repeat the original test conditions apart from thickness and outside diameter. This extends the qualification by two years.
c) Unlimited validity: if the welder works for the same manufacturer that issued the qualification, if the manufacturer's quality programme has been verified to EN ISO 3834-2 or -3, and if production welds of acceptable quality are documented (confirming position, BW/FW and mb/nb conditions), then the certificate remains valid indefinitely as long as the 9.2 confirmations continue.

A critical constraint on pressure equipment: Annex ZA of EN ISO 9606-1:2017 does not permit option c) for Category II, III and IV pressure equipment; in those categories the examiner/examining body shall be a competent third party (a notified body or a recognized third-party organization). National forewords give the same practical advice: on PED work, choose option a) or b). Applying unlimited validity in these categories produces a direct nonconformity on the project.

One more distinction matters: if the six-month confirmation signature is missing, the certificate is automatically invalid under Clause 9.1 — this is not a suspension. Cancellation (Clause 9.4), by contrast, applies when there is reason to doubt the welder's ability, and it covers only the qualifications in doubt; the rest stay valid.

7. How the philosophy differs from ASME BPVC Section IX

Coverage logic: on the welder side, ASME Section IX handles parent material through the QW-423 substitution table; the real essential variables are F-No. (QW-404.15; grouping table QW-432, alternative/substitute F-No. table QW-433), position (QW-405), thickness and diameter (the QW-452 tables, QW-403), backing (QW-402.4), root protection and transfer mode. The P-No. group belongs to the procedure side. EN ISO 9606-1, by contrast, is more finely tuned on filler material group and type, weld detail and position; the ISO range calculation works finer-grained, while the ASME one aggregates.
Continuity: in ASME Section IX, qualification is not tied to the calendar. When a welder has not used a process for more than six months, the qualification expires (QW-322.1) — it is not suspended — and under QW-322.2 a single test piece (free choice of material, thickness, diameter and position) renews all previous qualifications for that process. In ISO, validity depends on an active regime of records and confirmations — although 9.3 c) converges noticeably on the ASME logic. The generalization that "in ISO the certificate dies when the clock runs out" is therefore incomplete.
Acceptance criteria: ASME contains its own acceptance criteria; EN ISO 9606-1 relies on an external quality level standard, EN ISO 5817.

If the same welder needs both certificates, design the test piece to satisfy the range and testing requirements of both standards at once, and track validity under two separate rule sets.

8. Certification and supervision

Under Clause 6.1, the test piece shall be welded under the supervision of the examiner or examining body, and the certificate is issued solely under the responsibility of that examiner or examining body (Clause 10). This is a requirement of the standard, not a contractual extra.

What belongs on the certificate: the welder's identity, the test standard and date, the process, the essential variables and their corresponding ranges of qualification, the test results, the non-essential variables — type of current/polarity, parent material group/sub-group and shielding gas, the chosen revalidation method, and the confirmation fields. The job knowledge test in Annex B is informative; it is recommended, not mandatory — although some countries make it mandatory nationally. If it is carried out, record the result on the certificate as "Acceptable / Not tested".

The order of the designation is fixed in Clause 11; neither the shielding gas nor the backing gas enters that string, and in autogenous welding the parent material group replaces the filler material group. If several test pieces are combined on one certificate, then — except for the dimension (thickness/diameter) ranges given in Clause 5.7 — only one of weld type, position or deposited thickness may differ between the pieces. Showing weld symbols correctly on the drawing to EN ISO 2553 (or AWS A2.4) makes matching qualification to production far easier.

From the field

On a pipeline project, every welder had qualified on plate in position PF, and at first glance the certificates looked fine. Production, though, was fixed pipe in H-L045. Clause 5.3, read line by line, settled it: a plate test piece covers fixed pipe only for D ≥ 500 mm, and the line diameter was nowhere near that. PF on plate never covers H-L045 on pipe either. Several welders had to retest on pipe. Three days went out of the schedule, and the lesson was clear: read a qualification certificate variable by variable, not as "pass/fail".

The second lesson concerned confirmation. One welder's certificate looked valid, but the six-month confirmation signature was missing. The inspector did not "suspend" it. Under Clause 9.1 the certificate was already invalid; it had lapsed on its own the day the confirmation fell due. The third lesson cost the most. None of the certificates stated which of options a), b) or c) in Clause 9.3 would apply, so none of them qualified for revalidation. Worse, the job was Category III pressure equipment, where option c) was off the table to begin with. Qualification is not earned on the day someone signs it; the welder re-earns it every day, in the production record.

Related standards

EN ISO 9606-1:2017 (ISO 9606-1:2012 + Cor 1:2012 + Cor 2:2013) — Qualification testing of welders, Part 1: Steels (supersedes EN ISO 9606-1:2013; technical content unchanged, Annexes ZA/ZB updated)
EN ISO 9606-2 / -3 / -4 / -5 — Aluminium, copper, nickel, titanium/zirconium parts (merger revision under way as ISO/DIS 9606.2)
EN ISO 14732 — Qualification of welding operators and weld setters for mechanized and automatic welding
EN ISO 14731 — Welding coordination: tasks and responsibilities
CEN ISO/TR 15608 — Grouping system for metallic materials
EN ISO 6947 — Definition of welding positions
EN ISO 4063 — Numbering of welding processes
EN ISO 5817 — Quality levels for imperfections in fusion-welded joints in steel, nickel, titanium and their alloys (beam welding excluded)
EN ISO 6520-1 — Classification of welding imperfections
EN ISO 857-1 — Welding terms and definitions
EN ISO 3834 — Quality requirements for fusion welding of metallic materials
EN ISO 15609-1 / -2 and EN ISO 15614 — Welding procedure specification and procedure qualification
EN ISO 17637 / 17636 / 17640 / 5173 / 9017 / 17639 — Visual, radiographic and ultrasonic testing; bend, fracture and macroscopic examination
EN ISO 2553 / AWS A2.4 — Welding symbols
ASME BPVC Section IX — Welding, brazing and fusing procedure qualification / personnel qualification (comparison reference)
Pressure Equipment Directive (2014/68/EU) — the legislation linked through Annex ZA of EN ISO 9606-1:2017

Take it to the field: To keep the welding standards, procedure logic and field control steps in this article in your pocket — completely offline and free — take a look at the Doawise Welding Eng Guide app.

At DoaWise we carry out welding engineering and inspection services to international standards such as EN ISO 9606, EN ISO 15614 and ASME BPVC Section IX. From welder qualification through procedure qualification to field inspection, we document every step and deliver results that stay traceable and auditable long afterwards. Our aim is to close the gap between compliance on paper and real quality in the field, and to leave a traceable chain of assurance behind every weld.