What Are WPS and PQR? The Logic of Procedure Qualification

In welded fabrication, "the weld looks good" carries no technical weight. Weld quality lives largely in a metallurgical region you cannot see: the heat-affected zone (HAZ) and the internal structure of the weld metal. Two documents make that invisible quality repeatable, auditable, and evidence-based — the WPS and the PQR. Together they form the backbone of modern welding engineering. Get them wrong and neither a pressure vessel certification nor a steel structure approval will hold. This article covers what a WPS and a PQR are, the logical chain that links them, the concept of the essential variable, the qualification routes that need no test weld, and how the ASME and EN ISO worlds approach the same problem differently — along with the field realities that come with each.

1. WPS and PQR: two documents, one logic

The Welding Procedure Specification (WPS) is the production instruction that answers the welder's question: exactly how do I make this joint? It carries every production parameter — welding process, filler metal classification, shielding gas, current and voltage ranges, preheat and interpass temperature, welding position, joint geometry, and bead sequence. The WPS is a recipe.

The Procedure Qualification Record (PQR) is the laboratory record that proves the recipe actually works. It contains the values actually used while welding a test coupon, plus the results of the mechanical tests run on specimens cut from that weld. In short, the WPS points forward as an instruction; the PQR points backward to a proven fact. The WPS states what is to be done; the PQR states what was done and what came of it.

One terminology point up front: in the ASME world this record is called a PQR, while in the EN ISO world it is a WPQR (Welding Procedure Qualification Record). On the ISO side, the definitions of pWPS, WPS, and WPQR — along with the general qualification rules — come from EN ISO 15607; EN ISO 15609 covers content, and EN ISO 15614 covers qualification by testing. The rest of this article uses "PQR" for readability, but on an ISO job the document is a WPQR.

One very common generalization also needs correcting up front. The statement "a WPS is not valid without at least one PQR behind it" is true only for the qualification-by-testing route. If you follow ASME BPVC Section IX or EN ISO 15614, then yes, an approved PQR must sit behind the WPS. But the standards allow other routes. On the EN ISO side, EN ISO 15610 (qualification based on tested welding consumables), EN ISO 15611 (based on previous welding experience), and EN ISO 15612 (adoption of a standard welding procedure) require no test weld at all. EN ISO 15613 belongs in a separate category: it is a pre-production welding test standard. It does involve a test weld and mechanical testing; the difference is that the test is made on a mock-up in a geometry that represents actual production, rather than on a standard test piece. It exists for complex geometries and for cases where a standard specimen would not represent production.

On the ASME side, AWS B2.1 Standard WPS (SWPS) documents can be used without the manufacturer producing its own PQR. That route is governed by ASME BPVC Section IX, Article V (the QW-500 series) and carries four restrictions: (1) only the SWPSs listed in Section IX Mandatory Appendix E may be selected, (2) the SWPS shall be used as written — it cannot be modified or combined with another SWPS, (3) before first use the manufacturer shall weld and test a demonstration coupon per QW-510 and document it on Form QW-485, and (4) SWPSs are not permitted where notch-toughness (impact) testing is required. The construction code must also permit SWPS use. So the PQR is the most common and the strongest form of evidence — not the only one. Which route is acceptable is always decided by the contract code and by the customer or the third-party inspection (TPI) organization.

2. The qualification chain: pWPS → test weld → PQR → approved WPS

On the qualification-by-testing route, think of procedure approval as a chain — and the order of the links is not arbitrary.

The process starts with a preliminary Welding Procedure Specification (pWPS). The pWPS is an unproven draft recipe, built from the welding engineer's experience, the filler metal manufacturer's data, and the limits set by the standard. A test coupon is then welded in strict accordance with that pWPS. During that weld, every value is measured and recorded: current, voltage, travel speed, preheat, interpass temperature, and heat input.

The two systems treat witnessing differently, and site teams confuse the two constantly. ASME BPVC Section IX does not by itself require procedure qualification to be witnessed by an independent party. Preparing the PQR and certifying its accuracy is the manufacturer's own responsibility. Any witnessing requirement arises from the construction code, the Authorized Inspector (AI), the contract, or the TPI organization appointed by the customer. On the EN ISO route, the rule of thumb is the opposite: qualification is expected to be witnessed by an examiner or examining body. In short — under ASME, witnessing depends on code or contract; under ISO, the standard itself calls for it.

After welding, the laboratory cuts the specimens, runs the mechanical tests, and enters every actual value and result on the PQR. If the tests meet the acceptance criteria, the PQR becomes approved. In the final step, one or more production WPSs are written within the valid range that approved PQR supports.

The most misunderstood point in this chain is simple: the PQR is born before the WPS. First you prove the recipe, then you write the production instruction on the strength of that proof. A single solid PQR can support many different WPSs — and the reverse holds too: one WPS may be supported by more than one PQR (ASME QW-200.1). That is exactly what happens on a multi-process joint: GTAW for the root, SMAW or FCAW for the fill and cap passes.

3. Variable classification: the ASME three-tier system

The concept of the "variable" sits at the heart of procedure approval. How that concept is categorized, however, differs between systems. The three-tier classification below is specifically the framework used by ASME BPVC Section IX, listed process by process in the QW-250 series variable tables (QW-252, QW-253, QW-255, and so on).

Essential variables are parameters that, when changed, significantly affect the mechanical properties of the weld — tensile strength and ductility — and therefore demand requalification with a new PQR. Typical examples: a change of process, a change in the filler metal F-number (and, where toughness is required, the A-number as well), a change in material group (P-number), and thickness changes beyond defined limits. Note one trap: an AWS classification change that stays within the same F-number is not essential for most processes. What is essential is the F-number, not the classification designation.

Nonessential variables are parameters that do not significantly affect mechanical properties. Changing them requires no new PQR, but the WPS shall be revised. Minor changes in joint detail, certain adjustments in bead technique, and — to the surprise of many readers — welding position fall into this group.

Supplementary essential variables come into play only where notch-toughness (impact) testing is required. They cover parameters critical to toughness, such as heat input and interpass temperature. A variable treated as nonessential in an application without toughness requirements becomes supplementary essential once toughness is specified, and can then trigger requalification.

EN ISO 15614 does not split variables into these three categories. ISO uses the term "essential variable" too; what it has no counterpart for is the nonessential / supplementary essential split. In the European system, each essential variable gets a directly defined range of qualification, and toughness or hardness requirements are called out separately depending on the application class, material group, and testing level. The outcome is similar; the way it is recorded and classified is not. Both systems ask the same question — does changing this parameter degrade a mechanical property? — they just encode the answer differently. These classifications are not arbitrary. They are decades of welding metallurgy codified into standards.

4. PQR mechanical tests: the step that ties paper to reality

What makes a PQR real is the set of demanding laboratory tests the specimens from the test weld must survive. Which tests are mandatory, though, depends on the system and the construction code. Miss that distinction and you end up with either unnecessary cost or an incomplete qualification.

Nondestructive testing — a mandatory step before destructive testing (on the ISO side). Under EN ISO 15614-1, before any specimen is cut, the test piece undergoes visual testing, volumetric testing on butt welds (radiographic testing per EN ISO 17636 or ultrasonic testing per EN ISO 17640), and surface crack detection (magnetic particle, MT, or penetrant, PT). Acceptance is assessed against EN ISO 5817 quality level B, with level C permitted for excess weld metal, excessive convexity, excessive throat thickness, and excessive penetration. No specimen is taken until the NDT is complete. ASME Section IX, by contrast, does not require NDT for procedure qualification — apart from special cases such as the SWPS demonstration coupon. This is one of the most concrete practical differences between the two systems.
Tensile test — a core mandatory test in both systems. Transverse tensile specimens verify the tensile strength of the weld. The acceptance criterion is not where the specimen breaks. Under ASME QW-153.1 and EN ISO 15614-1 (test method EN ISO 4136), the criterion is that the tensile strength obtained shall not fall below the specified minimum for the base material. Even if the specimen breaks in the weld metal or the HAZ, it passes as long as the strength meets that minimum. In ASME this rule has three branches: (1) when two different base materials are joined, the specified minimum of the weaker one governs, (2) where the construction code permits, the specified minimum of the filler metal may be used, and (3) per QW-153.1(d), if the break occurs in the base metal outside the weld and the weld interface, the result is acceptable provided the measured strength is not less than 95% of the specified minimum. A break in the base metal is a useful observation, but it is not an acceptance condition on its own.
Bend test — a core mandatory test in both systems. Root and face bends, or side bends on heavier sections (ASME QW-451 substitutes side bends for root and face bends at t ≥ 19 mm (3/4 in.); on the ISO side the test method is EN ISO 5173), probe the ductility of the weld and the HAZ and the soundness of internal fusion. Per ASME QW-163, no open discontinuity exceeding 3 mm (1/8 in.) in any direction is acceptable on the convex surface within the weld metal or the HAZ; discontinuities outside the weld are not evaluated. Corner cracks get a separate and more generous allowance: unless there is definite evidence of slag inclusion or another internal discontinuity, corner cracks are disregarded unless they exceed 6 mm (1/4 in.).
Charpy V-notch impact test — conditional. This test does not measure fracture toughness. It measures the impact energy absorbed by the weld metal and the HAZ at a specified test temperature (notch toughness, in joules) and gives a comparative indicator of toughness. True fracture toughness in the fracture mechanics sense (K_Ic, CTOD, J-integral) is a separate class of testing (the EN ISO 12135, BS 7448 / BS 7910, ASTM E1820 route). The test temperature is not the operating temperature either; it is the temperature defined by the construction code or the application standard. Under ASME Section VIII, that is the minimum design metal temperature (MDMT). In ASME, impact testing applies only when the construction code demands it — for example Section VIII or the ASME B31 piping codes (B31.1, B31.3, and so on). Under EN ISO, it becomes mandatory not only through the application standard but also on the basis of material group and a test piece thickness threshold. For the thickness threshold and the required specimen sets, work from the table in the current edition of the standard; the test method is EN ISO 9016.
Macro examination and hardness — system and level dependent. These two items are not a universal PQR requirement. In ASME Section IX, the mandatory tests for groove procedure qualification are tensile and guided bend only; macro and hardness are called for solely where the construction code, a material property, or a special case such as overlay requires them. In EN ISO 15614-1, macro examination (EN ISO 17639) is the general rule, while hardness testing (EN ISO 9015-1) is required for ferritic and hardenable material groups and generally not for austenitic stainless and nickel-based groups. The level distinction introduced in the 2017 edition also governs here: at Level 2, macro sections and hardness are routinely required; at Level 1, macro is required and hardness only when the application standard calls for it. A macro section reveals bead sequence, fusion discontinuities, and penetration at low magnification.

The reason hardness is measured is often explained badly. In C-Mn and low-alloy steels, the primary justification for the HAZ hardness limit is hydrogen-induced cold (delayed) cracking. Rapid cooling produces a hard martensitic structure, and hard martensite combined with diffusible hydrogen (H_D) and residual stress is extremely prone to delayed cracking.

The hardness ceilings set by EN ISO 15614-1 are not a single number; they vary with material group and heat treatment condition. In the as-welded condition, the limit is on the order of 380 HV10 for ISO/TR 15608 groups 1–2 and 420 HV10 for group 3; in the heat-treated condition the limits are lower (typically 320 HV10 for groups 1–2 and 380 HV10 for group 3). For groups 4–5, the 2017 revision raised the limit from 320 to 350 HV10. Because these values are updated from edition to edition, always take the exact number from the hardness table in the standard currently in force.

The 250 HV10 limit for sour (H₂S-containing) service does not come from EN ISO 15614-1. Its source is ANSI/NACE MR0175 / ISO 15156-2, which sets a maximum of 250 HV10 (roughly 22 HRC) for the base metal, weld metal, and HAZ in carbon and low-alloy steels. For refinery equipment in wet H₂S / HIC conditions, service documents such as NACE SP0472 may impose lower hardness targets on top of that. If you need a separate hardness ceiling for hydrogen service, take that number from the relevant code or the project specification too — from an API 941 assessment, for instance — and do not present it as a general rule. Excessive hardness also lowers toughness and raises the risk of sulfide stress cracking (SSC) and stress corrosion cracking (SCC), but those are secondary, service-specific justifications.

If even one of these tests fails to meet its acceptance criteria, the PQR is not approved. The welding engineer revises the pWPS and the chain starts over.

5. Coverage logic: range of qualification

A PQR is usually produced with a limited set of thicknesses, diameters, and parameters — although a single PQR may cover more than one test coupon. Production, meanwhile, throws dozens of thicknesses and configurations at you. This is where the range of qualification comes in. The standard sets rules for how much production a single test weld covers.

For example, a test weld made at a given thickness normally qualifies a defined range above and below that thickness. Material grouping systems (P-numbers and Group numbers in ASME; ISO/TR 15608 groups and subgroups under EN ISO — adopted in Europe as CEN ISO/TR 15608) allow a single qualification to cover similar materials within the same group.

In ASME, these two numbers do not play the same role: the P-number provides the basic grouping, while the Group number becomes restrictive only as a supplementary essential variable, when notch-toughness testing is required. On a job with no toughness requirement, a change of Group number alone does not require requalification. This is one of the most frequently confused points in the field.

Welding position works differently in the two systems as well. In ASME Section IX, position is a nonessential variable for procedure qualification (QW-405.1, for all the main processes — SMAW, GMAW, FCAW, GTAW, SAW, PAW, OFW) and may be changed on the WPS. That means a PQR qualified in 1G can support a WPS covering all positions. (Where toughness was required, the QW-405.2 supplementary essential variable used to apply; that paragraph was removed in the 2019 edition.) By contrast, position does directly limit coverage in welder qualification (the QW-461 tables, EN ISO 9606) and in EN ISO 15614-1. A position that is free at procedure level but restrictive at welder level is one of the asymmetries most often missed on site.

One more critical warning belongs here: the grouping logic of the two systems does not map one-to-one. In ASME, the P-number and the Group number work independently; under ISO, the group/subgroup structure follows a different chemical-mechanical rationale. Shortcut equivalences of the "P-No 1 = Group 1" kind are not safe.

This mechanism keeps both cost and schedule under control by removing the need to test from scratch on every new job. Misreading the range of qualification is one of the most common causes of nonconformity in the field — discovering after production that the PQR in your hand does not actually cover the work is an expensive mistake.

6. The ASME Section IX approach

ASME BPVC Section IX carries the rules of a pressure vessel and piping world with North American roots. Its approach is highly systematic: materials are classified by P-numbers and Group numbers, filler metals by F-numbers and A-numbers. For each welding process, essential, nonessential, and supplementary essential variables are separated cleanly in the QW-250 series tables.

The Section IX philosophy emphasizes proving that the procedure delivers mechanical strength and soundness. Toughness tests such as Charpy V-notch enter the picture, together with supplementary essential variables, only when the construction code demands them. The mandatory test set is kept narrow, and matters such as NDT and witnessing are left to the construction code. The result is a flexible but disciplined system, well suited to deriving a wide production envelope from a single PQR.

7. The EN ISO 15614 and 15609 approach

In the European world, EN ISO 15607 defines the general qualification rules and terms, the EN ISO 15609 series defines the content of the procedure specification (with separate parts by process), and the EN ISO 15614 series defines qualification by testing. EN ISO 15614-1 is the primary reference for procedure tests covering arc and gas welding of steels and arc welding of nickel and nickel alloys. Gas welding covers steels only; gas welding of nickel alloys falls outside the scope of this standard.

Some distinct features of the EN ISO approach: it prescribes the test piece geometry and the number of specimens to be taken far more rigidly, it places nondestructive testing ahead of destructive testing as a mandatory step, it makes macro examination and — for the appropriate material groups — hardness a routine requirement, and it brings toughness requirements to the fore more often, tied to application class and thickness. The thickness and diameter ranges of qualification are set by different formulas than those in ASME.

For the purposes of this article, the most important recent development is the Level 1 / Level 2 testing-level distinction introduced by EN ISO 15614-1:2017. Level 1 was written directly on the basis of ASME BPVC Section IX requirements: fewer tests, narrower ranges of qualification. Level 2 follows the earlier editions of the standard, with broader test coverage and ranges that are more flexible for some variables and more prescriptive for others. The relationship between the two levels runs one way only: a qualification carried out to Level 2 automatically satisfies Level 1, but not the reverse. Where the contract does not state a level, Level 2 applies. So when you inherit an ISO WPQR, one of the first questions to ask is "which level?" That single piece of information changes both the list of tests performed and the production envelope the document covers.

The European system also positions procedure qualification within the broader quality assurance framework of EN ISO 3834 (quality requirements for welding), and gains flexibility from the 15610–15612 routes without testing and the 15613 pre-production test weld route discussed earlier.

8. Why the difference between the two worlds matters

The ASME and EN ISO systems rest on the same laws of physics. What differs is how they calculate ranges of qualification, how they classify materials and filler metals, which tests they make mandatory, where nondestructive testing sits, what they expect regarding witnessing, and what triggers a toughness testing requirement. That is why an ASME PQR does not automatically validate an EN ISO WPS.

Level 1 of 15614-1:2017 was written to close that distance — directly on the basis of Section IX requirements — and it does make the crossing technically easier. But that is not mutual recognition in itself.

Whichever system the project's contract code names, that is the system the qualification must be built on. An equivalence assessment between the two is theoretically possible, but it is not something a manufacturer can decide unilaterally. For an equivalence to hold, the contract and the applicable construction code have to permit it, and the customer plus the TPI organization — or, on EU-scope work, the notified body — have to accept it in writing. On international projects, one of the biggest time sinks is trying to substitute a document from one system for the other without securing that approval.

From the field

The most frequent error we encounter on site is a mismatch between the actual values on the PQR and the written range on the WPS. Say the test weld was made at a certain heat input and the impact test passed at that heat input — but the production WPS was written to allow a far wider current and travel-speed window, well past the supplementary essential variable limit. The welder raises the current and the weave width to cut down the number of passes and finish faster. Travel speed does not rise proportionally, so heat input climbs — that is simply Q = (U × I × k) / v at work. Slow cooling produces a coarse-grained HAZ, and toughness drifts outside the range the PQR proved. On paper everything looks compliant; in reality, production no longer matches what the qualification demonstrated.

There is a neglected other half to this picture: the lower limit on heat input is every bit as critical as the upper one. Run the same formula the other way and raising travel speed lowers heat input. Very low heat input, or inadequate preheat, means rapid cooling and a hard martensitic HAZ — which is a direct hydrogen-induced cold cracking risk. So the supplementary essential variable cuts both ways: push heat input up and toughness suffers; pull it down and hardness and cold cracking risk go up. Production has to stay inside the range the PQR proved, in both directions.

The way to prevent this is to derive the WPS from the actual PQR values with lower and upper limits clear enough for a welder to apply at the joint without interpretation, and to measure heat input during site surveillance rather than assume it. The second common error is trying to hold the range of qualification in your head. Verify the thickness, diameter, position, testing level, and material group ranges in writing at the start of every job, and you head off the expensive nonconformities that arrive later.

Related standards

ASME BPVC Section IX — Qualification of welding and brazing procedures and of welders and operators (WPS/PQR, QW variable tables; Article V and the mandatory appendices for SWPS use).
AWS B2.1 — The Standard Welding Procedure Specification (SWPS) system.
EN ISO 15607 — Specification and qualification of welding procedures: general rules and terms (definitions of pWPS / WPS / WPQR).
EN ISO 15609 (series) — Content of the welding procedure specification (parts by process).
EN ISO 15610 / 15611 / 15612 — Qualification routes requiring no test weld (tested welding consumables, previous welding experience, adoption of a standard welding procedure).
EN ISO 15613 — Qualification based on a pre-production welding test (includes testing; performed in a geometry representing production rather than on a standard test piece).
EN ISO 15614 (series, particularly 15614-1) — Qualification of welding procedures by testing; the scope of 15614-1 is arc and gas welding of steels and arc welding of nickel and nickel alloys; the 2017 edition introduced Level 1 / Level 2 testing levels.
EN ISO 4136 / 5173 / 9016 / 9015-1 / 17639 — Weld test methods: transverse tensile, bend, impact, hardness (Vickers), and macro/micro examination.
EN ISO 17636 / EN ISO 17640 — Radiographic and ultrasonic testing of welds (volumetric testing of the test piece).
EN ISO 9606 (series) — Welder qualification testing (welder qualification, not procedure).
EN ISO 14732 — Qualification of welding operators and weld setters for mechanized and automatic welding.
ISO/TR 15608 (adopted in Europe as CEN ISO/TR 15608) — Grouping of materials for welding purposes.
EN ISO 3834 (series) — Quality requirements for fusion welding of metallic materials.
EN ISO 5817 — Fusion-welded joints in steel, nickel, titanium and their alloys: quality levels for imperfections (EN ISO 10042 for aluminum).
ANSI/NACE MR0175 / ISO 15156 (particularly part 2) — Materials for use in H₂S-containing (sour) environments; hardness limit for carbon and low-alloy steel welds.
EN ISO 2553 / AWS A2.4 — Welding symbols and representation.

Take it to the field: To keep the welding standards, procedure logic, and site 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 — ASME BPVC Section IX, EN ISO 15614/15609, and EN ISO 3834 foremost among them. From preparing the pWPS to witnessing the test weld, and from evaluating PQR mechanical tests to transferring the approved WPS into production, we document every step, so the results stay recordable, traceable, and auditable.