Why a single assembly may need several different kinds of engineering evidence. A fitting can look like one product on a drawing, yet contain several pressure-retaining parts. Some may be covered by recognized standards. Others may need calculations, further analysis, or testing. This is where a CRN application for a nonstandard assembly can become more involved than expected. The experience below comes from an anonymized five-component fitting registration in Alberta. The aim is to explain the approach, not disclose the design or reproduce its project records.
Before choosing a calculation method, define the pressure boundary. Is the application for the machined body alone or for the complete unit, including its purchased connections and instruments? That decision affects which components need supporting information.
CSA B51 separates fittings into categories. For example, Category A covers certain conventional fittings, while Category F includes measuring and indicating devices. Category H applies to particular pressure-retaining components or assemblies that meet its provisions. An unusual shape does not, by itself, make a fitting Category H.
A clear boundary drawing should identify the pressure-containing parts, their connections, materials, and the proposed design conditions. It gives the reviewer a shared starting point and reduces the risk of later misunderstandings.
One of the most useful early checks is whether each purchased component actually complies with a standard recognized by the governing construction code. Being sold as a “standard fitting” is not the same as being covered by a listed standard with a suitable pressure rating.
For a design using ASME B31.3, the component standard, size, material, dimensions, and rating basis need to be checked. If the part is already registered, its CRN documentation must match the specific model and proposed pressure-temperature conditions. Where neither route establishes the rating, further evidence is needed.
In the completed registration, the main body was machined from steel and had an irregular internal pressure passage. Several smaller parts were connected mechanically. Some arrived with usable registration documents; others came with catalogue information that did not fully establish the pressure-design basis for the proposed assembly.
The main body called for finite element analysis (FEA). Its changing wall sections and intersecting passages were not readily represented by simple wall-thickness equations. The FEA therefore focused on the geometry that needed it, with the material, loads, constraints and stress checks documented for review.
The regular-shaped parts were handled differently. For those items, direct code calculations were clearer than a complex model. The checks included wall thickness at critical locations and the ability of threaded connections to carry pressure-related loads. Material grades, thread dimensions and actual engagement mattered more than a broad catalogue description.
FEA can be helpful when the shape is complex, but it should not automatically be the first step. Where an applicable code formula can answer the design question, a direct calculation is often easier to explain and review.
A detailed simulation also depends on good inputs. It cannot make up for unknown material properties, hidden load paths or uncertain support conditions. This became especially important for one purchased instrument in the assembly.
Part of the instrument was exposed directly to process pressure. Its electrical purpose did not remove the need to establish that its pressure-wetted element could withstand the specified conditions. The manufacturer did not have an applicable CRN for the configuration under review.
The supplier had routine factory pressure-cycling records. These were useful quality records, but they were not the same as a code-based proof test used to establish an allowable pressure rating. A simplified calculation could not adequately represent the instrument’s internal load support. An FEA model would also have required assumptions that could not be confirmed from the available information.
The chosen path was therefore a formal proof test, performed and documented under the applicable code provisions. For certain unlisted piping components, ASME B31.3 paragraph 304.7.2(c) allows specified proof-testing methods, including ASME Section VIII, Division 1, UG-101 where applicable.
Under the relevant UG-101 method, a test is not judged by the highest pressure reading alone. The reported result must be considered with the prescribed calculations, material properties, temperature factors, and any other applicable requirements. Depending on the method and its conditions, testing may be stopped at a sufficient pressure without rupture; neither stopping the test nor achieving a pressure hold, on its own, establishes an arbitrary design rating.
In this case, the accepted proof-test evidence supported a lower pressure than initially proposed. The assembly’s design basis was revised to the pressure that could be demonstrated, and the related calculations, drawings and descriptions were brought into line.
A successful test is only part of the evidence. The reviewer must be able to tell exactly what was tested, how it was tested, and whether the production item matches the tested configuration.
ABSA guideline AB-381e sets out useful documentation expectations: drawing and revision, material details and traceability, test setup, measuring instruments and calibration, observed pressure, reasons for stopping, the applicable calculation, and required inspection or witness records. The authority of the person witnessing or accepting the test should also be clear.
A proof test used to establish a component rating is different from the routine hydrostatic or leak testing of a completed assembly. The tests may involve pressure, but they do not answer the same question.
The final review often comes down to consistency. A well-supported calculation can still lead to questions if it uses a different material grade, drawing revision or design pressure from the application forms.
An Alberta fitting-registration package normally combines the appropriate ABSA application and declaration documents with a scope statement, detailed assembly drawings, component information, and the necessary calculations or acceptable proof-test evidence. Existing CRNs must be tied to the actual products being relied upon. ABSA’s submission page provides the current administrative requirements and permitted form alternatives.
The registration did not depend on one sophisticated calculation. It came together because each pressure-containing part had a clear verification route: code calculations where they applied, FEA for the difficult geometry, existing CRN documentation where valid, and proof testing when the other evidence was not enough.
For future nonstandard fitting applications, a part-by-part evidence list is a useful first document. Record each component, its proposed rating, the method used to verify it, and where the supporting evidence is filed. Any gaps then become visible before the package reaches the reviewer.
The important question is not “Have we analysed the assembly?” It is “Can we show how every pressure-containing part has been verified?”
As of ABSA bulletin IB24-002 Rev.13 (October 1, 2026), the relevant in-force editions for this discussion include CSA B51:19, ASME B31.3:2024 and ASME BPVC:2025. CSA B51:24 was not adopted in Alberta under that bulletin. The governing edition, registration route and applicable provisions should be confirmed for each actual submission.
Technical note: This article is educational and describes a generalized case. It is not a design approval, code interpretation or substitute for qualified engineering and regulatory review.
Mr. / Mrs.