Technical Evaluation of Three Candidate GRP Pipe Suppliers
For a glass-reinforced plastic (GRP) pipe system on a project, the client selected three potential GRP pipe suppliers, each to be evaluated from a technical perspective. Each supplier submitted a set of technical documents describing its component design, qualification testing and system design practice. The task was to review these documents against the project’s governing code of practice and to indicate the missing aspects. This case study covers the review of one supplier whose component design seems to be based on German DIN type standards.
A Performance-Based Code and a Four-Aspect Review Framework
The review examined four aspects of the submission: the component design and its basis, the qualification testing performed to date, the system design and its basis, and compliance with the project code. The project code is performance-based: all properties claimed and used by the supplier or a third party in the stress analyses have to be demonstrated by appropriate medium-term tests. The supplier’s correspondence cited experience with DIN 16965.
The basis of the DIN type standards is the short-term properties of the laminate or component, knocked down by part factors. The applicable part factors are not mentioned in the submitted documents, but since reference is made to DIN, the value needs to be 6. DIN 16965 imposes a lower confidence limit (LCL) of 60 MPa, based on a maximum circumferential tensile strength of 360 MPa and a safety factor of 6. This short-term value of 360 MPa needs to be validated by tensile tests.
How the Submitted Documents Compare With the Manufacturing and Qualification Requirements
The submitted package comprised email correspondence, a scope-of-work quotation for a similar project, a product handbook, a piping programme overview, piping design calculations, and a system stress analysis received as a late document. The handbook states that reinforcement is applied by filament winding or hand lay-up, as the code’s manufacturing chapter requires. Flanges are produced to ANSI B16.5 for standard 150#. The code requires GRP flange outside diameters and hole spacing to meet ASME B16.5, and requires the flanges to be flat faced. Flange manufacture to ASTM D5421 (hand lay-up), ASTM D4024 (filament winding) or by post-curing, as described by the code, is not mentioned explicitly for fittings.
The code’s qualification chapter defines a three-part programme: full regression-line testing on family representatives to establish the LCL, 1000-hour survival testing on product-sector representatives, and quality-control baseline testing on product-sector representatives. None of the submitted documents shows that medium-term or long-term tests on pipe or fittings have been performed, and no regression data were included. The handbook cites proof testing to ASME B31.3, which prescribes a burst test method and a hydrostatic test method, of which one has to be executed, but the results are not present in the information provided. The catalogue data are mainly ultimate (short-term) data, and it is unclear whether they are based on tests. The design therefore apparently rests on short-term property values, and compliance with the required short-term stress values still needs to be demonstrated.
Baseline Data and System-Design Inputs Under Review
No glass content data were provided. DIN 16965 requires a glass content of 60 ± 5%, whereas the project code prescribes 65 to 80% for filament-wound pipe, 65 to 75% for filament-wound fittings and 50 to 65% for hand lay-up fittings. If the supplier complies with DIN, the glass content for the pipes will be lower than the code requires. The code requires the degree of cure (Tg) to be qualified, and the Tg of the qualified component cannot be lower than the resin manufacturer’s minimum. No degree of cure data were provided, so the code requirement that the design temperature is at least 30 °C lower than Tg cannot be evaluated. For the metric pipes produced to DIN 16965-2, several pipe thicknesses are smaller than the 3 mm code minimum, so the minimum reinforced wall thicknesses do not comply. No baseline dimensional measurements were provided; basic dimensions for bends and tees are given to DIN 16966.
On system design, the submitted documents indicate that the stress and flexibility analysis is often outsourced. A typical analysis used CAESAR II and was based on ISO 14692. Its input allowable stress data were based on catalogue values, with no information on their substantiation, and it is unclear whether they are long-term data and how they have been qualified. The handbook support spans, based on DIN 16965, are slightly larger than the code’s tabulated values for DN 50 and smaller, and shorter for larger diameters. A second span table in the installation manual, for filament-wound dual laminate pressure pipe, gives spans that differ strongly with the selected FRP wall thickness, and its acceptability has to be demonstrated during system design.
Required Qualification Scope and Allowable-Stress Basis
Because none of the submitted documents shows medium-term or long-term testing, all pipes and fittings used in the project have to be qualified by 1000-hour tests on family-representative pipes and fittings at the applicable test pressure and test temperature, which for GRVE is at least the project design temperature. The 1000-hour qualification tests also qualify the stress that can be used in the stress analysis, and the tests have to be repeated for representatives of each different manufacturing process. The code also requires baseline data on degree of cure, glass content, minimum thickness and key component dimensions, the last measured in accordance with ASTM D3567.
Although the component design refers to DIN (60 MPa), the value for the long-term allowable stress used in the system stress analysis received as a late document is different, and the basis for this value is unclear.
Review Conclusions
The review concluded that the component design seems to be based on German DIN type standards, whose basis is the short-term properties of the laminate or component knocked down by part factors. Under the performance-based project code, all properties claimed by the supplier have to be demonstrated by appropriate tests.
All pipes and fittings therefore have to be qualified by 1000-hour tests, which also qualify the stress used in the stress analysis. If the manufacturing process complies with DIN, the glass content of the pipes will be lower than the code requires, and the long-term allowable stress used in the system stress analysis differs from the DIN value with an unclear basis.