The client experienced multiple unexpected failures in the underground firewater (FW) network at one of their facilities. The network utilizes fiberglass-reinforced plastic (FRP) piping. To determine the root causes of these failures, a material investigation was conducted by Dynaflow Research Group on pipe samples. This case study presents the methodology, testing activities, and key findings from the investigation performed.
Sample Selection and Preparation
The investigation centered around two distinct pipe sections, IRM1 and IRM2, both DN500 diameter but differing in configuration. IRM1 represented a clean, straight section of FRP pipe, while IRM2 comprised two short pipe sections joined by a cemented coupler joint. Both were taken from used but undamaged portions of the firewater network to ensure representative sampling. Samples were extracted using a water cutter and grouped based on their origin.
Loss on Ignition (LOI) Testing
LOI testing, performed in accordance with ASTM D2584, quantified the glass-resin ratio in the FRP laminate. Samples from IRM1 were subjected to ignition at 550–650°C, burning off the resin and leaving only glass fibers. Mass measurements before and after ignition enabled calculation of the glass-resin ratio. The average result was 78% by weight (60% by volume), which falls within the ISO 14692 recommended range (65–80% by weight) for filament-wound pipes. This confirmed that the as-installed laminate composition met industry quality criteria.
Thermal Analysis (DSC)
The glass transition temperature (Tg) of the laminate resin was established using Differential Scanning Calorimetry (DSC) on a random sample from IRM1. The measured Tg was 151°C, a typical value for epoxy-based FRP systems.
Chemical Analysis (FTIR)
Fourier-transform infrared spectroscopy (FTIR) was performed on two samples from IRM1 to assess the chemical integrity. The spectra obtained matched reference spectra for epoxy composites, with no evidence of unexpected chemical degradation or contamination.
Visual and Microscopic Examination
A combination of visual inspection (in accordance with ASTM D2563-94) and optical microscopy was used to assess macro- and microstructural integrity. IRM1 exhibited no visible defects without magnification. IRM2, however, revealed incomplete bonding between the pipe and its coupler joint, as visually confirmed and highlighted in Figure 7.1. This incomplete bonding affected joint strength and was further correlated with mechanical test results. Microscopic examination of IRM1 samples revealed small air bubbles (~0.3 mm diameter) at the cut surface.
Water Permeation Testing
Permeation tests, following ASTM D570-98, evaluated the water absorption characteristics of the laminate. Two IRM1 samples were immersed in either substitute ocean water or demineralized water at 23°C for 24 hours. The observed weight increases were 2.1% and 0.6% respectively.
Barcol Hardness Testing
Barcol hardness measurements were performed at multiple locations on both IRM1 and IRM2 according to the procedure in ISO 14692. Average hardness values were 45 (IRM1) and 44 (IRM2).
Mechanical Testing: Compressive and Tensile Properties
Compressive ring stiffness, measured per ASTM D2412, yielded a specific initial tangential stiffness (STIS) of 12,830 Pa for the tested rings, with a back-calculated hoop-direction bending modulus (EH,bending) of 22,000 MPa.
Tensile testing was performed on samples from both pipe sections. For IRM1, both axial and hoop direction samples were assessed; for IRM2, tests focused on the glued coupler joint. The tensile strength results highlight that the incomplete bonding at the IRM2 joint adversely impacted tensile performance.
Conclusions and Key Technical Takeaways
The in-depth material investigation of FRP firewater pipes from the client’s facility identified the following:
- The FRP pipe wall laminates (IRM1) conform to industry standards in terms of glass-resin ratio, thermal properties, chemical integrity, and mechanical performance. No evidence was found of systemic manufacturing flaws or resin degradation.
- The primary vulnerability identified was incomplete bonding at the coupler joint in IRM2. This defect directly reduced tensile strength at the joint and likely contributed to the observed network failures.
- Microvoids of typical size were observed, but their distribution appeared consistent with standard FRP manufacturing and not excessive.
- Water absorption and Barcol hardness tests indicated satisfactory resin cure and moderate, acceptable permeation characteristics.
In summary, the root cause of the underground firewater network failures was not due to inherent deficiencies in the as-manufactured FRP pipe but was instead attributable to joint construction quality, particularly incomplete bonding in coupler joints.