Analysis of Premature Failure of a PVC Water Pipe

A case study of the failure analysis of a PVC drinking water pipeline is presented. One week after installation in a construction zone of a building, a longitudinal cracking, and pressure drop occurred in an underground pipeline, and an investigation was carried out to determine the causes of the event. Based on the results of visual inspection, dimensional metrology, plastography, microhardness, fractography, FTIR analysis, and stress analysis, the failure mechanisms of the pipe were identified as creep and slow crack growth. The cause of the development of these failure mechanisms was the exposure of the pipe, to a mixture of paint thinner and polyurethane paint, which generated plasticization, with the corresponding deterioration of the mechanical properties. The source of the plasticizer could have been an accidental spill or an incorrect final disposition of this chemical residue during the construction process of the building.


Introduction
In pipes for potable water in buildings, it is common to use thermoplastic materials such as PVC (polyvinyl chloride), PP (polypropylene), ABS (acrylonitrile butadiene styrene), and PE (polyethylene), among others. The advantages of using these materials instead of metals are lower installation cost, greater resistance to corrosion, lower density, greater flexibility, lower hydraulic losses (interior finish with low roughness), and low tendency for scale formation [1]. For cold or hot potable water pipes, rigid PVC is usually used, with an expected lifespan of at least 50 years.
Rigid PVC pipes can conduct water, since there are no changes in the properties of the material, and the fluid is not contaminated; however, some organic solvents can produce adverse effects on PVC. Aliphatic alcohols in general have good chemical compatibility with PVC at room temperature; however, at high temperatures, there may be changes in the mechanical properties of the material. Aromatic compounds generally have a severe effect on PVC, changing its mechanical properties, and the same thing happens with acetone [2]. Thermoplastic polymers, including PVC, are susceptible to creep deformation at room temperature under constant load (cold flow) [3]. PVC pipes exhibit less creep deformation compared to other polymeric materials, which allows them to undergo only a small increase in diameter, and therefore, it takes a long time for them to exhibit cracking under the action of internal pressure [4]. If plasticizers are added during the manufacture of the PVC pipe, or if an aggressive environment during its service diffuses an organic solvent into the polymer, the deformation speed will increase, and the length of time before cracking will decrease. Creep associated cracking is caused by SCG (slow crack growth).
Environmental stress cracking (ESC) is a failure mode in which a glassy thermoplastic polymer exhibits cracking at a lower stress level than its tensile strength, in the presence of a chemically aggressive environment. It is estimated that between 15 and 40% of polymer failures could be due to this failure mode [5,6]. In this failure mechanism, the chemical compound can dissolve part of the material and diffuse into it, lowering the amount of stress necessary for the formation of crazes or cracks and generating residual stress due to the swelling of the material. The diffusion of the chemical compound and the plasticization of the material are facilitated in stress concentrators and crack fronts [7][8][9][10]. PVC is susceptible to ESC when attacked by organic liquids (some alcohols and solvents) and alkaline solutions [11].
The case under study in this paper corresponds to the cracking failure of an underground PVC potable water pipe, with a nominal size of 50.8 mm SDR 21 according to ASTM D2241 (maximum working pressure of 13.8 bar at 23°C) [12]. The pipeline was installed new and passed the hydrostatic test, and later, it was covered with fine stone and machine compacted sand backfill. Finally, a concrete sidewalk was poured. The pipeline was pressurized to 8.3 bar with potable water at 10-15°C, exhibiting a pressure drop approximately one week later.
Once the pressure drop was detected, the soil was excavated until the pipeline was reached, finding a cracked pipe in a zone with a change in color (chemically affected area), while other areas without cracking did not show this change (Fig. 1). The cracking was longitudinal and located in the lower part of the pipe (Fig. 2).

Experimental Procedure
To identify the root cause of failure, two sections of pipe were cut for analysis in the laboratory, one containing the crack (chemically affected area) and the other in the unaffected area. These samples were used to perform visual inspection, dimensional metrology, plastographic analysis, optical and SEM fractography, microhardness measurement, and FTIR spectroscopy.

Visual Inspection
The crack that formed in the chemically affected area of the pipe followed a longitudinal trajectory, which indicates that it propagated under the action of the stresses derived from internal pressure. Additionally, there was a presence of deformation by creep and swelling, or in other words ballooning by softening of the material [13] (Fig. 3), which are symptoms of PVC plasticization.

Dimensional Metrology
Using a Mitutoyo caliper with a resolution of 0.05 mm, measurements of external diameter and pipe wall thickness were taken in the two analysis sections. In the chemically unaffected area, the average external diameter was 60.36 mm, and the wall thickness was 3.24 mm, while in the crack area, the external diameter was 63.29 mm with a wall thickness of 2.51 mm. This was a consequence of creep deformation and swelling.

Plastography
A section for plastographic analysis designated AA adjacent to the main crack (Fig. 4), which contained chemically affected (with color change) and unaffected material, was cut, mounted in epoxy resin and polished and was observed with a reflected-light optical microscope (LECO 500 microscope with brightfield illumination). A dividing line was observed between the two materials, indicating the presence of a diffusion front of the plasticizing chemical (Fig. 4).    Table 1. The chemically affected material exhibited a marked decrease in microhardness, which is evidence of the plasticization of the material.

Fractography
The surface of the longitudinal crack showed high roughness, crack origin in the interior of the pipe (water side), with radial and beach marks (Fig. 5). Nucleating and propagating the cracks through craze formation (Fig. 6). This is characteristic of slow crack growth (SCG). For the fractographic inspection, an Olympus SZX-12 stereo microscope and a FEI Quanta 200-r electronic microscope were used. In SEM micrographs with secondary electrons, an acceleration voltage of 25 kV and ESEM mode were obtained.

FTIR Analysis
FTIR spectra were taken on the chemically affected material near the longitudinal crack and on the unaffected material, Fig. 1, using a Thermo Scientific Nicolet iS10 infrared spectrometer with an ATR module. The spectra and the main absorption peaks are shown in Fig. 7.
In the FTIR spectra of the unaffected material, the characteristic vibration peaks of PVC can be seen: CH stretch (2919.67 cm À1 ), CH 2 deformation (1326 cm À1 ), CH rocking (1245.26 cm À1 ), trans CH wagging (961.48 cm À1 ), C-Cl stretch (875 cm À1 ), and cis CH wagging (615.70 cm À1 ) [14,15], while in the spectra of the affected material, the same peaks are present, except that the peak corresponding to the C-Cl stretch disappears and now a new peak appears at 2350.76 cm À1 , probably corresponding to shifted isocyanate group band, due to the dipole-dipole interactions with PVC. The isocyanate group stretch absorbs infrared strongly in the region of 2300-2250 cm -1 [16].
In the FTIR spectra of the affected material, there was a marked presence of peaks associated with alcohols: OH stretch (3387.51 cm À1 ), C-C-O stretch (1089.68 cm À1 ), and C-O stretch (1032.61 cm À1 ); additionally, there also were characteristic peaks of aromatic compounds: C=C ring stretch (1646.20 and 1427.69 cm À1 ), C-H out-ofplane bend (775.11 cm À1 ), and CH stretch (2854.91 cm À1 ).
The spectra showed that the substance that affected the material of the pipe in the cracking zone was probably rich in alcohols, aromatic compounds, and isocyanates. In addition, this substance also reached the unaffected material, although to a lesser extent.

Stress analysis
In the original pipe, the inner radius was r 1 ¼ 26:94 mm and the thickness was e ¼ 3:24 mm, so the radius-tothickness ratio was 8.31, which means that for stress analysis, the pipeline should be considered a thick-walled cylinder. Since the pipe was pressurized before plasticization to p ¼ 8:3 bar and the outer radius was r 2 ¼ 30:18 mm, the value of the maximum hoop stress in the inner surface of the pipe can be estimated through Equation 1 [17], from which a value of r h;max ¼ 7:31 MPa is obtained (Fig. 8): The longitudinal crack was presented for r 1 ¼ 29:135 mm and thickness e ¼ 2:51 mm (creep deformation and swelling, consequence of plasticization), which for stress analysis classifies as thin-walled cylinder (r h ¼ 9:6 MPa), Eq 2: According to ASTM standards D2241 and D1784 [18], an SDR 21 pipe with a nominal diameter of 50.8 mm must be made of rigid PVC, with a minimum tensile strength of 34.5 MPa. The foregoing indicates that the pipe failure occurred under a stress between 21.19 and 27.83% of tensile strength, which is consistent with the creep deformation and slow crack growth (SCG) of plasticized PVC.

Discussion
From the experimental results and the stress analysis, it can be inferred that in the PVC pipe, two failure modes operated simultaneously: accelerated creep deformation and SCG. Both failure modes had their origin in a previous plasticization that the material underwent in the failure zone, a product of the absorption from the soil of a plasticizing chemical. This chemical was absorbed and diffused in the pipe material, producing a decrease in the mechanical strength (plasticization or softening associated with the presence of solvents and changes in the PVC structure), an increase in the creep deformation speed under the internal fluid pressure, and swelling and cracking by nucleation and propagation of crazes (SCG). The failure process was rapid (approximately one week) and occurred under a hoop stress between 7.31 and 9.6 MPa, which is one order of magnitude lower than the expected minimum tensile strength of the material (34.5 MPa). This failure mode could also be classified as environmental stress cracking (ESC); however, in general, ESC is not a ductile fracture as in this case [19]. From the FTIR spectra, it was determined that the plasticizing chemical was a mixture of alcohols, aromatic compounds, and isocyanates, which may correspond to a mixture of paint thinner and polyurethane paint. Commonly the paint thinner is mixtures of alcohols, xylene, toluene, ketones, and esters in variable quantities, and the polyurethane paint contains isocyanate compounds. Since the installation of the pipe was done in a building during the construction phase, it is likely that this substance had been poured into the soil near the pipe during the covering and compaction with fine stone and sand. This spill could have been accidental or due to an incorrect final disposal of

Conclusions
Failure analysis revealed that the PVC pipe failed due to creep deformation and SCG.
The failure mechanism initiated in the plasticization of the material, probably produced by the absorption of a mixture of paint thinner and polyurethane paint.
The plasticizer affected the PVC pipe due to an accidental spill or an incorrect final disposition of this chemical residue during the construction process of the building.
Funding Open Access funding provided by Colombia Consortium.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/.