GRP pipes have a lifetime of 150 years – here’s the proof

Högni Jónsson, head of product development and support of the Amiblu Group

“Engineered for the next 150 years”: it’s quite a bold claim that Amiblu makes about its GRP products. But how can anyone possibly know, given that the company itself is just over 60 years old? Amiblu Product Development Manager Högni Jónsson provides insights into some interesting research.

Amiblu promises a lifetime of 1.5 centuries for its GRP products. Where’s the proof?

Jónsson: To answer this, we need to look at the reasons why most pipes do not reach this mature age. These reasons carry names such as Thiobaccillus concretivorus and Thiobaccillus ferrooxidans – tiny bacteria that decompose sewage and form hydrogen sulphide gas. When combined with moist air, the gas forms sulphuric acid, which is highly corrosive to materials like concrete, steel, and cast iron and causes significant damage over time. Corrosion is also a major problem in seawater applications, where the contained sodium chloride eats away at e.g. metals and causes them to fall apart.

And with GRP products, the situation is different?

Jónsson: Exactly. Plastics are inherently more robust than both concrete and metals in acidic environments. To prove this, we literally put our pipes to the acid test: Several pipe samples are exposed to sulphuric acid for a considerable time, while being subjected to artificially high tensile strains. The idea is to simulate the chemical conditions in aggressive sewage, but under an excessive strain in order to cause failure within a reasonable timeframe. To determine the pipes’ long-term properties, the measured data is analyzed statistically and extrapolated into the unknown to predict a limiting strain for use in pipe design.

Did Amiblu invent this test method?

Jónsson: No. The acid test for GRP pipes was first standardized by the American Society for Testing and Materials in 1978 (test method ASTM D3681) and has been in use ever since. The same procedure is also specified in EN 1120 and ISO 10952. The test method calls for minimum 10,000 hours of testing with at least 18 samples. However, since the test is relatively simple to conduct and does not require much space, a great number of samples have been left exposed to acid for much longer time periods. We studied the results from over 40 years of continuous testing of Flowtite and Hobas GRP pipes, involving more than 1800 test samples.

This sounds comprehensive. Please tell us a little more about these samples and the test setting.

Strain corrosion tests according to ASTM D3681 in the Amiblu R&D center in Norway.

Jónsson: Our tests cover a variety of pipe designs which have been in continuous use. The samples, most of them DN 600, were taken from a number of Flowtite and Hobas manufacturing plants. According to ASTM D3681, we subjected each 300 mm long pipe sample to a vertical force causing tensile bending strain in the pipe invert, while exposing it to a 5% concentration of sulphuric acid.

A typical test series consists of 18-25 samples, usually from a single production batch, at various strain levels. The strain is measured after the load has been applied and then the sample is stored under controlled conditions until failure occurs, detectable as leakage through the pipe wall. With at least one data point exceeding 10,000 hours and the rest relatively evenly spread over the time range, and with an appropriate coefficient of correlation, the data can be safely used and extrapolated with classic statistical methods.

After 33 years exposure to sewage and seawater in the Norwegian town of Sandefjord, a Flowtite GRP diffuser was brought ashore for analysis of its condition and mechanical properties.

How about the results?

Jónsson: The longest, still-running test was set up on October 4, 1978. The sample has now been exposed to the acid test for more than 40 years at 0.91 % strain. What’s really interesting here is the bilinear behaviuor: up to strains of about 1.6 %, most samples fail within relatively short time periods. At strains between 0.9 % and 1.3 %, the time to failure is much longer. Only a handful of data points fall below this range, meaning that below a certain “threshold strain”, the samples simply do not fail. For this set of data, the threshold appears to be around 0.9 % strain. A classic regression analysis of the data points up to 1000 hours results in a line with a mild slope, while regression analysis of data points after 1000 hours through to over 350,000 hours show an almost horizontal line. By extrapolating this line by only ½ of a decade, which is less than one third of what classical statistics allows, we reached a 150-year strain value of 0.93 %.

Is this a typical strain in sewer or similar applications?

Jónsson: No, and that’s the fantastic conclusion: the typical long-term operating strain of such a pipe is merely 0.27 %. This means that, in real-life applications, we even reach an excellent safety margin of 3.4, and the pipes can serve several future generations.

This sounds really impressive! Any project example that proves this durability?

A sample of the GRP diffuser in Sandefjord. After cleaning, the pipe looked almost brand-new. Its mechanical properties were also basically unchanged.

Jónsson: In northern Norway, a Flowtite GRP diffuser was installed as part of a subaqueous marine outfall in 1975. In 2008, it was brought ashore for analysis of its condition. After cleaning parts of it, the pipe looked almost brand-new and its mechanical properties were also basically unchanged.