Landkarte mit zwei schwarzen Ringen und roten Kennzeichnungen

Figure 1: Transport pipelines for water supply (main network). (Bild: alle Kiwa Nederland B.V.)

Querschnitt in blau mit positioniertem Ring in rot markiert.
Figure 2. Principle of pipe joint. (Bild: alle Kiwa Nederland B.V.)

In The Netherlands there are large transport networks for the supply of drinking water and industrial water. The majority of these pipelines have been installed in the years between 1950 and 1980. The main two networks were realised in the densely populated western part of the country. One network, with pipelines dating from 1954 to 1957 and from 1963 to 1966, consists of large concrete pipelines for the transport of water from the Rhine and from the IJsselmeer to the large sand dunes at the coast. The water is filtrated in the dunes. The thick sand layers act as a filter for purification of the raw water into drinking water that is distributed to the Amsterdam area. The other network is for the water transport from large water collection basins in the Biesbosch to the Rotterdam area. This network was realized in 1972. Besides these two networks there are other water transport networks of concrete pipes in The Netherlands, in total comprising over 1.200 km.
In the last 16 years there were several opportunities to sample rubber sealing rings out of these pipelines during maintanance works and to investigate the remaining quality and sealing function of rubber rings that have been in use in large concrete pipes for water transport [1].

2. Construction of the pipelines

Aufgegrabene Erde mit innen liegender Pipeline.
Figure 3. Preparation of dismantling. (Bild: alle Kiwa Nederland B.V.)

The pipelines consist of 6 m long concrete segments with a spigot and a socket end. The pipes of the two networks have a diameter larger than 1.000 mm.
In these systems thousands of rubber rings are used and replacing them is impossible without dismantling the joints. The cross section of the rubber rings is in the range of 20 to 24 mm, depending on the specific pipeline type and size. The ring cross section was selected to enable a compression of about 40 % on the mounted rings.

3. Experimental - 3.1. Sampling of rubber rings

Pipeline mit Anschnitt.
Figure 4. Dismantling of joint. (Bild: alle Kiwa Nederland B.V.)

From time to time maintenance works have been carried out on the two mentioned networks. On several locations rubber rings have been sampled from the pipe joints after disconnection of the joints. This was possible at several occasions between 2002 and 2018. The sampled rings were marked for their mounted position and flow direction of the water. The rings had been in service for various periods, from 46 up to 64 years.

3.2. Investigation program

Pipeline mit Dichtungsring.
Figure 5. Spigot with ring.

The appearance of all rings was investigated and recorded, including outer surface condition, deformations, cracks, ruptures and other damages and irregularities. Dimensions of cross sections were measured (area, ring thickness in several directions) on different positions along the circumference. The hardness of the outer surface was measured in Shore A according to ISO 48-4 (former ISO 7619-1). The material from the bulk of the rings was tested on tensile strength and elongation at break according to ISO 37 and on tear strength according to ISO 34-2. The compression set was determined by calculation, using the measured ring thicknesses in the compressed direction, the cross sections and the clearances in the joints. In a few occasions parts of the rings were put in closed glass bottles, immediately after disconnection of the joints on location. The samples were directly sent to a specialized laboratory for analysis of microbiological activity. Also in a few occasions the rubber type and the composition of the rubber material was analyzed by pyrolysis fourier-transform infrared spectroscopy (pyr-FT-IR) and thermogravimetric analysis (TGA).

4. Results - 4.1. Appearance of rubber rings

Quite often rings appeared to be permanently twisted. This is probably caused at the construction of the pipeline where the ring is irregularly twisted during the placing of the socket over the spigot end. The twists caused an initial stress in the rubber rings. Although a relaxation of the stress has been taken place during the long service life, the stress was apparantly still high enough and the rubber was still elastic to flip back after the dismantling. In some occasions rings showed some surface damages. These damages can be explained by the fact that the surface of the rings has been in intense contact with the relatively rough concrete surface of the pipe during decades, giving a more or less mechanical bond. During dismantling the rubber surface likely could be damaged at the distortion of this bond. Most of the rings showed degradation and disappearance of rubber material in a straight line along the circumference. This line is the part of the ring that hade been in contact with the water between the spigot and the socket (the water side). The other part of the ring (at the soil side) showed no degradation. The level of degradation and disappearance showed a large fluctuation between the investigated rings as well as within individual rings. Rings sampled from the same location could show hardly any degradation as well as a severe loss of material. Within individual rings there could be a severe loss of material on a part of the ring, while another part of the ring could still be completely intact. There is no relation between the intensity of material losses and the mounted positions of these parts of the rings, like e.g. top or bottom of the pipe. There is also no relation with the service life; some of the oldest rings have no damage at all, while rings with the shortest service life show severe material loss. A small minority of rings had a severe loss of material. At the initial stage of the degradation there is a hardening of the rubber and often there are cracks in radial direction, pointing on a severe reduction of the elongation at break. In the next stage there is a clear presence of hard and brittle, thus degraded rubber. At a progressing degradation the brittle material is broken from the surface and disappeared. The degradation looks like oxidative degradation that causes a change of the molecular structure leading to a dramatic increase of the hardness and stiffness and a dramatic decrease of the strength and elongation at break. This effect only happens at the outer surface.

Ein blauer, gedrehter Dichtungsring.
Figure 6. Example of a twisted ring. (Bild: alle Kiwa Nederland B.V.)

4.2. Dimensions and compression set

Zwei Dichtungsringe im Querschnitt.
Figure 7. Strong degradation of water contact area. (left) Figure 8. Severe degradation of water contact area. (right) (Bild: alle Kiwa Nederland B.V.)

Cross sections were cut out of each ring at three different locations, one from an area without or with only minor degradation, one from an area with medium degradation, one from an area with the most degradation. De areas of the cross sections and the diameters in the compressed direction were measured. The initial diameters were calculated from the measured areas. The compression set values were calculated using the clearances from the pipeline specifications.
Calculated material losses were described as percentages of the total cross section areas. A small minority of the rings showed severe degradation at a part of the ring; the highest calculated material loss was 63 % on a ring that had been in service for 46 years. Calculated compression set values are in the large range from low to medium, between 7 % (a ring being used for 46 years) and 50 % (a ring being used for 52 years). Concerning this, it can be said that there is still sufficient sealing. From a compression set of about 80 % and higher the sealing gets critical.

4.2. Dimensions and compression set

Cross sections were cut out of each ring at three different locations, one from an area without or with only minor degradation, one from an area with medium degradation, one from an area with the most degradation. De areas of the cross sections and the diameters in the compressed direction were measured. The initial diameters were calculated from the measured areas. The compression set values were calculated using the clearances from the pipeline specifications. Calculated material losses were described as percentages of the total cross section areas. A small minority of the rings showed severe degradation at a part of the ring; the highest calculated material loss was 63 % on a ring that had been in service for 46 years. Calculated compression set values are in the large range from low to medium, between 7 % (a ring being used for 46 years) and 50 % (a ring being used for 52 years). Concerning this, it can be said that there is still sufficient sealing. From a compression set of about 80 % and higher the sealing gets critical.

Vier schwarze Dichtungsringe mit unterschiedlichen Gebrauchsschäden.
Figure 9. Cross sections of rings with different stages of degradation and material loss.

4.3. Physical-mechanical properties

Messergebnis vom Härtetest.
Figure 10. Hardness ISO 48-4 (Bild: alle Kiwa Nederland B.V.)

The figures 10 to 13 show the values of the hardness of the unaffected outer surface and the tensile strength, elongation at break and tear strength respectively for the bulk material of rings from the same pipeline as a function of the service life. The hardness of the outer surface shows a very small increase over the years. The values are still within the specification limits (the red lines) for the initial properties [2]. The tensile strength shows a very low decrease over time. After over 60 years of service, the values are still above the specification limit. The elongation at break shows a slow decrease. The values are higher than 300 % after over 60 years of service. The tear strength shows a decrease. Most of the values are still higher than the initial specification limit. The obtained values for the tensile strength, elongation at break and tear strength show that the quality of the rubber in the bulk of the rings is still good, even after long service times. There is no or hardly any loss in properties. The outer non-affected surface is still elastic and the hardness is hardly increased.

4.4. Pyr-FT-IR

Auswertung vom Infrarot Spectogramm.
Figure 14. Infrared spectrogram (Bild: alle Kiwa Nederland B.V.)

The rubber type of various rings was analyzed by pyrolysis fourier-transform infrared spectroscopy (pyr-FT-IR). An example of an infrared spectrogram is presented in Figure 14. In all cases the rubber type was natural rubber.

4.5. TGA

Diagramm mit roter und grüner Linie.
Figure 15. TGA graphs of degraded (red curve) and non-degraded parts (green curve) from the same ring. (Bild: alle Kiwa Nederland B.V.)

The global rubber composition of some rings was analyzed by thermogravimetric analysis. The analysis was carried out on samples from degraded and non-degraded parts from the same ring. Figure 15 shows the results from a ring that had been in service for 55 years. The red curve is the TGA graph from the centre of the ring, the green curve from degraded surface. The elastomer content is decreased at the degraded surface.

4.6. Analysis of microbiological activity

Weiße Punkte von Mikro Organismen in rotem Träger-Material.
Figure 16. Colonies of micro organisms grown on nutrient medium.

Samples from degraded parts of rubber rings were analyzed for the presence of microbiological activity and especially for the presence of micro organisms that are known for their ability to attack rubber molecules. Samples of degraded surfaces were washed out with pepton physiological saline solution, applied on petri dishes and incubated at 35 °C for 7 days. Thereafter the colony forming units were counted. From each sample the 2 numeric dominant colonies were selected for identification by 165 rRNA gene sequence analysis. The identified micro organisms were strains of bacteria that belong to the order of bacteria known to be able to attack natural rubber.

5. Discussion

The rings show degradation and loss of material at the areas that have been in contact with the water side, as described in section 4.1. This type of degradation and loss of material on rubber sealing rings in water distribution pipelines was detected already before 1960: a large investigation was carried out on this phenomenon, the report was published in 1961 [3]. It was discovered that certain bacteria were able to degrade the rings. Only natural rubber showed loss of material by microbiological attack, synthetic rubbers (SBR, NBR) were not affected. Tests showed that this degradation was not caused by oxidation or abrasion. Considering the typical appearance of the damages on the investigated rings, results of the FT-IR and TGA analyses and the analyses of microbiological activity it is clear that the material losses are caused by microbiological attack. Given the large variation of degradation stages it is not possible to predict the rate of material loss. Most of the rings have still a mild stage of degradation that does not lead to problems. The sealing area between the ring and the concrete is still complete. If the loss of material comes to an extend where about half or more of the ring material has disappeared, the risk for leakage dramatically increases. Because of the material loss and the changing shape factor (the ratio between the loaded and non-loaded surface of the remaining part of the ring) the sealing stress decreases, as well as the friction force on the smaller contact surface between ring and concrete. At that stage of severe degradation the ring could be displaced at rapid changes of the water pressure, leading to a loss of the sealing. With obtained hardness values and areas of the cross sections of the rings and maximum clearances in the joints, the initial sealing pressure (excluding additional water pressure) can be calculated, using the model of Lindley [4]. This initial sealing pressure is about 6,9 bar for one type of pipeline and 7,7 bar for the other type of pipeline. Assuming an estimated stress relaxation of 5 to 7 % per decade, the remaining sealing pressure is about 3,5 to 5,0 bar and 3,9 to 5,5 bar respectively after 45 years. The remaining sealing pressure is less than 2,5 bar at a ring with over 50 % material loss by degradation.
Angle rotations could furthermore lead to displacements of socket and spigot ends of the pipe segments. At a settlement of the pipeline the joint of the pipe segments will be displaced from each other at the bottom side. At a rotation of 1° between the segments the displacement of the socket end over the spigot end is about 21 mm at the bottom side. In case this rotation happens after the rubber ring in this joint has reached a state of severe loss of material (about 50 % or more), the ring could turn and this could lead to leakage. The risk of twisting of the ring is higher at a sudden settlement than at a slow one.

 

Schema von der Ringprüfung.
Figure 17. Left the complete ring, compressed between socket and spigot. The sealing stress by compression is Pi1. The water pressure Pw gives an additional pressure on the rubber, resulting in a higher sealing pressure Pi1 + Pw between ring and socket/spigot. Right the degraded ring, compressed between socket and spigot. The sealing pressure by compression is Pi2. Less rubber, therefore less sealing pressure at equal compression as on a complete ring. The ratio between sealing pressure by compression of the ring and the water pressure can change to an extent that the friction force between ring and socket/spigot can be overruled, especially at (rapid) pressure changes. This can cause displacement and twisting of the ring. (Bild: alle Kiwa Nederland B.V.)

6. Conclusions

Most of the rings show degradation at the water contact area, in a straight line along the circumference. Degradation is in different stages, from surface hardening in the initial stage to loss of material in proceeding stages. The degradation is caused by microbiological attack. The sealing function of the rings is still sufficient except at rings with severe loss of material (50 % or more of the ring has dissapeared). In that case there is an increased risk for leakages. The physical-mechanical properties of the rubber in the bulk of the rings are still good.

 

Quelle: Kiwa Nederland B.V.

Acknowlegdements
Kiwa Lab-C test laboratory for plastic and rubber, Apeldoorn (NL)
Kiwa Polymer Institut, Flörsheim-Wicker (D)
TNO Earth, Life and Social Sciences, Zeist (NL)

References
[1] KIWA investigation reports for clients (not public).
[2] KIWA, Requirement for approval No. 61 – Natural rubber rings for joints in transport pipes for drinking water, sewage water and gas – 1971.
[3] KIWA, Report degradation of rubber rings for water transport pipes - 1961.
[4] Lindley, P.B., 1967, „Compression Characteristics of Laterally Unrestrained Rubber O-ring,“ J. IRI, Vol. 1, pp. 202.

Sie möchten gerne weiterlesen?

Registrieren Sie sich jetzt kostenlos:

Mit der Registrierung akzeptiere ich die Nutzungsbedingungen der Portale im Industrie-Medien-Netzwerks. Die Datenschutzerklärung habe ich zur Kenntnis genommen.

Sie sind bereits registriert?