Increasing the life of NR pipe joint rings.

The microbiological degradation (MD) of natural rubber (NR) pipe joint tings was first reported by Leeflang in 1963. It has since been found to be a worldwide problem. Since it was first reported, MD has been observed in watermains, sewers and stormwater drains. Research by rubber manufacturers in New Zealand and elsewhere led to the introduction, from the mid-1970s, of rings believed to be reasonably immune to MD. These are referred to here as improved rings.

However, all drainage systems with rubber joint rings laid prior to about 1975 will have been prone to MD. This article describes a recent research project on the topic. The project was intended to (a) assess the condition (severity of degradation) of rubber joint rings from many parts of New Zealand; (b) establish the likely life of the improved rings; and (c) examine the in situ chemical modification of exposed surfaces of the pre-1970s rings to increase their longevity and hence delay the need for replacement. Shortly after the project commenced in 1996, its history and aims were published (Hills, 1996).

Historical information

Since the early 1960s, MD of NR pipe joint rings, often complying with BS 2494, has been reported in watermains and drainage systems (both sewers and stormwater drains). MD of NR pipe joint rings in Holland was reported by Leeflang in 1963, although it had been observed during the 1950s and as early as 1949. Leeflang reported that the degradation was due to organisms of the genus Streptomyces. The form of attack by Streptomyces (and other organisms) on NR pipe joint rings consists of the organism consuming the rubber hydrocarbon but retaining the structure by a form called the "mycelium". This formation of a mycelium (vegetative system rather than spores) is one reason for classifying Streptomyces, commonly reckoned among the bacteria, as a transition between bacteria and fungi. Figure 1 shows the overall process of MD of NR pipe joint rings. This is a diagrammatic representation of microbiological degradation of NR. Area 1 shows the new rubber surface. In area 2, some of the rubber hydrocarbon has been consumed and replaced by the mycelium. In area 3, attack has proceeded to a greater depth. Finally, in area 4, exposed fillers have been attrited by the action of flowing water.

[Figure 1 ILLUSTRATION OMITTED]

Reaction to Leeflang's work varied from panic in some quarters, to complete indifference. Many interested parties were confident that there was no problem in their country since most rings were by that time made to the requirements of a standard such as BS 2494. This view appeared to be supported when eroded rings were located in many countries, initially from older drains (the most likely to be excavated for alterations) where early rings were regarded as being of poor quality. Most contained crumb (ground-up waste rubber), the use of which was outlawed by BS 2494 in 1955. Many theories were advanced to explain the observed MD. One suggested that attack took place on an oxidative breakdown product, rather than on the rubber hydrocarbon itself. Before long, however, eroded rings manufactured to the requirements of BS 2494 were found in New Zealand and elsewhere.

In New Zealand, it appeared that at the time, only one drainage authority actively checked rings that were recovered during remedial work. They have continued to do so since 1964, but more authorities have become aware of the problem. One of many rings examined during this latest research is shown in figure 2. It had been manufactured to the requirements of BS 2494, but had failed due to MD long before its removal.

[Figure 2 ILLUSTRATION OMITTED]

Among the local pipe joint ring manufacturers, only Skellerup's Empire Rubber Mills commenced research on the subject, having recognized the need to develop rubber formulations immune to attack. Due to commercial sensitivity, it was possible for limited material only to be published by Hills in 1967. Subsequent funding permitted a PhD student to research the microbiological aspects, working closely with the writer. Several papers on this combined work were published, for example, Cundell and Mulcock (1973); Cundell, Hills and Mulcock (1973).

Leeflang (1963) had set up a test bath, and had confirmed that all NR formulations (with one exception) were prone to MD, but that most synthetic rubbers (or blends containing less than 50% NR) were not. However, it was generally considered that the mechanical properties of NR were superior to those of most synthetic rubbers in terms of pipe joint rings, and that the use of NR should be retained if at all possible.

The presence of Streptomyces associated with degradation had already been confirmed here and elsewhere, although some British work recorded the involvement of Nocardia, as reported by Brister (1981). Dickenson (1965) carried out work in England on rubber formulations, and there was a private exchange between him and the writer. It was found that through the use of a particular combination of compounding ingredients (antioxidants or antiozonants of a certain type, and chlorine-containing materials) rubber formulations that were regarded as immune to attack could be produced by the mid-1970s. Two observations instrumental in this development were made in New Zealand.

The first was that, in eroded rings containing crumb (usually obtained from tire scrap), the crumb particles projected above the eroded surface, as illustrated in figure 3. They had therefore been less prone to MD than the supposedly-good rubber around them, probably because tire formulations contained carbon black and a higher level of antioxidants and antiozonants (Hills, 1983; Hills, 1992). The second observation was that, in his test batch set up in 1965, the writer found that samples of rubber which were prone to MD, but which had been immersed in a saturated solution of chlorine, became immune to surface attack, due to the chlorination of the rubber. Penetration of the film allowed MD to proceed in the rubber that had not been chlorinated.

[Figure 3 ILLUSTRATION OMITTED]

Incidence of microbiological degradation

In order to quantify the situation regarding early (i.e., 1940s to mid-1970s) pipe rings and improved pipe tings (i.e., produced from the mid-1970s on), it was necessary to examine rings removed from service from all parts of New Zealand. This was intended to yield two distinct pieces of information. The first was a measure of the awareness and severity of MD, and the second was an assessment of the improved rings.

A request was made to all 73 local authorities in New Zealand by way of a questionnaire (Hills, 1997). From a total of 55 replies, 25 confirmed that MD had been observed, and 30 reported that it had not been observed. From those 30, 16 respondents had looked for MD, but 14 had not. Twenty-two of the 25 positive responses also reported failures due to MD. As those local authorities included many of the major urban areas, they represented a far greater percentage of the population than would be indicated by 22 out of 55 respondents or 22 out of a total of 73 questionnaires sent.

Some of the 22 positive responses reported dozens or more than 100 failures. Negative responses included one from an area with a reticulation system mainly less than 20 years old. Another was from a region where it is understood that pipe tings in bell-and-spigot joints were not in common use at the crucial time; external pipe bands had been popular. These pipe bands had been manufactured by extrusion, using a formulation based on chloroprene robber which, in common with most synthetic rubbers (such as EPDM), has been found to resist MD.

Samples were received from many areas, and all were examined and in many cases photographed for the record. In common with rings that the writer has examined in the Christchurch area since 1964, there were more rings showing MD on the inside of the joint than on the outside. Many rings from areas with a high water table were heavily attacked from the outside. The existence of the problem had usually been discovered due to infiltration of water. Because the bulk of MD was internal, the effect of different soil and water types did not appear to be significant.

One interesting aspect of the present work related to the types of organisms present on rings from different centers. In almost all cases, Nocardia was present, but Streptomyces was rarely found. This was a major change compared with microbial identifications carded out in the late 1960s, where Streptomyces species were found in all cases. It is possible that modern detergents and washing products may have some influence in this regard, but this is by no means certain. An impression was also gained that the severity of MD may now be lower than in the 1960s and 1970s, which may suggest that Nocardia is less destructive than Streptomyces. Whether this impression is real or not, MD is still a major problem in some localities. Some rings had failed in less than 15 years, while others were still in good condition after more than 50 years.

In summary, MD of NR pipe joint rings has been observed throughout New Zealand, and there have been many joint failures as a result of this degradation. Awareness of its likely existence is far higher than in the 1970s, when few local authorities seemed willing to look for it.

Life of improved (post-1970s) rings

Virtually all the samples received during the project had come from sewer and stormwater systems laid between 1960 and 1970. In only one case were degraded rings received from an installation dating from the late 1970s. However, when these V-section rings were tested, they were found to be manufactured from the old formulation, and so must have been held in stock for several years before being installed.

To date, no post-1970s rings have been sighted in a degraded condition. In the 1960s, rings less than five years old were found with MD corresponding to levels 2 and 3 in (surface embrittlement of the ring when in a dry condition). Improved rings that have been in service for about 20 years have not shown MD to any extent.

The conclusion that can be drawn from the rings examined, and from field reports, is that improved rings (and other seals) will perform their function for a period well in excess of 50 years.

Chemical modification of NR surfaces

Natural rubber has been modified in many ways to yield grades with enhanced properties. For example, chlorinated robber is used in paints, and crosslinked rubbers are used as processing aids for other rubbers. Some epoxidized rubber grades offer oil resistance properties not normally associated with NR. The bulk of the modifications are carded out at the latex stage, or in a solvent solution of the solid rubber coagulated from the latex. In other words, the modifications are carded out when the rubber hydrocarbon is in the unvulcanized state.

Much literature exists on the modifications that are, or can be, made to NR. This project is concerned with possible in situ modifications that may be carded out on vulcanized NR in the form of a pipe joint ring. The literature was singularly sparse on the subject of chemical modifications of vulcanized rubber, other than a few references to chlorination or bromination to improve adhesion to other materials. As noted earlier, the writer had used surface chlorination to' produce MD resistance over 30 years ago. As reported in his 1967 publication, the MD-resistant layer was thin, and if perforated in some way, did not prevent attack on the rubber below. In addition, the silky nature of a chlorinated robber surface was such that a treated pipe ring would no longer grip and roll in a bell-and-spigot joint.

Bearing in mind the nature of the chemicals involved, a short list of possible chemical modifications was drawn up. This was (a) chlorination, (b) bromination, (c) hydrochlorination, (d) cyclization, (e) oxidation, (f) chlorosulfonation and (g) crosslinking. Further study highlighted some interesting, and rather frightening, aspects to rubber industry processes of the past. It had been hoped to use chlorosulfonic acid for item (f), and sulfur dichloride or monochloride for (g). Regarding the latter, Hofmann (1989) noted that sulfur dichloride vulcanization "was very popular during the twenties (and) is still employed to a limited degree." All three chemicals were mentioned on suppliers' lists, none were available from local stock, and all three were described by BDH (1991) as "corrosive substances that can destroy living tissue." It was decided to delete these items from the list of trial modifications.

The production of cyclized rubber involves the use of sulfuric acid, and to be effective needs the acid to be quite concentrated. The attempted bromination of samples involved a trial with methyl bromide in kerosene. Both these trials were considered to be of academic interest, but not practical in terms of the in situ modification of rubber surfaces.

A rubber formulation known to be susceptible to MD, and the same as the base formulation in the writer's work in the 1960s, was produced and molded as very thin sheets. Strips were cut, and subjected to a range of chemical treatments under known conditions of concentration, temperature and time.

MD-resistance testing

The treated strips were suspended in a microbial test bath similar to the one used by Leeflang (1963) and Hills (1967). During the course of the work, the test bath was upgraded from operating at ambient temperature to an average operating temperature of 23 [+ or -] 2 [degrees] C, and by the addition of an air pump supplying a visible diffuse air stream throughout its length. It now meets the general requirements of Appendix B, Method of test for microbiological deterioration of BS 2494:1990. The water source is untreated water from the Christchurch City supply, rather than the raw river water specified in BS 2494.

Each set of test samples included a control sample that had not been subjected to any form of surface treatment. For example, one set comprised a control sample and others which had received 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15 and 20 minutes immersion in a saturated chlorine solution (abbreviated to SCS below). Other sets of samples were fewer in number. Samples were examined at monthly intervals, and MD was apparent to the naked eye as dulled patches on the surface of the test strip and as cracking of the edge or areas on a stretched (dried) strip observed under a low power (10 - 20 x magnification) stereo microscope. The tests continue at the time of preparation of this article.

Results to date confirm that cyclization (using sulfuric acid) produces surface embrittlement that cracks when under tension, and hence does not permit the presence of MD to be observed. Hydrochlorination caused the samples to adsorb considerable quantities of water, such that they swelled and wrinkled. The use of acidic or alkaline detergents with SCS did not appear to enhance the effectiveness of the solution. In fact, immersion periods as short as one minute in SCS produced a visible surface difference, and the samples were considerably more resistant to MD than untreated ones. Other treatments resulted in varying degrees of resistance to MD.

The use of an immersion period greater than two minutes in SCS produced a surface which could be regarded as MD-resistant, or at least far more MD-resistant than the formulations in use prior to the mid-1970s. Diluted SCS would require a longer immersion period for the same effect.

In situ treatment of pipe rings

The final phase of the work, nearing completion, is to finalize the details for the in situ treatment of exposed surfaces of pipe rings that were placed into service before about 1975. This will include the best method of cleaning (to permit maximum access to the internal exposed surface of each pipe joint ring) and the containment of the SCS for the required time in small-diameter pipes. For large-diameter pipes, a spraying technique is being developed, and for both methods, the process for the subsequent neutralization and disposal of the used solution is being documented. The time-concentration factor will be taken into account as chlorine is utilized, such that a minimum factor of 1.6 will be maintained. This is based on a time period of two minutes and an SCS concentration of 0.8 g/liter of chlorine at about 15 [degrees] C. Test results suggest that a concentration as low as 0.2 g/liter can be utilized for the appropriate time period (eight minutes). This in situ treatment can be applied to mains and adjacent parts of laterals for small-diameter systems, or to larger-diameter mains. Methods for the preparation and storage of the SCS will be included in the commercial process being developed, along with details of measuring chlorine concentration and neutralizing the solution for disposal.

Conclusions

MD of NR pipe joint rings is widespread throughout New Zealand. Awareness of its existence is far higher than in the 1970s, when few local authorities seemed willing to look for signs of its presence. Many joint failures have occurred as a result of MD, and the severity of attack on many samples sighted indicated that they would have failed within a few years had the affected joints not been replaced. As no examples of improved rings being subjected to MIC were located, it is reasonable to assume that these rings (and seals of other materials such as EPDM) will perform their function for periods far in excess of 50 years.

The in situ treatment of rings placed into service before 1975 appears to be possible through the use of a saturated chlorine solution, which can be neutralized for disposal after use. This process will markedly increase the remaining service life of the rings concerned, permitting a controlled program of replacements.

References

(1.) BDH Limited (1991) "BDH Chemicals 1991'-92."

(2.) Brister, D. (1981) "Developments in pipe joint rings," Progress of Rubber Technology, Applied Science Publishers, 105.

(3.) Cundell, A.M. and Mulcock, A.P. (1973) "Microbiological deterioration of natural rubber pipe joint rings," Material Organismen, 8, 165.

(4.) Cundell, A.M., Hills, D.A. and Mulcock, A.P. (1973) "The influence of antioxidants and sulfur level on the microbiological deterioration of vulcanized NR," Rubber J., 155, 22.

(5.) Dickenson, P.B. (1965) "Natural rubber and its use in underground pipe sealing rings," Rubber Developments, 18, 85.

(6.) Hills, D.A. (1967) "The degradation of natural rubber pipe joint rings," Rubber J., 77, 14.

(7.) Hills, D.A. (1983) "Aspects of degradation of rubbers," Proc. Degradation of Materials Symposium, Aust. Corrosion Assn., 26.

(8.) Hills, D.A. (1992) "Rubbers and plastics - how they may fail," monograph of Aust. Corrosion Assn., 12.

(9.) Hills, D.A. (1996) "Increasing the longevity of rubber rings in pipe joints," Water & Wastes in NZ, #93, 56.

(10.) Hills, D.A. (1997) "Microbiological degradation of rubber pipe joint rings in New Zealand," Water & Wastes in NZ, #97, 40.

(11.) Hofmann, W. (1989), Rubber Technology Handbook, Hanser, 452.

(12.) Leeflang, K. W.H. (1963), "Microbiological degradation of rubber," J. AWWA, 55, 1,523.

D.A. Hills, Materials & Quality Consultancy Ltd.