TESTS VALIDATE FIBER GLASS CEMENT TO PROTECT SUBSEA FBE COATING

Sept. 17, 1990
C. Traulsen, N. J. R. Nielsen, T. S. Nielsen Maersk Oil & Gas AS Copenhagen Rock-shield coating on subsea pipelines is a viable alternative to concrete coating when protection against rock dumping is required. Tests and installation experience of Maersk Oil & Gas AS, Copenhagen, indicated this use, provided that special attention is paid to the priming process, cleanliness of the coating applicator, handling of the line pipe, and selection of a suitable plow for trenching.
C. Traulsen, N. J. R. Nielsen, T. S. Nielsen
Maersk Oil & Gas AS
Copenhagen

Rock-shield coating on subsea pipelines is a viable alternative to concrete coating when protection against rock dumping is required.

Tests and installation experience of Maersk Oil & Gas AS, Copenhagen, indicated this use, provided that special attention is paid to the priming process, cleanliness of the coating applicator, handling of the line pipe, and selection of a suitable plow for trenching.

The company gained this experience with a polymer-modified, glass-fiber-reinforced cement as coating for a 12.75-in., 19.05-mm (0.750 in.) W.T. subsea pipeline installed in the Danish sector of the North Sea.

DANISH GROUP

Maersk Oil & Gas is operator for the Danish Underground Consortium (DUC); other partners are Shell and Texaco. Fig. 1 shows the platforms and interfield pipelines operated by Maersk Oil & Gas. The water depth is about 40 m (131 ft).

The 11-km (6.8 mile) uninsulated pipeline is designed for the transport of unstabilized hydrocarbons from the Skjold satellite field to the central process facility at Gorm (Fig. 1). As a result of the adopted heavy-wall design, no weight coating was required to ensure the on-bottom stability of line for the time period between laying and trenching.

The pipeline was installed with conventional laybarge techniques, and the lowering was performed with a plow.

The steel line pipe was fabricated in Japan at Nippon Steel Corp.'s Yawata plant in January and February 1986 and shipped to Key & Kramer in Holland for coating. All line pipe was coated with fusion-bonded epoxy (FBE) as an anticorrosion coating; an additional 85 pipe sections were coated with the polymer-modified, glass-fiber-reinforced cement on top of the FBE.

All this coating application took place in the spring of 1986. Thereafter, the project was postponed for about 2 years.

In the spring of 1988, the remaining 850 lengths of line pipe were coated with the polymer-modified, glass-fiber-reinforced cement as a result of changes concerning the installation of the pipeline.

The pipeline was installed in July 1988.

CHOOSING THE SYSTEM

On the basis of technical and economical considerations, the FBE coating was selected as corrosion protection for the pipeline.

Due to the relatively high design inlet temperature (82 C.), the pipeline expansion of a purely FBE-coated pipe would be excessive at the Skjold platform. Consequently, the longitudinal frictional resistance of the pipeline was increased by application of a 4-mm thick layer of rockshield coating on top of the FBE to the first 1 km of the pipeline (85 pipe joints).

Prior to pipeline installation in 1988, the pipeline design was reassessed as a result of reported upheaval-buckling incidents of hot pipelines. Thus, the original installation procedure was modified to include subsequent rock dumping on the lowered pipeline.

As a consequence, the 4mm thick rock-shield coating was applied to all the remaining 10-km line pipe for mechanical protection of the FBE coating.

The polymer-modified, glass-fiber-reinforced cement is a rock-shield coating built of the following components: Forton 2000 rock-shield dry mix, Forton 2000 liquid blend, and water.

These components must be correctly batched and mixed and effectively deposited on a primed pipe.

The primer was Forton 2000 RS primer (spray grade) DP 93-0485 (copolymer emulsion) suitable for application over FBE. The primer was developed both to increase bonding characteristics and to provide corrosion protection, having low oxygen and water permeability.

The appearance of the primer is a white, milky liquid with about 60% solid content. The pH is 4.5-6.0.

The Forton 2000 rockshield dry mix composition is glass fiber (1.3-1.6 wt %), cement plus additives (32.0-35.0 wt %), and sand (63.4-66.7 wt %).

The Forton 2000 liquid blend is based on a styrene acrylic emulsion with additives.

The cleanliness of the water should be in accordance with BS 3148 or similar.

COATING APPLICATION

The fact that the FBE-coated line pipe was stored outdoors for a period of 2 years and that some developments with rock-shield coating were made between 1986 and 1988 would indicate that slight differences between the line pipe coated in 1986 and in 1988 can be expected.

Because the priming process for the line pipe requires a preheat temperature of a 60 C. minimum, all line pipe was heated in a hot-water bath (tap water) at a temperature of at least 80 C.

The line pipe was immersed in the hot-water bath until reaching a temperature of approximately 70 C.

For the line pipe coated in 1986, no cleaning of the epoxy surface was carried out prior to coating. Due to the long storage period for the line pipe coated in 1986, all surfaces were wiped off with a clean, dry cloth to remove all loose soil and dirt.

Thereafter, the lengths of pipe were given a high-pressure, soap-and-water cleaning and rinsed in a water bath. The original idea was to have the pipe rinsed in a cold-water bath (10-20 C.). But because the pipe sections immediately after this bath were going to be heated to approximately 70 C., it was decided to preheat the pipe by rinsing in a hot-water bath (uncontrolled temperature) in order to speed up production.

The pipe sections were thereafter immersed in the 80 C. water bath.

After the hot-water bath, the pipe sections were primed while rotating. In 1986, the pipe was primed manually; priming for the pipe in 1988 was automatically applied by means of a spray nozzle at the end of a moving boom.

The primer should be sprayed in a such way that a thin continuous layer is obtained. Fig. 2 shows the automatic primer station.

After the primer station, the temperature of the pipe was measured, and if the temperature was cooler than 60 C., the lengths of pipe were lifted back into the hot bath. The primer was still intact after the reheating of the pipe, and thus a repriming was unnecessary.

For coating, the pipe sections coming from the primer station were loaded onto a bogie on which they could rotate. The lengths of pipe were carried along a coating applicator where the coating was sprayed onto the rotating pipe. The speed of rotation and translation was adjusted in such a way that a uniform coating thickness was ensured.

Any lumpy spots or other irregularities which appeared on the surface of the coating were removed by the wiping of a wet brush over the surface. The obtained coating thickness of the dry coating was about 4 mm.

Immediately after the application of the coating, the wet film thickness was measured by means of a thickness gauge.

After application of the coating, each length of pipe was moved to an indoor curing storage area, where it had to be stored for a minimum of 24 hr at ambient temperature. Fig. 3 shows pipe during indoor curing with the ends supported by wooden blocks.

The pipe sections were then stored outdoors for a minimum of 48 hr covered with plastic sheets (Fig. 4) and then stored at the main storage yard without any coverage. The pipe was stored one layer high for the whole period. The minimum storage period for a length of pipe was about 1 month in total.

The ranges of temperatures and relative humidities during the curing period were -2.1 C to +26 C., with an average of +10 C., and 25-100% relative humidity with an average of 75%.

TESTING

To verify that the requirements for adhesion strength and impact resistance of the coating were met, the following types of tests were carried out: elcometer adhesion testing, V-cut testing, and impact testing.

ADHESION TESTING

The elcometer adhesion test of the final coated pipe was made according to the year of coating.

  • 1986. Laboratory tests were carried out with an Instron tensile-test machine on lengths of pipe selected randomly during coating from the production coating.

    Test 1, 1986 pipe. The adhesion tests were made after 7 days of curing at a temperature of 20 C. and a relative humidity of 95% followed by a normalization period of 24 hr at a temperature of 20 C. and a relative humidity of 65%.

    Test 2, 1986 pipe. The adhesion tests were made after 14 days of curing at ambient conditions.

    Test 3, 1986 pipe. The adhesion tests were made after 4 days of curing at a temperature of 20 C. and a relative humidity of 99% followed by 2 days at 20 C. and 65% relative humidity.

  • 1988. Laboratory tests were again carried out with an Instron tensile-test machine, but on two pipe sections coated in the laboratory. Additionally, manual adhesion testing of production-coated pipe was carried out (both for pipe coated in 1986 and 1988) at the coating yard.

Generally, the procedure for the laboratory adhesion testing was as follows:

A 38-cm long pipe section was cut from a pipe selected by inspection. This sample was cut through its axis and one half selected.

Three adjacent areas of coating with dimensions of 38 cm x 12 cm (15.2 cm x 5 cm in 1986 pipe coating) were marked on the pipe section selected. The 38 cm was along the pipe axis.

For each area, three equidistant cores with dimensions of 5 cm x 5 cm were cut through the rock-shield coating layer to the FBE coating. A 5-cm square contra-piece, or "dolly," whose inside radius of curvature was similar to the outside radius of curvature of the rock-shield coated pipe, was glued to the test sites with an araldite epoxy resin and cured for 24 hr at 20 C.

During the adhesion test, an hydraulic tensile force was uniformly built up until break away. In Figs. 5 and 6, the pipe samples and equipment can be seen. Fig. 7 (Test 1, 1988 pipe) shows the pipe samples and dollies after testing.

As can be seen, the testing did not fail 100% by adhesion between the epoxy coating and the Forton coating.

Two laboratory test programs were carried out where the cleaning of the FBE coating varied.

Test 1, 1988 pipe. The pipe surface (coated with FBE) was cleaned with soft paper before priming and coating application. The test samples were then cured for 7 days at ambient temperature (15 C.) and a relative humidity of greater than 55%.

Test 2, 1988 pipe. The pipe surface (coated with FBE) was cleaned with hot water before priming and coating application. The curing condition was the same as for Test 1, 1988 pipe.

The manual adhesion tests made in 1988 pipe at the coating yard were performed in the following way:

A core (2 cm diameter) was drilled through the rockshield coating, then the toe of the core was filed to a flat shape. A dolly of 2 cm in diameter was glued on with araldite epoxy glue and allowed to cure for 24 hr. The dolly was then pulled off with a manual elcometer.

This test was performed both on the pipe coated in 1988 and on pipe coated in 1986 to indicate whether the adhesion was still acceptable after 2 years of storage:

Test 3, 1988 pipe. A manual adhesion test was made after a curing period of 7 days at ambient temperatures and humidity.

Test 4, 1988 pipe. A manual adhesion test was made.

Table 1 shows test results observed.

V-CUT TESTING

The V-cut test is also an adhesion test but much easier to perform.

The procedure is to make two incisions of approximately 12 mm through the deposited rock-shield coating to the FBE coating with a razor knife. The two incisions should form a "V" shape with an angle of approximately 30.

An attempt is then made to force the rock-shield coating by starting at the apex of the "V." If the rock-shield coating is acceptable, it should show good bonding characteristics.

At the coating yard, every tenth pipe was tested in this way, and all tests showed very good bonding characteristics.

This test was made after an initial curing period of 7 days.

IMPACT TEST

The impact test was carried out to verify that the impact resistance of the cured rock-shield coating was sufficiently high so that the coating could withstand normally encountered impact forces during handling and installation, in particular rock dumping.

The equipment was a Gardner heavy-duty, falling-ball impact tester in accordance with ASTM G14-77.

Every fifth pipe was tested with this equipment after the initial curing period of 48 hr. The specification did specify a minimum of 7 days, but it showed that all of the tested pipe could withstand the impact after 48 hr.

The ball struck each pipe in a radial direction at three equispaced locations along the pipe length. The impact energy was a minimum of 18 joule. None of the areas tested showed any damage to the rock-shield coating.

HANDLING, INSTALLATION

All the pipe coated in 1986 was transported by truck to a storage yard about 20 km from the coating yard for storage for 2 years before being transported back to the coating yard by trucks again.

The pipe lengths were inspected for damage after they arrived at the storage yard and again after arrival at the coating yard in 1988. A total of about 40 damage areas in 85 rock-shield coated pipe sections were present.

Some of the damaged areas were very small (50 sq cm), but elsewhere damage areas were quite large. For some of the pipe lengths with large damage areas, the coating along about one third of the length was very easily removed by hammering at the coating, because the adhesion was very poor.

The damaged areas were all repaired so that all 935 pipe sections loaded out in 1988 were intact.

During the handling of the pipe, some of the coating was again damaged. In this sense, it should be mentioned that there was a lot of handling of the pipe on the lay barge: lifting of pipe from supply vessel to the lay barge, lifting of pipe from storage to double-joint station, lifting of pipe from double-joint station to double-joint storage, and lifting of pipe from double-joint storage to main line.

It was planned to trench and rock dump the pipeline at a later stage. This trench was performed by plow along the pipeline at the bottom of the sea. Video tapes taken both before and after the plowing showed some areas of damage to the rock-shield coating. The largest area was about 1 sq m.

DIFFERING RESULTS

As can be seen in the section describing the testing of the coating material, different types of tests have been performed both on the pipe coated in 1986 and on the pipe coated in 1988, and different results were found on the adhesion tests of the coating.

Reasons for the different values (relatively low values found in 1986) could be numerous, but we think that it is related to how the pipe sections were primed.

The primer in 1986 was sprayed manually, which means that the primer was applied in a more uncontrolled manner as compared to the automatic priming in 1988. The test samples from 1986 showed that the primer was applied in a relatively thick layer.

Further, the priming took place only 1-2 sec before the actual rock-shield coating was applied. We think that the primer should be dry before application of the rockshield coating, which means that the primer must be applied at least 10 sec before the coating process.

During the coating process, it was very important that the water content in the wet mix be in the lower end of the acceptable range to avoid cracks in the dry coating.

The coating applicator was clogged up very easily, and it was therefore very important that the applicator be cleaned before every work break (lunch, for example).

If this was not done, the pipe-coating application was made so that the surfaces were very uneven along a helical path, and a lot of high peaks were present on the surface.

The handling of line pipe was one of the main reasons for repair of the coating. Typical damage occurred at the end of a pipe and at an anode doubler plate.

It should be noted that each pipe section was fitted with four circumferential ropes with a distance of 3 m between every rope.

After the trenching of the pipeline, some large damaged areas were found. On video tape taken underwater, it could be seen that the plow had scratched off the coating.

Thus, special attention must be paid to the selection and design of a plow for a pipeline with this type of coating.

ACKNOWLEDGMENT

The authors wish to thank the management of Maersk Oil & Gas AS and the DUC partners, Texaco and Shell, for permission to publish this article.

Copyright 1990 Oil & Gas Journal. All Rights Reserved.