FREEZE PLUG PROVES SAFE, ECONOMICAL IN RISER REPAIR

Michael J. Nelson
Exxon Co. U.S.A. New Orleans

In October 1992, Exxon Pipeline Co., Houston, performed in the Gulf of Mexico what the company believes to have been -the first underwater freeze-plug procedure.

To form a plug, water in a small section of the pipe is frozen with liquid nitrogen.

In partially replacing a 10-in. riser at South Marsh Island Block 6A (Fig. 1), Exxon Pipeline worked closely with a freeze-plug service company to minimize environmental and personnel exposure and to avoid the chance of an oil spill.

The freeze plug reduced the time the pipe was open-ended during the repair, and hydrotesting the freeze plug area and repair section ensured integrity.

Exxon Pipeline had successfully used freeze plugs in many applications onshore, particularly in pipeline repairs, and in hydrotesting applications to facilitate leak location. However, the company believes the technology had not been used on a subsea pipeline.

In 1991, Shell U.K. performed tests on similar under-water freeze plugging for a pipeline repair in the North Sea, but the procedure was not used.

Shell opted instead to use displacement batches with high seal pigs.1

OPTIONS

Traditional industry practice for similar repairs has been to leave the pipeline open-ended throughout the pipe replacement procedure. The risks of an open-ended pipeline include unintended injection by a third party and possible expansion of gas that could be trapped in the pipeline.

Before Exxon Pipeline incorporated the freeze plug into the repair procedure, it considered several options for plugging the pipeline prior to repairs.

Stopple fittings were evaluated, but the potential exposure to third-party damage and an additional possible spill source due to the additional fitting left on the riser ruled out this option.

Another consideration was to use a series of batching pigs and displace part of the pipeline with water, leaving the riser open-ended during the repair. Pipeline volume, however, made displacement of the entire line less practical due to water-handling expenses.

Another option Exxon Pipeline considered was to cut the pipe after water displacement and install a vent packer into the pipe immediately after cutting while the pipe was open-ended. The vent packer would direct any release of product safely to a vessel on the platform and could be used to hydrotest the replacement fitting.

A fourth option considered was to use a freeze plug after the line was partially displaced with water to reduce spill of product when the riser was replaced.

Exxon elected to use a combination of three procedures to minimize risk to the environment and personnel: the riser would be displaced with water, the freeze plug installed, and a vent plug would be used while the pipe was open ended.

SHORE TESTING

Before the freeze plug was used underwater, the procedure was tested onshore in a diving tank with a test spool. Because sea water would be used to displace the pipeline, saltwater was used for testing the freeze plug.

The spool was lowered into the tank, and a diver installed a 2-ft long, 10-in. diameter freeze jacket around the pipe. Separate lines were then installed by the diver to the jacket for supplying liquid nitrogen, venting the nitrogen, and monitoring the internal temperature of the jacket.

The water in the jacket was removed, and the hoses were purged with air before the flow of nitrogen. The internal temperature of the jacket was monitored until it stabilized at -320 F. after approximately 2 hr.

After the temperature had stabilized, the pressure was bled on one side of the plug to test that the freeze plug would hold. The plug held a differential pressure of 500 psi which was planned for the test.

Nitrogen flow continued through the jacket during the test. The heat transfer to the water kept ice from forming on the jacket.

The exterior surface of the jacket in contact with the constant movement of the surrounding water created a considerable heat sink which eliminated a buildup of ice on the equipment.

This heat sink had little effect on the capability of the equipment to establish a positive plug on this job. Also, the construction of the jacket provided an insulation barrier between the exterior surface of the chamber and the pipe wall.

In previous Exxon Pipeline freeze-plug installations onshore, the pipe typically was hydrotested after the plug was removed to verify the integrity of the pipe in the freeze plug area.

In this offshore application, a vent packer would be used for hydrotesting both the freeze plug area and the new forged flange connection. After installation of the flange, the vent packer would be set and vented.

The frozen plug would be melted by discontinuation of the nitrogen flow to the jacket. Any pressure behind the freeze plug would relieve through the vent packer which would then be reset to a point below the location of the freeze plug and the new flange.

A blind flange with a connection to accommodate the vent line to the packer would be used to hydrotest the section between the packer and the new flange, including the portion of pipe which was frozen. The pipe upstream of the packer would be vented throughout the hydrotest procedure.

Following the successful hydrotest, the blind flange and packer would be removed and the remaining part of the new riser installed. This procedure would be the only time the pipe would be open ended.

Exxon Pipeline deemed minimal the risk of oil spill while the pipe would be open-ended and the flanged joint was being installed. Contingency plans would include boom and skimmers in the area for containment and clean-up in the unlikely event of a spill.

SURFACE CLEANING

Pre-work involved removing the concrete coating and non-destructively testing the pipe where the freeze would take place.

The freezing process can impose significant longitudinal and circumferential stresses to the pipe, possibly causing an existing defect in the pipe to fracture. To avoid failure, the pipe was cleaned and examined to 1 ft outside the planned freeze area.

Externally, the pipe was examined for corrosion, mechanical damage, or high stress concentration areas that could induce a brittle fracture. Internally, an ultrasonic normal beam was used to verify thickness and absence of laminations and internal corrosion.

After the pipe integrity was verified, the coating was removed in the area where the pipe would be cold cut and the new flange installed (approximately 5 ft above the freeze plug).

The system was shut-in, locked out, and allowed to stabilize for approximately 2 hr. A 10-in. double dish (bi-directional) batching pig was then launched (Fig. 2) with water pumped from the firewater system.

A magnet was imbedded in the batching pig to trip an external pig-signal to verify the pig was launched and passed beyond a block valve. The valve was then closed and a second pig with magnet was launched.

Then a third batching pig was launched, followed by 200 bbl of water (approximately 2,000 ft of displacement).

While the water was being pumped, a freeze jacket which was made longer to ensure a good seal was installed (Fig. 3). The nitrogen circulating lines and temperature sensor lines were also modified by being placed in a single umbilical to aid the diver in installation (Fig. 4).

When the displacement was compete, 200 psi of pressure was kept in the line during the freezing operation to ensure that the riser contained a full column of water.

After about 4 hr, which allowed the temperature to stabilize, the freeze plug's integrity was verified by bleeding the pressure on the upstream side of the plug and observing that the pressure on the opposite side held constant at 200 psi.

Next, all pressure was removed, and the line was tapped above the plug to verify the line contained clean water. The line was cold cut above the freeze plug, while nitrogen continued to circulate in the jacket.

The new subsea flange was installed followed by setting the vent packer above the freeze plug (Fig. 5). After the vent line from the packer was connected to the platform relief system, the freeze plug was melted by stopping the flow of nitrogen and removing the jacket.

This allowed the surrounding water quickly to melt the frozen plug. After about 1 hr, the packer was de-pressured and lowered 2 ft below where the freeze plug had been located, keeping the relief vent line connected at all times so the packer could be re-set quickly if needed.

The packer was re-activated and the test flange bolted to the new flange. The pipe between the packer and the test flange was then hydrotested to verify the integrity of the new flange and the area where the freeze plug had been located (Fig. 6).

When the test was complete, the packer was deactivated and removed so that the new pre-tested riser spool could be installed and commissioned. During commissioning, the pressure was brought up slowly, and the connections were observed to verify no leakage (Fig. 7).

INCIDENT FREE

The entire procedure took about 40 hr from displacement through installation of the new riser. The work was done without accidents or spills.

Costs for introducing the freeze plug and vent packer to more standard procedures were about 10% of the entire job.

Exxon Pipeline considers use of the freeze plug in this application to have been successful and will consider it for future offshore work where a pipeline requires isolation for repairs.

The company, however, does not endorse or recommend this method for use by others, believing that each company needs to determine whether the method is safe or appropriate under other circumstances.