Gary B. Vogt
Project Consulting Services Inc.
Metairie, La.
Lee D. Walker
Southern Natural Gas Co.
Birmingham, Ala.
W.F. Lammert
OPI International
Houston
Gas began flowing this spring in the deepest underwater, large-diameter pipeline in the U.S. Gulf of Mexico.
Southern Natural Gas (SNG) Co., Birmingham, completed installation of the 60-mile, 20-in. line from Exxon Co. U.S.A.'s Alabaster platform in Mississippi Canyon Block 397 offshore Louisiana to a shore interconnect near Venice, La.
Water depth along the route of the pipeline varies from approximately 460 ft at the Alabaster platform, increasing to the record depth of 1,220 ft in the Mississippi Canyon area, and decreasing to negligible water depth at the landfall site southwest of Venice.
PROJECT TEAM
The SNG Mississippi Canyon Block 397 pipeline project exemplifies how a pipeline project can encounter an array of conditions which prompt special design considerations and installation techniques.
Important considerations for this project were related to pipe properties, anti-corrosion and weight coatings, span and buckle considerations, and installation equipment. A team effort was used to study, research, test, design, and install this pipeline.
The SNG project team consisted of personnel from Guillot-Vogt Associates Inc., New Orleans, for permitting, design, and project management; Hudson Engineering, Houston, for design and stability analysis; Materials & Welding Technology Inc., for review of welding and pipe specifications; and numerous personnel throughout the SNG organization for project management, materials management, inspection, and overall operations.
Preliminary pipeline design suggested various pipeline routes between the proposed Exxon Mississippi Canyon Block 397 platform and a shore facility near Venice. The proposed platform was to be installed on a large topographic rise caused by a diapiric intrusion known as the Alabaster Knoll.
Water depth at the platform location, near the top of the knoll, is approximately 460 ft. The slope and surface features of the bottom in the area made routing the pipeline down the knoll critical.
Within a radius of 10,000 ft in almost every direction from the knoll the water depth increases to a depth of 1,200 ft to 1,500 ft. Routing considerations studied the use of existing pipeline systems in the area compared with a direct route to a shore facility.
When it became clear that an existing pipeline transmission system to transport the natural gas could not be arranged, a direct route to shore was pursued.
Once it was determined that the pipeline would be installed directly to shore, various pipeline routes were analyzed. The offshore terrain between the proposed Exxon Alabaster platform and the proposed shore tie-in point ranged from treacherous sea floor conditions in the Mississippi Canyon area to an area congested with existing pipelines and platforms in the West Delta area.
To determine the preliminary route, an economic evaluation was performed on various routes which considered the installation and material cost of the pipe footage, burial footage, water depth, pipeline crossings, and subsea terrain.
Upon completion of the preliminary routing studies, a hazard survey was performed along the selected pipeline route. This survey included core samples of the bottom along the entire route. The survey contractor had to modify his coring equipment to be able to take the samples in the greater water depths.
Adjustments were made to the preliminary route based upon data obtained during the hazard and subsequent surveys. Because of severe bottom conditions in the Mississippi Canyon area, additional survey work was performed with a 3-D bathymetry system which provided a more detailed overview of the sea floor.
The survey data confirmed numerous escarpments and severe bottom conditions throughout the Mississippi Canyon area. A 9 slope was observed along the north face of the Alabaster Knoll.
The area with bottom topography of most concern was on the route's uphill approach in Mississippi Canyon Block 266. This area has numerous fault scarps, outcrops, and a shelf-edge separation scar that had to be avoided.
In addition to sea floor routing considerations, a pipeline corridor was chosen to avoid other pipelines proposed for the Exxon Alabaster platform.
A span analysis was performed on the proposed route in areas conducive to spans. Although the analysis showed the potential for spans in a few areas, soils analysis confirmed that the soil in these areas was very soft. The ROV as-built survey showed some spans in these areas, with the longest being 350 ft with an off-bottom height of approximately 6 ft.
Once the pipeline was flooded, the spans subsided and the length of the longest span decreased to 60 ft, well within design operating parameters.
PIPE DESIGN, COATINGS
Analysis for steel yield strength concluded that 60 ksi, API-5L specification was appropriate considering the wall-thickness requirement for the critical loading cases to resist internal pressure, buckle propagation, and installation loads.
Double submerged-arc welded (DSAW) pipe was used for the offshore portion of the pipeline for both technical and economic reasons. Considerable cost savings were realized over seamless pipe.
The optimum wall thicknesses along the length of the pipeline were determined with the Battelle Research Group interaction equation and a "Combined Loading Only" design criterion. This criterion is based upon the wall thickness being sufficient to resist collapse from the combination of hydrostatic pressure and of installation loads but being insufficient to prevent buckle propagation.
There were seven different wall thicknesses used on this 20-in. pipeline as outlined in Fig. 1. Buckle arrestors were required at all water depths greater than the buckle-propagation depth which was determined as 325 ft. Using this method of design allowed a significant cost savings over a buckle propagation-resistant wall-thickness design which would depend upon additional wall thickness in lieu of buckle arrestors.
Because of the limited application history of the design methods for buckle arrestors, tests were conducted to determine the incremental pressure to force the propagating buckle past the proposed buckle-arrestor design. The tests proved the design analysis to be conservative and showed that the actual crossover pressure was higher than expected.
At intervals of 400 ft, 146 buckle arrestors were installed in the pipeline. This interval was based upon an analysis of the present technology and equipment available to perform an effective repair in the deepwater areas along the proposed pipeline route.
To simplify installation, pipeline anodes were installed over the buckle arrestors as an integral unit, and the anodes were also utilized as a convenient means of identifying the buckle-arrestor joints.
Studies were performed on various types of anti-corrosion and weight coatings.
Fusion-bonded epoxy (FBE) was chosen as the anti-corrosion coating and was applied to a thickness of 17 mils. Compression-applied concrete was chosen as the weight coating at a density of 140 lb/cu ft.
Although other densities of concrete were considered, the 140 lb/cu ft concrete density was the most cost effective weight coating. All of the pipe coating operations were performed in New Iberia, La.
In order to meet the demanding pipe coating schedule, two weight coating plants were utilized. Only one plant was needed for the FBE application to stay ahead of the weight-coating application. The pipe-coating contractors had to handle more than 8,000 joints of 20-in. OD pipe consisting of nine different wall thicknesses of steel and eight different thicknesses of concrete.
In all, the offshore portion of the pipeline required 26 material barge loads to transport the pipe to the lay barge. This step represented transporting 42,000 tons of pipe to the job site.
The field-joint coating for the pipeline consisted of FBE for the anti-corrosion coating and polyurethane-foam filler material held in place by a tin shroud. This system performed well during all aspects of the installation.
The on-bottom stability analysis for the pipeline determined that a reduced specific gravity for the deepwater segment of the pipeline was acceptable. This element was important because the lower submerged weight allowed the installation contractor to lay at reduced installation tensions, although the tension applied to the pipe was not low by any means.
Distribution of the eight different concrete thicknesses as determined by the on-bottom stability analysis is shown in Fig. 2. Fig. 2 also shows a comparison between the water depth and the required concrete thickness to satisfy the stability analysis at this depth.
Because of high installation tensions, a means of holding the concrete to the pipe was required. Tests concluded that a spiral spray application of epoxy adhesive over the entire joint of pipe between the FBE coating and the concrete would maintain the integrity and reduce slippage of the concrete coating over the FBE coating.
During pipeline installation, it was observed that the width of a typical stress crack in the concrete coating was reduced as a result of the epoxy adhesive. Only hairline cracks were observed in the concrete of joints with adhesive between the weight coating and the anti-corrosion coating.
The only disadvantage with the adhesive was the additional time required for any concrete removal.
SCHEDULE ADJUSTMENTS
The Exxon Alabaster platform location and installation schedule challenged the SNG Mississippi Canyon pipeline project.
The original design of the platform had a J-tube for the SNG gas pipeline. But for scheduling problems, the J-tube would have been feasible. Using the J-tube would have delayed the pipeline installation until completion of the Exxon platform installation in late September 1991.
Another potential conflict was the Exxon oil pipeline which was to be installed immediately after the platform installation.
And this proposed J-tube scenario would have scheduled the SNG gas pipeline installation in the winter months with additional weather exposure and installation risk.
Because of the large diameter of the pipeline and the extreme water depths to be traversed, the decision was made to eliminate the J-tube and pre-install a riser on the platform. This plan allowed pipeline installation during the summer months when the weather was more favorable.
The elimination of the J-tube and the plan to pre-install the pipeline required the addition of a subsea riser tie-in at the base of the platform in order to connect the pipeline to the pre-installed riser.
The tie-in operation was scheduled after the platform installation and would be the only work performed during an unfavorable weather period, thus significantly reducing the weather exposure for the entire project.
This plan also improved the completion date of the project.
Another consideration was presented by the Block 397 pipeline having to cross many offshore areas with existing facilities in production as well as proposed fields that may have future production to be transported.
As a result, five subsea side taps have been installed in the pipeline at locations accessible to the developing fields. Locations of the side taps are shown in Table 1.
All of the sidetaps have been buried with 3 ft of cover for protection and have a 10.75-in. OD flange connection and a block valve.
Another feature of the SNG 20-in. Mississippi Canyon Block 397 pipeline is a subsea interconnect between the new 20-in. pipeline and the existing SNG 18 in. West Delta Block 105 pipeline. This interconnect is located in West Delta Block 75 and allows gas to flow between the two pipeline systems, depending on the available capacity in each of the lines.
A mechanical hot tap fitting was installed on the SNG 18-in. pipeline near the point of crossing the SNG 20-in. pipeline. A subsea tie-in assembly was later installed between the two pipelines.
A major portion of the interconnect assembly was installed during the lay operations, consisting of a 20-in. ball and check valve, a sidetap, flanges, and ball flanges. The design (Fig. 3) required placement of the assembly over the existing SNG 18-in. pipeline in order to ensure proper installation of the interconnect assembly.
INSTALLATION REQUIREMENTS
Installation of this pipeline was a significant event in the Gulf of Mexico and included many different operations. Because of the project's size and the tight schedule, the installation contractor had to respond with a plan involving numerous pieces of equipment and people.
OPI International was chosen as the general contractor for this project.
Before the bidding of the project, OPI pre-qualified its equipment and personnel because this project was the deepest large-diameter pipeline project the company had ever undertaken.
Detailed procedures for all of the installation operations were developed along with drawings and engineering analysis, which had to be approved by SNG and Exxon.
OPI used five construction barges to install the offshore section of this project. This portion consisted of 272,000 ft of 20-in. OD pipe distributed in water depths as shown in Fig. 4.
The OPI derrick lay barge (DLB) 423 and LB-289 were used to lay the deep water and shallow water segments of the pipeline. The BB-316 and BB-356 were used for the pipeline burial operations and foreign-pipeline crossing operations. The Pipeliner 5 was used for the beach approach tie-in with the shore segment of the pipeline.
The capabilities of the various barges were identified and changes in the equipment were made at an appropriate point with consideration for the location of the equipment at the particular time.
The DLB-423 installed the deepwater segment of the pipeline as well as the subsea riser tie-in at the Exxon Alabaster platform. This semisubmersible barge is equipped with 12 anchors with 9,500 ft of 2 in. cable and three 100-kip track-mounted tension machines.
The barge was refurbished and upgraded just before installation of this pipeline. On July 19, 1991, the DLB-423 went to sea trials prior to starting the work on this project.
On July 21 it arrived on location to begin the SNG pipeline project.
The pipeline was started at the proposed Exxon Alabaster platform end in Mississippi Canyon Block 397. The start-up procedure involved setting two 30-ton anchors with approximately 2,700 ft of 3 in. cable to hold back the pipeline for tensioning purposes.
The second anchor was a safety factor to ensure redundancy because of the critical requirements of laying the pipeline in a proposed 30 X 50-ft target area.
The placement of the pipeline was not only important for the eventual subsea tie-in but critical to the installation of the Exxon platform. Incorrect placement of the pipeline at this location could have been detrimental to the project.
Careful installation ensured proper placement of holdback anchors and positioning of the pipeline in the target area.
Extensive use of an ROV and positioning systems ensured the accurate placement of both the anchors and the pipeline.
Planning and execution successfully allowed the pipeline to be placed near the center of the target area. A remotely operated vehicle (ROV) was used to assist the start-up operations by monitoring the holdback-anchor placement and pipeline position on the sea floor.
DEEPWATER PIPELAY
Installation of the deepwater section of the pipeline used the DLB-423 with a 234 ft rigid stinger with the rollers set at a 575 ft radius of curvature. The stinger was ballasted to the proper depth and the pipe tension adjusted to the corresponding water depth as shown in Fig. 5.
Actual installation tensions were slightly higher than predicted which could be attributed to the weight-coating tolerances and other minor variables. The three tensioners were beneficial because of the potential for concrete slippage.
With tension applied to the pipe through three tension machines, the load on the concrete coating was distributed over a large area. Thus, a lower load was exerted on any given section of concrete coating.
A few isolated instances of minimal concrete slippage occurred which may have indicated that the limits of the adhesive bond capabilities was being approached. This experience may be critical for applications in which tension is applied to the concrete with fewer machines or a on smaller surface area.
There were six welding stations, one X-ray station, and two field-joint coating stations used on the DLB-423.
The pipe was welded with AWS-A569 E801OG(70+) welding rods in accordance with the American Petroleum Institute (API) 1104 welding code and the SNG specification.
Because of the critical nature of deepwater installations, SNG welding personnel-along with the installation contractor's personnel-conducted numerous tests to develop special weld-repair procedures for this project.
Internal X-ray which used an X-ray crawler radio-graphed the welds. Excessive heat in the pipeline affected the electronics in the X-ray crawler, but the slight problem was corrected by the crawler being moved further from the last welding station when not being used.
The pipeline was installed through a water depth of 1,220 ft, with anchors being placed in more than 1,500 ft of water. In these water depths, the DLB-423 experienced no weather downtime during pipelay operations.
The "as-built" diagram of the pipeline confirms that the line was installed inside of the 200 ft right-of-way throughout the route. This difficult task was accomplished with the assistance of an ROV monitoring the position of the pipe on bottom and reporting real-time positioning data to the lay barge.
The bottom conditions in most of the areas along the route consisted of soft mud and clay. In the deepwater sections of the pipeline, the ROV was able to monitor touchdown.
In water depths of 300 ft or shallower, however, monitoring the pipe at touchdown was impractical because of poor visibility. The ROV was able to inspect the pipeline after a short waiting period of a few hours in the shallow areas.
FOREIGN CROSSINGS
The offshore segment of the SNG 20-in. pipeline crossed 13 foreign pipelines in water depths ranging from 30 ft to 340 ft.
For the deepest crossings, the contractor installed 80 ft of 30-in. OD pipe parallel to the foreign pipeline as a temporary support called a "sleeper" to protect the line from potential damage during the crossing operation.
The ROV was used to assist the lay barge with the positioning of the sleeper. Once the sleeper was in place, the contractor continued the lay operation and verified with the ROV that the line had been laid across the sleeper.
This operation was repeated with the ROV monitoring the operation at three crossings. Divers working in saturation from a jet barge later removed the sleepers and installed sand-cement bags at the crossings.
Many of the foreign pipelines were lowered in place to provide the separation between the two pipelines. In some cases this was cost effective by reducing the number of sand-cement bags required at the crossings. If the foreign line could not be lowered, a typical hump crossing was performed.
The DLB-423 laid approximately 34 miles of pipeline to a water depth of 185 ft. The lay barge LB-289 which was better suited for the shallow water installation picked up the pipeline at this point and installed the remaining 18 miles of the offshore segment of the pipeline to the 15 ft water depth contour.
The LB-289 utilized a three-section articulated stinger to a water depth of 50 ft and laid the remainder of the pipeline without the use of a stinger.
Simultaneous operations were being conducted with the BB-316 working on the side-tap burial and deepwater crossings while the LB-289 completed the lay operation. The BB-316 buried the deeper section of the pipeline to the 180 ft contour utilizing the on-board saturation system for this work.
The BB-356 which was better suited for the shallow burial completed the remainder of the pipeline burial and crossing operations. A total of four sidetaps and 25 miles of 20-in. pipeline were buried by these two barges.
TESTING; FINAL TIE-IN
Upon completion of the burial operations, the line was filled and hydrotested from the shallow end with a jack up barge supporting the operation. An ROV was used to operate ROV-friendly valves on the subsea test head located at the offshore end of the pipeline and to monitor the fill and test operations.
Marking dye mixed with the test water and injected into the pipeline proved helpful with the ROV monitoring. At the start of test operations, the ROV was able to detect a leak on the subsea test head valve. The leak was repaired, and the hydrotest was completed successfully.
The subsea tie-in near the beach was performed by the Pipeliner 5, a 180 x 54 ft lay barge. This work involved a wet lift of the 20-in. pipeline and a subsea flange tie-in. The end of the pipeline was then tied into an existing flange on the shore segment of the pipeline.
This entire operation was performed in less than 2 days with the use of flotation buoys to reduce lifting stresses. The DLB-423 was mobilized back to Mississippi Canyon Block 397 to install the subsea riser tie-in assembly after the completion of the Alabaster platform installation by Exxon.
This work was performed in November 1991 during marginal sea conditions, showing the rough weather capability of this vessel. The subsea assembly was approximately 120 ft long with valves, misalignment flanges, and a sidetap.
The assembly tied into a flange on the riser, approximately 10 ft off bottom on one end and a flange on the pipeline end, located in the proposed target area.
This subsea work was done while crews from the DLB-423 completed the associated pipeline platform piping.
Copyright 1992 Oil & Gas Journal. All Rights Reserved.
