TAPS REROUTE MEETS SCHEDULE, BEATS COST ESTIMATE

E. Stephen Newcomer, Patrick G. McDevitt
Alyeska Pipeline Service Co.
Anchorage

Heavy corrosion concentration prompted Alyeska Pipeline Service Co., operator of the TransAlaska Pipeline (TAPS), last year to replace 8.5 miles of 48-in. OD main line in Alaska's Atigun River floodplain.

The Atigun project was one of the most successful projects ever undertaken by the company since TAPS' initial construction in 1977.

It was completed on schedule and at less cost than the budget estimate, which was based on historical costs for similar work in the area.

Additionally, the project received an award for Project of the Year from the Alaska Chapter of the Project Management Institute and is nominated for the national PMI award.

More than 8 billion bbl of North Slope crude oil have been transported through TAPS; current throughput represents 25% of U.S. domestic crude-oil production.

The project's remote location (Fig. 1), 135 miles north of the Arctic Circle in the heart of the Brooks Mountain Range, required that temporary living quarters for approximately 500 people and offices, shops, and a communication system be constructed in addition to the new pipeline.

The construction site was subject to flooding, rockslides, avalanches, and temperatures that reach -60 F. in the winter.

The new line began service 27 months after the project commenced in spite of environmental limits and the rigors of international procurement of materials.

The tie-in of the new section was a complex and challenging project within itself. Eight stopples and two bypass lines were employed to reduce pipeline shutdown time and maximize crude production during the tie-in operation.

A team of Alyeska employees managed the entire project, from conceptual engineering through construction and commissioning of the new pipeline segment. They completed conceptual engineering, selected and managed prima engineering and construction contractors, and managed the international effort to procure materials and transport them to the site.

The project was highly scrutinized by Alyeska management, regulators, and the public. Any deviations from a safe and environmentally sound completion were unacceptable.

INSPECTION PROGRAM

In 1978, shortly after its construction, Alyeska initiated a corrosion-monitoring program for TAPS. The original program used instrumented, in-line inspection devices ("pigs") with conventional magnetic flux leakage (MFL) technology to detect metal loss in the pipe wall. These original instrumented pigs could reliably detect a 50% pipe wall loss.

In 1984, Alyeska embarked on a worldwide search for enhanced corrosion-detection technology which could identify pipeline corrosion in its early stages. This search led to the redesign of the MFL pig developed by IPEL (now Pipetronic) and to the development of a new ultrasonic pig by Japan's NKK Corp.

Alyeska's corrosion-monitoring program, including the use of intelligent pigs, did not detect substantial external pipeline corrosion until late 1988 when the company ran the IPEL pig through the line. The findings were further verified in 1989 by an NKK pig.

Data from these two runs identified concentrated external corrosion within the 8.5 mile Atigun floodplain pipeline section.

Initial investigation by the project team showed that coating on the existing pipe had failed. The pipe had originally been coated with a fusion bonded epoxy (FBE) and wrapped with a specially designed polyethylene-based tape.

The existing cathodic-protection system consisted of two zinc ribbon anodes placed on each side of the pipe. These anodes were attached to the pipe at 500 to 1,000-ft intervals.

Examination of the failed coating indicated that angular bedding and padding material had penetrated the tape and allowed water to migrate beneath the tape and the FBE. The cathodic-protection system was therefore ineffective because this coating shielded the electrical current and thus allowed the pipe to corrode.

OPTIONS

The project team studied various options for correcting the corrosion problems. Options included repairing the original pipe in place, installing new pipe aboveground, installing new pipe belowground, and various combinations of these options.

Repairing the existing pipe was not economically feasible. An elevated line for the entire length was impossible because of the potential for debris flow and snow avalanches.

Alyeska selected a buried alignment with a typical minimum offset of 30 ft to enable placement of the line within the thaw bulb of the existing pipe. This approach would minimize any thaw settlement of the replacement pipe and disturbance of the ground surface.

Because the pipeline operates at a temperature of 120 F., thawing of ice-rich permafrost soils can cause substantial differential pipeline settlement. This is a primary concern during selection of a route for belowground arctic pipelines.

The 30 ft minimum offset was also based on maintaining sufficient soil restraint along the existing pipeline during excavation of the replacement ditch and permitting blasting for frozen ground and bedrock excavation without damaging the existing pipe.

Concurrent with the selection of the replacement method, Alyeska initiated a geotechnical drilling program between October and December 1989. This investigation included a total of 32 boreholes and two test pits.

Instrumentation, including thermistors and piezometers, was installed in the boreholes to indicate the groundwater and thermal conditions. The results of this investigation confirmed the geotechnical suitability of the replacement pipe route.

Once the replacement concept had been developed, the project team hired Williams Bros. Engineering Co., Tulsa, to provide the engineering and quality assurance for the job.

DESIGN CHALLENGES

The design of the replacement section of pipeline complied with TAPS' original design criteria and government stipulations, ASME B31.4 ("Liquid Transportation Systems for Hydrocarbons, Liquid Petroleum Gas, Anhydrous Ammonia and Alcohols"), and U.S. Department of Transportation (DOT) CFR Title 49 Part 195.

One of the major challenges at the beginning of design was to determine the best coating system for this application. A task force of coating experts from Alyeska and several of its owner companies set out to determine the best possible coating for TAPS' operating conditions and at the same time the most suitable coating for installation in arctic weather conditions.

Battelle Laboratories conducted extensive laboratory testing for various coating systems. The criteria for the coating system were the following:

Belowground use and wet service up to 145 F.
Effective moisture and water-vapor barrier
Electrical isolation of the pipe from its environment
Adequate shear resistance
Bendability according with API standards (1-1/2/diameter length)
Resistance to damage from ultraviolet rays
Compatibility with coating system for field-weld joints.

Battelle's test results determined that Valspar Inc.'s D1003LD FBE met all of these requirements.

In addition to the FBE coating, the pipe was coated with 1.25 in. of concrete to provide both mechanical protection and an enhanced pH environment for the pipe (Fig. 2). Alyeska selected concrete over other materials because it is an ineffective insulator for the cathodic-protection system.

The field joints were also concrete coated, thus resulting in an homogeneous coating system for the replacement pipeline.

Field testing of concrete grouting of field-weld joints occurred in December 1990. Tests explored variations in grout temperature, water content, mixing procedures, form preheat vibration techniques, and curing.

The grouting of field-weld joints would occur in winter, and the grout would need to obtain sufficient strength rapidly to permit lowering of the pipeline into the ditch.

The grout was mixed in a heated trailer and placed in heavy-gauge steel forms with adjustment bolts at the top. The forms and pipe were preheated by a forced air heater with a flexible duct and customized nozzle which directed the air in the annulus formed between the pipe and metal form.

After placement of the grout, heat was maintained for 24 hr. From the results of the tests came a final grout-mix design and procedure for placement of the concrete grout during construction.

IMPROVED CP

The cathodic-protection design improved the system used on the existing pipe by using four magnesium ribbons, two on the top and two on the bottom of the pipe. Each magnesium ribbon provided 50% more current than one of the original zinc ribbons, thus giving three times as much cathodic protection as the original installation.

Test stations at approximately every 500 ft would monitor the cathodic-protection system.

The pipe installed originally in this section of the TAPS pipeline was API 5L X-65, 48 in. OD and 0.462 W.T. The replacement pipe was API 5L X-70, 48 in, OD and 0.562-in. W.T. The additional wall thickness and strength provided a greater corrosion allowance.

The new pipe was DSAW (double submerged arc welded) high yield strength with high impact toughness. The Charpy impact values were greater than 50 ft-lb at -20 F.; the carbon equivalent was less than 0.40 wt %.

Pipe bends required for laving of the pipeline were prebent and coated with concrete in a shop in Fairbanks, then shipped to the job site for installation. Because the pipeline was concrete coated, the right-of-way (ROW) and ditch had to be constructed to specified grades to permit the prebent concrete coated pipe to be installed to the design grade.

RIVER TRAINING

Another significant design feature was the use of river training structures. The entire pipeline alignment within the floodplain of the Atigun River was subject to main channel flow and scour.

After research of available river flow and local precipitation records, the design flood conditions were established: flow--3,042 to 13,640 cfs; flow depth--6 ft; flow velocity--16 fps.

Based on the design flood criteria and river scour studies, Alyeska established a minimum burial criterion for the Atigun floodplain of 5 ft below the thalweg elevation (low point of the stream cross section). The elevations along the floodplain cross section are variable relative to the thalweg.

This criterion resulted in many areas of design burial depths of 15-20 ft. Existing pipe was installed at these burial depths. Alyeska designed and used an alternative design solution to deep burial for the Atigun pipe replacement.

River training structures were used to protect the replacement pipe from river scour and permit shallow burial. This approach enabled a 5-10 ft reduction in the burial depth.

The river training structures were gabion mats and articulated concrete mats placed above the pipe and flush with or below the riverbed level (Fig. 3).

The gabions consisted of a wire basket (1-ft deep x 30-ft wide) filled with 3-8 in. cobbles and placed approximately 2 ft above the pipeline. The cobbles were readily available as by-product from the screening of alluvial gravels in the production of bedding and padding materials.

The articulated concrete mats consisted of 1 ft x 1 ft x 6 in. individual concrete blocks cabled together to form mats that were 8 ft x 24 ft (Fig. 4). These mats were placed approximately 2 ft above the pipeline and were used in the areas where the river velocity was the greatest. The mats were fabricated in Fairbanks and shipped to the jobsite for installation.

Use of river-training structures in conjunction with the shallow burial had the following advantages over deep burial:

Quality of the pipe placement; a dry or readily dewatered ditch resulted in significant quality improvements in placement of the pipe and backfill.
Easier and faster pipeline construction; significantly reduced ditch excavation and expedited placement of the pipeline in the ditch. The river-training structures were independent of pipe laying and thus reduced the construction schedule for pipe laying.
Less environmental impact; river training reduced the amount of area to be distributed and greatly reduced dewatering activities.
Cost; river training reduced the construction cost as a result of less excavation, dewatering, and shorter construction schedule.

EXPLOSIVES; AVALANCHES

Excavation of frozen soils and bedrock for the replacement pipeline ditch required the use of explosives.

Blasting, used for almost the entire replacement pipeline ditch, was conducted close to the existing line. In many cases the blasting took place less than 30 ft from the operating pipeline.

Detailed technical procedures for this blasting were developed based on experimental work conducted by the American Gas Association.

As part of these procedures, verification test blasts were conducted at the project site prior to production blasting. The test blasts were carefully monitored to determine the actual peak ground particle velocity in both the frozen soil and bedrock materials. These data were used to determine the explosive charge weight limitation for specific distances from the operating pipeline.

Ground-particle velocities were carefully monitored for each production blast and adjustments were made to the blasting plan based on this monitoring. Nonelectric detonators were specified to reduce the risk of accidental detonation. Blasting was successful without damaging the existing pipe and resulted in a significant cost savings when compared with other excavation methods in frozen soils and bedrock.

The southern portion of the project is within an area of potential avalanches. They periodically closed the only road to the project from the south. Before construction, avalanche experts provided a better understanding of the risk and developed forecasting and control methods. The area was extensively mapped and potential avalanche paths identified before construction. During construction within the avalanche areas, a full time avalanche control and forecasting person was present. Explosives were utilized to shoot down potential avalanches and unstable snow deposits before construction personnel worked in the risk area.

The hydrotest of the replacement pipe occurred in one 8.5 mile section which had an elevation difference of 895 ft. The test was scheduled for early June just after winter breakup to ensure an adequate supply of water.

The water for the hydrotest came from construction of a small pond area in the Atigun River. This pond was also intended to be used for the discharge of the hydrotest water. The replacement pipe section was qualified to a maximum allowable operating pressure (MAOP) of 970 psig. The maximum test pressure was governed by a hoop stress of 100% of the specified minimum yield strength (SMYS) of the pipe. The minimum test pressure was 125% of the MAOP. Cleaning and gauging of the pipe section took place before the hydrotest.

STOPPLING

Initially, tie-in of the new 8.5 mile pipe segment was to take place by a complete shutdown of the existing pipeline and a draindown of the replacement segment into pump station storage tanks north of the construction site. This tie-in method was undesirable because of the revenue loss of 4 million bbl of oil associated with a 2-day pipeline shutdown to accomplish the task.

In lieu of this, a pipeline stoppling and bypass system was evaluated and used for the tie-in of the replacement pipe section. This system stem used four stopples at each tie-in location to isolate the existing pipe and permit the tie-ins to be accomplished. Pipeline throughput was maintained through two bypasses that were constructed around both of the tie-in areas (Fig. 5). Stoppling entails the hot tapping of the 48-in. pipeline under normal flow and pressure conditions and insertion of a plug which internally seals and accommodates the full pipeline operating pressure. Such stopple systems had been successfully used in smaller lines.

Alyeska's experience with 48-in. stopples prior to the Atigun project had been minimal. This past experience dealt with pipeline bypass operations using two stopples (single bypass with a stopple on each end), whereas the Atigun project used four stopples at each tie-in location (eight stopples total) for a double block on each side of the tie-in points. Developing new procedures for installation of the stopples involved full size testing of actual hardware. This effort was conducted by the same personnel who would be employed under field conditions. Before the stopples were set, the 36-in. bypasses were constructed around each of the tie-in points. The north and south bypasses were approximately 1,000 ft and 300 ft in length, respectively (Fig. 6).

The initial four stopple plugs were set over a 1 hr 48 min period. During the initial stopple setting operation, Alyeska shut down the pipeline flow between the pump stations bordering the replacement segment. The remainder of the pipeline continued to operate into tankage at a reduced rate of 25% of normal or about 450,000 b/d.

The setting of the initial four inboard stopple plugs allowed routing of the entire pipeline flow through the two bypasses around each of the tie-in areas. After the initial stopples were set, the pipeline flow was increased to normal and the remaining stopples were set. Vacuum trucks then drained the tie-in areas between the stopples. To identify any stopple seal problems, the pressure between the adjacent stopples was reduced and carefully monitored before and after the pipe was cut.

The tie-ins were completed, replacement pipe filled with oil, stopples removed, and the replacement pipe commissioned 10 days later after the initial stopples were set.

Draindown of the 8.5 mile section of the existing pipe to be taken out of service was accomplished with a high-pressure pump to reinject the oil into the operating line (operating pressure, 760 psig). The draindown of the existing pipe took place over a period of 4 days at a rate of approximately 350-500 gpm. The existing line was thoroughly cleaned with scrapper pigs, solvent, and foam pigs and abandoned in place.

CONSTRUCTION

Detailed engineering was completed by the end of May 1990 and bids were issued to prequalified projects.

Alyeska carefully screened potential contractors by issuing prequalification documents to contractors and followed up with selective interviews. The objective was to invite only those contractors Alyeska viewed as qualified and had resources to undertake the project.

Alyeska sought and received lump sum bids covering construction of the replacement line. Project definition and documentation was of sufficient detail to enable all bidders to respond with lump sum bids. This was the first time Alyeska had constructed main line pipe under a lump sum contract. Price/Northland, a joint venture consisting of Price Construction Co. and Northland Construction Co., received the work.

Before construction could begin, Alyeska had to obtain 35 separate permits from nine different government agencies. These permits were complete in August of 1990 and construction of facilities and right-of-way was begun in the same month. ROW construction was completed in December.

Environmental constraints had an effect on the construction schedule.

To take advantage of thawed soils, mild temperatures, and to minimize wildlife disturbance, mobilization, camp construction, and right-of-way work was scheduled for the fall of 1990. Pipeline construction was to proceed in February through April when surface waters are virtually nil and subsurface waters are deepest, thus minimizing the ditch dewatering effort.

Field investigation regarding dewatering requirements indicated that dewatering was necessary for about half of the ditch. A relatively dry ditch was required for the installation of the cathodic-protection system.

A ditch dewatering plan was developed that utilized wells and ditch sump pumps. Discharge of sump pumps to upland areas was the primary method used to meet stringent discharge permit requirements. Spoil from the ROW construction was placed between the pipeline route and the river to serve as a temporary dike. This also minimized environmental disturbance.

During this time period, ambient temperatures were typically well below 0 F. with as little as 3 hr of available daylight. However, temperatures and available daylight increased from January to June, which improved working conditions and morale. Work in the canyon area of the project was completed first to reduce the exposure of personnel to spring avalanches. The new pipe had to be backfilled before spring thaw and potential flooding in May.

Ditching, dewatering, and pipeline laying began in February 1991. The contractor welded an average of 30 joints (60 ft long)/day. Installation of the welded pipe and backfilling were completed in April. The installation of the articulated concrete mats and gabions took place in May; hydrotesting of the new tine, in June. Final tie-in of the new pipeline and commissioning of the line took place in August 1991. Rehabilitation of the Atigun River floodplain and work area occurred in October.