Alyeska Pipeline Service Co. plans the petroleum industry's most expensive corrosion repair campaign on its Trans Alaska Pipeline System (TAPS) carrying North Slope crude oil to a tanker terminal near Valdez in southern Alaska.
"We are in no danger of an immediate leak, but it's something we need to address," Bill Howitt, Alyeska's manager of engineering, told Oil & Gas Journal.
In addition to external corrosion along portions of its 800 mile, 48 in. main line system, Alyeska has found internal corrosion at some of the system's pump stations.
Detection of both types of corrosion was made in early stages, but repairs will still be costly.
The budgeted amount for main line corrosion inspection, repair, and replacement for 1990 is about $215 million. The budget for all corrosion activities by Alyeska in 1990 is $300 million.
"We estimate in the next 5 years we will be spending $600-800 million on corrosion," Howitt said.
MAIN LINE SYSTEM PROBLEMS
It took a newly developed corrosion detection pig to find the external corrosion.
Alyeska performs standard corrosion mitigation programs required of all pipelines by the U.S. Department of Transportation, but the standard programs, including annual cathodic protection surveys, did not reveal corrosion problems on TAPS.
Unlike most other pipelines, TAPS is subject to a set of additional requirements. For example, the pipeline's federal right-of-way grant requires Alyeska to develop and regularly run a corrosion detection pig program.
A corrosion detection pig has been sent through the system every year since the first run in 1978, less than 1 year after Alyeska started up.
All early corrosion detection pigs and most still in service today are based on magnetic flux leakage. The pig carries either magnets or electromagnets through the pipe. Magnetic flux leakage is measured to determine if a loss of metal has occurred in the pipe wall.
Howitt said a drawback of magnetic flux leakage technology is that many noncorrosion type defects and nonmetal loss defects have a signature close to corrosion metal loss. Laminations in the pipe, metallurgical inclusions, dents, and internal weld beads can be confused with metal loss.
THE 50% THRESHOLD
Corrosion indicated by magnetic flux pigs can be determined only by visual inspection-digging up underground pipe or taking off the insulation of aboveground pipe.
Further, prior to 1986 a magnetic flux pig operating in 48 in. pipe could detect only pipe wall losses of 50% or more on a reliable basis, Howitt said.
The 50% threshold was determined by vendor benchmark tests. Alyeska provided vendors with 48 in. pipe with defects milled into the pipe by Alyeska. The vendors pulled their pigs through 60 ft, 100 ft, or 200 ft sections of the pipe and gave Alyeska an interpretation of their findings. Alyeska compared the report to actual metal loss in the pipe wall.
Through 1986, Alyeska ran Vetco magnetic flux pigs.
The Vetco pigs found a number of anomaly locations. In the early years, each was dug up. Then Alyeska switched to digging up representative samples of the best signals.
"We found no corrosion," Howitt said. "There were no exceptions from 1978 through 1987. "And that's not surprising because it would take an aggressive corrosion rate to reduce wall thickness by 50%."
RESEARCH PROGRAM NEEDED
In 1984, Alyeska conducted a major review of its corrosion program, including trial digups of underground pipe to audit the detection programs. The company also assessed what was needed in corrosion detection pig technology to move into the 1990s and beyond.
"We decided a 50% of wall loss detection threshold was clearly inadequate," Howitt said. "We needed to do better than that."
The goal was to get the detection threshold down to 10% of wall loss to provide the company with enough early warning so a major problem could be avoided. After interviewing vendors, the company began a two phase research and development program.
The first phase involved improving magnetic flux leak technology, particularly with better detection sensor technology and the use of rare earth magnets that more thoroughly saturate the pipe wall.
Alyeska in 1985 hired International Pipeline Engineering Ltd., now known as Pipetronix, to develop within 2 years an enhanced pig able to detect 30% wall loss reliably.
Alyeska also signed a longer term agreement with NKK Corp. for development of an ultrasonic based pig with the potential of going to 10% wall loss detection. The agreement was that NKK would conduct the development work, while Alyeska would run the development pigs through the pipeline. Production was targeted for 1989.
Ultrasonics was chosen because it provides a direct measurement of remaining wall thickness, while magnetic flux has a lot of analog signals that require a technician to interpret data much like seismic data interpretation.
But ultrasonic technology required a breakthrough to package it into a pig in which the electronics could survive as much as a 450 mile trip through a 48 in. pipeline.
FIRST RUNS SHOW CORROSION
Alyeska made the first run with the improved magnetic flux pig from Pipetronix in 1987.
The run indicated 14 anomaly locations. All were dug up and, again, no corrosion was found.
During the 1987-88 winter, the pig was returned to Pipetronix's shop for tuneup and minor electrical and mechanical modifications. The focus, however, was to obtain better interpretation of data.
The pig was run again in summer 1988. The report, which came out in December 1988, showed 241 anomalies.
In January 1989, four representative aboveground and four representative underground sites were examined. Nothing was found above ground, but for the first time corrosion was found on the outer wall of the pipe in all four underground sites.
"That told us we needed to take the 241 anomalies very seriously," Howitt said. "We made a commitment to examine them all."
A major program with multiple field crews began in March 1989.
In the meantime, Alyeska ran the first NKK production ultrasonic pig through the system in July 1989. NKK began feeding Alyeska data by sections in September 1989.
By December 1989, Alyeska had 408 anomaly sites indicated by the NKK pig. Because of things such as slight differences in longitudinal positions, Alyeska could not tell if the NKK anomalies were repeats of the Pipetronix anomalies, so the NKK anomalies were added to the growing data base.
A GROWING DATA BASE
Alyeska had finished examining about 100 of the Pipetronix anomalies by the fall of 1989.
Data from the 100 examinations were given to Pipetronix with instructions to factor and correlate them into the company's interpretational data base. Alyeska wanted all the data tapes reinterpreted to see if the corrosion problem was less or greater than the 241 original anomalies.
"They came back to us with more than 100 additional sites," Howitt said.
Alyeska now had 827 anomalies to inspect, and the company began a major inspection program.
It decided that the absolute depth of the corrosion would not be the deciding factor in which piece of corrosion to go after first. The three factors that decide priority are depth, shape, and operating pressure of the pipe at the site of the anomaly.
All the anomalies were put in either Priority 1, 2, or 3.
Priority 1 sites are those where, if there is general corrosion, pipe strength could be affected under normal operating pressure.
Priority 2 is where pipe strength could be affected under maximum operating pressure, if in fact general corrosion exists. Priority 3 is superficial corrosion if corrosion exists at all.
By the end of last January, Alyeska had completed inspection of all 116 Priority 1 sites. In all, the company completed inspections at 308 anomaly sites, including some Priority 2 and 3 sites that were adjacent to Priority 1 sites.
WHAT ALYESKA FOUND
Alyeska found that the serious areas of exterior corrosion on the main line are concentrated in about 13 total miles of pipe in four major areas: three in or near the Brooks Range and one just north of Fairbanks.
There was no significant internal corrosion problem detected on the main line and no exterior corrosion problem in the aboveground portion of the main line.
"We found the average metal loss for the areas that had corrosion to be about 10-20%," Howitt said, which confirmed the no-find reports provided by earlier pig corrosion detection technology.
The main area of corrosion is concentrated in the Atigun River floodplain, north of Atigun Pass on the northern slope of the Brooks Range between Mile Posts 157 and 165.5 on the main line. The section has more than 300 of the anomalies detected by the Pipetronix and NKK pigs.
South of the floodplain, there is a 2 mile segment in Atigun Pass that contains 88 anomalies. In the Chandalar area south of Atigun, there is a 1.8 mile segment of pipe with 77 anomalies.
The Wilbur Creek area, about 70 miles north of Fairbanks, has a half mile segment with 50 anomalies.
Repair and replacement will be concentrated in fairly small areas.
"This doesn't mean to say that the rest of the corrosion is trivial," Howitt said. "It means we have some very defined pockets and then a couple of other places in the remainder of the main line.
"The big news is that you don't have a general corrosion problem throughout TAPS. You have a concentrated problem that can be addressed-caught early enough so it can be addressed in a phased manner."
ALYESKA'S PLAN
Alyeska plans to replace 9 miles of pipe in the Atigun River floodplain where there has been a general coating failure. Water in the area flows year round in the subsurface.
The company will lay a new line offsetting the existing line by 30-50 ft with a tie-in at either end. Pipe for the repair has been purchased and is expected to be rolled in May.
The coating will be fusion bonded epoxy with a concrete overcoat for mechanical protection. All engineering for the replacement is scheduled to be done by July, with a construction contract to be awarded in July or August.
The work is scheduled for the spring and summer of 1991. Alyeska plans to tie in the new line during third quarter 1991, which will require a 1-2 day shutdown of TAPS.
Alyeska considered refurbishing the pipe in place, which is a common practice elsewhere. However, there is no alternative transportation for North Slope production.
For the other sections of corrosion, Alyeska plans to complete a feasibility study in June. A final decision is expected in July or August, but construction isn't expected until 1992-93.
At each inspection site on the main line system, pipe is sandblasted down to near white metal. The company then puts a two part zinc rich epoxy phenolic coating on the pipe, overwrapped with a shrink sleeve. The cathodic protection is reinforced.
Alyeska, which has returned to the field with five crews this year, plans to finish inspecting all 827 anomalies this year.
MORE PIG RUNS
Alyeska has determined that the Pipetronix pipe registered a 70% hit rate and the NKK pig a 99.9% hit rate for corrosion detection, based on field inspection confirmations to date.
The company, which plans to run the NKK pig again in late May and early June, expects to find more anomalies due to development work on the pig.
On the NKK pig, there are 255 3/4 in. transducers arranged in a spiral. As the pig travels down the pipe, it sounds an ultrasonic pulse through each transducer that sends back echoes as it enters and exits the pipe wall.
Based on the echoes, an onboard computer calculates the wall thickness. External or internal corrosion is determined, as well as the shape of the corrosion.
The pig provides a 1/2 in. by 1/2 in. grid of the pipeline. Of the three microprocessors on board, one controls the sequencing, another captures the data and determines if there has been a wall thickness loss, and the third stores the data if there has been a 10% or greater loss of wall thickness.
"We were right in developing the NKK pig because its threshold is 10%," Howitt said. "We have done what we set out to do: develop a tool that would give us an early warning."
DEAD LEG CORROSION
Another corrosion problem on TAPS is at the pump stations, which have dead leg piping where formation water in the oil stream can drop out.
There is not much water in the oil that flows through TAPS, Howitt said, "but since we're shipping about 1.9 million b/d of oil, a lot of water comes along for the ride."
TAPS has dead leg piping at sites where connections were installed for future use and where flow is intermittent, such as relief piping to tanks.
"We have a corrosion problem there on the very bottom 4 in. of the pipe," Howitt said.
The problem is concentrated in northern pump stations because there is very little water left to drop out of the oil when it reaches southern pump stations.
The two main types of corrosion are conventional oxidation corrosion from slightly salty formation water laying on the pipe and sulfur reducing bacteria (SRB) induced corrosion. SRBs are metabolized compounds of sulfur that are naturally present in the crude oil. The byproduct is hydrogen sulfide.
Howitt said North Slope producers are combating SRBs from enhanced oil recovery projects by using biostatic agents. But it is only a mitigation method, not a prevention method.
"We have now had a breakthrough into the pipeline system," Howitt said. "It was an expected thing. It happens on every pipeline that ships water from an EOR field."
Alyeska is combating the pump station corrosion problem with several methods. One is by getting rid of dead legs.
"We are at the point in TAPS where we have reached peak flow," he said. "We know what dead leg connections are surplus to our requirements, so as we have the opportunity we are eliminating the dead legs. For instance, replacing a T with a smooth elbow."
USE OF INHIBITORS
Another anticorrosion method being used by Alyeska is oxygen scavenger, filmer, and biostatic inhibitors.
The pipeline has installed inhibitor systems in all its northern pump stations, Pump Stations 1-4, and is in the process of establishing inhibitor systems in all its southern pump stations, Pump Stations 5-12.
Another part of the pump station program monitors the rate of corrosion, strength of inhibitors, depletion of inhibitors, and active colonies of SRBs. The monitoring program will determine if Alyeska must make major repairs.
The inhibitor program, however, came too late at Pump Station 3, a critical station for TAPS.
This summer the company plans a $40 million project to replace most of the main crude piping at Pump Station 3, remove most of the dead legs, and install an advanced inhibitor system, among other things.
To avoid downtime for the main line, the company plans to bypass the station and install a contingency pump that will allow it to maintain the flow unimpeded at anticipated summer rates.
"That gives the crews 21-30 days to replace the manifold piping and take out some of the dead legs," Howitt said.
Another pump station project involves reinsulation. The original insulation system for belowground piping at permafrost pump stations relied on staying dry, a condition that did not exist.
After a number of attempts to work on piping and use various waterproofing system, Alyeska decided to redesign the system to work when wet. A prototype installed in 1983 at Pump Station 3 worked, so a phased program was begun in 1984 to reinsulate all permafrost pump stations.
The $5-10 million/year reinsulation program is scheduled for completion in 1992.
VALDEZ TERMINALS
Alyeska also has a major tank inspection program under way at Valdez terminal.
The terminal has 18 tanks, each with 500,000 bbl capacity. The company takes three tanks out of service each summer for draining, cleaning, corrosion inspection, and repair if warranted.
The program is scheduled for completion in 1994.
Copyright 1990 Oil & Gas Journal. All Rights Reserved.