Fluid Handling

The Alaska oil pipeline: What went wrong?

Source: PlantServices.com

By Ken Schnepf

Aug 29, 2006

Maintenance changes are behind the recent partial shutdown of the oil pipeline at Prudhoe Bay, Alaska. Choices made about how often, at what expense, and what level of maintenance is appropriate are being questioned in hindsight.

“In a processed crude oil transit pipeline, which had operated for almost 29 years without a spill, we discovered areas of severe internal corrosion,” Steven Marshall, president, BP Exploration (Alaska) Inc., said August 18, before the Alaska Senate and House Resource Committee. “Given our decades of past operating experience, we did not expect to see the degree of corrosion we found in the eastern transit line. Our engineers and corrosion experts also were concerned that the corrosion program we employed had not indicated the problem sooner.”

The portion of the pipeline in question was last smart-pigged in 1998, according to Marshall. Results of that test, and earlier smart-pig data, indicated the pipeline was in good condition. Those test results and that the oil  running through that part of the pipeline is more highly processed than in other locations led BP to change to ultrasonic testing there. Smart-pigging of that pipeline was scheduled for later this year, but didn’t happen before the problems occurred.

“It appears to be major process deficiencies that have placed the pipeline at risk,” says Richard Kuprewicz, president, Accufacts Inc., Redmond, Wash., an energy consulting firm that focuses on pipelines in sensitive areas. “Even though the oil is highly processed, it seems very odd to go to that extreme.”

Although the cost of continuing a regular maintenance program of pigging and smart-pigging would be more than $100,000, it is far less than what it is costing the company now, says Kuprewicz. “Cleaning pigs are relatively inexpensive,” he says. “It doesn’t make sense not to run the cleaning pigs.”

The pipeline is constructed of X60 Grade, 0.344-in. wall-thickness pipe manufactured between 1976 and 1979, according to BP. Various segments of the pipeline are built from 30-in. or 34-in. nominal diameter pipe. Of the total 20-mile pipeline, 16 miles  are scheduled for replacement.

“Corrosion can be caused by a number of conditions or circumstances, including the presence of carbon dioxide, water, solids and microbes as well as the geometry of the lines, whether there are low spots, and fluid velocity,” explains Marshall. “It is only through laboratory testing that we will be able to confirm the corrosion mechanism. This transit line is downstream from facilities that separate crude oil, natural gas, carbon dioxide and produced water, and the oil it carries is sales quality. With that situation, we did not expect the severe corrosion we found.”

A visual inspection of the corroded pipeline by an experienced person can determine with 90% certainty what type of corrosion occurred, says Kuprewicz. Given the high level of liability involved, the company would want a full-blown engineering study. The laboratory testing should take no more than a month to complete, he adds.

Problems with the pipeline first surfaced on March 2, with a small leak. That incident lead to smart-pigging part of the pipeline on July 22, with results available on August 4. The data from the pigging identified 16 anomalies representing pipeline wall loss in excess of 70%, including two of more than 80% at 12 areas of the pipeline, according to BP.

The approximately 1.5-in. by 1.5-in. anomalies were located in the lower quadrant of the pipe.

BP started performing direct visual and ultrasonic inspection of those locations on August 5. The next day, a leak of crude oil through the pipe wall and onto insulation material was discovered and the shutdown of the pipeline was ordered. Another leak of about five barrels of processed crude oil along with multiple holes in the pipe wall were detected the same day. Additional pinhole leaks were found at four other locations.

BP uses inline inspection (ILI) tools, or smart pigs, regularly on the North Slope. The company has run between five and six inline inspections each year since 2000, with more than 180 smart pigs on various pipelines since 1986. Most pipelines on the North Slope are not piggableand the significant majority are above grade and accessible=, says BP.

This means other inspection techniques(e.g. ultrasonic inspection and radiographic inspection) are used to detect and monitor corrosion.

Between August 1998 (the date of the last smart pig run) and February 2006, BP conducted approximately 1,200 inspections of various kinds on the oil transit line. There were another 1,600 inspections done on other oil transit lines between January 1, 1998 and March 2, 2006, according to BP. One-hundred percent of the samples from Prudhoe Bay oil transit lines tested between 1995 and 2005 showed corrosion rates below the company’s target of 0.002 in. per year. Annual ultrasonic inspection data showed manageable corrosion rates, although there was some evidence of increased corrosion rate in the GC-2 oil transit line in September 2005. BP expects to spend $72 million on corrosion control this year.

“We are still seeking to understand exactly what caused the pitting of the line, but we won’t know for sure until we can conduct laboratory tests,” says Marshall. “Not knowing exactly what we were up against, the only responsible option was to protect the environment from potential spills by shutting down the field in an orderly fashion.”

Checks and balances to reinforce the maintenance decisions are missing, says Kuprewicz. Cost reductions should never be done at the expense of the business.
“We identified a gap in our corrosion inspection system and we will correct it,” says Marshall. “In the future, we will have a better system to protect our pipelines and we have already gained important new operating knowledge. Through adversity, we will enhance our operating capability.”