BP: Learning from oil spill lessons
Johnson, Beaudo explain what they think happened in Prudhoe Bays largest spill and what the company is learning from the incident
On March 2 a BP well pad operator discovered a leak in the transit line that delivers oil to the trans-Alaska pipeline from Gathering Center 2 in the western operating area of the giant Prudhoe Bay oil field on Alaskas North Slope.
The leak occurred in the transit line segment between GC-2 and the point where the production from Gathering Center 1 enters the line.
BP launched an immediate response to what, with an estimated volume of around 200,000 gallons, proved to be the largest spill in the history of Prudhoe Bay.
The cause of the leak became obvious within a few days of its discovery: internal corrosion had caused a one-quarter-inch hole in the bottom of the transit pipeline.
The hole had formed in a section of line buried under what is termed a caribou crossing, a culvert designed to allow animals to cross over a pipeline as opposed to going under an elevated pipeline.
The winter snow covered the leaking oil, so the spill remained undetected, probably for several days. It was odor the smell of oil that ultimately exposed the leak to a worker.
Cleanup of the spill site has now been completed, according to the Alaska Department of Environmental Conservation. To date BP has spent $8 million responding to the incident, Daren Beaudo, spokesman for BP in Alaska, said in a recent interview with Petroleum News.
Beaudo and Maureen Johnson, BP senior vice president for Greater Prudhoe Bay, talked to Petroleum News about how the leak developed and BPs actions to prevent a similar problem in the future.
Microbiological corrosionAll indications are that the corrosion that caused the hole in the transit line was biological in origin, caused by sulfate reducing bacteria inside the pipeline, Johnson said.
The evidence has mounted that that is true, she said. Weve scanned what the (corrosion) pits look like in the bottom of the pipe.
Also the way in which the corrosion in the pipeline accelerated over time is characteristic of the way in which microbiological corrosion develops, as the bacteria grow and multiply.
The bacteria form in water, so that problems associated with microbiological corrosion tend to be associated with water carrying pipelines, such as the lines that are used for waterflood operations.
In fact, BP treats its water injection lines at Prudhoe Bay with anticorrosion chemicals and runs maintenance pigs down the lines at frequencies ranging from weekly to monthly (a maintenance pig is a device that passes through the inside of a pipeline, scraping and cleaning the inside walls of the pipeline). The company pigged its Prudhoe Bay pipelines more than 350 times in 2005, Johnson said.
Transit lines less corrosion proneBP has viewed oil carrying transit lines, such as the line from GC-2 that developed a leak, as much less susceptible to corrosion than a water bearing line. But the company has regularly monitored the Prudhoe Bay oil transit lines for internal corrosion using two techniques: ultrasonic testing and the use of corrosion coupons.
Ultrasonic testing involves the use of an ultrasonic device to measure the thickness of the pipeline wall a thinning of the wall indicates the presence of corrosion. A corrosion coupon is a small metal plate placed inside the pipeline and inspected for corrosion every 90 days.
We had 29 years of history that said these systems were not a great concern, Johnson said. All of those things told us not a big worry here.
In addition to the program of frequent inspections, BP runs what is known as a smart pig down each transit line every few years a smart pig contains instruments that can measure and test the condition of the pipeline, including the detection of corrosion damage.
The company ran a smart pig through the GC-2 transit line in 1990 and again in 1998. According BPs incident investigation report for the transit line oil spill, the 1990 pig run noted nothing of significance and the 1998 pig run showed moderate internal and external corrosion. The evidence for corrosion in 1998 was confirmed by ultrasonic testing BP subsequently used the ultrasonic testing to monitor any continuing development of corrosion in the line.
However the BP report says that the corrosion detected in 1998 was well within the BPXA fit for service criterion and that the subsequent ultrasonic testing up to 2004 indicated that corrosion was not highly active during this period. It was only in an inspection in September and October of 2005, that evidence of increasing corrosion activity started to appear.
The increasing amount of corrosion found in the fall of 2005 caused BP to step up the inspection program on the pipeline the company increased the number of inspection points, increased the frequency of inspections at some points and scheduled a smart pig inspection for the summer of 2006, according to the BP report.
However, an inspection of the line after the March 2006 leak showed evidence of high rates of corrosion, even in place that had been free of corrosion in the fall 2005 inspection. Clearly, there had been an exponential growth of corrosion, culminating in the hole that caused the oil spill.
Whether it (the exponential corrosion growth) was over six months or 12 months or 18 months, that becomes a little bit more of a judgment call, but it was pretty recent, Johnson said.
And because the rate of corrosion accelerated so rapidly, more frequent smart pigging would probably not have detected the problem unless, perhaps, the pigging had been done in the summer of 2005, Johnson said.
So why did the corrosion accelerate so rapidly?No one knows for sure. But the BP investigation report theorizes two main factors that came together at the same time.
The first factor relates to corrosion inhibitors. BP adds about 3 million gallons per year of these inhibitors to Prudhoe Bay production fluids; the fluids carry the inhibitors into production facilities such as GC-2.
We have been injecting something like 3 million gallons of corrosion inhibitor a year
And its going up every year as the amount of water we handle goes up, Johnson said.
But the corrosion inhibitors appear to have been present in relatively low concentrations in the GC-2 production facilities, when compared with the other Prudhoe Bay facilities. It is, therefore, likely that the fluids passing down the GC-2 transit line from GC-2 contained only small amounts of the inhibitors, thus providing opportunities for corrosion-causing bacteria to grow.
The corrosion inhibitor shortfall may have occurred because GC-2 is the only facility at Prudhoe Bay that processes viscous oil. The viscous oil production introduces more solids into the processing facilities than traditional production and BP thinks that these additional solids may have adsorbed some of the inhibitor.
The second possible factor was the relatively low flow rate in the GC-2 transit line, upstream of GC-1. With Prudhoe Bay production in decline, the transit line was carrying much smaller volumes of oil than the line was designed to handle the resulting sluggish flow may have enabled an increased build up of water in the line and provided an environment conducive to the incubation of bacteria. There was much less corrosion downstream from GC-1, where the addition of the fluids from GC-1 would have increased flow rates in the line.
Downstream from GC-1 we dont see real high corrosion, Johnson said.
There was also a build up of solids in the GC-2 transit line over a period of several years, as sediment carried by the fluids from the GC-2 processing facilities settled in the pipeline. Although it is not possible to discount the possibility that these solids contributed to the corrosion, it is unlikely that the solids were a significant factor, Johnson said. Johnson pointed out that there are about two inches of solids in the bottom of the east transit line, but that line does not appear to be corroding.
Also the BP incident investigation report says under-deposit corrosion is usually a longer term corrosion issue and by itself would not be expected to suddenly increase and produce the (corrosion) rates observed.
Leak detection systemIn the wake of the transit line leak there has been much discussion about the lines leak detection system. The leak detection system measures the volumes of fluid entering each pipeline segment and the volumes of fluid leaving each segment. The system triggers an alarm if the volume measurements dont match up.
So why did the system not detect a leak of the magnitude of the transit line oil spill?
In fact the leak detection alarm did sound four times during the week before the spill was discovered, although the alarm did not go off during the two days prior to the discovery.
Both before and after the discovery of the leak BP interpreted the leak detection alarms as false alarms, Johnson said.
It was a false alarm, Johnson said. What we thought then and what we think now
is that that was a process upset.
Beaudo explained that a process upset consists of a sudden change in fluid flow, rather like the situation when you turn on a garden hose. The sudden pressure change in the system causes false readings in the metering at the ends of a pipeline segment. However, although a process upset can cause an apparent loss of fluid in one segment of the pipeline, the fluctuation in flow causes an equal and opposite volume discrepancy in the next pipeline segment. The resulting volume loss in one segment balanced by a mirror image volume gain in the next segment is exactly what was observed when the false alarms sounded in the GC-2 transit line.
Johnson said that the GC-2 facility also confirmed that process upsets had occurred.
In fact none of the investigators of the oil spill incident were able to find any evidence of the pipeline leak when they examined the data from the leak detection system, Johnson said.
Weve looked at our leak detection data and even in hindsight we cannot tell when the leak started, she said.
It now seems clear that the leak occurred slowly over an extended period of time. As a result, the leak rate was below the regulated and practical threshold of the leak detection system, Beaudo said. Beaudo also said that BP is now investigating the potential for developing a more sensitive leak detection system.
Next actionsAs part of its reaction to the Prudhoe Bay oil spill and in response to corrective actions required by the U.S. Department of Transportation, BP has started taking a series of steps to ensure that another pipeline leak does not occur. The company has done 2,000 inspections of oil transit lines at Prudhoe Bay since the leak, Johnson said.
Weve looked at all of the oil transit lines
none other has the same combination of factors
bacteria in the facility, low flow rate and low corrosion inhibitor carry over, Johnson said.
But BP has committed to ratchet up its corrosion prevention program by running maintenance pigs through its oil transit pipelines at regular intervals, by increasing the frequency of smart pigging and by injecting corrosion inhibitors into the lines in effect the company will apply similar corrosion prevention measures to its oil transit lines as it already applies to its water systems.
Going forward well pig out oil transit lines on a regular basis and well look at our pigging practices on all of our pipelines and say Does this make sense? Johnson said.
BP is also continuing with a $17 million upgrade to GC-2, to improve the handling of solids from viscous oil production, Beaudo said.
BP is also now using infrared heat detectors to improve leak detection on its pipelines.
We have already initiated weekly overflights with infrared, Johnson said, adding that the company has ordered handheld infrared detectors for security people to use on the ground.
But the company has yet to determine what to do about the segment of the GC-2 transit line where the leak occurred the pipeline segment remains shutdown, with GC-2 production partially restored through a bypass line.
We arent clear at this point whether were gong to be able to patch it up well enough to be able to smart pig the whole thing
or whether it will be permanently out of service, Johnson said.
BP has committed not to put the line back into service until the company, the DOT and the Alaska Department of Environmental Conservation are all confident that oil can safely flow through it, she said.
Pigging issuesDOT has served BP with corrective actions requiring maintenance pigging on the transit lines. DOT also requires smart pigging within three months, Beaudo said. DOT has granted an extension of the period within which BP has to do the maintenance pigging because of some practical problems in carrying out the pigging.
BP plans to do the smart pigging within the DOTs specified time frame. However, the smart pig run will require clearing of the lines of solids using the maintenance pigs. So, depending on when it is possible to do the maintenance pigging, the company may also have to apply for a time extension for the smart pigging.
Theres been a lot of good, steady, regular dialogue between our people and DOT weve had some very constructive conversations, Beaudo said, adding that BP remains in compliance with DOT orders and that DOT agrees with BPs plan of action.
The problems with the maintenance pigging result from the quantities of solids currently lying in the transit lines BP will have to remove the solids in stages, using a series of increasingly aggressive pigs.
And there are no separator facilities between the transit lines and the trans-Alaska pipeline. Consequently, the sudden removal of a large quantity of solid material from the lines would result in an abnormally large quantity of solids suddenly flushing into the trans-Alaska pipeline. BP is working with Alyeska Pipeline Service Co., the trans-Alaska pipeline operator, to assess the risks involved in moving the solids and to plan a course of action. One solution might be to install separator tanks to remove the solids before they enter the Alyeska facilities.
All of this takes time.
Id rather take the criticism on taking longer, and do it right, and thats what were doing, Johnson said.
Future strategiesJohnson went on to talk about BPs longer term strategy for the maintenance of its pipelines on the North Slope. With the potential for the export of natural gas from the North Slope, the company is talking about a 50-year future in Alaska, she said. Senior company management is encouraging people to think about the technologies, equipment and people that will be needed over that time frame.
And Johnson likened the Prudhoe Bay field to a Cadillac that is 29 years old and has 120,000 miles on the clock.
We can drive it some more, safely, she said. If were going to go another 200,000 miles, another 50 years from where we are now, then we need to think about not only the car but the people driving it.
In the past year the company has embarked on a program of hiring and training people in appreciable numbers, she said.
Weve also been on this journey, looking at the kit, the facilities themselves, Johnson said.
And when it comes to pipelines the companys long-term strategy translates to doing things in a bigger broader way, although the company has yet to work out the specifics of what this means, Johnson said.
You can count on us to not only do the reactive things weve done already the inspections, the additional inhibitor, maintenance things, the smart pigging, the solids removal, Johnson said.
(But) the response will be bigger and more connected than that.
And thoughts about the transit line spill, a month after the spill was discovered?
Johnson said that her feelings about the incident go much deeper than just regret, sadness or disappointment.
We know that a spill of that size is completely unacceptable. If I could find the best way of apologizing to the people of Alaska, so they get how we feel, thats what Id do, Johnson said. A spill of that magnitude just doesnt fit BPs values as a company.