Pipeline Whistleblower: Cracks in The System
Insider ties poor weld inspections to rising rate of ruptures. Part two of a Tyee investigation.
Evan Vokes, a 46-year-old Calgary pipeline engineer, is a man with a mission, and a conscience.
While building natural gas pipelines in Canada, Mexico and the United States for TransCanada Corporation, he started raising concerns about industry practices.
Vokes had an important inside job: he was the guy that ensured the pipelines were constructed safely.
His specific duties included metallurgy and welding. He also specialized in an important accountability process known as non-destructive examination. And he didn't like what he was seeing.
At the invitation of Russ Girling, TransCanada's CEO, Vokes provided documents to senior executives of the company (it is Keystone XL's controversial sponsor) that allegedly documented systemic failure to follow code and regulations in 2011.
Shortly afterwards, the engineer lodged a complaint with regulators in Canada and the U.S. Last May TransCanada fired the engineer without cause.
Drawing on examples from the records of Enbridge and Kinder Morgan (the CBC is investigating TransCanada's record) Vokes is going public with his concerns about an industry facing unprecedented growth and what even the National Energy Board (NEB) describes as "an increased trend in the number and the severity" of pipeline incidents.
Vokes has stellar company. In particular, the U.S. National Transportation Safety Board (NTSB) has accused Enbridge, a Canadian company jointly regulated by the NEB and the U.S. Pipeline Hazardous Materials Standard Administration (PHMSA), of nurturing a "culture of deviance" on safety and integrity issues after a dramatic Michigan pipeline rupture in 2010. That debacle caused the largest and most expensive onshore oil spill in U.S. history.
Moreover, a lengthy 2008 audit of the company by the National Energy Board documented similar flaws two years before the event.
It found that company was not upholding the rules and regulations on pipeline integrity and safety in Canada either.
"Enbridge's integrity management program for pipelines and facilities do not meet some of the provisions required by" Onshore Pipeline Regulations and CAS-Z662 Oil and Gas Pipeline Systems, said the extensive audit which the NEB did not make public at the time.
In addition to "multiple findings of non-compliance and non-conformance" with regulations, the NEB also documented that Enbridge didn't have a process for "defining and evaluating the level of qualification and competence of contractors and consultants."
The company also didn't know how valid and effective its assessments of corrosion and cracking were in its pipeline safety program.
As a result the NTSB concluded that Enbridge's Michigan spill was partly the result of weak regulations, weak enforcement and a corporate disregard for learning from past mistakes.
In attempt to catch-up with events, the National Energy Board released a discussion paper on pipeline safety that pointedly echoes the very issues raised by Vokes.
The paper says that "accident prevention requires active leadership by management on safety issues" and adds that "there must be effective implementation of the right controls to manage, mitigate or eliminate hazards and risk."
'Someone is going to die'
It's exactly these kind of problems and accountability failures that Vokes is now trying to highlight as Canada prepares to double its pipeline capacity with controversial bitumen and diluent highways across the continent.
"Someone is going die and they just don't know it yet," explains Vokes, a large, intense and careful man who spoke to both the Tyee and the CBC over the last several weeks.
He's also filed his concerns and allegations with the National Energy Board, the Association of Professional Engineers and Geoscientists of Alberta (APEGA) and U.S. Pipeline Hazardous Materials Standard Administration (PHMSA). Documents have also been sent to the office of the prime minister.
The NEB told the Tyee that the board is taking the allegations and complaints made by Vokes seriously and is investigating them. In contrast, AGEGA, a self-regulating professional body, did not answer two separate queries from the Tyee.
"My motivation is to prevent unnecessary death and environmental damage," adds the engineer who has also been a welder and millwright.
"The controls for the industry are there but they are not being implemented or enforced. We have the technology to do things right, but we don't have the willpower."
Adds Vokes: "The pipeline industry must take accountability for the true safe construction of pipelines rather than a risk based approach based on faulty data sets on threats to integrity."
Risk-filled enterprise
Pipeline builders depend on high quality steel, careful engineering, expert welding and competent safety programs that are all subject to a variety of strains and stresses including commercial pressures to get pipe in the ground as fast as possible.
Dense professional jargon, detailed engineering codes and intricate metal science often make pipeline construction and integrity "a difficult subject to understand," adds Vokes.
"The public has little protection from engineering decisions on pipelines, whether or not they are made by professional engineers," says the engineer.
The most critical issue is not what companies do after a pipeline has been built, explains Vokes, but the quality of materials, welding and inspection performed during the construction.
In fact, the near doubling of pipeline incidents on Canadian pipelines (from an average of 95 to 161 in 2011) in some ways mirrors British Columbia's leaky condo scandals.
Several codes now govern the construction of pipelines carrying hydrocarbons in North America, including the American Society for Mechanical Engineering B31.4 and B31.8 and the Canadian Standard's Association Z662 also known as Oil and Gas Pipeline Systems.
These codes are good says Vokes, but "do not contain a blanket statement for permitting a violation when a company is in a hurry. Those violations happen everyday in this town. But there is no 'get er done' clause."
Case examples: Cracks in the system
In 2008, Enbridge built a 504-kilometre long oil pipeline from Cromer, Manitoba to Clearbrook, Minnesota called Southern Lights.
Shortly afterwards, the National Energy Board, which oversees the safety of interprovincial pipelines, heard about numerous welding quality problems along the pipeline.
"Given the potential systemic nature of defects associated with pipe manufacture and pipe field joining" an NEB letter asked Enbridge for more information about the cracks popping up in its girth welds, a growing epidemic throughout the industry.
A girth weld joins the individual sections of the pipe. If it is not done properly it can break or crack either during construction or later, resulting in leaks and ruptures. PHMSA flagged the problem with a major advisory in 2010.
Enbridge replied to NEB's request for more information with an unsigned report on girth weld cracks. The four-page document noted that there were 21 cracks and two hydrostatic testing failures in the Southern Lights pipeline on the Canadian side of the project as well as cracks in the U.S. portion. (Hydrostatic testing fills a pipeline with water under high pressure and is a rudimentary way of determining if a pipeline will rupture in service.)
Enbridge's anonymous 2009 report (like any professional group such as doctors and lawyers, engineers must authenticate and validate documents by signing them) explained that the cracking problems "occurred in high wind chill conditions brought about by ambient temperatures combined with strong Prairie winds." It added the pipeline had been built according to code and duly repaired "with best welding practices."
But Vokes says that's probably not the whole story as pipelining is an outdoor endeavour. A properly supervised welding operation takes the weather conditions into account and modifies welding procedures accordingly. "If you have a high repair rate on a pipeline then you are not following proper welding procedure," he explains. "Pipeline welding is a manufacturing process on the move."
Implementation is everything in this business, adds Vokes. "Quality plans count. If you don't make your welders follow the specified plan, you have a fuck-up."
Industry experts as well as a 2002 paper on the integrity of pipelines make exactly the same point: "Cracking in pipelines is not usually a defect assessment problem; it is usually an indication that operation, product or environment is a major problem."
In fact a natural gas pipeline (Rocky Express) built by Houston-based Kinder Morgan across the Great Plains in 2007 and 2008 experienced endless repair work due to shoddy welding practices and commercial pressures to get the pipe in the ground.
In 2012, PHMSA fined Kinder Morgan, which wants to expand its TransMountain pipeline across the Canadian Rockies, nearly half a million dollars for 13 specific violations of pipeline construction codes and regulations. (The NEB currently has no system for fining companies that violate regulations but has proposed one.)
The list of Kinder Morgan's transgressions is long.
According to PHMSA, Kinder Morgan did not follow quality welding procedures properly; nor did it perform welding "in accordance with proper procedures." It also "did not adequately inspect the welding." In addition, the company failed to prevent damage to pipe while backfilling trenches. Nor did the company remove defects in the pipe properly. It also didn't use the properly designed pipe along one section.
But poor welding isn't the only cause of cracked pipe in the industry. External and internal corrosion play major roles as does dented and damaged pipe. The National Energy Board now reports that nearly 30 per cent of all pipeline failures Canada are due to cracking.
Non-destructive examination
During the construction of a pipeline, inspectors must confirm and validate a number of procedures to ensure the integrity of the welds on an ongoing basis.
Manual welds with a cellulosic rod are common for pipes going up and down steep slopes. But a bad weld, say, at the top (12 o'clock position at the start of the weld) or at the bottom (six o'clock position) on a high strength steel pipe made by a cellulosic rod, can cause what the industry refers to as delayed cracking, cold cracking or hydrogen cracks.
Hydrogen, the first element on the periodic table, can migrate in solid steel to an area of stress at warm temperatures. When the steel cools, the hydrogen can get stuck and cause delayed cracking. It has long been a major issue in pipeline and building construction around the world.
To check for such cracks the industry uses a variety of different tools after the weld is completed. (In engineer jargon, "non-destructive examination" (NDE) checks the quality of pipeline welds and materials without damaging them.)
Or as Vokes puts it: "Welding determines the speed of construction and NDE holds it hostage."
The primary tools include radiography (it looks for defects in pipe density with gamma rays); manual ultra sonic, which looks for defects by sending a signal into the pipe with a fixed angle probe; or automatic ultra sonic (AUT). It uses sophisticated probes that look at the pipe from many angles.
Of the three tools AUT is the most effective for scanning the whole pipe and identifying the nature of defects and validating the integrity of the weld. "With AUT you can inspect any pipe wall, a quarter inch or thicker. It's the best."
Hugh Kerr, a retired Waterloo professor who specialized in welding processes and metallurgy and has trained some of Canada's top pipeline engineers, agrees.
"Automatic ultrasonic testing is much better at picking up defects from different orientations and lots of people are saying that."
Yet it costs more and can slow down the pace of construction. "Industry would rather get the pipe in the ground and deal with the problems afterwards," explains Vokes.
When industry opts not to use best practices such as AUT, the reasons can be accounting or scheduling decisions.
Kinder Morgan's Rocky Express pipeline again makes a perfect example. Although Kinder Morgan did employ AUT for the main line of the pipe, all repairs were inspected with a less reliable method that didn't catch hundreds of cracks.
The following eyewitness industry email obtained by the Tyee documents the scale of the negligence (hundreds of digs to repair the line) as well as the inadequacy of the company's welding inspection system.
"The repair process was set in place that a 24 hour delayed inspection was not necessary due to the metallurgical properties of the welding rod that was supposed to be used. Repair welders on the pipeline thought it was much easier to weld using Cellulosic welding rods and made the executive decision themselves to go that route not realizing the ramifications of their actions. Obviously the client was not aware of the change in the procedure and therefore did not mandate a post 24 hr inspection requirement and when it came time to Hydro-Test there were multiple failures all propagating from the repair areas due to post 24 hr cracking caused by the use of Cellulosic welding rods without proper inspections. All repairs done using the Cellulosic rod had to be excavated and re-repaired or cut-out causing major delays and massive financial repercussions also giving a black eye to the industry."
The National Energy Board issued a safety order against the company and its TransMountain pipeline in 2011, saying its inspection system for cracks was inadequate. Moreover, its chosen system was "not presently an accepted practice by industry."
Northern Gateway inspection issues
Similar issues dog the proposed Northern Gateway project. Enbridge, for instance, has told the National Energy Board that it may or may not use radiography and a process technically known as sectoral scan ultrasonics to check for bad welds.
The technology resembles the ultrasound machine used to scan fetuses in the womb: "Just as determining a baby's sex can be as problematic, finding a crack with a sectoral scan is more difficult than AUT -- the resolution for ultra sonic isn't that good," explains Vokes.
Although gamma radiography is a traditional technique in pipeline construction, it is also known as "the welder's friend" because of its inability to pinpoint defects. "You can see them but you don't know what they are." Companies proposing technically difficult projects in challenging northern geographies should not be allowed to use radiography as their only examination tool, says Vokes. Nor does he think regulators should allow only radiography to test welds.
A second problem with non-destructive examination concerns conflict of interest. An independent third party should do it, and not the contractor hired by the company to weld the pipe.
Yet in many cases pipeline companies don't hire an independent contractor to inspect the welding procedures or weld quality during the construction process. Alberta's provincial regulation, for example, does not prohibit contractors from hiring all inspection in house -- an inherent conflict of interest.
However federal regulations require all non-destructive examinations under Construction of Pipelines CSA Z662 to be performed "by an independent contractor retained by the company." (Yet the same federal regulation allows the companies to set the techniques and qualifications of inspection, adds Vokes.)
There are other problems with the state of non-destructive examination. The skill requirements under some codes for inspectors allows for inexperienced people to conduct and interpret welding inspections. After a short course, just about anyone can write a qualifying test in pencil, explains Vokes and he doesn't think that sort of training is rigorous enough. "It's another inherent conflict of interest in the whole process."
The failure of in line inspections
Industry lobbyists claim that safety programs routinely protect the public and the environment from pipeline failures with scheduled in-line inspections that might include x-rays, machines called pigs, special monitors or aerial surveys.
Yet companies fail to detect problems on their pipelines all the time, says Vokes.
A recent review of U.S. pipeline data by InsideClimateNews found that 19 out of 20 pipeline leaks aren't detected by remote sensing systems. Landowners and employees at the scene of ruptures report the majority of all incidents.
The U.S. National Transportation Safety Board also makes a similar point in its sweeping critique of Enbridge safety practices on Line 6B prior to the Michigan bitumen spill.
Reports the NTSB: "Despite their sophistication, the detection capabilities of in-line inspection tools have limitations. Each tool technology has a stated minimum defect size that can be detected and the tool can be subjected to interference from nearby anomalies or geometry."
As a consequence pipelines routinely fail and crack without warning.
In 2007 Enbridge's Line 3, for example, ruptured, spilling oil near Glenavon, Saskatchewan.
Both the company and the NEB had identified the line as susceptible to cracking because of 10 major incidents from seam failure to metal fatigue on the line since 1989.
Nevertheless a 2002 field inspection and a 2006 ultrasonic test failed to detect the anomaly responsible for the 2007 rupture because it "was not identified as requiring immediate repair since its depth was in the 12.5 to 25 per cent range" of wall thickness. (Cracks grow to become like half ellipses, before fracture occurs.)
Investigators concluded, "It is apparent that a degree of uncertainty can exist during the non‑destructive examination of the pipe in the field" while it is in service.
In 2009 Enbridge's Line 2 crude oil pipeline from Edmonton to Cromer, Manitoba, also cracked and spilled oil. A landowner found the leak thanks to a strong petroleum odor downstream of the Odessa pump station.
Despite a battery of inspections in 1998, 2004, and 2008 for cracks and metal loss and other problems, Enbridge missed the dent and gouge that resulted in the crack.
In fact, the dent that eventually turned into a crack had escaped detection since the pipe was laid in 1953 until 2004. Noted a 2009 incident report by the Transportation Safety Board of Canada: "The dent was detected during the 2004 metal loss in-line inspection. However, the metal loss associated with the gouge was below the detection limit of the ILI tool and the location was not flagged as a dent with secondary damage. No follow-up was performed after the 2004 inspection as the size of the dent was within company specifications. In 2008, the dent was again detected during the geometrical discontinuity in-line inspection. Again, as it did not meet Enbridge's damage reporting criteria, it was not flagged for further follow-up."
In 2011 an Enbridge pipeline carrying oil from Norman Wells to Alberta cracked and leaked more than 1,000 barrels of oil. The leak was due to a small crack on a girth weld on the pipeline and is currently under investigation. The Board ordered the company to conduct aerial surveys of the pipeline as well as reduce the pressure at which oil is pumped through the line.
NEB 'knows' inspections miss defects: Vokes
Vokes says the Norman Wells leak not only represents another failure of in-line inspection but also highlights the difficulty of keeping pipe safe in challenging terrain and climate.
"Does the National Energy Board know that in line inspection misses defects? It knows," says Vokes. "In line inspection won't take care of all latent defects found near girth welds."
In other words the best defense against leaks, cracks and ruptures lies in proper welding procedures. Next comes an immediate and rigorous inspection process enforced by transparent regulators, says Vokes.
Given that the pipeline industry proposes to double the nation's pipeline capacity over the next decade, the issues raised by Vokes deserve careful regulatory and political scrutiny.
"The engineers are doing the best they can. But the companies push them to get the pipe down by so many kilometres a day," adds Kerr, the retired welding expert.
"We are all shareholders in this process and want larger returns and that happens by cutting costs in the industry. It seems to me that this process can be a race to the bottom unless we make sure that the pipeline companies fabricate and properly maintain well designed pipelines."