It was my memory that it had been decades since a natural gas pipeline had ruptured due to external corrosion. To verify I reviewed a listing of pipeline failures in the USA and found what appears to have been the last documented pipeline corrosion related rupture.
1985 A 30 inch diameter gas pipeline operating at about 960 psi, weakened by atmospheric corrosion, ruptured, and tore out about 29 feet (8.8 m) of the carrier pipe, blew apart about 16 feet (4.9 m) of a 36-inch-diameter casing pipe, blasted an opening across Kentucky State Highway 90, and cut out a pear-shaped crater approximately 90 feet (27 m) long, 38 feet (12 m) wide, and 12 feet (3.7 m) deep near Beaumont, Kentucky. 5 people were killed in one home, and 3 injured. The fireball from the incident could be seen 20 miles away.(April 27, 1985)
1986 A 30 inch gas pipeline ruptured due to corrosion near Lancaster, Kentucky. 3 people had serious burns, and 5 others had lesser injuries. External corrosion made worse by difficulties of Cathodic protection in rocky soil was the cause. (February 21, 1986)
It has been over 26 years since a natural gas transmission pipeline has had a rupture due to external corrosion. The most recent external corrosion ruptures were both in Kentucky. The Beaumont Kentucky rupture was due to atmospheric corrosion inside a road casing under Kentucky State Highway 90 and the Lancaster Kentucy failure was on buried pipe that was insufficiently protected by cathodic protection. Cathodic protection cannot prevent atmospheric corrosion inside a road casing.
While I was working as Corrosion Specialist for a natural gas transmission pipeline, an emergency repair situation was found by a ILI tool inside a state highway casing next to a large farm house. The high pressure natural gas pipeline had a farm tap that supplied gas to the farm house. The pipeline passed through the front yard of the farm house and it was under the driveway to the house. The wall thinnnig reported by the ILI tool was significant enough that a rupture of the pipeline inside the casing was possible. The farm house was so close to the casing with the thin wall pipe that a rupture may have destroyed the farm house.
It became necessary to put the farm house family in a hotel while the replacement of the carrier pipe was being done as large pipeline construction equipment would be working in the house's front yard plus there would be no driveway. When the pipeline crew was close to pulling the carrier pipe out of the casing, the farmer's family arrived to find out how much longer they would have to live in a hotel. The farmer and his wife was talking to the pipeline foreman next to the highway and a little girl with her cute little dog was running around looking at all of the pipeline construction equipment. Several of the pipeline crew had suddenly the job of watching a little girl to make sure that she didn't get too close to something that could get her hurt. The girl and dog put the construction work on hold until she got back into the car with her parents and drove away to the hotel.
The bad pipe was pulled from the casing and the new replacement pipe installed. The pipeline was put back in service, the farm house driveway repaired, and the family moved back into their beautiful farm house. Since this was not a Sissonville WV like event where a rupture occurred that destroyed houses and damaged a major highway, it didn't make the national news. However, it happened even if it didn't make the news. For every natural gas pipeline rupture that occurres that makes the national news; there are many pipeline integrity repairs done that never makes the news.
Some generalized statements on the Sissonville WV pipeline failure since the investigation by the NTSB (National Transportation Safety Board) is still on-going and insufficient information has been released for factual statements.
Other than SCC (stress corrosion cracking) and internal corrosion ruptures of pipelines; it has been a long time since typical external corrosion has caused a pipeline rupture. Pipeline coatings, cathodic protection, CIS (Close Interval Survey), gas leak surveys, and ILI (inline inspection) has been very effective in almost totally eliminating typical external corrosion as a cause of pipeline ruptures. Therefore the news release from NTSB that this pipeline rupture was due to typical external corrosion and not SCC, Pre-1970 ERW pipe seam fatigue failure, or internal corrosion was a surprise. Most consider typical external corrosion now days to be only the cause of leaks.
Pipeline ruptures due to external corrosion outside of pipe seams are now rare. For a rupture to occur the corrosion pitting has to be both deep and long. Deep Corrosion pitting of short length will result only in a gas leak which is typically found by dead plants over the pipeline or a gas leak survey. Typically the pipe coating will limit the length of corrosion pitting to a size that only results in a gas leak.
Only if there is a significant coating failure will corrosion pitting be capable of having sufficient length with deep pitting to result in a pipeline rupture. ILI (Inline Inspection) tools care capable of finding corrosion pitting before a gas leak or pipeline rupture occur, but not all pipelines are designed for the passage of an ILI tool. Information released so far indicates that the pipeline that failed had not been upgraded for the passage of an ILI tool.
The following is required for upgrading a pipeline built before federal minimum pipeline safety rules required that all new pipelines be designed for the passage of ILI tools:
replace mainline gate valves with full opening ball valves that allow the passage of a ILI tool
replace pipeline bends that are too short radius for the passage of ILI tools
replace unbarred tees with barred tees that may not allow the passage of ILI tools
remove any known pipeline feature that may damage or not allow the passage of an ILI tool
run a caliper, plate tool, or geometry tool that will verify the pipeline is ready for the running of an ILI metal loss tool or find features that need to removed before running a ILI metal loss tool
For pipelines that cannot be inspected by ILI tools, ECDA (External Corrosion Direct Accessment) with ICDA (Internal Corrosion Direct Assessment) is performed. Since the pipeline that failed was downstream of a gas storage field, odds are an internal corrosion threat existed on the pipeline that would have required ICDA. The pipeline per news releases had a ERW (electric resistance weld) seam which would have been a pre-1970 ERW seam per the construction date of this segment of the pipeline. Since this type of seam has had service issues due to quality control issues during manufacturing, the pipeline would have had a manufacturing defect threat which would require evaluation to determine if this threat was stable or not stable. Typically, manufacturing defects like Pre-1970 pipe seams are stable and the rupture of this pipeline was outside of the ERW seam.
Per the NTSB investigation so far, the failure was due to external corrosion on the bottom of the pipe that resulted in significant wall thinning. The wall thinning was described as being approximately 6 feet long with as much as 70% wall thickness reduction is some locations. Wall thinning of this length and depth is sufficient to result in a rupture of the pipeline. There has been nothing stated to date that internal corrosion on the bottom of the pipe may have contributed to the wall thinning that resulted in a rupture of the pipe.
External corrosion does favor the bottom of the pipe due to a oxygen concentration cell effect that makes the pipe bottom anodic to the rest of the pipe surface. The bottom of the pipe has less oxygen in the soil next to the pipe surface. This corrosion cell can result in deep corrosion pitting which on large diameter pipelines may not be detected by over-the-line-survey CIS (Close Interval Survey) as such surveys record an average pipe-to-soil potential of the pipe surface and not the lowest pipe-to-soil potential on the pipe surface. In addition, over the line surveys tend to not detect corrosion cells on the bottom of the pipe as effectively as corrosion cells are top of the pipe. Corrosion cell activity under disbonded coatings can decrease the effectiveness of over-the-line surveys in detecting possible corrosion activity. MIC (microbiologically influenced corrosion) may accelerate the oxygen concentration cell corrosion cell activity by deplorizing the pipe surface, increasing the amount of cathodic protection current required to create a polarized pipe surface at coating holidays, and requiring an increase in the cathodic protection criterion to a value greater than -850 mV. From photos of the failure site, MIC probably was not a contributing factor in the pipeline failure but cannot be totally ruled out.
ACVG (Alternating Current Voltage Gradient) and DCVG (Direct Current Voltage Gradient) pipe coating surveys may sometimes detect coating damage on the bottom of a large diameter pipeline that a typical CIS inspection may miss due to the average pipe-to-soil potential effect. All methods of finding coating damage that may result in corrosion damage to the pipeline have limited effectiveness if the pipe coating has significant disbondment to the pipe surface with trapped water and few coating holidays. All pipeline coating surveys tend to detect coating holidays more effectively than disbonded coating. The best way of detecting disbonded pipe coating is a ILI EMAT tool.
The photos of the failure site suggests that the backfill around the pipe that failed may have had sufficient rock and stone to damage coal tar and asphalt pipe coatings used at the time the pipeline was built. The amount of clean backfill placed around the pipe during construction would influence how much coating damage may have occurred during backfilling. Cold Applied Tapes were not in much use when the pipeline was built and probably was not the coating system applied to the pipe. Asphalt pipe coatings were popular during the period the pipeline was built and tend to have coating disbondment issues which may result in shielded corrosion cell activity. At this time the pipe coating system on the pipeline that failed has not been stated. The condition of the pipe coating where the failure occurred will be impossible to know absolutely as the explosion and fire totally destroyed it; but the condition of the pipe coating near the failure may give some indication. The amount of corrosion damage to the pipe joint(s) that failed will give some idea of coating condition too.
Highway salts used to melt ice can cause chloride concentration to increase in the soil near the highway resulting in accelerated corrosion of buried steel. Since the failure was near a major highway, this may have been a contributing factor.
The casing under the highway if shorted to the carrier pipe may have caused a depression of cathodic protection levels on the pipe that failed. The electrical isolation of the casing to the carrier pipe is presently unknown. Casing use started due to the oxygen concentration corrosion cell effect of the pipe under the highway being more anodic than pipe by the highway. Corrosion failures under highways and railroad crossing caused the requirement of cased pipeline crossings of roads and railroads. Modern construction practice is to use a high performance coating system for the pipe under the highway or railroad and eliminate the use of casings. In addition, pipeline crossings under railroads and highways are done deeper by directional boring which isolates the pipeline from stresses caused by railroad traffic sufficiently to eliminate the need for a casing for load shielding of the carrier pipe.
Desire to declare that I have not worked for Columbia Gas Transmission, but I did interview for the position of Corrosion Engineer in their Charleston, WV office. Discussed corrosion issues with Columbia Gas Transmission staff during the Purdue Corrosion Short Course and NACE International technical committee meetings. Consider their corrosion issues to be not unique to those I had to deal with during my career with other natural gas transmission companies.
I do not claim to be an expert in this new technology My knowledge limited to attending a Technical Toolbook class in Houston, Texas on this subject. I will post useful internet presentations on this topic as I find them in this blog posting.
I found this internet based presentation useful in my preparation for a smart pig presentation for the NACE International Los Angeles section. You can view it here.
When I worked for Chevron Pipeline, I attended a ILI (Inline Inspection) seminar in Sandi Arabia as a representative of Chevron. On the way back to the USA, I stopped in Karlsrue, Germany to visit the Pipetronix research center where crack detection smart pigs were developed for inspection of Russian pipelines. I believed that crack detection ILI tools would soon be required in the USA pipeline inspection market. Due to the recent Enbridge pipeline rupture in Michigan due to SCC (Stress Corrosion Cracking) and the questions raised about the accuracy of the smart pigs used on that pipeline, I decided to research how much crack detection technology has improved in Russia since my visit to Germany.
The information posted in this blog article comes completely from Russia. The national natural gas pipeline operator in Russia found TFI MFL tools sufficient to locate near neutral SCC colonies in their pipelines. The Argentina national natural gas operator used Russian TFI MFL tools to inspect their pipelines for SCC colonies and found the smart pigs unable of properly detecting and sizing classic (high pH) SCC colonies. The Argentina pipeline operator developed EMAT smart pigs for inspection of their pipelines. The Russians modified their TFI MFL tools to have Argentinian developed EMAT technology to improve detection and sizing of SCC colonies.
Attached is information on the existing Russian smart pigs for finding corrosion and crack defects in natural gas pipelines. The information is posted without comment.
On 24 June 2011, the California Public Utilities Commission had a "In-Line Inspection Symposium. The presentations given at this event can be viewed here.
Beyond all of the issues raised about ECDA in the San Francisco Bay Area newspapers after the San Bruno Incident; I will list some additional problems.
The problem of skips due to road crossings, road casings, and areas that are impossible to do a surface survey due to paved surfaces over the pipe. It was hoped that the government regulatoy agencies would approve of the use of guided long wave ultrasonic inspection to eliminate the skips due to road crossings but that has not happened. Some pipeline operators have dealed with the ECDA skips by running smart pigs. Smart Pigs as a means of eliminating ECDA skips is a unique solution since ECDA was designed for use on non-piggable pipelines. A new technology is robo-pigs (self propelled robot inspection tools). Since government regulatory agencies desire 100% inspection of HCA (High Consequence Areas), the ECDA skips become an issue in reaching compliance.
In paved areas, it may be too much trouble to drill holes in the paved street so that ECDA survey can be done. In addition, ECDA requires more digs to prove that the pipeline has no significant corrosion damage than a smart pig survey. As expensive as digs are to do in city streets, it doesn't take too many ECDA digs to justify upgrading the pipeline to run a smart pig to reduce the number of inspection digs required to prove pipeline integrity.
After San Bruno, manufacturing and construction defects that may grow and fail in service is a major issue. ECDA is not able to find such defects but smart pigs do have the ability to find some manufacturing and construction defects before a failure occurs. This issue will be covered in soon to be published blog articles.
ECDA can find external corrosion damage but not internal corrosion activity. A smart pig finds both internal and external corrosion damage. For pipelines with an internal corrosion risk, ECDA is simply not an acceptable technique. There is internal corrosion direct accessment (ICDA) but one has to have complete information on where all of the low points are on the natural gas pipeline where liquids can become trapped resulting in internal corosion activity. For some pipeline operators, a GeoPig tool may be the only way of knowing were all of the liquid trap locations may be located. It may be cheaper to upgrade the pipeline for the running of smart pigs rather than attempting to prove pipeline integrity using internal corrosion direct accessment.
If the pipeline has a SCC (stress corrosion cracking) risk, ECDA per current NACE Interational SCCDA (Stress Corrosion Cracking Direct Assessment) recommended practices is successful in finding areas of classic (high pH) SCC but not near neutral SCC. The only effective way of finding near neutral SCC colonies is at a minimum a TFI smart pig or hydrotesting with a EMAT smart pig being more effective per some pipeline operators experience in locating and sizing SCC defects. I plan on posting a more detailed blog article on this mattter soon.
Shielded corrosion cells are best found by smart pigs than ECDA.
I will probably add more items to this blog posting in the future.
Since the existing economy in California is difficult for a corrosion engineer trying to make a living doing water and waste water projects, I find myself back in the natural gas pipeline business. Currently doing pipeline integrity projects related to inline inspection (smart pigging).
Once upon a time, smart pigs were expected to find corrosion pitting and third party damage (dents and gouges), but not expected to see much construction or manufacturing defects. That was before recent events that now have smart pigs expected to do more integrity inspection tasks.
The pressure is to make smart pigs capable of accurately sizing defects that a majority of existing smart pigs cannot detect. This creates a market for improved smart pigs that will soon make many existing smart pigs obsolete.
I plan on posting some blog articles on various topics soon.
My current gig is a one year long contract job scheduled to end in March 2013. Naturally, I am looking around for my next gig. However, it doesn't appear I have to do all that much searching to get a new assignment somewhere.
