PIPELINE CROSSING FAULT LINE

[eureka]

I need some guidence on how to cope with the movement in a pipeline which transverses a fault line. We were given the movement as approximately 1 inch per year. References and personal experience will be appreciated.

 

[rconner]

I think many publications and/or research programs have discussed installations of piping in geologically hazardous areas, including fault crossings. While you have not mentioned your specific utility application, I remember e.g. one such paper was presented by then DIPRA Engineer, Mr. Duane Ford, "Effects of Ground Failure on Ductile Iron Pipe" at the 1988 ASCE Int'l Pipeline Division "Pipeline Infrastructure" conference in Boston, MA. I think I have heard several design suggestions in this or other papers over the years, including various concepts of "trench softening" (including "crushable" surrounding backfill such as foams etc.), making the crossings at transverse angles to the fault as opposed to directly perpendicular to same, use of loose polyethylene encasement or even "double" loose polyethylene encasement in such areas to provide for minimal stressing of the pipe by ground movements, and/or also the use of special flexible restrained joints and coupling devices avaialble for such pipes that can even provide for substantial relative movements between individual pipes if this is necessary (I guess all to at least in theory minimize stresses in the piping as ground masses move). An example of such a joining mechanism is a "ball joint", that may allow e.g. for a little axial movement in or out and also as much as 15 degrees of joint deflection between pipes (I believe this amount of joint deflection translates by the trigonmetry to several feet of lateral "off-set"/from the alignment of the preceding pipe between e.g. 20 feet long individual pipes!) I think also some contemporary researchers have looked at the strengths of various common piping materials and joining mechanisms with regard to static and dynamic pushing and pulling etc. motions (e.g. see some such test results at

http://bric.ce.unr.edu/networking/Pipest/Test_Results/tables.html

-- I believe the ductile iron pipe joints referred to generically in this Table 1 as "bell-spigot, retaining ring joint" and exhibiting the substantial tension test results shown are AMERICAN Flex-Ring pipe joints).
I believe there have been numerous test and/or perhaps other installations of ductile iron piping across known fault lines over many years, and I suspect most crossings at least in recent years have employed at least some special provisions by the designers. I am not aware of any specific results of any such installations (other than I have not heard of any problems where provisions such as multiple ball joints in the area etc., have been employed). For more/better information you might wish to contact the Ductile Iron Pipe Research Association (DIPRA / dipra.org). I suspect a very strong piping with some significant flexibility built into the design may be advantageous for minimal risk, and particularly in known fault areas.

 

[rconner]

Subsequent to my post above I attended the 2005 ASCE Pipeline Division Specialty Conference that was held several weeks ago in Houston, Texas. Mr. Doug Nyman was chosen as the recipient of the 2005 Stephen D. Bechtel Pipeline Engineering award, and as is customary with this award he delivered the "Bechtel Lecture" associated with this award. Mr. Nyman chose as the subject of this lecture fault crossings etc. of energy pipelines, with which he appeared very familiar. His cover slide and some other slides in his presentation I attended were interestingly of an aboveground fault crossing design area for the Alaska Pipeline, that was I think was designed/constructed now more than three decades ago. He let us know in his lecture that an area of this pipeline where it was exposed crossing the Denali fault reasonably endured, if I understood him correctly, several meters of ground displacement in a rather recent extreme (sesqui “hundred year"?) earthquake event a quarter of a century later, in the magnitude 7.9 2002 Denali earthquake (see some more information at

http://www.aeic.alaska.edu/Seis/Denali_Fault_2002/,

that indicates this earthquake is the largest earthquake to hit North America in 150 years). I believe this fault crossing area incorporated some specially designed, quite wide “Vertical Support Member structures (VSM’s), as well as some intentional pipeline zig-zags or bends to accommodate thermal expansion/contraction and future movement. It appears a detailed account of this design and performance is also now at

http://www.alyeska-pipe.com/Inthenews/techpapers/3-TAPS AboveGround Denali EQ.pdf.

This site appears to indicate some pipeline “damage” noted in the first site I referenced above apparently consisted of some cross-beams being dislodged, bracket bolts broken, and VSM’s bent some laterally etc. by the reported “violent” ground movement in the event etc. Reportedly, the pipeline was shut down for only 66 hours as a precautionary measure and to accomplish support etc. repairs!

While I don’t think noted in Mr. Nyman's lecture, at the time this was perhaps one of the greatest if not the greatest pipeline project in history, and perhaps most significantly as noted in multiple recent publications, "More than 14 billion barrels of oil have moved" through this pipeline. I think the carrier pipe for this project was produced in Japan and American Steel Pipe (a division of ACIPCO) furnished significant other piping material for some 78,000 (most if not all) of the VSM’s for this massive pipeline project.

I believe Mr. Nyman also mentioned in his presentation that some other contemporary fault crossings at least of some more recent buried energy pipelines are now being designed buried in special wide and rather shallow trapezoidal cross-section trenches, apparently with some special (uncompacted) backfill.

 

[stanweld]

eureka,
Was the 1" per year average rate of displacement based on a creeping fault or based on a 100 year average, where primary displacements occur at significant siesmic events. I previously worked for a gas utility that had numerous pipelines crossing numerous faults, including the San Andreas. Only conventional pipeline design was used in their construction. In significant seismic events, pipe near the epicenter was unearthed and invesigated for damage. In one case nearly two miles were unearthed and the pipe allowed to spring back (up to 5 feet)into place to remove lateral earth loadings.

It is noted that as of 1980 when I left that company, no electric arc welded pipe had ever failed during a quake, although numerous failures occurred in oxy-fuel gas welds due to their inherent brittle nature.