Did Faulty Welds Cause Pipeline Explosion?

Below is an exchange between me, a retired Steamfitter Instrument Tech, and a genuine Moon Shot Scientist, a PhD in the relevant area. We advocate two improvements in pipeline safety. I would install rupture discs at each end of each pipeline segment, and Doctor Liu would use precision meters, developed for space, to detect leaks by mass balance. The long seam that opened up was an awful weld, but if the control systems are not improved, we will continue to have Colonial, Bellingham, and San Bruno type incidents. Each of these, and many more I am sure, arose through design, operation or maintenance faults in the system controlling the valves on the pipeline. If, as is reported in the news, new regulations require automatic or even remote operating valves, the communication links are an added vulnerability. Bob Woolery, DC stateoftheartchiro.com miraclechiro.com 326 deAnza dr. Vallejo, CA 94589 707 557 5471 -------------------------------------------------------------------------------- From: safepipelines@yahoogroups.com [mailto:safepipelines@yahoogroups.com] On Behalf Of zumath@aol.com Sent: Tuesday, February 01, 2011 8:05 PM To: safepipelines@yahoogroups.com Subject: Re: [safepipelines] Bad design and slow instruments were important in San Bru... In shock tube experiments, the amplitude of the shock wave can be up to 10x that of the upstream pressure, and are typically of only several milliseconds duration before equilibration begins to take place. Pressure shock waves are typically studied in detonation/deflagration experiments. A great deal of research into the effects of pressure shock waves was done back in the days of the "cold war," typically using high response quartz pressure transducers with "Transductron" type electronic data recovery systems, which attempted to reconstruct the pressure wave by solving a second order differential equation for the change of the pressure using analog circuitry. An classical example of the devastating effect of gaseous shock waves precipitated the demise of natural gas vehicle (NGV) programs at SoCalGas. A naive mechanical engineer with no real experience in fluid mechanics nor thermophysics was put in charge of the refueling station program. He implemented the three bank rapid fill cascade using fast opening solenoid valves. This essentially subjected the fiber wound tanks to three "hits" as the cascade valves were opened. When I pointed out the shock wave problem to him, he was incredulous, even though I showed him a very specific "thou shalt not use fast opening valves at high pressure differentials" statement in the Air Force Design Handbook "System Safety," stating that their pressure gauges showed no such phenomena... Well, a receiving tank failure on an RTD bus subsequently occurred at the RTD depot during refueling, blowing up the entire bus. The RTD Citizens Advisory Committee requested my testimony as an "expert witness" at the hearing on the accident. I read excerpts out of not only the USAF Design Handbook, but also several textbooks on compressible fluid dynamics -- yet the SoCalGas engineer continued to maintain that "That's only Dr. Liu's theory..." However, SoCalGas -- clearly a "deep pockets" organization -- subsequently withdrew as the lead organization for promoting NGVs. This eventually led to the discontinuation of factory NGV programs by all of the USA manufacturers... All because one naive "engineer" wouldn't go to the trouble of implementing compressed gas fueling cascades using controlled opening valves to ramp up the flowrate into vehicle receiving tanks... I personally still believe that NGVs have great potential, given our abundant supplies of natural gas -- if fueling is properly implemented. However, here we have seen the devastating effects of sudden pressure spikes in compressible media. In a message dated 1/31/2011 7:19:30 P.M. Pacific Standard Time, drbob@comcast.net writes: The pressure spike that was the proximate cause of the rupture was likely a great deal higher than the 386 psi reported, apparently by PG&E. Remember that the precipitating event was a UPS failure which allowed a pressure regulating valve feeding from an unspecified higher pressure to flop open. The sudden change in flow caused a shock wave to run down the pipeline until it found the defective long seam. The peak pressure of this shock wave could well have been higher even than the pressure in the transmission line feeding the valve. The instruments used to monitor pressure in these applications cannot detect fast transients. An industry workhorse, the Rosemont 3051 pressure transmitter specifies a minimum of 100 milliseconds to register a full scale change of pressure. Recorders, still mostly recording on rolls of paper, are even slower. The design of the control loop for the regulating valve was faulty. Such valves can be configured for various responses to a loss of power and/or air pressure. They can be set up to open, close, or stick where they are in the event of a supply loss. Clearly, the valve in question was set up to open on power failure. This is a major design error. It is truly a surprise that there were not storage batteries on the 24 volt instrumentation supply, but it may be that it was exactly this feature of the system that was being worked on. Events of this nature can be protected against, but regulatory agencies have not required the installation of the appropriate protective devices. The spikes from rapid valve opening are too fast to control with mechanical pressure relief valves, but several manufacturers offer rupture discs, precision "fuses" that can positively insure that the pressure in the pipeline NEVER exceeds the specified pressure. I would require the installation of such a device at each end of each pipeline segment, guaranteed to open at a pressure no greater than the qualifying hydrotest pressure. Provision for these could be made when installing pig traps and hydrotesting. I am a retired Steamfitter instrumentation technician, with some engineering background. I grew up in a trailer house with my father, a specialist in welding on these large diameter gas pipelines and organizer for the Pipeliner's union. I have retrained as an upper cervical Chiropractor, but continue to be interested in pipeline issues. For a better credentialed source, contact Arnold Liu, zumath@aol.com. He is a bona fide PhD rocket scientist, responsible for the successful metering of fuel in the moon landings. We have been in contact on the safepipelines discussion list, and he has given his permission to make referrals such as this. Please stay on this story, as around half the pipelines in the US are from before 1950 or of unknown age, and the hazards are real. The design and operations issues uncovered here are even more important than this particular long seam failure, tragedy though it was.