{"operation":"document","citation":"12-0121","title":"National Transportation Safety Board — Hazardous Materials Safety Interpretation","source_type":"guidance","agency":"Pipeline and Hazardous Materials Safety Administration","status":"guidance","official":true,"published_on":"2012-08-16","effective_on":null,"summary":"12-0121 response to National Transportation Safety Board concerning 177.840, 180.416.","machine_formats":{"json":"https://regulus.evalyn.ai/document/phmsa-interpretation-12-0121.json","markdown":"https://regulus.evalyn.ai/document/phmsa-interpretation-12-0121.md"},"app_url":"https://regulus.evalyn.ai/document/phmsa-interpretation-12-0121","source_url":"https://www.phmsa.dot.gov/sites/phmsa.dot.gov/files/legacy/interpretations/Interpretations/2012/120121.pdf","body":"<<<PAGE 1>>>\n\nU.S. Department\nof Transportation\nPipeline and Hazardous\nMaterials Safety\nAdministration\n1200 New Jersey Avenue, SE\nWashington, D.C. 20590\nAUG 16 2012\nThe Honorable Deborah A.P. Hersman\nChairman\nNational Transportation Safety Board\n490 L'Enfant Plaza, SW\nWashington, DC 20594\nRef. No.: 12-0121\nDear Chairman Hersman:\nThank you for your May 4, 2012letter requesting clarification of the inspection and\nmaintenance requirements for the discharge system on a cargo tank motor vehicle in\nliquefied compressed gas service under the Hazardous Materials Regulations (HMR; 49\nCFR Parts 171-180). Specifically, you request clarification of the meaning of\"in service\"\nas it applies to the monthly inspection of a delivery hose assembly (i.e., \"a cargo transfer\nhose assembly\") under§ 180.416(d)(l).\nAs part of an investigation of an accident involving the rupture of a hose used to transfer\nanhydrous ammonia, the National Transportation Safety Board (NTSB) included among its\nfindings the belief that lack of clarity of§ 180.416(d)(l) regarding monthly inspections of\neach \"in service\" hose assembly can lead a cargo tank motor vehicle operator to mistakenly\ndefer monthly inspections.1 The NTSB concluded that \"if a motor carrier does not believe\nthat [a hose assembly] will be used, it will not necessarily consider it to be 'in service' and\nas a result, the hose assembly may not be inspected, as was the case in this accident.\" It is\nthe understanding ofNTSB that a hose assembly carried on a cargo tank motor vehicle has\nthe potential to be used on any given day, and consequently, is \"in service.\" Therefore, in\norder to provide greater clarity of the monthly inspection requirement of a hose assembly\ninstalled or carried on a cargo tank, in its Safety Recommendation H-12-5, the NTSB\nrequests that PHMSA provide an interpretation of when a hose assembly is \"in service.\"\nYour understanding is correct. A hose assembly installed or (to be) carried on a cargo tank\nmotor vehicle in liquefied compressed gas service must be inspected monthly regardless of\nwhether it is used in any given month. Although a hose assembly must be checked prior to\n1 Hazardous Materials Accident Summary Report: Cargo Hose Rupture and Release of Anhydrous Ammonia\nDuring Offloading of a W emer Transportation Services Cargo Tank Motor Vehicle at the Tanner Industries\nPlant, Swansea, South Carolina, July 25, 2009.\n\n<<<PAGE 2>>>\n\neach unloading operation, this check involves only those components readily visible (see\n§ 177.840(m)). The requirement of§ 180.416(d)(l) ensures that at least once a month each\nhose assembly assigned to a cargo tank motor vehicle in liquefied compressed gas service\nwill undergo a thorough visual inspection whether or not it has been used. \"In service\" as\nused in § 180.416, is associated with qualification and maintenance for purposes of\nhazardous materials transportation and should not be misconstrued as \"in use.\" Just as a\ncargo tank in liquefied compressed gas service is subject to inspection and testing\nrequirements and must be removed from hazardous materials service (i.e., placed out of\nservice) if it fails a test or inspection, a hose assembly must be inspected and tested, and\nremoved from service according to rejection criteria found in§ 180.416. A repaired hose\nassembly successfully retested in accordance with§ 180.416(±) may be placed back in\nservice and then must again be inspected monthly regardless of whether it is used in any\ngiven month.\nIf I can provide further information or assistance, please feel free to contact me.\n\n<<<PAGE 3>>>\n\nW1nter-\n~ 1 ~o, ttl to(J)\nCarao Tun~ .\n\"'' IZ---012.{\nNational Transportation Safety Board\nWashington, D.C. 20594\nSafety Recommendation\nDate: MAY 0 4 2012\nIn reply refer to: H-12-2 through -6\nThe Honorable Cynthia L. Quarterman\nAdministrator\nPipeline and Hazardous Materials\nSafety Administration\nWashington, D.C. 20590\nOn July 15, 2009, about 8:00a.m., a cargo transfer hose ruptured shortly after transfer of\nanhydrous ammonia began from a Werner Transportation Services, Inc. (Werner) cargo tank\ntruck to a storage tank at the Tanner Industries, Inc. (Tanner) facility in Swansea, South Carolina.\nA white cloud of anhydrous ammonia, a toxic-by-inhalation gas,\n1 moved from the parking lot of\nthe facility across U.S. Highway 321 to a largely wooded area, where it eventually dissipated.\nAbout the same time, a motorist traveling north on the highway drove into the ammonia cloud,\napparently tried to get away from the cloud, then got out of her car and died of ammonia\npoisoning. Fourteen people reported experiencing minor respiratory problems or dizziness as a\nresult of the anhydrous ammonia release and were evaluated by emergency medical services\n(EMS) on scene. Of those 14, 7 displayed symptoms that required EMS to transport them for\nfurther evaluation at an emergency department; they were treated and released the same day. The\nanhydrous ammonia cloud caused temporary discoloration of vegetation in the area, including\nthe leaves on the trees. Residents in the area sheltered in place, and U.S. Highway 321 was\nclosed until about 2:00 p.m. on the day of the accident. The Lexington County Fire Service\narrived on scene about 8 :07 a.m. Property damage and losses were limited to the ruptured hose\nand about 6,895 pounds of the anhydrous ammonia that was released.2\nThe National Transportation Safety Board (NTSB) determined that the probable cause of\nthe accident was Werner Transportation Services, Inc.'s use of a cargo hose assembly that was\nnot chemically compatible with anhydrous ammonia. Contributing to the accident was the lack of\nexplicit requirements by the Pipeline and Hazardous Materials Safety Administration (PHMSA)\n1 The U.S. Department of Transportation (DOT) classifies anhydrous ammonia as a Hazard Class 2\nnonflammable gas. Anhydrous ammonia is a colorless liquid or gas that is both poisonous and corrosive and that has\nan intense, pungent, suffocating odor.\n2 See Cargo Hose Rupture and Release of Anhydrous Ammonia during Offloading of a Werner Transportation,\nInc. Cargo Tank Motor Vehicle at the Tanner Industries Plant near Swansea, South Carolina, July 15, 2009,\nHazardous Materials Accident Report NTSB/HZM-12/01 (Washington, D.C.: National Transportation Safety Board,\n2012) on the NTSB website at <http://www.ntsb.gov>.\n8391\n\n<<<PAGE 4>>>\n\n2\nthat the motor carrier and the facility carrier verify that the cargo hose assembly is chemically\ncompatible with the product to be transferred before transfer operations begin.\nAccident Hose Assembly\nThe ruptured transfer hose was manufactured by Durodyne, Inc} in 2005. The hose,\nDurodyne product number DD-G-220, was designed for liquefied petroleum gas (LPG) transfer\nonly and was constructed of several different layers of material that are chemically compatible\nwith LPG, but not with anhydrous ammonia. The innermost layer ofthe hose was made of nitrile\nrubber, followed by three sequential layers of polyethylene terephthalate (PET) yarn braids\nencased in chloroprene rubber and an outer layer of neoprene rubber. The PET yarn braids are\nused to mechanically reinforce the hose, and they provide the hose with the majority of its\nstrength. The hose had been approved by Underwriters Laboratories Incorporated (UL) and met\nthe UL 214 standard for LPG hose.\nThe hose was imprinted with text. One side of the black neoprene cover of the hose\nfeatured a blue Mylar stripe extending along the length of the hose with the Durodyne logo and\n\"DD-G-220 LPG TRANSFER ONLY 350 PSI MAX WP\" printed in black. The phrases \"To\nprevent serious injury or property damage use for intended purpose only,\" \"Warning: Use of\ndamaged hose could be hazardous; inspect hose and couplings prior to each use,\" and \"Textile\nreinforcements meet UL21\" were also embossed along the imprinted blue Mylar stripe. The\nopposite side of the hose had \"DURODYNE DD-G-220 LPG HOSE UL21 ISSUE E-7874\nMH29239 SPEC DD-G-220 TEXTILE BRAID WP 350 PSIG 4Q05 INSPECT HOSE BEFORE\nUSE\" embossed on an imprinted stripe extending the length of the hose. (See figure 1.)\nFigure 1. Blue Mylar and imprinted stripes on accident hose.\n3 In August 1999, Durodyne was purchased by Argo Tech Costa Mesa. In 2007, Eaton Corporation purchased\nArgo Tech, including the Durodyne unit.\n4 UL standard 21, for LPG hose, covers hose and hose assemblies in sizes up to and including a nominal internal\ndiameter of 4 inches for conveying LPG.\n\n<<<PAGE 5>>>\n\n3\nSmart-Hose Technologies of Philadelphia, Pennsylvania, purchased the LPG transfer\nhose from Durodyne and installed the Smart-Hose safety system to fabricate the complete\naccident hose assembly as a Lifeline III LPG Transfer Hose. The Smart-Hose safety system\nconsists of an internal cable running through the bore of the hose connected to specially designed\nunseated flapper valves located on each end of the cable. In the event of hose or coupling\nseparation, catastrophic hose rupture, or excessive hose stretching, the system is designed to shut\noff the flow of LPG in both directions as the flapper valves release and instantly seat.\nThe Smart-Hose assembly consisted of a 238-inch-long 1/4-inch lxl9 galvanized steel\nstrand cable with a nylon coating down the bore of a 216-inch-long piece of Durodyne LPG\ntransfer hose. Each end of the cable was connected to a valve flapper made from 316 stainless\nsteel. 5 The flappers were connected to 2-inch-diameter 316 stainless steel female national pipe\nthread end fittings on each end of the hose. The end fittings were secured to the hose by 2-inch\nferrules crimped onto each end of the hose, which completed the Smart-Hose assembly and held\nthe cable in compression. Records indicate that the hose assembly was 222 inches (18 1/2 feet)\nlong. Smart-Hose issued a new hose test certification for this hose assembly, serial number\n10573, on October 18, 2005.\nEach end of the hose also had a male acme hammer lock coupling made of cast iron and\ncarbon steel that threaded into each end fitting. This type of coupling is not acceptable in LPG\napplications because of sparking issues, but it is appropriate for anhydrous ammonia\napplications. According to Smart-Hose Technologies, these couplings were not installed by or\npurchased from Smart-Hose at the time Smart-Hose completed the hose assembly. Werner\npurchased the transfer hose assembly from Gas Equipment Company, Inc. of Indianapolis,\nIndiana, on December 20, 2005. The invoice for the purchase does not include any information\nabout the couplings, such as whether the couplings were purchased from or installed by Gas\nEquipment Company. Additionally, no other records or receipts were found that identified the\ncompany that installed the couplings. As a result, NTSB investigators were unable to determine\nwhen the hammer lock couplings were installed and who installed them.\nPostaccident Testing and Analysis of Accident Hose Assembly\nOn September 9, 2009, the LPG transfer hose assembly was examined at the NTSB's\nMaterials Laboratory in Washington, D.C., in the presence of the parties to the investigation. The\noverall length of the hose assembly was 18 1/2 feet. The measured length of the hose from end\nfitting to end fitting was 18 feet. At the time of construction, Smart-Hose certified the length of\nthe hose assembly, not including the couplings, as 18 112 feet.6 The rupture in the hose was about\n5 1/2 inches long. The centerline of the rupture was located 131 inches from the A end7 and\n91 inches from the Bend of the hose assembly.\n5 Type 316 stainless steel is an austenitic chromium nickel stainless steel containing molybdenum. Because of\nits superior corrosion and oxidation resistance, good mechanical properties, and fabricability, 316 stainless steel has\napplications in many sectors of industry, including its use for the manufacture of tanks and storage vessels for\ncorrosive liquids.\n6 Smart-Hose literature states that these hoses may contract up to 3 percent when pressurized. Therefore, an\n18 I /2-foot hose could contract 6 inches or more (that is, 3 percent of 18 I /2 feet, or 222 inches, is 6.66 inches).\n7 The \"A\" and \"B\" ends of the hose assembly were arbitrarily chosen and labeled by the NTSB's Materials\nLaboratory for reference use only.\n\n<<<PAGE 6>>>\n\n4\nThe exterior surfaces of the hose assembly contained only superficial abrasions. No\ngouges, slices, or other defects were noted on the surface of the hose, including the areas\nadjacent to and abutting the rupture area. The black text on the blue Mylar stripe was abraded in\nmultiple locations so that the stenciling along the length of the hose could not be read; however,\nseveral sections of the hose assembly were free of abrasions or stains, and the text could be read.\nThe embossed text on the imprinted line could be read along the length of the hose assembly.\nFractographic8 evidence indicates that the rupture in the accident hose assembly initiated\non the interior wall of the hose and propagated outward. The NTSB's Materials Laboratory\nidentified a definitive fracture origin on the surface of the fracture that was indicative of\nrelatively slow crack growth. Several secondary cracks were noted in the interior wall of the hose\nnear the fracture origin; however, the interior surface did not appear to be degraded from\nanhydrous ammonia exposure.\nThe reinforcing fibers in the two innermost PET braid layers of the hose assembly\nappeared to be severely damaged on both halves of the fracture surface along the entire length of\nthe ruptured area. In some parts of the ruptured area, the fibers were clumped together and\nappeared to be encased in salt-like particles. Laboratory analysis of the hose assembly revealed\nthat the fibers in the two innermost PET braid layers were degraded to the point that they were\nbrittle and friable when strained or mechanically flexed. (See figure 2.)\n8 Fractography is the study of the fracture surfaces of materials.\n\n<<<PAGE 7>>>\n\n5\n3rdbraid\n2\"dbraid\nl 5tbraid\n3rd braid\n2nd braid\nFigure 2. Cross-section of accident hose showing brittleness of fibers in two innermost PET\nbraid layers.\nThe PET used for the yam braids is not chemical~ resistant to anhydrous ammonia or\nammonium hydroxide. As documented in several studies, exposure to anhydrous ammonia or\nammonia-related compounds results in a chemical reaction (ammonolysis) that can cause PET\nfibers to degrade and lose strength.\n9 (a) C. Lorenzetti, et. a!., \"Chemical Recovery of Useful Chemicals from Polyester (PET) Waste from\nResource Conservation: A Survey of State of the Art,\" Journal of Polymers and the Environment, vol. 14, no. 1\n(2006), pp. 89-101. (b) V. Sinha, et. a!., \"PET Waste Management by Chemical Recycling: A Review,\" Journal of\nPolymers and the Environment, vol. 18, no. I (2008). (c) M. Khaddaj, et. al., \"Processing of New Materials Using\nThermal and Thermo-Vaporous Treatment of Terephthalates,\" Journal of Physics, Conference Series 121 (2008).\n(d) R. Lamparter, et. a!., \"Process for Recovering Terephthalic Acid from Waste Polyethylene Terephthalate, United\nStates Patent 4542239, September 17, 1985. (e) W. Murdoch, \"Production ofTerephthalic Acid and Ethyl em: Glycol\nfrom Polyethylene Terephthalate by Ammonolysis,\" United States Patent 6723873, April20, 2004.\n\n<<<PAGE 8>>>\n\n6\nThe outer rubber layer of the accident hose had an array of small pinpricks along its\nlength that was intended to allow the product that it was transferring to permeate through the\nrubber layers of the hose and escape to the atmosphere. The purpose of this is to prevent gas\nfrom becoming trapped in the hose wall and damaging the hose; this pinprick design is standard\nin rubber hoses for LPG and anhydrous ammonia service. When the anhydrous ammonia\npermeated through the rubber layers of the accident hose, it collected in the interstitial spaces of\nthe fibers within the PET braids. Also, the accident hose was likely exposed to moisture,\nincluding humidity and rain, throughout its life cycle. Any absorbed and dissolved moisture\ncontained in the accident hose likely would have converted the trapped anhydrous ammonia to\nammonium hydroxide, leading to chemical degradation of the PET fibers.\nTesting completed by both the NTSB 's Materials Laboratory and an independent\nlaboratory10 confirms that the PET fibers in the accident hose had sustained chemical degradation\nthat dramatically reduced the strength of the PET fiber. The NTSB concluded that the accident\nhose failed because it was not chemically compatible with the anhydrous ammonia in the cargo\ntank and that caused the chemical degradation, loss of mechanical strength, and ultimate failure\nofthe cargo hose.\nUse of Chemically Incompatible Hose\nWhen making deliveries, Werner drivers sometimes used facility-owned hoses instead of\nthe hose on the cargo tank vehicle. It is not known how many times the accident hose assembly\nwas used to transfer anhydrous ammonia before it failed.\nFollowing the accident, a NTSB investigator discovered that the accident hose was the\nLPG transfer hose that was originally assigned to trailer 2322, not the accident trailer (that is,\ntrailer 3002). Further investigation revealed that an anhydrous ammonia transfer hose\nmanufactured by Goodall Canada Inc. was carried on board trailer 2322 at the time of the\naccident. According to records and statements from Werner, trailers 2322 and 3002 were stored\non the same secure lot in Tampa, Florida, for a several hours on May 17, 2009. Although Werner\nstated that neither of its drivers had admitted to exchanging the transfer hoses, no other\nopportunity existed for the LPG transfer hose to be placed on the accident trailer. Based on its\nrecords, Werner estimated that the accident hose was used to unload anhydrous ammonia\nbetween 2 and 12 occasions.\nThe physical properties of hazardous materials vary so greatly that cargo hoses are\nconstructed and intended for use with specific products and cannot be used interchangeably. As\npreviously noted, the hose in the accident hose assembly had internal fibers made of PET, which\nis not chemically compatible with anhydrous ammonia. Therefore, any cargo hose containing\nP.ET fibers would not be suitable for anhydrous ammonia service. The need for chemical\ncompatibility applies not only to the hose material, but to all components of the completed hose\nassembly, including end fittings and couplers. The accident hose assembly had cast iron and\ncarbon steel couplings that were appropriate for anhydrous ammonia service but inappropriate\nfor LPG service, which requires spark-resistant materials such as brass, bronze, and stainless\nsteel.\n10 Trace Laboratories also tested the accident hose.\n\n<<<PAGE 9>>>\n\n7\nTherefore, the NTSB concluded that because of the chemical incompatibilities of the\nhose material with anhydrous ammonia and of the couplers with LPG, Werner Transportation\nServices, Inc.'s hose assembly was not suitable for use with either anhydrous ammonia or LPG\nand should not have been carried on the cargo tank motor vehicle that was involved in this\naccident. The PHMSA Office of Hazardous Materials Safety issues safety advisory notices to\nhelp the public understand significant safety risks. PHMSA's safety advisories are published in\nthe Federal Register and provide a description of the safety issue and a recommended action to\nresolve the issue. The Federal Motor Carrier Safety Administration (FMCSA), in conjunction\nwith its duties enforcing rules and regulations, conducting inspections, and licensing hazardous\nmaterials carriers, also issues safety advisory notices pertinent to cargo tank safety. The two\nagencies working together could provide the necessary outreach to assist carriers and facility\noperators in avoiding the hazards associated with the use of chemically incompatible hoses and\ncouplers during loading and unloading operations. Therefore, the NTSB recommends that\nPHMSA, with the Federal Motor Carrier Safety Administration, jointly issue a safety advisory\nbulletin to inform cargo tank motor vehicle owners and operators, registered inspectors of these\nvehicles, and transfer facility operators about the circumstances of this accident and actions\nneeded to prevent the occurrence of a similar accident.\nFacility and Carrier Information\nTanner, of Southampton, Pennsylvania, owned and operated the Swansea,\nSouth Carolina, facility at which this accident occurred. Tanner was registered with the FMCSA\nand PHMSA as a private carrier, transporter, and shipper of various hazardous materials,\nincluding anhydrous ammonia.\nThe cargo tank truck involved in this accident was owned and operated by Werner, of\nGainesville, Georgia. ·The company's business consists primarily of bulk transportation of\nanhydrous ammonia to its customers. Werner also transports flammable gases, including LPG\nand butane. At the time of the accident, all 21 of Werner's hazardous materials cargo trailers\nwere DOT specification MC 331 cargo tanks that are authorized for the transportation of\nliquefied compressed gases, including both LPG and anhydrous ammonia.\nActions Preceding the Accident\nAbout 7:40a.m. on July 15, 2009, after the Werner cargo tank truck arrived at the Tanner\nfacility in Swansea, South Carolina, the driver parked the truck, picked up the vapor hose\nassembly that was lying on the ground next to the manifold, and connected the hose assembly to\nthe vent line on the cargo tank. He then removed the cargo transfer hose assembly from the\nstorage tube on the trailer, checked the pressure valves on the tank, and connected the hose\nassembly to the liquid discharge fitting on the tank. A Tanner plant employee fastened the other\nend of the transfer hose assembly to the facility piping manifold. The driver then engaged the\npower take-off unit, turned on the pump, and checked the tank volume gauge on the side of the\ncargo tank to make sure product was flowing from the cargo tank to the storage tank. The driver\nwatched the reading on the cargo tank volume gauge drop from 71 percent full to 64 percent full,\nat which point he told the trainee to oversee the unloading while he completed paperwork in the\ntractor. A driver trainee, who was accompanying the cargo tank truck driver, watched the gauge\nto make sure the unloading was proceeding correctly. The last gauge reading observed by the\n\n<<<PAGE 10>>>\n\n8\ntrainee before the accident indicated that the cargo tank was about 53 percent full. The trainee\nlater estimated that at that time, 1 ,500 gallons had been transferred.\nJust before 8:00a.m., about 7 or 8 minutes after the transfer had begun, the trainee heard\na pop sound followed by a loud rush of gas. He looked underneath the truck and saw a rupture in\nthe transfer hose assembly directly in front of him. He stated that he saw anhydrous ammonia\npluming upward from the rupture, forming a dense white cloud. The trainee immediately pushed\nthe emergency shutdown button on the rear of the cargo tank. Upon doing this, he noticed that\nmovement of the ammonia cloud quickly shifted in the direction of highway 321. He then turned\nto the facility manifold to find a shutdown switch, but did not see one because the white\nammonia cloud was surrounding the manifold. He then evacuated through the south gate and ran\nwest into a wooded area. Eventually he heard the hose stop and saw the cloud lift soon after.\nThe driver stated that he also heard a loud pop about 8:00 a.m. and saw a cloud of gas\noutside the driver's side window. He then turned off the engine to stop the trailer pump; put on\nhis half-face respirator; got out of the truck on the passenger side; and escaped through the white\ncloud in front of the building along the fence, and exited through a gate.\nTo reduce the likelihood of using a hose assembly that is not chemically resistant to the\nhazardous material to be loaded into or unloaded from a highway cargo tank, the motor carrier\nand/or the facility carrier should not only visually inspect the cargo hose assembly for defects,\nbut also verify the chemical products that can be safely transferred through the hose assembly.\nVerification can be accomplished by noting markings on the hose assembly or through a written\ncertification that lists acceptable products for the hose assembly and/or restrictions provided by\nthe owner of the hose assembly. Verification that a cargo hose assembly is appropriate for its\nintended use also should be incorporated into the required pretransfer procedures. The NTSB\nrecommends that PHMSA require cargo tank motor vehicle carriers and transfer facilities to\nverify (1) that cargo transfer hose assemblies, whether carried on the vehicle or provided by the\nfacility, are chemically compatible with the hazardous material to be transferred and (2) that\ndrivers verify hoses are marked as compatible with the material to be transferred before either\nloading or unloading operations begin.\nInadequate Passive Emergency Discharge Requirements\nTitle 49 CFR 173.315(n)(2) requires that bulk transport vehicles transporting certain\nliquefied compressed gases, including anhydrous ammonia, be outfitted with passive emergency\nshutdown control equipment. The passive shutdown system serves as a means to shut off\nautomatically the flow of product from the cargo tank motor vehicle-without the need for\nhuman intervention-within 20 seconds of an unintentional release caused by \"a complete\nseparation of a liquid delivery hose.\" The two types of passive shutdown systems commonly\nused in industry to satisfy this requirement are the Smart-Hose system found on the accident\nhose and a permanently mounted, computer-controlled leak detection/shutdown system. The\ncomputer-controlled leak detection/shutdown system activates in the event of pressure change,\nwhereas the Smart-Hose system activates as a result of mechanical failure of the hose assembly\nthat leads to the internal cable's being stretched to a predetermined length. Consequently, a\ncomputer-controlled leak detection/shutdown system can shut off the flow of product as a result\nof either partial or complete hose separation, whereas the Smart-Hose system can do so only in\n\n<<<PAGE 11>>>\n\n9\nthe event of complete hose separation and the tensioning of the internal cable. Both systems\nsatisfy the current requirement to stop the flow of product from the cargo tank if there is a\ncomplete separation of the cargo hose assembly. However, in this accident, the hose assembly\ndid not experience a complete separation, and the internal cable inside the hose assembly was\nunaffected. As a result, the internal cable was not stretched to the predetermined length necessary\nand the flow rate was insufficient to activate the flapper valves in the ends of the hose assembly,\nwhich would have cut off the flow of anhydrous ammonia in all directions.\nThe current regulation does not take into account a rupture of a cargo hose without\ncomplete separation. For toxic or flammable gases, such as anhydrous ammonia and LPG, the\nconsequences of a hose rupture without separation can be just as severe as the consequences of a\ncomplete hose separation. The outcome of either scenario is the uncontrolled and free flow of the\ntoxic or flammable gas from the cargo tank. The impact of the hose rupture in Swansea was\nslightly mitigated because the driver trainee was able to trip the emergency shut-off valve on the\ncargo tank and thereby prevent the complete release of ammonia from the cargo tank. Even so,\nthe cloud of ammonia gas generated by the accident was sufficient to cause a passing motorist to\nsuffer a fatal injury.\nThe current requirement11\nwas established to prevent catastrophic human loss and\nproperty destruction that may result from the failure of a hose assembly while hazardous\nliquefied compressed gases are loaded into and unloaded from highway cargo tank vehicles.\nHowever, emergency discharge control should function under all-rather than only select-\ncircumstances; that is, any hose assembly failure rather than only failures that result in complete\nhose separations. The NTSB concludes that given the unique hazards of toxic and flammable\nliquefied compressed gases, the requirements in 49 CFR 173.315(n)(2) for passive emergency\ndischarge systems on highway cargo tanks fail to provide an acceptable level of protection\nagainst all types of cargo hose assembly ruptures.\nThis accident demonstrates that hose assemblies do not always separate completely if\nthey fail, and the NTSB believes that passive shutdown systems designed to function as a result\nof a complete separation of a hose assembly alone should not be permitted to satisfy the\nemergency discharge control requirement. Therefore, the NTSB recommends that PHMSA\namend the provisions of 49 CFR 173 .315(n)(2) to require that passive emergency shutdown\ncontrol systems for highway cargo tanks activate in the event of a partial or complete failure of a\ncargo hose assembly.\nHose Assembly Inspection and Testing Requirements\nTitle 49 CFR 180.416 establishes standards for inspecting and testing cargo hose\nassemblies that are installed or carried on specification MC 330 and MC 331 cargo tank motor\nvehicles that transport liquefied compressed gases such as anhydrous ammonia and LPG.\n11 On September 8 1996, in Sanford, North Carolina, during delivery ofpropane to a bulk storage facility by an\nMC 331 bulk transport, more than 35,000 gallons of propane were released. The discharge hose separated from its\nhose coupling at the delivery end of the hose. Mos~ of the !ransport's 9,800 gallon~ o~propane and mo:~ than ~0,000\ngallons from the storage tanks were released. Ifth1s quant1ty of released propane 1gmted, local authontles estimated\nthat about 125 emergency response personnel could have been injured or killed. Federal Register Volume 62,\nNumber 159 (Monday, August 18, I 997). http://www.gpo.gov/fdsys/pkg/FR-1997-08-18/html/97-21865.htm\n(accessed July 5, 2011).\n\n<<<PAGE 12>>>\n\n10\nElements of the inspection and testing program include requirements for monthly inspections,\nannual leak tests, testing of new and repaired hose assemblies, and a safety check of each hose\nafter unloading.\nUnder section 180.416( d), cargo tank motor vehicle carriers must visually inspect each\ndelivery hose assembly at least once each calendar month in which the hose is in service and\nrecord the inspection date, the inspector's name, the identification number of the hose, the\ncompany name, the test date of the transfer hose assembly, and the result of the inspection (that\nis, pass or fail). The monthly hose inspection records are to be retained by the motor vehicle\ncarrier until the next test of the same type is completed.\nAlthough section 180.416( d) specifies that a hose assembly must be inspected each\nmonth it is in service, the regulation does not define what is meant by \"in service.\" It seems\nlogical that a hose assembly carried on a cargo tank motor vehicle has the potential to be used on\nany given day, and, therefore, it should be considered to be \"in service.\" If a motor vehicle\ncarrier does not believe that it will be used, it will not necessarily consider it to be in service and\nas a result, the hose assembly may not be inspected, as was the case in this accident. The intent\nof this regulation presumably was to ensure that each hose assembly carried on a cargo tank\nmotor vehicle is inspected on a monthly basis. However, the wording in the regulation seems to\nhave created a loophole, because the term \"in service\" could be interpreted differently by various\nmotor vehicle carriers. Because monthly inspections are an important preventive measure for the\nidentification of physical deterioration, damage, and excessive wear, it is critical that these\ninspections are performed routinely.\nTypically, if PHMSA receives an industry inquiry regarding a specific hazardous\nmaterials regulation, it publishes a formal interpretation of the regulation in question to clarify or\nexplain the intent of the regulation. These interpretations are disseminated to the respective\nparties and posted on the PHMSA website. PHMSA has not published an interpretation on this\nsubject matter. The NTSB concludes that the lack of clarity of 49 CPR 180.416(d) regarding\nmonthly inspections of \"in service\" cargo hose assemblies can lead motor carriers to mistakenly\ndefer monthly inspections of transfer hose assemblies that are carried on cargo tank vehicles but\nbelieved not to be used regularly. Therefore, the NTSB recommends that PHMSA publish and\ndisseminate a formal interpretation of 49 CPR 180.416( d) that includes the criteria that determine\nwhen a cargo transfer hose assembly is \"in service.\"\nInadequate Annual Hose Assembly Leakage Test Requirements\nTitle 49 CFR 180.407 specifies the requirements for annual leakage tests for specification\nMC 330 and MC 331 cargo tanks that are used to transport liquefied compressed gases such as\nanhydrous ammonia and LPG. In accordance with section 180.407(h)(l), the leakage test is to\ninclude product piping \"with all valves and accessories in place and operative.\"\nSection 180.407(h)(4) requires registered inspectors of MC 330 and MC 331 cargo tanks to\ninspect visually the delivery hose assembly for noticeable defects while the hose assembly is\nunder the same test pressure as the tank. This paragraph further states that hose assemblies that\nare not permanently attached to the cargo tank motor vehicle can be inspected separately from\nthe cargo tank motor vehicle. Title 49 CPR 180.416( e) requires that the \"owner of a cargo hose\nassembly that is not permanently attached to a cargo tank motor vehicle must ensure that the\n\n<<<PAGE 13>>>\n\n11\nhose assembly is annually tested\" in accordance with section 180.407(h)(4).12 Under section\n180.407(h)(4), in addition to the written record of the inspection of the cargo tank motor vehicle,\nthe registered inspector conducting the leakage test of the hose assembly must record the hose\nidentification number, the date ofthe test, and the condition of the hose assembly.\nThe most recent annual external visual inspection and leakage test records for Werner's\nnine cargo tank motor vehicles that were equipped with hose assemblies were reviewed by\nNTSB investigators; these records were found to be incomplete. Those tests were performed by\n(1) L&L Repair and Testing, Inc., and (2) Boyd Service, Inc. Although each of the reports\nindicated that the hose assembly had been visually inspected and found to be in an acceptable\ncondition, only three of the reports included a separate leakage testreport. The test reports for the\nother six trailers indicated that the hose assemblies had been visually inspected and found to be\nin acceptable condition; however, no indication existed that any of these hose assemblies had\nbeen leak tested, and none of the required hose identification information was included in any of\nthe reports.\nAdditionally, although none of the 12 other vehicles that Werner owned at the time of the\naccident were equipped with hose assemblies, the test records for those cargo tank vehicles\nindicated that they were carrying hose assemblies that had been visually inspected and, in some\ncases, the test records included hose identification numbers. The inconsistencies found in the\ninspection and testing records strongly indicate that the registered inspectors had not been\nconsistently conducting leakage tests on the cargo hose assemblies on the cargo tank vehicles nor\nwere they completing annual leakage test reports as required.\nThe FMCSA also uncovered deficiencies in Werner's test records. During a posta,:cident\nfacility review, the FMCSA cited one of the companies contracted by Werner to perform\ninspections because it failed to include required information on test and inspection reports and to\nretrain hazardous materials employees every 3 years. The FMCSA reviewed 50 test and\ninspection records and found that all of them were missing some ofthe required information. The\nNTSB concludes that Werner's incomplete and incorrect inspection records of cargo tank and\nhose testing suggests that the accident cargo hose assembly may not have been inspected and\ntested properly before the accident. The NTSB believes that the compliance reviews conducted\nby the FMCSA following this accident, and its subsequent enforcement actions, satisfied the\nneed for an audit of Werner and its contracted registered inspectors. Therefore, the NTSB is not\nissuing any safety recommendations for this purpose at this time.\nNotwithstanding the hose inspection deficiencies of Werner and its contracted registered\ninspectors, the lack of clarity of the regulation (section 180.407(h)(4)) is also a factor in this\naccident. After stating the requirement for registered inspectors to inspect the hose assemblies\nwhile under leakage test pressure, the regulation states that \"Delivery hose assemblies not\npermanently attached to the cargo tank motor vehicle may be inspected separately from the cargo\ntank motor vehicle.\" Although PHMSA has not published a formal interpretation of this\nlanguage, it has indicated to investigators that \"inspected separately\" is intended to mean that a\nhose assembly does not have to be physically attached to the cargo tank to be tested for leaks. As\n12 These leakage tests include visual inspection of the hose assemblies while they are under leakage-test\npressure (that is, 120 percent of maximum working pressure).\n\n<<<PAGE 14>>>\n\n12\nis the case with the language contained in section 180.416 regarding monthly hose inspections\nbeing completed on hose assemblies that are in service, the NTSB believes that this language\ncould also be interpreted differently depending on the individual. One possible misinterpretation\nwould be that a hose assembly may be tested for leaks at a time other than during the annual\ninspection of a cargo tank motor vehicle. PHMSA told investigators that a cargo tank motor\nvehicle should not pass an annual inspection without its hose assembly being leakage tested,\nsince the hose assembly is considered to be part of the vehicle. However, the fact that registered\ninspectors allowed 12 of Werner's vehicles to pass annual inspection when none of them were\nequipped with hose assemblies indicates that industry and the regulators do not agree on the\nscope and procedures for leakage testing. Therefore, the NTSB concludes that the lack of clear\nrequirements for testing cargo hose assemblies and cargo tank motor vehicles for leaks has\nadversely affected the accuracy of the test records. The NTSB recommends that PHMSA issue\nguidance to motor carriers and registered inspectors that clarifies the testing and the\nrecordkeeping requirements of 49 CFR 180.407 for cargo hose assemblies and cargo tanks that\nare used to transport liquefied compressed gases to ensure that all hose assemblies are tested for\nleaks on an annual basis.\nTherefore, the National Transportation Safety Board makes the following safety\nrecommendations to the Pipeline and Hazardous Materials Safety Administration:\nWith the Federal Motor Carrier Safety Administration, jointly issue a safety\nadvisory bulletin to inform cargo tank motor vehicle owners and operators,\nregistered inspectors of these vehicles, and transfer facility operators about the\ncircumstances of this accident and actions needed to prevent the occurrence of a\nsimilar accident. (H-12-2)\nRequire cargo tank motor vehicle carriers and transfer facilities to verify (1) that\ncargo transfer hose assemblies, whether carried on the vehicle or provided by the\nfacility, are chemically compatible with the hazardous material to be transferred\nand (2) that drivers verify hoses are marked as compatible with the material to be\ntransferred before either loading or unloading operations begin. (H-12-3)\nAmend the provisions of Title 49 Code of Federal Regulations 173.315(n)(2) to\nrequire that passive emergency shutdown control systems for highway cargo tanks\nactivate in the event of a partial or complete failure of a cargo hose assembly.\n(H-12-4)\nPublish and dissemi","truncated":true,"body_characters":41240}