The power of a vacuum… caused only by the condensing of the hazardous material. The loading temperature of the mono-isopropylamine, a flammable liquid, was reported to be between 70°F and 80°F. When the tank collapsed, the ambient air temperature was approximately 10°F…
On January 4,1988, an MC 307 tank trailer departed with 39,000 pounds of mono-isopropylamine, a flammable liquid. The truck was en route to a plant in Wyandotte, Michigan. The truck driver reported that about 5:30 p.m. on January 5, as the truck was proceeding northbound on Interstate 65 about 3 miles north of Sonora, Kentucky, he heard a loud noise similar to a blowout and the truck jerked hard to his left. In the side view mirror, the truck driver saw sparks coming from the rear of the vehicle. Although difficult, the truck driver moved the vehicle onto the shoulder of the road and stopped. He disembarked and extinguished a fire beneath the trailer which he believed was being fueled by grease on the support pads of the trailer’s dolly legs.
The truck driver said that the tank had collapsed inward between the fourth and fifth stiffening rings causing the tank bottom to sag such that the dolly legs and spare
tire rack were dragging on the roadway and creating severe folds in the top and upper sides of the tank shell. Although none of the tank’s contents was released, the
catastrophic failure of the tank severely impaired the driver’s ability to control the vehicle, creating the potential for a serious highway accident which would have resulted in the release of the hazardous cargo. Following the accident, the tank was off-loaded, cleaned, and transported to Deer Park, Texas, for inspection and repair.
The loading temperature of the cargo was reported to be between 70°F and 80°F. When the tank collapsed, the ambient air temperature was approximately 10°F. With a large decrease in temperature, such as experienced in this incident, the volume of a liquid material can be significantly reduced. Calculations by the shipper estimated that the mono-isopropylamine contracted from an initial volume of 6,905 gallons to a final volume of 6,540 gallons, a volume reduction of 5.3 percent. (For comparison, anhydrous hydrogen fluoride, a material with an extremely high coefficient of thermal expansion, would contract approximately 9 percent over a comparable decrease in temperature, whereas gasoline would contract approximately 3.5 percent.) The internal pressure within the vapor space in the tank would have decreased from 19.7 to 10.8 pounds per square inch, absolute (psia), to produce a vacuum of 4.1 pounds per square inch (psi). The shipper assumed in these calculations that no air was introduced into the tank vapor space as the liquid contracted.
The tank was built in 1979, was 41 feet long and 68.5 inches in diameter. It was constructed of stainless steel, designed for a pressure of 30 pounds per square inch, gauge (psig), and had a nominal capacity of 7,500 gallons (58,000). The tank did not have a vacuum design rating. The tank was equipped with a single-flanged pressure relief valve and was rated at 32 psig. The valve was equipped with a non-rated vacuum arrester that consisted of a steel ball and spring housed in Teflon.
In December 1987, the owner of the trailer conducted the Department of Transportation (DOT) required biennial inspection and its own 60-day required inspection. To comply with the DOT requirements, the owner inspected the external areas of the tank for evidence of corrosion; defects in welds, piping, valves, and gaskets; and signs of leakage. Additionally, the pressure relief valve was bench-tested because it was rated in excess of 7 psig. The 60-day inspection exceeds DOT requirements by including routine inspection of the tank interior, sonic tests to measure shell thickness, and cleaning or replacing all components of the pressure relief valve. New internal components for the pressure relief valve were installed in December 1987.
Between the December 1987 inspection and the January 5 incident, the tank had transported five shipments: two were mono-isopropylamine, one was monoisopropanolamine, one was alcohol not otherwise specified (n.o.s.), and one was a combustible liquid n.o.s. identified as n-methyl pyrrolodine.
Sonic measurements of the shell taken in Deer Park on January 13,1988, in the presence of a Safety Board investigator were consistent with the sonic measurements taken during the inspection in December 1987. The sonic measurements indicated that the tank shell thickness complied with the DOT design specifications. The interior tank did not show evidence of pitting or other signs of corrosion. The failure area was not centralized along a weld.
The internal components looked clean and new. During a bench-test of the valve, the pressure relief device opened at 27 to 28 psig (41.7 to 42.7 psia). The operation of the vacuum arrester was not consistent. It opened at vacuum pressures ranging from 1 to 11 psia. Upon completion of the bench-test, the Teflon housing of the vacuum component felt tacky and the steel ball appeared to be sticking slightly to the housing.
The Safety Board was not able to determine whether the vacuum arrester on the pressure relief valve had malfunctioned, had the capacity to relieve the vacuum formed due to the drop in temperature experienced during transportation, was affected by previous cargo residue, or did not perform due to some other factor. The vacuum arrester was not required to be installed on the MC 307 tank. Also, the manufacturer’s brochure cautions that the vacuum arrester, due to its limited size, should not be relied upon wholly to prohibit the implosion of a tank vehicle.
The MC 307 design specifications in 49 Code of Federal Regulations (CFR) 178.342-4 did not establish vacuum design criteria or require the use of vacuum relief devices to prevent the development of a critical vacuum at the time of this incident. The regulations did, however, require protection against overpressuring a tank and specify performance criteria for devices used to achieve this protection. The pressure relief valve also was removed and inspected.
CLICK HERE for this NTSB report