My idea of purgatory is driving the New Jersey Turnpike on an hot August day in a crowded car in eight lanes of bumper-to-bumper traffic, the air conditioning out, the heat gage on high, the gas gage on empty and no service for 20 miles. The only worse thing would be for the traffic to stop, which is just what happened in the case at hand.
Scene of the Crime
A tanker truck was traveling the turnpike on an August afternoon 24 years ago. The driver heard a “snapping” sound and stopped to investigate. The tank had split in half, spilling some 5,000 gallons of concentrated hydrochloric acid across several lanes of traffic. Hydrochloric acid, also known as muriatic acid, is widely used in cleaning, etching and the like. It is strong stuff that will rapidly eat away most kinds of steel and should be kept well away from living creatures.
The traffic of course stopped, both in the acidulated lanes and in the opposite lanes as well due to gawkers. A police officer on the scene was doing pretty well directing traffic until he noticed his breath was condensing, a phenomenon most unusual in August weather. The officer proceeded to panic. Magnesium wheels on cars near the spill corroded visibly in minutes.
Ultimately, the truck was towed away and the experts descended upon it to analyze the failure. The failure scenario would play a key role in determining what size checks the various parties involved in the manufacturing, use and maintenance of the tank would write. I was retained as a metallurgist by a local consulting firm.
The tank had nine reinforcing ribs around the circumference to prevent crumpling due to the 20-ton liquid load. The ribs were channels about six inches wide by four inches deep with the flanges welded around the circumference of the tank. The crack was underneath the central, fifth rib, which terminated at a circular filling port on the top of the tank.
We wanted to know the full extent of the crack, which was largely hidden by the rib. A radiography firm with a cobalt-60 source was retained to determine the thickness of the steel under the middle rib. This test was a waste of time and money. Radiography is useful for showing discontinuities in density, as might occur with a brittle crack. Corrosion had reduced the metal to a knife edge, which was undetectable by radiography.
A key issue was whether the tank had corroded from the inside or the outside. The inside of the tank had a quarter-inch thick rubber coating to protect the steel from attack. It was possible the coating had somehow failed, allowing corrosive attack.
The only way to determine the condition of the rubber lining was for someone to enter the fume-filled tank and investigate. This task fell to me as the metallurgist so I donned a rubber suit with an air supply and dropped into the tank, Swiss Army Knife in hand. I am a bit claustrophobic and also did not much like being immersed in poisonous fumes. It was the longest 15 minutes of my life.
The Smoking Gun
My probing showed the rubber to be tightly bonded to the steel all along the edges of the crack, so the corrosion had not come from the inside.
I did an extensive investigation on samples of steel from the tank but found nothing wrong. The tank was of sound low-carbon mild steel with the expected microstructure and mechanical properties. However, such steel is not meant to operate in contact with concentrated hydrochloric acid.
There was clearly a corrosive attack from the outside in, with the bottom of the central rib being corroded through. In fact, some of the corroded metal on the outside had been painted over. Painting was, presumably, in lieu of eliminating the errant acid flow.
So what happened? The failure had occurred due to acid spills during filling. The construction of the filler port was such that the spilled acid could get into the region under the rib, flow around the outer circumference and corrode the tank without the corrosion being detectable. This was sort of a time bomb in operation that finally resulted in the tank cracking.
The problem and ensuing calamity could have been prevented by welding a pair of steel plates over the 4 × 6 inch end openings in the central rib. The cost would have been less than $100.