Oils become viscous and hard to pump at low temperatures. In the case at hand, refined mineral oil was being stored in a large tank equipped with a heat exchanger to keep the stuff pumpable. Fairly low temperature steam circulated through an array of Ύ-inch copper tubes immersed in the oil and then discharged into a dry well. Several of the tubes broke, which allowed the mineral oil to leak into the dry well and the surrounding ground and finally into Mount Hope (RI) Bay. Local environmentalists and waterfront residents were not pleased by the leak.
The Scene of the Crime
I was retained by the operator of the tank to find the cause of the piping fracture. My client, of course, hoped to be exonerated.
The tubing was just plain copper that, for some reason, had broken after only two years of service. Neither refined mineral oil nor clean steam attack copper, so the cause of failure was a mystery.
I did only a hand examination of the broken tubing the main study had been done by an expert for another party. His report had excellent electron and optical micrographs of the fractured region. I relied heavily on his results. Unfortunately, I was unable to locate the author of the report to obtain permission to use his figures in this column.
The fracture surfaces showed fatigue crack striations, which are caused by cyclic stresses. However, the thermal stresses-induced during heating and cooling were not nearly big enough to give failure in sound tubing. The scanning electron microscope pictures identified the origin. The fatigue cracks had started at intergranular cracks through the grain boundaries between the individual copper crystallites. Something corrosive attacked the grain boundaries separating the grains. The resulting cracking weakened the tubing and made it susceptible to fatigue failure.
It would have been easy to postulate a contaminant in the steam causing cracking. But the cracking started at the outer surface of the tubing that contacted only the supposedly benign mineral oil. The steam played no part in the failure. What on earth could have attacked the usually inert copper?
The Smoking Gun
Ammonia is one of the few substances to give intergranular cracking on copper and brass. But how did ammonia get in contact with the tubing? Ammonia was not involved in manufacturing or installation. I scratched my head a good bit and came up with two possibilities that could have occurred during storage. One was that the droppings from birds in warehouse rafters fell on the pipe. The other was that passing dogs decorated the piping with their urine, as dogs are wont to do. In either case, the damage was done before the piping was installed. The explanation seems far-fetched and I even got a call from a member of my client's staff asking if I were serious. I assured him I was.
The heat exchanger tubing actually suffered an attack by an old malady. Ammonia-induced cracking of copper and brass was observed in the early days of the automobile in America when farmers would store their prized horseless carriages in their barns for the winter. Often in the spring, the once-gleaming brass headlights were marred by a network of cracks, a phenomenon that came to be known as season cracking. The barns contained animal manure that gives off ammonia, which in turn attacked the grain boundaries. I recently learned the same term was used in India to describe cracking in brass cartridge cases that occurred during the monsoon season. Rotting vegetation produced ammonia. The cracked cases rendered the cartridges useless as ammunition.