In a midwestern refinery, complete penetration of a 400-foot long 4-inch hot water (170 F) line occurred within two years in several sand back- filled areas. The attack was not so evident on the pipes in areas backfilled with loam-cinders.
The initial observations made are given in Table 1. The pipe to reference electrode potentials remained steady and no potential gradient effects related to stray current were noted. The line currents measured did not account for the intense attack in the sand areas. Why was it then that the attack occurred primarily on the cathodic area?
The investigation was continued by removing samples of the crust of corrosion product and soil adhering to the pipe on representative areas and examining them. Findings are given in Table 2.
The author has observed that mill scale and rust scale removed from alternately exposed and submerged fresh water flumes and diesel...
