Corrosion of Autoclave Jackets Subjected to a Steam-Water Cycle*

Abstract

The corrosion rate of steel autoclaves which are alternately heated by steam and cooledby water was found, after about 18 months’ use, to be greater than was considered acceptable for this type of equipment. The corrosion combined with the stresses caused by differential expansion of the jacket and the autoclave calls also caused crackmg of the welds between the Jacket and the body. The corrosive conditions were simulated in the laboratory and a considerable number of inhibitor systems for the steam and water were investigated. Steel test pieces were exposed alternately to steam for about 2½ hours and water for 1½ hours, provision being made to dose the steam and/or water. In screening tests lasting 20 cycles, addition of a few ppm of octadecylamine, hydrazine or morpholine to the steam (no treatment being given to the water) had practicallyno effect on the corrosion rate of mild steel. Sodium benzoate and mercaptobenzthiazole(NaMBT) additions to the water reduced the corrosion by about 80%; sodium nitrite and sulphite additions were not so beneficial. The most effective inhibitor was 0·1% sodium chromate which reduced corrosion by 90%. Some crevice attack occurred even in this solution, but this could be reduced by a heavy (25ppm) dosage of octadecylamine to the steam. This was confirmed during a test lasting 5 weeks.

Stressed welded specimens were exposed in the laboratory to the water-steam cycle, the water being inhibited with 0·1 % sodium chromate for one year, but although there was some crevice attack, even control specimens in contact with the untreated water did not crack. Over the year, weight-loss measurements showed a corrosion rate of about 0·005 in./yr in the inhibited system, and roughly 10 times this in the control specimens in an untreated system.