Page 316 - Failure Analysis Case Studies II

Failure Analysis Case Studies II
D.R.H. Jones (Editor)
0 2001 Elsevier Science Ltd. .All rights reserved 30 1

CREVICE CORROSION OF 316L CAUSED BY CHLORIDE
PARTITION IN WATER-BUTANONE MIXTURES

J. H. CLELAND
Cambcor Ltd, 30 Windsor Road, Cambridge CB4 3JW, U.K.
(Received 9 May 1997)

Abstract-Chloride-induced crevice corrosion of a 316L pump was found to have occurred in what was
nominally a very dilute solution of sodium chloride in a single-phase solution of water in butanone at 25 "C.
The hypothesis formed was that water had entered the system in a sufficient amount to form an immiscible
mixture of a water-rich phase and a butanone-rich phase. This separation into two liquids was followed by
preferential partition into the water-rich phase of the sodium chloride to give a brine. A simple test demon-
strated that the hypothesis was tenable. 0 1997 Elsevier Science Ltd.
Keywords: Crevice corrosion.

1. BACKGROUND
Product contamination first alerted the operators of a production unit that they had a corrosion
problem. The unit was a mixing unit used for batch processing at 25°C a single-phase mixture
which contained, among other components, a high proportion of butanone in which small amounts
of water and chloride ions were dissolved. The only metallic components in the unit, which was
built of Pyrex glass, and had given trouble-free service for some time, were four identical 316L
stainless steel pumps.
The production unit was dismantled, and the problem was traced to crevice corrosion in one of
the stainless steel pumps. No corrosion was found in the other three pumps. The pump which had
suffered the crevice corrosion was located at almost the lowest point in the unit (the lowest point
was a drainage tap). It was reported that the liquid collected when the system was drained was an
emulsion. It therefore appeared that the liquid in the pump may have been a mixture of two
immiscible liquids: an aqueous solution of butanone as well as the expected solution of water in
butanone. Unfortunately, no samples were retained for analysis.

2. THE MATERIALS
2. I. The pump body
The stainless steel used for the pump body was UNS S31603 (316L) and chemical analysis (Table
1) showed that the metal conformed to the requirements of this specification.

2.2. The solution
The solution pumped was basically an 8wt% solution of distilled water in analytical grade
butanone (methyl ethyl ketone). The phase diagram for the water-butanone system is shown in Fig.

Table 1. Chemical analysis of a section from the corroded pump: composition in wt%
C Mn Si P S Cr Ni Mo
-~
~
Pump body 0.019 1.73 0.59 0.028 0.004 16.5 10.2 2.14
UNS S31603 0.03 2.00 1 .oo 0.045 0.030 16.&18.0 10.0-14.0 2.W3.00
maximum maximum maximum maximum maximum
Reprinted from Engineering Failure Analysis 4 (4), 287-29 1 (1 997)