^ Fig. 1. The inside surface (the acid side) of a leaked tube.

Article by Dr. Elayaperumal Kailasanathan, Corrosion and Metallurgical Consultant, India


The acid vaporizer
The tubes of the heat exchanger were of 316L stainless steel. The
material was tested and confirmed. The tubes had a wall thickness of
2.11 mm and frequently leaked during operation (steam on the shell side
and organic acid on the tube side) after only four months of service.
The designed service conditions were as follows: The organic acid, which
was preheated in an upstream preheater, entered the first pass tubes of
the vaporizer at 70°C, and was vaporized in the heat exchanger. The
vapor left the heat exchanger at 185°C and at a pressure of 4.0 kg/cm2g
and entered a further downstream super-heater.
There was no leakage in the upstream preheater or in the downstream
super heater. Leakage only occurred in the intermediate vaporizer.

The acid, bought from outside sources, was of 99.8 % purity. The
levels of impurities, such as lower-molecular-weight organic acid,
sulfuric acid, chlorides, and moisture in the as-received acid (which
usually lead to organic acid corrosion), were found to be in traces
insufficient to induce corrosion. From a purity perspective, therefore,
the organic acid used was not corrosive to 316Lstainless steel. The
evaporator was a 4-pass straight-tube heat exchanger having 120 tubes in
total. The leakages occurred only in the second and third pass tubes.
The leaking tubes were inspected closely on the inside surface
(tube-side, acid side) after being split longitudinally.

It was observed that:

  • The tubes removed from the first and fourth passes – about 30 tubes per pass, showed a good, smooth, corrosion-free surface.
  • The tubes removed from second and third passes had a rough
    inside surface. The roughness was not uniform in all the 360 degree
    positions of the tubes. There were longitudinal streaks of roughness
    along the length of the tube, separated by a few degrees in the angular
    direction. On close examination, using a magnifying lens, it was found
    that the streaks contained a series of circular and elliptical cavities
    along the length of the tubes. These were mainly elliptical, with a long
    axis parallel to the length of the tubes. See Fig. 1.
Samples were examined macroscopically using a depth focus stereo microscope under a low magnification (8x). See Fig. 2.
In Fig. 2 circular and elliptical ‘cavities’, can be seen on the
inside surface aligned in a longitudinal stretch along the length of the
tube. This is not uniform corrosion by an organic acid. The
longitudinal stretching of the cavities is nearer to the outlet end and
farther from the inlet.
The sample was further cut and polished to reveal the cross section
of the wall of the tube of the inside surface and examined under an
optical microscope at a magnification of 100x. See Fig. 3.

The following are inferences from the above site information and sample examinations:

  • There is no metallurgical quality problem in the 316L stainless steel items used for the entire process-wetting parts.
  • The acid used is not corrosive to 316L stainless steel at 70°C
    in liquid form at the operating pressure of 4.2 kg/ cm2 g (Pass No.1).
  • At temperatures approaching 185°C and an operating pressure of
    4.0 kg/cm2 g, the acid vapor is also not corrosive to 316Lstainless
    steel (Pass No.4).
  • Tube leakage occurred at passes 2 and 3 only, where the phase change from liquid-to-vapor occurred. (Passes 2 and 3)
  • This leakage is not due to any uniform chemical corrosion, but
    to mechanical cavitation assisted by velocity-induced erosion of
    high-intensity vapor bubbles. These bubbles formed during evaporation at
    points of high-temperature gradient where they burst on the metal
    surface (cavitation) assisted by localized high velocity moving the
    vapor bubbles. In turn they made elliptical cavities on the inside
    surface, elongated in the flow direction. This cavitation-erosion
    process ate away the metal over a period of time resulting in leakage.
    This is also not erosion–corrosion.

The remedial measures suggested are as follows:

  • Reduce the overall velocity of the fluid passing through the
    thirty tubes of each pass to the extent that the production rate allows.
  • Upgrade the metallurgy of the tubes to duplex stainless steel
    2205 (UNS S31803) or 2507 (UNS32750). These are austenitic+ferritic
    duplex stainless steels with mechanical yield strengths double that of
    316Lstainless steels. Minimum values are 220 and 450 N/mm2 for 316L and
    2205, respectively. At the same time, their corrosion resistance to
    organic acids is far better than that of 316Lstainless steels. As their
    mechanical strength is higher, the duplex stainless steel tubes can be
    thinner, thereby also reducing velocities.