After analysis, we found that the selection of worng type of pipe is causing typical failure problem.
- First we checked the MOC from a third party lab & got it tested for SS-316L. So that everybody is cleared about the doubt on its inferiority. This is the first reaction normally when any material failure occurs.
- Then we did boroscopy of tubes & found the following pics.
- Now this led us to the direction of Erosion caused phenomenon or some surface / Localized / crevice boiling related phenomenon which might have been the reason for this.
- For confirmation we checked the pipe line downstream of this exchanger & found following.
- Based on this we decided not only to calculate but analysed the mixture for its IBP in the lab. It was calculated to be 120°C whereas lab measured it ~123°C at Atmospheric condition.
- The pressure in the system was supposed to be ~3 bar, but to increase the plant capacity operator has fully opened a valve on the reactor inlet which was downstream of this exchanger. So the pressure in the system was having a back pressure only equal to system pressure drop.
- This led to a situation where minor pressure drop can cause vapor formation.
- So whenever, there was a crevice or rough surface found in the flow path probably one of the component was getting vaproized causing higher velocity of flow.
- The toal flow was around 5-6000 Kg/Hr out of which only 100 Kg/hr vaporization was sufficient to reach an erosion velocity of ~8 m/sec.
- Therefore, the starting point was some errosion and then crevice corrosion in the area where SS grainular structure is destroyed.
- This is supporting the fact that there was no corrsion / erosion in the entire system except this section.
a. The failure was starting at welding joints where welding slag was creating rough surfaces, Crevices and sufficient localized drop in pressure to help vaporization of liquid e.g. water, acetic acid, or toluene.
b. The marks of erosion were higher on bottom side than on the top portion of the horizontal pipe.
c. The marks were appearing upto a certain length but not in the whole pipe.
d. Apart from exchanger failure, pipe failure has occured three times and in all cases it started from welding point upto a finite length.
After this we are planning to install Seamless socket welded pipes to avoid such failures in future.