Analysis of the causes of cracks in distillation column bodies
Thread Content
During the PT inspection of the distillation tower, we found numerous cracks in the welds of the tower body, manholes, connections, insulation rings, and other areas that are welded to the tower body. The tower is 75 meters tall, with an inner diameter of φ1600 and wall thicknesses of 30/22/16/10. The operating temperature ranges from 90 to 100°C, the pressure is 0.6 MPa, and the material used is 304. The substances contained within the tower are trichlorosilane, dichlorodihydrosilane, and silicon tetrachloride. The preliminary cause of the crack is 304 stainless steel, whose resistance to chloride corrosion is relatively low. When the unit is under maintenance or a leak occurs, a large amount of chloride ions in the smoke accumulate within the insulation layer of the distillation tower. These chloride ions, attached to the insulation layer, cause corrosion to the metal surface of the tower’s exterior due to the effect of rainwater. Rainwater containing chloride ions accumulates for long periods in areas with steps (such as above the insulation support rings, manholes, connectors, and above flanges); corrosion in these areas is significantly more severe, and stress corrosion cracks occur as a result of welding residual stresses and stresses generated during the operation of the tower. We would like to conduct further analysis of the cracks in order to take preventive measures when designing and manufacturing new towers. From which aspects does the analysis need to be conducted? Is there a way to conduct stress testing on the key parts of the tower on-site? Are there any organizations that work in this area? Is it possible to conduct a qualitative analysis of corrosion without taking samples? I consulted experts here.
You’ve analyzed the reasons so clearly that I’m not sure what else needs to be analyzed. I’m really interested in knowing what solutions you plan to take.
This post was last edited by kareale88 on 2019-3-7 09:24. Corrosion and water resistance – I want to analyze the stress levels in those areas where cracks have appeared
Stress is one of the causes of cracks, but the magnitude of stress is not; I still want to know what measures you plan to take to address this.
It was detected in time, and a new tower was installed promptly; this represents a serious safety hazard. Is it inside the tower or outside?
You’ve analyzed the reasons pretty well; I wonder how you plan to fix it? Replacing it with a new one is costly and troublesome; not replacing it poses too great a risk
I encountered a similar crack in the transverse weld of a tower during an inspection I did on one tower before; around the weld joint where the tower body meets the lower head, the density of cracks was even higher than in the image you showed. We thought it was polyoxysulfate stress corrosion cracking at that time. It was caused by improper parking procedures; after polishing it clean, it was monitored and used for half a year.
It indeed resembles chloride stress corrosion cracking very much. But if you try to make a qualitative judgment without taking samples, it is still a bit too arbitrary. Furthermore, since it is known that 300 series stainless steels are sensitive to chlorides, why choose 300 series stainless steels then? The problem of 300 series stainless steel with chlorides is a defect inherent in the material itself, and it cannot be avoided through manufacturing or maintenance measures. Don’t say that you can’t carry out heat treatment on such large towers; even if it’s possible to conduct solution treatment after the equipment is built, ignoring the issue of high-temperature deformation, this can only prevent intergranular corrosion, but stress corrosion cracking still cannot be avoided.
So I’m wondering if there’s any way to detect stress online? The equipment hasn’t been scrapped yet, but it’s difficult to take samples
What is the use of online stress monitoring? What is the purpose? The magnitude of stress doesn’t tell us much. Moreover, the device behaves completely differently when it is in operation and when it is not.