Corrosion, Failure Analysis and Materials Selection Specialists

 

 

 

 

 

Corrosion Testing / Failure Analysis

 

 

 

Corrosion Testing

Failure Analysis

Field Investigations

Litigation

Metallography

Technical Papers

CTL Profile

Pricing & Policies

Contact CTL

Quality Assurance

Return to Failure Analysis Case Histories

Polythionic Stress Corrosion Cracking of Stainless Steel Tube

ENVIRONMENT:

Oil Refinery

EQUIPMENT:

Heater

MATERIAL:

 Type 316H Stainless Steel
SERVICE TIME: Seven years

FAILURE MODE:

Polythionic Stress Corrosion Cracking

 

Background

A hydrotest at 735 psi revealed numerous leaks in four tubes from a heater. All of the leaks were observed to be on top or fireside of the tubes.  

The heater was newly installed seven years ago. The design tube metal temperature was 1000F.  In the past, tube metal temperature was measured in the range of 1200 to 1400F during the past run in areas of coke deposits and less than 900F in areas where there were no coke deposits. The tubes were reported to be Type 316H stainless steel with minimum molybdenum content of 2.5%. 

The tubes had been pigged and flushed with a soda ash treatment to remove coke deposits and inhibit the formation of polythionic acids. 

 

Findings

Visual examination of the tube revealed two areas on the outside diameter (OD) where leaks had been observed. Both areas had been ground to reveal the underlying base metal. On the inside diameter (ID) surface a longitudinal strip of coke deposits approximately 2 -inch wide was present along the entire length of the tube sample. The leaks identified on the OD were within the area of these deposits.  

The outside and inside diameters were measured with a digital caliper to determine if the tube was out of round. The OD measurements determined that the tube was approximately 0.045 out of round. The largest OD measurement was obtained in the area of the coke deposits. The smallest OD measurement was obtained 90 to the coke deposit. The ID measurements were all 6.050 0.003. A portion of the coke was chipped off and the thickness measured to be from 0.040 to 0.050. 

An approximately 10 long section was removed which included the two identified leaks. This section was split longitudinally to reveal the ID surface. The section represented by Figure 1 indicates the ID of the leak in Area A of Figure 1. This portion was cut out and used for cross-sectional analysis. The remaining parts of this longitudinal section were cleaned and polished with abrasive grinding to a 240 grit finish on the ID surface. The polished surfaces were tested using liquid dye penetrant. One crack was identified, Figure 2. This crack was located underneath the coke deposits associated with known leak. 

 

 

Figure 1. Close up of apparent crack on ID surface at leak indication.

Figure 2.  Dye penetrant test on ID crack.

   
Figure 3. Cross-section of crack , as polished (20X Original Magnification)

 

When the tube was cut longitudinally, it was observed that substantial hoop stresses were present in the tube. When the longitudinal cut was finished the tube popped open increasing in diameter from 6.611 to 6.750. 

 

A metallographic cross-section of the crack in Figure 2 was prepared. Three separate cracks were observed in this area. The polished cross-section revealed that the cracks had initiated on the ID surface and appeared to have propagated intergranularly, Figure 3. The cracks were relatively narrow with no significant branching. No indications of creep voids were observed.

The polished cross section was placed into a JEOL JSM 35 scanning electron microscope equipped with Energy Dispersive x-ray Spectroscopy (EDS). An EDS analysis of the crack tip did not reveal any indication of the presence of chloride or sulfur compounds. 

The cross-section was electrolytically etched in oxalic acid and the microstructure viewed at up to 500X magnification. The microstructure observed was typical of sensitized austenitic stainless steel with deep ditches present. The intergranular nature of the cracks was confirmed.

Chemical analysis by OES (optical emission spectroscopy) confirmed the alloy to be within the chemical composition for 316H stainless steel.

 

Discussion

Based on this analysis the leaks were caused by cracks that initiated on the ID surface. All of the cracks observed were found underneath the coke deposits. The cracks are intergranular and narrow with no significant branching. These characteristics are consistent with stress corrosion cracking due to polythionic acids.  

Site Index

Site Copyright 1995 - 2007, All Rights Reserved,

Corrosion Testing Laboratories, Inc.

60 Blue Hen Drive

Newark, Delaware USA 19713

Phone: 1-302-454-8200

Fax: 1-302-454-8204

web@corrosionlab.com