Failure of Carbon Steel Superheater Tubes

Joseph P. Ribble, Betz Laboratories, Inc.


From: Handbook of Case Histories in Failure Analysis, Vol 2, K.A. Esakul, Ed., ASM International, 1992

Abstract: Two superheater tubes from a 6.2 MPa (900 psig) boiler failed in service because of creep rupture. One tube was carbon steel and the other was carbon steel welded to ASTM A213 Grade T22 (2.25Cr-1.0Mo) tubing. The failure in the welded tube occurred in the carbon steel section. Portions of the superheater were retubed five years previously with Grade 722 material. The failures indicated that tubes were exposed to long-term overheating conditions. While the carbon steel tube did not experience temperatures above the lower transformation temperature 727 deg C (1340 deg F), the welded tube did experience a temperature peak in excess of 727 deg C (1340 deg F). The long-term overheating conditions could have been the result of excessive heat flux and /or inadequate steam flow. In addition, the entire superheater bank should have been upgraded to Grade 722 material at the time of retubing.

Keywords: Boiler tubes; Creep rupture; Mechanical properties; Overheating

Materials: 2.25Cr-1Mo (Chromium-molybdenum alloy steel), UNS K21590; ASTM A213 grade T22 (Chromium-molybdenum alloy steel), UNS K21590; ASTM A192 (Nonresulfurized carbon steel), UNS K01201

Failure type: Creep fracture/stress rupture


Background

Two tubes from the superheater section of a two-drum boiler failed while in service. The tubes were identified as being the “old tube” (tube 1) and the “new tube” (tube 2). Tube 1 was an original equipment tube and was in service for 19 years. Tube 2 was in service for approximately five years.

Applications

The superheater tubes had different nominal outer diameters and nominal wall thicknesses. Tubes 1 and 2 measured 44 mm (1.75 in.) and 50.8 mm (2.00 in.) in diameter, respectively. The wall thicknesses of tube 1 and the failed end of tube 2 the measured 4.5 mm (0.176 in.) and 3.8 mm (0.150 in.) thick, respectively. The wall thickness at the end of tube 2 opposite the failure measured 6.0 mm (0.235 in.). These tubes were steam bearing tubes for a 6.2 MPa (900 psig) boiler.

Pertinent specifications

Tube 1 and the failed end of tube 2 were fabricated from low-carbon steel consistent with ASTM A192 specifications. The non-failed end of tube 2 was fabricated from ASTM A123 Grade T22 seamless ferritic alloy steel superheater tubing.

Performance of other parts in the same or similar service

Several other superheater tubes in a boiler of identical design and manufacturer, at the same plant, had failed two years prior to this failure by the same apparent mechanism. Portions of the superheater in the present boiler were retubed five years ago. The two submitted tubes were in parallel in the superheater section at the time of failure.

Selection of specimens

Two selections approximately 406 mm (16 in.) long were submitted for laboratory examination.

Visual Examination and General Physical Features

The external surface of the superheater tubes is shown in Fig. 1. The failures in tube 1 consisted of two ruptures in the middle of localized bulges. The two failure sites are indicated as “a” and “b” in Fig. 1. The ruptures were oriented approximately 90° apart with respect to the tube circumference, and were separated by approximately 70 mm (2.75 in.). The tube 2 sample consisted of two tubes welded together with the failure entirely within the carbon steel portion. The tube 2 failure was a thin-lipped rupture displaying a high degree of deformation. The low-alloy portion of tube 2 had no failures associated with it.
Figure 1

Fig. 1  The submitted super heater tube sections. 0.23×

Testing Procedure and Results

Surface examination

The tube 1 outer diameter surface exhibited two ruptures parallel to its longitudinal aids. Wall thinning to 0.76 mm (0.030 in.) and secondary creep cracks were observed at both tube 1 failure sites (Fig. 2 and 3). Secondary creep cracks and wall thinning to 0.51 mm (0.020 in.) were observed at the tube 2 failure lip (Fig. 4). The internal surface of tube 1 was covered with a thin layer of yellow-brown deposits over a tightly adherent black scale. The internal surface of tube 2 was covered with a thin layer of tightly adherent black magnetite scale.
Figure 2

Fig. 2  Tube 2 failure site a (see Fig. 1). Note the heavy oxidation and longitudinal secondary creep cracks. 0.60×

Figure 3

Fig. 3  Tube 1 failure site b (see Fig. 1). Note the heavy oxidation and longitudinal secondary creep cracks. 0.60×

Figure 4

Fig. 4  The failure edge of tube 2. Note the secondary creep cracks. 1.06×

Metallography

Representative sections from select areas of both tubes were prepared for metallographic examination. Creep voids were observed along the fracture lip b of tube 1 (Fig. 5). The microstructure observed at the mid-wall location of the tube 1 b failure and opposite failure a was in a very advanced stage of spheroidization (Fig. 6 and 7). In addition, isolated graphite nodules were also observed in the microstructure.
Figure 5

Fig. 5  Photomicrograph of the b failure lip in tube 1. Note the creep voids along the fracture surface. Nital etch, 38×

Figure 6

Fig. 6  Photomicrograph of the mid-wall microstructure observed in the tube 1 b failure area shown in Fig. 5. Note the spheroidization and large graphite nodule. Nital etch, 285×

Figure 7

Fig. 7  Photomicrograph showing the mid-wall microstructure opposite the failure of Fig. 5 and 6. Note the spheroidization. Nital etch, 285×

Elongated grains were observed at the failure lip of tube 2 (Fig. 8). The microstructure at the midwall location of the tube 2 failure consisted of ferrite and reformed pearlite (Fig. 9). The presence of transformation products (reformed pearlite) at the failure lip indicated that this portion of tube 2 experienced temperatures above its lower transformation temperature, 727 °C (1340 °F), at the time of failure. Opposite the tube 2 failure, the mid-wall microstructure consisted of spheroidal carbides and occasional islands of pearlite in a ferrite matrix (Fig. 10). The mid-wall microstructure observed in the low-alloy portion of tube 2 that did not fail consisted of a fine dispersion of spheroidal carbides (Fig. 11).
Figure 8

Fig. 8  Photomicrograph of the microstructure taken at the failure lip of tube 2. Note the creep voids and elongated grains. Nital etch, 57×

Figure 9

Fig. 9  Photomicrograph of the mid-wall microstructure at tube 2 failure lip. Note the reformed pearlite in the ferrite matrix. Nital etch, 285×

Figure 10

Fig. 10  Photomicrograph of the mid-wall microstructure opposite the tube 2 failure. Note the spheroidal carbides and pearlite islands. Nital etch, 285×

Figure 11

Fig. 11  Photomicrograph of the tube 2 mid-wall microstructure on the failure side, in the non-failed portion opposite the weld. Nital etch, 285×

Chemical analysis/identification

Material analysis. Portions of tube 1 and 2 from either side of the weld joint were tested for chemical analysis using optical emission spectroscopy (OES) to determine the alloy composition. The results (Table 1) confirmed that tube 1 and the portion of tube 2 containing the failure were fabricated from low-carbon steel consistent with ASTM A192 tubing specification. The similarity of the chemical composition of the failed tube sections suggest they were from the same heat of steel. Results of the alloy analysis on the non-failed portion of tube 2 (Table 2) indicated that it was fabricated from ASTM A213 Grade T22 tubing.

Table 1   Alloy analyses of failed tube sections

Element
Composition, %
Tube 1
Tube 2
ASTM A192 specification
C
0.11
0.12
0.06 to 0.18
Mn
0.49
0.46
0.27 to 0.63
P
0.006
<0.005
0.035 max
S
0.026
0.25
0.035 max
Si
0.02
0.01
0.25 max
Cr
0.03
0.03
ns
Ni
0.07
0.07
ns
Mo
0.02
0.02
ns
Cu
0.14
0.11
ns
Fe
rem
rem
rem
ns, not specified

Table 2   Alloy analysis of not-failed section of tube 2

Element
Composition, %
Non-failed section
ASTM A213 Grade T22 specification
C
0.11
0.15 max
Mn
0.49
0.30 to 0.60
P
0.016
0.30 max
S
0.017
0.30 max
Si
0.27
0.50 max
Cr
2.30
1.90 to 2.60
Ni
0.22
ns
Mo
1.02
0.87 to 1.13
Cu
0.09
ns
Fe
rem
rem
ns, not specified
Internal deposit analysis. Scanning electron microscope-energy dispersive spectroscopy (SEM-EDS) was used to determine the elemental composition of the internal deposits on both tubes.
The results (Table 3) revealed significant quantities of calcium, silicon and sodium species in tube 1. The same species were also present in tube 2 but in smaller quantities. This indicated that some carryover of boiler water was taking place in the superheater section.

Table 3   Analyses of the internal deposits

Element
Composition, %
Tube 1
Tube 2
Fe
47.1
83.5
Ca
22.2
5.5
Si
16.6
4.5
Na
9.8
4.6
Cl
0.7
S
1.0
<0.5
Al
1.7
Mn
0.9
Cr
<0.5
Mg
1.6

Discussion

Heavy oxidation and longitudinal cracking parallel to the failures in both tubes indicated a creep rupture mechanism. Chemical analysis revealed that tube 1 and the failed section of tube 2 were fabricated from plain carbon steel. The non-failed side of tube 2 was fabricated from ASTM A213 Grade T22 low-alloy steel tubing. In steam environments, the generally accepted oxidation limit for plain carbon steel is 454 °C (850 °F), while for Grade T22 it is approximately 580 °C (1075 °F). A spheroidized microstructure with graphite nodules indicated that tube 1 was operating at a temperature significantly above its oxidation limit and below its lower transformation temperature of 727 °C (1340 °F) for an extended period of time.
Transformation products in the failed region of tube 2 indicated that the failure region experienced a peak temperature in excess of 727 °C (1340 °F). However, secondary creep cracks around the failure and creep voids along the rupture lip indicated that it experienced long-term overheating effects prior to rupture. The presence of a transformed microstructure in the carbon steel portion of tube 2 made an assessment of the degree of long-term overheating impossible. The superior oxidation resistance and microstructural stability of Grade T22 tubing, via its chromium and molybdenum additions, resulted in the virtual absence of significant scale and fine dispersion of spheroidal carbides in the microstructure

Conclusion and Recommendations

Most probable cause

The tube 1 failure was caused by creep mechanism via long-term overheating. Microstructural analysis indicated that the tube 1 peak metal temperatures were greater than 454 °C (850 °F) but less than 727 °C (1340 °F). Secondary creep cracking indicates long-term overheating conditions were present prior to the failure of tube 2. However, microstructural analysis indicated that tube 2 experienced a brief high-temperature excursion greater than 727 °C (1340 °F) at the time of failure. The long-term overheating of tubes 1 and 2 was likely caused by an excessive heat flux and/or insufficient steam distribution. The high-temperature event coinciding with the violent rupture of tube 2 most likely resulted from a sudden loss in steam distribution through this tube circuit. Although SEM-EDS analysis identified some carryover deposits on the internal surfaces, only a small amount of transported deposit was present. Internal deposits were not a primary cause of the tube failures.

Remedial action

A review of steam flow in the superheater should be conducted to ensure adequate steam distribution in the circuit. The entire superheater should be upgraded to ASTM A213 Grade T22 tubing.

How failure could have been prevented

Alloy analyses of the submitted tube samples indicates the overheating failures occurred in low-carbon steel sections. The portion of tube 2 that did not contain the failure tested as being consistent with ASTM A213 Grade T22 boiler tubing specification. The similarity of the chemical composition of the failed sections suggests these sections were installed at the same time. When the superheater was retubed, the entire superheater assembly should have been upgraded to Grade T22 material.

Related Information

Creep and Stress Rupture Failures, Failure Analysis and Prevention, Vol 11, ASM Handbook, ASM International, 2002, p 728–737
D.N. French, Failures of Boilers and Related Equipment, Failure Analysis and Prevention, Vol 11, ASM Handbook, ASM International, 1986, p 602–627