Study on the characteristics of high temperature steam oxidation layer of T91 and TP347H steel pipes
In order to realize the energy saving and consumption reduction of thermal power units, the unit parameters continue to improve [1,2,3], the superheater and reheater should operate in the water steam environment with higher temperature and pressure, and the subsequent oxidation problems become more and more prominent [4,5,6,7,8]. In recent years, the economic and safety accidents caused by the rapid growth of oxide layer and a large number of peeling have become more and more serious [9, 10, 11]. Its harm involves the following four aspects:
- (1) The formation of oxide layer causes the decrease of effective wall thickness of steel pipe, the increase of actual working temperature, the decrease of anti creep ability and the shortening of service life of steel pipe;
- (2) The large amount of oxide layer peels off and accumulates, which reduces the flow cross section in the pipe, causes the increase of flow resistance, the decrease of steam flow volume and the local overtemperature of the pipe wall, thus causing the explosion of the pipe;
- (3) The peeled oxide layer With the steam flow carried into the turbine, the nozzle, blade and diaphragm are eroded to reduce the output and efficiency of the turbine;
- (4) The stripped oxide layer is easy to cause the main steam valve to jam, resulting in the main steam valve can not be closed normally when the unit is shut down. Therefore, it is very urgent and important to master the characteristics of various types of steel pipe steam oxide layer.
In this paper, a 350MW supercritical unit (water supply is treated by AVT (o), running about 9.8a) T91 high strength martensitic heat-resistant steel and TP347H austenitic stainless steel final superheater tube are taken as the objects to discuss and compare the composition structure, morphology characteristics and formation rules of the oxide layer formed when the two typical materials are in the same elevation, working condition and time, so as to effectively control the speed of oxide layer Rapid growth and stripping provide technical reserves and basis.
Experimental method
By machining the cut steel pipe sample on a lathe, a pipe section with a length of about 40 mm and a wall thickness of about 1.5 mm is obtained. Acetone and ethanol are used to clean the steel pipe samples to remove the oil and other impurities on the inner surface of the steel pipe samples. Cut and saw 10 mm × 10 mm test piece for phase analysis. The test piece of 10 mm × 10 mm is cut and sawed, sealed with conductive resin, grinded and polished until there is no obvious scratch to make a metallographic sample for metallographic analysis.
The phase composition of oxide layer was analyzed by Ultima IV type X-ray diffraction (XRD). The microstructure of oxide layer was analyzed by xjp-6a digital metallographic microscope. The microstructure and element distribution of the oxide layer were analyzed by scanning electron microscopy (SEM) of Fei quanta 400 with EDs.
Results and discussion
Phase composition of oxide layer
Figure 1 shows the X-ray diffraction pattern of the steam oxide layer of T91 and TP347H steel pipes. It can be seen from the figure that the characteristic diffraction peaks of each curve match the standard cards (jcpds-no.01-085-0599, jcpds-no.01-079-0418), indicating that the main phase compositions of each sample are Fe2O3 and Fe3O4. However, the diffracted peak intensities of the two groups of curves are quite different, which indicates that the composition proportions of Fe2O3 and Fe3O4 in the samples are different. Compared with T91, the diffraction peaks of TP347H curve are obviously enhanced at 2 θ = 24.2o, 33.2o, 35.6o, 40.8o, 49.5o and 54.0o, indicating that Fe2O3 accounts for more in the oxide layer of TP347H steel tube.
Fig.1 XRD patterns of T91/TP347H tubes oxide scale
Morphology and structure of oxide layer
Macroscopic appearance
Fig. 2a and B are the macromorphology of the steam oxide layer of T91 and TP347H steel pipe samples, respectively. It can be seen from the figure that the steam oxide layer of the steel tube sample is steel gray. The difference is that the oxidation layer of T91 steel tube sample appears the phenomenon of spot peeling with different sizes; the oxidation layer of TP347H steel tube sample remains relatively complete without obvious peeling signs.
Fig.2 Macroscopic images of tubes oxide scale: (a) T91; (b) TP347H
Micro morphology
Fig. 3a and Fig. B are the optical micrographs of the vapor oxide layer sections of T91 and TP347H steel tube samples, respectively. It can be seen from the figure that the oxide layer thickness distribution of T91 steel pipe sample is relatively uniform, and the average thickness exceeds 0.20 mm critical stripping and appropriate cleaning thickness specified in t / CEC 144-2017 chemical cleaning guide for superheater and reheater. The oxide layer can be divided into: mottled and uneven outer layer with embedded distribution, thick and relatively uniform middle layer with continuous distribution, thin and uneven inner layer with nodal distribution (near the substrate side). The boundary between the outer layer and the middle layer is fuzzy, there are loose holes in the middle layer, and there are continuous and clear microcracks at the boundary between the middle layer and the inner layer. These holes / cracks are almost parallel to the horizontal surface.
Fig.3 Optical microscope images of tubes oxide scale: (a) T91; (b) TP347H
The thickness distribution of the oxide layer of TP347H steel tube sample is not uniform, and the thickest value is different from the thinnest value. The maximum thickness is 0.165 mm, which exceeds the critical peeling and cleaning thickness of 0.09 mm specified in t / CEC 144-2017. The oxide layer can be divided into thin, uniform and continuous outer layer and thick, uneven and continuous inner layer. The boundary between layers is clear and well combined.
The results show that the structure and morphology of the steam oxide layer of T91 and TP347H steel tube samples are obviously different. The main features are as follows: (1) the two steel tube samples have the oxide layer (often called Fe2O3 layer), but TP347H oxide layer has more continuous distribution, more complete growth, more uniform distribution and clearer interface than T91. In view of the consistence of the in-service elevation, working condition and time of the two steel pipe samples, it can be seen that the material of the steel pipe samples has a significant influence on the growth of Fe2O3 layer. (2) Although the Fe2O3 layer of T91 steel tube sample is less than that of TP347H steel tube sample, the oxide layer of T91 steel tube sample has peeling signs, and the TP347H steel tube sample remains complete. It can be seen from the analysis that the occurrence and content of Fe2O3 as the judgment basis for the peeling tendency of oxide layer is not universal [12]. (3) There are a lot of holes or microcracks in T91 oxide layer, which may form larger size cracks. With the frequent start and stop of the boiler and the sudden change of load, in addition to the different sizes of stress between the oxide layer and the metal matrix, and between the oxide layer and the cracks, there will also be a large range of fluctuations, which will lead to the peeling of the oxide layer, and these cracks are often the locations where the peeling of the oxide layer is induced [13,14]. The analysis shows that the existence of a large number of holes or cracks will increase the risk of oxide layer peeling, which can be used as one of the important signals to predict oxide layer peeling.
Element distribution of oxide layer
2.3.1 T91
Figure 4A and table 1 show the backscattered electron image of the vapor oxide cross section of T91 steel tube and EDS results of each region. It can be seen from the chart that the main elements of the outer oxide layer (positions 1 and 2) are Fe and O; when the outer oxide layer transits to the middle layer, the Cr content increases sharply, and the content of the middle layer (positions 3, 4, 5 and 7) is basically kept in the range of 11.42% ~ 15.78%. Meanwhile, a small amount of Mn and Mo are also found in the layer.
Fig.4 Backscattred electron images of tubes oxide scale: (a) T91; (b) partila enlarged datail
Table 1 EDS analysis result of the regions in Fig.4(a) (mass fraction / %)
| Area | O | Na | Si | K | Ca | Cr | Mn | Fe | Mo |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 35.07 | — | 0.05 | — | — | 0.04 | 0.02 | 64.82 | — |
| 2 | 32.99 | — | — | — | — | 0.13 | 0.16 | 66.72 | — |
| 3 | 33.47 | — | 0.13 | — | — | 11.42 | 0.76 | 54.10 | 0.11 |
| 4 | 33.39 | — | 0.44 | — | — | 12.10 | 0.62 | 51.97 | 1.48 |
| 5 | 33.08 | — | 0.53 | — | — | 14.00 | 0.70 | 50.32 | 1.36 |
| 6 | 34.15 | 3.22 | 0.98 | 0.85 | 2.34 | 13.91 | 1.71 | 42.20 | 0.63 |
| 7 | 32.99 | — | 0.65 | — | — | 15.78 | 0.44 | 48.60 | 1.54 |
| 8 | — | — | 0.41 | — | — | 9.47 | 0.41 | 88.53 | 1.17 |
In order to further explore the distribution of elements in the inner layer of oxidation, the area is enlarged as shown in Fig. 4B. It can be seen from table 2 that the Cr content in this area (positions 2 and 3) increased to 22.91% and 24.78% respectively, which is significantly higher than that in the oxidation intermediate layer, i.e. a discontinuous nodal chromium rich area was formed, which to some extent hindered the inward diffusion of O or the outward diffusion channels and speed of Fe, thus slowing down the growth of new oxides [15, 16, 17]. However, due to the intermittent and thin chromium rich area, its blocking effect was weak.
Table 2 EDS analysis result of regions in Fig.4(b) (mass fraction / %)
| Area | O | Si | Cr | Mn | Fe | Mo |
|---|---|---|---|---|---|---|
| 1 | 33.24 | 0.67 | 16.34 | 0.41 | 47.67 | 1.67 |
| 2 | 25.93 | 0.34 | 22.91 | 0.80 | 47.55 | 2.47 |
| 3 | 25.88 | 0.20 | 24.78 | 0.76 | 46.71 | 2.67 |
| 4 | — | 0.39 | 9.19 | 0.46 | 88.81 | 1.15 |
Fig. 5 is the EDS line scan result of the oxide layer section of T91 steel tube sample. It can be seen from the figure that the change trend of Mo and Mn content is basically the same, and there is no obvious floating, indicating that Mo and Mn basically have no outward diffusion movement. The content of Cr in the inner layer is unstable and fluctuated obviously, which may be caused by more pores. In Fig. 5, all element content spectra show a sharp drop, which is the microcrack near the boundary between the middle layer and the inner layer. However, at the position below the crack, the elements basically returned to the original content range.
Fig.5 EDS scanning line analysis images of T91 tubes oxide scale: (a) O; (b) Fe; (c) Cr; (d) Mn; (e) Mo
2.3.2 TP347H
Figure 6 and table 3 show the backscattered electron image of the oxide cross section of TP347H steel tube sample and EDS results of each region. It can be seen from the chart that the main elements of the outer oxide layer (positions 1 and 2) are Fe and O. the closer the outer oxide layer is to the metal matrix (positions 3 and 4), the uneven oxidation front occurs, and the Cr content increases sharply to 25.41%; meanwhile, the Ni content also increases to 15.50%. It is worth noting that the Cr content at the junction (position 5) of the oxide layer and the matrix increases to 43.53%. Compared with T91 steel tube samples, the chromium rich zone is more widely distributed and can wrap the oxide layer inside, which is conducive to blocking the diffusion movement of Fe and O and effectively slowing down the deep oxidation.
Fig.6 Backscattred electron image of TP347H tubes oxide scale
Table3 EDS analysis result of regions in Fig.6 (mass fraction / %)
| Area | O | Si | Cr | Mn | Fe | Ni |
|---|---|---|---|---|---|---|
| 1 | 32.68 | 0.19 | 0.30 | 0.21 | 66.62 | — |
| 2 | 31.01 | 0.18 | 0.65 | 0.75 | 66.61 | 0.80 |
| 3 | 31.82 | 0.53 | 19.53 | 1.61 | 37.67 | 8.83 |
| 4 | 27.46 | 0.36 | 25.41 | 2.52 | 28.76 | 15.50 |
| 5 | 34.33 | 0.60 | 43.53 | 3.16 | 17.16 | 1.22 |
| 6 | — | 0.35 | 18.15 | 2.01 | 70.39 | 9.09 |
Fig. 7 is EDS line scan result of oxidation layer section of TP347H steel tube sample. It can be seen from the figure that the change trend of Mn and Ni content is basically the same, and the inner layer of the oxide layer increases slightly. When it is close to the inner oxide layer, the Cr content increases sharply, and when it is close to the critical position, it drops sharply, that is, there is a CR poor zone [18], and then there is a short small peak (corresponding to the results in Fig. 6 and table 3). It can be seen from the analysis that when the operation time of the unit is long and the structure of the oxide layer tends to be uniform and complete, the chromium rich belt will effectively block part of the diffusion channels of Fe and O, reduce the oxidation speed and prolong the oxidation time; however, the existence of the chromium poor belt is often regarded as the weak area, which is the priority channel for element diffusion. As the rich CR belt is thicker and more effective than the poor CR belt, the overall oxidation rate shows a slower trend, that is, with the continuous running time of the unit, the oxidation rate shows a downward trend.
Fig.7 EDS scanning line analysis images of TP347H tubes oxide scale: (a) O; (b) Fe; (c) Cr; (d) Mn; (e) Ni
The results show that there are obvious differences between T91 and TP347H steel tube samples in the morphology, structure and element distribution, mainly as follows: there are chromium rich areas at the junction of the oxide layer and the matrix of the two steel tube samples, but the continuity of the chromium rich zone of TP347H steel tube samples is better, and there is a very thin chromium poor zone near the junction. The oxidation layer of TP347H steel tube is wrapped in the chromium rich protective belt along the junction of the matrix, which can inhibit the rapid diffusion of Fe, Cr, O and other elements, and then effectively control the deep oxidation, which is one of the reasons for the relatively thinner oxide skin of TP347H steel tube.
Conclusion
- (1) There are obvious differences in phase composition, morphology and element distribution between T91 and TP347H.
- (2) The appearance and content of Fe2O3 as a sign of the peeling behavior of steel tube oxide layer is not universal.
- (3) The oxidation layer of T91 steel pipe is easy to produce holes or microcracks, which increases the risk of oxide layer peeling.
- (4) TP347H steel tube is easy to form a continuous and complete chromium rich band, which can hinder the diffusion channel and rate of elements and inhibit the rapid growth of oxide layer.
Source: China TP347H Steel Pipe Manufacturer – Yaang Pipe Industry Co., Limited (www.pipelinedubai.com)
(Yaang Pipe Industry is a leading manufacturer and supplier of nickel alloy and stainless steel products, including Super Duplex Stainless Steel Flanges, Stainless Steel Flanges, Stainless Steel Pipe Fittings, Stainless Steel Pipe. Yaang products are widely used in Shipbuilding, Nuclear power, Marine engineering, Petroleum, Chemical, Mining, Sewage treatment, Natural gas and Pressure vessels and other industries.)
If you want to have more information about the article or you want to share your opinion with us, contact us at [email protected]
Please notice that you might be interested in the other technical articles we’ve published:
- How to get high quality steel pipes
- What is the difference between a steel pipe and a steel tube
- How to get high quality stainless steel pipes
- How to get high quality carbon steel pipes
- How To Distinguish Inferior Steel Pipe
- How to get high quality alloy steel pipes
- What is a steel pipe
- What Is A LSAW Steel Pipe
- What is a spiral steel pipe
- What is a SSAW pipe
- Identification Method of Seamless Steel Pipe and Welded Steel Pipe
Reference:
- [1] Tang F, Dong B, Zhao M.USC unit development and application in China [J]. Electr. Power Constr., 2010, 31(1): 80
- [2] Xu T M, Yuan Y C, Chen G J, et al.Developmental trend of super-large capacity and ultra supercritical boilers [J]. Power Eng., 2010, 23: 2363
- [3] Liu J R.The analysis and optimization program about energy-saving potential of ultra-supercritical boiler unit [D].Beijing: North China Electric Power University (Beijing), 2010
- [4] Kritzer P.Corrosion in high-temperature and supercritical water and aqueous solutions: A review [J]. J. Supercrit. Fluids, 2004, 29: 1
- [5] Zhang D Q, Xu J J, Zhao G Q, et al.Oxidation characteristic of ferritic-martensitic steel T91 in water-vapour atmosphere [J]. Chin. J. Mater. Res., 2008, 22: 599
- [6] Li Z G.Study on high temperature oxidation of water vapor[R] . Xi’an: Xi’an Thermal Power Research Institute Co. Ltd., 2006
- [7] Liu J, Ding S F.Evaluation research on iron scale generation of high temperature heat surfaces in supercritical boiler[R] . Shanghai: Shanghai Power Equipment Research Institute, 2013
- [8] Xia T F, Zhang D Q, Gao L X.Advances in high temperature steam oxidation behavior and control technology of heat resistant steel [J]. Corros. Sci. Prot. Technol., 2015, 27: 199
- [9] Meng X M, Zhang L, Lai Y T, et al.Analysis on bursting reason of super-heater tubes of a supercritical boiler [J]. Phys. Test. Chem. Anal., 2015, 51A: 353
- [10] Huang W, Li Y Q, Xiong W L, et al.Invalidation analysis of final stage super heater on super critical boiler [J]. Power Syst. Eng., 2008, 24(4): 32
- [11] Wright I G, Dooley R B.A review of the oxidation behaviour of structural alloys in steam [J]. Int. Mater. Rev., 2010, 55: 129
- [12] Jia J M, Chen J G, Li Z G, et al.Countermeasures against massive exfoliation of oxidation scale on the internal surface of coarse grained 18-8 type stainless steel boiler tubes [J]. Electr. Power, 2008, 41(5): 37
- [13] Yuan J T, Wang W, Zhu S L, et al.Oxidation behavior of super 304H steel in steam at 700~900 ℃ [J]. J. Chin. Soc. Corros. Prot., 2014, 34: 218
- [14] Chen Y, Sridharan K, Allen T.Corrosion behavior of ferritic-martensitic steel T91 in supercritical water
- [15] Huang X D, Zhou X Y, You Z, et al.Oxide scale growth and exfoliation behavior on high temperature heat-absorbing surface exposed to steam for supercritical (Ultrasupercritical) boilers [J]. J. Power Eng., 2009, 29: 602
- [16] Robertson J.The mechanism of high temperature aqueous corrosion of stainless steels [J]. Corros. Sci., 1991, 32: 443
- [17] Guo Y, Jia J M, Hou S F, et al.Steam oxidation behavior of domestic TP347H FG steel [J]. Corros. Sci. Prot. Technol., 2011, 23: 505
- [18] Tang L Y, Jia J M, Cai H, et al.Microstructure and growth mechanism of the oxide scale formed on HR3C at the steam side in ultra supercritical (USC) units [J]. Corros. Prot., 2011, 32: 775
- [18] GUO Dan, ZHANG Weike, PENG Zhanghua, SHI Qinglin, WANG Mengjiao, YI Zengqiang. Characteristics of Oxide Scales Formed on T91 Martensite Steel and TP347H Austenite Steel in HighTemperature Steam [J]. Corrosion Science and Protection Technology, 2019, 31(6): 609-614 doi:10.11903/1002.6495.2019.059