Y Gong, Z-G Yang/ Materials and Design 32(2011)671-681 800 600 f叫 w V NW 2/° Fig.14.XRD results of the solid corrosion product (a) the black one and(b)the yellow e one. 3. 23. SEM and EDs Through optical microscope, it can be learned from Fig. 16a that 90 two different morphologies existed on the cross-section of the pe 190℃,829) foration, i.e. the bright silver one and the gray one. In order to fur ther oscopic morphologies ompositions, SEM and edS were employed In Fig. 16b, the bright silver one was compacted while the grey one was pitted. By means of EDS. it was determined that the former one was the matrix me- tal of casting aluminum, and the latter one was the corrosion prod uct containing a high content of sulfur element, seen in Fig. 16c and d and Table 5. This was in accordance with the IC, xrd and XRF re- 280℃45%0) sults, and further identified that causes of the perforation on the manhole door was concerned with sulfur related corrosion Fig. 15. TGA result of the black corrosion 3.3. Air preheater By means of XRF, element compositions of the brown rust accu- mulating on the bottom of the air preheater was determined in Ta- In this step, ferrous sulfate hydrates that performed lik ble 6. In it, Fe and o were the two predominant elements, and green liquid corrosion products dehydrated the crystal consequently it can be confirmed that the rust was iron oxides. seen in Eq.(1)[28, and the weight loss of them was In terms of the concrete compounds, the XRD results displayed in 18% at 190C in Fig. 15 Fig. 17 showed that ferroferric oxide (Fe3 O4) and ferric oxide FeSOg-6H20- FeSO4.4H20+2H20 (1) (Fe2O3)were the two primary ones among the many kinds, which were usually the ultimate products of uniform corrosion[29]. FeSO4. 4H,0 FeSO4. H20+3H20 Step two(190°C~280°): decomposition 4 Discussion After being dehydrated in step one, the ferrous sulfate hydrates 4.1. Fracture of nozzle hen decomposed in this step, particularly for the product Fes O4. H2O. The primary reactions mainly consisted of two succes- As is shown in Fig. 2b that perforation occurred at the juncture sive transformations: firstly the FeSOaH20 decomposed virtually the welded joint between the nozzle and the inlet tube, FeOOH(hydroxyl ferric oxide), seen in Eq. (2): and then the then it can be inferred that unqualified welding may be accounted FeOOH further decomposed to Fe203, seen in Eq.(3). Conse- for this failure incident. As a result, the weakest sites after welding. quently, with generation of the gases as SO2 and SO3, only such as the tiny interspace between the nozzle and the tube, and 65% of the original weight was left at 280C in Fig. 1 or the concaves on the weld seam due to insufficient filler, were most prone to be attacked under aggressive environments. As the 2FeOOH +SO2+ SO3 (2) fuels fired in this CFB boiler were predominantly the high-sulfur 2FeOOH 03+H20 trol coke 5.5%, wt%), also the temperature of the bed mate (3) rials was above 850%C, consequently high-temperature sulfur cor rosion should be primarily blamed for the perforation on the · Step three(>280°: dehydration nozzle. Concretely speaking, initially high-temperature fuels accu- mulated at the weakest sites on the juncture, where the heat could Above 280C, only some remnant FeSO4H20 after step two not be easily released in such semi-closed zones With increase of dehydrated to ferrous sulfate, Eq.(4), and it can be learned in temperature, the surrounding materials were melted, i.e. ablation Fig 15 that nearly no weight loss occurred in this step was engendered near these localized defects. Under this condition, FeSO4·H2O FeSO4+H2O (4) corrosion resistances of the melted materials vanished, and would be subsequently corroded by the enriched sulfur element in theIn this step, ferrous sulfate hydrates that performed like the green liquid corrosion products dehydrated the crystal waters, seen in Eq. (1) [28], and the weight loss of them was about 18% at 190 C in Fig. 15. FeSO4 6H2O !70—100C FeSO4 4H2O þ 2H2O ð1Þ FeSO4 4H2O !95—190C FeSO4 H2O þ 3H2O:  Step two (190 C 280 C): decomposition After being dehydrated in step one, the ferrous sulfate hydrates then decomposed in this step, particularly for the product FeS￾O4H2O. The primary reactions mainly consisted of two succes￾sive transformations: firstly the FeSO4H2O decomposed to FeOOH (hydroxyl ferric oxide), seen in Eq. (2); and then the FeOOH further decomposed to Fe2O3, seen in Eq. (3). Conse￾quently, with generation of the gases as SO2 and SO3, only 55% of the original weight was left at 280 C in Fig. 15. 2FeSO4 H2O !>200 C 2FeOOH þ SO2 þ SO3: ð2Þ 2FeOOH !>200 C Fe2O3 þ H2O: ð3Þ  Step three (>280 C): dehydration Above 280 C, only some remnant FeSO4H2O after step two dehydrated to ferrous sulfate, Eq. (4), and it can be learned in Fig. 15 that nearly no weight loss occurred in this step. FeSO4 H2O ! 245—310 C FeSO4 þ H2O: ð4Þ 3.2.3. SEM and EDS Through optical microscope, it can be learned from Fig. 16a that two different morphologies existed on the cross-section of the per￾foration, i.e. the bright silver one and the gray one. In order to fur￾ther study their microscopic morphologies and micro-area compositions, SEM and EDS were employed. In Fig. 16b, the bright silver one was compacted while the grey one was pitted. By means of EDS, it was determined that the former one was the matrix me￾tal of casting aluminum, and the latter one was the corrosion prod￾uct containing a high content of sulfur element, seen in Fig. 16c and d and Table 5. This was in accordance with the IC, XRD and XRF re￾sults, and further identified that causes of the perforation on the manhole door was concerned with sulfur related corrosion. 3.3. Air preheater By means of XRF, element compositions of the brown rust accu￾mulating on the bottom of the air preheater was determined in Ta￾ble 6. In it, Fe and O were the two predominant elements, and consequently it can be confirmed that the rust was iron oxides. In terms of the concrete compounds, the XRD results displayed in Fig. 17 showed that ferroferric oxide (Fe3O4) and ferric oxide (Fe2O3) were the two primary ones among the many kinds, which were usually the ultimate products of uniform corrosion [29]. 4. Discussion 4.1. Fracture of nozzle As is shown in Fig. 2b that perforation occurred at the juncture, virtually the welded joint between the nozzle and the inlet tube, then it can be inferred that unqualified welding may be accounted for this failure incident. As a result, the weakest sites after welding, such as the tiny interspace between the nozzle and the tube, and/ or the concaves on the weld seam due to insufficient filler, were most prone to be attacked under aggressive environments. As the fuels fired in this CFB boiler were predominantly the high-sulfur petrol coke (4.5–5.5%, wt.%), also the temperature of the bed mate￾rials was above 850 C, consequently high-temperature sulfur cor￾rosion should be primarily blamed for the perforation on the nozzle. Concretely speaking, initially high-temperature fuels accu￾mulated at the weakest sites on the juncture, where the heat could not be easily released in such semi-closed zones. With increase of temperature, the surrounding materials were melted, i.e. ablation was engendered near these localized defects. Under this condition, corrosion resistances of the melted materials vanished, and would be subsequently corroded by the enriched sulfur element in the Fig. 14. XRD results of the solid corrosion product (a) the black one and (b) the yellow one. Fig. 15. TGA result of the black corrosion product. 678 Y. Gong, Z.-G. Yang / Materials and Design 32 (2011) 671–681