Wastage rate of water walls in a commercial circulating fluidized bed combustor.

Kim, Tae-Woo ; Choi, Jeong-Hoo ; Shun, Do Won 等

Wastage rate profiles of water walls were determined in a large (200 tonnes steam/h, 4.97 m x 9.90 m x 28.98 m tall) commercial circulating fluidized bed furnace by measuring tube thickness with an ultrasonic thickness gauge. The wastage rate was most significant in the region just above the refractory lining on all water walls, decreasing with increasing height. Wear in this region was enhanced at the centre of the sidewalls, on some sidewall tubes near the front wall, and on outer areas of the front and rear walls. The wear pattern was complex around the gas exit level. Wastage increased with increasing height on the front wall below the gas exit. Wastage decreased, then increased, with increasing height at the top of the front wall. Above the gas exit level, the wastage on the rear wall was less than on the front wall. Tubes just below the gas exit were subject to appreciable wear on the rear wall. High wastage rates also occurred on some tubes near the centre of the sidewalls at the gas exit level. On all wing walls the wastage rate decreased, then increased, with increasing height. The wastage rate of many tubes below 7 m on all water walls exceeded the maximum acceptable wear rate for steel boiler tubes in a coal-fi red utility plant, while for some tubes at the gas exit level on all walls it was close to the maximum acceptable rate.

On a determine les profils de taux de corrosion generale des parois d'eau dans un four a lit fluidise circulant commercial de grande dimension (200 tonnes de vapeur/h et 4,97 m x 9,90 m x 28,98 m) en mesurant l'epaisseur des tubes au moyen d'une jauge d'epaisseur ultrasonore. Le taux de corrosion est plus important dans la region juste au-dessus du garnissage refractaire sur toutes les parois d'eau et diminue avec l'augmentation de la hauteur. L'usure dans cette region est augmentee au centre des parois laterales, sur certains tubes des parois laterales pres de la paroi frontale et sur les parties exterieures des parois frontales et arriere. Le profil d'usure est complexe au niveau de la sortie de gaz. La corrosion augmente avec la hauteur sur la paroi frontale au-dessous de la sortie de gaz. La corrosion diminue, puis augmente, avec l'augmentation de la hauteur au sommet de la paroi frontale. Au-dessus du niveau de la sortie de gaz, la corrosion sur la paroi arriere est moindre que sur la paroi frontale. Les tubes juste au-dessous de la sortie de gaz sont sujets a une usure appreciable sur la paroi arriere. Des taux de corrosion eleves sont egalement observes pres du centre des parois au niveau de la sortie de gaz. Sur toutes les parois laterales, le taux de corrosion diminue, puis augmente, avec l'augmentation de la hauteur. Le taux de corrosion de nombreux tubes au-dessous de 7 m sur toutes les parois d'eau excede la vitesse d'usure maximale acceptable pour des tubes de rebouilleur en acier dans une installation alimentee en charbon, tandis que pour certains tubes au niveau de la sortie de gaz sur toutes les parois, celui-ci est proche de la vitesse maximale acceptable.

Keywords: tube wear, erosion, wastage, circulating fluidized bed, combustor

INTRODUCTION

Circulating fluidized bed (CFB) combustors facilitate burning a wide variety of fuels with high combustion efficiency while meeting restrictive S[O.sub.2] and N[O.sub.x] emission requirements. However, wastage of materials, both refractory and metallic, is a source of concern because it can be responsible for shutdowns and high maintenance costs.

In CFB boilers, particulates leave the combustor and enter gas-solids separators (usually cyclones). Particles larger than about 0.04 mm in diameter are removed and recirculated to the combustor bottom, while smaller particles and flue gas continue on through the convection tube bank. Wastage of tubes and other materials can occur anywhere along the gas path due to entrained particles. Stringer and Stallings (1991) and Solomon (1998) identified erosion-prone areas in circulating and bubbling fluidized bed boilers and described current methods for erosion protection. The water walls of CFB furnaces are of special concern. Metal wastage has commonly been found in lower furnace walls immediately above the refractory interfaces, on water walls near cyclone inlets, and on furnace roof tubes.

A number of studies have been conducted in an effort to understand metal wastage in fluidized bed combustion environments and to find ways to reduce it. Metal wastage is caused by simultaneous erosion and corrosion (Wang et al., 1990; Johnson, 1991; Dutheillet and Prunier, 1998; Solomon, 1998; Nafari and Nylund, 2002). The rates and mechanisms of metal loss are complex functions of the characteristics of the particles, i.e., composition, shape, size, and strength; of the conditions inside the combustors, such as temperature, gas or particle velocity and impact angle, gas composition, solids loading, and location; and of the characteristics of the metal, i.e., composition and morphology (Wang et al., 1990). Higher wastage is associated with higher ash content and with a greater proportion of hard erosive species, such as Si[O.sub.2], [Al.sub.2][O.sub.3] and [Fe.sub.2][O.sub.3], in the coal ash (Lindsley et al., 1995; Wang and Luer, 1998). The greater the proportion of weaker and softer limestone and the larger the CaO content, the less erosive is the bed material (Wang et al., 1990). CaO mechanically mixed with Fe oxide forms an intimate mixture that protects the metal surface (Geng et al., 1991). Ca and S constituents in bed material can protect surfaces from erosion by forming a thick, continuous, protective layer on the target (Lindsley et al., 1995).

Higher wastage is associated with lower moisture content in the bed material (Wang and Luer, 1998). Metal loss increases with increasing particle size (Wang et al., 1990, 1992; Rogers and Boyle, 1993). Angular particles erode metal surfaces more than round ones (Ninham et al., 1989; Zhu et al., 1990; Wang et al., 1990, 1992; Lindsley et al., 1995). Metal wastage increases with increasing gas and particle velocities (Wang et al., 1990; Geng et al., 1991; Seitzinger, 1995). The particles must be strong enough not to shatter upon impact in order to be erosive (Wang et al., 1990, 1992). Impacts at shallow angles tend to cause more damage than those at steep angles (Wang et al., 1990). Erosion rates are higher on tube crests than on the fins for membrane water walls (Lockhart et al., 1995). The influence of temperature on metal wastage is complex because of a number of combined physical and chemical effects on the properties of gas, particles and target materials (Wang et al., 1990; Zhu et al., 1990; Holtzer and Rademakers, 1991; Stringer and Stallings, 1991; Wang, 1995, 1996; Rogers et al., 1997; Wang and Lee, 1997).

The erosion-corrosion resistance of coatings is primarily dependent on their composition and microstructure, rather than on their hardness. Coatings of favourable composition and fine structure exhibit the greatest erosion-corrosion resistance (Wang and Lee, 1997). Other chemical factors also influence wastage. For example, chlorides in deposits can react with the scale to form chlorine that enters the scale and accelerates oxidation (Grabke et al., 1995). S[O.sub.2] can cause a minor increase of active corrosion by sulphation of chlorides and generation of chlorine (Grabke et al., 1995). HCl can transform sulphates into chlorides, thereby enhancing active oxidation (Grabke et al., 1995; Hou et al., 1999). In the presence of balanced concentrations of HCl and S[O.sub.2], however, corrosion was limited (Grabke et al., 1995). Wastage has been found to increase with increasing proportions of NaCl and KCl in the bed material (Wang, 1996). Biomass-fired boiler fly ash has been reported to have relatively high erosivity due to high concentrations of chemically active compounds containing alkali, sulphur, phosphorus and chlorine (Wang, 1995). Austenitic steels suffer predominantly from selective corrosion that resulted in depletion of chromium. Selective corrosion increases with increasing chromium content of the alloy (Montgomery and Karlsson, 1999). Ninham et al. (1989) and Yamamoto et al. (2001) found that the acid resistance and erosion resistance of metals depends on their composition and on temperature. Vincent et al. (1989) compared the relative erosion resistance of several metals.

Erosion-prone areas have been addressed by design changes (Stringer and Stallings, 1991; Solomon, 1998). Where these are not feasible, protection methods such as refractory, weld overlay, erosion blocks, stainless steel shelves, steel ball studs and fins, and stainless steel shields have been employed (Fan et al., 1990; Stringer and Stallings, 1991; Solomon, 1998). Despite such protection, problems have often been experienced with reliability and/or longevity. As a result, more attention has been focused on thermal spray coatings, as summarized by Solomon (1998).

Most studies of tube wastage have been conducted under model conditions. Several results have been reported from large-scale fluidized bed combustors with long test times (Vincent et al., 1989; Holtzer and Rademakers, 1991; Stringer and Stallings, 1991; Jestin et al., 1992; Seitzinger, 1995; Nafari and Nylund, 2002). However, there have been few reports of overall tube wastage profiles on water walls, including wing walls, for large commercial CFB furnaces. This type of information is needed to gain more detailed understanding of wastage patterns, to assist in early detection of problems, and to develop procedures for proper furnace maintenance.

The purpose of this study was to determine and explain the overall tube wastage pattern on water walls, including wing walls, in a large commercial CFB furnace. The wastage rate was found by measuring local thicknesses in 1998 and 2002, taking differences and dividing by the time of operation. The tube wastage profiles not only help to provide qualitative understanding, but give detailed wastage patterns. In the current paper, we are able to report precise wastage patterns on whole water walls, especially near the exits.

EXPERIMENTAL METHODOLOGY

The tube thickness profile was measured in a combustor of a commercial CFB boiler generating 200 tonnes steam/h, shown schematically in Figure 1. The combustor is 9.90 m x 4.97 m in cross-section and 28.98 m tall. It is equipped with two hot cyclones and two loop seals. The combustor walls consist of vertical heat exchanger tubes separated by fins. Therefore, heat is absorbed directly over most of the height. However, the tubes are lined with refractory in the lower section of the combustor (up to 3.5 m above the distributor). The tubes are straight, not offset, at the refractory and tube interface (i.e., there is no "kickoff tube").

[FIGURE 1 OMITTED]

Figure 2 summarizes the configuration of the water walls of the combustor and the tube-numbering scheme. The width of the cross-section increases gradually from 2.5 m at the distributor plate to 4.97 m in the lower part of the combustor, with the diverging section ending 3.23 m above the distributor. The combustor walls, including four wing walls, contain 372 tubes of diameter 63.5 mm, connected side-by-side by bridging fins. Each wing wall includes nine tubes and is attached to the front wall. Wing walls are not equally spaced, but are located a little closer to both sides of the front wall. The wing walls are each 28.98 m high and lined with refractory in the lower section (again up to 3.5 m above the distributor). Table 1 summarizes the specifications of the heat exchanger tubes.

[FIGURE 2 OMITTED]

Two gas exits leading to a pair of hot cyclones (barrel diameter: 4.95 m, barrel length: 6.43 m, length of cone section: 6.23 m, dipleg diameter: 0.9 m, dipleg length: 10.5 m) are located at opposite ends of the rear wall at the top of the combustor. Four recycle inlets for return of particles collected in cyclones penetrate the rear wall in the lower part of the combustor. Fresh coal is fed there, together with recycled solids. Tubes are numbered clockwise, and the lateral position of each tube can be estimated from the relations in Table 2.

Tube thicknesses were determined with an ultrasonic thickness gauge (Krautkramer, DM4E). In order to obtain axial profiles of thickness, measurements were carried out at various heights along each tube. The tube thicknesses were measured at the crest of the tube. This combustor started commercial operation in November 1988, whereas the measurements reported here were carried out twice, in 1998 and 2002. The duration of operation between the two sets of measurements was 1228 d. The tube wastage rate at each point was determined by subtracting the tube thickness in 2002 from that in 1998, and then dividing the difference by the duration of operation.

The tube wastage was so considerable between 3.5 m and 7 m that overlay coating had been applied annually. Therefore, tube thicknesses for heights above 7 m show the cumulative loss of tube thickness, whereas those between 3.5 m and 7 m include maintenance effects. Hence, the wastage is considered here only for heights above 7 m.

The average operating conditions were as follows: superficial gas velocity based on freeboard area and average furnace temperature: 4.40 m/s; furnace temperature: 843[degrees]C; pressure drop between distributor level and gas exit (23 m above the distributor): 5224 Pa; average volumetric solids holdup between two levels (see Table 3): 0.0101; particle density: 2300 kg/[m.sup.3]; coal feed rate: 20.1 t/h; limestone feed rate: 0.145 t/h; sand feed rate: 14 kg/h; steam production rate: 162 t/h. Bituminous coal was the fuel, with limestone as sorbent for S[O.sub.2] capture, and sand added to maintain a steady solids inventory in the furnace. The external solid circulation flux could not be measured because of the large size of the commercial unit. Tables 4 to 6 show the coal properties; composition of limestone, sand, bed particles, and fly ash; and mean diameter and size distribution of the particles, respectively. Tables 7 and 8 show typical temperature and air supply profiles in the combustor. Note that the axial pressure drop on the west wall was somewhat greater than on the east wall. All measurements for Tables 3 and 7 were made at the centres of the respective walls.

The fluidizing velocity, temperature and feed rates of coal and limestone increase with the steam production rate. However, the steam production rate was not varied significantly. The present combustor was operated at a relatively dilute solids loading (5224 Pa pressure drop across the combustor). The main cause may be due to the poor cyclone effi ciency. This is consistent with the requirement for continuous sand injection during operation. According to the correlation of Lee and Kim (1988), the transition velocity to turbulent velocity is 3.27 m/s and the bed operates in the turbulent fluidization flow regime.

RESULTS AND DISCUSSION

Figure 3 shows tube wastage rate profiles on the front wall of the combustor and on three typical single tubes. The four wing walls (tube numbers 113-148) are excluded in this figure. The highest wastage rate (1.14 x [10.sup.-3] mm/d) occurred at the bottom of the region included in this study (7 m above the distributor). The wastage rate decreased, followed by an increase up to 0.81 x [10.sup.-3] mm/d at the gas exit level, with increasing height. The minimum wastage rate (0.32 x [10.sup.-3] mm/d) occurred at a height of about 17 m. The wastage rate decreased, followed by an increase, with increasing height above the gas exit. At lower heights the tube wastage rate was significantly greater at both sides than towards the centre of the wall.

[FIGURE 3 OMITTED]

A transition region generally exists between a lower dense bed region and an upper lean freeboard region in gas-fluidized beds. In this transition region, particles are ejected upwards, causing vigorous collisions with the wall. In addition, there is a huge down flow of particles along the wall in this height interval (Stringer and Stallings, 1991; Kunii and Levenspiel, 1991). The mean particle size and particle concentration decrease with increasing height in this region. Therefore, particle impacts with the wall decrease accordingly. The wastage rate decreases with increasing height in this region.

The freeboard cross-section of this combustor can be considered as two adjacent connected nearly square cross-sections, each about 4.97 m x 4.97 m in size. Van der Meer et al. (1996, 2000) found larger solids down-flow near the centre of the wall and at the corners in a riser of square cross-section having an exit of the same cross-section. Their findings also show wastage in the transition region, enhanced towards the outside of the front wall, since wing walls in the present combustor are not equally spaced, but located a little closer to both sides of the front wall. Because the two exits are located towards the outside of the rear wall in the current study, the solids up-flow seems to have been skewed towards both sidewalls. Therefore, wastage near the centre was smaller in the transition region of the front wall than at the outside.

In Figure 3, the wastage rate at the refractory termination (3.5 m above the distributor) was about 2 x [10.sup.-3] mm/d, predicted by extrapolating the present trend. This is less than reported by Stringer and Stallings (1991): 5.28 x [10.sup.-3] mm/d on average just above the refractory termination at Nucla, Colorado; 8.4 x [10.sup.-3] mm/d after adding shelves for protection at Stockton, California; 0.115 mm/d 25 mm above the refractory shelf at Chatham, New Brunswick; and 0.125 mm/d 100 mm above the refractory termination at Stockton, California.

High wastage rates were also found on some tubes in the top section of the front wall. This combustor has two abrupt gas exits at the top, requiring the gas stream to turn sharply. Relatively large entrained solid particles are unable to follow the gas streamlines, and instead move to the outer part of the stream, causing increased particle concentrations and collisions with the wall near the exits. Therefore, particle momentum flux to the wall increases locally, augmenting the wear on the top section of the wall (Stringer and Stallings, 1991; Van der Meer et al., 1996; Harris et al., 2003). Van der Meer et al. (1996) and Harris et al. (2003) found a large eddy opposite the exit. A large eddy of this nature likely caused the present trend that the wastage rate decreased again, followed by an increase, with increasing height above the gas exit. The solids holdup seemed to increase with height between 17 m and the gas exit due to the eddy effect, resulting in the wastage rate increasing with height in this region.

Figure 4 shows the profile of tube wastage rate on the east sidewall. The right side of the top section near the gas exit was covered with refractory. The wastage rate on the wall decreased from 0.98 x [10.sup.-3] mm/d with increasing height above the distributor, except for the centre at the gas exit level. In the lateral direction at low heights, wastage seemed to be enhanced close to the front wall and near the centre of the wall. In particular, high wastage rates occurred on some tubes near the centre of the wall at the gas exit level. The cause of tube wastage in the transition region seemed to be similar to that on the front wall (Stringer and Stallings, 1991; Van der Meer et al., 1996, 2000). Van der Meer et al. (1996, 2000) found that solids down-flow in the freeboard was larger at the corners than elsewhere. However, in our case it seemed that the down-flow rate of solids was larger near the centre of the wall due to secondary flow of the second kind, which was stronger toward the centre than corners (Prandtl, 1952; Van der Meer et al., 1996, 2000).

[FIGURE 4 OMITTED]

Figure 5 shows the tube wastage rate along the rear wall. Both sides of the top section contain gas exits. The axial profile of wear rate (maximum rate 1.46 x [10.sup.-3] mm/d) is similar to that for the front wall in Figure 3, except at the top. The wastage rate decreased with increasing height, except just below the gas exits (tube number: 206 to 224, 297 to 315). At lower heights the wastage rate was again higher at both sides than towards the centre of the wall. Particularly high wastage rates occurred just below the gas exits. Except in this region, the wastage rate at the top of this wall was less than on the front wall. Tube wastage seems to have been caused by colliding particles in the transition region, as on the front wall in Figure 3, even though no wing walls exist. The effect of secondary flow of the second kind (Prandtl, 1952; Van der Meer et al., 1996, 2000) seems to be stronger again toward the centre than corners, as shown in Figure 4. According to Van der Meer et al. (1996), recirculation of solids to the front wall is greater than to the rear wall at the top of the combustor. That seems to account for the lower wastage rate above the gas exit on the rear wall, compared with the front wall. Coarse particles descending towards the edge of the gas exit or from dunes in horizontal ducts, as shown by Van der Meer et al. (1996) and Harris et al. (2003), may explain the high wastage rate just below the gas exits.

[FIGURE 5 OMITTED]

Profiles of tube wastage rate on the west sidewall appear in Figure 6. The left side of the top section near the exit is covered with refractory. Except in the top section, the axial and lateral profiles of tube wastage rate on the west sidewall (maximum wear rate: 1.3 x [10.sup.-3] mm/d) were similar to those on the east sidewall (Figure 4). In the lateral direction, wastage seems to have been enhanced on tubes near the front wall and near the centre of the wall at lower heights. High wastage rates occurred on some tubes near the centre of the wall at the gas exit level. The main causes of wastage on this wall are therefore likely to be the same as for the east sidewall. Wastage was faster at the top section than on the east wall, probably resulting from asymmetry in gas and solids flow between the two gas exits, as can be seen in Table 3 where solids holdup on the west side was greater than on the east side.

[FIGURE 6 OMITTED]

The wear rate for the rear wall just above the refractory was as high as for the other outside walls in Figure 5, even though the solids recirculation re-entry ports were located on the rear wall. Water wall wear just above the refractory was found around the entire circumference of the combustor. This finding is not in agreement with results for the Nucla CFB furnace (Stringer and Stallings, 1991), implying that the gas and solids flow patterns in the present combustor differ from those in the Nucla unit.

Figure 7 shows the profile of tube wastage rate for the wing walls extending into the interior of the furnace from the front wall. The tube wastage rate on all wing walls decreased, followed by an increase, with increasing height, as for the front wall. The wastage rate for tube numbers 124 and 127 on the second wing wall at the refractory termination (3.5 m above the distributor) was about 4 x [10.sup.-3] mm/d, similar to that reported for the Nucla unit (Stringer and Stallings, 1991).

[FIGURE 7 OMITTED]

In measurement of tube thicknesses, very thin regions, that could be misunderstood as serious erosion, were found at tube joints about 15 m above the distributor. Wear on tube walls is strongly influenced by the flow pattern of the gas-solid mixture and, therefore, by the geometry of the combustor. In order to understand better the tube wear patterns in the transition region and around the gas exit level where wastage was pronounced, the micro- and macroscopic flow patterns of gas and solids in the combustor require elucidation.

CONCLUSION

Tube wastage of water walls, including wing walls, has been quantified in a large commercial CFB combustor for heights exceeding 7 m. The wastage rate was largest in the transition region just above the refractory lining on all water walls, including wing walls, as well as outside walls. The wastage rate decreased with increasing height in this region. Water wall wear just above the refractory lining occurred around the entire circumference of the combustor, especially at the centre of the sidewalls, on some sidewall tubes near the front wall, and in the outer areas of the front and rear walls. The wear pattern was complex near the gas exits. Wear increased with increasing height below the gas exit on the front wall. The wastage rate on the top section of the front wall decreased, then increased, with increasing height. The wear rate above the gas exit level on the rear wall was less than on the front wall. The wastage rate of tubes just below the gas exit was appreciable on the rear wall. High wastage rates occurred on some tubes near the centre of the sidewalls at the gas exit level. The axial profiles on wing walls were similar to those on the front wall to which the wing walls were attached.

The wastage rate of many tubes below 7 m on all water walls appear to have exceeded the maximum acceptable wear rate (1.2x[10.sup.-3] mm/d) for steel boiler tubes in a coal-fi red utility plant (Lyczkowski and Bouillard, 2002). Some tubes of all walls at the gas exit level also experienced wear close to the maximum acceptable rate.

ACKNOWLEDGMENT

This research was sponsored by the Korea Energy Management Corporation in the Korea Ministry of Commerce, Industry and Energy under a Clean Energy Technology Development Project.

REFERENCES

Dutheillet, Y. and V. Prunier, "Evaluation of the Erosion-Corrosion Resistance of Coated Metallic Materials for CFBCs Applications," Schriften des Forschungszentrums Juelich, Reihe Energietechnik/Energy Technology 5, 779-787 (1998).

Fan, J., D. Zhou and K. Cen, "Study of the Fin-Tube Erosion-Protection Method in the Circulating Fluidized Beds," Huaxue Fanying Gongcheng Yu Gongyi 6, 37-44 (1990).

Geng, G. Q., B. Q. Wang and A. V. Levy, "The Effect of the Composition of Fluidized Bed Combustor Bed Materials on the Erosion-Corrosion of Carbon Steel," Wear 150, 125-134 (1991).

Grabke, H. J., E. Reese and M. Spiegel, "The Effects of Chlorides, Hydrogen Chloride, and Sulfur Dioxide in the Oxidation of Steels below Deposits," Corrosion Science 37, 1023-1043 (1995).

Harris, A. T., J. F. Davidson and R. B. Thorpe, "Influence of Exit Geometry in Circulating Fluidized-Bed Risers," AIChE J. 49, 52-64 (2003).

Holtzer, G. J. and P. L. F. Rademakers, "Studies on 90 MWth AKZO and 4 MWth TNO FBC Show Excellent Erosion-Corrosion Results," in "Proc. 11th Int. Conf. on Fluidized Bed Combustion," (1991), pp. 743-753.

Hou, P. Y., J. T. Sum, Y. Niu and J. Stringer, "HCl Effect on In-Bed Tube Wastage in Bubbling Fluidized Bed, A Laboratory Study Under Simulated Dense Particle Impact Conditions," in "Proc. of 15th Int. Conf. on Fluidized Bed Combustion," (1999), pp. 1335-1353.

Jestin, L., P. Meyer, G. Schmitt and J. X. Morin, "Heat Transfer in a 125 MWe CFB Boiler," in "Fluidization VII," O. E. Potter and D. J. Nicklin, Eds., Engineering Foundation, New York (1992), pp. 849-856.

Johnson, R. C., "Preformed Refractory Modules for Circulating Fluidized Bed Boilers," in "Proc. 11th Int. Conf. on Fluidized Bed Combustion," (1991), pp. 1445-1449.

Kunii, D. and O. Levenspiel, "Fluidization Engineering," Butterworth-Heinemann, Boston, MA (1991).

Lee, G. S. and S. D. Kim, "Pressure Fluctuations in Turbulent Fluidized Beds," J. Chem. Eng. Japan 21, 515-521 (1988).

Lindsley, B. A., A. R. Marder and J. J. Lewnard, "The Effect of Circulating Fluidized Bed Particle Characteristics on Erosion of 1020 Carbon Steel," Wear 188, 33-39 (1995).

Lockhart, C., J. Zhu, C. M. H. Brereton, C. J. Lim and J. R. Grace, "Local Heat Transfer, Solids Concentration and Erosion around Membrane Tubes in a Cold Model Circulating Fluidized Bed," International Journal of Heat and Mass Transfer 38, 2403-2410 (1995).

Lyczkowski, R. W. and J. X. Bouillard, "Scaling and Guidelines for Erosion in Fluidized Beds," Powder Technology 125, 217-225 (2002).

Montgomery, M. and A. Karlsson, "In-situ Corrosion Investigation at Masnedo CHP Plant. A Straw-Fired Power Plant in Denmark," Materials and Corrosion 50, 579-584 (1999).

Nafari, A. and A. Nylund, "Erosion Corrosion of Steel Tubes in the Loop Seal of a Biofuel Fired CFB Plant," Schriften des Forschungszentrums Juelich, Reihe Energietechnik/ Energy Technology 21, 969-978 (2002).

Ninham, A. J., M. J. Entwisle, I. M. Hutchings and J. A. Little, "A Laboratory-Scale Fluidized Bed Rig for High Temperature Tube Wastage Studies," in "Proc. 10th Int. Conf. on Fluidized Bed Combustion," (1989), pp. 583-589.

Prandtl, L., "Essentials of Fluid Dynamics," Blackie & Son, London (1952), pp. 145-149.

Rogers, W. A. and E. J. Boyle, "Wear Prediction in a Fluidized Bed Combustor," in "Proc. 12th Int. Conf. on Fluidized Bed Combustion," (1993), pp. 811-817.

Rogers, P. M., I. M. Hutchings, J. A. Little and F. Kara, "Microstructural Characterization of Surface Layers Formed on a Low Alloy Steel During Erosion-Oxidation in a Fluidized Bed," Journal of Materials Science 32, 4575-4583 (1997).

Seitzinger, D. L., "Atmospheric Fluidized Bed Combustion Gas Erosion Solution," in "Proc. 13th Int. Conf. on Fluidized Bed Combustion," (1995), pp. 585-595.

Solomon, N. G., "Erosion-Resistant Coatings for Fluidized Bed Boilers," Materials Performance 37, 38-43 (1998).

Stringer, J. and J. Stallings, "Materials Issues in Circulating Fluidized-Bed Combustors," in "Proc. 11th Int. Conf. on Fluidized Bed Combustion," (1991), pp. 589-608.

Van der Meer, E. H., R. B. Thorpe and J. F. Davidson, "The Influence of Exit Geometry for a Circulating Fluidized Bed with a Square Cross-Sectional Riser," in "Proc. 5th Int. Conf. on Circulating Fluidized Beds," (1996), pp. Eq2.1-2.6.

Van der Meer, E. H., R. B. Thorpe and J. F. Davidson, "Flow Patterns in the Square Cross-Section Riser of a Circulating Fluidized Bed and the Effect of Riser Exit Design," Chem. Eng. Sci. 55, 4079-4099 (2000).

Vincent, R. Q., D. A. Canonico and J. M. Wheeldon, "An Evaluation Program for Metal Wastage in Fluidized Bed Combustors," in "Proc. 10th Int. Conf. on Fluidized Bed Combustion," (1989), pp. 927-935.

Wang, B.-Q., "Erosion-Corrosion of Coatings by Biomass-Fired Boiler Fly Ash," Wear 188, 40-48 (1995).

Wang, B. Q., "Effect of Alkali Chlorides on Erosion-Corrosion of Cooled Mild Steel and Cr3C2-NiCr Coating," Wear 199, 268-274 (1996).

Wang, B. Q., G. Q. Geng and A. V. Levy, "Surface Behavior of Heat Exchanger Tubes in Fluidized-Bed Combustors," Surface and Coatings Technology 42, 253-274 (1990).

Wang, B. Q., G. Q. Geng, A. V. Levy and W. Mack, "Erosivity of Particles in Circulating Fluidized Bed Combustors," Wear 152, 201-222 (1992).

Wang, B.-Q. and S. W. Lee, "Erosion Resistance of Cooled Thermal Sprayed Coatings under Simulated Erosion Conditions at Waterwall in FBCS," in "Proc. 14th Int. Conf. on Fluidized Bed Combustion," (1997), pp. 335-342.

Wang, B.-Q. and K. Luer, "The Relative Erosivity of Limestone, Dolomite and Coal Samples from an Operating Boiler," Wear 215, 180-190 (1998).

Yamamoto, K., I. Kajigaya, K. Sonoya and Y. Tsuji, "Material Selection for the Super-Heater and Re-Heater Tubes on PFBC, Based on the Results of Laboratory Test and Ex-Serviced Materials Survey," in "Proc. 16th Int. Conf. on Fluidized Bed Combustion," (2001), pp. 1451-1465.

Zhu, J., J. R. Grace and C. J. Lim, "Tube Wear in a Gas Fluidized Beds - I. Experimental Findings," Chem. Eng. Sci. 45, 1003-1015 (1990).

Tae-Woo Kim (1), Jeong-Hoo Choi (1) *, Do Won Shun (2), Bongjin Jung (3), Soo-Sup Kim (4), Jae-Ek Son (5), Sang Done Kim (6) and John R. Grace (7)

* Author to whom correspondence may be addressed. E-mail address: choijhoo@konkuk.ac.kr

(1.) Department of Chemical Engineering, Konkuk University, Seoul 143-701, Korea

(2.) Korea Institute of Energy Research, Daejeon 305-343, Korea

(3.) Department of Environmental Engineering, Suwon University, Suwon 445-743, Korea

(4.) SK Chemicals, Ulsan, 680-160, Korea

(5.) The Graduate School of Energy and Environment, Seoul National University of Technology, Seoul 139-743, Korea

(6.) Department of Chemical and Biomolecular Engineering, KAIST, Daejeon 305-701, Korea

(7.) Department of Chemical and Biological Engineering, The University of British Columbia, Vancouver, BC, Canada V6T 1Z3

Manuscript received March 28, 2006; revised manuscript received June 23, 2006; accepted for publication August 28, 2006

Table 1. Specifications of water wall tubes

Type Membrane wall

Original outside diameter 63.5 mm

Original wall thickness 6.1 mm

Material A210A1

Pitch 88 mm

Total number 372

Table 2. Equations for horizontal distances of tube centres
in Figure 2(b)

Wall Direction Range of Equation for horizontal
 tube no. distance in mm.

North A [right arrow] B 1-112 (tube no.-1) x 88+44

Wing from north wall 113-121 (tube no.-112) x 88

 122-130 (tube no.-121) x 88

 131-139 (tube no.-130) x 88

 140-148 (tube no.-139) x 88

East B [right arrow]C 149-204 (tube no.-149) x 88+44

South C [right arrow] D 205-316 (tube no.-205) x 88+44

West D [right arrow] A 317-372 (tube no.-317) x 88+44

Table 3. Typical pressure profile in the combustor

Height above distributor (m) East wall North wall West wall

 Gauge pressure (Pa):

0 5035 5211 5260

23 -52 -59

Table 4. Coal properties.

Proximate analysis (as received, wt.%)

Moisture 11.7
Volatiles 28.1
Ash 6.9
Fixed carbon 53.3

Ultimate analysis (dry basis, wt.%)

C 70.7
H 4.7
N 0.8
S 0.6
O (by difference) 15.4

Higher heating value (kJ/kg) 29110

Table 5. Composition of solid materials

 Lime- Sand Bottom Fly
 stone ash ash

Component Composition (dry basis, wt.%)

Si[O.sub.2] 1.1 80.3 51.7 29.1
[Al.sub.2][O.sub.3] 0.3 10.8 12.3 10.3
Ti[O.sub.2] 0.2 0.5 0.5
[Fe.sub.2][O.sub.3] 0.2 1.1 6.1 7.9
CaO 55.6 0.9 23.2 38.7
MgO 0.3 0.3 0.8 1.3
[Na.sub.2]O 2.1 3.7 1.4
[K.sub.2]O 4.2 0.9 0.4
MnO 0.2 0.3
[P.sub.2][O.sub.5] 0.1

Table 6. Size distribution of particles

Solids Lime- Sand Bottom Fly
 stone ash ash

Mean 0.39 0.362 0.361 0.174
diameter
(mm) *

Particle size distribution:

Screen size Weight fraction
range (mm)

5.7-4 0.030

4-2 0.050

2-1.18 0.001 0.072

1.18-0.71 0.399 0.007 0.103

0.71-0.5 0.272 0.204 0.105 0.050

0.5-0.212 0.275 0.747 0.463 0.336

0.212-0.125 0.017 0.032 0.160 0.359

0.125-0.075 0.008 0.009 0.017 0.252

0.075-0 0.028 0.001 0.002

* specific surface mean diameter

Table 7. Typical temperature profile in the combustor

Height above East Front and West
distributor (m) sidewall rear walls sidewall

 Temperature ([degrees]C):

0 842 851

0.3 882 *

0.75 837, 827

0.9 806

2.2 850

2.8 883

28(exits) 822 823

* Rear wall.

Table 8. Typical profile of air supply to the combustor

Height above distributor (m) Air supply fraction

0 0.50

0.5-0.6 0.24

1.5 0.07

2.3 0.12

3.0 0.07