416
Energy & Fuels 1997, 11, 416-420
Predicting Removal of Coal Ash Deposits in Convective Heat Exchangers C. L. Senior* Physical Sciences Inc., 20 New England Business Center, Andover, Massachusetts 01810 Received May 21, 1996X
Electric utilities are under pressure to reduce emissions and increase efficiency, and this means that coal-fired power plants must meet new challenges such as switching to coals for which the units were not designed. Ash deposition in coal-fired power plants reduces heat transfer and can cause the plant to be shut down. Ash deposits on convective heat transfer surfaces are generally removed mechanically by soot-blowers. On-line cleaning is practical as long as the ash deposits are not highly sintered. The sintering behavior of deposits is complex and depends on flue gas temperature, ash particle size, and ash composition. In this paper, a method is presented for estimating, for a specific coal, the maximum flue gas inlet temperature that allows the convective heat exchanger to be cleaned using conventional means. The calculation was carried out for conditions that represent the steam superheater section of a conventional pulverized coal-fired power plant, and the results are consistent with observations from existing plants. The method was then applied to the design of a novel air heater being designed for the Combustion 2000 HIPPS program.
Introduction Combustion of coal in furnaces and boilers releases ash, the deposition of which can cause problems such as reduced heat transfer, blockage or plugging of gas flow, and corrosion. These problems can be costly due to reductions in generating capacity, unplanned outages, lower plant efficiency, and equipment modifications. In the high-temperature furnace region, ash deposition on water walls, commonly referred to as slagging, reduces heat transfer by conduction (increased thermal resistance due to the deposit) and radiation (decreased emissivity of the deposit surface). In the convective section ash deposition or fouling can impede the gas flow and decrease heat transfer. Both slagging and fouling in coal-fired power plants can be controlled on-line using soot-blowers, devices that direct a jet of high-pressure air, steam, or water at the deposit surface. An important part of designing a furnace is determination of the location, number, and area of coverage of soot-blowers as well as consideration of the dimensions of the furnace and heat transfer surfaces. Coal properties are the most important factor in these choices.1 Operators of coal-fired boilers today are under pressure to decrease power production costs while at the same time complying with increasingly stringent emissions requirements. To satisfy both objectives, it is often desirable to burn “nondesign” coals, that is coals having properties that out of the range of the boiler’s design. Safe and efficient boiler operation requires methods to estimate the impact of nondesign coals on ash deposition, heat transfer, and deposit removability. * Author to whom correspondence should be addressed (e-mail
[email protected]). X Abstract published in Advance ACS Abstracts, January 1, 1997. (1) Borio, R. W.; Levasseur, A. A.; Chow, O. K.; Miemiec, L. S. Ash Deposition: A Boiler Manufacturer’s Perspective. Inorganic Transformations and Ash Deposition During Combustion; ASME: New York, 1991.
S0887-0624(96)00078-3 CCC: $14.00
The design of new types of coal-fired power-generating plants, such as the Combustion 2000 program sponsored by the U.S. Department of Energy, also requires methods for estimating the effect of ash deposition. The High Performance Power System (HIPPS) portion of this program, for example, uses a combined cycle system to increase the overall efficiency of a coal-fired power plant to 47%.2 The HIPPS concept features a novel hightemperature convective air heater. Conditions in this convective air heater are very different from those in conventional power plants’ steam cycles. The motivation for the work described in this paper is the design of the HIPPS plant. However, both conventional and innovative coal-fired power plants can benefit from an increased understanding of the impact of ash characteristics on operation. The question to be addressed in this paper is the following: what is the maximum flue gas temperature in the convective heat transfer section of a coal-fired boiler such that ash deposits on heat transfer surfaces can be removed easily? Answering this question involves a three-step process. First, it is necessary to establish a criterion for deposit removability by soot-blowers. A minimum deposit porosity will be identified, based on measurements of deposit strength. The second step will be to calculate sintering rates in the deposit to make sure that the deposit does not reach the “critical” porosity too quickly for easy removal. The result will be a maximum allowable deposit surface temperature. Finally, using the maximum allowable deposit surface temperature, a heat transfer calculation will be performed to arrive at the maximum allowable flue gas temperature as a function of deposit thickness. The results depend on the composition of the coal ash. In this paper, a (2) Sangiovanni, J. J.; Seery, D. J.; Chiappetta, L.; Senior, C. L. CoalFired High Performance Power Generating System; ASME Paper 94JPGC-PWR-3, 1994.
© 1997 American Chemical Society
Removal of Coal Ash Deposits
Figure 1. Variation in compressive strength for 10 slags at porosity values from 0 to 50 vol % (source: ref 3).
bituminous and a sub-bituminous coal will be used as examples. Deposit Removability Criterion Soot-blowing of deposits in coal combustion systems produces both thermal and mechanical shock resulting from the jet impingement, and these are responsible for fracturing the deposit. The strength of the deposit depends on the porosity of the deposit3 and on the crystallinity of the deposit.4 These properties are, in turn, related to the composition of the ash particles in the deposit. What are the deposit characteristics that relate to successful deposit removal? This question was addressed by Wain et al.3 They measured thermal and mechanical properties of 10 slags and tried to relate the measurements to soot-blowing experience. The compressive strength was measured as a function of porosity for 10 slags as shown in Figure 1. As porosity decreased, the compressive strength increased. The change in compressive strength was most dramatic as the porosity dropped below 25%. Above 25%, all of the slags had similar compressive strengths. The elastic modulus showed a similar trend with porosity. It is well-known that reduction in porosity, i.e., sintering of the deposit, makes the deposit harder to remove by soot-blowing. Since the deposit strength is fairly constant for porosity >25%, we can conclude that this is a critical porosity for deposit removal by sootblowing. Ash Sintering Rate The most important criterion for determination of the maximum allowable gas temperature entering the convective heat exchanger is that the deposit must be removable by soot-blowing. Since an ash deposit has a finite thickness, we can specify that the “outer” surface of the deposit must not be so sintered as to prevent easy removal by soot-blowing. The criterion for easy removal will be that a porosity of 25% or more will be obtained in >8 h, which was selected as a minimum soot-blowing frequency. (3) Wain, S. E.; Livingston, W. R.; Sanyal, A.; Williamson, J. Inorganic Transformations and Ash Deposition During Combustion; ASME: New York, 1991; pp 459-470. (4) Nowok, J. W.; Benson, S. A.; Jones; Kalmanovitch, D. P. Sintering behaviour and strength development in various coal ashes. Fuel 1990, 69, 1020-1027.
Energy & Fuels, Vol. 11, No. 2, 1997 417
Models for the sintering of coal ash deposits are based on viscous flow of the ash. The earliest work by Frenkel5 concerned the rate of coalescence of two spheres such that the energy dissipated in viscous flow equals the energy gained by a decrease in surface area. Frenkel’s model is applicable to the early stages of deposit sintering and can provide a qualitative indication of the degree of sintering.6 Mackenzie and Shuttleworth7 developed a model for the shrinkage of closed pores, which occurs in the later stages of densification. Scherer8 developed a more general approach for sintering of open pores using an array of intersecting cylinders as a model. He showed that his model reproduced the behavior of the Mackenzie and Shuttleworth model for values of porosity j70%. For the purposes of modeling sintering of coal ash deposits from approximately 50% to 0% porosity, the model of Mackenzie and Shuttleworth will be used. This model describes the change in porosity as a function of time as follows:
/0 ) e-3tσ/2rp µ
(1)
where 0 and are the deposit porosities initially and at some time t, respectively. The initial pore radius is rp. The ash surface tension and viscosity are denoted by σ and µ, respectively. In this analysis, viscosity will be calculated on the basis of the bulk composition of the ash. As discussed below, using bulk viscosity provides only an approximate answer since the composition of individual ash particles in the deposit can differ greatly from the bulk composition. A constant surface tension will be assumed. For the sintering calculation, the ratio of surface tension to viscosity is the quantity of interest. Both surface tension and viscosity are functions of temperature. Surface tension depends approximately on the square root of temperature, while viscosity shows an exponential dependence on temperature.9 Given that the viscosity changes by several orders of magnitude over the temperature range of interest, it is reasonable to neglect the variation of surface tension with temperature. We have picked two coals, Illinois No. 6 and WyodakRochelle, for detailed analysis. The first is an Illinois basin coal with high iron content, giving high ash fusion temperature under oxidizing conditions but low ash fusion temperature under reducing conditions. The second is a Powder River Basin (PRB) sub-bituminous coal with high calcium and low ash fusion temperature. Ash compositions are reported in Table 1. According to Raask,9 viscosities in the range of 1071011 P are relevant to the formation of sintered deposits in coal-fired boilers. Senior and Srinivasachar10 developed a model for viscosity of silicates in the range of (5) Frenkel, J. Viscous Flow of Crystalline Bodies Under the Action of Surface Tension. J. Phys. (Moscow) 1945, 9, 385-391. (6) Hiram, Y.; Nir, A. A Simulation of Surface Tension Driven Coalescence. J .Colloid Interface Sci. 1983, 95, 462-470. (7) Mackenzie, J. K.; Shuttleworth, R. Phenomenological Theory of Sintering. Proc. Phys. Soc. (London) 1949, B62, 833-852. (8) Scherer, G. W. Sintering of Low-Density Glasses: I, Theory. J. Am. Ceram. Soc. 1977, 60, 236-239. (9) Raask, E. Mineral Impurities in Coal Combustion; Hemisphere Publishing: Washington, DC, 1985. (10) Senior, C. L.; Srinivasachar, S. Viscosity of Ash Particles in Combustion Systems for Prediction of Particle Sticking. Energy Fuels 1995, 9 (2), 277-283.
418 Energy & Fuels, Vol. 11, No. 2, 1997
Senior
Table 1. ASTM Ash Analysis (Weight Percent) oxide
Illinois No. 6
Wyodak-Rochelle
SiO2 Al2O3 TiO2 Fe2O3 CaO MgO Na2O K2O P2O5 SO3
45.0 18.0 1.0 20.0 7.0 1.0 0.6 1.9 0.2 3.5
32.1 17.8 1.0 5.2 24.8 10.0 1.0 0.3 1.2 7.3
total
98.2
100.7
Table 2. Ash Particle Size Distributions (in Weight Percent) Illinois No. 6
Wyodak-Rochelle
1-2.5 2.5-5 5-10 10-20 20-40 40-60
1.5 9.0 19.9 34.8 29.5 5.3
16.8 34.8 25.6 19.2 3.2 0.6
weighted av
10.2
4.2
diameter, µm
Figure 3. Densification of Illinois No. 6 deposit.
Figure 2. Viscosity-temperature curves for oxidizing conditions.
105-109 P. Other models of slag viscosity work well for flowing slag (i.e., viscosity j103 P) but cannot be applied to the much higher viscosities found in deposits. Using the ASTM ash composition (Table 1), the viscosity was calculated as a function of temperature from 1000 to 1400 K. Figure 2 displays the viscosity-temperature curves for oxidizing (all iron as Fe+3) conditions calculated as described in ref 10. Densification calculations were performed using the method of Mackenzie and Shuttleworth.7 The bulk ash composition was used to calculate the viscosity, and a surface tension of 400 dyn/cm was assumed on the basis of measurements of surface tension in a number of coal ash samples.11 The initial deposit porosity was assumed to be 50% since deposits are observed to have low strength (implying little sintering) at this value of porosity.3 On the basis of Scherer’s structural model for densification of a cubic lattice,8 the pore radius was set equal to half the average particle diameter (Table 2). Figures 3 and 4 show densification as a function of time for Illinois No. 6 and Wyodak-Rochelle ash, respectively. The temperature at which porosity reaches the critical value of 25% in ∼8 h is 1170 K for Illinois No. 6 and 1015 K for Wyodak-Rochelle. These points correspond to a viscosity of approximately 3 × 1010 P, which is consistent with Raask’s assessment of the range of viscosity for deposit sintering. Heat Transfer in the Deposit Having defined a maximum deposit surface temperature, Ts, such that the deposit can be easily removed, we can now compute the maximum flue gas tempera(11) Nowok, J. W.; Benson, S. A. Inorganic Transformations and Ash Deposition During Combustion; ASME: New York, 1991; pp 405424.
Figure 4. Densification of Wyodak-Rochelle deposit.
ture in the convective heat exchanger. The primary motivation for the present work is the design of a new type of convective air heater.2 To verify the approach, however, an example from a conventional coal-fired power plant will be presented first. Because we wish to find a maximum gas temperature, we will perform a steady state heat balance on the heat exchanger inlet using the maximum temperature for the working fluid. Figure 5 defines the system to be considered. The gas temperature can be calculated by solving the steady state heat transfer problem given the air and deposit surface temperatures. This involves solving three equations for three unknowns:
hfluid(Tfluid - Ti) )
ktube
(Ti - To) )
ri ln(ro/ri)
kash ri ln[(x + ro)/ro]
(To - Ts) )
hg(x + ro) (Ts - Tg) (2) ri
Variables are defined in Figure 5 and Table 3. The equations are solved for gas temperature, Tg, as a function of ash thickness, x. As a check on the calculational approach, the maximum allowable flue gas inlet temperature was estimated for a conventional steam convective air heater. Since the first bank of tubes in a power plant, the secondary superheater, is subject to a lot of radiation, the second bank, the reheater, was modeled. The gas
Removal of Coal Ash Deposits
Energy & Fuels, Vol. 11, No. 2, 1997 419
Figure 6. Maximum allowable flue gas inlet temperature for a steam reheater (Illinois No. 6 coal).
Figure 7. Maximum allowable flue gas inlet temperature for a steam reheater; comparison of Illinois No. 6 coal and PRB (Wyodak-Rochelle).
Figure 5. Temperature profile for steady state heat transfer calculation. Table 3. Values Used To Calculate Steady State Heat Transfer parameter working fluid working fluid temp tube material tube thermal conductivity tube i.d. tube o.d. ash thermal conductivity gas heat transfer coefficient working fluid heat transfer coefficient
symbol [units]
steam reheater
HIPPS air heater
ktube [W/m‚K]
steam 617 carbon steel 34.6
air 978 Inconel 16
ri [m] ro [m] kash [W/m‚K]
0.027 0.032 0.65
0.022 0.025 0.65
hg [W/m2‚K]
47.1
170
hfluid [W/m2‚K] 1039
545
Tfluid [K]
temperature entering the reheater is typically 1325 K (1920 °F) burning an Illinois No. 6 coal.12 Ash deposits of 0.5 in. in 8 h are typical and can be removed easily. Table 3 gives the parameters used for the steady state heat transfer calculation applied to a conventional steam reheater. Values of the gas and steam heat transfer coefficients were taken from the example cited in ref 12. We consider the case in which deposits are not assumed to sinter strongly because the deposit surface temperature remains below the critical point for (12) Stultz, S.; Kitto, J. Steam: Its Generation and Use, 40th ed.; Babcock and Wilcox: Barberton, OH, 1992; pp 21-8- 21-11.
densification. Thus, we chose a value for the ash thermal conductivity that is representative of nonfused deposits.13 The coal was assumed to be Illinois No. 6 coal having a maximum deposit surface temperature of 1170 K. The shaded region in Figure 6 shows that at a maximum gas temperature of 1325 K, deposits
