close

Вход

Забыли?

вход по аккаунту

?

Mineral Matter Transformation During Pulverized Coal Combustion.

код для вставкиСкачать
Dev. Chem Eng. Mineral Process., 9(3/4),pp.313-327,2001.
Mineral Matter Transformation During
Pulverized Coal Combustion
Eric G. Edding;, Kevin A. Davis’, Michael P. Heap’,
James R. Valentine’ and Adel F. Sarofim
Department of Chemical and Fuels Engineering, University of Utah,
Sdr Lake City, Utah 84112, USA
#
Reaction Engineering International, Salt Lake City, Utah 84101, USA
~~
~~~
Significant progress has been made in understanding and quantr&ing the processes
governing the transformation of the inorganic constituents of fiels, leading to the
promise of being able to predict the behavior of mineral matter in real systems. In
addition, modern computational models have progressed to the point that combustion
simulations can be now be realizedfor industrial-scale systems. The combination of
current understanding of mineral matter transformations with state-of-the-art
computational jluid 4namics (CFD)models provides a powerful tooI for analying
utiIity boilers.
A case stu& is presented illustrating how knowledge of the
transformation of pyrites coupled with CFD models can be used to explain problems
of corrosion that have been encountered in some utility boilers as a result of
modificarions to reduce emissions of nitrogen oxides.
Introduction
Coal-fued power plants provide more than one third of the electricity generation
worldwide (Eliasson, 1995) and will remain the prevailing source of electricity
generation well into the twenty-first century. Although the future of coal is different
in developing and developed countries, both are subject to economic and
environmental driving forces in the operation of existing power plants or construction
* Authorfor correspondence.
313
E.G.EaZings, K.A. Davis, M.P.Heap, et al.
of new ones. Few new coal-fired plants will be constructed in developed countries,
such as the U S A . Existing power plants, however, represent a large amount of
embedded capital, and financial considerations require that their useful life be
extended as long as possible. The combination of deregulation and the entrance into
the market of independent power producers has resulted in increased competition and
greater pressure to produce power at the lowest possible cost. In parallel with
increased competition, the industry is faced with increased stringency of
environmental regulations that increase the cost of power and will place additional
pressures to reduce costs on operators of existing plants. The major present concern
in the U.S. is for finding inexpensive methods for meeting the tighter emission
regulations. At present, the major focus is on NOx regulations, but increasing
attention is being paid to the emissions of mineral matter in response to the newly
promulgated regulations for fine particles and air toxics. The ability to predict
mineral matter transformations in real systems using computer simulations has been
shown to have great engineering benefits (Eddings et al., 2001).
The pressures for reduced costs and increasing environmental regulation will also
apply to the new plants that will be built primarily in developing countries such as
China and India (Takahashi, 1998). Traditionally, these plants have been subcritical
pulverized coal-fired units with electrostatic precipitators. Competition and the
scarcity of investment capital have decreased the price of such plants to %800/kW
(Takahashi, 1998).
There are increasing pressures by international lending
institutions to make new units cleaner and more efficient. Sulfur dioxide regulations
have been introduced in China (Xu et al., 1999). Supercritical units, which are 2-5
percent more efficient than subcritical units, are being introduced in China and their
rate of introduction will increase when local vendors become suppliers (Takahashi,
1998). However, the increase in steam temperatures to as high as 650°C in
supercritical units compared to about 540°C in subcritical units will require the use of
hi&er temperature steels as well as improved operation and maintenance to avoid
corrosion problems.
Boiler corrosion is an area that plagues nearly every stationary power plant and is
costly and poses a potential safety risk.
314
Corrosion in the boiler can result in
Mineral Matter Transformation During PCC
unscheduled outages. Corrosion, as well as slaggins and fouling problems, are major
contributors to the loss of availability. Maintenance costs can thus depend on the size
and composition of the larger ash particles which deposit on the walls.
The challenges of meeting the competitive and environmental targets can be
guided by the use of computer simulations. Although computer models for predicting
furnace performance have been under development for some time, it is only recently
that they have reached a state of maturity where industry is relying upon computer
simulations to solve their problems. The present paper will present a case study to
illustrate the value of computer models in evaluating a specific problem (waterwall
corrosion) associated with boiler modifications in response to NOx regulations.
Transformation of Iron Pyrite
Iron is one of the most significant elements to consider when discussing the effect of
mineral matter transformations on the operation of a coal-fired boiler. The presence
of iron in an ash deposit will often increase the stickiness or lower the melting point
of an ash deposit (Bool, et al., 1995). In addition, the presence of iron pyrites on steel
surfaces is suspected of enhancing waterwall corrosion rates over direct gaseous
attack (Eddings, et al., 1997). Iron is most often found in untreated coal samples in its
pyritic form; and is generally present in both included and extraneous forms.
Included pyrites are associated with other mineral matter and are distributed
throughout a given coal particle. Extraneous pyrites are typically isolated pyrite
particles resulting fkom the comminution process that have not been rejected in the
mill.
Table 1. Size distribution in microns for diflerent minerals in a Western Kentucky
#I 1 bituminous coai.
Mineral(wt%)
c2.j
2.5-5.0
5-10
10-20
20-40
> 40
Quartz (19)
19
33
34
26
10
25
16
24
14
9
0
15
23
28
21
20
14
4
18
16
12
8
12
6
9
14
21
41
15
5
4
4
Kaolinite (8)
Iliite (1 6)
Mix. Sil. (23)
Pynte (24)
Calcite (5)
All Min. (1 00)
1
24
15
14
54
13
11
33
0
16
315
E.G.Eddings, K.A. Davis, M.P. Heap, et al.
Table 1 shows size distributions for different mineral constituents of a Kentucky
coal. From the table it can be seen that pyrites have the largest number of large
particles. The greater size of the pyrite particles results in longer sulfide oxidation
times to reach the more benign iron oxides. When considering iron deposition on
boiler tube walls, large pyrite particles will have a greater tendency to deposit on
boiler tube walls while still retaining some sulfur, thus leading to potential corrosion
problems.
The transformation of pyritic iron varies depending on whether it is included or
extraneous. A summary of these transformations is given in Figure 1. The extraneous
pyrite reaction pathway follows the work of Srinivasachar and Boni (1989) and
initiates with a decomposition step of pyrite to pyrrhotite at around 800-900 K. The
pyrrhotite, which can undergo bgmentation if a critical porosity is achieved, will
then begin to oxidize at the available reaction surfaces. If temperatures are high
enough, the pyrrhotite can melt and the oxide layer will dissolve to form an Fe-0-S
melt. The Fe-0-S melt can be further oxidized to an iron oxide melt. If the particle
temperature is then cooled to below 1600 K, the iron oxide will crystallize as
magnetite.
g
Pyliloutc
Decomposiuon
to
oate
8W%OK+
efitatio
pm
Eg
%*a*
Pyrrboatc
Pynte Core
*w
Magnmte
Fra
0.
Fynhottte Melrr
0xid;gyIvc)
PyrrhOUtC core
Magncnte
Clysdllzes
Magnmte
0
Fc-0-S Melt
/
Q /MeltOUd12g
Iron oudc Melt
(a) Reaction pathway for extraneous pyrite (after Srinivasachar and Boni, 1989).
-
(b) Reaction pathway for included pyrite (after Bool, et al., 1995)
Figure 1. Reaction pathways for excluded (a) and inherent (3) pyrite during coal
cornbustion.
316
Mineral Matter Transformation During PCC
For included pyrite, there are some similarities but there is the added complication
of the mineral matter being affected by the burning char, as well as other materials in
the char. The reaction pathway for included pyrite, as described by Boo1 and coworkers (1999, also initiates with a decomposition to pyrrhotite; however, the
pyrrhotite cannot be oxidized until the char surface recedes sufficiently to allow
exposure to oxygen. The pyrrhotite can also melt while still encapsulated if the
particle temperature rises as high as 1356 K. Once the surface is exposed, the
pyrrhotite can either oxidize or mix with surrounding silicate materials to form a
glassy material. The iron oxide can be either molten or solid. If molten, the iron
oxide can crystallize to magnetite at temperatures below 1600 K, as with the
extraneous pyrite mechanism.
Coupling with CFD Calculations
As mentioned previously, computational fluid dynamics (CFD) tools have been
developed to the point where many practical problems of industrial interest can be
solved. For coal-fired utility boilers, the problems are typically related to evaluating
the viability of a modem retrofit, such as a low NOx firing system, and estimating the
potential for adverse side effects such as increased levels of unburned carbon,
additional depositiodfouling problems, and the potential for increased waterwall
corrosion. A case study will be presented here demonstrating how fundamental
models of ash behavior (specifically pyrite transformation), incorporated into such
CFD tools, can be used to address a problem related to mineral matter transformations
under low NOx conditions.
Overview of CFD Model. The computational tools used in this study were developed
by Reaction Engineering International (REI) to address a wide range of problems
involved in the operation and design of many combustion systems. These include
utility boilers, pyrolysis furnaces, gas turbine combustors, rotary kilns, waste
incinerators and smelting cyclones. The current models simulate both reacting and
non-reacting flow of gases and particles, including gaseous diffusion flames,
pulverized-coal flames, liquid sprays, coal slurries, injected sorbents, and other
oxidatiodreduction systems.
Emphasis has been placed on simulating coal
317
E.G. Etidings, K.A. Davis, M.P.Heap, et aL
combustion and pollutant formation. This three-dimensional, two-phase reacting flow
code (GLACIER) includes several capabilities necessary for accurate simulation of
coal-fired boilers. These capabilities include turbulent particle transport with full
coupling of particle and gas-phase mass and momentum; coal reaction processes such
as devoiatilization, char oxidation and gas-particle interchange; NO, formation and
reduction chemistry; particle convection and radiation with absorption, emission and
anisotropic scattering; full coupling of gas-particle energy exchange; and ash
deposition. In addition, boiler-side waterwall and radiant panel surface temperatures
can be predicted as part of the computation, given a back-side (i.e. steam) temperature
and surface resistance (from the deposit thickness and thermal conductivity, for
example).
GLACIER solves the governing gaseous fluid mechanical and reaction equations
in a Eulerian h e w o r k . Turbulence is incorporated with a k-E turbulence model for
closure that has been modified for the presence of particles. The turbulent fluid
mechanics affect many of the properties in the flow field. The gaseous reaction rates
are assumed to be limited by molecular-scale mixing. Different gas-phase mixture
hctions are defined for the local mass fraction of the inlet gas fuel, and the local
mass fiaction of gas evolved from the solid coal particles. Transport equations are
solved for each mean mixture fraction and its variance about the mean. Statistical
probability density functions (PDF’s) are used at each point in the flow field to obtain
mean properties of chemical composition, temperature, and other variables based on
local instantaneous equilibrium and then by convolution of the PDF. The PDF’s
account for the effects of turbulent fluctuations on the mean properties.
The particle mechanics are solved by following the trajectory of a discretized
group or cloud of particles in a Lagrangian fiame of reference. Coal devolatilization
and heterogeneous reaction rate processes are included using a variety of suggested
submodels fiom the literature. Heat, mass, and momentum transport effects are
included for each particle. The transport of thermal radiation is solved with a discrete
ordinates model that includes absorption, emission and non-isotropic (forward)
scattering from the particles. Particle deposition rates are calculated and include the
mechanisms of inertial impaction, turbulent displacement and thermophoretic
318
Mineral Matter TramformationDuring PCC
transport. Particles experience an elastic collision unless either the particle or wall
surface temperature exceeds a specified critical sticking temperature.
Pyrite Reaction Mechanism. Initial simulations performed in the case study described
below indicated the potential importance of understanding pyrite behavior in
identifying root causes of increased corrosion rates under low NOx firing. As a result,
consideration was given to the inclusion of a pyrite model in GLACIER. Evaluation
of pyrite transformation during coal combustion should address both inherent and
excluded sources. The key issue is whether the iron sulfide compounds are fully
oxidized prior to depositing on a waterwall surface. If the deposit consists solely of
iron oxides, it is unlikely that the deposit will result in any enhancement of corrosion
rates; however, if there are iron sulfide particles present in the deposit, the local sulfur
concentration will increase and there is potential for enhanced corrosion.
In order to incorporate the complexity of pyrite transformation into a multi-phase,
turbulent reacting CFD code such as GLACIER, some assumptions were required. As
a fist approximation, it was assumed that the inherent pyrite oxidizes at least as
rapidly as the char in the ash particles. If upon analysis of the simulation results
deposits of unburned carbon were found on the waterwall tubes, it could be assumed
that there was potential for unoxidized inherent iron sulfide species as well.
In order to examine the deposition of excluded particles of iron sulfide, it was
necessary to incorporate pyrite decomposition and iron sulfide oxidation models. The
model of Srinivasachar and Boni (1989) was found to be consistent with laboratory
data, and its formulation lent itself to implementation in GLACIER. The overall
pyrite conversion rate as described in their model includes the following mechanisms:
thermal decomposition of pyrite to pyrrhotite; pyrrhotite oxidation; and inclusion of
boundary layer and liquid layer diffusion. This model was adapted for use in
GLACIER and, as a verification step that the model described above was implemented
correctly, was used to simulate bench-scale data fiom Levasseur and co-workers
(1 989). The comparison between the GLACIER predictions and the experimental data
is shown in Figure 2 and illustrates that the model does a reasonable job of
representing the experimental results. The ignition point is not accurately represented
319
E.G. Eddings, K.A. Davis, M.P.Heap, et al.
at the lowest temperature, but the rate and level of sulfur release is accurately
reflected after about 150 ms.
-
/.=,a
I
I
-
5% 02,
95% N2
,,
d = 41.5 pm
-- 1311 K,
K,
- 1450
1722 K,
Predicted
Predicted
Predicted
r -
l
,d
0 1311 K, Measured
o
1450K, Measured
0.2
A 1722 K, Measured
0.1
0
0.1
0.2
0.3
0.4
0.5
Residence Time (s)
Figure 2. Comparison of GLACIERpredictions. with bench-scale data of Levassew
et al. (1989).
Case Study: Tangentially-Fired Pulverized Coal Boiler
A computational simulation was performed for a tangentially-fired (T-fired)
pulverized coal boiler which had undergone a low NOx firing system retrofit, and had
subsequently experienced a dramatic increase in waterwall corrosion rates. The unit
is a supercritical, divided furnace steam generator rated at 900 MW,. It consists of
two identical tangentially-fired furnaces sharing a common center waterwall. Each
furnace has two columns of wall-mounted burners located near the comers on the
front wall and two columns on the rear wall. The unit also has separated overfire air
(SOFA) ports above the burner columns that were added during the low-NOx retrofit.
The burner columns themselves consist of eight levels of coal burners alternating with
over- and under-fire auxiliary air inlets. The burner columns are located on the front
and rear walls and operate with the firing pattern shown in Fi,we 3. Note that
although this unit is tangentially-fired, it is not corner-fired. This figure also illustrates
a cross section of the computational mesh for the simulation, where each small arrow
represents a node point in the calculation. Although difficult to distinguish in the
320
Mineral Matter Tranrfomtion During PCC
resolution of this figure, there are over 5000 node points in the horizontal plane
shown, with much of the resolution in the near-burner region. For this simulation,
only one of the twin furnaces was modeled encompassing the lower end of the hopper
up to the boiler nose, and the simulation used an overall computational grid of
660,000 nodes.
Primary Firing Angles
Secondary Firing Angles
Figure 3. Primary (coal carrying) and secondav firing pattern.
Simulations were performed of both the pre-low-NOx-retrofit furnace, and the
post-low-NOx-retrofit furnace. The purpose for running both baseline conditions was
to provide insight on significant changes within the boiler that might have lead to the
sudden increase in waterwall corrosion.
Baseiine operating conditions were
somewhat variable, but were selected based upon discussions with plant engineering
and by comparisons with predicted observables for the baseline conditions.
Subsequent simulations were run utilizing modifications to the initial conditions
including excess air, wall deposit thickness, burner tilt, and close-coupled overfire air
distribution for comparison with historical observations based on plant experience.
These results provided confidence in both the initial input conditions used and the
model predictions, and a foundation was developed for evaluating various strategies
for improving furnace behavior. Expected and predicted NOx and carbon-in-ash are
presented in Table 2 for both pre- and post-low-NOx retrofit conditions.
Table 2. Expected and predicted NOx and carbon-in-ash values.
~
Pre-Retrofit
Post-Retrofit
Expected NOx
(Ppm)
Predicted NOx
(Ppm)
505
420
4-5
5
295
200
8 - 12
16
Expected C-inash (%)
Predicted Cin-ash (%)
321
E.G. Eddings, K.A. Davis, M.P. Heap, et al.
As part of this unit’s maintenance routine, water wall wastage measurements were
performed. Measured waterwall wastage rates were provided by the plant for both
furnaces (even though only one of the theoretically identical furnaces was simulated
in this study). Figure 4 shows the measured waterwall wastage rates for one of the
furnaces. Rates were similar for its twin. The measured data are plotted as contour
lines with the lowest line representing a wastage rate of 15 mildyear with increasing
intervals of 5 mils/year. This rear wall image is shown from the perspective of
standing inside the boiler. Measurements on the side and center walls, not pictured,
indicate much lower wastage rates. Wastage rate information from this and similar
h a c e s allows the following general conclusions to be made concerning the
observed waterwall wastage rates.
Figure 4. Measured waterwall wastage rates and predicted concentration of Fe as
FeS in material depositing on rear wall.
-
The areas of high wastage rates are localized not only with respect to their
relationship with the main heat release zone, but also with respect to particular
walls.
-
The lowest wastage rates were observed on the center wall with maximum rates of
29 and 14 mildyear for the two furnaces. The tube metal temperatures on the
center wall are lower than the other walls due to the water circulation pattern.
322
Mineral Matter Traqformatwn During PCC
- The highest wastage rates were observed on the fiont and rear walls, in all firing
confi,wtions.
The highest rates were measured in the area of the SOFA and not
in the burner zone; and the wastage patterns for the fiont and rear walls of both
furnaces are remarkably similar showing only minor differences.
-
There is a region of moderate wastage between the bumers and the adjacent
sidewall for four of the eight burner columns. The location of this area is the same
in both furnaces when the direction of rotation is taken into account.
Results and Discussion
GLACIER was not confi-wed to predict actual waterwall wastage rates but it could
indicate the conditions that might lead to increased waterwall wastage rates on the
furnace wall. These conditions are normally related to the presence of reducing
conditions generally indicated by high CO concentrations. Simulations of the pre- and
post-retrofit conditions were compared in order to examine the changes in conditions
on the waterwall that might affect wastage rates and that could be attributed to the
installation of the low NOx f
~ system.
g
As indicated earlier, the simulation of only
one furnace was attempted, but examination of the measured wastage rates indicated
that there is no significant difference in the wastage patterns in the two furnaces,
justifying the decision that it was unnecessary to simulate both furnaces.
The conditions on the furnace wall that can create aggressive waterwall corrosion
rates have been attributed to several factors, it is convenient to divide a discussion of
the results into three categories: Thermal Conditions; Gaseous Species Concentration;
and Deposition.
Thermal Conditions. Deposit temperature can affect waterwall wastage by altering
diffusion rates of aggressive species through the deposit, increasing the waterwall
surface temperature and if thermal cycling occurs, causing changes in deposit
morphology. This produces a tendency to spall, thus exposing fresh tube surface to
attack.
The gas temperatures in the computational cell next to the wall and the incident
heat flux for the rear and side walls of the furnace show two significant differences
between the pre- and post-retrofit conditions, and these are:
323
E.G.Eddings, K.A. Davis, M.P.Heap, et al.
-
The gas temperature adjacent to the wall in the region of high waterwall wastage
to the side of the burner stack is higher in the post-retrofit case. This is also one of
the two comers with relatively high CO concentrations in the burner zone.
-
The incident heat flux in the high wastage region at the SOFA level appears to be
higher in the pre-retrofit case.
Based upon these comparisons it could be argued that thermal effects alone are not
accountable for the high waterwall wastage rates observed in this T-fired unit, but that
they could contribute to the moderate wastage rates observed in the “hot” comer
beside the burner stack.
Gaseous Species Concentration on the
Wails. Waterwall corrosion in the lower
bate is generally associated with reducing conditions in the gases in close
proximity to the wall. This is associated with hydrogen sulfide being the species
responsible for the sulfidation reactions that lead to waterwall wastage. Hydrogen
sulfide is only favored under fbel rich conditions; therefore, examining the local gas
phase stoichiometry should give an indication of the presence of corrosive agents in
the region close to the wall. The firing system was designed to maintain an oxidizing
atmosphere on the furnace wall in the bumer region, and the post-retrofit simulation
indicates that with the exception of the hot comer and two other regions, most of the
rear wall is bathed in oxidizing gases. However, the gradients are very steep close to
the wall, passing from reducing to oxidizing in the space of a few inches.
In these simulations, the hydrogen sulfide concentration that is reported is the
thermodynamic equilibrium concentration for the local conditions. As such, when the
gas phase stoichiometry is close to around 0.9, the predicted hydrogen sulfide
concentration probably exceeds the actual concentration. In the high wastage region
at the SOFA level, there is almost no hydrogen sulfide on the rear wall. The comer
high wastage region however does correspond to the higher hydrogen sulfide
concentrations, but it should be remembered that the SOFA region has the highest
measured wastage rates. Thus, based upon the simulations, it appears that hydrogen
sulfide alone cannot account for the high wastage rates observed in this T-fired unit.
One factor not accounted for in this thinking is intermittency. If hydrogen sulfide
concentration gradients in the SOFA region are very steep it is possible that: (a) the
324
Mineral Matter TransformationDuring PCC
furnace walls are constantly being swept by gases of varying composition sometimes
very fuel rich and sometimes very lean; and (b) this intermittency could affect
wastage rates. It is possible that GLACIER does not model the large-scale turbulent
fluctuations very well and does not accurately portray this intermittency. It is thought
that fluctuations fiom reducing to oxidizing could continually remove an outer tube
layer and expose fiesh tube surface to corrosive species. Certainly there are very
steep concentration gradients close to the wall in the SOFA region, and it is probably
such a region that will experience this intermittent exposure to oxidizing and reducing
conditions.
Particle Deposition. Ash and partially reacted coakhar are shown to be deposited on
the furnace walls in the simulation. In agreement with plant observations, the
simulation also indicates that the lower furnace walls experience decreased deposition
under post-retrofit conditions. When significant char is deposited it will create locally
reducing conditions. As it also contains sulfur, it will deposit material that can
transport corrosive agents directly into the solid layer coating the waterwalls. The
coal s u l k can be organic, associated within the hydrocarbon coal matrix, or it can be
inorganic, present primarily as the included or excluded iron pyrite. Deposition of
iron sulfide can thus create aggressive corrosion conditions.
Figure 4 shows the hction of the total mass of iron being deposited on the rear
wall that is iron sulfide. In the post-retrofit condition there is less deposition (on a
mass basis) but more of it is unburned material.
The regions of high pyrite
concentration in the material correspond to both of the high waterwall wastage
regions. Particles are transported to the wall by impaction, turbulent dispersion and
thermophoresis. Particles experience an elastic collision unless either the particle or
wall surface temperature exceeds a specified critical temperature. In all of this study,
this critical temperature was assumed to be the Tsso of the whole coal fly ash.
Variation of this critical temperature would vary the deposition pattern of unburned
coal.
An advantage of a CFD model is that the source of the deposit can be easily
identified. By following particle trajectories, the burners that are contributing most
significantly to the deposit can be identified. In excess of 98% of the total unburned
325
E.G. Eddings, K.A. Davis, M.P. Heap, et al.
coal (qualitatively very similar to the unoxidized sulfide) that is deposited on the rear
wall between the SOFA ports originates fiom the ffont wall burners. Sixty-eight
percent originates from level six (there are eight levels) and the burners adjacent to
the center walls account for 84% of the unburned carbon. The largest particle size,
250 microns, accounts for 62% of the unburned coal deposition. 175 micron particles
account for 28% and 120 micron particles account for 15%. There is even a small
contribution fiom 85 micron particles originating fiom the center burner, sixth level.
Conclusions
CFD-based models have been developed to the point where they are viable tools for
analyzing industrial systems. Complex systems, such as the transformation of mineral
matter in coal-fired utility boilers, can be simulated when a fundamental knowledge
of controlling phenomena aie understood. Based on a fundamental understanding,
engineering simplifications can be made to allow the inclusion of important physical
processes in a coupled yet tractable manner in a reacting CFD code.
A review of the current understanding of fundamental mechanisms associated with
mineral matter transformation during pulverized coal combustion was given. A case
study was presented illustrating how an appropriately confi,pred CFD model could be
utilized to identify the relative contributions of various parameters in enhanced
.
corrosion of waterwall tubes. The key findings fiom the case study include:
In areas of high waterwall wastage the characteristics of the solids being deposited
are more important than the gaseous reducing conditions on the wall. There is a
possibility that the most aggressive conditions are created when the deposits are
rich in unburned coal and iron sulfide and possibly, although not discussed in this
.
paper, where the gas phase fluctuates between oxidizing and reducing.
Deposition of unburned coal and iron sulfide in fuel-rich regions produce less, but
still very significant, aggressive conditions.
Significant but lower waterwall wastage rates can occur under a combination of
circumstances which involve gaseous reducing conditions on the wall, deposition
of unburned material and high incident heat fluxes.
326
Mineral Matter TransformationDuring PCC
References
Edding, E.G.,K.A. Davis, M.P.Heap, J.R. Valentine, A Facchiano, R. Hardman, N. Grigas. 1997. Effect
of Low NOx Firing Conditions on Increased Carbon-in-Ash and Water-Wall Corrosion Rates.
Proceedings of the EPRVEPA/DOE Mega Symposium on NWSWAir Toxics, Washington, D.C.,
A u ~ U 25-29.
&
Eddings, E.G., A.F. Sarofim, C.M. Lee, K.A. Davis and J.R. Valentine. 2001. Trends in Predicting and
Controlling Emissions fiom Coal-Fired Boilers, Fuel Processing Technology, In Press.
Eliasson, B. 1995. Energy: Statistics and Scenarios, 1994. ABB Corporate Research Center, Switzerland.
Levasseur, A.A., Goetz, G.J., Lao, T.C., and Patel, R.L. 1989. The Fundamental Role of Iron in Coal Ash
Deposition. Combustion Engineering, Connecticut, USA, Draft RepoR submitted for publication.
Srinivasachar, S. and Boni, A.A. 1989. A kinetic model for pyrite transformations in a combustion
environment. Fuel, Vol. 68, pp. 829-836.
Takahashi, H. 1998. Technologies for Reducing Emissions in Coal-Fired Power plan?^. Energy Issues,
The World Bank,Vol. 14, pp. 1-8.
Xu,X., Chen, C., Qi, H., He, R., You, C., and Xiang, G. 1999. The Air Pollution Control Measures in
China. Conference on International Aid and Clean and Effrcient Coal Technology in China, Tokyo,
Japan.
327
Документ
Категория
Без категории
Просмотров
0
Размер файла
763 Кб
Теги
pulverized, transformation, matter, coal, combustion, mineraly
1/--страниц
Пожаловаться на содержимое документа