EPA/600/A-96/112
Equilibrium Speciation of As, Cd, Cr, Hg, Ni, Pb, and Se
in Oxidative Thermal Conversion of Coal
- A Comparison of Thermodynamic Packages
Flemming Frandsen^, Thomas A. Erickson^, Vit Kiihnel^,
Joseph J. Helble^, and William P. Linak^
1) Department of Chemical Engineering, Technical University of Denmark, Building 229,
DK-2800 Lyngby, Denmark.
2) Energy & Environmental Research Center, University of North Dakota, Grand Forks,
ND, USA.
3) Chemical Engineering Department, University of Connecticut, Storrs, CT, USA.
4) U. S. Environmental Protection Agency, National Risk Management Research
Laboratory, Research Triangle Park, NC, USA.
0: Corresponding author.
Abstract
The equilibrium speciation of the trace elements As, Cd, Cr, Hg, Ni, Pb, and Se, in a
model system, is predicted assuming ideal gas and pure condensed phases and taking into account
the presence of a sorbent material for sulfur capture. The model system is based on a
demonstration-scale pressurized fluidized-bed combustor, firing a Pittsburgh No. 8 bituminous
coal, using dolomite as sorbent for sulfur capture, and equipped with a ceramic filter for particulate
removal.
Results from four different thermodynamic packages (MINGTSYS, NASA-CET89,
FACT, and SOLGASMEX) are compared over the temperature range [700 -2000 K], at pressures
of 1 and 20 atmospheres. For the elements As, Hg, and Se almost identical equilibrium
distributions were predicted by the four packages, while for the elements Cd, Cr, Ni, and Pb,
differences in the output from the four packages were observed. These differences may be due to
different solution techniques, convergence criteria, and/or thermodynamic input data utilized in the
thermodynamic packages.
This paper contains an outline of the equilibrium distributions, the comparison, a
discussion of the reasons for the different results obtained with the different thermodynamic
packages, and a comparison between measured and predicted partitioning data.
Keywords
Coal, Combustion, Emission, Flue Gas Cleaning, Thermodynamic Calculations, PFBC, Trace
Elements
1. Introduction
The elements contained in fossil and biomass fuels can be grouped into three concentration
-------
levels: 1) the major elements, C, O, H, S, and N, building up the organic matrix of the fuel; 2) the
ash forming elements, Al, Ca, Fe, K, Mg, Na, and Si, typically present in the concentration range
from about 1000 ppmw to a few %(w/w) on a dry fuel basis; and 3) the trace elements (e.g., As,
B, Cd, Cr, Hg, Ni, Pb, and Se) present typically in concentrations below 1000 ppmw [Swaine
(1990)]. Several of the trace elements are vaporized during pyrolysis and combustion and
recondensed on the surface of fly ash particles during subsequent cooling of the flue gas [see e.g.,
Davidson etal. (1974) and Swaine and Goodarzi (1995)]. Full-scale measurements have revealed
significant amounts of some trace elements (e.g., B, Hg, and Se) in the flue gas leaving the stack
[Meij et al. (1984), Clarke and Sloss (1992), and Sander (1993)]. Direct gaseous emission of these
trace elements is undesired because of suspected toxicological effects on the environment and
potential genetic or biological changes in living creatures [Swaine (1990), Clarke and Sloss (1992),
Swaine and Goodarzi (1995), and Davidson and Clarke (1996)].
As a first approach, Global Equilibrium Analysis (GEA) has been used for several years in
order to understand the various subprocesses in thermal fuel conversion systems. Frandsen et al.
(1994) provided an introduction to the application of GEA on combustion systems, by utilizing the
Gibbs energy minimization code MINGTSYS and reporting equilibrium distributions for 18 trace
elements of concern with respect to coal utilization [Swaine (1990)]. Reducing and oxidizing
conditions were considered, and the results were compared qualitatively to experimental results
published in the literature [Frandsen et al. (1994) and Frandsen (1995)].
Numerous thermodynamic packages have been developed for the purpose of minimizing
the Gibbs free energy of mass balance constrained combustion systems containing ash forming and
trace element species [Vuthaluru etal. (1994)]. Differences in the output from those packages may
be due to: 1) use of different thermochemical data, 2) different numerical solution techniques, or 3)
different convergence criteria utilized. Thus, a GEA test was performed in order to compare the
output from different available thermodynamic packages. This paper contains a summary and the
major conclusions of this GEA test.
2. Description of the Modeling Approach
The total Gibbs energy, G[, of a chemical system is given by:
G' N G°fi
R • T = n> * [ R • T + ln^]
where G symbolizes a Gibbs energy, superscript t denotes total, R is the universal gas constant, T
is the absolute temperature, N is the total number of species, n-, is the number of moles of species
i, superscript o denotes standard state, subscript fi denotes formation of species i, and a, is the
activity of species i. The function Gl is combined with the mass balance constraints of the system
and minimized, using the method of undetermined Lagrangian multipliers [Eriksson (1975)].
A GEA model, based on an actual operating combustion system, has been set up. The
temperature, pressure, and total elemental composition are specified. The gas is assumed ideal, and
all condensed phases are considered pure. Among the ash forming elements, only calcium, Ca, is
taken into account as an As-, Cr-, and S-capturing sorbent, in the form of CaCOj. The trace
elements As, Cd, Cr, Hg, Ni, Pb, and Se are considered one at a time. Four different
thermodynamic packages, MINGTSYS, NASA-CET89, FACT, and SOLGASM1X, have been
used to minimize the total Gibbs energy of a model system with a well-defined chemistry; i.e., a list
2
-------
of chemical species to be considered, and elemental composition. MINGTSYS has been chosen
since it provides a baseline study of trace element equilibrium chemistry in hot flue gases [Frandsen
et al. (1994)], while the other packages are commercially available and commonly used worldwide.
A detailed description of the combustion system, a list of the combustion products and trace
element species, and an introduction to each of the thermodynamic packages utilized in this study
will be given in a future paper.
3. Major Results of the GEA Test
In this section, the thermodynamic stable species of the trace elements As, Cd, Cr, Hg,
Ni, Pb, and Se are outlined and compared internally. Concentrations used by the models are
presented in Table 1.
For the elements As, Hg, and Se at standard pressure, 1 atm.. all four codes predicted
equilibrium distributions equal to those reported by Frandsen et al. (1994). At elevated pressure, 20
atm.. the equilibrium distributions of As, Hg, and Se, predicted by the four codes, were also
consistent. For the elements Cd, Cr, Ni, and Pb, observed differences in the equilibrium
distributions predicted by the four codes are briefly outlined below.
3.1. Cadmium (Cd)
At standard pressure, 1 atm.. the MINGTSYS code found Cd to exist as CdCh at
temperatures up to 1100 K: CdCl2(cr) below and CdCl2(g) above 750 K. Above 1200 K Cd(g) and
CdO(g) are the stable species, Cd(g) accounting for more than 95 %(mol/mol) of the Cd present at
all temperatures. CdO(g) is gradually decomposing as the temperature is raised above 1280 K,
forming Cd(g) and 02(g).
The NASA-CET89 code found CdCl2(cr) to be stable below and Cd(g) to be the major
stable form of Cd above 750 K. Small amounts [< 2 %(mol/mol)] of the Cd were present as
CdO(g) in the temperature range 750 -1000 K.
The FACT code found CdCl2(cr) to be stable up to 750 K, above which temperature
CdO(cr) is stable up to 1000 K, see Figure 1. Formation of Cd(g) and CdO(g) begins at 850 K,
Cd(g) being the major species above 1000 K. CdO(g) gradually decomposes forming Cd(g) and
02(g) above 1000 K.
The SOLGASMIX code found CdO(cr) to be the stable form of Cd at temperatures up to
1400 K. Formation of Cd(g) begins at 1200 K. Above 1600 K, Cd(g) is the major stable form of
Cd present. In addition, minor amounts of CdO(g) and CdCl2(g) were formed in the temperature
ranges [1300 - 2000 K] and [700 -1500 K], respectively.
At elevated pressure, 20 atm.. all four codes found equilibrium distributions of Cd equal
to the ones reported for 1 atm., but with an increase in temperature of about 100 K; i.e., the
MINGTSYS and NASA-CET89 codes found CdCl2(cr) to be stable below 850 K. The relative
amount of CdO(g) formed in the system is increased when the pressure is increased from 1 to 20
atm. Thus, elevated pressure seems to favor formation of CdO(g) at high temperatures.
3.2. Chromium (Cr)
At standard pressure, 1 atm.. MINGTSYS and NASA-CET89 found Cr203(cr) to exist at
temperatures up to approx. 1200 K, where it is decomposing, forming Cr02(OH)2(g) (see Figure
2). The latter has a maximum occurrence around 1300 K. Above 1300 K, chromium showed a
3
-------
very complex equilibrium chemistry, forming the gaseous components, CrO(OH), Cr02(0H),
CrO(OII)2, Cr02, and Cr03.
The FACT code showed a completely different distribution of Cr at 1 atm., primarily due
to 1) a lack of thermodynamic data for the gaseous chromiumoxides and -hydroxyoxides provided
by Ebbinghaus (1993) and 2) an implementation of the composed oxide, CaO.Cr2C>3(cr), which
was shown to be thermodynamically stable between 800 and 1800 K (see Figure 3).
The SOLGASMIX code found Ca0.Cr203(cr) to be the major Cr-species stable in the
temperature range [700 - 2000 K]. In addition, minor amounts of Cr2C>3(cr), Ci02(g), and CrC>3(g)
were formed at temperatures above 1450 K. The thermochemical data for the gaseous
chromiumoxides and -hydroxyoxides reported by Ebbinghaus (1993) have not been included in
these calculations.
At elevated pressure, 20 atm., MINGTSYS and NASA-CET89 found Cr02(OH)2(g) to be
the major stable form of Cr in the temperature range [900 - 1800 K], Below 900 K, Cr2C>3(cr) was
stable, and below 1500 K, Cr02(OH)2(g) was decomposing, forming primarily the gaseous Cr-
species CrO(OH), Cr02(OH), CrO(OH)2, and CrO(OH)3.
The FACT code found almost the same equilibrium distribution of Cr as reported for 1
atm., but with an increase in temperature of about 150 K. Ca0.Cr203(cr) was found to be stable in
the temperature range [900 - 1900 K],
In case Ca0.Cr203(cr) was not taken into account, the SOLGASMIX code found
Cr203(cr) to be the major stable form of Cr in the temperature range [700 - 2000 K]. In addition,
minor amounts of Cr02(g) and CrC>3(g) were formed above 1600 K. If taken into account,
Ca0.Cr203(cr) was the major Cr-species formed in the temperature range [700 - 2000 K],
3.3. Nickel (Ni)
At standard pressure, 1 atm.. the MINGTSYS and FACT codes found Ni to form
NiO(cr), being stable up to approximately 1500 K, where NiO(g), Ni(g), and NiCl(g) were
formed, see Figure 4. Formation of NiCl2(g) begins at 1000 K. NiO(g), NiCl(g), and NiCl2(g)
were stable up to 1800 K, where the last trace amount of NiO(cr) disappeared. Above 1825 K,
existing equilibrium relations among NiO(g), NiCl(g), NiCl2(g), and Ni(g) were shifted towards
Ni(g) as the temperature was increased. Calculations performed with the FACT code indicated that
implementation of Ni(OH)2(g), which was not included in the common list of species taken into
account in this study, could affect the stability of NiO(cr). The NAS A-CET89 code also
predicted NiO(cr) to be stable at temperatures up to 1500 K, where formation of NiO(g), Ni(g),
and NiCl(g) has started, but this code predicted a much higher concentration of Ni(g) and
correspondingly lower concentrations of NiO(g), NiCl(g), and NiCl2(g) above 1750 K, than the
MINGTSYS and FACT codes.
The SOLGASMIX code found NiO(cr) to be the major stable form of Ni. Various gaseous
Ni-species (Ni, NiO, NiCl, and NiCl2) were formed at temperatures above 1600 K. As with the
FACT code, the output from the SOLGASMIX code indicated that Ni(OH)2(g) may affect the
stability of NiO(cr).
At elevated pressure, 20 atm.. all four codes found the crystalline nickel oxide, NiO(cr), to
be stable at temperatures below 1850 K, and the SOLGASMIX code, even up to 2000 K. Above
1100 K, varying amounts of NiO(g), Ni(g), NiCl(g), and NiCl2(g) were formed. The NASA-
4
-------
CET89 code again predicted a much higher concentration of Ni(g) and correspondingly lower
concentrations of NiO(g), NiCl(g), and NiCl2(g) above 1750 K than the other codes.
3.4. Lead (Pb)
At standard pressure, 1 atm.. MINGTSYS and NASA-CET89 found PbCl4(g) to be the
major stable form of Pb below 1100 K (see Figure 5). Above 1300 K, PbO(g) was the major stable
Pb-species, but it was gradually decomposed with increasing temperatures above 1500 K, forming
Pb(g). Peaks of PbC^g) and PbCl(g) were in the temperature ranges [1000 - 1500 K] and [1100 -
2000 K], respectively. PbChCg) has a maximum occurrence of approx. 48 %(mol/mol) at 1200 K,
while PbCl(g) has a maximum occurrence of approx. 8 %(mol/mol) at 1250 K.
The FACT and SOLGASMIX codes showed an equilibrium distribution of Pb similar to
that reported above, but the relative amounts of PbCl(g) and PbCl2(g) formed were smaller than
predicted by the MINGTSYS and NASA-CET89 codes.
The output from the SOLGASMIX code indicated formation of significant amounts of
PbO(cr) in the intermediate temperature range [1050 - 1800 K], No condensed phases of Pb were
found below 1000 K.
At elevated pressure, 20 atm.. all four codes predict equilibrium distributions of Pb equal
to those reported at 1 atm., but at about 150 K higher temperature. And again, the relative amounts
of PbCl(g) and PbCl2(g) predicted by MINGTSYS and NASA-CET89 were higher than those
predicted by FACT and SOLGASMIX. Also, the SOLGASMIX code predicted formation of
significant amounts of PbO(cr) in the intermediate temperature range [1100 - 2000 Kj, but no
condensed phases were formed below 1100 K.
4. Comparison with Experimental Data
As part of this work, the calculated equilibrium distributions presented above have been
compared to trace element partitioning measured in the demonstration-scale 70 MWe full load Tidd
pressurized bubbling fluidized-bed combustor located in Brilliant, Ohio. This plant is operated by
Ohio Power Company, a subsidiary of American Electric Power.
Crushed Pittsburgh No. 8, bituminous coal is combined with water from a nearby river to
produce a coal paste of approximately 25 % (w/w) moisture. The paste is fed to the combustor
along with crushed dolomite. The material is fluidized by high velocity combustion air in the water-
cooled boiler. Mean temperature in the combustor was controlled at approximately 1100 K [Radian
(1994)]. After releasing heat to the in-bed, water-cooled tubes, the particulate-laden combustion
gases are led into seven parallel, two-stage cyclones. The cyclones remove approximately 93
%(w/w) of the entrained solids (primarily sulfated lime, unreacted lime, ash, and unburned
carbon). After the cyclones, the combustion gases flow through a gas turbine, to an electrostatic
precipitator before the gases are released to the atmosphere. Bed ash, comprising about 45 % (w/w)
of the total ash produced, is removed from the bottom of the combustor periodically through a lock
hopper system [Radian (1994)].
A research feature of the Tidd plant is a demonstration scale hot gas cleanup system.
Treated gas from one of the seven cyclone systems is diverted to a ceramic barrier advanced particle
filter (APF) and backup cyclone, and directed back to the outlet header of the secondary cyclone.
The APF uses Schumacher silicon carbide candles in a cluster/plenum arrangement developed by
5
-------
Westinghouse Corporation to filter the gas [Radian (1994)]. In Table 2, the total trace element
concentration in |ig/Nm3 including gas and particulate material at the APF inlet (1000 K and 9.2
atm.) and outlet is shown.
Notice in Table 2, that the total concentration of the trace elements As, Cd, Cr, Ni, and Pb
is reduced significantly in the APF. For the elements Hg and Se, a moderate reduction in the total
concentration is observed. These results correspond well with the trace element partitioning
outlined in section 3, since (as also indicated in Table 2) the trace elements As, Cd, Cr, and Ni are
expected to be in a condensed form at 1000 K and 9.2 atm. Anyhow, the output from the four
codes indicate that gaseous forms of Cd and Cr could be stable at 1000 K. The predicted behavior
of lead is in discrepancy with observations. All four codes predict Pb to be in the gas phase in the
form of PbCl4(g) at 1000 K. However, the high reduction of Pb across the APF suggests a
condensed Pb species. The condensation of PbO(g) on solid S1O2 is the likely removal mechanism
[Owens and Biswas (1996)]. For Hg and Se, all four codes predict thermodynamic stable gaseous
forms at 1000 K. Thus, the moderate reduction in the total concentration across the APF is not very
surprising.
5. Summary and Discussion
The equilibrium speciation of As, Cd, Cr, Hg, Ni, Pb, and Se has been predicted at 1 and
20 atm. in the temperature range [700 - 2000 K], by use of the thermodynamic packages
MENGTSYS, NASA-CET89, FACT, and SOLGASMIX.
For the elements As, Hg, and Se, almost identical equilibrium distributions were predicted
by the four packages, while for the elements Cd, Cr, Ni, and Pb, differences in the output from the
four packages were observed in this study. The greatest differences were due to the presence or
absence of specific chemical species, such as Ni(OH)2(g) and the chromium oxyhydroxides, in the
respective databases of the four packages. Differences due to the thermodynamic data for individual
chemical species were also significant. For example, the amount of Ni(g) at 2000 K predicted by
the NASA-CET89 code suggests that other thermodynamic data have been used than in the
remaining packages.
When GEA is used on a thermal fuel conversion system one has to be aware of the
following [Frandsen et al. (1994), Linak and Wendt (1994)]:
1) All relevant chemical species occurring in the thermal fuel conversion system must be
taken into account, otherwise the output from the GEA will be misleading.
2) Consistent thermodynamic data must be used.
3) Appropriate mixing models (pure phases, ideal or non-ideal mixing) should be applied in
the condensed phases.
4) In thermal fuel conversion systems, mixing phenomena and/or boiler design characteristics
may introduce local conditions (e.g., temperature and/or composition gradients) ignored in
the GEA.
Item 1) above has been illustrated through the comparison of thermodynamic packages
performed in this study. All possible chemical species in an actual thermal fuel conversion system
must be included when performing a GEA, otherwise the results and conclusions will be
misleading. A complete proof of this will require a systematic combination of the different Gibbs
energy minimization codes and thermochemical databases utilized in the four thermodynamic
6
-------
packages compared. Thus, a certain standardization of the available thermochemical data for trace
element species may eliminate some of the differences outlined in section 3 of this paper. This will
also help to provide a relative accountability of the four packages.
In addition, this study has revealed the necessity of a systematic scanning for possible
compounds of trace element fixation by ash species; e.g., Ca0.Cr203 or (Pb0)n.Si02. This may
also help to clarify the equilibrium chemistry of trace elements in flue gases from thermal fuel
conversion systems.
Acknowledgement
The authors are grateful to the Combustion and Harmful Emission Control (CHEC)
research program at the Department of Chemical Engineering, Technical University of Denmark,
for financial support. The CHEC research program is cofunded by ELS AM (The Jutland-Funen
Electricity Consortium), ELKRAFT (The Zealand Electricity Consortium), the Danish Technical
Research Council, and the Danish and the Nordic Energy Research programs.
The authors are also grateful to: the National Center for Excellence on Air Toxics Metals at
the EERC, University of North Dakota, the Environmental Research Institute, University of
Connecticut, and the Air Pollution Prevention and Control Division, EPA/NRMRL. Grateful
acknowledgement is also made to Bonnie McBride, NASA Lewis Research Center, for providing
the NASA-CET89 code and other assistance.
References
Clarke, L. B. and L. L. Sloss (1992). Trace Elements Emissions from Coal Combustion and
Gasification, IEA Coal Research, Report No. IEACR/49, London, UK.
Davidson, R. L., D. F. S. Natusch, J. R. Wallace, and C. A. Evans, Jr. (1974). Trace Elements in
Fly Ash Dependence of Concentration on Particle Size, Environ. Sci. Technol., 8(13), 1107-
1113.
Davidson, R. M. and L. B. Clarke (1996). Trace Elements in Coal., IEA Coal Research, Report
No. IEAPER/21, London, UK.
Ebbinghaus, B. B. (1993). Thermodynamic of Gas Phase Chromium Species: The Chromium
Oxides, the Chromium Oxyhydroxides, and Volatility Calculations in Waste Incineration
Processes, Combust. Flame, 93, 119-137.
Eriksson, G. (1975). Thermodynamic Studies of High-Temperature Equilibria XII, Chemica
Scripta, 8, 100.
Frandsen, F. J., K. Dam-Johansen, and P. Rasmussen (1994). Trace Elements from Combustion
and Gasification of Coal - An Equilibrium Approach, Prog. Energy Combust. Sci., 20, 115-138.
Frandsen, F. J. (1995). Trace Elements from Coal Combustion, Ph.D.-Thesis, Dept. Chem. Eng.,
Technol. Univ. of Denmark.
7
-------
Linak, W. P. and J. O. L. Wendt (1994). Trace Metal Transformation Mechanisms during Coal
Combustion, Fuel Process. Technol., 39, 173-198.
Meij, R., J. Van der Kooij, J. L. G. van der Sluys, F. G. C. Siepman, and H. A. van der Sloot
(1984). Characteristics of Emitted Fly Ash and Trace Elements from Utility Boilers Fired with
Pulverized Coal, KEMA ScL & Technol. Reports, 2(1), 1-8.
Owens, T. M. and P. Biswas (1996). Reactions Between Vapor Phase Lead Compounds and In-
situ Generated Silica Particles at Various Lead-Silicon Feed Ratios: Applications to Toxic Metal
Capture in Combustors, J. Air & Waste Manage. Assoc., 46(6), 530 - 538.
Radian (1994). A Study of Hazardous Air Pollutants at the Tidd PFBC Demonstration Plant., Final
Report DCN 94-633-021-03, Radian Corporation, Austin, Texas.
Sander, B. (1993). Measurements of Trace Element Mass Balances in Coal-fired Power Plants
Equipped with Different Types of FGD Systems, Proc. 2nd Int. EPRI Conf. on Managing
Hazardous Air Pollutants, Washington D.C., July 13-15,
Swaine, D. J. (1990). Trace Elements in Coal., Butterworth and Co, Ltd., London, UK.
Swaine, D. J. and F. Goodarzi (1995). Environmental Aspects of Trace Elements in Coal., Kluwer
Ac. Pub., Dordrecht, The Netherlands.
Vuthaluru, H. B„ S. Eenkhoorn, J. H. A. Kiel, and H, J. Veringa (1994). Trace Element
Emissions - Literature Review, ECN Report No. ECN-C-94-096, ECN, The Netherlands.
8
-------
Reactor feedstock mass flows fkg/hrt:
Coal: Sorbent: Moisture =
15.2 : 8.7 : 5.2
%
-------
CdO (cr) Cd (g)
1 oo
oo go -
a)
o
a)
p
OO go -
~o
o
03
O
1_
-------
1 oo •
CO
O
CO
ex.
40
20
Cr203 (cr)
Cr02(0H)2 (g)
700
CrO(OH) (g)
Cr02(0H) (g)
CrO(OH), (g)
900 1100 1300 1500 1700 1900
Temperature ( K )
Figure 2: Equilibrium distribution of Cr at 1 atm. as found by the MINGTSYS and
NASA-CET89 codes. 1: Cr03{g) and 2: Cr02(g).
11
-------
Ca0.Cr203 (cr)
i oo
oo so -
_cu
*o
CD
CL
CO 60-
40 -
C
CD
O
CD
Q_ 20 --
700
900
1 100
1 300
1 500 - 1 700
1900
Temperature (K)
Figure 3: Equilibrium distribution of Cr at 1 atm. as found by the FACT code.
12
-------
NIO (cr)
100
co 80-
Q_ 20 -
NiCI (g)
1300
700
900
1 1 OO
1 500
1 700
t 900
Temperature (K)
Figure 4: Equilibrium distribution of Ni at 1 atm. as found by the MINGTSYS and
FACT codes.
13
-------
PbCL (g)
i oo
PbO (g)
00 80 -
a>
*o
CD
Q_
OO 60 -
1
_Q
Q—
4-0 -
C
CD
O
Pb (g)
-------
"MRMP T -DTD-D-1Q7 TECHNICAL REPORT DATA
IN X\ IVin riir sr 10 i (Please read Instructions on the reverse before completing)
1
1. REPORT NO, , 2.
EPA/600/A-96/112
3. RECIPIEf
4. TITLE ANO SUBTITLE
Equilibrium Speciation of As, Cd, Cr, Hg, Ni, Pb,
and Se in Oxidative Thermal Conversion of Coal--
A Comparison of Thermodynamic Packages
S. REPORT OATE
6. PERFORMING ORGANIZATION CODE
7. author(s)F. Frandsen (Tech Univ of Denmark); T.Erick-
son and V.Kuhnel (Univ of ND); J. Helble (Univ of CT);
and W. Linak (EPA)
8. PERFORMING ORGANIZATION REPORT NO.
9. performing oroanization NAME and address Xech Univ of Den-
mark, Lyngby, Denmark; Univ of N. Dakota, Grand
Forks, ND 58202; Univ of Conn., Storrs, CT 06268.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
NA (Inhouse)
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Published paper; 6/95-6/96
14. SPONSORING AGENCY CODE
EPA/600/13
16. supplementary notes A PPCD project officer is William P. Linak, Mail Drop 65, 919/
541-5792. Presented at 3rd Int. Symp. on Gas Cleaning at High Temperatures,
Karlsruhe. Germany. 9/18-20/96.
i6. abstract xhe paper predicts the equilibrium speciation of the trace elements As. Cd,
Cr, Hg, Ni, Pb, and Se in a model system, assuming ideal gas and pure condensed
phases and taking into account the presence of a sorbent material for sulfur capture.
The model system is based on a demonstration-scale pressurized fluidized-bed com-
bustor, firing a Pittsburgh No. 8 bituminous coal, using dolomite as sorbent for sul-
fur capture and a ceramic filter for particulate removal. Results from four different
thermodynamic packages (MINGTSYS, NASA-CET89, FACT, and SOLGASMX) are
compared over the temperature range 700-2000 K, at pressures of 1 and 20 atm. For
As, Hg, and Se, almost identical equilibrium distributions were predicted by the
four packages; while forCd, Cr, Ni, and Pb, differences in the output from the four
packages were observed. These differences may be due to different solution techni-
ques, convergence criteria, and/or thermodynamic input data utilized in the thermo-
dynamic packages. The paper outlines the equilibrium distributions, the comparison,
a discussion of the reasons for the different results obtained with the different ther-
modynamic packages, and a comparison between measured and predicted partitioning
data.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Gas Scrubbing
Coal
Combustion
Therm ody nami c s
Emission
Flue Gases
Pollution Control
Stationary Sources
Trace Elements
Pressurized Fluidized-
bed Combustion (PFBC)
13B 07A ,13H
21D
21B
20 M
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 <9-73)
------- |