United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-8 3-009 Apr. 1983
Project Summary
Disposal of Spent Sorbent from
Dry FGD Processes
P. M. Stephen, H. S. Rosenberg, and R. B. Bennett
The main by-product of a dry injec-
tion flue gas desulfurization (FGD) pro-
cess for a coal burning power plant
consists of a mixture of reacted sorbent
and fly ash. When sodium-containing
sorbents such as nahcolite (NaHCOa)
and trona (NaHCO3-Na2CO3-2H2O) are
used to capture the SOa in the flue
gases, the spent sorbent consists pri-
marily of sodium sulfate (NazSQt) and
sodium sulfite (NaaSOs) which are
highly water soluble and generally un-
suitable for disposal in a simple landfill.
A research program was undertaken to
study methods of stabilizing the fly
ash/spent sorbent mixtures through
agglomeration and sintering. A con-
ceptual process flow diagram and pro-
cess economics were determined for a
facility which would process and some-
what stabilize fly ash/spent sorbent
mixtures from a dry FGD process for a
500 MW power plant burning low sul-
fur western coal. While the stabili-
zation process tested successfully re-
duced leaching of the FGD waste/ash
mixture, its projected costs (about
$41/kW and 5 mills/kWh) appear to
make this process economically un-
attractive for a single 500 MW plant at
this time. Thus, the stabilized waste
would cost nearly $88/tonne of waste
disposed of, about five times the cost
for conventional FGD waste.
This Project Summary was developed
by EPA's Industrial Environmental Re-
search Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
Recent trends in S02 emission controls
have included the use of fabric filters or
electrostatic precipitators (ESPs) for par-
ticulate control, and wet lime (or lime-
stone) scrubbing for control of S02 from
the boiler flue gas. Alternative FGD pro-
cesses in various stages of research,
development or commercialization cur-
rently include spray drying and dry injec-
tion of sorbents into the boiler flue gas
stream. Dry injection of sorbents is desir-
able in arid regions where water is scarce
and western coal, with lower sulfur con-
tent than eastern coal, is available.
The Buell Emission Control Division of
Envirotech Corporation* with support of
the U.S. EPA has tested a dry FGD process
at the Martin Drake Power Plant in Colo-
rado Springs, CO. This process involves
injecting nahcolite (NaHCOs) or trona
(NaHCO3-Na2C03-2H2O) into a flue gas
slipstream from the plant's No. 6 boiler
and collecting the resulting fly/ash spent
sorbent mixtures in a pilot baghouseorthe
baghouse at the facility.
The primary problem associated with
the dry injection FGD process using sodium
compounds is disposal of the spent sor-
bent This reacted sorbent containing
sodium sulfate and sodium sulfite formed
in the reaction with the S02 in the flue
gases, is highly water soluble and has a
high potential for leaching into gound
water if disposal is via unlined landfill.
Objectives of the research program con-
ducted at Battelle/Columbus Laboratories
(BCL) were: (1) to study the sintering
and leaching mechanisms of fly ash/spent
sodium sorbent mixtures from a dry injec-
tion FGD process, and (2) to determine
the process economics for a pelletizing
and sintering facility which would handle
the output of fly ash and spent sorbent
from a full-scale power plant
•Now General Electric Environmental Services, Inc.
-------
A previous study at Battelle, sponsored
by Industrial Resources, Inc. (IRI), ad-
dressed the problem of disposal of spent
nahcolite from dry FGD processes. The
laboratory program included glass melting
studies and sintering experiments, which
eventually led to development of the
patented Sinterna process for consoli-
dating and sintering material by-products
containing spent nahcolite sorbent.
The Program
The laboratory research program in-
cludes pelletizing experiments, sintering
experiments, use of pellets as aggregates,
and an economic analysis of the process.
Palletizing Experiments
The agglomeration technique used for
the spent sorbent/fly ash stabilizing pro-
cess was disc pelletizing, due to its wide-
spread use for agglomerating a large
volume of similar materials. Objectives of
the pelletizing technique were: (1) to
determine the pelletizer operating param-
eters required to produce pellets from the
spent sorbent/fly ash mixtures in a satis-
factory size range and with sufficient
handling strength, and (2) to determine
the pelletizing moisture and binder require-
ments for the feed material. Disc pelletizing
followed by sintering of the fly ash pellets,
used for many years for lightweight aggre-
gate production, is mentioned in two
recent patents as the preferred agglomer-
ation method for fly ash waste disposal.
Disc (or pan) pelletizing involves feeding
a powdered material onto a rotating disc
and spraying water or another liquid onto
the powder to start agglomerating the
individual powder particles. The liquid
acts as a binder while the powder is rolling
and tumbling as it is palletized into spheres
of a narrow size distribution determined
by the operating parameters and raw ma-
terial variables. The operating parameters
include pan angle and depth, rotational
speed, water spray and feed material rates
and location, and scraper location. Raw
material variables include particle shape
and size distribution, compositional varia-
bility, binder type and amount and particle
surface conditions.
The initial pelletizing experiments used
small batches (about 2 kg) of the fly
ash/spent sorbent mixtures with and
without additional binders added for green
strength. These batches were pelletized in
a small pelletizer, 36 cm in diameter and
16 cm deep. Pellets, 5-15 mm in diameter,
were tested for strength using a hand-held
spring compression device after drying or
firing. The force required to break the
dried or fired pellets was measured and
recorded. The binders selected for com-
parison of their green strength contribu-
tion included calcium lignosulfonate,
bentonite, hydrated lime, and limestone.
Binders were added in amounts of 2-5
percent of the original batch weight
Pilot-scale production of about 100 kg
of pellets was conducted using two larger
pelletizing units with different rim-height-
to-diameter ratios. A deep drum pelletizer
(Mars Mineral Corporation Series20) with
a rear feed screw, a diameter of 61 cm, and
a rim depth of 45 cm was used initially in
attempts to pelletize the fly ash/trona
sorbent powder received from Envirotech
from a pilot test run at the Martin Drake
Power Plant in Colorado Springs. More of
the trona-containing mixtures and the fly
ash/nahcolite mixtures were later pellet-
ized in a shallow disc pelletizer, 106 cm in
diameter and 22 cm deep. The pellets
from the later trials were dried at 1 50°C,
fired at 925°C, and used as aggregate in
the concrete and asphalt mixture testing.
Table 1 shows analyses of pure fly ash
(average of several samples), the fly ash/
spent nahcolite from Envirotech test 12 R1,
and the fly ash/spent trona from Envirotech
test 22 R1 at its pilot facility at the Martin
Drake Power Plant The nahcolite was
obtained from the U.S. Bureau of Mines
from a deposit in Rio Blanco County, CO;
the trona was from Owens Lake in Cali-
fornia The weight ratio of fly ash to spent
sorbent was roughly 70/30 in the mix-
tures received from Envirotech.
Several studies have been completed on
the characterization of fly ashes from coal
burning power plants. TVA reports that fly
ash consists primarily of the oxides of
silicon, aluminum, calcium, iron, and other
minor elements. Petrographic analyses
indicate that fly ashes are generally 50-90
percent glass by weight; the remaining
material is primarily quartz, mullite, hema-
tite, magnetite, and carbon. The glassy
phase contains a considerable amount of
material known as cenospheres, which are
rounded particles containing nitrogen and
C02, comprise 5-20 percent by volume of
the total fly ash, and are 20-200 ju.m in
diameter.
The particle diameters of the fly ash/
spent nahcolite received from Envirotech
were 1 -60 /*m, with a median diameter of
15 mm, measured by sedimentation tech-
niques. The surface area of the powder,
measured in a Micrometrics Model 2205
analyzer, was 4.5 mVg, which would be
higher than that calculated from the par-
ticle size distribution, but which may be
due to porosity of the nahcolite-containing
particles.
To increase pellet strength, binders
used included calcium hydroxide at 2 and
5 percent by weight calcium lignosulfonate
at 2 and 4 percent by weight and bentonite
at 4 percent by weight A high calcium
limestone was added to some batches
because of its potential as a binder at
higher sintering temperatures (as a glass
network modifier and flux). These binders
Table 1. Sample Chemical Analyses3
Description
Fly Ash from
Colorado Springs
Plant - Averages
Fly Ash/
Spent Nahcolite -
Test 12R1
Weight Percent
SiOz FezOa -4/2Q3 CaO MgO 80s KzO NazO PzOs TiOz UzO LOIaooc Total
66.80 4.00 17.50 4.40 1.50 0.26 1.07 1.50 0.80 0.80 0.02 - 98.65
51.22 3.23 22.83 3.60 1.49 2.90b 1.21 3.42>> 1.31 0.78 0.02 7.55 99.56
Fly Ash/
Spent Trona -
Test22R1
52.14 5.23 21.08 3.63 1.21 2.78C 1.08 4.93* 1.26 0.86 0.02 5.33 99.55
"Analyses provided by General Electric Environmental Services, Inc.
>>BCL Analyses SO4 = 3.59, SOa = 0.16, NazO = 3.05.
cBCL Analyses SO* = 7.0ft SOj =
-------
were tested for strength separately and in
combination with other binders in the
batches.
Loads required to break the dried pellets
were 5-55 newtons*; 15 newtons was
the cutoff point below which pellets ex-
hibited unacceptable breakage while rolling
and tumbling through the various pro-
cessing steps. Low dried pellet strength is
also related to low green pellet strength
(as discharged from the pelletizer), as
measured by dropping pellets from vari-
ous heights immediately after palletizing
to determine their initial resistance to
breakage. The pellets with low dried
strength (< 15 newtons) would normally
break after one or two drops from a height
of 20-45 cm onto a hard flat surface.
Pellets with higher dried strengths usually
survived multiple drops from heights up
to 40 cm.
The dried fly ash/spent nahcolite pel-
lets broke readily with moderate handling
and exhibited considerable attrition loss
during handling in the sample container.
Most of the binders increased pellet
strength at some level of addition to the
basic starting mixtures; the spent trona
batches had higher comparative dried pel-
let strength than those containing spent
nahcolite. Increases in fired pellet strength
indicated that some sintering and forma-
tion of glassy bonds was taking place
during thermal treatment of the pellets. A
5 percent addition of calcium hydroxide
[Ca(OHJ2l was used in the larger scale
palletizing tests due to its contribution to
pellet strength, comparative cost and avail-
ability, and for its possible role in the
fixation of the sulfates in the mixtures. The
binder also significantly reduced dusting
of the pellets during handling.
Initial attempts to pelletize larger quanti-
ties of the fly ash/spent trona powder
were made using the Mars Series 20
pelletizer. The pan angle was set at 73°,
and the pan was operated with a rim speed
of 40 to 50 m/min. The powder feed rate
was 30 - 50 kg/hr and water flow rate was
7-10 l/hr. A stable condition could not be
reached in this unit since the weight of the
powder tumbling in the drum crushed the
seed pellets.
The next attempt at palletizing the fly
ash/spent sorbent powders involved the use
of the 1-m diameter disc pelletizer. The
angle was set at 60°, and the rim speed
was set at 70-80 m/min. The powder
feed rate was 40-50 kg/hr, while the
corresponding water addition was 13-16
l/hr. Pellets were formed readily in this
•1 newton = 0.225 pound-force.
pelletizer, ranging from 1 to 2 cm in
diameter. Nearly 100 kg of pellets were
made in this unit, dried at 150°C, and
then sent for testing as aggregate in
concrete and asphalt mixtures.
Sintering Experiments
Temperature and heating rate most sig-
nificantly influence the amount of S02
evolved from the spent sorbent-containing
pellets during thermal processing. The
development of a glassy microstructure
and/or sintering of the particles in the
pellets during thermal processing influ-
ence the solubility of the pellets in a liquid
S02 evolution during heating of the pellets
was measured up to 1200°C and leaching
of sodium and sulfates from pellets in
water was measured after the pellets had
been heated. Most of the tests were
conducted on pellets containing spent
nahcolite and 5 percent lime; these were
considered most representative of the
material that would be processed in a full-
scale operating facility using the dry injec-
tion FGD process.
Figure 1 consists of micrographs of
mounted and polished sections of fly ash/
spent trona pellets made on a disc pellet-
izer and either dried at 150°C or fired for 1
hour at 1000°C in an electrically heated
furnace. The pellets dried at 150°C have
distinct particle and matrix phases. In the
pellets fired at 1000°C, most of the glassy
particles have melted to form a glassy
matrix with crystalline inclusions. The
changes in crystalline content of the pel-
lets with thermal treatment were studied
by x-ray diffraction (XRD) analysis of pow-
der from crushed pellets.
The XRD analysis results for fly ash/
spent sorbent mixtures and for pellets of
these compositions fired at 1000°C are
shown in Table 2. Included in the table are
the crystalline compounds found in the
patterns, their powder diffraction file num-
bers, and normalized concentrations. The
normalized concentrations show relative
amounts of the various crystalline com-
pounds. Best estimates for actual amounts
(not shown in the table) would be based on
factors for the ratio of crystalline to non-
crystalline material in the samples. Petro-
graphic examination of the powders indi-
cates that roughly half of the material in
both as-received and fired pellets is crys-
talline, indicating no substantial glass
formation with thermal treatment up to
1000X although there has been sub-
stantial melting and sintering of the glassy
particles at the higher temperatures.
An interesting difference occurs in the
XRD analysis results for the as-received
powders. All of the sulfur in the pattern for
the fly ash/spent nahcolite powder occurs
as sodium sulfate; both burkeite [NaeSOa
(804)2] and sodium sulfate are in the fly
ash/spent trona sample. As expected,
quartz and mullite are the major crystalline
compounds in the as-received samples.
The complexity of the XRD analysis
patterns for the samples fired at 1000° C
indicates that several new crystalline com-
pounds are formed during thermal treat-
ment of the pellets, while the relative
amounts of quartz and mullite decrease.
The most interesting and unexpected
compound to form is a complex sulfate
[Na6Ca2AI6Si6024(S04)2], which is pres-
ent to some extent in all of the fired
samples but is the major crystalline com-
pound in the Batch A-(trona) and Batch 2a
(nahcolite plus lime) compositions. Two
sodium aluminosilicate compounds (albite
and nepheline) are also present in major
amounts in the fired samples. The crystal-
line compounds in the fired samples are a
combination of small amounts of unre-
acted crystals(primarily quartz and mullite)
and recrystallization products which form
during cooling of the molten constituents.
The compounds or glasses in the therm-
ally treated pellets influence the degree of
solubility or stability of the pellets under
leaching conditions. Processing at high
temperatures (>900°Q evidently converts
most of the soluble alkaline sulfates in the
pellets to less soluble compounds or ab-
sorbs them partially into the glassy matrix.
To determine the optimum time and
processing temperatures to maximize sul-
fur retention and minimize leaching of the
pellets in water, individual pellets of the
various compositions made during the
pelletizing studies were subjected to a
controlled heating cycle during which the
amount of S02 evolved was measured.
These pellets were then leached for vari-
ous times up to 1000 hours in distilled
water with moderate stirring.
Figure 2 represents the percentage of
sodium leached from pellets of Batch 2 a
which had been heated to different tem-
peratures. As expected, the dried pellets
exhibited the highest sodium leaching
rates, up to 60 percent of the available
sodium Pellets heated to 1000°C and
leached upto 100 hours had sodium leach
rates of — 10 percent, indicating the
stabilization of the sodium through in-
corporation into the glassy network
Figure 3 is a composite figure which
includes the sulfates leached at 100 hours
of stirring, S02 evolved after 1 hour of
heating, and the calculated percentage of
sulfur retained in the pellets after being
-------
a. Dried at 150°C I285X)
b. Fired at 1000°C (28SX)
subjected to both tests. The curves for
total sulfur retained show the optimum
temperature for thermal processing of the
pellets. For Batch 2a (nahcolite plus 5
percent lime), the optimum temperature is
near 1000°C with 50 percent sulfur reten-
tion, indicating the influence of the lime
addition which increased the temperature
at which there was significant SO2 evolu-
tion during thermal processing.
Use of Pellets as Aggregates
Pellets containing fly ash, spent nahco-
lite, and 5 wt percent of hydrated lime
were tested as aggregates in standard
Portland cement and asphalt mixes. Pel-
lets fired at 1 50°C or fired at 925°C were
substituted at 20 percent by volume for
similarly sized limestone in the mixes.
Both cement and asphalt mixes exhibited
unacceptable leaching rates of sulfates
and/or sodium. Also, in the comparative
strength tests under ASTM procedures,
samples made from the cement and as-
phalt mixes had lower than acceptable
strength levels after curing.
Economic Analysis
The conceptual process flow diagram
for agglomerating and sintering fly ash/
spent sorbent mixtures at a power plant is
shown in Figure 4. For a 500 MW power
plant burning 227 tonnesVhour of low
sulfur coal with 9.7 percent ash, the fly ash
collected would be 17.6 tonnes/hour,
with 7.3 tonnes/hourof nahcolite injected
and 1.3 tonnes/hour of hydrated lime
added as a pelletizing binder. Nearly 20
percent by weight of water is needed for
pelletizing.
Process equipment included two 5-m
diameter disc pelletizers, a conditioning
unit for lime addition and de-dusting, and a
traveling grate furnace, roughly 3 m wide
and 54,5 m long (based on a processing
rate of 1.5 m3 m-2 hr-1).
The total process capital for this facility
was estimated at $7,102,000 (1 982 $}.
The addition of other direct costs and fac-
tored capital charges results in a total capi-
tal investment estimate of $20,277,000
($40.55/kW of generation capacity) for
the 500 kW plant
First year revenue requirements for the
facility were calculated as $8,567,845
(1984 $), or 3.12 mills/kWh. Levelized
capital charges for the plant assuming a
30-year life, were calculated as $ 13,518,038,
or 4.92 mills/kWh. Based on waste
disposal of 154,000 tonnes/year of ash
plus spent sorbent the levelized annual
Figure 1. Batch A pellet microstructures (fly ash and trona).
4
*1 tonne= 1000kg= 1.1025 tons
-------
Table 2. X-Ray Diffraction Analyses
Sample
Fly Ash +
Nahcolite
Fly Ash +
Nahcolite
Batch 1
Fired 1000° C
Trona +
Fly Ash
Batch A
Raw Powder
Trona +
Fly Ash
Batch A
Fired WOO'C
Nahcolite +
5% Ca(OH)2+
Fly Ash
Batch 2a
Fired 1000° C
for Fly Ash Spent Sorbent Mixtures
Compound
e-Quartz (Site)
Mullite (AI6Si2Oi3)
NB2SO4 (Form III)
Natron (Na2C03- 10H2O)
b Unknown
Magnetite (FesO*)
Albite (NaAISisOsJ
e-Quartz (SiOz)
Mullite (AI6Si2Oi3)
Nepheline (NaAISiO4>
Na6Ca2A/6Si6O24(SO4)2
Magnetite {FesO^
NazSO4 (Form III)
Unknown
e-Quartz (SiO2)
Mullite (AleSi2Oi3)
Burkeite (NaeCOsf 804)2)
N32SO4 (Form III)
Unknown
Magnetite (FeaO4)
NaeCa2Al6SieO24(SO4)2
Nepheline (NaAISiO4)
e-Quartz (SiO2)
Albite (NaAISisOa)
Mullite (A/6S/2O13J
Unknown
Magnetite (Fe3O4)
Na6Ca2Al6Si'eO24(S04J2
Nepheline (NaAISi04)
e-Quartz (SiO2)
Unknown
Mullite (A/6Si2Oi3)
Albite (NaAISisOs)
Na2SO4 (Form III)
Magnetite (FesO4)
Anhydrite (CaS04)
PDFNo.0
5-490
15-776
24-1132
15-800
-
19-629
20-554C*
5-490
15-776
19-1176
25-802
19-629
24-1132
-
5-490
15-776
24-1134
24-1132
.
19-629
25-802
19-1176
5-490
20-554C
15-776
.
19-629
25-802
19-1176
5-490
.
15-776
20-554C
24- 1 132
19-629
6-226
Normalized
Concentration
fWt Percent)
45
40
6
5
2
1
30
28
22
7
6
2
2
2
35
35
18
7
3
1
45
18
14
8
7
7
1
38
25
12
12
6
3
2
1
0.4
*PDF = Powder Diffraction File.
b Unknown = intensity of the strongest unidentified line in the pattern
CC= a calculated pattern, giving a better fit than one produced from a natural mineral specimen.
cost is nearly $88/tonne of waste, or
$11/tonne of coal burned.
Note that SOa evolved during sintering
of the pellets would be captured via recir-
culation of gases from the furnace to the
power plant FGD system. The costs, both
capital and operating, of removing this
S02 are not included in these estimates.
Conclusion
Technically, results of the program are
mixed in terms of feasibility or desirability
of using a Sinterna-type process for dis-
posal of fly ash/spent sodium compound
mixtures from dry FGD systems. TheS02
evolved during the processing (heating)
must be balanced against the sodium
sulfate leached from the processed waste,
and the evolved S02 must be accounted
for in controlling total allowable S02 emis-
sions from the plant Economically, the
process appears to be too costly for dis-
posal of the FGD by-products from a single
500 MW power plant The disposal cost
is nearly $88/tonne of waste, about five
times greater than for conventional FGD
waste.
Other options exist for disposing of
waste solids from dry injection of sodium
compounds into flue gas for S02 control.
These options include disposal in lined
pits with controlled runoff, or processing
the material removed from the baghouse
by methods similar to a double alkali
process, where the spent sorbent would
be dissolved in water, and then reacted
with lime and converted to gypsum. Feas-
ibility and cost estimates for these alterna-
tives might be gained through the analysis
of similar existing systems, but are beyond
the scope of this study.
Acknowledgments
The U.S. Bureau of Mines supplied the
nahcolite for the pilot plant experiments
from their research tract near Grand
Junction, CO.
The City of Colorado Springs, CO, through
Ronald L Ostop in the Department of
Public Utilities, was most helpful in provid-
ing the power plant facilities required to
conduct the pilot plant dry injection
experiments.
-------
100
40 60 80 100
Leaching Time, hours
120
Figure 2. Percent Na leached from Batch 2a after heating at various temperatures.
(Batch 2a contains fly ash and nahcolite product solids with 5 wt. percent
Ca(OH)2 added.)
100
• SO4 leached after 100 hours
O SO2 evolved after 1 hour
O Sulfur retained
40 -
20 -
105 700
800 900
Temperature, °C
1000
Figure 3. Sulfur retention in Batch 2a after heating and leaching tests. (Batch 2a contains fly
ash and nahcolite product solids with 5 wt. percent CafOHh added.)
-------
Fly Ash •
Boiler
Bag
House
Nahcolite
Injection
Storage
Bin
Storage
Bin
v V
Blending
Storage
Bin
Figure 4. Process flow diagram.
P. M. Stephen, H. S. Rosenberg, and R. B. Bennett are with Battelle/Columbus
Laboratories, Columbus, OH 43201.
Theodore G. Brna is the EPA Project Officer (see below).
The complete report, entitled "Disposal of Spent Sorbent from Dry FGD
Processes," (Order No. PB83-165 266; Cost: $11.50, subject to change) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
U. S. GOVERNMENT PRINTING OFFICE: 1983/6S9-095/1927
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
PS 0000329
U S ENVIR PROTECTION AGENCY
REGION 5 LIBRARY
250 S DEARBORN STREET
CHICAGO IL 60604
------- |