United States
                   Environmental Protection
                   Agency
Industrial Environmental
Research Laboratory
Research Triangle Park NC 27711
                    Research and Development
EPA-600/S2-84-064  May 1984
f/EPA          Project Summary
                   Laboratory  Feasibility  Studies for
                   the  Fluidized-Bed  Combustion  of
                   Spent  Pot I'm ing  from  Aluminum
                   Reduction
                    B.C. Kim, E.J. Mezey, D.R. Hopper, A. Wensky, R. Heffelfinger, and S.E.
                    Rogers
                     This study was undertaken with the
                    objective of providing a preliminary
                    assessment of the technical feasibility
                    and environmental acceptability of a
                    fluidized-bed combustion (FBC) process
                    for the disposal of spent potlining waste
                    from the aluminum reduction process. In
                    accomplishing the above, technical
                    efforts were directed to two main areas
                    of laboratory  studies: (1) differential
                    thermal analysis experiments to establish
                    the operating temperature range to
                    prevent agglomeration in a FBC process,
                    and (2) fixed-bed combustion experi-
                    ments to determine cyanide destruction
                    and gaseous emissions expected from a
                    FBC process.
                     The results from the  differential
                    thermal analysis and the fixed-bed
                    combustion experiments indicated that
                    FBC should be limited to temperatures
                    below 760°C due to agglomeration of
                    potlining waste that occurs above that
                    temperature. Analysis of the combus-
                    tion residues showed that 99.9 to
                    99.99 percent of cyanide was removed
                    from the potlining and that substantial
                    burning of the carbon in the potlining
                    could be achieved at temperatures as
                    low as 760°C. The results on cyanide
                    removal and carbon combustion are
                    very encouraging and indicate that
                    potential application of a FBC process
                    to the disposal of spent potlining
                    wastes is technically feasible.
                     A small fraction (0.01-2 percent) of
                    CM and a substantial fraction (15-35
                    percent) of F were found in the gaseous
                    effluent from the combustion of potlin-
ing. The gaseous emissions, however,
can  be adequately controlled and
should not pose an insurmountable
technical barrier in the FBC process. No
significant emissions of SO2 or Na2O
were found during combustion  of
potlining waste.
  Results  of the laboratory studies
indicate that further investigation of the
FBC process, through a pilot-plant
scale test program (particularly efforts
to prove the operability of a fluidized
bed over a range of combustion tempera-
• tures to accomplish effective destruc-
tion of cyanide in the  spent potlining
waste),  would provide valuable infor-
mation in this technological area.  "

  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
  Approximately 230,000 tons of spent
potlining is generated each year in the
production of aluminum metal in the U.S.
This discarded carbon potlining material
contains significant quantities of alumi-
num,  and  sodium metals, aluminum
oxides, fluorides of sodium and aluminum.
carbides and nitrides of aluminum, and
quantities of cyanide which have pene-
trated the potliners over their 3 to 5 year
useful life.

-------
  Because of the significant energy,
chemical, and environmental implications
of the generation and storage of spent
potlining,  efforts by  the aluminum
industry have been directed toward
developing environmentally sound re-
source recovery alternatives. In December
1981, The  Environmental Committee of
the Aluminum Association, Inc. sponsored
a workshop on the storage, disposal, and
recovery of spent potlining.
  Although in July 1980, spent potlining
was  listed by EPA as a hazardous waste
under RCRA, its  listing  is  currently
suspended because  of relevant court
challenges. However, EPA's interest and
responsibility in working with concerned
groups in addressing the potlining dispo-
sal problem remains active.
  The scope of this  program was to
conduct laboratory studies to evaluate
the thermal destruction of cyanide in the
potlining waste and  potential gaseous
emissions from FBC  of spent potlining.
The objective was to assess the technical
feasibility and environmental acceptabil-
ity of the FBC process and provide a basis
for recommending and evaluating test
plans for  follow-up pilot-scale FBC of
spent potlining.

Conclusions
  The study results have identified
agglomeration as a potential problem for
a FBC process due to softening or melting
of the  potlining  from the aluminum
reduction process during the combustion
process. Although the combustion residue
(ash) from the potlining could be main-
tained free flowing without agglomeration
at temperatures above 900°C using  a
kaolin clay as an additive,  the potlining
itself  was observed to agglomerate at
temperatures as low as 760°C even with
the clay additive. Whether  the turbulent
mixing obtained in the FBC process could
overcome the agglomeration problem
appears to be the key technical issue that
must be resolved to prove the FBC process
for application in spent potlining disposal.
  The results from fixed-bed combustion
experiments  have also indicated that
99.9  to  99.99 percent of the cyanide in
the potlining  can be removed at 760°C.
Small quantities of HCN were found in
the gaseous combustion products from
fixed-bed combustion tests performed on
potlining samples. It is expected that the
HCN generated during combustion of
potlining waste can be  destroyed by
providing sufficient gas-phase residence
time at combustion temperature in a FBC.
No detectable emission of SOa was found
in the  flue  gas  from combustion of
potlining.
  A «ignificafrt evolution of HF has been
.found  from •combustion of potljning.
Therefore, the flue gas from potlining
waste combustion will need to be treated
to control  HF emission and  possibly
reclaim the fluoride value.
  Results from  fixed-bed combustion
experiments have indicated that the
carbon in potlining waste can be ignited
and burned efficiently at temperatures as
low as 760°C. The reactivity of the carbon
in spent  potlining  waste appears  to be
sufficiently high to obtain a high combus-
tion efficiency  (>90%) and possibly
sustain combustion without an auxiliary
fuel in a  FBC.
  No  significant volatilization  of  Na20
from possible reactions between potlining
and CaO or water vapor was detected at
760°C. Therefore, generation  of  Na20
from combustion of potlining and  asso-
ciated slagging and fouling problems are
not expected in the FBC process.

Recommendations
  Based on  the laboratory feasibility
studies, a FBC test program should be of
value. A  major technical  issue that could
be addressed in such a program would be
demonstration of the operation of a
fluidized  bed over a practical  range of
temperature to effectively destroy cyanide
without  bed  agglomeration.  The target
temperature range would be at least 50°C
(preferably  100°C) to allow  for normal
temperature fluctuations and process
upsets that can be anticipated  in a full-
scale FBC  unit.  The  clinkers formed by
molten potliner waste are extremely
hard, and formation of such clinkers in a
fluidized  bed by temperature excursions
would cause  a major  shutdown and
possibly extensive  damage to the refrac-
tory in a FBC. If bed agglomeration causes
a severe problem in fluidized bed opera-
tion, other combustion technologies (e.g.,
a rotary  kiln incinerator) could be con-
sidered as an alternative.
  Other  technical  issues that  could be
addressed in a fluidized-bed test program
are:
  1. Multiple sampling of the feed and
     ash during the test program, due to
     the  inhomogeneous nature of pot-
     line
     waste.
  2. The use of kaolin clay as an additive
     to increase combustion temperature
     to the suggested  600 to  800°C
     combustion temperature range.
  3. Flue gas analysis for HCN and HF
     emissions to further establish the
     need for downstream treatment of
     the flue gas. (The level of HF  in the
     flue gas has an important bearing on
     possible recovery ,of fluoride «alue
     and waste heat.)
  4.  Confirmation of cyanide destruction
     by FBC  by leaching  tests on .the
     combustion residue.

Results
  Laboratory studies were conducted in
two parts: (1) differential thermal analyses
(DTAs) to establish the operating tempera-
ture  for the FBC process dictated by ash
agglomeration, and (2) fixed-bed combus-
tion  experiments  to determine  cyanide
destruction and gaseous emissions at the
operating temperature established in the
experiments  of  differential  thermal
analysis.

Potlining Samples
  Potlining samples  were supplied by
Aluminum Company of America (Alcoa)
and Reynolds MetaJs Company Chemical
analyses of  the potlining samples are
shown in Table 1. Both samples exhibited
a strong odor of ammonia, probably
produced by  hydrolysis reaction between
moisture in  ambient air and aluminum
nitride present  The samples  contained
some coarse (about 1/4-m or 0.635 cm)
particles and appeared nonhomogeneous.
Therefore, the samples were dumped
from  the shipping containers, mixed,
coned, and split into smaller aliquots by
sequential quartering and coning or by
using a sample splitter. The head samples
prepared in this manner were the 1/32
cut of the Alcoa sample and the  1/128
cut  of the Reynolds  sample.  The head
samples  were pulverized and stored in
sealed containers  for use in the DTA
studies.

Table 1.   Composition of Potlining Samples"
                  Composition. Percent

c
Al
Si
Fe
Na
F
CN
Al Carbide
Ash
Alcoa
28.6
142
4.76
3.23
139
18.2
009
046
68.7
Reynolds
42.7
47
0.11
0.27
20.0
17.6
0135

571
^Analyses provided by Alcoa and Reynolds,
 except ash analyses performed by Battelle
Differential Thermal Analyses
  DTAs were carried out on two potlining
samples and the ash samples prepared
from the potlining samples. Ash samples
were prepared by burning off the combus-
tible matter from the potlining samples in
a muffle furnace in ambient air at 750°C

-------
to constant residue weight. DT&s were
.also carried out <®n mixtures of the
potlining ash samples with  a reagent-
grade fcaikewm 3
TiO2
CaO
MgO
Na2O
Loss on Ignition at 1000°C
469
38.2
0.35
1 42
0.43
0.58
004
13.9
  Results from the ash agglomeration
tests are presented  in  Table 3  and
summarized below:
  • Alcoa  potlining ash  has higher
     agglomeration temperature than
     Reynolds potlining ash.
  • Adding CaO has little or no adverse
     effect on raising ash agglomeration
     temperature.
  • Adding clay significantly increases
     the ash agglomeration  temperature
     to above 900°C.

Fixed-Bed Combustion Studies
  Fixed-bed combustion experiments
were performed to determine cyanide
destruction and emissions of cyanide and
fluwr*de in  ithve gaseous combustion
products. The reactor was a quartz tube,
2 in. IDt>y44 in. long(6.08 by 111.76 cm),
heated by a3 that is formed by hydrolysis
of AIN in potlining. For this hypothesis,
agglomeration  could  be avoided  in a
fluidized  bed,  since  dehydration  of
AI(OH)3 is commercially practiced  in
fluidized-bed calciners.  However, the
probability  of  the AI(OH>3 sintering
process' being the sole cause of agglom-
eration is remote in view of the complex
salts included in the potlining composi-
tions. Therefore, agglomeration of potlin-
ing poses a potential problem m a FBC
process. Results of the fixed-bed combus-
tion tests on bed agglomeration  are,  in
summary:
  1. The Alcoa potlining sample showed
    greater tendency for agglomeration
    than the Reynolds potlining sample.
    (Compare Tests 5 and .9, or Tests 7
    and 8.) This behavior of the potlining
    samples is opposite from that
    observed with the ash samples  in
    the DTA tests which showed higher
    agglomeration temperatures for the
    Alcoa  sample compared with the
    Reynolds sample.
  2. Reducing the sample size decreased
    the apparent extent of agglomeration
    of potlining. (Compare Tests 4 and
    7.) However, even a smaller quantity
    of test samples showed agglomera-
    tion at 760°C (Tests 8 and 9).
Potlining
Test
1



2
3
4
5
6
7
8
9
10
;;"
Type
Alcoa



Alcoa
Reynolds
Reynolds
Reynolds
Alcoa
Reynolds
Alcoa
Alcoa
Reynolds
Reynolds
Sample Wt.. g
75



75
75
75
12
12
12
12
12
12
12
Wt.,g
7.5



7.5
7.5
37.5
6
4.1
6
7.2
7.2
6+6(CaO)
6
% of
Ash"
15



15
17
87
87
50
87
87
87
87+87(CaOj
87
Temperature,
°C
911



762
760
762
760
875
760
760
76O
760
760
Residence
Time, mm
30



120
120
30
120
30
30
30
120
120
120
Results
Very severe agglomeration:
Quartz reactor cracked.
apparently caused by
fused potlining
Very severe agglomeration
Very severe agglomeration
Very severe agglomeration
Slight agglomeration
Very severe agglomeration
Slight agglomeration
Severe agglomeration
Severe agglomeration
Slight agglomeration
Slight agglomeration
"Based on 68.7 percent ash in the Alcoa potlining and 57.1 percent ash in the Reynolds potlining.
''Combustion air was humidified to approximately 2 percent water vapor concentration.

-------
    Due to the severe agglomeration of the
  bed  that  resulted in poor gas/solid
  contact, the ash  residue and  impinger
  train samples that were selected for
  analysis were chosen from those tests
  performed during the latter part of the
  combustion test  program in which bed
  agglomeration was substantially reduced
  by decreasing the combustion  tempera-
  ture
  and the amount of potlining sample used.
  Material  balance results describing the
  fate of cyanide, fluorine, carbon, and
  sodium during the fixed-bed combustion
  of potlining were calculated from these
  sample analysis results.
    Tests 5, 7, 8, and 9 were performed to
  determine cyanide destruction, emissions
  of cyanide and fluoride in flue gas, and
  carbon combustion efficiency. Results
  are:
    1. Cyanide was removed  from the
       potlining  in  the  range of 99.9  to
       99.99 percent with both Alcoa and
       Reynolds potlining samples at 760°C
       and 30  to 120 minutes residence
       time. Small quantities of cyanide in
       the range of 0.01 to 2 percent of feed
       were found in  the flue  gas. It is
       expected that a FBC should provide
       sufficient gas-phase residence time
       in the bed and in the freeboard to
       destroy the cyanide evolved as HCN
       from the potlining.
    2. Approximately 20 to 30 percent loss
       of fluoride from potlining  was
       observed  under the conditions
       employed in the combustion tests.
       Fluoride collection from the flue gas
       was measured to be in the range of
       0.2 to 8.7 percent. The fluoride
       unaccounted for was probably lost
       by reaction with the quartz reactor
       and other glass surfaces  that con-
       tacted the flue gas.  Nevertheless,
           the results strongly suggest that a
           significant fraction of fluoride is
           generated  as HF  from  potlining
           during the combustion  process.
           Therefore, flue gas should be treated
           to control HF emission and possibly
           recover the fluoride value.
        3. Results  on carbon  balance show
           approximately 70 to 90 plus percent
           carbon combustion  efficiencies in
           the combustion tests.  The  total
           carbon found in the ash and the flue
           gas was found to be approximately
           25 to 35 percent above the carbon
           introduced with the potlining, indi-
           cating a reasonable closure on the
           carbon balance. Combustion effi-
           ciency in a fluidized bed is expected
           to be higher due to improved mixing
           and gas/solid contact in a fluidized
           bed. Therefore, combustion of pot-
           lining in a fluidized bed at tempera-
           tures as low as 760°C appears to be
           feasible.
        4. No detectable emission of SOz was
           found in the flue gas from combus-
           tion of potlining.
  Tests  10 and  11  were  performed to
determine volatilization of Na20  by
reactions of a potlining sample with CaO
and  with H20  vapor. The tests were
carried out with the Reynolds potlining
sample that has a higher Na content than
the Alcoa potlining sample. Test 10 was
carried out with CaO addition and dry
combustion air. Test 11  was carried out
with the combustion air humidified to
approximately 2 percent  water vapor.
Results on sodium balance show essen-
tially no volatilization of  Na2O due either
to CaO or HaO vapor under the conditions
employed in the combustion tests.
         B. Kim, E. Mezey, D. Hopper, A. Wensky. Ft. Heffelfinger, and S. Rogers are with
          Battelle-Columbus Laboratories, Columbus, OH 43201.
         David C. Sanchez is the EPA Project Officer (see below).
         The complete report, entitled "Laboratory Feasibility Studies for the Fluidized-Bed
          Combustion of Spent Potlining from Aluminum  Reduction," (Order No.  PB
          84-168 764; Cost: $10.00. 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
                                                                              •fr U S GOVERNMENT PRINTING OFFICE, 1984 — 759-015/7685
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300

-------