.V
United States
Environmental Protection
Agency
Hazardous Waste Engineering
Research Laboratory
Cincinnati, OH 45268
Research and Development
EPA/600/S2-85/012 May 1985
Project  Summary
Full-Scale  Carbon  Adsorption
Applications  Study

T. P. Nelson, J. R. Blacksmith, and J. L. Randall
  The full report  presents the results
of theoretical and field test investiga-
tions of full-scale carbon adsorption
applications. The  program focused on
the  performance testing  of several
fixed-bed, steam-regenerated carbon
adsorption systems used  in the rub-
berized  fabric, magnetic tape, and
flexible  packaging industries. The test
results  showed  that  a  96  percent
volatile   organic  compound  (VOC)
reduction can be expected for most
common industrial solvents when ex-
haust gases are processed through a
properly designed and  operated  ac-
tivated  carbon bed.  The test program
was also designed to measure the ef-
fective carbon life for each of the test
facilities.  Carbon life was found to
vary from plant to  plant and ranged
from as  low  as  9   months  at a
magnetic tape facility to over 5 years
at a rubberized fabric coating opera-
tion. The full report also includes a
brief discussion on the theory of car-
bon adsorption  and  on  computer
simulations of carbon  bed systems.
  This   Project  Summary  was
developed by EPA's Hazardous Waste
Engineering Research Laboratory, Cin-
cinnati,  OH, to announce key findings
of  the  research project that is fully
documented in a separate report of
the  same  title  (see  Project Report
ordering information at back).

Introduction
  The   U.S.  Environmental  Protection
Agency  (EPA) is currently evaluating the
environmental   and  health  impacts  of
volatile organic compound (VOC) emis-
sions from a variety of industrial sources.
As  part of this effort,  Radian Corporation
has contracted a program (EPA Contract
 No. 68-03-3038) to evaluate the perform-
 ance of full-scale vapor-phase carbon ad-
 sorption VOC control systems for solvent
 recovery.  Inherent  in  any  cost-
 effectiveness evaluation of a  carbon ad-
 sorber installation is some decision on the
 effective life of the carbon. The Agency
 was  unable to find substantive  data to
 reflect  the  effect  of irreversible fouling
 that  may take  place as a result of im-
 purities or byproducts of some industrial
 operations. This study was initiated to in-
 vestigate the change, if any, in efficiency
 of the carbon bed with time.
  Carbon   adsorption-based  solvent
 recovery is a commonly used VOC control
 technology  in  the  surface  coating  in-
 dustry. Economic and/or regulatory incen-
 tives contribute to its  wide application.
 Limited   data  are  available  on
 measurements of carbon service  life and
 VOC removal efficiencies for full-scale, ex-
 isting  carbon bed systems. Therefore, to
 evaluate the VOC control performance,
 the following  program objectives  were
 established: (1) to  determine  the effects
 of carbon service life on long-term system
 performance and (2) to compare the per-
 formance  of the tested systems and the
 major factors affecting performance.

Summary of Results
  Six full-scale activated carbon systems
were tested at various locations across
the United  States. The test sites included
two plants from each of the following sur-
face coating industries:  rubberized fabric,
magnetic  tape, and  flexible  packaging.
These categories were  selected because
they  represented  a  variety  of  coating
solvents and control  system  sizes. The
coating solvents examined in the study in-
clude  hexane,  toluene, tetrahydrofuran
(THF),  isopropyl  acetate,  n-propyl

-------
acetate, and  methyl ethyl ketone (MEK).
The  volumetric flow  rate  of  the  gas
streams treated ranged  from 3.7  to  32
NmVsec (7800 to  68,000 scfm). The ad-
sorption systems at the  rubberized  fabric
and  flexible   packaging  plants  feature
horizontal  carbon   beds,  although  SLA
flow direction  through the beds  differs
(downward   for  the  rubberized   fabric
plants and upward for the flexible packag-
ing plants). Conversely, the magnetic tape
plants feature adsorption  systems with an-
nular carbon  beds, although again, SLA
flow direction through the bed differs (in-
ward for one plant and  outward for the
other). The number of carbon beds per
system  also   varied from  two  to  six,
depending  on  the  volume of gas  being
treated.
  The relative  SLA flow rates,  as com-
pared to the bed carbon capacity, average
approximately 0.0015 Nm3/sec-kg carbon
(1.5  scfm/lb  carbon) for the six  tested
systems  with the  rates  ranging  from a
relatively conservative  0.00069 Nm3/sec-
kg carbon (0.65 scfm/lb carbon) to a high
of   0.0026  Nm3/sec-kg  carbon  (2.5
scfm/1b carbon). Also, the horizontal bed
systems generally  feature average design
superficial  bed velocities of  0.45  m/sec
(89 fpm), while the thin-bed designs of
the  annular  bed  systems feature  lower
velocities, nominally 0.33 m/sec (65 fpm).
The   six  designs   are  characterized  by
significant  differences   in  inlet  solvent
loadings,  with  average  inlet concentra-
tions ranging from less than 1000 ppmv to
over  4000 ppmv.  On   a carbon  basis,
design solvent loadings vary over an order
of magnitude,  from 0.0066  to  0.097 kg
solvent/kg carbon (0.0066 to 0.097  Ib sol-
vent /Ib carbon).
  Cost data were also developed for each
of the tested facilities  based on plant in-
formation.  In 1983 dollars, the total capital
costs for the  activated  carbon  systems
range from $267,000 to  $2,685,000.  The
capital cost  normalized to the SLA flow
rate  ranges from $51,000 to $173,000  per
Mm3 ($24 to  $82 per scfm).  The normal-
ized  annual costs range from 0.19 to 0.66
$/kg (0.09 to 0.30  $/lb)  of recovered sol-
vent.
  The  carbon  adsorption  systems  were
tested using EPA and ASTM methods.  In-
let and  outlet  VOC concentrations were
measured   semicontinuously  with  flame
ionization  detector  (FID)  total  hydrocar-
bon (THC) instrumentation as described in
EPA  Method  25A.  Volumetric flow rates
were  measured   according  to   EPA
Methods 1 and 2. All  test results were
verified  by  EPA-specified  quality
assurance/quality   control   procedures.
Other   process   information,   including
steam   rate,   steam  temperature,  SLA
temperature,   and  carbon  weight,  was
determined from  process instrumentation
or design specifications.
  All  six  plants  were originally  tested in
early 1982. During these tests, data  were
taken to characterize the VOC removal  ef-
ficiency of the carbon bed systems and to
characterize the conditions of the carbon.
Four of the six plants were retested using
the same test methods approximately  18
to 22 months later. A comparison of the
changes in performance was  made in  an
attempt to define the effects of  long-term
carbon  degradation. In  general, the data
indicate significant differences in perform-
ance  from plant  to plant  and  from test
period to test period. However,  these dif-
ferences were not  always attributable to
carbon  degradation.  A  summary  of the
test results for all six plants is  shown in
Table 1.
  The  average  VOC control  efficiency
data indicate that,  with  the  exception of
one plant, 'all tested adsorption systems
were  capable of  achieving efficiencies of
95  percent or greater. (Mechanical prob-
lems  at two  plants  caused  considerable
decreases in the VOC control performance
during   the   repeat  testing.)  Average
removal efficiencies exceeding 99  percent
were  measured  for two of  the systems
during the original  testing.
  The useful  carbon  life of the activated
carbon systems  examined  in this study
ranged from  a  few months to  over 6
years. During the repeat testing  at four of
the original six test sites, carbon degrada-
tion was definitely detected at only one of
the  facilities. The other  three facilities
either did not  indicate  any  substantial
reductions in VOC  removal  performance
or  were  experiencing reductions  in per-
formance not related to carbon degrada-
tion.   It  was concluded  that the  test
methods currently used for measuring car-
bon bed  performance cannot be used to
predict useful carbon life.

-------
Table 1.     Comparison of Test Results—Plants 1-6



                              Rubberized Fabric
                                                                        Magnetic Tape
                                                                                                                      Flexible Packaging
PLANT 1 5
Original Original
Test Test
Solvent Type MEK toluene
Adsorption mode 95* 150
length, min
SLA flow rate. 5.4 4.3
Nrrf/seclscfm) (11,400) (9,100)
Inlet solvent concen- 2, 190 1,940
tration, ppmv
VOC control 84.9 97.6
efficiency, %
Desorption mode 55 50
length, min
Steam/ recovered sol- 20120) 8.218.2)
vent, kg/kg (Ib/lb)
Carbon age, years 0.4 5.0
Cost Information
(1983 dollars)
Total Capital, tlO> 622 267
Relative Capital. $ 10*/ 110(56) 51(24)
Nrrf/sec (f/scfm)
Total Annualized 262 125
Cost, $10>
Relative Annualized 0.53 (0.24) 0. 19 (0.09)
Cost, S/kg
Recovered solvent
2 3
Repeat Original Repeat Original
Test Test Test Test
toluene THF/to/uene3 THF/toluene* THF/toluane/
MEK/MIBK/
cycfohaxanone
200 64 65 47
3.7 4.6 4.5 9.4*
17,8001 (9.800) (9.500) (19.800)
896 1.470 1,140 3,220
80.5 99.7 ' 95.2 94.7
40 32 32 35
15115) 5.2(5.2) 6.5(6.5) 6.0(6.0)
6.5 2.0 3.8 0.4
650 1,800
108(51) 173182)
325 1,430
0.61(0.28) 0.66(0.30)
4 6
Original Repeat Original Repeat
Test Test Test Test
hexane hexane toluene/ If AC toluene IIP AC
120 90 470* 580"
8.3 10.7 15.8* 16.0^
(17,700) (22.600) (33,400) (33.900)
1,260s 940* S04 847
99.0s 4S.7e-i 97.5 97.9
25 25 30 30
4.6(4.6) 11(11) 1.0(1.01 0.8(0.8)
3.0 4.5 0.2 1.6
1,334 2,685
129 161) 72 (34)
494 839
0.51 (0.23) 0.28 (0. 13)
^Original THF/ toluene percentages were 50/50: repeat test percentages were 75/25.
^Cycle timing controlled by exhaust gas (breakthrough) hydrocarbon analyzer.
cFlow rate measured when 4 of the 6 beds were in service (2 beds were on permanent standby).
^Process Train 1 conditions (total system capacity was twice that of Process Train 1).
^Measurements made at common inlet/common outlet.
' Removal efficiency is low due to steam valve leakage (99 percent efficiency was measured after repair of the valves).
T. P. Nelson, J. R. Blacksmith, andJ. L. Randall are with Radian Corp., Austin, TX
78766.
Ronald J. Turner is the EPA Project Officer (see below).
The complete report, entitled "Full-Scale Carbon A dsorption Applications Study, "
(Order No. PB 85- 1 72 906/AS; Cost $20.50, subject to change) will be available
only from:
National Technical Information Service
5285 Port Royal Road
Springfield, V A 221 61
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Hazardous Waste Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268

                                             ft U.S. GOVERNMENT PRINTING OFFICE: 1985-559-016/27054

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
        OCOC329   PS

        U  S  ENVIR PROTECTION  AGENCY
        REGION 5  LIBRARY
        230  S  DEARBORN  STREET
        CHICAGO              IL   60*04

-------