vvEPA
United States
Environmental Protection
Agency
Industrial Environmental
Laboratory
Research Triangle Park NC 2771 1
EPA-600/9-82-017
September t982
Research and Development
Symposium
Proceedings:
Environmental
Aspects of Fuel
Conversion Technology-
VI, A Symposium on
Coal-Based Synfuels
(October 26-30, 1981)
-------
EPA-600/9-82-017
Symposium Proceedings:
Environmental Aspects of Fuel Conversion Technology - VI,
A Symposium on Coal-Based Synfuels
(October 26-30, 1981)
F. A. Ayer and N. S. Jones, Compilers
Research Triangle Institute
P. 0. Box 12194
Research Triangle Park, NC 27709
Contract No. 68-02-3170
Task No. 56
Program Element No. CCZN1A
EPA Project Officer: N. Dean Smith
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
These proceedings for the symposium on "Environmental Aspects of Fuel Conver-
sion Technology-Vl" constitute the final report submitted to the Industrial
Environmental Research Laboratory, U.S. Environmental Protection Agency,
(EPA/IERL-RTP), Research Triangle Park, NC. The symposium was conducted in
Denver, Colorado, October 26-30, 1981.
This symposium provided a forum for the exchange of ideas and for discussion
on environmentally related information on coal gasification and liquefaction.
The program included sessions on environmental source test and evaluation
results for gasification and indirect liquefaction, and for direct liquefac-
tion, on water-related environmental considerations, on solid waste-related
environmental considerations, on multimedia environmental considerations,
and on product-related environmental considerations.
Process developers and users, research scientists, and state and federal
officials participated in the symposium, the sixth to be conducted on this
subject by IERL-RTP since 1974.
N. Dean Smith, Gasification and Indirect Liquefaction Branch, EPA/IERL,
Research Triangle Park, NC, was the Project Officer and the Technical
Chairman. William J. Rhodes, Synfuels Technical Coordinator for EPA/IERL-RTP,
was General Chairman.
Franklin A. Ayer, Manager, and N. Stuart Jones, Analyst, Technology and
Resource Management Department, Center for Technology Applications, Research
Triangle Institute, Research Triangle Park, NC, were symposium coordinators
and compilers of the proceedings.
In these proceedings, the title of each paper that has resulted from an EPA-
funded project is marked with a (t) to indicate that it has been reviewed in
accordance with the U.S. Environmental Protection Agency's peer and admin-
istrative review policies and approved for presentation and publication.
The absence of a (t) in the title of a paper in these proceedings indicates
that the paper is not the result of EPA-funded work and, therefore, its con-
tents do not necessarily reflect the views of the Agency, and no official
endorsement should be inferred.
ii
-------
TABLE OF CONTENTS
Page
OPENING SESSION 1
Session I: ENVIRONMENTAL SOURCE TEST AND EVALUATION RESULTS. ... 1
PART A: GASIFICATION AND INDIRECT LIQUEFACTION 1
Robert C. Lagemann, Chairman
William C. Yee, Cochairman
Characterization of Process Liquids and Organic Condensates
from the Lurgi Coal Gasification Plant at Kosovo, Yugoslaviat ... 2
Karl J. Bombaugh,* Kenneth W. Lee, Ronald G. Oldham, and
Slobodan Kapor
Application of Kosovo (Lurgi) Gasification Plant Test Results
to Pollution Control Process Designt 23
G. C. Page, W. E. Corbett, and R. A. Magee*
Environmental Aspects of the GKT Coal Gasification Processt .... 42
R. E. Wetzel,* K. W. Crawford,and W. C. Yee
Source Test of the Texaco Gasification Process Located
at Oberhausen-Holten, West GermanyT 57
Robert G. Wetherold,* Robert M. Mann, John Morgan,
William Yee, Peter Ruprecht, and Ranier Diirfield
Source Test and Evaluation of a Riley Gas Producer Firing
North Dakota LigniteT 66
Fred L. Jones,* William P. Earley, M. R. Fuchs, and
V. A. Kolesh
PART B: DIRECT LIQUEFACTION 94
W. Gene Tucker, Chairman
Morris H. Altschuler, Cochairman
Environmental Program and Plans for the EDS Coal Liquefaction
Project 95
Richard L. Thomas
Sampling and Analysis of Process and Effluent Streams from
the Exxon Donor Solvent Coal Liquefaction Pilot Plantt 107
Mark Notich* and Jung Kim
^Denotes speaker
iii
-------
Page
Health and Environmental Studies of H-Coal Process 124
K. E. Cowser,* J. L. Epler, C. W. Gehrs, M. R. Guerin,
and J. A. Klein
Chemical Characterization and Bioassay of SRC Process
Materials 148
W. Dale Felix," D. D. Mahlura, B. W. Wilson, W. C. Weimer,
and R. A. Pelroy
Session II: WATER-RELATED ENVIRONMENTAL CONSIDERATIONS 149
N. Dean Smith, Chairman
William E. Corbett, Cochairman
Coal Conversion Wastewater Treatment/Reuse - An Overview 150
F. E. Witmer
Characterization of Coal Conversion Wastewaters Using On-Site
GC/MSt 170
C. J. Thielen* and R. V. Collins
Treatment of Wastewater from a Fixed-Bed Atmospheric Coal
Gasifiert - 186
Philip C. Singer" and Eli Miller
Treatment of Fossil Fuel Derived Wastewaters with Powdered
Activated Carbon/Activated Sludge Technology 203
R. B. Ely" and C. L. Berndt
Land Treatment of Coal Conversion Wastewaters 218
R. C. Sims* and M. R. Overcash
Session III: AIR-RELATED ENVIRONMENTAL CONSIDERATIONS 231
Theodore G. Brna, Chairman
Removal of Acid Gases and Other Contaminates from Coal
Gas Using Refrigerated Methanolt 232
J. K. Ferrell, R. M. Kelly, R. W. Rousseau, and
R. M. Felder*
Advanced Techniques for Flue Gas Desulfurizationt 256
Charles C. Masser, Theodore G. Brna, and Michael A. Maxwell*
Health and Environmental Studies of Coal Gasification Process
Streams and Effluents 282
C. A. Reilly, Jr." A. S. Boparai, S. Bourne, R. D. Flotard,
D. A. Haugen, R. E. Jones, F. R. Kirchner, T. Matsushita,
M. J. Peak, V. C. Stamoudis, J. R. Stetter, and K. E. Wilzbach
"'Denotes speaker
iv
-------
Page
Gaseous Fugitive Emissions from Synfuels Production - Sources
and Controls!" 293
R. L. Honerkamp
Control Systems for Air Emissions from Coal Gasification 313
Sid Thomson
Session IV: SOLID WASTE-RELATED ENVIRONMENTAL CONSIDERATIONS . . .328
David A. Kirchgessner, Chairman
Kimm W. Crawford, Cochairman
Health Effects Bioassay Results from Coal Conversion Solid
Wastes 329
M. P. Maskarinec,* F. W. Larimer, J. L. Epler, and C. W. Francis
A Comparison of RCRA Leachates of Solid Wastes from Coal-Fired
Utilities and Low- and Medium-Btu Gasification Processest 341
Michael R. Fuchs,* Donnie L. Heinrich, Larry J. Holcombe,
and Kishore T. Ajmera
Characterization of Solid Wastes from Indirect Liquefaction
Facilitiest. 358
Cora A. Hunter, Kar Y. Yu, and Kimm W. Crawford"
Ash/Slag Residuals and Wastewater Treatment Plant Sludges
from Synfuels Facilities: Characterizations and Implications
for Disposal 380
Ronald D. Neufeld,* Georg Keleti, J. Bern,
C. Moretti, S. Wallach, and H. Erdogen
Update of EPA's Regulatory Views on Coal Conversion
Solid Wastes? 397
Yvonne M. Garbe
Session V: MULTIMEDIA ENVIRONMENTAL CONSIDERATIONS 398
T. Kelly Janes, Chairman
John T. Dale, Cochairman
A Permitter's View of Synfuel Commercialization! 399
George L. Harlow
Comparison of Environmental Design Aspects of Some Lurgi-Based
Synfuels Plants t 400
Milton R. Beychok" and William J. Rhodes
Session VI: PRODUCT-RELATED ENVIRONMENTAL CONSIDERATIONS 421
Robert P. Hangebrauck, Chairman
Minh Triet-Lethi, Cochairperson
'''Denotes speaker
-------
Page
Risk Assessment of Synfuel Technologyt 422
A. Alan Moghissi
Premanufacture Review of Synfuels under TSCAt 423
Matthew Hale, Jr.* and Carl Mazza
Methanol as a Clean Major Fuel 440
Paul W. Spaite
Methanol as an Alternative Transportation Fuelt. 466
Richard Rykowski, Dwight Atkinson,* Daniel Heiser,
John McGuckin, David Fletcher, Jeff Alson, and
Murray Rosenfeld
Project Summary - A Compendium of Synfuel End Use Testing
Programs! 489
Masood Ghassemi,* Sandra Quinlivan, and Michael Haro
Comparative Testing of Emissions from Combustion of Synthetic
and Petroleum Fuelst 509
W. Gene Tucker* and Joseph A. McSorley
UNPRESENTED PAPERS 523
Problems Associated with the Analysis of Synfuels Product,
Process, and Wastewater StreamsT 524
H. C. Higman, D. K. Rohrbaugh, R. H. Colleton, and
R. A. Auel
Solvent Extraction Processing for Coal Conversion Wastewaters . . .535
James R. Campbell, Richard G. Luthy, and Manuel J. T. Corrondo
APPENDIX: Attendees 545
'^Denotes speaker
VI
-------
Session I: ENVIRONMENTAL SOURCE TEST AND EVALUATION RESULTS
Part A: Gasification and Indirect Liquefaction
Chairman: Robert C. Lagemann
U.S. Environmental Protection Agency
Research Triangle Park, NC
Cochairman: William C. Yee
Tennessee Valley Authority
Knoxville, TN
-------
CHARACTERIZATION OF PROCESS LIQUIDS AND ORGANIC t
CONDENSATES FROM THE LURGI COAL GASIFICATION
PLANT AT KOSOVO, YUGOSLAVIA
by: Karl J. Bombaugh, Kenneth W. Lee and Ronald G. Oldham
Radian Corporation
8501 Mo-Pac Blvd.
Austin, Texas 78766
and
Slobodan Kapor
Institut za Primenu Nuklearne Energy
Baranj ska 15
11080 Beograd - Zemun
Yugoslavia
ABSTRACT
Process liquids and gaseous stream condensates from the Lurgi Coal
Gasification plant at Kosovo were characterized to define their organic compo-
sition. Samples of entrained liquids and condensates were collected during
Phase II of the Kosovo source test that was described at the preceeding Syn-
fuel Symposium. These samples were characterized by Liquid Chromatographic
fractionation using EPA's protocol for a Level I source assessment. In addi-
tion, GC-MS analyses were performed on key samples to quantify their levels of
potentially hazardous PNA's, and GC with selective detection was used to
characterize sulfur and nitrogen bearing species.
This presentation will provide a discussion of the analytical results and
of the impact that these condensates have on the plant's discharge stream
severity. It will also include a comparison of the composition of liquids
from the Lurgi process with the compositions of liquids from other processes.
INTRODUCTION
Process liquids, gas stream condensates and solid wastes from the Kosovo
Coal Gasification Plant were characterized to determine their organic composi-
tion. Samples were taken from fourteen gas streams, plus five liquid and two
solid phase streams during Phase II of the source test that was sponsored
jointly by the United States Environmental Protection Agency and the Govern-
ment of Yugoslavia. Sampling and analyses were conducted as a cooperative
effort by American and Yugoslav scientists (1, 2).
-------
The 21 streams selected for organic characterization are identified in
Table 1. These streams provided a representative cross section of the Lurgi
technology that is used at Kosovo. The locations of these streams are included
in the plant's description.
TABLE 1. ANALYSES PERFORMED ON KOSOVO GAS STREAM CONDENSATES, PROCESS LIQUID AND SOLID WASTES
Analyses
Stream
Particulate
Determination
TCO
Grav
LC
GC
Sulfur
GC
Nitrogen
GCMS
PNA'S
Gas Stream
Fleissner Autoclave Vent x
LP Coal Lock Vent x
HP Coal Lock Vent x
Start-up Vent x
H2S-Rich Waste Gas
C02-Rich Waste Gas
Crude Product Gas
Tar Tank Vent
Medium-Oil Tank Vent
Tar Separation Waste Gas x
Phenolic Water Tank Vent
Ammonia Stripper Vent
Naphtha Storage Tank Vent
Waste Gasses to Flare x
Other Streams
Fleissner Condensate
Gasifier Ash
Heavy Tar x
Phenolic Water x
Tar
Medium Oil
Naphtha
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
©
©
©
x - analyzed in Yugoslavia.
+ - analyzed in US using grab samples taken at random.
©- Data not included in this report, but Included In Reference 2.
The gasification plant consists of nine operational units as illustrated
in Figure 1. The plant consumes dried lignite and produces two primary pro-
ducts (a medium Btu fuel gas and hydrogen) plus four by-products (tar, medium
oil, naphtha and crude phenol). The plant's operation is as follows: run of
the mine coal is dried in steam autoclaves by the Fleissner process and then
sized to select particles suitable for the Lurgi gasifiers (dp =6-60 mm).
The dried coal is fed through a pressure lock system (Coal Lock Vents) to a
3.5 M diameter Lurgi gasifier where it is reacted with steam at 2.5 MP (25 atm)
pressure to produce a crude gas which is quenched, cooled and then cleaned by
the Rectisol process prior to its transport to the utilization site. As the
hot gas is quenched and cooled, organic matter consisting of phenols, tars and
oils are removed with quench liquor and hot gas condensate. In the gas clean-
ing operation, condensable organics are removed from the cooled gas by refri-
geration after which the acid-gases (H2S and COz) are removed by sorption in
cold methanol. The acid-gas rich methanol is regenerated by depressurization
and heating, releasing HzS-rich Waste Gas which is flared and COa-Rich Waste
Gas which is vented into the atmosphere.
-------
Mtdlum
» BluQH
10 P
By-Product!
.to Sttam and
^ Prwir
Qamratlon
Figure 1. Simplifed Flow Diagram of the Kosovo Coal Gasification Plant.
Tar, heavy tar and medium oil are each separated from their aqueous phase
by decantation after which the combined waters are depressurized (Phenolic
Water Tank) then stripped to remove ammonia (Ammonia Stripper Vent) and then
extracted with disopropyl ether to remove extractable organics (Crude Phenol)
prior to disposal.
EXPERIMENTAL
The sample characterization program consisted of a combination of the
following methods: gas and liquid chromatographic fractionation using EPA's
protocol for a Level 1 Environmental Source Assessment to determine the mass
distributions of volatile and non-volatile organics; gas chromatographic
analyses with both universal and selective GC detections to characterize or-
ganics; and GC-MS analysis to quantify certain potentially hazardous poly-
nuclear aromatic compounds. The distributions of volatile and non-volatile
organics were determined by the EPA protocol as Total Chromatographable
Organics (TCO) and Gravimetrically Determined Organics (Grav) respectively (3).
Condensable organics were collected from gaseous streams with a sampling
train that consisted of an entrainment separator, an ice cooled condenser, and
a resin filled absorber in series. In some cases, sample collections were
made in conjunction with a particulate measurement for which the entrained and
condensed liquids were combined and then divided equally for the particulate
determination and the organic characterization. In most cases, collections
were made specifically for the organic characterization. All samples were re-
frigerated during the storage period between collection and work-up for analy-
sis.
Sorbed vapors from the respective streams were recovered from their
collection resin by soxhlet extraction with methylene chloride and were
-------
combined with the organic extracts from their stream's condensates prior to
analysis by the EPA protocol for TCO. The strategy followed for these deter-
minations is illustrated in Figure 2.
To supplement the information provided by the TCO and Grav determinations,
the extracts from selected streams (Table 1) were analyzed by gas chromato-
graphy with element specific detection to obtain profiles of the sulfur and
nitrogen-containing species.
Polynuclear aromatics (PNA's) were determined on several streams by GC-MS.
A liquid crystal GC column was used to isolate Benzo(a)pyrene from other
isomeric PNA's.
The streams sampled and the analyses performed are summarized in Table 1.
RESULTS AND DISCUSSION
The concentrations of organics in the thirteen major gaseous streams of
the Kosovo gasification plant (shown in Table 2) indicate that the phenolic
water tank discharges the highest concentration of organics in the Kosovo plant
(1.2 x 10s mg/m3) and that a major portion (92%) of the^emission is due to TCO.
The TCO value obtained from this measurement is a factor of five higher than
the level indicated by light aromatic concentrations that were determined
during the Phase II test. The significant difference between these two values
is not explained. The discharge from the ammonia stripper vent also contains
a high concentration of organics but in this stream the mass concentrations of
the TCO hereafter called volatile organics (VO) and Grav hereafter called non-
volatile organics (NVO) are about equal (57% volatiles). Excluding the two
high concentration streams, the average concentration of combined organics in
tested streams was 5,800 mg/m3 with most values falling in a range between
1,000 and 20,000 mg/m3. As expected, the lowest concentration of combined
organics was found in the COa-rich waste gas. The value obtained in this
stream for non-volatile organics may not be significant since it is based on
a single determination and probably is within the noise level of the analyti-
cal method. A more accurate measurement is required to establish, definitively,
the level of condensable organics in this stream.
A comparison of the concentration levels of VO's with the concentrations
of light aromatics, as determined during Phase II of the source test, is shown
in Table 3. The light aromatic values shown are the sum of phenol plus Ci and
Ca alkyl aromatics. These light aromatics are included in the volatile or-
ganics determination and appear to represent a significant portion (30 to 60%)
of the materials determined as VO. In cases where the concentrations of light
aromatics are greater than the concentration of VO, it is not known whether
the difference is due to variations in the stream's composition or to analyti-
cal error. However, the VO and light aromatic concentration values, when
considered together, provide a reasonable indication of the quantities of
volatile organics in the respective streams.
A similar comparison is made in Table 4 between the concentration values
of NVO and the Tar and Oil (T&O) values that were reported for particulates in
several streams. In this comparison, most values fall within a factor of 2
-------
JconvSi^Kl-LT p^l'tll. U Concentrate
/I Record Volume |^ Starti to Form || •' to 5.0 mf
I IStop ifI IDilute ToI
centrate I 1 precipitate 1 Convenient Volude k
' *•" "^ I | Starts to Form | | Record Voluae |\
*~~ ^~~- f~^ s-^S^ . 1 ^
Take A mi-
Aliquot
*Crimp-seal Sample
( GRAV J
Figure 2. Strategy Followed for the Characterization of Kosovo Organics.
-------
TABLE 2.
CONCENTRATIONS OF VOLATILE AND NONVOLATILE ORGANICS IN KOSOVO
PLANT GAS STREAM CONDENSATES
Source
Fleissner Autoclave Vent
LP Coal Lock Vent
HP Coal Lock Vent
Start-up Vent
H2S-Rich Waste Gas**
C02-Rich Waste Gas
Tar Tank Vent
Medium Oil Tank Vent
Tar Separation Waste Gas
Phenolic Water Tank Vent
Ammonia Stripper Vent
Naphtha Storage Tank Vent
Combined Gas to Flare
Volatile
Organics1
(TCO)
306
3,732
1,622
2,670
40
5
10,785
19,921
2,335
115,012
56,167
5,089
312
mg/m3
Non-Volatile
Organics2
(Grav)
807
4,007
1,250
7,053
90
9
3,628
1,197
967
9,869
43,051
499
290
Total
Organics3
1,113
7,739
2,872
(9,723)
130
14
14,412
21,118
3,302
124,881
99,218
5,588
602
%
Volatile
Organics
27
48
57
27
31
36
75
94
71
92
57
91
52
% Non-
Volatile
Organics
73
52
43
73
69
64
25
6
29
8
43
9
48
'Volatile organics were determined as total chromatographable organics (TCO's) using EPA's Protocol
which is based on a gas chromatographic determination of substances eluting in the range of C? to Cie
hydrocarbons representing a boiling range between 100 and 300°C.
2Non-volatile organics were determined gravimetrically using EPA's Protocol for "Grav" which includes all
substances retained from a 24-hour ambient evaporation. The Protocol may allow the same mid-range vola-
tiles to be included in both determinations; consequently the total may be higher than the true value.
3These values are a summation of toluene, xylene, and phenol as determined during the Phase II test and
are included for comparison.
**No XAD-2 value included.
TABLE 3. COMPARISON BETWEEN VOLATILE ORGANICS (TCO) AND SUMMED
LIGHT AROMATICS (Ci + C2 ALKYL AROMATICS + PHENOL)
AS DETERMINED DURING PHASE II OF THE KOSOVO SOURCE
TEST
Source
LP Coal Lock Vent
HP Coal Lock Vent
H2S-Rich Waste Gas
C02-Rich Waste Gas
Tar Tank Vent
Medium Oil Tank
Tar Separation Waste Gas
Phenolic Water Tank
Ammonia Stripper Vent
Naphtha Storage Tank
Volatile
Organics
3,732
1,622
40
5
10,785
19,921
2,335
115,012
56,167
5,089
Z Light
Aromatics
1,170
1,730
30
10
3,790
6,060
5,190
21,670*
23,800
7,410
*Phenol value was calculated from its vapor pressure over a saturated aqueous
solution ?t 65°C.
-------
TABLE 4. COMPARISON BETWEEN ORGANICS COLLECTED BY" PARTICULATE
TRAIN AND. THOSE COLLECTED VIA ORGANIC TRAIN
Source
Fleissner Autoclave Vent
LP Coal Lock Vent
HP Coal Lock Vent
Start-up Vent
Tar Separation Waste Gas
Tar &
Oil
534
7,920
953
9,800
723
Non-Volatile
Organics
808
4,007
1,634
7,051
967
Total
Organics
E VO + NVO
1,114
7,739
3,753
9,721
3,302
of each other (mean =1.2+0.5). As with the previous comparison, these
results provide a reasonable indication of the level of non-volatile organics
(Tars and Oil) which are transported by the respective streams.
The mass flow of combined organics in each stream is shown in Table 5.
These results show that two streams (the start-up vent and the ammonia strip-
per vent) transport, by far, the greatest quantity of organic matter (98%).
Either of these streams, when operating, transports more condensable organics
than all of the other streams combined.
TABLE 5. MASS FLOW OF CONDENSABLE ORGANICS IN KOSOVO GASEOUS
STREAMS
Source
LP Coal Lock Vent
HP Coal Lock Vent
Start-up Vent
H2S-Rich Waste Gas
C02-Rich Waste Gas
Tar Tank Vent
Medium Oil Tank Vent
Gar Separation Waste Gas
Phenolic Water Tank Vent
Ammonia Stripper Vent
Naphtha Storage Tank Vent
Stream Flow
m3/hr
21
230
12,500
3,600
3,600
0.55
1.7
28
5.5
260
4.5
Mass Flow
g/hr
163
660
121,538
468
50
8
36
92
687
25,797
25
%
Non-Volatile
52
43
73
69
64
25
6
i 29
8
43
9
Since both the composition and the flow rate of the discharge from the
start-up vent varies considerably over the start-up period, the values pre-
sented here may represent a worst-case; nevertheless, the discharge is signi-
ficant because a large gasification complex having many gasifiers can be
expected to have at least one gasifier in a start-up mode at all times. Under
such conditions, the start-up stream could flow continuously.
The ammonia stripper vent, reported previously as the most environ-
mentally significant stream in the Kosovo Plant (1), is also the major
-------
source of condensable organics. This stream is intermittent at Kosovo because
the Phenosolvan plant is operated on demand, whereby phenolic water is accu-
mulated in two large tanks and processed at a rate that is independent of
the gas production rate. However, when the plant is operated at design capa-
city, continuous operation is necessary and the stripper vent is then a
continuous discharge stream.
The phenolic water tank discharge contains the highest concentration of
condensable organics, but because of its lower stream volume, its organic mass
discharge rate is comparable to lower concentration but higher volume streams,.
e.g., the HP coal lock vent and the HzS-rich waste gas stream.
The naphtha tank discharge which contained high concentrations of benzene,
contained comparatively little volatile organics as defined by the protocol for
Total Chromatographable Organics.
LC FRACTIONATION OF CONDENSABLE ORGANICS FROM KOSOVO STREAMS
The mass distribution of organic matter in the Kosovo condensates, as
determined by the EPA Level 1 fractionation protocol, is shown in Table 6.
All data are given as stream concentration, expressed in mg/m3. In this form,
the values shown for each fraction do not indicate the mass recovered from the
column, but rather the computed mass concentration in one cubic meter of gas.
Therefore., fraction concentrations are directly relatable to stream concen-
trations.
TABLE 6. CONCENTRATIONS OF ORGANICS FOUND IN EACH LIQUID CHROMATOGRAPHIC FRACTION AND IN THE SAMPLE STREAM
LC Fraction (ra
Source
Fleissner Autoclave Vent
LP Coal Lock Vent
HP Coal Lock Vent
Start-up Vent
H2S-Rlch Waste Gas
Tar Tank Vent
Medium Oil Tank Vent
Tar Separation Waste Gas
Phenolic Water Tank Vent
Ammonia Stripper Vent
Naphtha Storage Tank Vent
1
229
167
395
458
19
3,734
2,275
266
32,190
1,880
342
2
47
79
285
430
5
449
229
105
3,335
2,607
25
3
80
426
145
1,743
12
410
246
154
2,458
2,222
40
4
47
1,456
168
160
7
652
250
129
3,650
16,923
44
g/m3)
5
73
1,281
282
268
8
753
335
183
4,185
17,692
34
6
203
2,297
563
1,595
18
2,179
1,061
649
10,847
27,949
380
7
20
266
67
302
4
225
76
38
857
4,145
37
Total
Recovered
mg/m3
699
5,981
1,912
4,956
73
8,402
4,471
1,525
51,737
73,419
902
Concentration*
in Stream
mg/m3
1,114
7,739
2,872
5,540
130
14,412
21,118
3,302
124,884
99,218
5,589
All values computed to stream concentration and expressed as milligram per normal cubic meter of gas
*From Table 2.
Recoveries of organics from the LC fractionation averaged about 50%
(Table 7). Recoveries of VO and NVO were computed separately, consequently
the values shown for combined organics are weighted values. These results
show that NVO recovery, generally, was better than VO's recovery; probably
because the component loss to evaporation was more significant than loss
through non-elution.
The composition of each chromatographic fraction is defined, in part, by
the polarity of the eluting solvent. Consequently, all components in a given
chromatographic fraction should have similar polarity, but they may represent
widely differing chemical classes. Some chemical classes which could be found
in the respective fractions are shown in Table 8, Fraction 5, which shows no
entry should contain overlap from adjacent fractions.
-------
TABLE 7. RECOVERIES OF ORGANICS FROM THE LIQUID CHROMATOGRAPHIC PROCEDURE
Stream
Fleissner Autoclave Vent
LP Coal Lock Vent
HP Coal Lock Discharge
Gasifier Start-up Vent
HzS-Rich Waste Gas
Tar Tank Vent
Medium Oil Tank Vent
Tar Separation Waste Gas
Phenolic Water Tank Vent
Ammonia Stripper Vent
Naphtha Storage Tank Vent
Volatile
Organics
41
109
66
45
40
58
17
38
42
98
9
Percent Recovery
Non-Volatile
Organics
70
47
67
54
65
58
93
93
91
43
83
Combined
Organics
62
77
66
49
58
58
21
46
46
74
16
Mean 43
SD +25
69
+18
53
+19
TABLE 8. PROBABLE LC FRACTION IN WHICH VARIOUS COMPONENT CLASSES APPEAR
Eluting
Solvent
Possible
Chemical
Classes
in
Fraction
Fraction
Number
tUlog«a*c*d
Aliphaclci
Uacerocyc,
FuMd-Rlng irdrocartaon*
Aroutlc H
Aliphatic*
dro carbon*
Thioph*n««
IndolM &
CarbotolM
NltrllM
Ic Qryf«n
Ald«hr
-------
The relative distributions of the eluted organics across the seven
fractions are shown in Table 9. These results indicate that fraction 6,
which should include phenols and nitrogen heterocyclics, contains the largest
portion of the eluate (Ave = 31% + 8) followed by fraction one [Ave = (27% +
19)] which should contain only paraffins.
TABLE 9. PERCENT OF ELUTED ORGANICS FOUND IN EACH LC FRACTION AND
PERCENT RECOVERY FROM THE LC SEPARATION PROCESS
Source
Fleissner Autoclave. Vent
LP Coal Lock Vent
HP Coal Lock Vent
Start-up Vent
H2S-Rlch Waste Gas
Tar Tank Vent
Medium Oil Tank Vent
Tar Separation Waste Gas
Phenolic Water Tank Vent
Ammonia Stripper Vent
Naphtha Storage Tank Vent
Tnt" al VI tit"i nn
lULdJ. dJ.ULJ.On
mg/m3
699
5,981
2.499
-
59
8,402
4,471
1,525
57,522
59,880
902
Fraction Number
1
33
3
21
9
25
44
51
17
56
3
38
2
7
1
15
(9)
7
5
5
7
6
4
3
3
11
7
7
(35)
17
5
5
10
4
3
4
4
7
24
9
(3)
10
8
6
8
6
23
5
5
10
21
15
(5)
11
9
7
12
7
24
4
6
29
38
29
(32)
24
26
24
43
19
38
42
7
3
4
4
(6)
6
3
2
3
2
6
4
P£ IT cent
Recovered
63
77
67
51
56
58
21
46
46
74
16
A three-dimensional view of the relative distribution of chemical classes
across the key streams in the Kosovo plant is shown in Figure 3. From this
view, it is evident that a significant portion of the condensable organics
from each stream is found in fraction 6 which contains phenols and nitrogen
heterocyclics and that most streams contain relatively large proportions of
alkyl aromatics as found in fraction 1. Fraction 3 is seen to be larger in
the Start-Up Vent and H2S-Rich Waste Gas's condensate. This fraction could
contain thiophenes, indols, nitriles, and oxygen heterocyclics which are nor-
mally found in fraction 4 but probably overlap.
The low level of alkyl aromatics in the ammonia stripper vent condensate
supports substantially the Phase II source test results which showed virtually
no benzene or toluene in that vent's discharge.
A COMPARISON OF CONDENSABLE ORGANICS IN KOSOVO STREAMS WITH THOSE OF OTHER
TYPES OF GASIFIERS
A comparison of the levels of condensable organics in vent gases from
the Kosovo plant with those in "similar" streams from other types of gasi-
fiers indicates that the organics concentration levels generally are com-
parable. The comparison data are shown in Table 10. The lower level of
volatile organics in the Chapman coal vent discharge may be because the
Chapman gasifier was using bituminus coal whereas all others listed were
using lignite.
11
-------
Figure 3. Distribution of Chemical Classes across the LC Fractions of
Kosovo Gas Stream Condensates.
TABLE 10. A COMPARISON OF ORGANICS IN "SIMILAR" STREAMS FROM
DIFFERENT GASIFICATION PLANTS (4-6)
Source
Lurgi (Kosovo) LP Coal Lock
Lurgi (Kosovo) HP Coal Lock
Chapman (Holston) Coal Feeder Vent
Riley Product Gas
Wellman-Galusha (Ft. Snelling)
Product Gas
Kosovo Tar Separation Offgas
Chapman Separator Vent
Volatile
Organics
3,732
2,121
378
3,643
5,900
2,335
1,897
mg/m3
Non-Volatile
Organics
4,007
1,632
2,002
2,186
2,100
967
2,303
Total
Organics
7,739
3,753
2,380
5,829
8,000
3,302
4,200
12
-------
CHARACTERIZATION OF KOSOVO PROCESS LIQUIDS AND SOLID WASTES
The concentrations of organics in Kosovo process liquids and solid wastes
are summarized in Table 11. These results indicate that phenolic water is
transporting approximately 11 g/l of organics of which 32% are volatile.
Heavy tar consists of about 86% organics - the balance probably being a com-
bination of error in the determination and inorganics in the coal; heavy tar
is known to contain a considerable amount of coal dust (26% insoluble parti-
cles). Gasifier ash, also listed in Table 11, contains minimal organic matter
(.04%) of which none was volatile.
TABLE 11. CONCENTRATIONS OF VOLATILE AND NON-VOLATILE AND TOTAL ORGANICS
IN KOSOVO PROCESS LIQUIDS AND SOLID WASTES
Phenolic Water
Medium Oil
Tar
Heavy Tar
Gasifier Ash
Units
mg/1
mg/1
mg/g
mg/g
mg/g
mg/g
mg/kg
Volatile
Organics
3,569
3,774
389
334
567
460
Non-Volatile
Organics
7,556
7,371
439
452
778
404
400
Total
Organics
11,225
11,145
828
786
1,345
864
400
Percent
Volatile
32
34
47
42
42
53
0
Percent Non-
Volatile
68
66
53
58
58
47
100
For definition of volatile and non-volatile, see Table 2.
The by-products, tar and medium oil, show very similar distributions be-
tween volatile and non-volatile organics. When their vastly differing boiling
point ranges are considered, this similarity is surprising. However, the low
recovery of total organics from the medium oil suggests that a significant
amount of sample was lost in the determination (^20%). If a correction for
this loss were applied to the volatile organics, a more reasonable value
would be obtained (55%).
The distribution of organics in various Kosovo tars and oils as deter-
mined by liquid chromatography is shown in Table 12. Although these materials
can be expected to have vastly different compositions, as would be indicated
by their differing solubilities and boiling point ranges, they have surpris-
ingly similar chromatographic profiles.
The close similarity in the LC profiles from the Kosovo streams, whose
chemical compositions may differ significantly indicates that more discriminat-
ing methods of separation and detection are needed to obtain descriptive pro-
files of these streams.
GAS CHROMATOGRAPHIC ANALYSIS WITH ELEMENT SPECIFIC DETECTORS
As a supplement to the information that was provided by the LC Fractiona-
tion, the condensable organics from several streams were examined by gas
13
-------
TABLE 12. PERCENT ELUTED ORGANICS FOUND IN EACH LC FRACTION FROM
THE LC SEPARATION OF SEVERAL KOSOVO LIQUIDS AND HEAVY
TAR
Percent in Each Fraction
Fraction
1
2
3
4
5
6
7
Phenolic
Water
19
15
19
4
10
31
2
Heavy
Tar
10
6
16
11
14
39
4
Tar
17
2
15
9
9
39
9
Medium
Oil
12
6
16
4
10
45
7
chromatography with element specific detectors. Chromatograms of the sulfur-
containing species and of the nitrogen-containing species that were obtained
in this manner show that the Kosovo streams contained complex mixtures of both
types of compounds.
The sulfur-specific chromatograms of condensable organics from three
streams in Figure 4 show both similarities and differences in these materials.
For example, the chromatograms of the LP and the HP coal lock vent have many
peaks in common, while the chromatogram of the ammonia stripper condensate is
distinctly different from those of the coal lock condensates. The stripper
condensate may contain a relatively larger amount of the more water soluble
sulfur-containing species.
The condensate from the HP coal lock shows considerably less background
matrix effect which may relate in part to the influences of the Venturi scrub-
ber through which the gas had passed. The peaks labeled 1-4 have been iden-
tified as thiophenes. Peaks labeled 5-8 are unidentified; however, their
intensity suggests that they represent materials which contain a higher sulfur
to carbon ratio than the thiophenes (possibly two or more sulfurs per molecule,
e.g., disulfides or dithiols). Their uniform difference in retention suggests
that they may be an homologous series of icomcrs.
Nitrogen-specific chromatograms of the stream condensates indicate that
the Kosovo condensable organics contain several classes of organic nitrogen
compounds. The chromatogram of the condensates from four streams, shown in
Figure 5, indicate that these samples contained many of the same components.
Several peaks have been identified tentatively as isomers of pyridine and
quinoline. In contrast, the chromatograms in Figure 6 indicate that the nitro-
gen species in some streams differ significantly from the others. For
example, the chromatogram of the condensable organics from the stripper vent
(A) and the medium oil tank (C) differs significantly from that of the phenolic
water tank vent (B). (The latter chromatogram (B) is similar to those in
Figure 5). These results indicate that the major components in the two groups
of samples are different compounds rather than the same compounds which have
distributed differently between the two groups of streams.
14
-------
10
20
100
200 300
°C boiling point
Figure 4. Chromatograms of Sulfur Species in
Condensates from 3 Kosovo Gas Streams.
Top - LP Coal Lock
Center Ammonia Stripper Vent
Bottom - HP Coal Lock
115
150
200
250
minutes
°C boiling Point
Figure 5. Chromatograms of Nitrogen Species in
Condensates from 4 Kosovo Gas Streams-
A - LP Coal Lock Vent
B - HP Coal Lock Vent
C - Tar Separation Waste Gas
D - Tar Tank Vent
'10
40 minutes
15
20
25 minutes
115 150 200 250 °C boiling point
Figure 6. Chromatograms of Nitrogen Species
in Condensates from 3 Kosovo Streams*
A - Ammonia Stripper Vent
B Phenolic Water Tank Vent
C Medium Oil Tank Vent
115 150 250 350 °C boiling point
Figure 7. Chromatograms of Nitrogen Species
in Kosovo Medium Oil.
A Neutral Fraction
B (2) Base Extractable Fraction
(contains acids 5, phenols)
C Acid Extractable Fraction
(contains organic bases)
15
-------
A most surprising result is the chromatogram of the condensable organics
from the medium oil tanks. This nitrogen-specific chromatogram differs from
those of the other streams and from previously prepared chromatograms of the
nitrogen species in Kosovo medium oil. Since the chromatogram of the condensed
discharge differs from that of the tank's oil, this condensate must not be a
product of simple vaporization.
INVESTIGATION OF KOSOVO MEDIUM OIL
A brief study was conducted on Kosovo medium oil using a combination of
physical/chemical separations and gas chromatography with element specific
detectors in an attempt to gain some insite into the cause of the observed
difference between the elemental chromatograms of the various stream conden-
sates. The Kosovo medium oil was separated into an acid-extractable fraction,
a base-extractable fraction and a neutral fraction. A water-oil co-distillate
was also obtained. Element specific chromatograms of each of these fractions
provide significant information about the medium oil which can be applied to
the condensable organics which are transported by the various Kosovo gaseous
streams.
The nitrogen-bearing species in Kosovo medium oil were found to be more
complex than the sulfur species in that they consisted of a complex mixture of
several different classes of compounds whose solubilities were markedly af-
fected by pH. The chromatograms in Figure 7 show that three distinctly dif-
ferent sets of compounds are found in the respective neutral, base extractable
and acid extractable fractions, and that each fraction contains numerous
(40-60) compounds. Since very few peaks show common retention times between
fractions, it is reasonable to conclude that there is minimal, if any, overlap
between fractions.
The acid extractable fraction should contain proton acceptors such as
pyridines and quinolines: the neutral fraction should contain the pyrazine
(diazines) and/or other more neutral nitrogen species including possibly
oxazoles and thiazoles; and the base extractable fraction should contain pro-
ton donors such as acids, alcohols and phenols.
The chromatogram in Figure 8 was obtained with the NaOH soluble material
that precipitated during the acid-extraction of the medium oil. Since pyroles
are known to polymerize in an acid medium and since the peak at 35 min. matches
that of cabazole (dibenzopyrole) it is conceivable that this fraction contains
pyroles.
A chromatogram of the water/oil co-distillate is shown in Figure 9 along
with a chromatogram of the vapor over a closed container of medium oil at
about 50°C. This comparison shows that except for the broad early eluting
peaks in the vapor, the two samples contain many of the same compounds. The
broad early eluting peaks in the vapor suggest that the vapors contain low
boiling, strongly polar nitrogen species which are interacting with the column.
These materials have not been identified or quantified in the Kosovo gas
streams.
16
-------
40 minutes
115 150 250 350 °C boiling
point
Figure 8. Chromatogram of Acid Insolible Fraction
from Kosovo Medium Oil. Identified peak
(+) has same retention time as Carbazole.
minutes
115
Figure 9.
150 200 250 300
°C boiling
point
Chromatograms of Nitrogen
Species in Medium Oil Head
Space Vapor (top) and in
Water-Oil Co-distillate
(bottom).
40 minutes
J
115 150 250 350 °C boiling
point
Figure 10. Chromatogram of Nitrogen Species in
Residual Water from Co-distillation
of Kosovo Medium Oil (top) Compared
with Chromatogram of Nitrogen Species
in Water-Oil Co-distillate (bottom).
10
20
minutes
100 150 250 350 °C boiling point
Figure 11. Chromatograms of Sulfur
Sp.ecies in Fractions of Kosovo
Medium Oil •
A Acid Extractable Fraction
B Base Extractable Fraction
C Neutral Fraction
17
-------
A chromatogram of the residual water from the co-distillation (still pot)
is shown in Figure 10 along with a chromatogram of the steam distillate.
These chromatograms also have many peaks in common, but the two materials
differ greatly in composition. Significant quantities of high molecular weight
materials have partitioned into the water layer from the medium oil producing
a fraction that is rich in high-boiling water soluble components. A portion
of the water layer chromatogram (5-15 min) is very similar to the chromatogram
of the condensate from the ammonia stripper vent shown in Figure 6.
The chromatograms of the several fractions of medium oil shown in Figures
6 through 10 show that the organic nitrogen species in the Kosovo organics re-
present an extremely complex mixture of compounds covering a wide range of
boiling points, solubilities, polarities and dissociation potentials. How-
ever, when these compounds are fractionated according to solubility, pH, and
finally vapor pressure, a set of profiles is obtained which defines the compo-
sition of the mixture. Because these fractionation methods are not absolute,
the composition of each fraction depends upon the separation method used, as-
well-as on the sample's compositions. Consequently, the influence of the
separation method must be considered when interpreting results.
A major factor in the complexity of the mixture of nitrogen compounds is
that the nitrogen species are influenced by so many different properties. The
stream's composition can be altered significantly by a slight change in pH, or
by the presence or absence of water or by an increased organic layer in an
aqueous process stream. Further, a sample's composition can also be in-
fluenced by a vapor collection method as well as by a vapor recovery method.
As was demonstrated here, exposing a sample to a strong acid can remove an
entire class of components. Consequently, a characterization based on a well
defined methodology, that takes into consideration the specific properties of
these nitrogen compounds, is needed before the composition of the Kosovo or-
ganics can be defined quantitatively.
The sulfur species in Kosovo medium oil also represent a range of solu-
bilities and dissociation potentials. However, the mixture appears to be
somewhat less complex than the mixture of nitrogen compounds. The neutral
fraction of medium oil contains many more components (Figure 11) than either
the acid extractable or the base extractable fractions. The neutral fraction
contains primarily the thiophenes and the mercaptans. The base extractable
fraction which should contain organic acids, and phenols probably contains bi-
functional (oxygen and sulfur) compounds also. The acid extractable fraction,
which was shown to contain the nitrogen bases, must also contain bifunctional
compounds such as thiozoles. .Both the acid extractable and the base extract-
able sulfur compounds appear to fit into a comparatively well defined boiling
point range.
When the sulfur specific chromatograms of Kosovo medium oil fractions are
compared with those of the plant's discharge stream condensates, it is ap-
parent that the neutral components are dominant in the gas streams associated
with the product gas. In contrast, the chromatogram of the ammonia stripper
vent shows numerous peaks which match the profiles of the acid and base
extractable fractions of medium oil, supporting the earlier premise that the
ammonia stripper condensate contains higher concentrations of the water
18
-------
soluble sulfur species. Bifunctional compounds such as thiozoles (S & N) and
thiophenols (S & 0), exhibit higher water solubilities than thiophenes and may
well be present in the stripper vent's discharge.
POLYNUCLEAR AROMATICS (PNA'S) IN KOSOVO STREAM CONDENSATES
The concentrations of four hazardous polynuclear aromatics (PNA's) in
several Kosovo discharge streams are shown in Table 13. Highest concentration
levels were found in the LP coal lock discharge. The higher sensitivity levels
achieved for the naphtha storage tanks were achieved by concentrating the
sample enough to obtain a measurable level of PNA. The concentration of BAP
in the naphtha storage tank was thereby measured at a level of 0.085 yg/m3.
For the remaining streams, PNA concentrations were measured to a sensitivity
level of 0.1 ppm on the extracts as defined by the protocol shown in Figure 3.
The resulting minimum detectable concentrations calculated for each stream are
listed in Table 13.
TABLE 13. CONCENTRATIONS (yg/m3) OF SELECTED HAZARDOUS POLYNUCLEAR
AROMATICS IN KOSOVO GASEOUS DISCHARGE STREAMS
S°Urce BaA BaP dBahA BhF
LP Coal Lock Vent
Ammonia Stripper Vent
Naphtha Storage Tank Vent
Start-up Vent
Tar Tank Vent
Phenolic Water Tank Vent
Medium Oil Tank Vent
H2S-Rich Waste Gas
COa-Rich Waste Gas
163
85
<0.06
-
-
-
-
-
—
670
20
0.085
139
252
<50
<6.5
<0.6
<0.7
52
<2.1
0.06
<2.1
<10
<50
<6.5
<0.6
<0.7
670
12
0.11
-
-
-
-
-
—
All (<) values are calculated from a minimum detectable concentration of
0.1 ppm in the measuring solution.
- not determined.
BaH - Benzo(a)anthracene
BaP - Benzo(a)pyrene
dBahA - dibenz(ah)anthracene
BhF - Benzo(h)Fluoranthene
In a previous report (1) the concentration level of PNA in the Kosovo coal
lock discharge was estimated from the concentration of PNA in by-product oil
using the measured level of tars and oils in the discharge as a base for cal-
culation. The level for BAP in the LP coal lock discharge was estimated to be
between 500 yg/m3 (the oil based value) and 1500 yg/m (the tar based value).
The measured value on the stream is 670 yg/m which is very close to the esti-
mate that was based on medium oil. This agreement indicates that the reason-
able estimates of the levels of PNA in discharge streams can be made with the
combined use of the concentration of tar/oil aerosols in the streams and the
PNA concentration of medium oil.
19
-------
Estimated values that were reported previously for 7,12 Dimethylbenz(a)-
anthracene could not be confirmed because difficulty was experienced in the
determination. Reproducible results could not be obtained.
ENVIRONMENTAL SIGNIFICANCE OF CONDENSABLE ORGANICS IN KOSOVO STREAMS
In a previous work, the environmental significance of each of these
Kosovo streams was determined using the SAM-l/A model to compute the streams'
Total Discharge Severities (TDS). To make these determinations, the concen-
trations of each potential pollutant in the gas phase was divided by its D-MEG
value to obtain component DS values which could be summed to obtain the
streams' TDS value (7).
To include the contributions of the condensable organics in the stream's
TDS determination, it is necessary to adopt a representative D-MEG value with
which to calculate Discharge Severity (DS) values for these heterogeneous
mixtures of organic substances. Use of a representative D-MEG value is the
only reasonable alternative since a rigorous treatment based on individual
component concentrations and D-MEG values is impractical, if not impossible.
A D-MEG (air-health) value of 2.5 x 10H was adopted as being representa-
tive of higher molecular weight components thought to be present in the Kosovo
organic condensates. This value was used to calculate DS values as shown in
Table 14 for condensates in each stream. Also shown are: the previously re-
ported TDS values that are based on the stream's major and minor components,
and the percentages by which each stream's TDS would be increased by the addi-
tion of the DS due to condensable organics. No percentage value exceeded 20%
and several were less than 1%.
TABLE 14. EFFECT OF CONDENSABLE ORGANICS ON TOTAL DISCHARGE SEVERITY
(TDS) OF KOSOVO STREAMS
Source
Fleissner Autoclave Vent
LP Coal Lock Vent
HP Coal Lock Vent
Start-Up Vent
HzS-Rich Waste Gas
C02-Rich Waste Gas
Tar Tank Vent
Medium Oil Tank Vent
Tar Separation Waste Gas
Phenolic Water Tank Vent
Ammonia Stripper Vent
Naphtha Storage Tank Vent
Previously
Reported
TDS
1.00E04
7.88E03
5.92E03
7.19E03
1.11E04
7.32E02
4.31E03
2.82E04
2.06E04
3.67E04
2.07E04
7.08E04
Added
DS*
4.50E01
3.10E02
1.15E02
3.89E02
5.20EOO
5.60E01
5.76E02
8.44E02
1.32E02
5.00E03
3.97E03
3.23E02
%
Increase
<1
4
2
5
<1
<1
13
3
6
14
19
<1
*Based on an average D-MEG (air-health) of 2.5E04
l.OOEOA represents 1.00 x 10"
20
-------
These results indicate that the "condensable" organics are environmentally
significant in all streams except the C02-Rich Waste Gas but, in all cases,
their contributions to the IDS values of these streams are relatively small.
ENVIRONMENTAL SIGNIFICANCE OF HAZARDOUS PNA'S IN KOSOVO STREAMS
DS (air-health) values for the potentially more significant PNA's are
listed in Table 15 along with the percentages by which these DS's would in-
crease the previously reported TDS values of each stream. As with the con-
densable organics, all DS values greater than unity are potentially significant
environmentally. However, significant increases in stream TDS value were found
in only three of the nine streams tested. The greatest increase (430%) was seen
in the LP coal lock vent; followed by the tar tank vent (300%) and the start-
up vent (100%). A most significant observation is that the PNA's contribution
to stream TDS is dominated by the contribution from Benzo(a)pyrene.
TABLE 15. DISCHARGE SEVERITY DUE TO HAZARDOUS PNA'S IN KOSOVO STREAMS
Source
BaA
BaP
dBahA
Total
LP' Coal Lock Vent
Ammonia Stripper Vent
Naphtha Storage Tank Vent
Start-Up Vent
Tar Tank Vent
Phenolic Water Tank Vent
Medium Oil Tank Vent
HzS-Rich Waste Gas
C02-Rich Waste Gas
3.62EOO
1.89EOO
-
-
-
-
-
-
-
3.35E04
1.00E03
4.25EOO
6.95E03
1.26E04
<2E3
<3E2
<3E1
<4E1
5.59E02
2.26E01
<1
<2E1
<1E2
<5E2
<7E1
<6EO
<8EO
3.41E04
1.02E03
4.25EOO
6.97E03
1.27E04
<2.5E3
<3.7E2
<3.6E1
<4.8E1
433
5
nil
97
297
<7
<1
<1
<7
All DS values for Benzo(h)fluoranthene were <1 and are not shown.
ACKNOWLEDGEMENT
This work was sponsored by the Industrial Environmental Research
Laboratory of the United States Environmental Protection Agency. The authors
express their thanks' to the following organizations and individuals for their
contributions to this work: U.S. EPA, T. Kelly Janes and W. J. Rhodes;
Radian Corporation, R. V. Collins and W. E. Corbett; Rudarski Institute,
M. Mitrovic and D. Petkovic; Kosovo Institute, A. Kukaj and M. Milesavljevic;
Elektroprevreda Kosovo, S. Dyla and E. Boti.
21
-------
REFERENCES
1. Bombaugh, K. J., W. E. Corbett, K. W. Lee, and W. S. Seames. An
Environmentally Based Evaluation of the Multimedia Discharges from the
Lurgi Coal Gasification System at Kosovo. In: EPA Symposium on Environ-
mental Aspects of Fuel Conversion Technology V, September 16-19, 1980.
Radian Corporation, Austin, Texas.
2. Lee, K. W., W. S. Seames, R. V. Collins, and K. J. Bombaugh. Environmental
Assessment: Source Test and Evaluation Report—Lurgi (Kosovo, Yugoslavia)
Medium-Btu Gasification. In press. Radian Corporation, Austin, Texas.
March 1981. 354 pp.
3. Lentzen, D. E., D. E. Wagoner, E. D. Estes, and W. F. Gutknecht. IERL-RTP
Procedures Manual: Level 1 Environmental Assessment, (Second Edition).
EPA-600/7-78-201, PB-293 795. Research Triangle Institute, Research
Triangle Park, NC. October 1978. 277 pp.
4. Page, G. C. Environmental Assessment: Source Test and Evaluation Report—
Chapman Low-Btu Gasification. PB-289 940. EPA-600/7-78-202. Radian
Corporation, Austin, Texas. October 1978.
5. Fuchs, M. R., R. A. Magee, and P. M. Jeans. Environmental Assessment:
Source Test and Evaluation Report—Riley Gas Producer, Low-Btu Gasifica-
tion. IERL-RTP Contract No. 68-02-31317. April 1981.
6. Kilpatrick, M. P., R. A. Magee, and T. E. Emmel. Environmental Assess-
ment: Source Test and Evaluation Report—Wellman-Galusha (Ft. Snelling)
Low-Btu Gasification. EPA-600/7-80-097. Radian Corporation, Austin,
Texas. January 1980.
7. Cleland, J. G. and G. L. Kingsburgy, Multimedia Environmental Goals for
Environmental Assessment, Volumes I and II (final report). Research
Triangle Institute. Research Triangle Park, North Carolina. Report Num-
ber EPA-600-7-77-136a, b, NTIS Number PB 276-920 (Volume II). EPA Contract
Number 68-02-2612. November, 1977.
22
-------
APPLICATION OF KOSOVO (LURGI) GASIFICATION t
PLANT TEST RESULTS TO POLLUTION CONTROL PROCESS DESIGN
by: G. C. Page, W. E. Corbett, R. A. Magee
Radian Corporation
8500 Shoal Creek Blvd.
Austin, Texas 78766
ABSTRACT
This paper describes a test program performed by Radian Corporation to
obtain process data to define the pollution control technology requirements
for Lurgi-based coal gasification plants. This program was sponsored by the
Industrial Environmental Research Laboratory (IERL) of the U.S. Environmental
Protection Agency conducted at a Lurgi-based gasification plant in the Kosovo
region of Yugoslavia. It should be emphasized that the Kosovo plant does not
reflect state-of-the-art Lurgi technology especially in pollution control
practices. However, the "uncontrolled" process discharge streams from the
Kosovo plant are representative of those from Lurgi-based gasification
plants.
From an assessment of the Kosovo data, the following discharge streams
were selected to be "key" based on flow rate and/or concentration of pollu-
tants (1) high- and low-pressure coal lock vent gases and ash from the gas
production section, (2) liquid depressurization gases and surge tank vent
gases from the tar/oil separation section, (3) I^S- and C02~rich vent
gases from the Rectisol acid gas removal section, and (4) extracted waste-
water from the Phenosolvan unit.
The conclusions reached from an engineering evaluation of the components
in those key discharge streams and the effects those components may have on
pollution control processes were as follows: (1) pollution control processes
are commercially available for treating these streams, (2) the effects of
minor and trace components on the performance of those control processes have
not been demonstrated and there may be problems in the direct transfer of
technology from other industries (e.g., coke ovens), (3) the design and
selection of pollution control processes during transient and normal
operation should occur in parallel with the base plant design, and (4) the
variability of the components in the discharge streams must be determined and
included in pollution control process design.
23
-------
INTRODUCTION
An international program, sponsored by the Industrial Environmental
Research Laboratory (TERL) of the U.S. Environmental Protection Agency, was
conducted in the Kosovo region of Yugoslavia. The major objective of this
program was to characterize process discharge streams associated with a
Lurgi-based gasification plant and to assess how components in those streams
may affect pollution control process design and operation. The study,
conducted over a three year period, was a cooperative endeavor between
scientists from Yugoslavia, the EPA, and Radian. The program was undertaken
because the Lurgi gasification process has significant potential for use in
the United States.
The purpose of this paper is to address the key process discharge
streams from the Kosovo plant that will require pollution control in a
Lurgi-based plant constructed in the United States. The potential impacts of
specific components in those key uncontrolled discharge streams on the design
and operation of pollution control process are examined.
It should be emphasized that the Kosovo plant does not reflect
state-of-the-art Lurgi technology and that it has essentially no pollution
control processes that would be acceptable for Lurgi-based plants built in
the U.S. However, the "uncontrolled" process discharge streams from the
Kosovo plant contain compounds that will be present in discharge streams from
U.S. Lurgi plants. Many of these species will affect the design and
operation of pollution control processes in the first generation Lurgi-based
gasification plants.
DESCRIPTION OF THE KOSOVO GASIFICATION FACILITY
The Kosovo gasification plant is an integral part of a large mine-mouth
industrial complex located near the city of Pristina in the Kosovo Region of
Southern Yugoslavia. The industrial complex consists of a coal mine, a coal
preparation plant, the gasification plant, an ammonia plant, an air separa-
tion plant, and a steam and power generation plant. The gasification plant
consumes dried lignite and produces two primary products: a fuel gas with a
net heating value of 14 MJ/nP @ 25°C (360 Btu/scf) and hydrogen for ammonia
synthesis. Several hydrocarbon by-products are also produced and are used as
fuel. These by-products include: light tar, medium oil, naphtha, and crude
phenols.
The design flow rates for the major streams in the Kosovo plant are
shown in Figure 1. These flow rates are for five of the six Lurgi gasifiers
in operation. As shown in Figure 1, the plant is designed to produce 24 Mg
(65,000 m3 @ 25°C) of product gas for 80 Mg of dried lignite consumed.
The unit operations employed at the Kosovo gasification plant are
typical of the operations in several proposed U.S. Lurgi-based plants.
Figure 2 is a simplified flow diagram of the unit operations in the Kosovo
facility.
24
-------
Rectisol
Acid Gases
(H2S-Rich and CO2-Rich)
(45)
1
Dried Coal (80)
Steam (65)
02(14)
KOSOVO
Gasification
Plant
T
GasifierAsh
(14)^
1
Waste-
waters
(68)
Heavy Tar
(.5)
Clean
Product
Gas
(25)
Light Tar (2.2)
Medium Oil (1.3)
Naphtha (.7)
Phenols (.4)
Ammonia (1)
Figure 1. Design Flow Rates of Major Streams for the Kosovo (Lurgi)
Gasification Plant (Flow Rates in Megagrams/hr.)
25
-------
WASTE
RASES
FLUE GASES
STEAM
RUN-OF-MINE
COAL
FINES TO
STEAM AND POWER
GENERATION
STEAM
f
COAL
PREPARATION
DRI
SIZED
ED,
COM
GAS
PRODUCTION
WASTEWATER
CRUDE
CAS
RECTISOL
PURIFICATION
H2 to NH3
SYNTHESIS
CLEAN
GAS
GAS
DISTRIBUTION
MEDIUM
*- BTU CAS
TO PIPELINE
NAPHTHA
GAS
LIQUOR
TAR/OIL
SEPARATION
TARS AND
OILS
PHENOLIC
WATER
PHENOSOLVAN
BY-PRODUCT
STORAGE
BY-PRODUCTS
TO STEAM AND
POWER
GENERATION
PHENOLS
- WASTEWATER
Figure 2. Simplified Flow Diagram of the Kosovo Coal Gasification Plant
-------
The following summarizes the function of each unit operation, its
discharge streams, and those discharge streams considered to be key. Key
discharge streams were selected based on two criteria: mass flow rate and/or
high concentrations of pollutants.
COAL PREPARATION
Run-of-mine coal containing approximately 50 weight percent moisture is
dried by the Fleissner process to approximately 25 weight percent moisture.
The dried coal is then crushed and particles between 6 and 60 mm are stored
in the gasifier coal feed hopper. Table 1 shows a typical composition of the
dried coal feedstock. Coal fines (less than 6 mm) are routed to the on-site
steam and power plant.
Although the Fleissner process is used in foreign countries, the process
is not likely to be used in U.S. Lurgi plants. Therefore, the discharge
streams from this process were not considered "key" with respect to the
evaluation of pollution control requirements for U.S. facilities.
GAS PRODUCTION
In the gas production section, dried coal is reacted with steam and
oxygen at a pressure of 25 atm to produce a crude product gas. This reaction
occurs in the gasifiers (3.5m in diameter) used at Kosovo. The crude product
gas is then cooled and scrubbed to remove coal fines, tars, oils, and other
condensibles. Ash produced during gasification is collected from the bottom
of each gasifier in a lock hopper. The ash is water-quenched and routed to a
landfill.
The discharge streams from the gas production section are: coal room
vent, high-pressure coal lock vent, low-pressure coal lock vent, start-up
vent, gas liquor tank vent, ash lock vent, gasifier ash, and ash quench
water. Of these discharge streams, the following are considered to be key
streams:
o low-pressure coal lock vent,
o high-pressure coal lock vent, and
o gasifier ash.
TAR/OIL SEPARATION
In the tar/oil separation section, heavy tar, light tar, and medium oil
are separated from the crude gas quench liquor and from the condensates
produced by cooling the crude gas. Depressurization of quench liquor and
condensates followed by a series of phase separators are used to accomplish
this. Light tar and medium oil are sent to by-product storage to be used as
boiler fuel. Heavy tar is landfilled at Kosovo; however, current plans for
U.S. plants are to recycle this tar to the gasifier or to use it as boiler
fuel. The aqueous phase from the separators is routed to Phenosolvan for
phenol recovery.
27
-------
TABLE 1. TYPICAL COMPOSITION OF THE DRIED COAL
FEEDSTOCK FOR THE KOSOVO PLANT
Ultimate Analysis (wt. %)
Moisture 20
Ash 14
Carbon 45
Sulfur 0.89
Hydrogen 3.5
Nitrogen 1.1
Oxygen 16
Heating Values (kcal/kg)
Proximate HHV 3900
Proximate LHV 3700
28
-------
The discharge streams from this unit operation are: depressurization
gases, surge tank vents, phenolic water, and heavy tar. The key discharge
streams include depressurization gases and surge tank vents.
ACID GAS REMOVAL (RECTISOL)
Acid gases, such as ^S, COS, C02, HCN, and mercaptans are selec-
tively removed from the cooled product gas by sorption in cold methanol. The
product gas enters the Rectisol process at about 22°C and 23 atm. The gas is
cooled by a cold water wash followed by a cold methanol wash. Condensates
from this initial cooling are a light organic phase (naphtha) and aqueous
phase. Naphtha is sent to by-product storage to be used as fuel while the
aqueous phase is routed to tar/oil separation. After the initial gas cooling
step, the gas is scrubbed with cold methanol in the ^S absorber. The
H2S-lean product gas is then sent to C02 absorption for final purifi-
cation. During methanol regeneration, the acid gases removed from the
product gas stream are stripped from the methanol which results in two waste
gas streams H2S- and C02~rich vent gases.
Discharge streams from the acid gas removal unit operation are:
o H2S-rich vent,
o C02~rich vent,
o naphtha, and
o aqueous condensate.
Of these streams, the l^S- and C02~rich vent gases are considered key.
PHENOL REMOVAL (PHENOSOLVAN)
In the phenol removal section, phenolic water from the tar/oil separa-
tion section is extracted with diisopropylether (DIPE) to remove phenolic
compounds. To accomplish this, residual tars and oils are removed by phase
separation and filtration, followed by removal of dissolved gases by steam
stripping. After dissolved gas removal, the water is extracted with DIPE to
remove phenolic compounds. Recovered phenol is sent to by-product storage to
be used as a fuel. The extracted water is discharged.
The discharge streams from the Phenosolvan section include:
o stripped gases,
o surge tank vents,
o by-product phenol, and
o wastewater.
The key discharge stream from this unit operation is wastewater from DIPE
extraction.
KEY DISCHARGE STREAMS
As discussed above the key discharge streams from the Kosovo plant that
are of primary concern for proposed Lurgi-based plants in the U.S. are:
29
-------
o Gas production
- High-pressure coal lock vent
- Low-pressure coal lock vent
- Gasifier ash
o Tar/Oil Separation
- Depressurization gases
- Surge tank vents
o Acid Gas Removal (Rectisol)
- H2S-rich vent gas
- C02~rich vent gas
o Phenol Recovery (Phenosolvan)
- Extracted wastewater
These streams were selected as key discharge streams based on their flow
rates and/or their concentration of pollutants. Tables 2 and 3 summarize the
flow rates and compositions of these streams. The data in these tables are a
portion of the results from the environmental test program performed at the
Kosovo plant.
The following text contains a discussion of these key discharge streams
which emphasize the components in each stream that will affect the design and
operation of processes used for pollution control in U.S. Lurgi-based plants.
The use of the hydrocarbon by-products as fuel and/or as feedstocks for
petrochemical manufacturing is also addressed.
GAS PRODUCTION
High-Pressure Coal Lock Vent
The flow rate of the high-pressure coal lock vent stream is approxi-
mately 2 percent of the crude product gas flow rate. This stream will
contain all of the compounds found in the crude product gas exiting the
gasifier including coal fines, tars, oils, reduced sulfur compounds, HCN,
NH3, CO, H2, etc.
Because of the high flow rate and energy content of this stream, viable
control alternatives include recycle to the product gas or use as a fuel.
For recycle to the product gas, entrained particulate matter and tar/oil
aerosols need to be removed prior to compression to product gas pressure.
Particulate and aerosol removal also will be necessary prior to combustion to
minimize equipment fouling and buildup of tar in the gas lines. Flue gas
control of gases resulting from the combustion of the high-pressure vent
stream also will be necessary because of the high levels of sulfur compounds
in this stream.
Low-pressure Coal Lock Vent
The flow rate of the low-pressure coal lock vent gas is low (less than
0.2 percent of the crude product gas flow), however this gas will contain
30
-------
TABLE 2. KEY GASEOUS DISCHARGE STREAM COMPOSITIONS
Stream Parameter
Dry Gas Flow Rate
(m3/gaslfier-hr S25°C>
Moisture Content (wt Z)
Dry Gas Molecular Weight
Dry Gas Composition
Fixed Gases (Vol Z)
H2
°2
N2
CH4
CO
C02
Sulfur Species (ppmv)
H2S
COS
Methyl Mercaptan
Ethyl Mercaptan
Hydrocarbons (vol Z)
C2»6
C2H4
C3's
C4's
C5's
C6+'s
Light Aromatlcs (ppmv)
Benzene
Toluene
Xylene, Ethylbenzene
Phenols
Higher Aromatlcs
Nitrogen Species (ppmv)
NH3
HCN
Partlculate Matter
(mg/m3 9 25"C)
Total Partlculates
Organlcs Contained in Partlculate
Low-PreMure
Coal Lock Vent
21**
44
23.5
37
0.27
0.18
8.6
14.6
36.5
13,000
110
420
220
0.22
<0.01
0.14
0.05
<0.01
D.,12
760
220
75
5.7
NF
2400
600
8100
Matter* 7300
High-Pressure
Coal Lock Vent
(after scrubbing)
230**
11
24.9
32
0.24
0.14
10.5
12
42
3500
120
460
210
0.42
<0.01
0.25
O.U
0.01
0.08
550
100
38
2.5
NF
NF
170
960
660
Tar/011 Separation
Depreaauricatlon
Gases
28**
7.7
39.0
11
<0.01
<0.01
3.5
1.1
77.5
9000
120
2500
1600
0.33
<0.01
0.41
0.41
0.09
1.3
9600
1200
150
4.2
4.9
19,300
64
920
660
Medium Oil
Surge Tank
Vent
1.7**
8.4
32.5
<0.01
0.45
1.1
7.6
5.9
56
26,000
96
5200
2100
0.34
<0.01
0.30
0.25
0.09
2.4
7650
1400
140
110
NF
19
57
NS
US
H2S-Rlch
Tall Gas
3600***
3.9
43.0
0.11
<0.01
<0.01
4.3
1.1
88
45,400
420
2100
780
0.82
<0.01
0.63
0.32
0.04
0.21
110
8
NF
< 1
NF
2200
200
NS
NS
C02-Rlch
Tall Gas
3600***
5.1
42.2
<0.01
<0.01
<0.01
1.2
<0.01
94
39
62
8.5
4.4
1.60
<0.01
0.28
<0.01
<0.01
NF
1.0
< 1
< 1
NF
NF
4.6
13
NS
NS
*0rganics include tars and oils which contain significant amount0 of polynuclar aromatics as found in the Koaovo tar and
medium oil by-products.
**Measured flow rates
***Deslgn flow rates
NF: Not found
NS: Sample not obtained during the test program
31
-------
TABLE 3. KEY LIQUID AND SOLID DISCHARGE STREAM COMPOSITIONS
Key Liquid S treat
Design FLow Rate
(m3/ga8lfler-hr)
PH
Solids Analysis (mg/L)
Total Solids
Suspended Solids
Dissolved Solids
Water Quality Parameters
COD (aa mg 02/L)
Permanganate (mg/L)
BOD5 (as mg 02/L)
Aqueous Composition Data (mg/L)
TOC
Total Phenols
Volatile Phenols
Free Ammonia
Fixed Ammonia
Cyanide
Nitrites
Nitrates
Pyrldlnes
Chlorides
Fluorides
Total Sulfur
Sulfltes
Sulfates
Sulfldes
Thlocyanates
Thlosulfates
PNA Analysis (mg/L)
fienz(a)anthracene
7,12-dimethylbenz(a)anthracene
Benzo(a)f luoranthrene
Benzo(a)pyrene
3-methylcholanthrene
Dlbenz(a,h)anthr scene
252 Group (as BaP)
Extracted
from Phenosolvan
13
9.6
1,350
1,160
190
7,910
4,040
2,350
1,470
230
130
Tr
205
0.019
Tr
11.4
—
60
Tr
84
110
-
<75
Tr
NF
NF
NF
NF
NF
NF
0.19
Key Solid
Design Flow Rate
(Mg/gaslfier-hr)
Ultimate Analysis (wt 1)
Moisture
Ash
Volatile
Fixed Carbon
Csrbon Dioxide
Total Sulfur
Free Sulfur
Fixed Sulfur
Hydrogen
Nitrogen
Oxygen (By Difference)
Chlorine
Proximate Analysis (wt Z)
Moisture
Ash
Volatlles
Fixed Carbon
Total Sulfur
Trace Elements (mg/kg) by AA
As
Be
Cd
Co
Cr
Cu
Hg
Mo
Nl
Pb
Sb
Se
Sr
Tl
V
Zn
Streu
Gasifler Ash
before Quenching
2.8
2.1
94
-
1.7
0.15
0.25
0.03
2.3
0.04
2.1
94
6.5
-
0.15
75
2.5
69
17
180
40
0.30
8.9
320
52
NF
24
370
NF
100
2.1
Tr • Trace
NF - Not Found
- ~ Not Analyzed
AA: Atomic Adsorption Spectroscopy
32
-------
pollutants found in the crude product gas (e.g., tars, oils, reduced sulfur
compounds, HCN, CO, etc.). Even if these gases are diluted with air, they
still will contain significant levels of tar/oil aerosols and reduced sulfur
species and, therefore, should not be directly vented to the atmosphere.
A viable control of the low-pressure coal lock stream involves first
minimizing its flow rate by controlling the pressure at which the low-
pressure lock vent is opened (approximately 2 atm) followed by particulate
and aerosol collection and then combustion (flaring). The major concern in
controlling this stream is the pressure drop required to remove particulates
and aerosols. To attain the required pressure for this, a blower may have to
be installed in the vent line. If a blower is required, an explosive gas
mixture may result due to the influx of air. Precautions must be taken to
eliminate exposing the gas mixture to ignition sources.
Gasifier Ash
The gasifier ash is a key waste stream because of its high flow rate and
the potential for the leaching of trace elements contained in the ash.
However, leaching tests (RCRA and ASTM, Ref. 1 and 2) performed on the
unquenched gasifier ash from Kosovo and on ashes from other gasifier
processes show that the concentration of trace metals in the leachates are
well below RCRA limits for hazardous wastes. Table 4 shows the results of
the RCRA and the ASTM leaching tests on the gasifier ash from the Kosovo
plant. As shown in this table, the trace element concentration in the
leachates was between 10 and 1000 times lower than the RCRA limits for
hazardous wastes.
There are two disposal aternatives for gasifier ash resource recovery or
disposal in a nonhazardous waste landfill. It should be emphasized that the
leaching tests were conducted on unquenched ash and the leaching properties
of the ash could be significantly different if untreated process wastewaters
were used to quench the ash. If process water is used for ash quench, the
water should be treated to remove any toxic organics or trace elements that
may render the ash to be classified as hazardous under RCRA.
TAR/OIL SEPARATION
Liquid Depressurization Gases and Surge Tank Vents
The combined flow rate of depressurization gases and surge tank vents is
less than one percent of the crude product gas flow rate. However, these
streams contain high levels of pollutants (e.g., H2S, COS, mercaptans, HCN,
and nonmethane hydrocarbons) which must be controlled.
Two viable control alternatives for these gaseous streams are.
o containment and collection followed by combustion
with flue gas controls and
o containment and collection followed by the sulfur
recovery process used to treat the H2S-rich gas from
Rectisol.
33
-------
TABLE 4. GASIFIER ASH TRACE ELEMENT LEACHING RESULTS
Leachate Concentration/RCRA Limits
Element RCRA Leachate Neutral Leachate
As
Ba
Cd
Cr
Pb
Hg
Se
Ag
0.001
0.03
<0.001
0.06
0.002
<0.005
0.01
<0.0002
0.002
0.001
<0.004
0.10
0.01
<0.020
0.007
<0.0008
34
-------
Use of the first control alternative may be complicated by the pressure
differences of the gas streams and the physical location of vents in the
plant. Routing these gases to sulfur recovery also may cause operational
problems in the sulfur recovery process due to the presence of reduced sulfur
compounds (COS, mercaptans), HCN, and hydrocarbons. These potential problems
are discussed in the control alternatives for the H2S-rich vent gas from
Rectisol.
ACID GAS REMOVAL (RECTISOL)
H2S-Rich Vent Gas
The H2S-rich vent gas from the Rectisol acid gas removal process has a
high flow rate (approximately 30 percent of the clean product gas flow rate)
and contains pollutants. Table 5 summarizes the compounds of concern in this
stream. As shown in this table, the ^S-rich gas contains significant
levels of H2S, mercaptans, COS, HCN, and nonmethane hydrocarbons.
Of the many viable methods to control this stream, two methods were
selected for discussion in this paper:
o H2S concentration (e.g., routing the gas to an
amine-based acid gas removal process to concentrate
the H2$ to approximately 10 to 15 volume percent
in the gas) followed by sulfur recovery using a
Glaus with a Glaus tail gas cleanup process, or
o H2S removal using a Stretford process followed by
tail gas combustion.
If the first control alternative is used, problems may be caused be the
production of organic sulfur compounds in the Glaus process because of the
high levels of hydrocarbons in the gas (ref. 3). Also, the effectiveness of
the Glaus and the Glaus tail gas treatment processes in removing high levels
of reduced sulfur and nitrogen compounds has not been demonstrated.
The Stretford process will not remove COS from the stream and HCN in the
stream will cause formation of nonregenerable compounds in the Stretford
solution (ref. 4). The fate of mercaptans in the Stretford process is
uncertain (e.g., the removal of mercaptans by the Stretford solution and how
mercaptans distribute between the Stretford tail gas and the oxidizer vent
gas). Another concern with the Stretford process is the effect of contami-
nants in the gas on the purity of the by-product sulfur (e.g., organics in
the gas may end up in the by-product sulfur) thus reducing the marketability
of the sulfur. ^
C0?-Rich Vent Gas
The C02~rich vent gas from the Rectisol process is a high volume
stream (approximately 30 percent of the clean product gas flow) and contains
pollutants as shown in Table 6. In certain designs of Lurgi-based plants,
the C02~rich vent stream is vented directly to the atmosphere. However,
35
-------
TABLE 5. POLLUTANTS IN THE H2S-RICH GAS FROM
RECTISOL TO BE CONTROLLED
Component Concentration
Major (vol. %)
H2S
CH4
Nonmethane Hydrocarbons
CO
Minor (ppmv)
COS
Mercaptans
NH3
HCN
2-5
4
2
1
400
3,000
2,000
200
TABLE 6. POLLUTANTS IN THE C02-RICH GAS FROM
RECTISOL TO BE CONTROLLED
Component Concentration
Major (vol. %)
Nonmethane Hydrocarbons 2
Minor (ppmv)
COS 50
Mercaptans 10
NH3 5
HCN 10
H2S 40
36
-------
based on the results obtained at Kosovo, the amount of nonmethane hydro-
carbons and mercaptans in the C02~rich stream may not allow direct
discharge of this stream.
Venting of the C02~rich tail gas directly to the atmosphere would
involve operating the Rectisol process such that the levels of nonmethane
hydrocarbons and reduced sulfur compounds are significantly lower than levels
found at Kosovo. If this cannot be accomplished, a nonselective Rectisol
process configuration may be used which has only one vent gas containing
components in both the H2S- and C02~rich vent gases. Viable control
alternatives for this stream would be similar to those for the f^S-rich
vent gas.
PHENOL RECOVERY (PHENOSOLVAN)
Extracted Wastewater
The extracted wastewater from the Phenosolvan process has a flow rate of
approximately 0.8 kg of wastewater per kg of coal fed to the gasifier and
contains significant levels of pollutants. These compounds include phenols,
organic acids, refractory organics, cyanides, and ammonia.
A viable control alternative for this wastewater is removal of organics
by biological oxidation followed by trace organic removal by carbon
adsorption and possibly incineration. Biological oxidation and carbon
adsorption will be effective in removing the major portion of the organics;
however, removal of high molecular weight "refractory" organics by these
processes has not been demonstrated.
Direct transfer of biological oxidation and carbon adsorption
technologies from other industries (e.g., coke ovens) may not be entirely
applicable. This is illustrated by comparing the organic fractions in
wastewaters from the Kosovo plant, the Chapman low-Btu gasification plant,
and a coke oven plant. Figure 3 shows the relative amounts of total organic
carbon (TOG) remaining in these wastewaters after extraction with DIPE to
simulate the phenosolvan process and after analytical extraction with
methylene chloride. Both of the gasification wastewaters contain signifi-
cantly higher amounts of nonextractable (refractory) organics than the coke
oven wastewater. These organics are very polar and/or ionic in nature since
extraction (including pH adjustment) would not remove them.
Some preliminary data on carbon adsorption of wastewaters from the
Chapman low-Btu gasification plant indicate that carbon adsorption cannot
reduce the level of total organic carbon (TOC) below 150 to 200 ppm (ref. 5).
Therefore, if significant levels of organics remain in the wastewater after
biological treatment and carbon adsorption, incineration of the wastewater to
destroy these organics may be necessary. It should be emphasized that
incineration is necessary if these organics are toxic. At this time there
are no data concerning the toxicity of these organic constituents.
37
-------
W///A Nonertractable
••H Analytical Ertractable
^^ DIPE Ertractable
10,000
o
E
o
o
I-
5,000
Lurgi Chapman
•5-iyiel>.w
f&i's&zs
Coke
Oven
Figure 3. Extractable and Nonextractable Organics in Gasification and
Coke Oven Wastewater (Ref. 5)
38
-------
By-Products; Tar, Oil, Naphtha and Phenol
The composition of the by-products (tar, medium oil, naphtha and phenol)
will affect their end use. Chemical analysis data for Kosovo by-products are
shown in Table 7. Analysis of the phenol is not reported because a sample
was not obtained during the test. Table 7 indicates that the sulfur contents
of the liquid by-products become progressively higher in the "lighter"
fractions. In contrast, the trend in the nitrogen values is reversed. These
data indicate that heavy hydrocarbon by-products similar to those generated
at Kosovo, could be used to satisfy some of the on-site fuel needs (e.g., for
steam generation) of a U.S. Lurgi plant without flue gas desulfurization.
This assumes that current SC>2 emissions standards consistent with those for
large fossil fuel fired steam generators are applicable.
Table 7 also shows that the naphtha by-product contained none of the
heavy polynuclear aromatlcs (PNA's) found in the tar and medium oil. Use of
the tar and oil as fuel would destroy those PNA's by combustion. However,
the naphtha could be used as a chemical feedstock with minimal risk of worker
exposure to heavy PNA's.
CONCLUSIONS AND RECOMMENDATIONS
The conclusions and recommendations derived from the results of the
Kosovo test program and from an assessment of the pollution control processes
to treat key discharge streams from a Lurgi-based plant are presented in four
areas:
o availability of controls,
o applicability of controls,
o integrated plant concerns, and
o discharge stream variability
AVAILABILITY OF CONTROLS
Pollution control processes for all of the key discharge streams are
commercially available. Most of these processes have been proven in related
industries (e.g., coke oven, refinery, etc.).
APPLICABILITY OF CONTROLS
Even though pollution control processes are commercially available,
minor and trace components in discharge streams from gasification plants will
affect the operation of those processes. Because many of these components
are unique to gasification technology, direct transfer of pollution control
process design from other industries may not be applicable.
INTEGRATED PLANT CONCERNS
Pollution control costs will be a significant portion of the base plant
cost. The control of discharge streams by recycling or process modification
should be considered in the total plant design, especially if significant
cost savings for pollution control can be realized.
39
-------
TABLE 7. CHEMICAL AND PHYSICAL DATA FOR KOSOVO BY-PRODUCTS
By-Product
Specific Gravity
(g/cm3)
Higher Heating Value
(kcal/kg)
Lower Heating Value
(kcal/kg)
Ultimate Analysis (wt. %)
Carbon
Hydrogen
Nitrogen
Sulfur
Ash
Oxygen (difference)
Moisture Content (wt. %)
PNA Analysis (mg/kg)
B e nz ( a ) a nt hr ac e ne
7, 12-dimethylbenz(a)anthracene
Benzo(b)f luoranthrene
Benzo(a)pyrene
3-methylcholanthrene
Dibenz(a,h)anthracene
252 Group (as BaP)
Light
Tar
1.06
8910
8280
82
8.4
1.3
0.49
0.22
7.8
1.1
490
1100
310
210
26
23
950
Medium
Oil
0.97
9500
9400
82
8.9
1.00
0.83
0.03
8.2
0.8
160
62
120
68
NF
6.6
280
Naphtha
0.85
9940
8925
86
9.9
0.18
2.2
2.2
NF
NF
NF
NF
NF
NF
NF
NF = not found.
= no data available.
40
-------
DISCHARGE STREAM VARIABILITY
Every Lurgi-based plant will have unique discharge stream character-
istics resulting from differences in coal feedstock properties, process
operation, and plant configuration. Pollution control processes must be
designed to handle worst case transients as well as "normal" operating
conditions. Therefore, during the testing phase to obtain operating
parameters on a specific coal feedstock, characterization of discharge
streams during steady-state and transient conditions should be performed to
define the bases for pollution control process design.
As stated previously, pollution control technology is available for
Lurgi-based gasification plants, but the application of this technology is
not completely straightforward. For this reason, pollution control process
designs must be based on design data obtained over a broad spectrum of
operating conditions. It is particularly important that data be obtained
under both transient as well as steady-state conditions.
ACKNOWLEDGEMENT
This work was sponsored by the Industrial Environmental Research
Laboratory of the United States Environmental Protection Agency. The authors
express their thanks to the following organizations and individuals for their
contributions to this work:
U.S. EPA - T. Kelly Janes, W. J. Rhodes
Radian Corporation - K. J. Bombaugh, K. W. Lee
Rudarski Institute - M. Mitrovic, D. Petkovic
Kosovo Institute - B. Shalja, A. Kukaj, M. Milesavljevic
REMHK Kosovo - S. Dyla, E. Boti
INEP - S. Kapor
REFERENCES CITED
1. Federal Register Vol. 46, No. 24, Proposed Rules for
Land Disposal Facilities, (Feb. 5, 1981).
2. Bombaugh, K.J., et al. "An Environmental Based
Evaluation of the Multimedia Discharges from the Kosovo
Lurgi Coal Gasification System" presented at the 5th
IERL/EPA Symposium on Environmental Aspects of Fuel
Conversion Technology, St. Louis, MO (Sept. 1980).
3. Goar, G. "Impure Feeds Cause Glaus Plant Problems"
Hydrocarbon Processing. 53(7), 129-32 (1974).
4. Cavanaugh, E.G. , et al. Environmental Assessment Data
Base for Low/Medium-Btu Gasification Technology. Vol. I
and II, EPA-600/7-77-125a and b (Nov. 1977).
5 Collins, R.V. , et al. "Comparison of Coal Conversion
Wastewaters" presented at the 5th IERL/EPA Symposium on
Environmental Aspects of Fuel Conversion Technology,
St. Louis, MO, (Sept. 1970)
41
-------
ENVIRONMENTAL ASPECTS OF THE GKT COAL GASIFICATION PROCESS!
R.E. Wetzel
GKT Gesellschaft fur Kohle-Technologie mbH
29 Moltkestralie
43 Essen, FRG
K.W. Crawford
TRW Systems and Energy
One Space Park
Redondo Beach, CA 9o278
W.C. Yee
Tennessee Valley Authority
3oo Corroon and Black Building
Chattanooga, Tennessee 37 H-ol
INTRODUCTION
Thermal Conversion of coal is always accompanied by the pro-
duction of pyrolysis products, of solid wastes like ash, slag or
flydust, of different sulfur compounds and a number of undesired
trace compounds in the gas. Quality and quantity of these compo-
nents depend on the applied process principle, the composition
of the reactants and on the main process parameters like tempe-
rature and pressure. The high temperature entrained coal gasifi-
cation process according to GKT, a well established commercial
process since 3o years, offers a coal conversion system with
minimum environmental impact.
To evaluate the operating conditions and the environmental im-
pact when gasifying American coal, TVA and GKT agreed to conduct
a large scale test with about 5ooo short tons of Illinois No. 6
coal in a commercial coal to ammonia plant in Greece. TRW, fun-
ded by TVA, cooperated in the extensive test programme regarding
the environmental aspects. The favourable results of the test
runs, carried out in March/April 1981, and the extraordinary low
42
-------
environmental impact demonstrated, have led to TVA's decision to
built its Murphy Hill plant basing on GKT's technology.
THE GKT PROCESS
In 1936/42 Friedrich Totzek and his coworkers developed a
new gasification principle, the gasification of pulverized coal
in an entrained bed reactor, using oxygen and steam as gasifica-
tion media.
In this GKT PROCESS coal dust and oxygen are reacted within
one second in a flame reaction to carbon monoxide and hydrogen,
at close to atmospheric pressure and temperatures of 14oo to
16oo C. Under these conditions, the coal ash converts to liquid
slag and flyash.
1 Coal Dust
2 Oxygen
3 Steam
4 Raw gas
5 Liquid Slag
6 Granulated Slag
Figure 1 The GKT gasifier
The reaction vessel is shown in Figure 1. Coal dust metered
via dosing screw conveyors, is injected with oxygen and steam
through opposite burners. Liquid slag accumulating at the reac-
tor wall, flows down through the bottom opening into a water
bath, granulates and is discharged via chain conveyor. The reac-
tor wall is cooled by raising steam in the double-wall jacket.
The product gas exiting at the top is quenched to about 1000 C
by water injection in order to solidify slag droplets before
entering the waste heat boiler.
The reactions of the coal with the gasifying agents are sum-
43
-------
marized in Table 1. Most important are Reactions (1) to (3)
which lead to a product gas with approximately 90 % 'CO and
at a ratio of 2/1 to 2.5/1. Formation of methane, Reaction (
is of no importance at the high gasification temperatures. How
ever, traces of methane are always present in the gas.
TABLE 1 REACTIONS OF THE SYSTEM C/H/0/N/S
Gasification reactions:
C +02 = C02 (1)
C + CO2 - 2 CO (2)
C + H2O - CO + H2 (3)
C + 2 H2 = CH4 (4)
Side reactions:
C +2S =CS2 (5)
CO +S =COS (6)
H2 +S =H2S (7)
2CO2+S =SO2+2CO (8)
3H2 +N2 =2NH3 (9)
C + NH3 = HCN + H2 (10)
O2 +N2 =2 NO (11)
The sulphur contained in the coal in form of organic and
inorganic compounds, is almost completely converted to H?S and
COS at a molar ratio of 9/1. Further reactions of the system
C/H/O/S permit the formation of traces of CS~, S0~ and elemental
sulfur, reactions (5) to (8). The nitrogen in the coal and the
nitrogen content of the gasification oxygen lead by a number of
side reactions to the formation of traces of HCN, NH., and NO,
reactions (9) to (11). J
The reactions shown in Table 1 represent the minimum number
which describe the system. Their combinations result in a large
number of heterogeneous and homogeneous reactions occuring simul-
taneously .
Besides these reactions, also coal ash components will
react under the prevailing conditions.
The coal ash contains practically all chemical elements in
small traces, similar to all natural ores and minerals used for
industrial purposes. Under gasification conditions accordingly
many trace element side reactions as reduction reactions, vola-
44
-------
tilization reactions, so-called transport reactions and conden-
sation reactions are experienced.
SYNTHESIS GAS VIA THE GKT PROCESS
The flow diagram of a GKT coal gasification plant for the
production of synthesis gas is shown in Figure 2. All important
streams entering and leaving the plant are marked.
<=>
1 Drying Mill
2 Gasifier
3 Waste Heat Boiler
4 Cooling/Dedusting
5 Compressor
6 Desulphurization
7 CO Conversion
8 CO2 Removal
9 Slag Extractor
lOCIarifier
11 Cooler
12 Settling Pond
—Water Circuit
Figure 2 Flowsheet of a synthesis gas from coal plant
The raw coal entering the plant is crushed and pulverized
to_ the^necessary particle size ( <• o. 1 mm) and is simultaneously
dried in a mill. The coal dust is pneumatically conveyed to the
feed bunkers of the individual gasifiers. With oxygen and steam
admixture the coal dust is injected into the gasifiers and gasi-
fied autothermally.
The slag leaving the reactor at the bottom is quenched in
the water seal of the gasifier and discharged as a granulate.
The raw gas after quenching is cooled to about 300°C in a waste
heat boiler, where saturated steam at loo bar is raised.
Cooling of the raw gas to ambient temperature and compres-
sor grade dedusting is attained in a series of steps comprising
a washer cooler, a disintegrator stage and a wet electrostatic
precipitator. The flyash and the sensible heat are absorbed by
the wash water which is recycled via a clarifier and a cooling
stage, The flyash is pumped as a slurry from the clarifier to a
p oiid •
45
-------
The cooled and dedusted raw gas is compressed to 2o-6o bar
for desulphurization by a chemical or physical wash, a CO shift
conversion and a C0~ removal. According to the application of
the synthesis gas the sulfur free gas is partly or totally ente-
ring the CO shift reactor and the C02~removal stage.
THE TEST UNIT
For the large-scale test operation the fertilizer plant of
"Nitrogenous Fertilizer Industry S.A." in Ptolemais, Greece, was
chosen. This plant started operation in 1963. Extentions increa-
sed production in 1971 and 1973. The coal gasification section
of this plant is shown schematically in Figure 3.
Grinding
Unit
Gasifier Waste Heat Cooling- Disintegrator
Boiler Washer
6
r
i
1
1
1
6
Side
Stream
Unit
Figure 3 The Test unit at the NFI plant
Three grinding units and six gasifiers are available. For
the test operation
the drying and grinding unit no. 3 and
the gasifier no. 6
of the plant were required.
From Figure 3 the limitations of the system are apparent.
The raw gas produced in the test gasifier is mixed with the gas
produced in the other gasifiers before entering the second
washing stage. The washing water entering the cooling washer is
recycle water from the operation of the total gasification and
46
-------
gas cleaning section of the plant.
For the determination of the chemical composition of the
washing water in a scrubbing system not affected by the parallel
operation of the gasifiers 1 to 5 with Greek lignite a wet gas
cleaning side stream unit including
a washer cooler
a Theisen disintegrator
a drop separator and
a closed water circuit with clarifier and cooler
was installed, Figure 4.
Cooling Water
<=>
Cooling Water
•<=>
Raw Gas
O—
from WHB
Figure 4 The side stream gas cleaning unit
TEST RESULTS AND DISCUSSION
The following discussion will focus on the streams which are
shown in Figure 2.
Especially
the coal feedstock
the slag
the flydust
the wash water from the washing unit and
the raw synthesis gas before entering the compression stage
will be discussed.
47
-------
Further streams, as
the final synthesis gas
the sulfur rich Glaus gas and
C02~rich off gas from the CO^-removal stage
will be characterized.
TABLE 2 TEST GASIFICATION DATA
Operating Conditions
Illinois No. 6 coal
O2/Coal (maf)1 '-ratio
Steam/Coal (maf)-ratio
Operating Results
6.300 kg/h
1.08 kg/kg (maf)
0.11 kg/kg (maf)
Raw gas (dry) 10,900 m^/h
(CO + HJ/coaKmafJ-ratio 1.73 rn^/kg (maf)
Coal conversion: t)C 94.6%
Efficiency: fyh.ctom.1 77.8%
1.) maf = moisture and ash free
2.) excluding steam production in waste heat boiler
For_orientation operational conditions and results are sum-
marized in Table 2. The flow rates of all product and discharge
streams are presented in Table 3. All the results presented are
obtained at these test conditions.
The composition of the coal gasified is given in Table 4.
The Illinois no. 6 coal has a high sulfur content, a medium ash
content and a low chlorine content of o.o5 to o.o7 %. The
moisture content of the coal dust prepared for gasification is
1 %. The solid by-products, slag and flyash comprise the coal
ash and the unconverted coal, their composition is shown in
Table 5.
The slag, recovered from the gasifier in granular form is
environmentally harmless as it has been fused at high tempera-
tures. According to the high SiO -content it has a vitreous and
dense structure. Traces of dissolved carbon are responsible for
its black colour. Leaching tests according to the RCRA procedure
demonstrate, that practically no priority pollutants are develo-
48
-------
TABLE 3 SPECIFIC FLOW RATES
Flow Rates of Product and Discharge Streams
Basis: 11 Illinois No. 6 Coal (moisture free):
Synthesis Gas 1730m3/t
Glaus Feed Gas 71 rrin/t
COj-Off Gas 831 mS/t
Discharge Water ' > 1.3 m3/t
Flydust(dry) 120 kg/t
Slag 46 kg/t
1' Side Stream Unit only.
Commercial Operation ~ 0.5 m3/t
TABLE
COAL ANALYSIS
Ultimate Analysis,
wt.-% mf
H 4.6
C 69.6
^combustible ^-"
N 1.4
Ash 12.7
O(by difference 1 9.1
Total Sulphur 2.7
Ash Analysis,
wt.-%
FeA
Si02
AIA
CaO
MgO
Na^
K^O
TiO2
PA
S03
17.0
49.8
21.5
3.2
1.2
0.4
2.4
1.0
0.1
3.4
Amax Delta Mine, Illinois No. 6 Coal
49
-------
TABLE 5 SLAG AND FLYDUST ANALYSES
Component
C
Ash
S
Ash: SiO2
AI2O3 + TiO2
Fed
Fe203
CaO
MgO
Na2O+K2O
P205
SO3
Concentration in wt.-% (dry)
Slag
0.2
99.5
0.3
50.1
25.5
10.8
4.5
3.9
1.2
2.5
01
Solids in Wash Water
32.9
66.3
1.3
49.3
23.5
16.8
3.6
1.2
3.0
0.1
1.9
TABLE 6 SLAG LEACHING TEST
Element
Ag
As
Ba
Cd
Cr
Hg
Pb
Se
NH3
SCN-
COD
Concentration in mg/kg
RCRA
Extract
PH5
<0.01
<0.4
<0.01
<0.007
<0.04
<0.0002
<0.05
<0.4
Neutral
Extract
PH7
<0.01
<0.4
<0.01
<0.007
<0.04
<0.0002
<0.05
<04
0.7
<0.1
4.8
RCRA
Standard
5
5
100
1
5
0.2
5
1
None11
None 'i
None"
1 ' No Standard Available
50
-------
ped, Table 6. Values below the analytical detection limit are
marked by the sign "smaller than" (< ) in Table 6 and the follow-
ing relevant Tables. Traces of NH., result from the adsorbed
quench water. Limited washing during extraction by chain convey-
ers will reduce this contaminant further. The use of the slag for
road construction or other applications is therefore principally
possible.
The flyash entrained in the raw synthesis gas is removed in
the wet washing and cooling stage of the 'process. The fine grain
size, the unconverted carbon content and the water content from
the wet cleaning operation hinder an economic application. There-
fore it has to be deposited. The exposition to the high tempera-
ture of the gasification results in an inert material. The carbon
content is mainly graphitized, volatiles are limited. A hydrogen
content up to o.l % was analysed. Leachable components from the
flyash but also from the raw gas are transferred to the discharge
water, which is an important stream from the environmental point
of view, as it is necessary at least to discharge surplus water
resulting from coal moisture and steam input into the gasifica-
tion stage as well as quenching water used for raw gas cooling
before entering the waste heat boiler.
TABLE 7 INORGANIC PRIORITY POLLUTANTS
Element
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
Concentration, mg/l
Boiler Feed
Water
<0.06
<0.1
<0.007
<0.015
0.025
<0.1
<0.0002
<0.03
<0.5
<0.01
<0.4
<0.005
Wash Water
to Clarifier
<0.06
<0.1
<0.02
0.015
< 0.01 5
0.02
<0.1
<0.0002
0.06
<0.5
<0.01
<0.4
0.29
The inorganic priority pollutants in the discharge water in
comparison to the boiler feed water, that was used as make up wa-
ter during the test operation, are shown in Table 7.
51
-------
It has to be stated, that a small increase in nickel most
probably results from the low pH-water attack on the steel piping
and the clarifier material. The increase in zinc content possib-
ly result from the volatilization of zinc traces in the coal and
the transportation as gaseous ZnCl2 to the washing system.
The analyses of the discharge water and the boiler feed
water are given in Table 8.
TABLE 8 WATER ANALYSES
Components
pH
COD
Dissolved Solids
ci-
F-
NH3
CN-
SCN-
s-
S203-
so3-
scv
Concentration in mg/l
Boiler Feed
Water
11.2
9
98
9
<0.1
<1
<1
<1
<1
<1
<1
25
Wash Water
to Clarifier
3.8
118
812
306
37
120
1.7-6
3
1
83
11
217
The main components needing further treatment are NHU, CN~
and also the sulfur compounds in lower state of oxidation. Espe-
cially from the relatively high amount of sulfate ions it is evi-
dent, that the dissolved sulfur compounds are oxidized finally
to the sulfate stage. The chlorine content in the coal is nearly
totally transferred to the discharge water. In actual operation
at the TVA plant the discharge water will be minimized to less
than half of the amount of the water that was actually dischar-
ged at the "Test Unit". The resulting concentrations in the dis-
charge water stream will increase accordingly. Ammonia and cya-
nide result from the earlier discussed gasification reactions
and are transferred from the raw gas to the discharge water.
The washing effect for the trace components in the raw gas
is limited, as is demonstrated in Table 9, which presents the
raw gas analysis after the washing stage in the side stream unit
The remaining impurities are removed in the downstream gas
52
-------
TABLE 9 RAW GAS ANALYSIS
Main Components
Vol.-% (dry)
CO2 9.4
CO 62.2
H2 25.6
N2+Ar 1.8
Impurities
mg/rrtf (dry)
HjS 13,896
COS 2,653
CS2 80
SO2 18
NH3 19
HCN 89
NO 4
Solids 50
TABLE lo PRODUCT AND BYPRODUCT GASES
Components
CO2
CO
H2
N2 + Ar
CH4
H2S
COS
CS2
HCN
Methanol
Concentration in vol.-%
Synthesis
Gas
3
28.48
67.52
0.99
0.01
CO2-Off Gas
75.82
0.02
0.02
24.14
4ppmv
300 ppmv
H2S Claus
Gas
73.39
1.32
0.08
22.21
2.47
0.09
0.19
0.25
53
-------
handling stages. In the Rectisol unit, according to GKT ' s con-
cept, H2S, COS, CS2 and also HCN are removed and transferred to
the Glaus unit for sulfur production. The components S0 and NO
are virtually completely reduced to
and N9 in the gas treat-
ment system. Remaining traces will be removed with condensates
and waste water from the gas treatment stages.
These combined water streams are used for slag quenching
and finally as raw gas quenching water before the waste heat
boiler, Figure 2. From the C02 removal a C02~rich stream is
generated, which is vented to the atmosphere. The final clean
synthesis gas stream, the C02~rich off gas and the H2S-rich
Glaus gas are characterized in Table lo. According to calcula -
ting results it seems possible to reduce the CO-content of the
C02 off-gas by changing the flash conditions in the C02 removal
stage to still lower values.
A further analytical effort was aiming for evaluation of
the organic compounds present in the waste streams of the GKT
PROCESS.
Table 11 shows the results obtained for the raw gas after
cooling and washing. Practically no higher hydrocarbon than
methane and this also at a rather low level were detected.
TABLE 11 ORGANICS IN RAW GAS
Components
CH«
C2\~\4~^~C'2\~\'2
C3H6
Benzene
Toluene
Xylene
C3H8
C2H4
CH3SH
C2H5SH
Concentration
ppmv
10
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
54
-------
To collect organics in the hot raw gas before entering the
washer group XAD resin traps were used. No resin sample contai-
ned sufficient extract to perform liquid chromatography. Infra-
red analysis of each gravimetric residue indicated no organics
other than those found in the resin blank. Low resolution mass
spectrometry revealed that elemental sulfur (Sg) was the only
species not found in the blank, Table 12.
TABLE 12 ORGANICS IN GASIFICATION STREAMS
Components
Oil/Grease
Formate
Phenols
TCO"
Grav21
S8 in Grav.
TCO+Grav31
Raw Gas
before
Washer
mg/m3,
1.6
3.8
-98%
1.7
Wash Water
to
Clarifier
mg/l
1.3
<0.1
<0.001
<0.1
13.2
97.2%
0.4
Slag
mg/kg
0.1
32.0
94.4%
1.9
Solids in
Discharge
Slurry
mg/kg
7.2
780
94.3%
52
1.) Total chromatographable organics
2.) Gravimetric organics
3.) Sulfur free
In the discharge water small amounts of formate are present
and traces of grease and oil were found in the side stream
washer system. Essentially all of the organic extracts of the
sample could be attributed to elemental sulfur (Sg). No other
organics could be identified in the extracts by infrared ana-
lysis and low resolution mass spectrometry. The resulting resi-
due was insufficient for liquid chromatography.
Slag and solids in the washing water (flydust) were subjec-
ted to methylene chloride extraction, followed by gas chromato-
graphy on the extract. Low resolution mass spectrometry again
indicated almost entirely elemental sulfur. The higher gravime-
tric residue of the flydust indicate also some organics with
boiling points higher than that of the C,g normal alkane, which
may result from the carbon content of the flydust, not fully
graphitised, or dust fines, having passed through the filter.
55
-------
CONCLUSIONS
From the results achieved it is evident, that special inter-
est is directed towards the flydust slurry and the discharge wa-
ter. The GKT concept for handling this streams is shown in
Figure 2. The flydust slurry is transferred from the clarifier
to a settling pond where it is slowly dewatered. The effluent
from the pond is partly recycled to the main washing system.
The surplus water in the system is discharged and will be trea-
ted further, especially for ammonia reduction. It has to be poin-
ted out, that the impurities of the wash water react with each_
other and with the flyash. Sulfide ions are converted to S203
and SOU . HCN reacts with sulfur compounds to form SCN and
with the flyash to form insoluble complexes. Additional oxida-
tion reactions occur in the settling pond. These oxidation reac-
tions, which are catalysed by flyash components, involve S~0., ,
SO ~ , CN~, SCN~ and also NH . This experience is demonstrated
in Table 13, which shows water analyses from the GKT coal gasi-
fication plant in Modderfontein, Republic of South Africa. This
Figure, excluding the wash water stream to the clarifier, was
presented by TRW and GKT in 1980.
TABLE 13 WATER ANALYSES FROM GKT COAL
GASIFICATION IN MODDERFONTEIN,RSA
Components
pH
NH3
CNT
SON"
s-
S203=
so3-
sor
Concentration in mg/l
Make-up Water
PSE"
6.8
73
<0.2
2.1
<0.1
<1
<1
584
CW21
8.5
2.4
1.2
2.1
<0.1
< 1
<1
853
Wash Water
to
Clarifier
8.0
134.0
8.6
6.9
0.7
3.6
1.4
746
Settling Pond
Discharge
8.0
38
<0.2
1.3
<0.1
<1
<1
752
1 ' Municipal Purified Sewage Effluent
2 'Cooling Water
The changes in composition from the wash water stream be-
fore the clarifier to the final water discharge stream from the
pond are remarkable, no further treatment of the discharge water
is carried out.
56
-------
SOURCE TEST OF THE TEXACO GASIFICATION PROCESS
LOCATED AT OBERHAUSEN-HOLTEN, WEST GERMANY
by:
Robert G. Wetherold and
Robert M. Mann
Radian Corporation
8501 MoPac Boulevard
Austin, Texas 78766
John Morgan and William Yee
Tennessee Valley Authority
1070 Chestnut Street
Chattanooga, Tennessee 37401
Dr. Peter Ruprecht
Ruhrchemie AG
4200 Oberhausen-Holten, West Germany
and
Dr. Ranier Diirfield
Ruhrkohle 01 und Gas GmbH
4250 Bottrop, West Germany
ABSTRACT
A comprehensive environmental characterization of water and solids from
the Ruhrkohle/Ruhrchemie Coal Gasification Pilot Plant in Oberhausen-Holten,
West Germany has been conducted. Coal is gasified at the plant with a
modified Texaco coal gasification process. The pilot plant tests were
conducted in November, 1980, during gasification of Illinois No. 6 coal.
A test plan was prepared including stream selection, sample collection and
chemical analyses. Multiple samples of six process water and four process
solid streams were collected during two twelve-hour environmental balance
periods. Normal pilot plant operation was maintained during the first
period; the second incorporated water recycle to observe dissolved levels
of components with minimized makeup water.
Samples of liquid and solid process streams have been subjected to
comprehensive analyses including water quality parameters, trace elements,
organic characterization, physical testing, radioactivity analyses and
bioassay testing.
These efforts, conducted under contract with the Tennessee Valley
Authority, have been performed to provide support information for both
process operation and environmental impact associated with a 10,000 ton per
day coal gasification plant proposed by TVA for a northern Alabama site.
Topics to be addressed in the presentation include pilot plant configuration
and operation, sample collection, analytical testing and results of the
characterization program.
57
-------
ENVIRONMENTAL TEST RESULTS
FOR THE RUHRKOHLE/RUHRCHEMIE/COAL GASIFICATION PILOT PLANT
Under a contract with the Tennessee Valley Authority, Radian Corporation
has performed a comprehensive environmental characterization of the process
water, effluent water and solid waste from the Ruhrkohle/Ruhrchemie Coal
Gasification Pilot Plant in Oberhausen-Holten, West Germany. At this plant,
coal is gasified with a modified Texaco coal gasification process.
During a recent test in November and December, 1980, samples of process
water streams, effluent water streams, slag, raw coal and coal slurry were
collected. The samples were obtained during two 12-hour environmental balance
periods on November 14 and November 18, 1980.
The liquid and solid samples have been subjected to comprehensive
analyses. Some analyses were carried out on-site at the Ruhrchemie plant in
Oberhausen-Holten while others have been performed at the Radian laboratories
in Austin, Texas.
Some of the results of the program are briefly summarized in this paper.
The process is described and the overall mass balance is presented. Many of
the key characteristics of the process effluent water are defined, and some
of the results from the wastewater treatability study are presented. The
solid wastes from the process were subjected to RCRA leaching tests, and
these results are given.
PROCESS DESCRIPTION
The testing took place at the Ruhrkohle/Ruhrchemie demonstration coal
gasification plant located in Oberhausen-Holten, West Germany. This plant
contains a Texaco gasifier and has a nominal coal capacity of 5,700 to 6,100
kg/hr [150-160 tons/day]. A simplified flow scheme of the Ruhrkohle/Ruhrchemie
demonstration plant is shown in Figure 1. The numbered points on this diagram
identify the locations from which solid and liquid samples were taken during
the environmental balance periods.
In the plant, coal is fed from a storage bunker to a grinding mill.
There it is pulverized and combined with fresh water or, alternatively,
effluent water recycled from the settler. The resulting coal slurry is pumped
to agitated run tanks and from there into the gasification reactor.
The Texaco Coal Gasification Process reactor is a pressurized, entrained
bed, downflow slagging gasifier. It operates under pressures of 2 to 10 Mpa
[300 to 1,500 psia] and at high temperatures, generally in the range of 1,200
to 1,400°C. These temperatures are above the melting point of the coal ash.
58
-------
Ln
VD
COAL FRESH
SLURRY WATER
V-
TOCOAL SLURRY
PREPARATION
RAW
WASTEWATER
FIGURE 1 FLOW SCHEME FOR THE RUHRKOHLE/RUHRCHEMIE COAL
GASIFICATION PLANT DURING THE ENVIRONMENTAL TESTING
-------
The coal slurry and oxygen are fed into the top of the gasification
reactor vessel. This vessel is lined with refractory, and it consists of two
zones, a gasification zone and a radiant zone. The basic combustion and
gasification reactions occur in the gasification section. The raw synthesis
gas from the gasification section passes into the radiant cooler or quench
section of the vessel. Here the gas is partially cooled and process steam
generated. Synthesis gas from the gasifier is routed into a quench tower,
where recycled process water is used to additionally cool the gas and remove
particulate matter. The synthesis gas leaves the quench tower and passes
into a water scrubber where the final gas cooling and particulate removal
takes place.
Most of the slag produced in the reactor is solidified in a water bath
in the bottom of the radiant cooler section. It is then removed from the
reactor through an electronically-controlled lockhopper system. The slag and
sluice water fall into a covered bin where the granulated slag is continuously
removed by a conveyor belt. The slag water, containing some fine particu-
lates, is sent to the settler. The water from the radiant cooler section of
the gasifier and from the quench tower is routed to a flash tank where dis-
solved gases are flashed off during depressurization. The water from the
flash tank then passes into the settler. The residence time in the settler
is sufficient to allow settling of most of the fine particulate matter.
Two water streams are taken from the settler. The overflow water stream,
taken off near the top of the settler, contains a relatively low concentra-
tion of solids. The underflow stream, taken from the bottom of the settler,
contains the residual fines at a much higher concentration.
The settler overflow is combined with the scrubber blowdown stream and
makeup water stream. This combined water stream is used as slag sluice water
and also recycled to the quench tower and the radiant zone of the gasifica-
tion reactor.
The underflow [or at least a portion of it] from the settler serves as
the water blowdown stream from the coal gasification unit. The blowdown
stream is first sent to an open holding tank before being routed to the plant
water treatment system.
MASS BALANCES
The environmental balance periods took place on November 14, 1980 [EB-1]
and November 18, 1980 [EB-2]. During the tests, Illinois No. 6 coal was pro-
cessed at a rate of 6,900 to 7,200 kg/hr [180 to 190 tons/day]. The composi-
tion of the test coal is shown in Table 1. The major difference in the
operating conditions between the two environmental balance periods was the
disposition of the clarifier underflow [bottoms] stream. During the first
environmental balance period [EB-1], all of the settler underflow stream was
sent to the holding tank and then to water treatment; no underflow was recy-
cled. During the second environmental balance period [EB-2], the fresh water
requirements to the plant were minimized. Approximately half of the settler
underflow stream was recycled to the coal slurry preparation area to satisfy
60
-------
the water requirements to slurry the fresh coal which was sent to the gasi-
fier. The remainder of the settler underflow was sent to the plant water
treatment facilities.
TABLE 1. COMPOSITION OF ILLINOIS NO. 6 COAL
Component
Ash
Volatile
Fixed Carbon
Energy Content [Btu/lb]
C
H
N
Cl
S
Ash
Oxygen (difference)
Concentration*
11.21
28.54
50.25
12686
70.94
4.94
1.28
0.05
3.37
11.21
8.21
100.00
*Values as percent [%] except Energy Content
Some of the operating conditions, as reported by Ruhrkohle/Ruhrchemie
for the environmental balance periods, are summarized in Table 2.
TABLE 2. OPERATING CONDITIONS DURING ENVIRONMENTAL
BALANCE PERIODS
Solids Feed Rate, kg/hr 6,000-6,300
Carbon Conversion [Once-Through], % 89-95
Dry Syngas Production, NM3/hr 11,800-12,200
[H2+CO] Production, NM3/hr 9,000-9,400
Total material balances for both environmental balance periods were
developed. These are shown in Table 3. The balances [or closures] are quite
good for both test periods. Ruhrkohle/Ruhrchemie reported material balances
for oxygen, carbon, water, sulfur, and nitrogen. The oxygen, carbon, water,
and sulfur balances closed within ±4%. The nitrogen balances were within ±7%
of closure.
61
-------
TABLE 3. RUHUKOHLE/RUHRCHEMIE COAL GASIFICATION PLANT -
TOTAL MATERIAL BALANCE
Stream Flow Rates, kg/hr
Environmental Balance
Period 1
Environmental Balance
Period 2
In
Out
In
Coal Feed
Water
Other Inlet Streams
Synthesis Gas
Effluent Water
Slag & Other Outlet Streams
7,157
8,312
6,226
TOTAL 21,695
6,923
4,600
5,921
11,597
7,708
1,863
21,168 17,444
Balance [Out/In], %
97.6
97.9
Out
11,542
3,706
1,833
17,081
WASTEWATER CHARACTERISTICS AND TREATABILITY
Samples of seven water streams were obtained during the two environmental
balance periods. These streams were:
Inlet Streams - fresh water
Process Streams - gasifier quench water
- quench tower bottom stream
- slag water
- settler inflow water
Outlet Streams
- settler overflow water
- settler underflow water
During the environmental testing, the settler overflow water stream was re-
cycled to the process. The settler underflow stream served as a purge stream,
and was sent to wastewater treatment. During the second environmental balance
period, some of the settler underflow was recycled to the coal slurry prepa-
ration area.
Several hundred gallons of the pilot plant settler underflow stream were
collected during the two environmental balance periods. Settler underflow
water from the second environmental balance period was subjected to treata-
bility tests in a screening study performed by AWARE, Inc., of Nashville,
Tennessee. The goal of this screening study was to define the treatability
of wastewater from the Texaco coal gasification process. A conceptual waste-
water treatment system was simulated on a laboratory-scale in the study. This
system utilized conventional existing water treatment technology. Bench-scale
62
-------
simulation of wastewater treatment processes is an accepted and often the
only means of obtaining treatment system design criteria when a "new" indus-
trial wastewater is involved.
In the study, each treatment process was individually simulated using a
portion of the wastewater to determine optimum process conditions. After the
optimum conditions for a process were defined, the remainder of the waste-
water was treated at those conditions. The next treatment process in the
sequence was then evaluated.
A summary of the water quality characteristics of the raw settler under-
flow water stream is presented in Table 4. The raw wastewater characteristics
are either average values of samples taken from the drums of wastewater to be
treated or the value of a composite of samples from the drums. The sulfide
and cyanide concentrations had the greatest variations among drum samples.
Sulfide results were biased low because the samples were hot when collected,
and some off-gasing occurred during cooling.
TABLE 4. CHARACTERISTICS OF WASTEWATER TO TREATMENT
Concentration
Parameter
COD
BOD
Phenolics
NH3-N
CN~
SCN-
S=
S04~
TSS @ 105°C
TDS @ 180°C
Alkalinity
[as CaC03]
pH [units]
540
202
<0.018
1, 550±98
15±14
11
128±80
5
152±42
960
4,070
8.3
A relatively high ammonia concentration was observed in the raw waste-
water samples. Minimal concentrations of nitrate or nitrite were found. The
total alkalinity was found to be high. Bicarbonate alkalinity was the pre-
dominant form.
The wastewater treatability screening study was directed toward the re-
duction of the major wastewater quality parameters including BOD, COD, TSS,
NHa, and foS levels. The laboratory treatment system, consisting of conven-
tional solids removal, steam stripping, and oxidation, was effective in reduc-
ing these parameters. The reductions are summarized in Table 5. The removal
efficiency across the total treatment process for both COD and TSS was in the
range of 70-80%. The removal efficiency for most other constituents was in
63
-------
excess of 90%. Total dissolved solids increased substantially due to caustic
and acid additions required for pH adjustment during the treatment sequence.
TABLE 5. TREATABILITY RESULTS
Reduction
Component t^J
BOD 91
COD 68-86
67-86
NH3
H2S
Fe 83
TDS *
Sulfate *
Arsenic 0
Fluoride 3
Selenium 0
Chloride 0
*The use of NaOH and HaSO^ for pH adjustment increased
the levels of sulfate and dissolved solids.
The effect of the treatment process on the levels of various metals was
also determined. Elements whose concentrations are reduced by the treatment
system include Al, Ba, Be, Ca, Fe, Mg, Mn, Pb, Si, and Ti. On the other hand,
the concentrations of several metals were virtually unaffected in the treat-
ment process. Included in this category are As, B, Cd, Ni, Sb, and Se. The
levels of a number of elements, including Ag, Co, Cr, Mo, Tl, and V, were
below detectable limits in both the raw wastewater and the effluent from the
treatment system.
SOLIDS CHARACTERIZATION
Two process solid wastes and two solid wastes from the bench-scale waste-
water treatment system were subjected to RCRA extractions. The solid wastes
were:
process wastes - slag
settler fines
treatment wastes - primary sludge
biosludge
An elemental analysis was performed on the RCRA leachates. The results
of these analyses are shown in Table 6. None of the RCRA limits for metals
was exceeded in the leachates of any of the solids stream.
64
-------
TABLE 6. RCRA LEACHING RESULTS OF SOLIDS*
Process Samples
Element
Ag
As
Ba
Cd
Cr
Hg
Pb
Se
Slag
<0.001
<0.003
0.094
0.008
< 0.001
<0.0005
<0.002
<0.004
Settler
Fines
<0.001
< 0.003
0.10
0.014
< 0.001
< 0.0005
0.024
<0.004
Treatability Samples
Primary
Sludge
< 0.001
0.031
0.21
0.010
0.074
< 0.0005
0.003
< 0.004
Bio sludge
<0.001
0.039
0.29
0.35
0.055
< 0.0005
0.008
0.078
RCRA
Limit
5.0
5.0
100
1.0
5.0
0.2
5.0
1.0
*Concentration in yg/ml (ppm)
SUMMARY OF RESULTS AND CONCLUSIONS
Some of the major results and conclusions of this study are:
The gasification pilot plant appeared to be operating at
steady-state conditions during the environmental testing.
The total mass balance showed very good closure.
• Existing conventional wastewater treatment technology is
effective in reducing the significant effluent water quality
parameters including BOD, COD, TSS, NHs, and
The concentrations of many of the trace metals are signifi-
cantly reduced in the wastewater treatment system.
The quality of most receiving waters should not be adversely
affected by the treated effluent if properly treated with
existing wastewater treatment technology.
The solid wastes are not classified as toxic wastes according
to RCRA extraction procedures.
ACKNOWLEDGEMENT
The coal test run was sponsored by the Electric Power Research
Institute and we wish to acknowledge the assistance which they provided during
our sampling activities. We would especially like to express our appreciation
and thanks to the staff and operating personnel at the Ruhrkohle/Ruhrchemie
gasification plant for their considerable efforts in support of our sampling
program.
The assistance and cooperation of the Texaco Development Corporation
is gratefully acknowledged.
65
-------
SOURCE TEST AND EVALUATION OF A RILEY T
GAS PRODUCER FIRING NORTH DAKOTA LIGNITE
by: Fred L. Jones
American Natural Service Company
Detroit, MI 48226
William P. Barley
Riley Stoker Corporation
Worcester, MA 01613
M. R. Fuchs
Radian Corporation
Austin, TX 78766
V. A. Kolesh
Riley Stoker Corporation
Worcester, MA 01613
ABSTRACT
A ten-foot six-inch diameter Riley Morgan gasifier was operated for 14
days to convert North Dakota lignite to low-Btu gas. During that period,
the gasifier was operated at a range of load conditions, and the product
gas was transported to a commercial-scale kiln burner mounted in a large
combustion test chamber. Process stream conditions and compositions were
recorded throughout the test and were submitted to an SAM/1A analysis.
Gaseous effluent streams were found to be well controlled due to the unique
Riley coal feed and poke hole systems. Solid wastes from the process
(gasifier ash and cyclone dust) were found to be nontoxic, noncarcinogenic
and nonmutagenie. Gasifier wastewater effluent (ash pan water) was simi-
larly found to be nonhazardous. Although combustion stack gases were not
monitored, sulfur and particulate loadings in the gasifier product gas
indicated that the stack gases would comply with current EPA New Source
Standards. If all reduced nitrogen compounds were converted to NO how-
ever, these emissions would exceed New Source Performance Standards'
66
-------
"SOURCE TEST AND EVALUATION OF A RILEY GAS PRODUCER
FIRING NORTH DAKOTA LIGNITE"
In 1979, American Natural Service Company commissioned the Riley Stoker
Company to perform a full scale test of the low Btu gasification of North
Dakota lignite in the Riley coal gasifier. Co-sponsors of the test were
the Riley Stoker Company, The Hanna Mining Company and the Environmental
Protection Agency through a contract with Radian Corporation. This report
summarizes the results of that test and more specifically the environmental
assessment of the Riley gas producer carried out by Radian Corporation.
PROCESS DESCRIPTION
The Riley gasifier is a modern, modified version of the Morgan gas
producer, workhorse of the steel, glass, lime, pulp and chemical industries
during the first fifty years of this century. Of the nearly 1,100 of these
first generation units manufactured, at one time installations could be
found on every continent in the world.
Since 1974, Riley Stoker Corporation has been operating a commercial
sized demonstration unit at its Worcester, MA., R & D facility, the primary
goal being the refinement of the successfully established first generation
technology to the standards imposed by both environmental and operational
constraints of the synfuels industry. A secondary, but no less important
goal, has been the gaining of hands on experience in an area long treated as
more an art than a science, and in which hard data is generally lacking.
During this time, a total of twenty-two demonstration runs with various
eastern coals was carried out on the gasifier, together with an equal amount
on a smaller one-fifth scale gasifier. Results of this program, together
with a summary of practical operating experiences have been presented before
other bodies (References 1, 2, 3).
THE GASIFIER
The Riley gasifier is depicted in Figure 1. This unit is an example
of that group of gas producers classified as thin-bed, atmospheric. The
entire fuel bed, including ash, never exceeds 140 cm. (55 inches), and slowly
rotates, at a speed of one revolution in six and one-half minutes.
The height of the active fuel bed, 76 to 91 cm. (30-36 inches),
differentiates this gasifier from moderately deep bed gasifiers (Wellman-
Galusha, Lurgi), and very deep bed gasifiers (two-stage units). This design
resulted from the need to accommodate swelling bituminous coals in the steel
67
-------
Figure 1. The Riley Gasifier
68
-------
industries of the United States and the United Kingdom.
As has been found (Reference 1), coal particle heating rate is the
controlling factor in managing swelling coals, and the ability to vary bed
height allows the time-temperature history of the individual coal particle
to be preselected and governed, resulting in minimal swelling. In general,
thin-bed gasifiers operate with considerably higher exit temperatures than
do the other classes of gasifiers, so that the distillation/pyrolysis en-
vironment which is first seen by the coal is more severe. This difference
must be kept in mind in comparisons of yields and distribution of some of
the minor families of compounds evolved from different gasifiers. Ash
retention characteristics may also be influenced.
Continual rotation of the entire 3.2 meter (10'-6") I.D. unit accomp-
lishes a number of purposes. Primarily, it is to assure even coal distri-
bution across the entire fuel bed, a crucial factor in thin-bed management.
This is accomplished without the use of an internal distributor by means of
a slot drum feed across an entire radius of the unit. Thus, a continuous
curtain of the fuel is evenly laid upon the advancing fuel bed.
Second, two horizontally fixed but vertically moveable water-cooled bars
perform the function of fuel bed agitation, another requirement for swelling
coals.
Third, ash is removed intermittently thru the use of a plow mechanism
which is periodically stopped, scooping ash from the integral pan and dis-
charging it over the ash pan lip. Some of the seal water will be carried
over with this ash, and means for its treatment must be considered.
The Riley gasifier utilizes a blast hood for air/steam admission, rather
than a grate, the ash bed acting as the diffuser.
THE SYSTEM
The demonstration facility at Riley Stoker is shown schematically in
Figure 2. Coal is fed to the unit from a 60 ton nitrogen sealed bunker thru
a three-valve lock hopper system, and is gasified by the countercurrent
air/steam mixture. Gas exits thru a .9 meter (36 inch) insulated line, is
cleaned of particulate in a high efficiency cyclone and transported to a
300 million Btu/hour test furnace where it is combusted. Char is removed
from the cyclone dry, thru a lock hopper arrangement. A photograph of the
installation is shown in Figure 3.
DESIGN FOR EMISSIONS
Much of the work at RSC during the past seven years has been devoted
to design improvements of those parts of the system responsible for fugitive
emissions. Historically, most gas producer manufacturers paid scant attention
to the two areas most responsible for such emissions of raw gas: the coal
feed system and access ports.
69
-------
N,SEALED
COAL BUNKEB
14 UPPER
FLOW CONTROL VALVE-*-
(Coal Shut oil)
Figure 2. Schematic of the Test Facility
70
-------
Figure 3. Riley Installation
t ' mr '
71
-------
Coal Feed System
In the Riley Stoker lock-purge coal feed system as shown in Figure 2,
fuel is inventoried up to a level just below the middle valve, with the
valves positioned as shown. The top valve must support a head of coal up
to 6.4 meters (21 feet). The middle valve is closed to contain system gases.
The lower valve is open to admit lock hopper coal to the gasifier feeder.
All valves are of a semi-ball type, with ground seats.
As gasification proceeds the fuel in the lock hopper falls to a level
just below the lower valve, where its absence is detected by means of a
sonar device. This triggers a sequence as follows: following closure of
the lower valve, a short burst of steam is admitted into the line from the
eductor to the lock hopper cleaning that line of any residual tar or dust
from the previous cycle. After an interval, the purge shut-off valve opens,
and the steam eductor begins evacuating the lock hopper of gas and dis-
charging it into the downstream gas piping system. The pressure in the lock
hopper eventually becomes sub-atmospheric (660 mm. Hg. or less) . At this
time, the steam and purge shut-off valves close, and the upper two flow
valves open, admitting coal from the main storage bunker to the lock hopper.
These valves remain open until the lock hopper is filled to the previous
level (approximately 1 ton). At this time the upper valve closes, inter-
rupting coal flow, followed by the middle valve closing, creating a gas
tight seal. The lower valve opens with middle valve closure, and the cycle
is completed. Total elapsed time for this entire cycle is approximately one
minute.
Throughout this sequence, gasifier feed has been maintained continuously,
from the inventory of coal located between the coal feeder and the lower lock
valve.
By the maintenance of a nitrogen blanket just slight above atmospheric
pressure in the storage bunker above this system, together with the eductor
system, migration of gases is always toward the gasifier, and never from it.
Access Ports
After a number of trials, RSC has perfected a nitrogen-sealed (any inert
will do), universal joint access port, to be used during those times when
access to the gasifier bed is a necessity. This unit is shown in Figure 4.
Mounted atop a machined ball that moves within a gland seal fixed to the
gasifier deck is another ball valve, a packing gland and a flexible hose con-
veying inert gas at approximately 1.4 kg/cm^ (20 psig).
Insertion of a rod thru the upper gland to the ball valve is followed by
opening the inert gas line. The ball valve is then opened and the rod in-
serted into the gasifier thru the assembly, inert gas flowing into the gasi-
fier, and also around the upper gland to the environment.
This arrangement assures no leakage of producer gas into the environment,
72
-------
LO
Figure 4. Nitrogen-sealed universal joint access port.
-------
and also makes possible viewing of the fuel bed by the substitution of a
plexiglass plate for the rod or other device.
ABSTRACT OF RUN
The gasifier test facility was operated from December 4 to 15, 1979.
During this period a total of 374 metric tons of North Dakota lignite was
gasified. Because of some non-gasifier problems, two interruptions caused
shutdowns early in the test. The major portion of the test fuel was pro-
cessed in the last five days of operation, and the results presented herein
represent data taken from this continuous period. More specifically, data
for the environmental assessment were taken over a 25 hour period extending
from 0900 hours on December 13 through 1000 hours December 14.
Tables 1 and 2 present the heat and material balance around the gasi-
fier at a time near the end of the 25 hour period. Table 3 shows a summary
of the concentrations of the major species of the product gas over the 25
hour operating period. Table 4 summarizes the chemical compositions of the
feed coal, tar and cyclone dust.
Over the total two week test period, the gasifier was operated at feed
rates up to 4550 Kg/hr (10,000 Ib/hr) of coal, and produced a high quality
low Btu gas with a heating value of approximately 160 Btu/cubic foot. A
summary of process conditions during the 25 hour sampling period is given
in Figures 5 and 6.
ENVIRONMENTAL RESULTS
For the purposes of a source analyses model evaluation, five effluent
or process streams are considered coming from the gasifier: product gas
vapors; product gas particulates, tars and oils; gasifier ash; cyclone dust;
and ash pan water. Product gas vapors are distinguishable from product gas
particulates, tars and oils by a characterization temperature of 115 C
(240 F). The product gas is separated into two fractions at this temperature
by the sampling procedure. The separation makes it possible to assess the
potential health and ecological effects of fugitive emissions. It also
allows an evaluation of appropriate control technologies. The phase conden-
sed and collected at 115 C (particulates, tars and oils) was collected in an
electrostatic precipitator. The remaining portion of the product gas (vapors)
was collected in a condenser at approximately 15 C (60 F), followed by an
organics absorption resin for organics collection, or an impinger train for
trace elements, ammonia or hydrogen cyanide.
The results of the source analysis model (SAM/1A, Reference 4) evalua-
tion of the five effluent or process streams are presented in Figure 7. By
this evaluation procedure there are potentially harmful health and ecological
effects for all total discharge severity (TDS) and total weighted discharge
severity (WDS) values above 1. Each of the five streams exhibited potenti-
ally harmful health and ecological effects. The SAM/1A approach indicated
that potential health and ecological effects were primarily due to organic
compounds. However, the total DS for the streams (except gasifier ash and
product gas vapors) include significant contributions from "worst case assump-
74
-------
TABLE 1. OVERALL HEAT & MATERIAL BALANCE
Ln
Mass Flow Rate
kg/s
Inputs
Coal
Net Stream
Air
Total
Outputs
Dry Gas
Moisture
Tars and
Oils
Cyclone Dust
Ash
Heat to Cooling
Water
Unaccounted for
Losses
TOTAL
1.
0.
1.
2.
2.
0.
0.
0.
0.
2.
045
274
554
837
299
440
0215
0043
108
873
(lb/hr)
( 8,292)
( 2,174)
(12,334)
(22,800)
(18,245)
( 3,492)
( 171)
( 34)
( 858)
(22,800)
Temperature
Op / O-ri\
\j \ r )
-2 ( 29)
164 (328)
-2 ( 29)
270 (518)
270 (518)
270 (518)
270 (518)
93 (200)
Type of
Heat
Potential
Sensible
Sensible
Sensible
Potential
Sensible
Sensible
Potential
Sensible
Potential
Sensible
Potential
Sensible
Heat Flow
Rate
Enthalpy*
kJ/kg
16,205
-33
2,847
-28
6,020
265
2,910
26,193
205
20,139
205
8,806
58
(Btu/lb)
( 6
(
( 1
(
( 2
(
( 1
(11
(
( 8
(
( 3
(
,967)
-14)
,224)
-12)
,588)
114)
,251)
,261)
88)
,658)
88)
,786)
25)
KW
16,932
-34
780
-43
17,635
13,840
609
1,280
566
4
86
1
952
6
128
163
17,635
(1000
Btu/hr)
(57,700)
( -116)
( 2,661)
( -148)
(60,167)
(47,218)
( 2,080)
( 4,368)
( 1,926)
( 15)
( 294)
( 3)
( 3,248)
( 21)
( 437)
( 557)
(60,167)
Percent
96.0
-0.2
4.4
-0.2
100.0
78.5
3.5
7.3
3.2
0.5
5.4
— — —
0.7
0.9
100.0
* Enthalpy is 25 C (77°F) and H20 liquid. Potential heats are based on higher heating value (HHV).
-------
TABLE
POTENTIAL ENERGY FLOWS BY LOWER HEATING VALUE
Lignite Feed
Dry Gas
Tars and
Oils
Cyclone Dust
Ash
TABLE 3.
Mass
Flow Rate
kg/s
1.045
2.299
0.0215
0.0043
0.108
MAJOR GAS
LHV
kJ/kg
14
5
24
19
8
,783
,636
,398
,821
,706
LHV Heat
Flow Rate
kJ/s
15
12
,448
,957
525
85
940
COMPONENTS
Time
Hrs
Dec. 13 1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
Dec. 14 2400
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
Volume Percent on
CO
NA
NA
NA
NA
26.9
24.1
26.1
28.0
27.1
27.4
24.6
25.3
27.3
27.4
27.6
26.6
27.0
27.7
28.2
28.9
25.7
28.8
29.2
28.9
26.4
C02 H2
NA NA
NA NA
NA NA
NA NA
7.0 16.6
8.6 16.6
7.2 16.5
6.8 16 . 7
6.6 16.7
6.7 26.9
8.3 16.5
8.0 16.5
7.1 16.8
7.2 16.9
7.1 16.6
7.7 16.0
7.1 17.1
6.4 18.0
6.2 17.4
6.3 17.3
7.7 18.3
6.3 18.7
5.9 19.0
5.3 17.3
8.9 17.2
°?
NA
NA
NA
NA
1.1
1.1
1.0
1.2
1.1
1.2
1.1
1.1
1.0
1.0
1.1
0.9
1.0
1.1
1.2
1.2
1.3
1.3
1.1
1.0
1.2
Dry Basis
N?
NA
NA
NA
NA
46.8
48.2
47.3
45.8
46.5
46.4
47.6
47.5
45.7
45.6
45.3
46.6
45.6
44.5
44.8
44.7
45.4
43.3
43.2
45.9
44.8
CH&
NA
NA
NA
NA
NA
0.6
1.0
0.7
1.2
0.7
1.0
0.7
1.2
1.1
1.4
1.4
1.3
1.5
1.5
0.9
0.9
0.9
NA
NA
NA
Notes:
* Compositions are Radian process gas chromatograph readings
normalized to 100 percent.
** Argon was not measured and is assumed to be 0.54 volume
percent for all periods.
76
-------
GASIFIER
EXIT
TEMP. "F
GAS AT
BURNER
BLAST
SAT.
TEMP. °F
10,000-
AIR RATE 9,500-
LBS/HR 9,000-
8,500 -
STEAM
RATE
LBS/HR
2,400 -
2,300-
2,200
2,100-
2,000 -
AIR TO
STEAM
RATIO
LBS/LBS
6.0
5.8
5.6
5.4
LOWER
HEATING
VALUE OF
DRY
PRODUCT
GAS
BTU/SCF
170-
160-
150-
140-
130-
SPECIFIC
GRAVITY
OF DRY
GAS
0.860 -
0.840-
0.820 -
GASIFIER EXIT TEMP.
GAS AT BURNER
1000 1400 1800 2200|0200 0600 1000
d) (2) (2) (2) ^) (2)
~^— 12/13/79112/14/79 —^>
PLOW TIME AND NUMBER
Figure 5. Process Variable for 24 Hour Period
77
-------
10,000.
AIR RATE 9,500
LBS/HR 9]000
8,500
8,000-
COAL 7,000
FEED
RATE
LBS/HR 6,000
5,000-
AVERAGE
FOR
PERIOD
AVERAGE
TOTAL
BED
HEIGHT,
INCHES
AVERAGE ASH BED HEIGHT 25 INCHES
BED aP
INCHES
WATER
COLUMN
10.0_
9.0_
8.0_
7.0_
6.0
T
TT
0900 I 1200
(t)
1600
I 2000 2200 2400 0200 0400
© (2)
-^—• 12/13/79112/14/79 —»•
PLOW TIME AND NUMBER
0600 0800 1000
Figure 6. Bed Related Process Variable Plots for 25 Hour Lignite Run
78
-------
1612-
1E11-
1E10-
1E9-
1E8-
1E7-
1E6 -
1E5-
1E4 -
1E3-
162-
161-
1612-
1E11-
1E10-
1E9 -
1E8-
1E7-
1EB-
1E5-
1E4-
163-
162-
1E1 -
J Haallh Impact
I Ecology Impact
Haallfi Impact
Ecology Imapct
Li I
Asn Pan wa(*r
Figure 7. Total Discharge Severities and Weighted
Discharge Severities of Effluent Streams
79
-------
TABLE 4. SELECTED CHEMICAL ANALYSES
oo
o
Moisture
Ash
Carbon
Hydrogen
Oxygen
Nitrogen
Sulfur
V.M.
F.C.
HHV, Btu/lb
HHV, kJ/kg
LHV, kJ/kg
Feed Coal
(as received) %
34.10
7.54
41.82
2.83
12.70
0.70
0.31
30.68
27.68
6,967
16,205
14,669
Cyclone
Dust
WT.%
0.59
38.21
54.48
1.48
2.82
0.83
1.59
—
—
8,607
20,020
Tar
WT.%
—
0.20
59.57
8.40
30.75
0.69
0.33
—
—
11,261
26,193
TABLE 5. CONTRIBUTIONS TO TOTAL DS AND TOTAL WDS BY WORST CASE ASSUMPTIONS
GC/HECD DATA, AND EXPERIMENTALLY DETERMINED RESULTS
EFFLUENT STREAMS
Health
WCA*GC/HECDEXP.ORG./INORG.
Ecology
WCA* GC/HECD EXP. ORG./INORG.
Product Gas: Particulates,
Tars and Oils ~97%
Product Gas: Gas and Vapors ~100%
Gasifier Ash ~1%
Cyclone Dust ~83%
Ash Pan Water ~74%
99.9%/0.1% ~99%
100%/0%
-97% 18%/82%
83%/17% ~23%
99.9%/0.1% ~76%
1% 99.7%/0.3%
"100% 100%/0%
~95% 5%/95%
~77% 23%/77%
79%/21%
*Worst Case Assumptions
-------
tions" for the organics, which should be greatly reduced by more extensive
analysis for specific organics (Table 5).
Inorganic elements, on the other hand, had the most significant con-
tribution to the health and ecological impacts of the gasifier ash stream,
with iron having the most significant health impact and phosphorus the
most significant ecological impact. Inorganic elements did not contribute
significantly to total DS or WDS for product gas or ash pan water health
effects. However, phosphorus did make a significant contribution to the
ecological effects of the ash pan water.
The findings in this source test evaluation indicate that the potential
health and ecological effects of the ash pan water are significant. Even
though the organic loading of the ash pan water was very low, the health
and ecological DS and WDS of the stream were the result primarily of or-
ganics, other than the contribution of phosphorus to ecological impact.
During the test program, ash pan water was continuously purged. Therefore,
the concentrations of many parameters of concern in the stream were possibly
held below anticipated levels of design operation using recycle, thereby
reducing the total DS. However, continuous purging of the ash pan water
provided a flow rate higher than design operation flow rate for the ash pan
water and thereby raised the total WDS values to a level representative of
commercial operation of the gasifier (Table 6) .
Since organics provide the most significant contribution to the total
DS of the product gas, these values would be reduced for the product gas
combustion effluent due to the vast reduction in organic content following
combustion. It should be noted that the SAM/1A approach treated the product
gas as an effluent stream, which it is not, other than as a fugitive emission.
The SAM/1A results for gasifier ash and cyclone dust also showed po-
tentially harmful health and ecological effects. On the other hand, bio-
assay tests conducted on the solid gasifier ash and cyclone dust indicated
little or no health hazard. A neutral leaching of the two solid streams pro-
vided a liquid for bio-assay testing that showed a high level of ecological
hazard. However, subjecting the gasifier ash and cyclone dust to (RCRA)
leaching procedures (Reference 5) resulting in the solids being classified
as non-hazardous (Tables 7, 8).
EMISSIONS FOLLOWING COMBUSTION
Presently, the main concern about the utilization of coal and coal-
derived fuels in industry centers above emissions of oxides of sulfur,
oxides of nitrogen, and particulates. While the actual emissions- of S02
and NOX due to low Btu gas combustion are dependent on application, the
Riley test provided some indication of what the expected levels of these
emissions might be.
During the two week test period, product gas from the gasifier was
passed through a single stage cyclone for clean-up and transported directly
to the large kiln burner. Because the temperature of the gas was maintained
very close to that observed at the gasifier exit, not one gallon of product
81
-------
TABLE 6. COMPARISON OF LIQUID STREAMS TO DRINKING WATER STANDARDS*
ASH PAN WATER**
As
Ba
Cd
Cl-
Cr
Cu
F-
Fe
Pb
Mn
Hg
Se
Ag
Zn
N03- (as N)
S04= (mg/1)
pH
TDS (mg/1)
NIPDWS***
Mg/1
50
1,000
10
50
1,800
50
2
10
50
NSDWR****
Mg/1
250,000
1,000
300
50
5,000
10,000
250
6.5-8.5
500
RANGE
Mg/1
11 -
91 -
30+
230
48,000-110,000
3 -
300 -
540 -
<1 -
27 -
<5 -
2 -
<3 -
540 -
10.2 -
1,250 -
10
710
3,400
2
96
10+
30
12
1,870
11.3
2,050
AVERAGE
Mg/1
21+
170
<1
78,000
<1
6
530
1,900
<1.3
58
<0.5+
<7
14
<5.3
<20
1,260
10.8
2,250
ASH PAN
CARRY-OVER
WATER
Mg/1
9+
250
<10
2,000
<1
<1
~1,000
910
300
11
<0.5+
<5+
20
<3
5.0
SERVICE WATER
Mg/1
<10
100
<10
300
<30
70
~20
3,000
<10
40
<10
<10
80
6.1
4.3
52
*ICPES analytical results unless noted otherwise. Analysis performed by SSMS.
**Samples RM-18, RM-23, RM-49
***National Interim Primary Drinking Water Standards (Federal Register, 8/27/80).
****National Secondary Drinking Water Regulations (Federal Register, 7/19/79).
+Analysis performed by AAS.
-------
TABLE 7,
BIOASSAY TEST MATRIX
Cyclone Dust
Gasifier Ash
Ames*
(Health)
Negative
Negative
CHO**
(Health)
Low Toxicity
No Detectable
Tr-lVl r*4 <-1T
RAM*** RAT****
(Health) (Health)
No Detectible
Toxicity
Low Toxicity
Fresh-
water
Alga*****
(Ecology)
Gasifier Ash
Neutral Leachate
Cyclone Dust
Neutral Leachate
Negative
Negative
Low Toxicity
No Detectable
Toxicity
Not
Toxic
Not
Toxic
Toxic
Toxic
*Salmonella Mutagenesis Assay (Ames)
**Chinese Hamster Ovary Clonal Toxicity Assay (In Vitro Cytotoxicity Assay)
***Rabbit Alveolar Macrophage Assay (In Vitro Cytotoxicity Assay)
****Rodent Acute Toxicity (Acute in Vivo Toxicological Test)
*****Freshwater Alga (Selenastrum capricornutum) Toxicity Assay
TABLE 8. COMPARISON OF SOLID EFFLUENT EXTRACTS
AND RCRA EXTRACT LIMITS
RCRA
Extract Limits*
(5/19/80)
Gasifier
Ash Leachate*
Cyclone
Dust Leachate*
As
Ba
Cd
Cr
Pb
Hg
Se
Ag
Endrin
Lindane
Methoxychlor
Toxaphene
2,4-D
2,4,5-TP Silvex
5,000
100,000
1,000
5,000
5,000
200
1,000
5,000
20
400
10,000
500
10,000
1,000
33
680
<0.5
<1
<2
<0.5
6
<0.5
<2.0
0.2
<2.0
<100
<0.8
<0.3
4
390
<0.5
<1
<2
<0.5
2
<0.5
<2.0
<0.2
<2.0
<100
<0.8
<0.3
Concentrations in yg/liter.
83
-------
gas tars and oils was condensed from the gas, nor was any significant amount
of water-based condensate produced. With the exception of a light coating
of dust which formed on the inside of the gas main, all product gas efflu-
ents leaving the gasifier cyclone outlet proceeded to the gas burner for
combustion.
Combustion of the product gases was achieved in a low pressure baffle
burner operating with combustion in the primary air zone at 33% of theoreti-
cal air and 213% of theoretical air overall. The theoretical adiabatic
flame temperature for this mixture was 1077 C (1970 F) . The flame produced
was a long diffusion flame, ranging from 1.2 to 2 meters (4-7 feet) in dia-
meter and 9 to 12 meters (30-40 feet) in length. Figure 8 illustrates sev-
eral temperature profiles measured in the flame using a suction pyrometer.
No measurements of actual stack emissions were made. However, by
analyzing the composition of the product gas being fed to the combustor,
it was possible to make the following correlations.
The North Dakota (Indian Head) lignite gasifier feedstock for the 24
hour test period had an average sulfur concentration of 0.44g/10^J (1.02 Ib
S/1Q6 Btu). Some 53% of the sulfur being fed to the gasifier was being
converted to reduced sulfur species in the product gas, with the majority
of the remainder being retained by the gasifier ash stream. If 100% of the
reduced sulfur species in the product gas were converted to sulfur dioxide
during combustion, the resulting S02 emission level would be 0.49g/lQ6j
(1.10 Ib S02/106 Btu) based on the heat value of the lignite feed. The New
Source Performance Standards emission limit for S02 is 0.52g/106J (1.20 Ib
S02/106 Btu) for coal-fired boilers (Reference 6, Subpart D: Fossil Fuel
Fired Steam Generators).
The average ammonia content of the product gas was 7.8 X 10 yg/NnH
and the average HCN concentration was 1.8 x 105 yg/Nm^. About 26% of the
nitrogen in the lignite feedstock was converted to reduced nitrogen species.
Assuming that 100% of the reduced nitrogen species in the product gas was
converted to NOX during combustion, the resulting NOX emission (as N02)
would be 0.36g NC>2/106J (0.841b N02/106 Btu) based on the heat value of the
lignite feedstock. The NSPS emission level (Subpart D) for NOX is 0.26g
N02/106J (0.60 Ib N02/1Q6 Btu) for coal fired boilers. While estimated
NOX emissions may be biased high, assuming 100% conversion of reduced nitro-
gen species to NOX, it does not provide for the additional NOX created due
to thermal reaction of nitrogen and oxygen during combustion, which for
many applications may be a significant contribution. For the specific
test described here, the low combustion temperatures observed would likely
minimize thermal production of NOX, but high excess air levels would likely
favor conversion of reduced nitrogen species to NOX.
The particulate loading of the product gas stream was 4.76 x 10^ yg/Nnr
downstream of the cyclone. Particulate was assumed to be of the same compo-
sition as the cyclone dust, and the cyclone dust ash content was used for
calculations to determine the particulate emissions after combustion. Basing
the adjusted particulate loading upon the heat value of the lignite feed-
stock, the particulate emission after combustion would be 0.026g particulate/
84
-------
00
fiiiiliiii'iiiil
J 5'8" from Burner
J—'—'—' Burner Face
5 0 5
Distance from Center!ine
Figure 8. Low Btu Gas Flame Temperature Profile (9300 Ib/hr rate)
-------
(0.06 Ib particulate/lO^ Btu). This particulate emission estimate does
not consider particulate resulting from incomplete combustion of tars and
oils. The NSPS (Subpart D) for particulate is 0.043g particulate/106J
(0.10 Ib particulate/106 Btu) for coal-fired boilers.
FUGITIVE EMISSIONS
The North Dakota lignite gasification test provided an opportunity for
testing the effectiveness of the unique Riley coal feed and poke hole designs
for minimizing fugitive emissions from the gasifier. Fugitive emissions of
hydrocarbons were measured in the vicinity of the gasifier by several meth-
ods. Hydrocarbon concentrations, reported as methane, were less than on
.part per million as summarized in Table 9. Hydrocarbons were also measured
in the off-gases from the nitrogen-pressurized coal bin. Concentrations
here were 5-6ppm as methane. Readings of two carbon monoxide monitors
maintained by Riley Stoker were recorded during the sampling period, and are
summarized in Table 10. The maximum recorded CO concentration was 24 ppm with
readings generally below this value. The Occupational Safety and Health
Administration (OSHA) permissable exposure limit for CO is 50 ppm (see
Reference 7). There is no OSHA regulation for hydrocarbons as a compound
class. Propane is the lightest hydrocarbon regulated by OSHA and has a
permissable exposure limit of 1,000 ppm.
Radian Corporation has performed STE's for a Chapman low Btu gasifier
with a bituminous coal feedstock and a Wellman Galusha gasifier with an
anthracite coal feedstock. The Chapman STE Report (Reference 8) presents
coal feeder vent gas hydrocarbons concentrations of 2.5 x 10& lag/NnP. Simi-
lar measurements at the Wellman Galusha facility (Reference 9) resulted in
coal hopper gas hydrocarbon concentrations of 1.4 x 106 ug/Nm3 as methane.
Related values measured at the Riley gas producer are many orders of magni-
tude less. These data demonstrate the relative reduction of fugitive emis-
sions achieved by the controls employed on the coal bin at the Riley gas
producer.
A sample was taken of the poke hole gas discharge during a simulated
poking operation on the gasifier, to determine the effectiveness of the
Riley poke hole design in keeping product gas sealed within the gasifier.
Table 11 summarizes the results of that test, and shows the high degree of
effectiveness of the Riley poke hole.
TRACE ELEMENTS
Trace elements enter the gasification process with the lignite feed-
stock and are subjected to the high temperatures of the process. Many ele-
ments, especially the more volatile ones, undergo volatilization in the hot
areas of the system, and may either remain a vapor in the product gas, con-
dense homogeneously, or condenses upon aerosol particles. Other elements
are chemically transformed into gaseous species and are emitted in the pro-
duct gas. Most trace elements remain in the coal solids and are emitted in
the gasifier ash. Even though the majority of most elements are emitted with
the solid effluent streams, RCRA extraction procedures analyses result in
the classification of these solids as non-hazardous. .
86
-------
TABLE 9. ORGANIC VAPORS ANALYSIS
DATE
12/13
TIME
LOCATION
2223 hrs
—top of gasifier during
poking operation
Coal Bin
—2-inch gate on top
CONCENTRATION
(ppm as
12/13
12/13
12/13
1100 hrs
2200 hrs
2200 hrs
Gasifier Building
— all walkways
Gasifier Building
— all walkways
Gasifier Building
1.5
1
1
ppm
ppm
ppm
5-6 ppm
TABLE 10. CARBON MONOXIDE MONITOR READINGS
Date
12/13
12/13
12/14
Time Monitor 1*
1100
1500 1 ppm
1502
1506
1750 1 ppm
1808
2213
0035
0235 1 ppm
Monitor 2**
20 ppm
12 ppm
15 ppm
24 ppm
1 ppm
<0 ppm
<0 ppm
O ppm
*Located in gasifier building on ground level, west wall.
**Located in gasifier building on gasifier poke hole level,
north wall.
87
-------
TABLE 11. POKE HOLE DISCHARGE DURING SIMULATED POKING OPERATION
DATE: 12/14
TIME: 0645 hours
FLOWRATE: 0.022 m3/sec (actual)
GAS ANALYSIS:
N2 95.4%
H2 1.1%
02 0.2%
CO Below detection limit
CH^ Below detection limit
C02 Below detection limit
88
-------
Minor and trace elements can be grouped according to the mechanism by
which each is emitted. The elements primarily in the product gas can be
considered highly volatile or transformed into gaseous compounds. Moderately
volatile elements are predominately in the cyclone dust or product gas par-
ticulate and can be evaluated on the basis of volatilization and recondensa-
tion. Elements emitted predominately in the gasifier ash can be considered
to be non-volatile elements.
For this source test evaluation, an element was considered to be highly
volatile if 25% or more of its total mass was found in the gas and vapors
portion of the product gas. These highly volatile elements were: bromine,
cesium, chlorine, fluorine, gallium, iodine, selenium, silicon, sulfur and
tellurium. An element was classified as moderately volatile if 25% or more
of its total mass was found in the cyclone dust and particulates, tars and
oils portion of the gas. These elements were: antimony, arsenic, chromium,
germanium, lead, tin, and zinc.
The following elements were possible volatile and will acquire addi-
tional data to characterize their behavior definitively: beryllium, bismuth,
cadmium, dysprosium, erbium, europium, gold, holmium, iridium, neodymium,
osmium, palladium, platinum, praseodymium, radium, rhodium, ruthenium, silver,
tantalum, terbium, thallium, thulium, uranium and ytterbium.
Figures 9 and 10 graphically present the elemental distribution in the
effluent streams. The elements are listed in the order of increasing boiling
points. In general, as the elemental boiling points increase, the predomin-
ance of elemental distribution shifts from the product gas to the gasifier
ash. Although a general trend is evident, there is no direct correlation
between elemental boiling point and distribution. The distribution of in-
dividual elements in the system is dependent not only on elemental boiling
point, but also on much more complex properties, including chemical reactions
within the gasifier, the volatility of compounds containing the elements,
and solubility of compounds in the tars and oils.
Most of the elements classified as highly volatile from their distribu-
tion in the effluent streams were depleted in both the gasifier ash and cy-
clone dust. Cesium and gallium were exceptions and were enriched in both
solids. This behavior is more characteristic of non-volatile elements.
Those elements considered to be moderately volatile from distributions
fell into two major categories. Lead, chromium and zinc follwed the expected
behavior of being depleted in the gasifier ash and enriched in the cyclone
dust. Arsenic and antimony, however, were depleted in both solids. The
distribution results show that both of these elements were found in the par-
ticulates, tars and oils fraction of the product gas. This indicates some-
what greater volatility than that of other moderately volatile elements.
The behavior of germanium and tin was more characteristic of non-volatile
elements, for germanium was enriched in both solids, and the enrichment ratios
for tin were very close to the ash contents.
89
-------
D
Product Gas
Ash Pan Water
Cyclone Dust
Gasllier Ash
0 100 n
t 25-75 H
"5
o
o
ra
50-50 H
75 25
1000
F Cl Br I P Hg S As Se Rb Cs K Na Zn Mg Ba LI Sb Yb Eu Ca Pb Sm Mn Sn Ga
Elements
Figure 9. Stream Elemental Distributions as Percentage of
Combined Effluent Streams Emission (Ordered by
Increasing Elemental Boiling Points)
90
-------
0 100i
^ 25 75 H
o^
To
•*•
o
EH Product Gas
Q Ash Pan Water
| Cyclone Dust
Hi GaslfierAsh
o
u
a
50 50 -I
75 25
1000]
Al Cr B Cu Dy Sc Nl Be Tb Ge Co Y Gd Fe V Pr Tl Ce La Zr U Th Nb Mo W
Elements
Figure 10. Stream Elemental Distributions as Percentage of Combined
Effluent Streams Emission (Ordered by Increasing
Elemental Boiling Points)
91
-------
CONCLUSIONS
The major findings of this program are summarized below.
- Acute bio-assay tests of the gasifier ash and cyclone dust
solids and neutral leachates of the gasifier ash and cyclone
dust indicated no adverse health effects. Environmental bio-
results of neutral leachates of the gasifier ash and cyclone
dust showed significant toxic effects.
- Leaching studies conducted on the gasifier ash and cyclone
dust to determine the effects of solid waste disposal, in-
dicated that the materials are non-hazardous according to
Resource, Conservation and Recovery Act (RCRA) protocol and
standards.
- Although the gasification process emits over 50% of the
lignite sulfur in the product gas, it will not require ad-
ditional sulfur removal to meet New Source Performance Standards
for coal fired boilers.
- Assuming that 100% of the ammonia and HCN present in the
product gas are converted to NO during combustion, and not
considering the contribution of thermally created NOX in
the boiler, additional NOX controls will be necessary to
meet the New Source Performance Standards for coal fired
boilers.
- Additional particulate control measures will not be
necessary to meet NSPS for particulate. This conclusion
is based upon the particulate loading of the product gas
downstream of the cyclone and the heat value of the lignite
feedstock.
- Enclosed and pressurizing the coal bin, together with the
nitrogen purge poke hole mechanism developed by Riley Stoker
significantly reduce fugitive emissions.
92
-------
REFERENCES
1. Barley, W. P., Lisauskas, R. A., and Rawdon, A. H. "Practical
Operating Experience on a Riley-Morgan Gasifier." 88th AIChE
Meeting, Phila., Pa. June 8-12, 1980.
2. Rawdon, A. H., Lisauskas, R. A., and Johnson, S.A. "Operation of
a Commercial Size Riley-Morgan Coal Gasifier." American Power
Conference Chicago, 111. April 19-21, 1976.
3. Lisauskas, R. A., Johnson, S. A., and Earley. W. P. "Control of
Condensible Tar Vapors for a Fixed-Bed Coal Gasification Process."
Fourth Energy Resource Conference, Institute of Mining and Minerals
Research. Lexington, Ky. Jan. 7-8, 1976.
4. Schalit, L. M., and K. J. Wolfe. SAM/1A: A Rapid Screening Method
for Environmental Assessment of Fossil Energy Process Effluents.
EPA-600/7-78-015. Acurex Corporation/Aerotherm Division, Mountain
View, CA, February, 1978.
5. Environmental Protection Agency. Hazardous Waste Management System-
Identification and Listing of Hazardous Waste, Federal Register,
45( 98):33084-33135, May 19, 1980.
6. Environmental Protection Agency. New Stationary Source Performance
Standards: Electric Utility Steam Generating Units. Federal
Register, 44(113):33580-33624, June 11, 1979.
7. Occupational Safety and Health Administration. Industrial Hygiene
Field Operational Manual. U.S. Department of Labor. Washington,
D.C., April, 1979.
8. Page, G. C. Environmental Assessment: Source Test and Evaluation
Report — Chapman Low-Btu Gasification. EPA-600/7-78-202 (NTIS-
PB289940)- Radian Corporation, Austin, TX, October, 1978.
9. Thomas, W. C., K. N. Trede, and G. C. Page. Environmental Assess-
ment: Source Test and Evaluation Report — Wellman-Galusha (Glen-
Gery) Low-Btu Gasification. EPA-600/7-79-185. Radian Corporation,
Austin, TX, August, 1979'.
93
-------
Session I: ENVIRONMENTAL SOURCE TEST AND EVALUATION RESULTS
Part B: Direct Liquefaction
Chairman: W. Gene Tucker
U.S. Environmental Protection Agency
Research Triangle Park, NC
Cochairman: Morris H. Altschuler
U.S. Environmental Protection Agency
Washington, DC
94
-------
ENVIRONMENTAL PROGRAM AND PLANS
FOR THE EDS COAL LIQUEFACTION PROJECT
by: Richard L. Thomas
Exxon Research and Engineering Co.
Florham Park, New Jersey
ABSTRACT
The Exxon Donor Solvent (EDS) coal liquefaction project is a unique
government/industry arrangement for developing EDS technology to the point
that commercial plants can be designed with an acceptable level of risk.
Project participants are the U.S. Department of Energy (DOE), Exxon Com-
pany, U.S.A., Electric Power Research Institute, Japan Coal Liquefaction
Development Company, Inc., Phillips Coal Company, ARCO Coal Company,
Ruhrkohle A.G., and AGIP S.p.A.
A broad environmental program is being advanced within the project to
address plant emission, occupational health, and product-related environ-
mental concerns associated with the direct liquefaction of coal. The
current plans, status and outlook for the EDS Environmental Program are
described to provide information on the overall strategy being followed for
the acquisition of data relating to these concerns.
PROGRAM ORGANIZATION AND MANAGEMENT
The EDS Cooperative Agreement forms the basis upon which the govern-
ment can participate in developing a technology in the national interest
with industrial partners who develop and are the end users of the tech-
nology (1). Thus, the EDS Environmental Program Organization and Manage-
ment reflects this arrangement in terms of the character and direction of
the work activities.
Exxon Research and Engineering Company, the developer of the EDS
process, has overall technical and execution responsibilities for the EDS
project. Construction and operation support is provided by Exxon Company,
U.S.A. The various contractual interfaces are shown in Figure 1.
Project direction is carried out by a number of committees consisting
of members of sponsoring organizations participating in the cost sharing of
the project as shown in Figure 2. The EDS Environmental Program draws upon
the various elements of the Exxon organization for carrying out work activi-
ties related to their specific areas of expertise. The EDS Project Director
has responsibility and authority for work direction, stewardship and com-
munications.
95
-------
U.S. DEPARTMENT
OF ENERGY
COOPERATIVE
AGREEMENT
EXXON RESEARCH AND ENGINEERING COMPANY
• PROGRAM MANAGEMENT
• TECHNICAL/COST STEWARDSHIP
• LADOHATOHY AND ENGINEERING STUDIES
• LARGE PILOT PLANT TECHNICAL PROGRAM
PARTICIPATION
AGREEMENTS
SUBAGREEMENT FOR
LARGE PILOT PLANT
EXXON COMPANY. U.S.A.
ELECTRIC POWER RESEARCH INSTITUTE
JAPAN COAL LIQUEFACTION
DEVELOPMENT COMPANY
PHILLIPS COAL COMPANY
ARCO COAL COMPANY
RUHRKOHLE AG
AGIP S.P.A.
EXXON COMPANY. U.S.A.
• OWNERSHIP
• ENGINEERING/PROCUREMENT
• CONSTRUCTION
• OPERATION
Figure 1. Contractual Interfaces EDS Coal Liquefaction Project
-------
SPONSORS MANAGEMENT
COMMITTEE
ER&E VICE PRESIDENT
SPONSORS TECHNICAL
COMMITTEE
EDS PROJECT DIRECTOR"
STC ADVISORY
SUBCOMMITTEES
I T 1
H h-, H
EDS PROCESS
RESEARCH
COAL LIQUEFACTION
LABORATORY
EXXON RESEARCH
AND DEVELOPMENT
LABORATORIES
ER&E PRODUCTS
RESEARCH DIV.
EDS ENGINEERING
EDS LIQUEFACTION
ENGINEERING
DIVISION
EXXON ENGINEERING
TECHNOLOGY DEPT.
LARGE PILOT PLANT
EXXON COMPANY USA
LABORATORY STUDIES
ANALYTICAL SUPPORT
PRODUCT EMISSIONS
PROCESS ENGINEERING
STUDY DESIGNS
DATA ANALYSIS
TECHNOLOGY DEVELOPMENT
COMMERCIALIZATION ASSESSMENT
INDUSTRIAL HYGIENE
COMPLIANCE TESTING
SOURCE TESTING
TOXICOLOGY TESTING
Figure 2.. EDS Coal Liquefaction Project Organization for EDS Environmental Program Activities
-------
DRY
COAL
WATER
AIR
SLURRY DRYING*
LIQUEFACTION*
PRODUCT DISTILLATION*
SOLVENT HYDROGENATION*
SOLVENT FRACTIONATION*
BOTTOMS PROCESSING
* BEING DEMONSTRATED AT ECLP
LSFO
VGO
• STEAM REFORMING
• CRYOGENIC H2
Recovery
• NH3 SYNTHESIS
• DEA OFFGAS
SCRUBBER*
• SOUR WATER TREATING
• DEA REGENERATION*
• SULFUR PLANT
• PHENOL EXTRACTION
t
C3+
•
i
"1
H2
PHENOL
WATER TO
OFFSITES
SULFUR
C3 LPG
»- C4 LPG
»• NAPHTHA
Figure 3. EDS Commercial Plant Study Design Features
-------
IMPLEMENTATION PLAN
The overall objective of the EDS Environmental Program is to assure a
safe and environmentally sound process. Bench-scale research, small pilot
unit operation, engineering design and technology studies, and operation of
a 250 ton-per-day coal liquefaction pilot plant (ECLP) are collectively
being utilized to provide an environmental data base to meet this overall
objective. This effort is summarized in Table 1.
A conceptual design for a commercial scale plant operating on Illinois
bituminous coal has been recently completed for a Western Illinois loca-
tion (2). This engineering study depicting the state of EDS technology in
1978, after approximately ten years of development work, was carried out
in sufficient detail to define environmental control needs and costs for
siting a commercial plant. Studies of this type are used for research
guidance in the environmental program. A similar study reflecting poten-
tial process improvements conceived after 1978, is currently underway for
a conceptual plant operating on Wyoming coal in a Western U.S. location.
The large 250 ton-per-day pilot plant at Baytown3 Texas plays an
important role in providing representative commercial streams for environ-
mental and health studies of EDS materials in the various stages of produc-
tion from raw materials to products and effluents. Chemical and physical
characterization of pilot plant materials in conjunction with bioassay
and occupational exposure data from the pilot plant constitute the data
base for making judgments on the potential environmental acceptability
of the EDS process for commercialization.
Program emphasis is on the aspects of the EDS process which con-
ceivably can be scaled to commercial size facilities. The basic EDS
process streams, plant products and commercial plant design features are
shown in Figure 3 along with those features undergoing demonstration at
the large 250 ton-per-day pilot plant (ECLP).
In the EDS process, coal is dried and slurried with hydrogenated
recycle solvent and reacted in a liquefaction reactor at approximately
800°F and 2000 psia. The three phase product stream from the liquefaction
is separated by a combination of atmospheric and vacuum distillation. The
liquid fuel products are naphtha, a middle distillate (LSFO) and a vacuum
gas oil (V60). If desired, the vacuum gas oil stream may be recycled to
extinction in the liquefaction reactors to provide a product slate with a
boiling range below 800°F. The basic environmental control units involve
sulfur, phenol and ammonia recovery.
The major operating units at ECLP, as they pertain to the EDS pro-
cess are the coal preparation section, the slurry drying section, the
liquefaction section, the product recovery section and the solvent hydro-
genation section. Other areas of ECLP are similar in nature to typical
support units of any petroleum refinery and include DEA regeneration and
gas treating, hydrogen compression, safety facilities, waste handling,
sour water collection facilities, utilities and tankage.
99
-------
TABLE 1. EDS ENVIRONMENTAL PROGRAM IMPLEMENTATION
• Carry out Conceptual Plant Design Studies
- To Identify Areas Requiring Additional Research
- To Develop Updated Investment Costs and Economics
- To Provide Base Point for Initial Commercialization of a Future
Pioneer Plant
§ Carry out Pilot Plant Demonstrations
- Stream Characterization and Source Testing
- Monitor Workplace Exposures
- Equipment Design and Scale-up Data
- Representative EDS Products for Combustion Emission Testing
• Develop Integrated Environmental and Health Assessment Data Base
- Chemical and Physical Properties
- Bioassay and Occupational Exposure Data
100
-------
Not included in the ECLP demonstration is processing of the vacuum
bottoms material which consists largely of 1000°F+ liquids, unconverted
coal and coal mineral matter. Work is in progress to evaluate the use of
bottoms partial oxidation processing for hydrogen/fuel gas generation and
direct combustion of bottoms for plant fuel. Conceptual Commerical Plant
Study designs carried out to date have utilized FLEXICOKING for vacuum
bottoms processing. FLEXICOKING, a commercial petroleum process that
employs integrated coking and gasification reactions in circulating beds,
recovers essentially all of the feed carbon from the bottoms material as
product liquid or plant fuel gas. A small amount of carbon is purged from
the unit with the coal mineral matter. Leachate tests have been performed
on the solids from FLEXICOKING to identify any problems requiring resolu-
tion (3). Environmental assessments will need to be carried out for the
other bottoms processing/ utilization options being developed for the EDS
process.
PROGRAM HIGHLIGHTS
Conceptually the program consists of related environmental and health
monitoring, testing, engineering studies and assessments. Specific activi-
ties within the EDS Environmental Program include monitoring and testing
of process streams and occupational exposures as well as engineering and
laboratory studies of environmental controls. The following summarizes the
major activities in each of the environmental areas; of air, water, solid
wastes, human health, ecology and product utilization highlighted in Table
2.
AIR EMISSIONS
The air emissions activity consists of compliance monitoring associ-
ated with the large pilot plant (ECLP) operations at Baytown, Texas, design
studies to define control technology options for criteria pollutants in
conceptual commercial plants and in-plant testing to characterize noise and
process emission sources. The focus of the pilot plant test program is to
assess fugitive, particulate, and potentially toxic emissions during both
normal and intermittent operations to provide a data base for environmental
assessments for future plants and the design of emission control facilities
where needed.
WASTEWATER TREATMENT
Treatment of all process and other water effluent streams from the
ECLP operations is being carried out in the adjacent Baytown Refinery
facilities as provided in the environmental permit for the pilot plant.
An extensive in-plant test program is underway to monitor and characterize
raw process water streams for variability, composition (including trace
101
-------
TABLE 2. EDS ENVIRONMENTAL PROGRAM HIGHLIGHTS
• Air Emissions
- Control Technology Options for Criteria Pollutants
- In-plant Testing
- Assessment of Fugitive H/C Emissions
• Wastewater Treatment
- Characterization of Raw Process Streams
- Treatability Studies
- Bench Scale Testing
t Solid Waste Disposal
- Physical and Chemical Properties
- Solid Waste Management Techniques
0 Occupational Health
- Seven Phase Worker Protection Program
- Workplace Monitoring
- Medical Surveillance
» Toxicity
- Acute, Subchronic, Chronic Testing
- Environmental and Product Streams
- Human and Ecological Systems
• Product Utilization
- Raw EDS Products
- Combustion Emissions
102
-------
metals), and treatability. Offsite bench-scale treating tests will be
carried out on samples from large pilot plants to establish the water
treatment requirements for a commercial plant. This work will serve to
confirm the basis for commercial plant design studies being conducted in
parallel with the test program to define quantities and streams for a
commercial plant. An independent inplant test program has been completed at
ECLP by the U.S. Environmental Protection Agency (EPA) to serve as a data
base for EPA's research activities in direct coal liquefaction.
SOLID WASTE DISPOSAL
Solid waste management techniques and requirements are being developed
as part of a study design activity for a conceptual commercial plant.
In addition, in-plant test work to characterize the solid wastes on all
project coals will be carried out to determine handling and disposal
properties.
OCCUPATIONAL HEALTH
A seven-phase program involving engineering controls, industrial hy-
giene, operations and laboratory work practices, personal hygiene, medical
surveillance, and health education forms the basis for the ECLP Occupa-
tional Health Program (4). Specific goals are to assure a safe and healthy
work environment at ECLP and to provide an expanded data base for future
production facilities. The industrial hygiene data base being generated
includes pre-startup and periodic baseline surveys, routine monitoring of
process and mechanical personnel, and area monitoring of special operations
such as maintenance. Over 1200 personal and area samples have been gener-
ated during the first of three program coals. An independent industrial
hygiene in-plant survey has been carried out at ECLP by the National
Institute of Occupational Health and Safety (NIOSH) to support NIOSH
research in direct coal liquefaction.
TOXICITY
The goals of the Toxicity Program are to 1) identify toxic hazards
to either human health or ecological systems, 2) assess the risks those
hazards present, and 3) assess the commercial readiness of the EDS process
technology in the light of those risks and hazards. The program provides
for analytical characterization, and in-vitro and invivo testing of samples
of EDS product, process and waste streams. The testing will encompass the
following: acute oral, dermal and inhalation toxicity; eye and skin irri-
tation; skin sensitization; mutagenicity; carcinogenicity; subchronic
toxicity, teratology and reproductive effects; and fish and daphnia
toxicity, daphnia and algae growth and inhibition.
103
-------
PRODUCT UTILIZATION
Downstream processing/refining of EDS products and subsequent mar-
keting and use of such upgraded products is outside the present scope of
the EDS project. However, the middle distillate (LSFO) and vacuum gas oil
(VGO) products can be used directly or as a blendstock for existing pe-
troleum products. Combustion emission testing of EDS fuel oil blends has
been initiated (5). In addition, all products will be evaluated for toxic
hazards to human health and ecological systems recognizing the need for
handling and transporting of EDS products from a production facility.
STATUS AND OUTLOOK
The environmental data base being generated within the EDS project
is designed to complement programs being carried out in cooperation with
government agencies. In this manner, the EDS process is expected to meet
commercial environmental design requirements and resolve present concerns
for the class of materials which exist in coal liquefaction plants.
The program status is highlighted in Table 3. To date, 3900 hours of
operation on Illinois No. 6 coal have successfully been completed at the
large EDS Pilot Plant in Baytown, Texas (6). All environmental data acqui-
sition objectives for this run have been met and a major data analysis and
laboratory investigation effort is underway. Present operating plans for
the pilot plant call for operation on a subbituminous and a lignite coal
with further environmental testing to establish a data base for three
different types of coals.
The EDS process is still evolving with the introduction of bottoms
recycle operations at ECLP in August, 1981, and the work in progress to
evaluate various bottoms processing and utilization options. Environmental
data acquisition efforts will be integrated into these process development
areas consistent with the overall strategy of the EDS Environmental Pro-
gram.
As presently funded, the EDS project will terminate June 30, 1982,
with the subsequent dismantling of ECLP and completion of the EDS environ-
mental work outlined in this paper. Under the terms of the EDS Cooperative
Agreement, work of a non-proprietary nature is to be made available to
the EDS Project Sponsors. The reporting system for the EDS Project con-
sists of monthly, quarterly, and annual technical reports and assures that
all technical contract data for the EDS Environmental Program will be in
the public domain through DOE sponsorship of the project.
104
-------
TABLE 3. EDS Environmental Program Status
t Illinois Coal Study Design and Pilot Plant Operations Complete
• Wyoming Coal Study Design and Pilot Plant Operations Underway
t EDS Process Still Evolving
- Bottoms Processing Studies
- Bottoms Recycle Under Demonstration at ECLP
- Product Utilization Emphasis on Distillate Fuels
• Data Analysis and Laboratory Work will Continue for Three Types
of Coals
105
-------
REFERENCES
1. Epperly, W. R., "Running Government-Industry Projects," CHEMTECH, 1981,
pp. 220-223.
2. , "EDS Commercial Study Design Update, Illinois Coal,"
FE-2893-62, February, 1981.
3. Boyer, G. T. et al, "Water Pollution Control in the Exxon Donor Solvent
Coal Liquefaction Process," AIChE 87th National Meeting, Boston, Mass.,
August 21, 1979.
4. Montgomery, C. Hunter, "Coal Liquefaction Occupational Health Program,"
65th Annual Meeting of the American Occupational Medical Association,
Detroit, Mich., April 21-25, 1980.
5. Ryan, D. F., Panzer, J., "PNA Emissions From Combustion of EDS Fuel
Oils," 6th Annual EPRI Contractor's Conference on Coal Liquefaction,
May 13-14, 1981.
6. Brackett, R. H., Clunie, T. J., Goldstein, A. M., "Exxon Donor Solvent
Coal Liquefaction Process: Exxon Coal Liquefaction Plant Operations,"
6th Annual EPRI Contractors' Conference on Coal Liquefaction, May 13-
14, 1981.
## ## ##
106
-------
SAMPLING AND ANALYSIS OF PROCESS ANDt
EFFLUENT STREAMS FROM THE EXXON
DONOR SOLVENT COAL LIQUEFACTION PILOT PLANT
by: Mark Notich and Jung Kim
Hittman Associates, Inc.
9190 Red Branch Road
Columbia, MD 21045
ABSTRACT
Under contract to the U.S. Environmental Protection Agency (EPA),
Hittman Associates, Inc. performed a sampling and analysis of process
discharge streams from the Exxon Donor Solvent (EDS) coal liquefaction
plant in Baytown, Texas. Twenty-four streams were sampled and 2,200 sam-
ples were returned to Hittman1s laboratory for analysis. The chemical
analyses of these samples included water quality parameters, GC/MS, GC/FID,
and bioassays. Analyses were also performed to determine the accuracy and
precision of the data and to determine the variability of stream components
due to process variations. Preliminary results are available and data
evaluation for the Source Test and Evaluation Report is underway.
INTRODUCTION
The EPA Industrial Environmental Research Laboratory is developing a
data base in support of EPA's synfuels program. This data base includes
data obtained through sampling and analysis of environmentally significant
waste and process streams from existing synfuels facilities. Environmental
data acquired in this program will be used to assess the environmental
impacts of synthetic fuels plants and evaulate the effectiveness of control
technologies.
The Exxon Donor Solvent process is one of several processes used to
convert coal to liquid fuel which is under investigation. In this process
a "donor solvent" is first hydrogenated and then mixed with pulverized
coal and hydrogen. Hydrogen is transferred from the donor to components of
the coal, thereby liquefying the coal. Subsequent fractionation of the
resulting mixture yields hydrocarbon products. The donor solvent is sepa-
rated and recycled for hydrogenation. The EDS process is being studied at
the Exxon Coal Liquefaction Pilot Plant (ECLP) in Baytown, Texas. Hittman
Associates, Inc. performed sampling and analysis of the plant's process
discharges. The results of the analysis will be used by EPA to assess the
environmental impacts of the EDS process. It should be noted that although
the pilot plant represents a commercial facility, there are significant
107
-------
differences. Three major differences are: (1) the pilot plant has no
wastewater treatment facility; all sour water streams are combined and sent
to the adjacent Baytown Refinery wastewater treatment plant; (2) acid gas
(H9S) removed from the gaseous streams also is treated by the refinery
sulfur recovery system, while a commercial facility would have its own
sulfur recovery system; and (3) the vacuum bottoms (carbonaceous residue)
is drummed and stored at the pilot plant, whereas in a commercial facility
this would be either treated in a Flexicoker® or gasified to produce
hydrogen.
The criteria used to select the ECLP streams to be sampled are pre-
sented in detail in the EDS Test Plan. (Hittman Associates, Inc. Envi-
ronmental Test Plan for the EDS Pilot Plant in Baytown, Texas. EPA Con-
tract No. 68-02-3147, February 1981). The intent was to select streams
which would be found in a commercial facility or would be similar to such
streams and were significant either to potential environmental impacts or
to control technology evaluation. No internal process streams were sampled.
The selected streams are listed in Table 1. They include 15 sour water
streams and the combined sour water that leaves the ECLP for treatment,
naphtha, light solvent fuel oil, combination product, feed coal, vacuum
bottoms, and several gaseous streams relevant to control technology evalu-
ation. The sampling program consisted of three separate efforts: (1)
collection of composite samples over a three-day period for each of the
selected streams; (2) collection of a set of samples from six of the
streams to determine process, sampling, and analytical variability; and (3)
collection and on-site analysis of the gaseous samples on a one-time only
basis. The primary liquid samples from the ECLP plant were split,
composited, preserved, and returned to the Hittman Laboratory for analysis.
The analytical program was based on a combined Level I/Level 2
methodology using a directed analytical approach. The combined methodology
was adopted because in conducting consecutive Level I/Level 2 analyses, the
time interval between the two efforts allows for major changes in the
facility, particularly in the case of pilot operations. A directed analyti-
cal approach was chosen because it permits complete analyses of a selected
group of high-priority streams which guide the analyses of components of
the priority streams. A complete discussion of the analytical program is
presented in the EDS Test Plan. A paper devoted to the EDS analytical work
is included in this symposium (Higman, et al. "Problems Associated with the
Analysis of Synfuel Products, Process, and Waste Water Streams").
PROCESS DIAGRAM AND SAMPLE POINTS
The first step of the EDS process is coal preparation. Figure 1 shows
the coal preparation area. Coal is transported to the plant via a bottom-
dump rail car and taken to a 5,000-ton storage silo. The coal is then
crushed and dried before entering the slurry drier tank.
The crushed coal is mixed with recycle solvent and fed to the slurry
drier (Figure 2). The coal-solvent mixture is pumped, along with hydrogen,
to the preheat furnace and then to the liquefaction reactors. These
reactors are kept at 840°F and 1,900 to 2,000 psig. The off-gas from the
108
-------
TABLE 1. hCLP SAMPLE FOiNTS
Aqueous Sample Points
Stream
Process Area
Sour Water - Recycle Gas Cold
Separator Drum
Rich DEA - Liquefaction DEA
Scrubber
Scrubber Water - Recycle Gas
Water Scrubber
Sour Water - Atmospheric
Fractionator
Cold Sour Water - Atmospheric
Fractionator
Sour Water - Steam Ejector Con-
densate Pump
Scrubber Water - Water Scrubber
Unit
Rich DEA - DEA Scrubber
Condensed Water - P-302 & P-304
Rich DEA - Hydrocarbon Skimming
Drum
Lean DEA - DEA Regenerator
Sour Water - Fuel Gas DEA
Scrubber Sour Water Pump
Scrubber Water - Acid Gas
Water Scrubber
Slurry Drying and Lique-
facti on
Slurry Drying and Lique-
faction
Slurry Drying and Lique-
faction
Product Distillation
Product Distillation
Product Distillation
Solvent Hydrogenation
Solvent Hydrogenation
Solvent Hydrogenation
Fuel Gas Treating and DEA
Regeneration
Fuel Gas Treating and DEA
Regeneration
Fuel Gas Treating and DEA
Regeneration
Fuel Gas Treating and DEA
Regeneration
109
-------
TABLE 1. (CONTINUED)
Solid Sample Points
Stream
Feed Coal
Vacuum Bottoms
Stream
Naphtha
Light Solvent Fuel Oil
Combined Product
Process Area
Coal Prep
Product Distillation
Product Sample Points
Process Area
Solvent Hydrogenation
Solvent Hydrogenation
Solvent Hydrogenation
Gaseous Sample Points
Stream
Offgas - DEA Regenerator
Offgas - Fuel Gas Condensate
Separator Drum
Acid Gas to Refinery
Offgas - Fuel Gas DEA Scrubber
Process Area
Fuel Gas Treating and DEA
Regeneration
Fuel Gas Treating and DEA
Regeneration
Fuel Gas Treating and DEA
Regeneration
Fuel Gas Treating and DEA
Regeneration
110
-------
COVERED CONVEYOR
DUST COLLECTION SYSTEM
COAL CRUSHING
TO SLURRY DRIER
1 - Feed Coal
Figure 1. Coal Preparation Area
reactors is separated into a vapor stream and a slurry stream. The
vapor stream is condensed, yielding sour water, hydrocarbons, and
an off-gas stream. The sour water stream goes to the sour water
disposal tank. The condensed hydrocarbons are mixed into the
slurry stream and sent to the atmospheric fractionator. The off-gas
from the separator drums is scrubbed with DEA and water and recycled
back to the process.
CRUSHED COAL
FRESH SOLVENT
HYDROGEN
VENT TO FLARE
t
SLURRY
DRIER &
PREHEATER
I
COAL/
SOLVENT
SLURRY
LIQUEFACTION
REACTORS
8kO°f
1920 psig
REACTOR
OUTPUT
~~~
r
WATE
RECYCLE
GAS
f
GAS/LIQUID
SEPARATION
® <£>
R CD T
* RICH
TO
ATMOSPHERIC
FRACTIONATOR
3FA
SOUR WATER
LEAN DEA
SOUR WATER
2 - Sour Water - Recycle Gas Cold Separator Drum
3 - Scrubber Water - Recycle Gas Water Scrubber
4 - Rich DEA - Liquefaction DEA Scrubber
Figure 2. Slurry Drying and Liquefaction Area
111
-------
The slurry stream is fed to the atmospheric fractionator, where it is
separated into atmospheric bottoms, naphtha, atmospheric light gas oil, and
off-gas (Figure 3). The off-gas is condensed and separated into sour
water, condensed hydrocarbons, and raw fuel gas. The atmospheric bottoms
are fed to the vacuum fractionator, where the off-gas, light and heavy gas
oil, and vacuum bottoms are separated. The products from the atmospheric
and vacuum fractionators are combined and fed to the solvent hydrogenation
section for further processing.
STEAM
UNREACTED COAL
AND SOLIDS
SLURRY
OFF GAS
OFF GAS
LIQUIDS FROM REACTORS
GAS OIL
PRODUCT
TO SOLVENT
HYDROGEN-
ATION
SOUR WATER
VACUUM BOTTOMS
5 - Sour Water - Atmospheric Fractionator
6 - Cold Sour Water - Atmospheric Fractionator
7 - Vacuum Bottoms
8 - Sour Water - Steam Ejector Condensate Pump
Figure 3. Product Distillation Area
The output from the product distillation area is mixed with hydrogen
and fed to the hydrogenation reactors (Figure 4). These reactors consist
of four fixed-bed reactors containing a nickel-molybdate catalyst. The
reactor output is separated into hydrogen-rich gas, sour water, and a
hydrotreated liquid stream after passing through hot and cold separator
drums. The hydrogen-rich gas is scrubbed with DBA and water and the
hydrogen is recycled back to the process. The solvent fractionator
separates the hydrotreated liquids into naphtha, light solvent fuel oil,
gas oil product, fresh recycle solvent, and raw fuel gas.
112
-------
RECYCLE HYDROGEN
RECYCLE GAS
LIQUIDS
FROM
PRODUCT
DISTIL-
LAT 1 ON
*•
H
GUARD
CHAMBER
AND
FEED
SURGE
DRUM
CDROG
I'l
EN
1 . REACTOR
HYDROGEN-
ATI ON
REACTORS
OUTPUT |
*" GAS/LIQUID
„ SEPARATION
WATER
SPENT CATALYST |N ,®
(INTERMITTENT ,, R,CH c
DISCHARGE) SOUR WATER
L 1 QU 1 D
CARBONS
1
SOLVENT
FRACTION-
ATOR
& 7
1
SOUR
EA WATER '
COME
PROC
LI
SOL
Wl
NAPH
INED
UCT
GHT
WENT
§>
3
.D
THA
9 - Scrubber Water - Water Scrubber Unit
10 - Rich DEA - DBA Scrubber
11 - Condensed Water from P-302 and P-304
12 - Light Solvent Fuel Oil
13 - Naphtha
14 - Combined Product from E-306
Figure 4. Solvent Hydrogenation Section
The rich DEA from the DEA scrubbers is pumped to the fuel gas treat-
ing and DEA regeneration section (Figure 5). The DEA is fed to the DEA
regenerator, where it is stripped of I^S and S0£ and then returned to the
process. The stripped acid gas is water scrubbed and sent to the refinery
for sulfur recovery. The raw fuel gas is water scrubbed and then DEA
scrubbed before being used as fuel gas for the process.
RICH DEA © *
WASH WATER *•
FUEL GAS
TREATING M
DEA
REGENERATIC
AREA
^
D
)N
ft
(fy $3) FUEL GAS
TO PROCESS
<& ^ LEAN DEA
TO PROCESS
15
16
17
18
19
20
21
22
23
SOUR WATER
Rich DEA - Hydrocarbon Skimming Drum
Wash Water Input
Offgas - Fuel Gas Condensate Separator Drum
Offgas - Fuel Gas DEA Scrubber
Lean DEA - DEA Regenerator
Sour Water - Fuel Gas DEA Scrubber Sour Water Pump
Scrubber Water - Acid Gas Water Scrubber
Offgas - DEA Regenerator
Acid Gas to Refinery
Figure 5. Fuel Gas Treating and DEA Regeneration Section
113
-------
All of the sour water and scrubber water generated by the process is
pumped to the sour water collection section (Figure 6). The sour water
is then pumped to the refinery's sour water stripper.
SOUR WATER
FROM PROCESS
w
SOUR WATER
COLLECTION
DRUM
TO
(%) SOUR
WATER
STRIPPER
24 - Sour Water - Sour Water Disposal Pump
Figure 6. Sour Water Collection Section
SAMPLING PROGRAM
PRE-SAMPLING ACTIVITIES
To accommodate a sampling effort of this size and scope, a field
laboratory had to be established. The chosen facility was an empty,
2,500 sq.ft. warehouse located 1/2 mile from the pilot plant. This
building was the central point for all sample splitting, preservation,
packaging, shipping, and on-site analysis.
To reduce the work load for the field team, as much preparatory work
as possible was done at the home office. A field manual was compiled
which provided exact instructions on the handling, preservation, and
shipment of each sample. Each sampling team member was assigned a spe-
cific task during the sampling effort. All sample bottles, 2,200 in all,
were pre-cleaned and labeled before shipment to the field laboratory.
The on-site analysis called for the use of a gas chromatograph to
analyze gaseous grab samples. These samples had to be analyzed within
one hour after sampling in order to meet holding-time requirements. An
experienced chemist with a GC background was assigned to these analyses.
All necessary equipment and chemicals were delivered to the field
laboratory at least four days before sampling began. This provided time
for the field team to check over the equipment and prepare any necessary
reagents.
SAMPLING SCHEDULE
With the exception of the gaseous samples, samples were collected
twice daily, at 8:00 a.m. and 8:00 p.m., on three consecutive days.
Samples for the process variability program were collected during the
appropriate sampling period along with the composite samples. The sam-
pling schedule is detailed in Table 2.
114
-------
TABLE 2. ECLP SAMPLING SCHEDULE
Day 1
A.M.
Composite Samples Composite Samples Composite Samples
Process Variability Process Variability
Samples Samples
P.M. P.M. P.M.
Composite Samples Composite Samples Composite Samples
Process Variability
Samples
Composite samples contained equal aliquots from all six sampling
periods. The analytical result for each component from this composite is
the average value of that component over the six sampling periods. Process
variability samples are not composited but are distinct samples represent-
ing individual sampling periods. The analytical results from these samples
track certain components to determine how the concentration varies with
changing process conditions and other factors.
IN-PLANT AND FIELD LABORATORY ACTIVITIES
Liquid samples were collected in 5-gallon and 1-gallon bottles. The
5-gallon bottles were used to collect composite and process variability
samples, while the 1-gallon bottles were used only for composite samples.
Volatile Organic Analysis (VOA) samples were taken in 40 ml septum-top
vials and sampled in duplicate. Feed coal and vacuum bottom samples were
collected in 2-liter, brown-glass, wide-mouth bottles.
Once all the samples from a given sampling period were obtained, they
were immediately returned to the field laboratory for processing. This
phase included sample splitting and preservation. Samples from the 5- and
1-gallon bottles were split into smaller bottles for two reasons, first, to
allow for required preservation steps, and second, to make sample handling
easier for laboratory personnel. Thus, there was less chance for sample
degradation and errors in handling and analysis. Each composite sample
bottle and process variability sample bottle was pre-labeled. These labels
contained the stream name, intended analysis, preservation method, and
aliquot volume required. Having all the bottles labeled with the proper
information enabled the field team to perform production-line sample split-
ting.
Preservation of the samples for shipment and subsequent analysis was
very important. Every precaution was taken to properly preserve the sam-
ples and to reduce the degradation of the chemical species of interest.
115
-------
Samples were preserved in accordance with the procedures defined in Manual
of Methods: Preservation and Analysis of Coal Gasification Wastewaters,
(Luthy, Richard G.).Each aliquot that was split into sample bottles had
to be preserved, and most of the 2,200 samples required chemical preserva-
tion. These preservation procedures were repeated six times on approximate-
ly 1,500 bottles.
Packaging and shipment was the last procedure that the samples were
subjected to at the field laboratory. Holding-time requirements dictated
that the volatile organic analyses samples be delivered overnight to the
analytical laboratory. The samples also had to be kept at 4°C during
shipping to meet preservation requirements. The samples were packed in
styrofoam shipping coolers with packing material and ice just prior to
pickup by the shipper. To avoid the loss of a sample due to breakage in
transit, all samples were prepared in duplicate and shipped so that dupli-
cates were in separate coolers. Composite samples were stored in ice
during the 3-day sampling period while compositing was being completed.
With these packaging procedures, only four of the 2,200 bottles were lost
or broken.
PROBLEM AREAS AND SOLUTIONS
There are many problems associated with a sampling effort of this
size. The best way of avoiding difficulties is to identify potential
problem areas and determine what precautions can be taken. Three areas
which Hittman identified as potential problems were:
• Fumes and vapors from the acidification of sour water samples
containing high levels of sulfur
• Keeping the samples at 4°C for an extended period of time
• Properly packaging and shipping the samples.
Since acidification with concentrated nitric or sulfuric acid is
required for several species, any evolution of HLS from the samples could
present a health hazard. A glove box was converted into a sealed-hood
system with vacuum pumps to draw the gas out of the box and through two
scrubbing bottles containing 15 to 25% NaOH. The scrubbed gas was pumped
to the outside of the field laboratory. Industrial fans were located so
that H«S fumes and other hazardous materials were prevented from accumu-
lating in the field laboratory.
Samples were kept in a large walk-in dumpster converted into a cooler.
Layers of 1-inch polystyrene were attached to the walls and floor and
covered with thick plastic. A roof was installed and insulated with poly-
styrene and plastic. The dumpster was 24 feet long, 6 feet wide, and 4
feet high. It required between 800 and 1,000 pounds of block ice per day
to keep the samples at 4°C. Refrigerated trucks were not suitable because
of the danger of contamination in the event of a sample spill.
The packaging and shipment of such a large quantity of bottles is
subject to both mishandling and breakage. This problem was addressed by
116
-------
having the samples duplicated, split, and shipped in different coolers. In
this way, if a cooler was lost in shipment or damaged, sufficient sample
would still be available in the other cooler. Two members of the sampling
team were assigned full-time to packaging and coordinating sample shipments.
ANALYTICAL PROGRAM
The EDS analytical program consisted of two areas: composite sample
analysis and variability sample analysis. The analyses of the composite
samples included a wide range of chemical tests, while the variability
analyses were limited to four tests. Results from the composite samples
will provide an overall picture of the plants operation during the three
days of sampling. Results from the process variability samples will pro-
vide information on the sensitivity of certain species to process variations
The analyses performed on the composite samples are listed in Table 3.
TABLE 3. EDS COMPOSITE SAMPLE ANALYSES
Inorganics and Water Quality Parameters
CN~ Cl" TSS Phenolics BOD
NH3+ Fl" TDS Oil & Grease Trace
Metals
S~ Alkalinity Total N TOC SCN~
NO /NO Acidity Total S COD SO,:
£• O •
Organics Bioassays
VGA GC/FID Ames Test RAM Test
GC/MS HPLC CHO Cytotoxicity Fathead Minnow
Daphnia
The variability analyses performed are listed in Table 4.
TABLE 4. EDS PROCESS VARIABILITY ANALYSES
Total sulfur
Total Nitrogen
Trace Metals
GC/FID
Organics
GC/MS
117
-------
These analyses will aid in defining the cause of differences in test
results due to process and sampling variability, analytical accuracy, and
analytical reproducibility.
Process variability is the result of variations in process operating
parameters during the sampling period. Variations are due to changes in
coal feed rate, solvent recycle rate, temperature, pressure, and other
operational parameters. If the plant has not reached process equilibrium
before sampling is initiated, sample variability will result from non-
steady state conditions.
Sampling variability results from non-reproducibile samping tech-
nique (e.g., non-isokinetic sampling or sampling of non-homogeneous
streams).
Analytical variability in precision results from non-homogeneity of
sample, minor variations in technique, etc., while variability in accuracy
is normally the result of poor recoverability during extractions.
The determination of the variability due to these four factors is
illustrated in the branch diagram in Figure 7.
Day 1
Day 2
Day 3
Process
Variability
Sampling
Variability
Analytical
Variability
Figure 7. Process Variability Branch Diagram
The application of this diagram can be more clearly seen when analy-
tical results are presented with it (Figure 8).
PRELIMINAEY RESULTS
Preliminary results for the sample sour water - atmospheric frac-
tionator are presented in Table 5. This stream is the condensed water
from the reflux drum of the atmospheric fractionator. A process block
diagram of this sample is provided in Figure 2, Product Distillation
Area. The results are from the six-period composite sample.
In the Source Test Evaluation Report on the EDS pilot plant, all
results will be presented as a range. This range will be determined on
the basis of the analytical error derived from the variability analyses.
118
-------
DAY 1 AM
DAY 2 AM
DAY 3 PM
Aluminum
Antimony
Arsenic
Barium
Beryllium
Bismuth
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Potassium
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Tungsten
Uranium
Vanadium
Zinc
Al
Sb
As
Ba
Be
Cd
Ca
Cr
Co
Cu
Fe
Pb
Mg
Ma
Hg
Mo
Ni
P04
K
Se
Si02
Ag
Na
Sr
Sn
Ti
W
U
V
Zn
4.64
L
1.60
L
L
L
0.089 0,091
0.060
0.112
L L
0.058 0.031
L L
0.38 2.92
3.82
L
1.72
0.004
L
L
(a) (b) (C)
0.22 0.25 L
L L L
0.004 0.003 0.004
0.023 0.020 0.024
L L L
L L L
474. 480. 486.
L L L
0.92 0.89 0.91
L 0.082
L L
L L
1.42 0.35
L L
L
L
L
0.32
L
0.078
0.026 0.021
0.097 0.087
L
L
L
0.37
2.76
L
1.90
0.003
L
L
L
0.038
4.45
L
2.32
0.004
L
L
0.23 0.28
L L
0.003 0.003
0.020 0.025
L L
L L
454. 448.
L L
0.95 0.89
0.14 L
0.22 0.20
0.10 L
0.110 0.095
0.019 0.019
0.086 0.098
L L
0.043 0.027
L L
0.45 0.38
4.52
L
1.66
L
L
0.010
(b)
0.24
L
0.003
0.025
L
L
452.
L
0.86
0.040
L
L
0.20
L
0.098
0.014
0.088
L
L
L
0.49
3.30
L
1.74
L
L
0.029
(c)
0.25
L
0.003
0.027
L
L
468.
L
0.96
L
L
L
0.25
L
0.110
0.015
0.088
L
L
L
0.56
5.21
L
2.11
0.003
L
0.010
detection
limit
0.15
0.15
0.002
0.001
0.003
0.50
0.01
0.025
0.01
0.03
0.02
0.015
0.03
0.08
0.001
0.003
0.002
0.04
0.025
0.40
0.01
0.08
0.03
0.10
0.001
0.03
0.006
L L L L L L L 0.01
0.068 0.088 0.066 0.240 0.110 0.150 0.072 0.015
(- cannot be analyzed by ICAP)
(L = less than detection limit)
(All units are mg/1)
Figure 8. ICAP Analysis of Sour Water From Recycle Gas Cold Separator
Drum, Process Variability EDS Samples
119
-------
TABLE 5. PRELIMINARY ANALYTICAL RESULTS, SOUR WATER -
ATMOSPHERIC FRACTIONATOR
Water Quality Parameters
COD
TOC
TDS
TSS
Alkalinity (as CaCO.,)
Cl"
Fl"
NH3
S=
Oil and Grease
N°3
so4=
SCN"
Phenolics
Aluminum
Boron
Calcium
Iron
Magnesium
Potassium
Sodium
Zinc
Total S
Total N
Concentration (mg/1)
93,700
27,000
678
31
5,020
122
8
1,730
188
<20
0.
63
240
18,000
0.
0.
2.
1.
0.
0.
2.
0.
1,640
1,990
15
024
054
61
90
085
16
16
091
Organic Analysis - Major Components
GC/MS - Acid and Base/Neutral Extracts Analysis
Pheno1 Aniline/Methyl Pyridine
GI Phenol Benzofuran
C Phenol
120
-------
Organic Analysis - Major Components (Continued)
GC/MS - Volatile Organic Analysis
Butane Propyl Nitrile
Pentane Toluene
C-6 Alkanes Methyl Pyrole
C-7 Alkanes Methyl iso-butyl ketone
Ethyl Nitrile
Bioassays E^cn
Ames Test Not determined
CHO Clonal Cytotoxicity <6 ul/ral
Assay
RAM Assay <6 ul/ml
Fathead Minnow (LC5Q) 0.047%
Daphnia 0.158%
Figure 9 represents the concentration of phenolics in several process
streams. The level of phenolics is the highest in the condensates from
the separation drums throughout the process. The sour water-atmospheric
fractionator has the highest level of phenolics. Several streams are not
represented in this process diagram, such as those from the fuel gas
treating and DEA regeneration area and the sour water collection section.
These omitted streams generally contain lower levels of phenolics than
indicated in Figure 9.
121
-------
TO FLARE SOUR WATER
ATMOSPHERIC
FUEL GAS TO FUEL RICH OEA TO
On: TRE'.TIIIG SOUR WATER OEA REGENERATOR
16.024
SOUR
"""" WATER
57
LEAN
~ DEA
557
^ SOUR
~ WATER
_^CAS OIL PRODUCT
TO STORAGE
Process areas not included:
Fuel Gas Treating and DEA Regeneration
Sour Water Collection
Figure 9. Phenolics in EDS Pilot Plant Process
(mg/1)
-------
SUMMARY
A pilot plant is not fully representative of a commercial facility.
To obtain the most representative data possible, we sampled only those
streams which we know will be present in a commercial facility. Streams
that are unique to the pilot plant were not sampled. The results obtained
from the analyses of these samples can be scaled up based on the expected
operational conditions of a commercial-scale facility.
The process variability analyses performed as part of this program
were mainly a quality control/quality assurance measure. The data obtained
from the process variability analyses will be evaluated to determine the
accuracy and precision of the analytical results. By identifying the
source of variations in the data, it is possible to reduce errors in future
sampling and analytical programs.
ACKNOWLEDGEMENTS
The authors would like to express their appreciation to the people who
have participated in this effort. We would like to thank Mr. Richard
Thomas and Ms. Diane Husa of Exxon Research and Engineering; Mr. Richard
Brackett and Dr. Tom Clunie of the Exxon Coal Liquefaction Plant; Mr. Bruce
Henschel and Dr. Ray Merrill of EPA; and Mr. James Farley of the Department
of Energy. We would also like to thank Mr. Frank Woods, owner of the
warehouse used as a field laboratory, and all Hittman personnel who con-
tributed to the effort.
123
-------
HEALTH AND ENVIRONMENTAL STUDIES OF H-COAL PROCESS*
by: K. E. Cowser, J. L. Epler, C. W. Gehrs
M. R. Guerin and J. A. Klein
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
ABSTRACT
With the implementation of the Energy Security Act of 1980, coal and
oil shale are expected to be principal sources for petroleum and natural
gas substitutes. H-Coal is one of several processes under intensive study
for the direct conversion of coal to the desired synthetic fuels.
In this paper we describe the health and environmental study program
of H-Coal, sponsored by the Department of Energy. Presented are the re-
sults of the chemical, biological, and ecological characterization of prod-
ucts and by products derived from the operation of a process development
unit. These initial results provide an informed basis for subsequent moni-
toring and testing activities of the nominal 200- to 600-ton/d pilot plant
at Catlettsburg, Kentucky.
*Research sponsored by the Office of Energy Research and the Division of
Environmental Technology, U.S. Department of Energy, under contract
W-7U05-eng-26 with the Union Carbide Corporation.
By acceptance of this article, the publisher or recipient acknowledges
the U.S. Government's right to retain a nonexclusive, royalty-free license
in and to any copyright covering the article.
124
-------
HEALTH AND ENVIRONMENTAL STUDIES OF H-COAL PROCESS
INTRODUCTION
In the next two decades the major production and use of products derived
from coal and oil shale is expected. The primary incentive for such develop-
ment is the imbalance between the domestic supply and demand for oil and gas
and the consequent dependence on oil imports. Clearly a synthetic fuels
(synfuels) industry will increase flexibility in dealing with any future
disruptions in the world oil market.
2
Over TO coal liquefaction processes have been proposed. These can be
classified as indirect liquefaction, direct liquefaction, and pyrolysis. The
Department of Energy (DOE) is devoting considerable attention to direct lique-
faction because of its potential for lower cost. H-Coal is one of the at-
tractive methods of reacting coal with hydrogen in the direct production of
liquid products such as naphtha and fuel oils.
Accompanying the development of energy-producing technologies is the
consideration of potential health and environmental impacts. Recognizing
this need, DOE asked Oak Ridge National Laboratory (ORNL) to develop compre-
hensive environmental and health plans to study the H-Coal process and in
particular the pilot plant at Catlettsburg, Kentucky. Components of the pilot
plant operation applicable to commercial size facilities are to be emphasized.
Similar studies of the solvent refined coal (SRC) process are in progress
elsewhere, complementing the H-Coal activity. Together they will provide a
basis for technology assessments.^
Our study of the H-Coal process is being carried out in two phases.
Phase I involves characterizing and testing materials produced by a process
development unit (PDU); Phase II is a study of the pilot plant. In this paper
we report the results of our Phase I activities and describe the Phase II pro-
gram, which has just begun.
PROCESS DESCRIPTION
H-Coal is a process for the catalytic hydrogenation of coal under high
pressure and temperature to produce liquid hydrocarbon products and fuel gas.
The process was developed initially by Hydrocarbon Research, Inc., with the
use of bench-scale units and a 3-ton/d PDU located in Trenton, New Jersey. A
pilot plant was subsequently constructed at Catlettsburg, Kentucky, with a
nominal capacity of 200- to 600-ton/d, depending upon the operating mode.
Operation of the plant began in 1980 to demonstrate the commercial viability
125
-------
of the process and develop data for the design of commercial units.
In the H-Coal process (Figure l), coal is slurried with a process-
derived oil, pumped to reactor pressure, mixed with recycle and makeup hydro-
gen, and fed through a preheater to the catalytic ( Co/Mo) ebullated-bed
reactor. Typical operating conditions are 2500-3000 psi and 850°F. Catalyst
activity is maintained "by the periodic addition of fresh catalyst and the
withdrawal of spent catalyst, and ebullition is provided by an external pump
that recycles the coal-solvent slurry.
The reactor products leave the reactor and are separated for subsequent
processing. The vapor from the reactor is cooled and scrubbed to produce a
Hn-rich recycle gas and a light hydrocarbon stream fed to the distillation
unit. The liquid-solid product from the reactor, containing unconverted coal,
ash, and oil, is fed to a liquid flash separator. The flashed-off material
is passed to the distillation unit to produce a variety of fuel gases and
light and heavy distillate products. The bottoms products from the flash
separator are further separated in a hydroclone and then in a vacuum distil-
lation unit. A portion of the heavy distillate is recycled to the reactor,
with the heavy bottoms stream from the vacuum distillation unit being utilized
for hydrogen production.
By varying the residence time in the reactor, the process can be designed
to operate in the synthetic crude (syncrude) or the fuel oil mode. To produce
syncrude, more hydrogen is required and there is a lower yield of residual
fuel oil. To produce a low-sulfur residual fuel oil as a major product, the
temperature and pressure in the reactor are lower and less hydrogen is re-
quired. However, a special liquid-solid separation unit, not shown in
Figure 1, will be required.
PROCESS DEVELOPMENT UNIT STUDIES
Elements of our synfuels research program concern the chemical, physical,
and biological properties of hazardous or toxic materials; the environmental
transport and systems to control the release of or to minimize the exposure
to such materials; and the assessment of the consequences of exposure. Num-
erous comparative studies of coal-derived liquids and other related materials
such as shale oils, petroleum crude oils, petroleum products, and various
polynuclear aromatic compounds have been completed and reported, including
research with materials from the H-Coal
The following discussion is limited to the results of several of the more
recent characterization and testing studies of samples from the PDU. These
samples are not necessarily representative of coal liquids that will eventual-
ly be produced in a commercial facility; consequently they are not adequate
for definitive process-specific comparisons. The results are valuable, how-
ever, as indicators of potential problem areas. As such, they provide a basis
for selecting samples and defining studies to be performed witlT pilot plant
materials .
Characterization and_ Cellular Bioassays
An important focal point of our research has been the identification of
126
-------
ORNL-DWG 81-5716
RECYCLE HYDROGEN
COAL
FEED.
FUEL GAS •
NAPHTHA
LIGHT DISTILLATE
HEAVY DISTILLATE
RESIDUAL FUEL
(TO HYDROGEN
PRODUCTION)
Figure 1. H-Coal Process Schematic.
127
-------
the chemical constituents responsible for the potential "biological effects of
synfuel products and processing effluents. An effective approach is
to integrate biological testing with a chemical manipulation of the test
material. Thus, in our comparative mutagenesis program ve have emphasized the
combining of chemical class fractionation with biotesting.
Samples examined in this study were provided by Mobil Research and De-
velopment Corporation and Hydrocarbon Research, Inc., and incorporated into
the Synfuels Research Materials Facility.5 Both raw distillates and products
upgraded by hydrotreatment (HDl) were included. The samples are identified
in Table 1, with information given on their boiling point ranges and ultimate
analyses. Because these samples were not necessarily representative of the
coal liquids that will eventually be produced in a commercial facility, they
were used for generic research into the chemical and biological properties
of petroleum substitutes.
All samples were treated according to the procedure shown in Figure 2.
After removal of the highly volatile matter, the residue was fractionated
into chemical classes with a diethyl ether-aqueous acid partitioning and a
subsequent Sephadex LH-20 separation of the neutral fraction. '' The result-
ing volatiles, insoluble matter, and acidic, basic, and neutral subfractions
were weighed and subjected to bacterial mutagenic testing. Although biologi-
cal screening studies with H-Coal materials have included tests in a bacterial
system (Salmonella typhimurium) and a protozoan system (Tetrahymena
pyriformis) ,°" only the former tests are discussed in this paper.
The results of characterization and mutagenic testing are summarized in
Table 2 by general chemical class and approximate weight and by mutagenic con-
tribution. These results, useful in identifying general trends as opposed to
absolute hazards posed by the test materials, have been discussed extensive-
ly in other publications."'10 For example, the total mutagenicities (the sura
of chemical fractions) of coal-liquid samples that are more volatile (sample
No. 1312) or that have been hydrotreated (sample Nos. 1603 and l6oU) tend to
be lower, and niutagenicity tends to increase with increasing vapor pressure
(e.g., sample Nos. 1313-1315)- These samples exhibit greater mutagenicity
than petroleum crude oils. In addition to the neutral subfractions, the
alkaline components can contribute significantly to the mutagenicity of coal
liquids. Recent evidence indicates that polycyclic aromatic amines and az-
aarenes are unusually bioactive alkaline constituents whereas polar-substituted
neutral polycyclic aromatics are occasionally responsible for high mutageni-
cities.
In-Vivo Mammalian Tests
Whole-animal studies have included the preliminary investigation of both
the acute and chronic toxic effects of coal-derived liquid materials. A total
of five acute toxicity tests was used: determination of the acute toxicity
following oral and interperitoneal administration of the test materials to
mice, acute dermal toxicity in rats, primary skin irritation and eye irrita-
tion in rabbits, and delayed-contact sensitivity in guinea pigs. Skin carcino-
genesis tests involved the repeated application of the test material to the
shaved skin of mice. Test and data analyses procedures have been described
elsewhere.11,12
128
-------
TABLE 1. PROPERTIES OF COAL LIQUID SAMPLES FROM PROCESS DEVELOPMENT UNIT
Identification
Operational Description
PDU ORNL
Run Repository Mode Coal Sample Point Location
1
1
5
7
1601
1602
1603
i6ol*
1617
1618
1619
1308
1309
1310
1311
1312
1313
1311*
1315
Fuel oil 111. H-Coal distillate (raw)
No. 6 HDT at low severity
HDT at medium severity
HDT at high severity
Fuel oil 111. H-Coal fuel oil
No. 6 HDT at low severity
HDT at medium severity
HDT at high severity
Syncrude 111. Atmospheric overhead
No. 6 (ASOH)
Atmospheric bottoms (ASB)
Vacuum overhead (VSOH)
Vacuum bottoms (VSB)
Fuel oil 111. Atmospheric overhead
NO. 6 (ASOH)
Atmospheric bottoms (ASB)
Vacuum overhead (VSOH)
Vacuum bottoms (VSB)
Characteristics
Boiling*
Range (°F)
300-700
200-650
375-1000
375-1000
11+6-590
1*58-650
(66%)
1*92-650
(3W)
172-565
1*10-650
(8155)
1*62-633
(1*650
Hydrogen
9-
9-
10.
10.
9-
9-
9.
11.
9.
8.
1*.
11.
9-
8.
1*.
65
87
5***
CJ***
17
22
67
&
1*
7
3
1+
5
7
7
wt.
Oxygen
1.62
0.95
0.39
0.21
0.91*
0.71*
0.1+1*
%
Nitrogen
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.
39
20
13
09
58
55
30
17
37
1*2
90
30
1*0
50
5
Sulfur
0.
0.
<0,
<0.
0.
0,
0.
0
<0
<0
3
0
0
0
2
,1017
,002
.002
,002
.059
.039
.011
.07
.03
.03
.0
.2
.5
.1
.2
(!)**
CA
57
52
1*5
39
58
52
51
1*1*
*Parenthetical values are percent of volume distillated at highest indicated temperature .
**Aromatic carbon as percent of total carbon .
***Estimated .
-------
U)
o
[ STARTING MATERIAL |
ROTARY EVAPORATION
VOLATILES
NONVOLATILES
I
ETHER/ACID
1
ETHER SOLUBLE
BASES (ETHER)
V
1
BASES (AQUEOUS)
1
pH/ETHER
1
1
j INSOLUBLE
WATER SOLUBLE
BASES (WATER)
T
"BASES"
1
INSOLUBLE
BASES
j
1
SEPHADEX LH-20/ISOPROPANOL
1
1
SATURATES ARC
l
1
3MATICS POLYAROMATICS
-
1
POLARS
)
1
[ACIDS AND NEUTRALS (ETHER)
\
pH/ETHER
1
1
NEUTRALS (ETHER) ACIDS (AQUEOUS)
pH/ETHER
1 1 1
ETHER SOLUBLE WATER SOLUBLE INSOLUBLE
ACIDS (ETHER) ACIDS (AQUEOUS) ACIDS
l. _ _ _ _ >
"NEUTRALS"
"ACIDS"
CHEMICAL CLASS FRACTIONATION FOR BIOTESTING
[MATERIAL!. STEP, (PHASE)
Figure 2. Swain-Stedman Acid-Base Fraction Coupled with LH-20 Separation of Neutral Fraction
-------
TABLE 2. APPROXIMATE MUTAC-EHICITY AND WEIGHT DISTRIBUTIONS BY CHEMICAL CLASS
Mutagenic Activity*
Percentage Contribution
Total
(revert
Acids
Bases
Weight Distribution
Percent of Total
Total
Heutrals Other** Acids Bases Heutrals Other** Recovered
Sample mg-1 )
H-Coal ASB (Syn)
Ho. 1309***
H-Coal VSOH (Syn)
Ho. 1310
H-Coal VSB (Syn)
Ho. 1311
H-Coal ASOH (FO)
Ho. 1312
H-Coal ASB (FO)
Ho. 1313
H-Coal VSOH (FO)
No. 1311*
H-Coal VSB (FO)
Ho. 1315
H-Coal 'Dist'
No. 1601
H-Coal 'Dist'
HDT-L Ho. 1602
H-Coal 'Dist'
HDT-M Ho. 1603
H-Coal 'Dist'
HDT-H No. l6ol*
Wilmington Crude
Ho. 5301
970
21*00
2000
0
ll*0
1*200
6300
350
5l*0
210
0
5
1
0
0
0
0
0
0
0
0
0
0
0
1*6
1*2
66
0
100
25
89
37
0
0
0
0
53
58
17
0
0
75
9
63-
100
100
0
100
0 <1
1
16 2
1
1
1
2 1
0 1
0 1
0 1
0 1
0 2
3
3
5
2
2
2
10
1
2
2
1
1
95
95
30
55
95
95
30
85
85
80
80
90
-------
The results of the acute oral toxicity tests are listed in Table 3.
The IiD5QS of 'the PDU materials were greater than those of the petroleum crude
oil, "but only of moderate toxicity. The trend suggests that oral toxicity
tends to "be lower for the more volatile and HDT coal liquids. No coal liquid
tested exhibited acute lethality in rats when applied to the skin at a dose
of 2 g/kg or produced skin sensitization when applied intradermally. Eye ir-,
ritation was noted with some materials, although it was a reversible effect.
Chronic dermal exposure studies revealed that coal liquids from the PDU
were carcinogenic to mouse skin.-1-5 The most carcinogenic materials were
those of higher boiling range, but a substantial reduction of skin carcino-
genic potential occurred even at the lowest severity of hydrotreatment involved.
Neurotoxic and systemic toxic effects are now being studied.
Ecological Tests
Parallel studies of the acute and chronic effects of PDU materials on
aquatic and terrestrial organisms of different ecological organizational
levels have also been completed.1°»17 This discussion will be limited to the
test results of liquid products in bioassays with freshwater algae (Selenastrum
capricornutum and Microcystis aeruginosa) and freshwater crustacean (Daphnia
magna), the basic screening tools for preliminary comparative studies.
Spills of liquid products derived from coal, oil shale, and natural
petroleum are a potential source of environmental impact. Comparative informa-
tion on transport, dissolution and effects is necessary to define the potential
impacts and the requirements for cleanup. Of considerable interest is the
primary toxic materials which dissolve rapidly into water in the event of an
aquatic spill. Thus, one element of research has focused on water-soluble
fractions (WSFs) of these materials; the results of testing several PDU
materials are listed in Table k.
The WSFs permit testing of the toxic components of oils, which were
prepared by gently stirring the mixture of oil floating on distilled water.
Their effect on photosynthesis by freshwater algae was measured as a concentra-
tion causing 20% inhibition (EC2o) of orSanic carbon uptake in h-h exposures.1"
Values for the coal-liquid WSFs were below those for petroleum WSFs or of
greater potential acute toxicity. Water soluble fractions were also tested
for acute toxicity to Daphnia in standard U8-h bioassays (LC^g) and for
chronic effects in examinations of the lowest concentrations at which signifi-
cant change to reproduction was observed in 28-d exposures (LOEC).1^'20 The acute
effects for the WSFs of coal liquefaction products were larger (LC5QS ranging
from 0.2 to H.6$) than those for the petroleum products; similarly, repro-
duction effects were also larger.
Generally the toxicity of chemical class fractions from coal-liquid
WSFs was found to increase as ether-soluble bases > ether-soluble acids >
neutral subfractions. Phenolic compounds and anilines were determined to
be the most important water soluble components of the coal liquids in terms
toxic effects.
132
-------
TABLE 3. ACUTE TOXICITY IN MAMMALIAN SYSTEM
Sample
H-Coal ASB (Syn) No. 1309**
H-Coal VSOH (Syn) No. 1310
H-Coal ASOH (FO) No. 1312
H-Coal ASB (FO) No. 1313
H-Coal ASOH (FO) No. 131^
H-Coal 'Dist' No. 1601
H-Coal 'Dist1 HDT-L No. 1602
H-Coal 'Dist' HDT-H No. l6oU
Wilmington Crude No. 5301
Oral LD *
(g/kg)5°
3.6
2.5
5.8
2.3
2.6
3.6
U.o
5.5
>16
95% Confidence
Limits
2.U-5.2
1.7-3.1
It. 7-7. 2
1.9-2.6
2.2-3.2
2.8-U.5
3.it-U.7
2.8-7.2
*Dose in grams of material per kilogram of body weight that kills 50% of
animals.
**Numbers following sample names are designations of the ORNL repository.
Abbreviations are identified in Table 1.
133
-------
TABLE U. TOXICITY IN ECOLOGICAL TEST SYSTEMS
Algae Acute
(EC2o
Selenastrum
H-Coal
H-Coal
H-Coal
H-Coal
H-Coal
H-Coal
H-Coal
H-Coal
H-Coal
H-Coal
Sample
ASOH (Syn) No. 130o****
ASB (Syn) No. 1309
VSOH (Syn) No. 1310
ASOH (FO) No. 1312
ASB (FO) Mo. 1313
ASOH (FO) Ho. 131U
'dist1 No. 1601
'dist' HDT-L No. 1602
'dist1 HDT-M Ho. 1603
'dist' HDT-H No. l6oU
Petroleum Ho. 2 diesel fuel
capricornutum
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
15.
.25
,30
.063
019
16
019
1*2
,57
,68
60
Toxicity
)*
Microcystis
aeruginosa
0.
0,
0,
1.
0.
0,
0,
0,
1.
1.
25
,1*6
.26
.13
,1*
.15
.13
.91
.75
,1*
,6
Crustacea Toxicity
Aeute (LC )**
Daphnla
magna
1*
2
0
1
2
0
0
1
1
30
.6
• 5
.2k
.0
.5
.1*
.5
.5
-7
Chronic (LOEC)***
Daphnia
magna
0.
0.
0.
0.
20
58
92
016
25
*WSF concentration (percent dilution) causing 20% inhibition of photosynthesis.
**WSF concentration (percent dilution) that killed 50$ of the organisms in U8-h.
***Lowest WSF concentration (percent dilution) at which significant effects on reproduction were observed in
28-d.
****Numbers following sample names are designations of the ORNL repository. Abbreviations are identified
in Table 1.
-------
PLANNED STUDIES OF PILOT PLANT
Major areas of health and environmental concern for synfuel development
have been described in detail.21"2^ These include consideration of facility
siting, potential degradation of air and water quality, solid waste management,
worker health and safety, and potential public health risks. Extensive federal
legislation exists to cope with these concerns, with the new legislation hav-
ing increased almost exponentially in number since the passage of the National
Environmental Policy Act of 1969.25 However, additional information in these
areas is needed; thus, research programs relate to the perceived issues in-
volving both regulated activities and yet-to-be-defined regulations.
The objective of the H-Coal Environmental and Health Program is to pro-
vide data and information to support analyses and assessments of coal lique-
faction technology. It is not intended as an environmental compliance
activity, because the protection of the worker and environment at the pilot
plant is the responsibility of Ashland Synthetic Fuels, Inc. (ASFl).
Program emphasis is on those aspects of the H-Coal process and those
units that can conceivably be scaled to commercial-size facilities. Process
sampling is thus keyed to the examination of products, effluents, possible
occupational exposures, and the information necessary for control technology
evaluation. Biological screening activities focus on samples representing
material of the greatest potential for human exposure or health effects,
tempered with the results of tests on samples from the H-Coal PDU. Environ-
mental studies complement the process and in-plant studies, with the thrust
on testing product oils and plant effluents, including solid wastes.
PROCESS MEASUREMENTS AND CONTROLS
Sufficient samples and analyses are provided to characterize a few points
in the process streams and nearly all the points of plant effluents and to
assess the efficiency of environmental control devices. The details of pro-
cess sampling and analyses are described in the H-Coal program plan. ° In
general, process sampling strategy provides for the characterization of mat-
erials introduced into the process; minimal sampling of intermediate-process
streams based on considerations of mass flow, scale-up problems, and the
potential for occupational exposure; and final product and waste streams.
Fifty-three sampling points are located to meet our sampling criteria;
twenty-four are built into the plant (e.g., hydroclone overhead), and the
others can be obtained at several preselected points (e.g., coal pile runoff).
Sample collection is targeted to steady-state operation, and because steady-
state operation cannot be determined a priori, several sample suites will be
collected during each coal run. After operational conditions are evaluated,
materials for testing are selected from samples that have been stored under
controlled conditions. The frequency and intensity of sampling and monitoring
are subject to modification as experience dictates.
The analytical procedures and the constitutents or parameters to be
measured were chosen to allow early measurement of traditionally monitored or
suspected materials and to maximize the likelihood of detecting unexpected and
135
-------
hazardous constituents. Results must be adequate to document process condi-
tions, to evaluate the efficiency of environmental control technology, to
identify limitations in sample size or analytical methodologies, to identify
possible biological hazards in potential fugitive emissions, and to assign
priorities to materials for subsequent bioassay.
Each sample can be identified as a process sample, a product (or final
effluent), a fugitive emission, or a solid waste. As shown in Table 5» each
sample is designated a process (I), product (II), fugitive emission (ill), or
solid-waste (IV) sample. Subsets of each category—gases, tars, solids,
etc.— can then be listed with the chemical and physical characterizations to
be performed. Thirty-two classes of analyses are specified, but not for all
samples. For example, the study of oils and tars comprises (l) elemental
analysis; (2) analyses of Environmental Protection Agency (EPA) priority
pollutant trace elements; (3) determination of filterable solids, moisture,
volatile organic compounds, volatile organosulfur compounds, benzo(a)pyrene,
polycyclic aromatic hydrocarbons, and organonitrogen compounds; (k) organic
class analyses; and (5) a bioassay preparation. As with sampling, we view
the analyses strategy as flexible because actual measurements may suggest
curtailing some studies or expanding others.
Environmental Control Technology—
A complete evaluation of two environmental control methods that are
scalable to larger systems.will be attempted. One is the diethanolamine
absorption towers for CQg and H^S removal from the sour fuel gas and vent
gas streams, the other is the sour water strippers for H2S and NR-^ removal
from the wastewater stream. Companion studies will also be made of the re-
maining wastewater system with the operating contractor. A special study of
the treatability of coal liquefaction wastewaters (described below) will also
be undertaken.
Sample Collection Status—
Present plans are to operate the pilot plant in the syncrude operational
mode with at least three different coals. During a ^5-d run initiated on
February IT, 1981, using Illinois No. 6 coal, two sets of samples were collect-
ed and placed in storage. Early in May the plant was brought on stream with
a Kentucky No. 9 coal, and an extensive sample set was taken for the environ-
mental program. Limited characterization and testing of these latter samples
began in late August. During September continuous plant operation with Il-
linois No. 6 coal was achieved, and two additional sample sets were collected.
As this run continues, additional samples will be collected. Selection be-
tween these sets for subsequent study will be made shortly on the basis of pro-
cess conditions.
OCCUPATIONAL EXPOSURE AND EFFECTS
The potential exposure of man in the working environment includes considera-
tion of plant area controls and the effects on man if exposures occur- Moni-
toring and testing activities thus involve the requirements of worker protec-
136
-------
TABLE 5 . DETERMINATIONS CONSTITUTING THE STUDIES
u>
o
l/l
c
-0 §
VI i —
4-> t/)
>» C fl
ro -r- -r- C
(U O C
4J.p-.r- U
•— r— -U
CO i — -4-1 +•» -r-
01 -r- 01 00
_i u- i— o. i— s:
X X
XXX
X X" X
X
X X
X
bAg, As. Ba, Be, Cd, Co, Cr, Cu, Hg, Mo, Ni, Pb. Sb. Se. Tl . and Zn
by the techniques used.
<;As, Ba: Cd, Cr, Pb, Hg, Se, Ag.
Includes extractables from suspended solids.
*:Pertairs to the particulate content.
"Analysis for chemical smog precursors (e.g.,
^Standard components determined in industrial
EP: RCRA protocol extraction procedure (EPA)
acetaldehyde).
hygiene surveillance.
•a
c a>
at — V)
a-r-
i- m to
O -M >)
o »—
O) ns (D
r- S- C
01 LTi
u £
IB eg \
L- 1 O
•*-* Q O
X <£
UJ X
J
Xj
V ytt
X
X
Q
xc x
X
X
X
•o
c
o
E
o
u
c
s_
0
•r- -f- -O
O I- O
r— >> O.
Q) O O
4-» -^ * U
10 C 0
•— <0 i- C
3 Ol C Of
VI (J U (D Ol
O
VI 4-» U S-
Ol «O i— •*-
Q. f^ . • rOi — i — tC CD
CMO 4->fOO1
0. C r—
gO Oi IS
•^ c c
u -*J •»- ra
(d .2*
c c c ui
a> o -r- vi
CT ^- i- IO
o 4-> -a •—
-E S- i- 0
C T3 C
^
V)
01
ID
.O
s_
o
•o
ID
><
Ol N IO CT X
1_ 'r- 4-> 1-
O l/l */> O
X
X
X
X
X
X
X X
which are
01
X
X
c
1o
L
a.
Ol
i.
a.
10
ul
> m
1— S-
c_ °-
IO Ol 01
UI i.
t/> s- a.
s: oi
•^» ^ — 4-J
C3 £ IO
m s-
o x
0) (O 01
t- OJ-S.
•r- i_ CL
Q
-------
tion and the potential effects of exposure to primary effluents and fugitive
emissions.
Plant Area Sampling and Characterization
The primary objective of an industrial hygiene program is to recognize,
evaluate, and control exposures that may have the capability of producing
untoward health effects. ASFI has prime responsibility for protecting the
health of its employees, and we have participated by complementing ASFI's
requirements and providing information for occupational health control as-
sessments .
Two types of monitoring of potential exposures are provided. Area moni-
toring for particulates, fugitive emissions, and various physical and chemical
stresses indicates possible exposures whereas personnel monitoring defines
the actual exposures. New capabilities in monitoring pollutants associated
with tars and oils have been demonstrated and will be used in program imple-
mentation.2''' These include portable instruments with the real-time measure-
ment capabilities listed in Table 6 to assist in the selection of sample sites
and in the determination of residual worker contamination. A variety of stand-
ard industrial hygiene techniques employing filter cassettes and gas badges
will also be used to define the time-weighted exposures to organic vapors and
particulate contamination.
Occupational Toxicology
The principal focus of our occupational toxicology studies is on the
testing of products, primary effluents, and potential fugitive emissions to
estimate the effects on man. Questions to be answered concern:
the relative toxicity of products, by-products, and effluent;
toxicity variation with process conditions; and
the potential for work-place toxicity.
A tiered or multilevel approach will be used in the investigation of
these questions of toxicity, which will be guided by the results with PDU
materials. Level one tests, or cellular mutagenic bioassays, are the initial
screens to ascertain the relative toxicitv of materials of interest and the
need for further testing and to correlate with whole-animal somatic effects.
Level two tests, or mammalian somatic toxicity tests, complement the mutagenic
and cytotoxic testing and provide validating or confirmatory information on
biological potency.
In Table 7 we list the bioassays to be employed, although not all tests
will be run on all samples collected at a given point. Cellular bioassays
make use of a variety of biological systems including bacteria, yeast, and
mammalian cells to investigate mutagenic effects. These shorter-term tests
will provide guidance in subsequent testing and be complemented by longer-
term validating assays using Drosophila, cultured mammalian cells, and whole-
animal (mouse) systems. Toxicity tests involve the use of whole animals to
characterize the acute, subacute, and chronic toxicity of products and ef-
fuents. They are used in the study of materials of likely high toxicity
138
-------
TABLE 6. PORTABLE MONITORING INSTRUMENTS
Control Pollutant
NH3, N0x, S02, C6H6!
Cx-H OH, naphthalene
its derivatives
Second-derivative ultraviolet absorption
spectrometer with multipass gas cell for
real-time monitoring of selected effluents
Tar and oil on contaminated
surfaces
Fluorescence spill spotter for general
surface contamination including BaP
Tar and oil on contaminated
skin
Lightpipe luminoscope for residual skin
contamination using very low intensities
of UV light
PNA vapors
Passive meter for area or personnel moni-
toring of selected PNA compounds using
room-temperature phosphorescence detection
techniques
139
-------
TABLE T. BIOASSAY TESTS FOR HEALTH EFFECTS ASSESSMENT
Purpose
Test
Screening
Screening and validation
Mammalian mutagenesis
Mammalian toxicity
1.
2.
3.
IK
5.
6.
7.
1.
2.
3.
U.
5.
1.
2.
3.
h.
5.
1.
2.
3.
k.
5.
6.
Bacteria — Salmonella typhimurium strains
Yeast — Saccharomyces cerevisiae
DNA repair — Bacillus subtilis
Mammalian cell (CHO) cytotoxicity
Invertebrate cytotoxicity — Tetrahymena
pyriformis
Embryo toxicity-- XenQpus laevis
Mammalian teratogenesis — mouse
Fruit fly — Drosophila melanogaster
Mammalian cell (CHO) gene mutation
Mammalian cell (CHO) cytogenetic damage
Mammalian cell (leukocyte) chromosomal change
In-vitro cell transformation
Mouse — dominant lethals
Mouse — heritable translocations
Mouse — specific locus
Mouse — spot test, somatic mutation
Mouse — reproductive capacity
Mouse — acute oral LD^Q
Mouse — intraperitoneal injection LDcQ
Rat — acute dermal toxicity
Rabbit-- eye and skin irritation
Guinea pigs - dermal sensitization
Mouse-- maximum tolerated dose
Mammalian carcinogenesis
1. Mouse-- lung tumors
2. Mouse— skin tumors
140
-------
about which little information is available but which have potential for human
exposure. A brief description of each bioassay is provided in the H-Coal pro-
gram plan.
ENVIRONMENTAL FATE AND EFFECTS
Environmental studies emphasize the data base requirements to assess the
H-Coal technology rather than the pilot plant. Consequently, the thrust of
the program is on characterizing and testing process and plant effluents,
solid wastes, and liquid products. The latter studies are concerned with the
effects of possible oil spills on terrestrial and aquatic systems.
Ecological tests to be performed on the various materials are identified
in Table 8. Toxicity screening tests provide an initial indication of poten-
tial ecological effects and include algal photosynthetic inhibition (U-h
exposure) and acute toxicity response (U8-h LCcg), the latter using three
different aquatic test organisms. Materials showing high toxicity and
high potential for environmental exposure will be tested further. These
activities include tests of the reproduction effects on Crustacea and insects
and of acute toxicity and abnormalities on fish embryo-larval life stages.
Chemical and physical characterization (Table l) is a part of the testing
protocol.
The transport and fate of products that may be spilled in aquatic en-
vironments will be studied in small field ponds. Aqueous extracts of vacuum
bottoms flaked product and filter cake consisting of solids from the waste-
water treatment system will be prepared to simulate on-site storage and landfill
disposal, respectively, and will be subjected to selected tests. Studies of
aqueous wastes will be limited to effluents from the wastewater treatment
plant and to the combined discharge (process water, sanitary effluents, and sur-
surface runoff from the plant site) to the Big Sandy River. All tests will be
replicated with equivalent petroleum crude and oil for comparison purposes.
SPECIAL STUDIES
Several special studies will be carried out to examine issues of particu-
lar importance in direct coal liquefactio'n, which involves the H-Coal process.
They are summarized as follows.
Advanced Wastewater Control^ Technology
A 1-gpm wastewater treatment PDU will be designed and constructed for
initial use at the H-Coal pilot plant.^° The treatment unit provides the
means to evaluate the efficiency and cost of advanced treatment techniques to
achieve zero stream discharge or meet future discharge regulations, to in-
vestigate the operational problems of existing systems, and to provide scale-
up data for larger facilities. Unit processes in the treatment train will be
constructed in transportable, self-contained modules that can be interchanged
or bypassed to achieve maximum flexibility- As shown in Figure 3, unit pro-
cesses will provide for pretreatment and conditioning by distillation stripping
of NH^ and H2S, setting and flotation for solids and oil removal, and solvent
extraction 'for phenol removal; biological oxidation; and polishing operations
141
-------
TABLE 8. ECOLOGICAL TESTS OF H-COAL MATERIALS
a i 1
cd i .cd
£q
p
i -P
1 3
H a
cd £-1
faO O
H 0
cd -H
FH
' P-
co cd
•H O
CO
0) g
r^H 3
-P Sn
Sample £ -^
i>j co
Description
cd ^
0) cd
0 H
cd
CO
O
a
•H
bfl
p"
?H
(D
cd
CO
•H
-P
CO
a'
o
^i
a
•H
S
cd
-p i P-
CU PJ O -p -p D
^3 03 CU -H g
-P W -p £4 O -H
cd CD co a -H PL,
'H H 3 T-I CO X CO
cd ^ J-j pS O 1 CD
" fi o a o -pi -H
>> P- -H ^ TJ
-PCD -cd --p HO ^
•H S Cd Ceo cdfl -p
0-H OM OCU >S co
•H PL, -H Cd -H Q fn -H
X -PS -po cdS -p
O 1 O O Tj H CO Jn
-pi pJ cd ^i I 13 cd O
Is tj-r-i T^cd OcdH P.
CUO O£ O-p i>jCDCD co
-PCl ^ rC S-lCD M^S
-------
COAL CONVERSION WASTEWATER TREATMENT PILOT PLANT
H,S NH,
OIL
t
,S AND NH3
STRIPPING
t
_». OIL/SOLIDS
REMOVAL
SOLIDS
GRAVITY
SEPARATION
INDUCED AIR
FLOTATION
PRETREATMENT
REMOVAL
PHENOL
SOLVENT
EXTRACTIO
J
N
BIO-
OXIDATION
SLUDGE
ACTIVATED
SLUDGE
FLUIDIZED-
BIOREACTO
OZONAI 1UN * .p
CARBON ». REVERSE
SORPTION OSMOSIS
EFFLUENT WATER _
i
SPENT T
CARBON CONCENTRATE
*
BED
R
1
1 POL
1
EVAPORATION
SOLID SALTS
ISHING 1
Figure 3. Coal Conversion Wastewater Treatment Process Development Unit
-------
with ozonation and carbon adsorption for refractory and residual organic re-
moval and reverse osmosis for dissolved salt removal.
Product Upgrading
Exploratory research indicates that hydrotreatment and selective dis-
tillation of H-Coal PDU liquid products may reduce significantly microbial
genotoxicity and mammalian toxicity. Similar findings are reported for
SRC-II liquids.29 Consequently, a systematic study has begun of the effects
of hydrotreatment and process conditions on the chemical, physical, and
biological properties of liquid products derived from pilot plant operations
of H-Coal, SRC-II, and Exxon Donor Solvent (EDS) processes. Blends of dis-
tillates will be hydrotreated to three levels of severity and characterized
and tested for toxicological response. In a companion effort the status and
preliminary cost estimates of process technology for hydrotreatment, boiling-
cut fractionation, and other methods of product upgrading (e.g., nitrosation,
acid-base extraction, organic solvent extraction, and chromatographic separa-
tion) will be investigated.30
CONCLUDING REMARKS
Samples of coal-liquid products from the H-Coal PDU have provided initial
information on important areas of continuing research. Comparative studies
show that crude petroleum substitutes, including the H-Coal materials, gener-
ally exhibit greater activity in biological and ecological test systems than
petroleum crudes, but this activity is reduced in samples that have been
hydrotreated and in low-boiling distillates. Constituents of the alkaline
and neutral fractions of coal liquids are responsible for mutagenicity whereas
phenolic compounds and anilines cause the greatest toxicity in freshwater algae
and zooplankton.
Based upon the initial results of studies with PDU materials and in
consideration of the scale-up requirements for a commercial-size facility, the
implementation of an extensive health and environmental study of the H-Coal
pilot plant has begun. Plans include the characterization and testing of
products, by-products, and effluents; collaborative studies with the operating
contractor involving plant area monitoring and worker protection; and investi-
gations of environmental controls for plant effluents. Systematic studies
have also begun of hydrotreatment and other methods of upgrading liquid
products to alleviate biological activity.
144
-------
REFERENCES
1. U.S. Department of Energy, Securing America's Energy Future: The
National Energy Policy Plan, DOE/S-0008 (July 1981).
2. R. S. Livingston et al. , Toward a Desirable Energy Future: A National
Perspective, ORNL/PPA-81/6 (March 198l).
3. W. D. Felix et al., Chemical/Biological Characterization of SRC-II
Product and By-Products', FNL-SA-8813 (September' 1980')'.
1*. L. W. Rickert and J. L. Seiber, Publications in Life Sciences Synthetic
Fuels of Oak Ridge National Laboratory, ORNL/TM-7680 (June 1981).
5. W. H. Griest, D. L. Coffin, and M. R. Guerin, Fossil Fuels Research
Matrix Program, ORNL/TM-73^6 (June 1980).
6. I. B. Rubin, M. R. Guerin, A. A Hardigree, and J. L. Epler, "Fractiona-
tion of Synthetic Crude Oils from Coal for Biological Testing,"
Environ. Res. 12:358-365 (1976).
7. A. R. Jones, M. R. Guerin, and B. R. Clark, "Preparative-Scale Liquid
Chromatographic Fractionation of Crude Oils Derived from Coal and
Shale," Anal. Chem. 1*9:1766 (1977).
8. T. ¥. Schultz, J. N.'Dumont, T. K. Rao, M. R. Guerin, C. Y. Ma, and
J. L. Epler, "Evaluation of Hydrotreatment as a Means of Reducing
Biological Activity of Synfuel Related Materials," Environ.
Res. (submitted September 1981).
9. J. L. Epler, "The Use of Short-Term Tests in the Isolation and Identi-
fication of Chemical Mutagens in Complex Mixtures," Chemical Mutagens:
Principles and Methods for Their Detection, New York, Plenum Press (1980).
10. M. R. Guerin, I. B. Ruben, T. K. Rao, B. R. Clark, and J. L. Epler,
"Distribution of Mutagenic Activity in Petroleum and Petroleum Sub-
stitutes," Fuel 60 (1*): 231-21*8 (1981).
11. H. R. Witschi, W. M. Huschek, and L. H. Smith, Level 2- - Mammalian
Toxicity, University of Minnesota-Duluth Coal Gasification Project
Quarterly Progress Report for the Period Ending December 31, 1979,
ORNL/TM-7268 (June 1980).
12. J. M. Holland, D. W. Wolf, and B. R. Clark, "Relative Potency Estimation
for Synthetic Petroleum Skin Carcinogens," Environ. Health
Perspect. 38:lU9-155 (1981).
13. H. R. Witschi, Acute Oral Toxicity Tests, H-Coal Pilot Plant Environmental
Project Semiannual Progress Report for the Period Ending December 31,
1980, ORNL/TM-77^1 (June 1981).
145
-------
lU. J. L. Epler, ed., Health Effects Research in Direct Coal Liquefaction:
Studies of H-Coal Distillates, Process Development Unit Samples,
ORNL/TM (in draft). ~ ' ' ~ ~ ~~
15. J. M. Holland, Toxicity and Carcinogenicity Associated with Chronic
Dermal Exposure to H-Coal PDU Distillated and Effect of Hydrotreatment
on the Chronic Dermal Toxicity of a Blended H-Coal PDU Distillate,
Life Sciences Synthetic Fuels Semiannual Progress Report for the
Period Ending June 30, 198l, ORNL/TM-7926 (in press).
16. C. W. Gehrs, Advanced Fossil Energy Program, Environmental Sciences
Division Annual Progress Report for the Period Ending September 30, 1980,
ORNL-5700 (March 198l).
IT. J. M. Giddings, Summary of Research on Goal Liquefaction Product Spills,
ORNL/TM-7966
18. J. M. Giddings and J. N. Washington, "Coal-Liquefaction Products. Sha,le
Oil, and_ Petroleum. Acute Toxicity to Freshwater Algae." Environ.
Sci. Technol. 15: (1981).
19. B. R. Parkhurst, C. W. Gehrs, and I. B. Rubin, "Value of Chemical Fraction-
ation for Identifying the Toxic Components of Complex Aqueous Effuents,"
in Aquatic Toxic ology, American Society for Testing Materials,
ASTM STP-66?.
20. U.S. Department of Energy, Fate and Effects of Coal Liquefaction Products
in Aquatic Ecosystems, Final Environmental Impact- Statement, SRC-I Demonstra-
tion Project, 2 DOE/EIS-0073 (July 1981).
21 U.S. Department of Energy, Environmental Development Plan: Coal
Gasification, DOE/EDP-OOU3 (October 1980).
22. U.S. Department of Energy, Environmental Development Plan: Coal
Liquefaction, DOE/EDP-00^ (August 1980).
23. U.S. Department of Energy, Environmental Readiness Document, Oil Shale,
DOE/ERD-0016 (September 1979).
2k. U.S. Department of Energy, Synthetic Fuels and the Environment: An
Environmental and Regulatory Impact Analysis, DOE/EV-0087 (June 1980).
25. C. R. Richmond, "Opening Comments," Proc. 3rd ORNL Life Sciences
Synrpo. on Health Risk Analyses, Gatlinburg, Tennessee, October 27-29,
1980 (in press).
26. K. E. Cowser, ed., Environmental and Health Program for H-Coal Pilot
Plant, Oak Ridge National Laboratory (November 1980).
146
-------
27. D. D Schuresko, Preliminary Field Testing of H-Coal Monitoring Instru-
ments, H-Coal Pilot Plant Environmental Project Semiannual Progress
Report for the Period Ending December 31, 1980, ORNL/TM-77^1 (June 1981):
28. G. E. Osvald, J. F. Walker, J. A. Klein, and R. K. Genung, Coal Conversion
Wastewater Treatment PDU, Life Sciences Synthetic Fuels Semiannual Pro-
gress Report for the Period Ending June 30, 198l, ORNL/TM-7926 (in press).
29. W. C. Weimer, R. A. Pelroy, and B. W. Wilson, Initial Chemical and
Biological Characterization of Hydrotreated Solvent Refined Coal
(SRC -II) Liquids: A Status Report, PNL-3^6^ (July 198(3).
30. R. Salmon, J. F. Fisher, and K. H. Lin, An Evaluation of the Technology
Available for the Reduction of Bioactivity in the Products of Direct
Coa'l Liquefaction, QRNL/TM-8039 '( "in" draft )'.
147
-------
CHKMTCAT, CHARACTERIZATION AND BIOASSAY OF SRC PROCESS MATERIALS
by: W. Dale Felix, D. D. Hahlum, B. W. Wilson,
W. C. Weimer, and R. A. Pelroy
Battelle Pacific Northwest Laboratory
Richland, WA 99352
ABSTRACT
Bioassay techniques have shown "chat certain coal liquefaction process
streams and products are both mutagenic (Ames assay) and carcinogenic.
These materials have been chemically fractionated using a number of tech-
niques (solvent extraction, alumina column separation, HPLC, Sephadex
LH-20) in an attempt to identify the constituents responsible for the
biological activity. These studies have shown that primary aromatic
amines (PAA'sJ account for more than 90% of the mutagenic response in the
Ames test. Long-term skin painting and initiation-promotion assays indi-
cate that the PAA's may also play a role in the carcinogenicity of the
coal-derived materials. However, while the PAA's can be designated as the
determinant mutagens in coal liquids, they cannot be assigned a determi-
nant role in skin carcinogenesis. Thus far, carcinogenicity appears to
better correlate with increasing molecular weight and boiling point. Our
results also suggest that benzo(a)pyrene is not a reliable marker compound
for carcinogenic activity.
(Only the abstract is published herein.)
148
-------
Session II: WATER-RELATED ENVIRONMENTAL CONSIDERATIONS
Chairman: N. Dean Smith
U.S. Environmental Protection Agency
Research Triangle Park, NC
Cochairman: William E. Corbett
Radian Corporation
Austin, TX
149
-------
COAL CONVERSION WASTEWATER TREATMENT/REUSE - AN OVERVIEW
F. E. Witmer
Environmental Technology Division
U.S. Department of Energy
Washington, D.C. 20545
ABSTRACT
Environmentally, the production of synfuels from coal can be classified into
two categories: (1) low temperature processes, and (2) high temperature pro-
cesses. Low temperature processes are characterized by the production of high
boiling liquids and tars which tend to retain the multiple-ring structure of
the original coal "molecule," while high temperature processes typically pro-
duce synthesis gas, methane and/or light liquids. Dry-ash moving-bed gasifi-
cation and direct 1-quefaction processes are representative of low temperature
conversion processes. Entrained gasification is an example of high temperature
processes. Fluid bed gasification processes that operate at temperatures just
below the ash slagging point may produce limited heavy liquids and fall inter-
mediate within the classification regime.
Depending on the process, process steam which is subsequently condensed and/or
gas clean-up quench waters come into direct contact with the raw gaseous pro-
duct stream. As a consequence, the resultant wastewater associated with the
low temperature processes is highly contaminated with organics. The production
of ammonia in the high temperature processes is generally suppressed and re-
duced due to "cracking." Condensate waters from high temperature processes
usually contain little or neglible NH_, while the condensate waters from low
temperature processes contain high levels of NH.,. The condensate waters from
both low and high temperature processes generally contain volatized and en-
trained mineral matter, trace elements and salts as well as adsorbed H«S, CO.,
and cyanates.
The treatment of the condensate waters from the low temperature processes poses
a special challenge due to the high and variable level and toxic nature of the
gross organics. A portion of the total organic carbon is biorefractory and
this also causes concern. Laboratory treatability tests have demonstrated that
with appropriate dilution and/or pretreatment (e.g., gas stripping, organic
extraction, and/or the addition of powdered activated carbon) activated sludge
treatment processes do a reasonable job of reducing biological oxygen demands
(BOD) and total organic carbon levels (TOC), and coupled with activated carbon
treatment, relatively high quality effluent can be produced. In a "zero dis-
charge" mode, subsequent concentration and reuse of the effluent must be ef-
fected to ultimately produce a concentrated brine or dry salt.
The questions that remain center on the capability of this rather elaborate
treatment train to accommodate variabilities in the raw feed and on the relia-
bility and costs of such a system, i.e., do viable alternatives exist? Options
will be outlined with special emphasis on: (1) improvements to biological
treatment, and (2) purely physical/chemical systems. The effect of more
stringent standards with respect to, say, the control of biorefractory ring-
structure compounds, trace elements, ammonia, etc., will be discussed relative
150
-------
to the state-of-the-art biotreatment and these environmental control options.
Areas of uncertainty and future research will be delineated based on a recent
synfuel wastewater workship, conducted in June 1981.
INTRODUCTION
The capability to adequately treat and discharge wastewaters associated with
coal conversion causes some apprehension primarily because of the lack of
treatability data from actual operating facilities, at scale and under
stringent discharge standards. A number of concerns exist for the tentative
wastewater control systems:
• the possible requirement to meet tighter future effluent discharge
standards for ring-structured biorefractory organics, trace elements,
ammonia, etc.;
• the high level of contamination, variability of composition, and
large and variable volumeric flowrate (variability being a special
concern with pioneer type plants);
• the vulnerability of biological treatment systems to toxic effects
(either due to high loadings in the feed and/or build-up of toxic
agents from recycle); and
• the desirability that the wastewater treatment/reuse system be
highly reliable to preclude shutdown or curtailment of production.
If one considers "chemical" pollutants of universal concern, adverse
environmental effects include (1) changes in pH by strong acid and bases,
(2) increase in water corrosivity and reduced suitability for irrigation
due to soluble salts, (3) toxicity caused by heavy metals, phenols and
cyanides, (4) depletion of dissolved oxygen by oxygen consuming organics,
(5) surface films from trace oils, (6) taste and odor problems associated
with phenols and chlorinated derivatives and (7) release of biorefractory
materials which can be fatal to fish and aquatic life (note—the effect on
man, especially any long term cumulative effects, has not been established).
Representative compositions of condensate waters resulting from low
temperature coal conversion processes contain each of these "chemical
pollutants" (Figure 1). It is apparent that intensive and specialized
treatments are required for such waters.
In the design of wastewater treatment facilities, a variety of sources are
encountered, although condensate waters typically account for over half the
wastewater produced (Figure 2). General design practice is to segregate
streams and use different methods of pretreatment tailored to the
composition of the individual streams. The current pollution control design
data base is such that the practice is to encourage the incorporation of
of enhanced design flexibility within the total system. In this context,
"flexibility" refers to parallel units and/or spares, conservative design
specifications, bypass lines and space to accommodate additional equipment,
151
-------
if necessary. If one examines the preliminary designs of a number of
wastewater treatment trains, one finds a high degree of variation between
individual process designs with provisions for "flexibility" reflecting a
common design philosophy (Figures 3, A, 5).
pH
BOD
TOC
COD
PHENOLS
SULFIDE, S
AMMONIA, N
THIOCYANATE, SCN"
CYANIDE. CN-
TDS
GASIFICATION
LURGI
9.0-9.5
4,000-15,000
4,000-20,000
15.000-30,000
2,000-6.000
100-600
2.000-10.000
20-200
0.1-10.0
1,000-5,000
DIRECT LIQUEFACTION
oflC I
8.4
17,000
11.000
60.000
1,900
16.000
18,000
-
-
16,000
H-COAL
9.5-10.8
-
-
28,000
70,000
30.000
15,000
-
4.0
-
EDS
-
-
-
-
6,000
12,000
9,500
10
4.0
-
RGURE 1. COMPOSITION OF REPRESENTATIVE RAW CONDENSATE WATERS FOR
"LOW TEMPERATURE" CONVERSION PROCESSES, PPM
WASTEWATER
STREAM
PROCESS CONDENSATE
HIGHLY CONTAMINATED
(LOW TEMPERATURE
PROCESSES)
SOURCE
MOVING BED GASIFIERS
DIRECT LIQUEFACTION
QUANTITY
Mgpd
1.5-7.0
.05-1.0
MAJOR
CONCERN
ORGANICS
MODERATELY CONTAMIN-
ATED (HIGH TEMPERATURE
PROCESSES)
ENTRAINED GASIFIERS
FISCHER TROPSCH."
MOBIL-M*
2.5-4.0
1.0-3.0
ORGANICS
CLEAN
SLOWDOWNS
COOLING TOWER
BOILER
RAIN RUNOFF
SANITARY WASTES
METHANATION STEP
COOLING TOWER SYSTEM
ION EXCHANGE REGENERA-
TION AND REVERSE
OSMOSIS CONCENTRATE
RAIN FALL FROM
IMPOUNDMENT
POTABLE WATER SYSTEM
0.2-1.0
0.5
0.1-1.0
VARIABLE
75% POTABLE
RATE
TDS
MIXED
•EXCLUSIVE OF GASIFICATION STEP
RGURE 2. SUMMARY OF COAL CONVERSION PLANT WASTEWATER STREAMS
(3x1011 Btu/d REF PLANT)
152
-------
EVAPORATION
PHENOLSOLVAN EXTRACTED
SOUR WATER FROM
AMMONIA STRIPPER
RECTISOL WASTES
SLUDGE
SLUDGE
ASH DISPOSAL
PROCESS AREA
RUN OFF
OILY WASTES
RAW WATER
PRETREATMENT
AND LOW PRESSURE
STEAM BLOW-DOWNS
FIGURE 3. WASTEWATER TREATMENT SYSTEM ANG COAL GASIFICATION PLANT
COOLING
TOWER
SLOWDOWN
SOUR WATER FROM
AMMONIA-SULFIDE
STRIPPER
LANDFILL
RUNOFF/
LEACHATE
GASIFIER
WASTEWATER'
SRC PILE AND
COAL PILE RUNOFF
PROCESS AREA RUNOFF
OILY WASTES'
S f
SURGE
BASIN
OIL
REMOVAL
SLUDGE
SLUDGE
DISINFECTANT
RIVER OR
FIGURE 4. TENTATIVE WASTEWATER TREATMENT SYSTEM -
SRC-I DEMONSTRATION PLANT
153
-------
(U-GAS FLUID BED GASIFIERt
UNTREATED WASTE
TO MUNICIPAL SEWER
RECOVERED OIL TO SLUDGE TO COAL PILE 1
OFF-SITE DISPOSAL DISPOSAL RUNOFF I
SANITARY
WASTEWATER
DECANT
DISSOLVED AIR PITS RUNOFF
SPENT SERVICE • FLOTATION , K " ^_ ... TBPATMFNT
WATER SIUDGE SI UDRF
t
TREATED jL
WATER ••»
UME RECARBONATORGAS """^ H°.|!^ 1
TO/FROM TAIL GAS NEUTRALIZATION •
PREPARATION i
eTRipppn W«TFR „ ._ TREATED |
FROM SOUR »- WASTEWATER A
WATER STRIPPING I
FINE" FILTRATE RECOVERED 1
FROM STEAM •- pnocf'- • WATER
GENERATION WIXT™!™ SLUDGE CHROMATE
CREDIT GENERATION »- SLUDGE SLUDGE
CONDENSATE i 1 Ht
T WAT
| MISSI
CLARIFIED WATER f POLYMER Rl\
ron?IN<^AND OXYGEN GRANULAR COOLING
SrBIJBRIhlO FROM AIR CARBON TOWER
SEpARAT|ON (MAKEUP> DEWATERED SLOWDOWN
SLUDGE TO
DISPOSAL
» MAKEUP TO
COOLING
TOWER
» MAKEUP TO
ASH SLURRY
WATER TANK
BOILER BOTTOM
ASH SLUICING
MED
ER TO
SSIPPI
EH
FIGURE 5. WASTEWATER TREATMENT SYSTEM - MEMPHIS INDUSTRIAL FUEL GAS PLANT
-------
DESIGN BASIS FOR WASTEWATER TREATMENT TRAIN
While there are many options for the wastewater treatment/reuse systems,
several criteria are overriding in determining the characteristics of the
raw influent and the ultimate basis of design:
• geographical location of the plant (i.e., availability of water
supply and nature of effluent acceptor);
• characteristics of conversion process; and
• environmental setting and/or controlling permits/standards.
These criteria are obviously interdependent. The geographical location also
determines the coal type and influences the product/by-product mix as called
for by the market place; this in turn dictates the configuration of the
plant. The site specific environmental constraints also influence the
process design—the type and nature of wastewater treatment system; the
propensity for wastewater reuse and disposal options for wastewater sludges
(both organic and inorganic).
GEOGRAPHIC SITING
Geographical water availability/discharge constraints strongly influence the
design philosophy for water use integral to the plant. In the arid West,
dry cooling and staged quenching may be considered to conserve water, while
"zero discharge," coupled with evaporation ponding, is likely to be
encouraged to preserve salt-taxed river basins. In the East a different
situation prevails; while the use of effluent discharge is considered to
provide an acceptable means of salt dispersion, the release of residual
trace biorefactory material and trace elements into potential drinking water
supplies even though highly diluted, causes anxiety. This is especially true
for biological-activated-sludge treatment systems which may experience upsets
and require extended periods (several weeks) to recover. During the recovery
period, adequate contingency must be available (holding ponds, plant derating,
activated carbon units, etc.) to allow the plant to continue to operate.
The large coal requirements and concomitant ash disposal needs for commercial
synfuel facilities dictates that the plant be located at or near the mine.
There are significant chemical and processing differences between the
western and eastern coals. While it is recognized that there are large and
overlapping variations in the composition and chemistry of different coal
types and that pretreatments can modify the coal structure, decrease ash and
reduce sulfur and nitrogen levels, some generalizations relative to plant
siting and feed stock requirements for the process can be made:
Western (lignite and sub-bituminous type coals) are geologically
younger than the eastern bituminous coals. The lignites, in
particular, contain high levels of moisture and inherently produce
net water during conversion. In the East, the ready availability
of water supplies and discharge acceptors makes once-through water
use preferred, although ideally the conversion processes can be
155
-------
designed to be a net water consumer. The bituminous coals have
agglommerating properties that generally preclude their use in
moving bed gasifiers without intensive pretreatment. As a
consequence, the large scale gasification of eastern bituminous
coal will likely rely on entrained and/or fluid-bed gasifiers
with a concomitant improvement in the quality of process condensate
waters. Condensate waters associated with direct liquefaction
processes will contain gross organic contamination for all coal types.
The composition of the runoff from coal storage piles is likely to vary
as function of coal pyrite content. The pyrite abets acid generation which
enhances the mobility of metals and total dissolved solids (TDS). The higher
pyritic coals are in the East, thus potentially aggrevating a concern in a
region where "zero discharge" is not contemplated. Ferric iron tends to
predominate the metal release (Figure 6). Conventional treatment practice is
neutralization followed by settling.
PROCESS CHARACTERISTICS
The reaction conditions and coal type in the conversion process strongly
affect the composition of the condensate water. The most important variable
is the temperature-residence time regime to which the coal/reaction products
are exposed. This is markedly illustrated by comparing condensate water
qualities for an entrained gasifier (bench scale) and a slagging moving bed
gasifier (pilot-scale) (Figure 7). While the residence time in entrained
gasifiers is very short (on the order of 20 millisec in the Eyring unit), the
very high temperatures obtained appear capable of precluding the formation
of ring-structured compounds and ammonia during a rapid devolatization/
pyrolysis step. At the onset, the extremely rapid exothermic carbon-oxygen
reaction predominates the slower endothermic steam-carbon and carbon
dioxide-carbon reactions (Figure 8, regions I and II, respectively). The
residual char has been demonstrated to effectively scavenger
for trace ring-structured compounds that- may be formed—the condensate
water has been found to be nearly devoid of organics. This is in marked
contrast to the condensate waters associated with lower temperature
processes such as direct liquefaction with residence times up to several
hours, which can contain practically all the organic compounds found in the
coal. Thus, from the standpoint of raw condensate water quality and
subsequent amenability to treatment, the temperature of coal conversion
processes represents a major variable and the condensate waters may be
classified under low or high temperature regimes.
A major variable effecting the low-temperature processes is coal rank—the
more easily pyrolyzed, more reactive lignite and sub-bituminous western coals
generally produce more phenols, given similar process conditions (Figure 9).
Process configurations, quantity and recycle of product gas quench waters
and/or staging also determines the quantity and quality of the condensate
waste stream. General gross differences between gasification and liquefaction
condensate waters are reflected in the sulfide and ammonia concentrations
(Figure 1). With respect to organics, laboratory treatability testing of
steam stripped waters indicates that biological substrate utilization rates
156
-------
for liquefaction condensates may be significantly (an order of magnitude or
more) less than gasification. The incentive for staged quenching and
concomitantly reducing water requirements have been found to reside mainly
in the use of less expensive materials of construction (carbon steel instead
of stainless) downstream of the initial quench which removes the strong acids,
Coals with a halide content of 0.15% Cl or greater, generally eastern coals,
are expected to benefit from such a configuration.
PARAMETER
pH
IRON
SULFATE
ZINC
COPPER
CHROMIUM
TDS
MEAN, PPM
2.7
20,000
9,000
3.6
2.1
3.3
16,000
RANGE, PPM
2.1-3.0
0.2-90,000
500-22,000
1.6-23
1.6-3.4
0-16
720-44,000
FIGURE 6. REPRESENTATIVE COMPOSITION COAL PILE RAIN RUNOFF, EASTERN COAL
(ANDERSON AND YOUNGSTROM, CORNELL UNIVERSITY)
PROCESS
COAL TYPE
ORGANIZATION
HIGH TEMPERATURE
ENTRAINED
BITUMINOUS
EYRING RESEARCH
(MOUNTAIN FUEL)
LOW TEMPERATURE
SLAGGING-MOVING BED
LIGNITE
GFETC
pH
BOD
TOC
COD
PHENOLS
SULFIDE, S
AMMONIA, N
THIOCYANATE
CYANIDE
TDS
-
NIL
NIL
-
NON-DETECTED «5)
1.5
35 (FIXED)
-
.02 MAX
330
8.6
26,000
11,000
32,000
5,500
100
6,000
120
2
2400
'NOTE-ABSOLUTE CONCENTRATIONS ARE DEPENDENT ON QUENCH WATER CONTACT/RECYCLE ETC.
WHICH DIFFER. RELATIVE PREDOMINANCE OF SPECIES IMPORTANT.
FIGURE 7. COMPARISON OF CONDENSATE WATER FROM "HIGH TEMPERATURE"
AND "LOW TEMPERATURE" GASIFICATION PROCESSES, PPM*
157
-------
LLI
CC
oc
UJ
Q_
5
LJJ
o
EC
z
o
to
o
0-
5
o
u
3000
2500
2000
1500
1000
500
0
60
50
40
30
20
10
0
J\
10
20
30
REACTOR LENGTH-CENTIMETERS
40
(OUTLET)
FIGURE 8. TEMPERATURE/COMPOSITION PROFILE ENTRAINED GASIFIER
(P. SMITH, BRIGHAM YOUNG UNIVERSITY)
PROCESS
LURGI/BRITISH GAS
SLAGGING LURGI
IWESTFIELD. SCOTLANDI
PHENOL LEVEL, PPM*
COAL.
RANK' LIGNITE SUBBITUMINOUS
BITUMINOUS
MONTANA
4,400
PITTSBURGH
ILLINOIS NO.6 NO. 8
1,900
2,100
PETC SYNTHANE
(FLUID-BED)
NORTH
DAKOTA
6.600
WYOMING
6,000
2,600
1,700
•NOTE-ABSOLUTE CONCENTRATIONS ARE DEPENDENT ON QUENCH WATER CONTACT/RECYCLE ETC. WHICH
DIFFER BETWEEN PROCESSES, TRENDS ARE IMPORTANT.
FIGURE 9. COMPARISON OF CONDENSATE WATER PHENOLIC CONTENT
FROM DIFFERENT RANK COALS
158
-------
Solid and semi-solid sludges and biosludges will result from wastewater
treatment. The composition of these potentially biohazardous sludges will be
variable. It is likely that the sludges will be rendered non-hazardous by
oxidizing them at high temperatures by incineration (direct or fluid bed) or
combining them with the coal feedstock or carbonaceous bottoms to be recycled
to a high temperature gasifier. The role of wet-air oxidation to detoxify
these sludges is under investigation.
ENVIRONMENTAL SETTING
Federal guidelines and standards along with state and local permitting
authorities set the effluent specifications. Special site specific
conditions and concerns can lead to stringent regulations which could
conceivably dictate the degree of treatment and even the requirement for
"zero discharge."
REPRESENTATIVE TREATMENT TRAIN
While a universal wastewater treatment train configuration does not exist, a
inventory of unit operations are generally available to the different systems
proposed for treating the condensate waters associated with low temperature
coal conversion processes (Figure 10). Subsequent discussion will be confined
to the treatment of low temperature condensate waters, representative of the
most difficult to-treat waters, since coal pile runoff and sanitary wastes
are susceptible to conventional treatment practice. The major unit operations
are arranged with wastewater of intermediate compositions and/or dilution
potential being interjected at various points along the treatment train. The
sequence of the various steps, in particular, the extraction and stripping
operations, may be interchanged:
Oil Separation
As a pretreatment to remove suspended oil, tar, grease and solids (includes
settling ponds with skimmers, API separators, centrifuges, etc.). These
pretreatments are not highly effective for emulsions, small particles, and
substances which possess densities near that of the aqueous phase; thus
dissolved air floatation which can remove these materials is sometimes
employed as a follow-up pretreatment.
Steam Stripping
Removes volatile material, namely dissolved gases (NH-, CO^, lUS, HCN and
COS). Light, low boiling organics may also be removed. Steam requirements
may vary from 0.05 to 0.2 Ib. steam/lb. wastewater. Means must be provided
to facilitate caustic addition to free fixed ammonia, because meeting free
ammonia effluent limits in the final effluent is difficult with poorly
stripped raw feeds. Process sewer streams typically require steam stripping
prior to biotreatment.
Solvent Extraction
Removes gross organics, phenols and polyhydric aromatics, in particular.
159
-------
Methyl isobutyl ketone (MIBK) and diisopropyl ether (DIPE) are preferred
solvents, the MIBK having the broader selectivity for organic material.
The requisite solvent recovery step generally involves stripping. When
used in conjunction with biological systems solvent extraction tends to
dampen fluctuations in organic loadings and potential toxic effects.
RAW WASTE WATER
APPROXIMATE COST
$71000 GAL
EQUALIZATION
STEAM-
NIL
AMMONIA
STRIPPING
SOLVENT-
NH3
2-5
PHENOL
EXTRACTION
pH ADJ
AIR-
NUTRIENT-
-PHENOLS
3-7
•*-REUSE ?
BIOOXIDATION
-SLUDGE
2-8
FILTRATION
OZONE-
0.1-0.2
•*-REUSE?
PARTIAL
OZONATION
2-5 (FUTURE STANDARDS)
CARBON
ADSORPTION
10-15
TOTAL 19-40
REUSE ?
EFFLUENT
FIGURE 10. REPRESENTATIVE WASTEWATER TREATMENT TRAIN
FOR COAL CONVERSION EFFLUENTS
160
-------
Biological Treatment
Applicable where effluent discharge and/or reuse of a low biological oxygen
demand (BOD) water is contemplated. Activated sludge treatment has a long
and successful history of dealing with coking and petroleum refinery wastes
which are similar in many respects to coal condensate waters. One of the
principal advantages is the forgiving nature of the biological system in its
ability to adapt to variable feed composition, provided abrupt changes are
not encountered. It is capable of removing all the BOD (by definition) and
approximately 75% of the chemical oxygen demand (COD) in the condensate
waters. The susceptibility of the biological process to upsets and toxicity
effects can be mitigated by introducing powder activated carbon (PAC) to the
reactor. This also improves the settlability of the sludge. Polynuclear
aromatics, some of which are refractory and collodial in nature, along with
heavy metals, may be incorporated with the sludge and must be dealt with
during disposal. The thiocynate content of the incoming feed can present
difficulties, if a stringent free ammonia discharge standard must be met.
The biological degradation of thiocynate releases NH3 which may require
subsequent air stripping and/or biological nitrification/denitrification.
The latter step generally requires long residence times, e.g., holding ponds.
Carbon Adsorption
A polishing step to remove low level refractory organics and color bodies and
may serve as a safeguard for process upsets. Prefiltration is normally
required to preclude fouling of the bed. Pollutants may be leached from the
bed immediately after carbon regeneration—recycle may be required.
While this touches on the more prominent conventional processes, there are a
large inventory of treatment processes that may be brought to bear for
special applications (Figure 11). The treatment of wastewater to discharge
quality by such a train is not cheap by municipal standards. Costs are in
the range of $20-35/1000 gals. This translates into an expense of 5-10% of
projected synfuel selling costs—hardly a barrier to commercialization.
A recent study has shown that, if raw water and an acceptable discharge
acceptor (large river or lake) is available to the plant site, e.g., an
Eastern location, the most cost effective and preferred approach is to use
a once-through water management plan based on PAC-biological treatment and
regeneration of the PAC by wet air oxidation, as opposed to water recycle/
reuse and/or "zero discharge" alternatives.
AREAS OF CONCERN AND FUTURE RESEARCH
The technology just discussed has assumed the availability of water and
steady-state operation. If one considers "zero discharge" and the facility
for handling process upsets, the representative treatment train needs some
refinement or possible replacement. Before addressing the more stringent
requirements imposed by "zero discharge" and unsteady operation, it appears
useful to outline areas were perhaps the conventional technology could be
improved or at least better understood.
161
-------
TREATMENT COST
UNIT OPERATION
GRAVITY SEPARATION
STEAM STRIPPING
SOLVENT EXTRACTION
BIOLOGICAL
CARBON ADSORPTION
OZONATION
PRECIPITATION W/IRON
WET-AIR OXIDATION
ION EXCHANGE
DESALINATION PROCESSES
INCINERATION/GASIFICATION
DEEP-WELL INJECTION
APPLICABILITY
DEPENDENCE ON
CONSTITUENT TREATED GROSS
BY-PRODUCT RANGE EFFLUENT EFFLUENT
ORGANIC INORGANIC REMOVAL POLISHING RECOVERY S/1000 GAL QUALITY QUANTITY
iS - V* - - 0.10-0.20 tS
BASIS FOR
RANGE
V*
-vs
10-20
6-12
-^25
0.5-2
5-10
10.000ppm BOD
( 2,000ppm COO
(CARBON REGEN
600ppm TOC
V*
FIGURE 11. SELECTED WASTEWATER TREATMENT OPTIONS
-------
STATE-OF-THE-ART TREATMENT PROCESSES
Reviewing the characteristics and concerns associated with the major unit
operations, several areas of research may be highlighted:
Oil Separation
The use of expendable/regenerable absorption media may be used to scavenge
for neutrally bouyant entrained oleophillic materials. Absorbents such as
coal, crushed slag, sand, etc., could be employed in a fluid-bed. Much of
the multi-ring structured organic material is sparingly water soluble and
is in the colloidal state; it appears that clarification and/or filtration
enhanced with appropriate flocculation aids and polymer addition could
significantly reduce the concentration of these materials.
Steam Stripping
Stripping is preferred for NH- concentrations >250 ppm. Subsequent, ammonia
recovery is economical via the PHOSAM W or the Chevron process at NH.,
concentrations of 10,000 ppm and flows of 250-500 gpm. The preferred
location of the steam stripping unit, before or after the extraction step,
if extraction is employed, needs to be determined. Volatile organics are
decreased if the stripper is located downstream of the extraction unit;
however, residual dissolved solvent may enter the stripping system. The
addition of lime to free fixed ammonia and reduce steam requirements also
can be used to precipitate heavy metals. The addition of the lime,
typically after the stripping of acid gases and free ammonia, increases
softening requirements, if intensive reuse and/or concentration of waste
brines is planned. A promising alternative under study is to use a liquid
cation exchanger to selectively recover ammonia as a by-product and enhance
stripping of the acid gases.
Solvent Extraction
As indicated, the preferred sequencing of the steam stripping and solvent
extraction is not clearly established. High pH, characteristic of intensive
ammonia stripping operations, causes appreciable ionization of phenols and
correspondingly leads to lower distribution coefficients, the requirement
of higher solvent to water ratios, and ultimately to a more costly process.
The most difficult-to-extract component normally dictates the controlling
solvent to water ratio. Extraction is capable of removing entrained
organics such as polynuclear aromatic micelles. With the proper solvent
or combination of solvents, extraction could also be effective at removing
the more polar, hydrophillic organics which comprise that significant
fraction of the TOC which is not extracted by commercial extraction
processes, e.g., DIPE, MIBK, etc., nor by conventional analytical procedures,
i.e., methylene chloride. For example, trioctyl phosphine oxide (TOPO), a
stronger Lewis base than ketones, has been shown to remove 90% of the COD
when used on a representative condensate water. TOPO is costly ($7-8/lb.)
and solvent recovery is critical. Concomitant with the development of
improved solvents and solvent systems, effective means of solvent
regeneration/recovery need to be stressed.
163
-------
Approximately 70% of the cost for extraction is equipment related, in contrast
to stripping where a major portion of the cost is for steam. Extractor costs
are nearly proportional to the number of stages. Thus there is incentive for
better solvents, improved stage efficiency and an approach to true counter-
current operation. One might consider the use of ultrasonics, cavitation, etc.,
to obtain intimate contact between solvent and solute (especially when a low
volume solvent to water ratio is used) and membranes to break the resultant
emulsion. The potential role of membranes in solvent recovery may be worth
exploring.
Biological Treatment
While biological treatment is effective, it is necessary to pretreat the feed
or dilute it to bring the high BOD loadings (phenol in particular) to an
acceptable level. With dilution, large volumes of bio-reactors are needed
due to high (recycled) influent flows and residence times of several days.
Solvent extraction becomes attractive for BOD levels of greater than 2000 ppm.
It also reduces difficulties due to foaming. There is incentive for reducing
the volume of the bioreactor systems. The use of oxygen enrichment (on-site
generated oxygen is available at most coal conversion plants) should be
considered to reduce volumes. The use of fluid bed bioreactors is being
studied to greatly increase volumetric loadings of biosubstrate with a
corresponding decrease in residence time requirements. Oxygen availability
becomes controlling in such a system—coupled with enriched air, approximately
an order of magnitude decrease in residence time can be achieved. While
fluid bed systems require pumping power to recycle the wastewater and maintain
the bed, the energy requirements are about one half those associated with
aeration for air-activated sludge systems. Because of the reduced residence
times, and availability of developed substrate to the process, it is
anticipated that fluid bed reactors will be more accommodating to process
variability and recover more quickly from process upsets. The potential
role of PAC to help mitigate possible upsets in fluid bed biosystems should
be investigated. Bioreactor staging can be considered as another means of
increasing specific bioactivity and better accommodating process variability.
In a single mixed reactor, concentrations are close to effluent concentrations;
consequently, reaction rates are low. By approximating plug flow through
staging, higher BOD loadings can be effected on the average, along with
higher reaction rates. Appropriate real time instrumentation to anticipate
toxic effects is desired, in lieu of monitoring completeness of thiocyanate
degradation as a lead indicator.
It has been shown that the major fraction of TOC resistant to biological
degradation has a molecular weight>30,000 and likely represents bio-
organism wastes. Research at characterizing of and determining techniques
for removing these materials is needed. It is quite likely that they are
sparingly soluble and enhanced clarification/flocculation techniques could be
applicable.
Carbon Adsorption
Because carbon consumption is related directly to the TOC of the water being
treated, activated carbon is generally used as a polishing step. The
164
-------
performance of carbon with respect to the high molecular weight organics,
touched on above, is uncertain. It has been found that ozonation prior to
treatment improves the adsorption efficiency for multi-ring compounds from
50-60% to 90% or better. A major uncertainty exists with respect to carbon
regenerability, especially in brackish waters. Acceptable carbon treatment
costs are based on the premise of complete regeneration, with secondary loss
and make-up of 5%. The actual capacity, effectiveness of regeneration and
costs, for activated carbon when used in a polishing mode on condensate
water needs to be better established.
PHYSICAL/CHEMICAL OPTIONS
Alternative treatment processes deserve consideration due to the relative
high costs associated with the series of five or more process steps that
comprise the typical state-of-the-art treatment train. An additional
consideration is the concern over system reliability resulting from
sequencing several unit operations, especially when a biological step, that
is vulnerable to upsets, is in the train.
As previously indicated "zero discharge" considerations can impose an
additional and overriding constraint. An end-of-the-pipe approach is to use
desalting technology to control the salt content of the effluent to render
it suitable for recycle/reuse at the front end of the process. The high
quality of the effluent from the representative treatment train should insure
the effluent is amenable to conventional desalting (distillation/reverse
osmosis) and, depending on the hardness, some softening may be desirable to
facilitate high product water recovery and reduce the quantity of brine that
requires further concentration/disposal. Cooling towers have been
universally used to cost-effectively reject process heat and perform the
initial concentration of process wastewaters. Typically, filtered effluent
from the activated sludge unit is used as make-up to the tower, although use
of DIPE extracted, steam stripped condensate water is contemplated for the
ANG dry-ash Lurgi plant at Beulah, North Dakota. Based on petroleum
experience, it appears that cooling towers can handle BOD loadings up to
500 ppm in the make-up water. An area of concern, in addition to potential
drift and odor difficulties, is the allowable concentration factor before
biofouling and corrosion problems become a detriment to the heat exchanger
loop. A study is underway to obtain a better handle on these
limitations.
Ideally, one desires a single process step that can take the raw wastewater
process stream and produce a moderate to good quality stream suitable for
reuse and a small highly contaminated stream that can be treated intensively
(Figure 12). Solvent extraction (previously discussed), distillation (vapor
recompression, in particular), and membrane processes represent candidates
for the major separation process, while wet-air oxidation, incineration,
gasification, dirty steam generation, etc., could be used to deal with the
resultant concentrate and render it acceptable for ultimate disposal.
Distillation and membrane processes also can be used to retain the salts in
the concentrated stream and, in this respect, are superior to solvent
extraction which must be coupled with ion exchange or another desalination
process to achieve this end. It should be noted that the product water may
165
-------
concentrate stream will have a very small volume, high unit processing costs
can be tolerated. The limitations and tradeoffs associated with the
distillation of poor quality wastewaters require better definition. The full
integration of these processes with the conversion plant, proper, should be
emphasized. Costs are estimated to be in the range of $6-12/1000 gallons,
thus there appears to be adequate leeway for system refinement and
optimization to be competitive with a conventional treatment train.
Membranes
The potential application of membrane processes (ultrafiltration and reverse
osmosis) to the concentration of raw coal condensate waters is relatively
long range. Progress has been made in producing composite membranes from
stable substrate polymers such as a porous polysulfone coated with an
exceedingly thin film (~200X) of salt rejecting membrane, for example a
highly crosslinked polyurea. Polyurea membranes exhibit good stability at
high and low pH's and in the presence of aromatic solvents. In general,
membrane rejection for ammonia is poor and, as a consequence, any membrane
concentration process will probably require steam stripping. The rejection
of phenols is improved at high pH's (~11); as a consequence, the addition
of lime to abet NH., stripping will also enhance rejection of ionizable
organics. This is in contrast to solvent extraction processes.
The raw condensate water will likely undergo oil separation and filtration
(essential, yet negligible cost pretreatments) prior to the reverse osmosis
units. The physical configuration of the membrane unit has a bearing on the
capability of handling a fouling/dirty feed—dead spots must be avoided to
preclude the deposition of material and progressive pluging of the unit.
Normally, tubular membranes are used (the influent flowing inside the tubes/
tubule bundle) to assure positive flow. Membrane units of this configuration
are in commercial applications on cheese whey and latex paints. In addition
to the preferred membrane composition, open to question is the degree of
concentration that can reliably be effected with a membrane unit—5 to
1 represents a conservative estimate, with 10 or 15 to 1 as probably an upper
limit. Very preliminary estimates of membrane separation costs are
$4-8/1000 gallon; thus although the recovery may not be as high as vapor-
recompression systems, the lower unit costs could more than compensate the
larger volume of concentrate subject to subsequent treatment, e.g., wet-air
oxidation. It would seem prudent to support a continuing research effort
to advance the application of membranes to condensate waters and solvent
recovery (previously mentioned).
Wet-Air Oxidation
Ideally wet-oxidation can convert pollutants to CO , N_ and H20 by reaction
with oxygen at high temperature and pressure. Because of the large flows
and expense for pressure vessels and heat exchangers, there is considerable
incentive for optimizing and moderating reaction conditions. Only limited
research is being directed at these tradeoffs. Costs are proportional to
water throughput and advantage can be obtained from the combustion heat
associated with high levels of organics. Costs are uncertain, but are
expected to be in the range of $20-30/1000 gallons. Wet-air oxidation
166
-------
require a polishing step to remove dissolved gases and/or hydrophillic
organics (generally of a low molecular weight) that could interfere with the
specific reuse application.
STRIPPED RAW WASTEWATER
REUSE
CANDIDATES
SOLVENT EXTRACTION
VAPOR RECOMPRESSION DISTILLATION
MEMBRANE PROCESSES
"CATALYZED" WET-AIR-OXIDATION
CANDIDATES
WET-AIR-OXIDATION
INCINERATION
TRICKLING FILTER
FIGURE 12. PHYSICAL/CHEMICAL OPTION
Distillation
The energy requirements for distillation processes can be greatly reduced by
staging (multi-effect evaporators) and/or using vapor-recompression systems.
The low quality heat, required for multi-stage evaporation, is readily
available on-site and distillation processes might be expected to serve as an
effective "bottoming cycle" to the conversion plant. High quality energy is
required to run the fan compressor deployed in the vapor-recompression
system; this, however, represents a small fraction (approximately 1/50) of
the energy required for single stage distillation. Concentration factors as
high as 25 have been achieved on raw condensate waters, with pretreatment
involving the sequential addition of acid to remove temporary hardness and
suppress carbonate scaling and caustic to ionize the phenolic compounds and
hold them in the concentrate during the distillation. The distillate may
require polishing (activated carbon treatment) for a high quality use such
as boiler feed make-up. The concentrate will probably require wet-air
oxidation or an equivalent treatment and evaporation to dryness. As the
167
-------
is deserving of continued attention to treat small flows of relatively
concentrated streams.
REUSE
A number of water reuse options which avoid release to a surface acceptor are
apparently evolving in commercial designs. As previously indicated, many
processes are net water consumers; thus there is incentive for closing the
"loop." The more prominent reuse options and their advantages and
disadvantages are compared in Figure 13. Ideally it is economically
desirable to use as poor a quality of water as the reuse application will
permit. Many of the recycle systems are merely paper designs and it is not
clear that special precaution has been taken to preclude the build-up of
trace extractables including organics and corrosion products which may
inadvertently react with or precipitate from the recycle loop, thereby
impairing the operation and reliability of the system. Many times such
difficulties are hard to anticipate. There appears to be room for more
systems engineering, tradeoff and optimization at the "tail" of the water
use cycle where the waste brine is typically concentrated for disposal.
Innovation should be encouraged in this part of the cycle.
REINJECTION INTO
CONVERSION PROCESS
MAKE-UP FOR COOLING
TOWERS
MAKE-UP FOR
FLUE GAS DESULFURIZATION
UNIT
POTENTIAL ADVANTAGES
• "ZERO" DISCHARGE
• DECREASES WATER USE
• CAN CONSERVE SENSIBLE HEAT
• LITTLE TREATMENT REQUIRED
• "ZERO" DISCHARGE
• DECREASES WATER USE
• SAME AS ABOVE
• COULD IMPROVE
FORCED OXIDATION OF
SLUDGE
POTENTIAL DISADVANTAGES
• CONCENTRATION STEP MAY BE
REQUIRED TO PRESERVE
WATER BALANCE
• PROCESS CHANGES MAY BE
REQUIRED TO ACCOMMODATE
• SIGNIFICANT PRETREATMENT
MAY BE REQUIRED
• MAY NOT BE APPLICABLE ON
YEAR-ROUND BASIS
• TRACE METAL PPT MAY
ADVERSELY AFFECT LIME
REACTIVITY
UNCERTAINTIES
• RELIABILITY AND COST
• SOME MATERIALS MAY
FAVOR RECYCLE LOOP.
MAKING PURGE NECESSARY
• NATURE OF SOLID WASTES
MAY BE MODIFIED
• RELIABILITY AND COST
• CONTAMINANTS IN DRIFT
MAY PRESENT PROBLEM
• CORROSIVE NATURE AND
SPARINGLY SOLUBLE
CONSTITUENTS MAY PRESENT
DIFFICULTY
' TRACE ELEMENTS IN
SLUDGE MAY PRESENT
PROBLEM
WET-DOWN FOR DUST
CONTROL AND IRRIGATION
BOILER FEEDWATER
FEED FOR SANITARY
SYSTEM
• "ZERO" DISCHARGE
• ASSISTS IN CONTROL OF
FUGITIVE EMISSIONS
• ABETS REVEGETATION EFFORT
• "ZERO" DISCHARGE
• SAVE POTABLE WATER
• SIGNIFICANT PHE-TREATMENT
REQUIRED
> INTRUSION OF TRACE ELEMENTS
AND REFRACTORY ORGANICS
INTO ECOSYSTEM AND AQUIFERS
• SIGNIFICANT TREATMENT AND
DEMORALIZATION REQUIRED
• TREATMENT REQUIRED TO
CONTROL COLOR AND ODOR,
ADDITIVES MAY WORK
• EXTENT OF NATURAL
"DETOXIFICATION" OF
CONTAMINATED WASTES
> COST BENEFIT
> COLOR AND ODOR MAY
CAUSE OBJECTIONS
• EFFECT ON BIOTREATMENT
PLANT
FIGURE 13. SELECTED WASTEWATER "REUSE" OPTIONS
168
-------
CONCLUSION
While existing wastewater treatment technology, which is primarily based on
biological oxidation, appears capable of meeting current discharge permit
requirements, there is a need for confirmatory operating and performance
data on large scale (low temperature) coal conversion facilities that
produce and treat large volumes of highly contaminated condensate water.
Considerable need and incentive exists for sustaining and expanding the
ongoing R&D on state-of-the-art wastewater treatment trains, physical/
chemical alternatives and sludge disposal options. The control and disposal
of secondary pollutants which has not been addressed in this presentation
should not be overlooked—the problem of salt disposal for "zero discharge
systems" remains a concern. Greater emphasis should perhaps be given to
system reliability, this aspect assuming greater import as plants get closer
to being operative.
ACKNOWLEDGEMENT
The technical contribution of a number of individuals and their respective
organizations which have worked for the Environmental Technology Division,
Office of Environmental Programs, Assistant Secretary for Environmental
Protection, Safety, and Emergency Preparedness, U.S. Department of Energy,
is acknowledged. Special appreciation is expressed to J. King, University
of California, for leading the recent wastewater workshop and providing
insight into solvent extraction processes; J. Klein, ORNL, for initiating
work on PNA control and the design and testing of an advanced treatment
train; C. Drummond, PETC, for comparative treatability studies on gasifica-
tion and liquefaction condensate waters; R. Luthy, Carnegie Mellon; and
F. Castaldi, Engineering Science, Inc., for an assessment of water treatment
options including solvent extraction and biological oxidation with PAC; and
D. Goldstein, WPA, for a survey of reuse alternatives and recommendations
for water management. Further details of these studies along with other
Governmental and industrial research is contained in the workshop proceedings
entitled "Processing Needs and Methodology for Wastewater from Coal, Oil
Shale and Biomass Synfuel Processes-II" and will be made available upon
request to the Division.
169
-------
CHARACTERIZATION OF COAL CONVERSION WASTEWATERS USING ON-SITE GC/MS t
by: C. J. Thielen and R. V. Collins
Radian Corporation
8501 MoPac Blvd.
Austin, TX 78766
ABSTRACT
This paper discusses a study which was done to characterize a wastewater
stream from a coal gasification facility using on-site extraction and GC/MS
analysis. The objectives of this program were to:
• Characterize the wastewater organic components primarily
for selected Priority Pollutants, Appendix C and Synfuels
compounds,
• Investigate the stability of these compounds under
refrigeration and ambient storage, and
• Evaluate the destruction of organics by wet oxidation.
Extractable material in the wastewater consisted primarily of phenols and
alkylphenols. These compounds accounted for about 98 percent of the total
organic mass identified. Several polynuclear aromatic (PNA) compounds were
also identified. Deterioration in the composition of the sample was observed
over a one month period. This was most evident in the concentration of
dimethylphenols which dropped approximately 75 percent during two weeks of
refrigerated storage. Ambient sample storage produced a greater decrease in
the concentration of phenol but did not appear to affect the alkylphenols or
the base/neutral compounds as much as phenol. It is expected that the
observed changes in composition would hamper any off-site wastewater treat-
ability studies with water of this type. Treatment of the wastewater by wet
oxidation was also evaluated and found to remove greater than 90% of the
extractable organics.
INTRODUCTION
The Chapman-Wilputte gasifier at the Holston Army Ammunitions plant in
Kingsport, Tennessee, has been the site of several environmental assessment
tests. This study deals with an effort to characterize more accurately the
aqueous process condensate (separator liquor wastewater) at this facility.
Previous studies have shown the Holston process condensate to be similar in
170
-------
composition to the Lurgi process condensate from the Kosovo plant* especially
with respect to phenolic compounds.
Lurgi gasification facilities have been proposed for commercial plants
in the United States but current operating facilities are not easily access-
ible. The Holston plant is located in the United States and provides a
readily accessible source ^f coal gasification wastewater for characteriza-
tion and treatment system development.
Successful treatment of wastewaters requires a good understanding of the
composition of the wastewater and the chemistry involved in any decomposi-
tion. Previous characterization studies may have been conducted on samples
that had deteriorated during shipment and cold storage. Analysis of a deter-
iorated sample can produce results which do not accurately reflect the com-
position of the water as it would be fed to a treatme'nt system.
This study included immediate, on-site extractions which were performed
in an attempt to minimize any sample deterioration. The on-site Hewlett-
Packard Model 5993B gas chromatograph/mass spectrometer (GC/MS) provided
immediate analysis of the sample extracts as well as the positive identifica-
tion of any compound present. It could also identify and track the appear-
ance and/or disappearance of compounds during decomposition.
The main objectives of this program were:
• to provide a more accurate characterization of the aqueous
process condensate;
• to investigate sample stability during refrigerated
and ambient storage; and
• to investigate the effects of wet oxidation on this
wastewater.
PROCESS DESCRIPTION
The Chapman-Wilputte gasification process uses an air-blown, atmospheric
pressure gasifier. It gasifies approximately one ton of bituminous coal an
hour to produce a fuel gas with average heat content of 150 Btu/scf. The
product gas is first quenched then cleaned via direct contact with the pro-
cess water. The resulting liquor collects in a liquor separator. Here the
aqueous layer is decanted from the oils and tars which were removed from the
gas stream. The tar layer is recovered for use as a supplemental boiler fuel
and the aqueous layer is recirculated through the gas quenching/scrubbing
system.
*Collins, R. V., K. W. Lee, and D. S. Lewis. Comparison of Coal Conversion
Wastewaters. EPA 600/9-81-006. Contained in the Symposium Proceedings:
Environmental Aspects of Fuel Conversion Technology V, St. Louis, MO,
(September, 1980). Radian Corporation, Austin, TX, January, 1981.
171
-------
Excess water is blown down via an over flow weir to a holding sump.
From there, the accumulated water is periodically pumped to a forced evapora-
tor system for ultimate disposal. The brine and tar resulting from the evap-
oration is returned to the separator. There is no fresh water make up to the
system and the net accumulation of water is minimized through the proper
operation of the gasification and cleaning systems. Grab samples of the pro-
cess condensate were collected from the aqueous layer in the separation tank
at the point indicated in Figure 1.
PROCEDURES
SAMPLING
Grab samples of the aqueous process condensate were collected from the
separator near the off take for the recirculating quench system. The water
at this point contained a minimum amount of tar. The pH and temperature of
the water were measured at the time of collection.
EXTRACTION
Samples were extracted using a base/neutral-acid extraction procedure.
The separator liquor was first basified to pH^ 12 with NaOH and extracted
with methylene chloride followed by diethyl ether to obtain the basic and
polynuclear aromatic compounds. It was then acidified to pH <2 with HC1 and
extracted as before to obtain the phenolic compounds. The pH adjustment
provided enhanced recovery of the basic and acidic compounds and the diethyl
ether provided a polar medium for enhanced extraction of phenols. Extracts
were concentrated by a factor of ten using a Kuderna-Danish apparatus.
Further concentration was not possible due to the large amount of material
present in the extracts.
ANALYSIS
Extracts were analyzed using EPA Priority Pollutant conditions* for
acid, base/neutral and purgeable componds on a Hewlett-Packard 5993B gas
chromatograph/mass spectrometer.
CHARACTERIZATION OF THE AQUEOUS PROCESS CONDENSATE
EXTRACTABLE COMPOUNDS
The base/neutral and acid extracts were analyzed for EPA base/neutral
and acid extractable Appendix A Priority Pollutants as well as those com-
pounds listed as Appendix C and Synfuels "priority pollutants". Additional
compounds that might occur in coal conversion processes were included based
*U. S. Environmental Protection Agency. Federal Register 44(233),
69464-69575, 1979.
172
-------
LIQUOR
SEPARATOR
VAPORS
LOW-BTUGASTO
PROCESS FURNACES
EVAPORATOR
GASES
t
SIZED COAL
LEGEND
SAMPLE POINT
EVAPORATOR
SUMP
1
EVAPORATOR
STEAM
Figure 1. Location of water sampling point in process
-------
on their documented behavior in biological oxidation systems* or their known
or suspected carcinogenic activity.
Table 1 summarizes the results of the characterization study. The
majority of the compounds identified were of a phenolic nature. Phenol,
methylphenol, and dimethylphenol account for 98 percent of the total extract-
able organics. Significant levels of PNA's were also found. The values
given represent the average concentration and the range of these values dur-
ing the six days of monitoring.
The ranges of values indicate that significant variability exists in the
data. In an effort to find the source of this variability, a followup study
was done using Hols ton separator liquor which had been collected 24 hours
prior to extraction and stored in amber bottles at 4°C since the time of col-
lection. Determinations of instrument variability and extraction variability
were made by replicate extractions and replicate analyses of the extracts.
Values for representative compounds for instrument variability are given in
Table 2. Values for representative compounds for overall (extraction and
analysis) variability are listed in Table 3. The instrument variability for
total chromatographable organics (TCO) is 1.7 percent while overall variabil-
ity is 13 percent. This indicates that essentially all of the variability
(12.9 percent) is due to extraction for chromatographable organics as a
group. This same trend is also seen in the representative compounds presen-
ted in the tables except for naphthalene which is close to the detection
limit. While the relative standard deviation (la) for the overall variabil-
ity of phenol is only 16 percent, the cresols and naphthalene vary by 47
percent and 34 percent, respectively. These same trends were also observed
in the samples which were extracted on-site.
Part of this variability may be due to the complexity of the wastewater
sample matrix. The extraction procedure does not produce a clean separation
between the base/neutrals (B/N) and the acids (A). Much of the phenolics
were extracted into the B/N fraction. The B/N extracts were also analyzed
for phenols and the concentrations of phenols found in this fraction added to
the values obtained in the acid fraction. The magnitude of this premature
extraction is shown in Table 4. The concentrations of phenol show the
greatest amount of variability in the base/neutral extract while the con-
centrations of phenol in the acid extract remain fairly constant. The
dimethylphenols, some of the least acidic phenolic compounds identified,
demonstrated the greatest amount of extraction into the B/N fraction.
VOLATILE ORGANICS
Volatile species (purgeable halocarbons and aromatics) were also deter-
mined in the raw water. These compounds, listed in Table 5, do not show
*Singer, P. C., F. K. Pfander, J. Chinchilli, A. F. Maciorowski, J. C. Lamb
III, and R. Goodman. Assessment of Coal Conversion Wastewaters:
Characterization and Preliminary Biotreatability. EPA 600/7-78-181, PB-294
338. University of North Carolina, Department of Environmental Sciences and
Engineering, Chapel Hill, NC, September, 1978.
174
-------
TABLE 1. CHARACTERIZATION DATA FOR EXTRACTABLE SPECIES IN SEPARATOR LIQUOR
Concentration
Acid
t
+
t*
Identifications /Compound
Extrac table Compounds
Phenol
Methylphenols
Dimethylphenols (total)
2 , 4-Dimethylphenol
Trimethylphenol
Indanol
1-Naphthol
2-Naphthol
Resorcinol/Catechol
Hydroxybenzaldehyde
in the Liquor (mg/L)
Average
2400
3200*
1200
420
0
1
5
6
30
5
.82
.7
.0
.7
.7
1900
1500
330
98
0
< 0
3
5
3
< 0
Range
.35
.07
.3
.4
.6
.18
3400
4700
1900
820
2
3
8
9
65
19
.2
.2
.5
.2
Base/Neutral ExCractable Compounds
t
t
t
j.
\
t
t
t
/t
/
/t
/
/
Naphthalene
Acenaphthylene
Fluorene
Phenanthrene/ Anthracene
Fluoranthene
Pyrene
Bis(2-ethylhexyl)phthalate
Chrysene
Benzo(b) f luoranthene
Benzo (a)pyrene
Pyridine
2-Ethylpyridine
Qu incline
4-Met hy Iquinol ine
1-Methylnaphthalene
2,3-Dimethylnaphthalene
2 , 6-Dimethylnaphthalene
Indole
2-Methylindole
3-Methylindole
8
3
2
2
5
5
12
0
0
0
1
18
3
0
2
2
2
12
12
2
.6
.6
.6
.3
.7
.7
.12
.10
.12
.2
.1
.11
.3
.3
.2
.4
1
1
0
0
0
0
1
1
0
0
0
< 1
8
2
0
.6
.2
.28
.7
.3
.4
.2
.3
.62
.43
.65
.3
.2
.58
17
< 6
< 6
< 6
< 9
< 9
32
61
5.
4.
< 4.
< 3
14
16
3.
.6
,2
,5
6
t Appendix A Priority Pollutants
-: One data point not included in the average was rejected due to extremely high value,
but compound was identified 6 of 6 times.
+ Includes 2,4-DMP.
£ The portion of 2,4-DMP from the B/N fraction was estimated from the amount found in the
acid fraction. The value presented here represents the sum of the acid and B/N fraction.
/ Identified only in one of six samples.
175
-------
TABLE 2. VARIABILITY IN ANALYSIS
Average* % Relative Std.
(mg/L) Deviation (£ x 100)
x
Phenol
Cresol
Naphthalene
Total Chromatographable Organics
2000 + 190
1200 + 210
3.2 + 1.8**
4200 + 70
9.5
17.5
56.0
1.7
* For three determinations.
**Close to detection limit.
TABLE 3. VARIABILITY IN EXTRACTION AND ANALYSIS
Average*
(mg/L)
% Relative Std.
Deviation (0_ x
Phenol
Cresol
Naphthalene
Total Chromatographable Organics
1700 + 230
800 + 380
3.5 + 1.2**
3700 + 480
16 %
47 %
34 %
13 %
* For three determinations of the combined variability (extraction and
analysis).
**Close to detection limit.
176
-------
TABLE 4. EXTRACTION OF PHENOLIC COMPOUNDS INTO BASE/NEUTRAL EXTRACT
Compound
Phenol (mg/L)
Acid
B/N
Total
B/N % of Total
Day
1
1900
370
2300
16
Day
2
2200
590
2800
21
Day
3
1600
1800
3400
53
Day
4
1800
470
2300
20
Day
5
1800
120
1900
6
Day
6
1800
100
1900
5
Methylphenols (mg/L)
Acid
B/N
Total
B/N % of Total
Dimethylphenols
Acid
B/N
Total
B/N % of Total
1800
2400
4200
57
(mg/L)
230
1300
1500
87
1500
3200
4700
68
400
1300
1700
76
690
30000
30700
98
44
1900
1940
98
1200
2900
4100
71
120
1100
1200
92
900
580
1500
39
130
220
350
63
1100
430
1500
29
81
250
330
76
177
-------
TABLE 5. CHARACTERIZATION DATA FOR VOLATILE ORGANIC COMPOUNDS
Average* % Relative Std.
(ug/L) Deviation (o_x 100)
Benzene
Toluene
Ethylbenzene
Total Xylenes
630
420
48
280
12 %
9 %
69 %
25 %
*For six determinations.
178
-------
the level of variability seen in the extractable compounds since the addi-
tional variability associated with extraction was not introduced.
INVESTIGATION OF SAMPLE STABILITY
The second objective of this program was to determine the effects of
refrigeration and ambient storage of the wastewater on the stability of its
composition. The refrigerated sample of water was stored in an amber bottle
at 4°C for one month, while the ambient sample was stored in an amber bottle
at ambient temperatures (up to approximately 35°C) for three weeks. These
samples were then extracted and analyzed. The results were then compared to
aliquots of the same sample which had been extracted immediately on-site.
Figure 2 illustrates how the concentrations of representative compounds de-
creased with storage. The figure shows that the concentrations of most com-
pounds appear to decrease more during ambient storage than when kept under
refrigeration. However, a high relative error associated with the analysis
may account for some of the differences observed in concentration between the
ambient and refrigerated samples.
Physical changes were also observed in the sample stored under ambient
conditions. These include a darkening of the color as well as an increase in
the turbidity of the water.
WASTEWATER TREATABILITY BY WET OXIDATION
The third objective of this study was to evaluate the efficiency of
removal of organics from process condensate by wet oxidation. The wet oxida-
tion apparatus used to treat the wastewater is shown in Figure 3. Immediate-
ly after collection, approximately one liter of the water sample was placed
in a stainless steel bomb, heated to 500°F and simultaneously pressurized to
1500 psig with zero air. The bomb remained under these conditions for about
30 minutes. After cooling and then depressurizing, an aliquot of the oxi-
dized water was extracted, analyzed and compared to an aliquot of the unoxi-
dized water sample which had been extracted immediately after sampling.
Table 6 compares the concentrations of each compound determined in the
fresh aliquot to those determined in an aliquot of the same sample after wet
oxidation. The amount of total extractable organic material is significantly
reduced, from 8000 mg/L to approximately 600 mg/L or 8 percent of the origi-
nal amount. The level of total phenols was reduced to approximately 10 per-
cent of the original concentration. Phenol itself showed the least loss with
an 85 percent reduction compared to methylphenols and dimethylphenols which
exhibited about a 95 percent reduction in concentration. The less than
values represent the detection limit of the instrument for each day of analy-
sis.
Wet oxidation significantly reduced the high concentration of the phenol
and alkylated phenols. Previous work by Singer, et al* shows that these
Singer, P. C., 1978, (op. cit.).
179
-------
OO
o
100%-
PERCENT OF
ORIGINAL
CONCENTRATION
REMAINING
50%-
25%-
PHENOL
METHYLPHENOLS
NAPHTHOLS
INDOLE
(REFRIGERATED STORAGE
| (AMBIENT STORAGE
Figure 2. Results of the stability study
-------
00
PRESSURE
GAUGE
CYLINDER
REGULATOR
AIR
CYLINDER
VENT
VALVE
f f
PRESSURE
RELIEF VALVE
HEATING
MANTLE
BOMB
THERMOCOUPLE
READOUT
316 STAINLESS
STEEL BOMB
INSULATION
Figure 3. Schematic of the wet oxidation apparatus
-------
TABLE 6. EFFICIENCY OF ORGANIC COMPOUND REMOVAL
FROM SEPARATOR LIQUOR BY WET OXIDATION
Concentration (mg/L)
Compound
Resistance
to Bio-
degradation t
Total Extractable Organics (as determined by
the sum of the total chromatographable organics
and the gravimetric residue after evaporation)
Phenol
Methylphenols
Dimethylphenols
2 , 4-Dimethylphenol
Trimethylphenol
Indanol
1-Naphthol
2-Naphthol
Resor cinol/ Cat echo 1
Hydroxybenz aldehyde
Naphthalene
Acenaphthylene
Fluorene
Phenanthrene/ Anthracene
Fluoranthene
Pyrene
Bis(2-ethylhexyl)phthalate
Chrysene
Benzo (b) f luoranthene
Benzo(a)pyrene
Pyridine
2-Ethylpyridine
Quinoline
4-Methylqu incline
1-Methylnaphthalene
2,3-Ditnethylnaphthalene
Indole
3-Methylindole
E
E
E/R
NR
M/R
R
M
E
E
E/M
E
NR
NR
NR
NR
NR
NR
NR
NR
NR
R
R
E
R
E
R
M
E
Fresh
Sample
7900
1900
1500
330
120
2.
1.
4.
6,
20
< 0,
1.
1.
0.
0,
0.
0.
1.
0,
0.
0.
1.
1.
0,
0.
0.
0,
8.
1.
.2
.2
3
.0
,3
,6
.5
.49
.70
,3
,4
.4
.12
.10
.12
2
.3
,62
.11
.43
,65
.0
After
Wet
Oxidation
600
280
80
12
< 2.8
< 0.63
< 0.27
< 0.63
< 0.27
< 2.0
2.2
< 1
1.0
< 2
< 2
< 3
< 3
0.34
< 1
< 1.6
< 1
< 1
< 1.7
< 1.1
< 1.1
< 1.2
< 1.6
< 3.4
< 1
% Removal
Efficiency
from Control
Sample
92
85
95
96
> 98
> 71
> 78
> 85
> 96
> 90
> 90
> 38
33
NA
NA
NA
NA
76
NA
NA
NA
NA
NA
NA
NA
NA
NA
> 58
> 29
7.
7.
7.
7.
7.
7.
7.
7.
%
%
%
7.
%
•/.
7.
t Reference 2
E Easily Degraded
M = Moderately Degraded
R - Resistant to Degradation
NR = Biodegradation data not reported
NA = Not Applicable
182
-------
compounds are also easily treated by bioxidation. However, wet oxidation
also has the ability to reduce the levels of organic compounds which are not
readily treated by biological systems. Figure 4 illustrates the efficiency
of this reduction in the levels of a few representative compounds which are
moderately biodegradable and/or resistant to biological treatment. The
concentration of 1-naphthol, which is moderately resistant, was reduced by
greater than 85 percent; trimethylphenol, which has both resistant and moder-
ately resistant isomers, was reduced by 71 percent and the concentration of
indanol, which is resistant to bioxidation, was reduced by greater than 78
percent. In all cases this reduction is greater than 70 percent. This value
is outside the limits of the analytical variability discussed previously,
indicating definite trends in the removal of organics. Table 6 also indi-
cates the resistance to biodegradation (where available) for each of the
other compounds not discussed in this section.
However, despite this efficiency, the overall feasibility of wet oxida-
tion is limited. This is because 1) this technology is still in the devel-
opmental stages and 2) there are high costs associated with this process. To
date, its usefulness is limited to a few specific applications where there is
a need for treatment of highly toxic and/or small volume organic laden
streams.
CONCLUSIONS
The following statements summarize the results of this study.
• A loss of sample integrity during sample handling and
storage is indicated even when samples are refrigerated.
• Much variability is associated with the complex matrix
of this aqueous process condensate. Better separation
procedures are required before these samples can be more
accurately quantified.
• It is possible to operate a GC/MS system under field
conditions.
• About 95 percent (by mass) of the identified compounds
are readily biodegradable.
• Wet oxidation reduced the levels of extractable organics
by greater than 90 percent.
• Wet oxidation reduced the levels of some compounds which
are not readily biodegradable.
RECOMMENDATIONS
The following recommendations, are presented in response to difficulties
encountered during sample analysis and data reduction. Since a large source
of variability seems to be associated with the extraction of phenols into the
183
-------
4.3 —
CONCENTRATION
(mg/L)
2.2 —
00
.6 —
1-NAPHTHOL
(MODERATELY)
TRIMETHYLPHENOL
(MODERATELY/RESISTANT)
INDANOL
(RESISTANT)
(FRESH SAMPLE
| | AFTER WET OXIDATION
( ) INDICATES RESISTANCE TO BIODEGRADATION
Figure 4. Wet oxidation results for compounds not easily biodegradable
-------
base/neutral fraction, an extraction technique which provides good partition-
ing of acid and base/neutral compounds is required. This might be achieved
by an acid/neutral followed by a basic extraction, then separation of the
acid and neutral compounds by liquid chromatography or a less vigorous
extraction of base/neutral compounds, using only methylene chloride for the
base/neutral compounds, but continuing with a methylene chloride/diethyl
ether extraction for the acidic compounds.
Sample analysis could also be facilitated by using a capillary column to
provide better chromatographic separation in place of a packed column specif-
ied by EPA protocol. Use of the capillary column would allow better specia-
tion of the compounds present.
The need for on-site extraction and GC/MS analysis has not been estab-
lished. The possiblity of on-site extraction/off-site analysis should also
be investigated. The stability of the extracted samples should be evaluated
by analysis of the extract immediately after extraction and at predetermined
intervals following the extraction to monitor any decrease in one or more
compounds. If the stability of extracted samples is adequate to allow trans-
port and storage, the expense of providing on-site analysis could be avoided.
185
-------
TREATMENT OF WASTEWATER FROM A FIXED-BED ATMOSPHERIC COAL GASIFIERt
by: Philip C. Singer and Eli Miller
Department of Environmental Sciences and Engineering
School of Public Health
University of North Carolina
Chapel Hill, North Carolina 27514
ABSTRACT
Previous studies using a simulated coal conversion wastewater have
demonstrated the feasibility of treating this type of waste by an activated
sludge process. Phenol concentrations were reduced to levels below 1 mg/1
and the toxicity and mutagenicity of the simulated wastewater were reduced
substantially by the biological treatment. This paper will present the
results of an evaluation of the biological and subsequent physical-chemical
treatability of a real coal conversion wastewater, along with a comparison
of the results with those obtained using the simulated wastewater.
Coal gasification wastewater was obtained from a Chapman gasifier at
the Hols ton Army Ammunition Plant in Kingsport, Tennessee. The wastewater
was diluted to 25% of full-strength, supplemented with phosphate, and
subjected to aerobic biological treatment in a 22.5-liter completely-mixed
activated sludge reactor. The reactor was operated at a solids retention
time of 20 days and a hydraulic detention time of 10 days. In addition to
characterizing the quality of the effluent using various chemical and bio-
assay procedures, the effluent from the biological reactor was subjected to
a series of physical-chemical treatment steps consisting of chemical coagu-
lation, ammonia stripping, ozonation, and activated carbon adsorption. The
chemical quality and bioassay characteristics of these various samples will
be presented.
INTRODUCTION
Previous research at the University of North Carolina has dealt pri-
marily with an assessment of the biological treatability of a simulated coal
conversion wastewater (1,2,3). A 25% dilution of the simulated wastewater
was fed to a series of completely-mixed activated sludge reactors, operated
at several different solids retention times (sludge ages). The results indi-
cated that TOG, COD, and BOD removal increase with increasing sludge age,
and that phenol is essentially completely removed with a sludge age of 5 days.
Cresols and xylenols required 10 and 20 days, respectively, for removal to
levels below 1 mg/1. Bioassays of the raw and treated quarter-strength simu-
lated wastewater showed that the acute toxicity of the wastewater to fish and
to mammalian cells is reduced markedly as a result of the biological treat-
186
-------
ment and that the extent of the reduction in toxicity increases with
increasing sludge age. Additionally, at the concentrations tested, biological
treatment reduces the mutagenic activity associated with the raw simulated
wastewater to undetectable levels.
BIOLOGICAL TREATMENT OF HOLSTON WASTEWATER
More recently, we were able to obtain a real coal gasification waste-
water from the Holston Army Ammunition Plant in Kingsport, Tennessee. The
Holston facility has a fixed-bed, atmospheric Chapman gasifier which produces
a low Btu gas which is used as fuel for process heaters. The wastewater
sample was collected by R. Collins of the Radian Corporation (4) from the
separator liquor tank which receives process condensate and condensed tars
and oils from the gas-quenching and scrubbing steps at the Holston facility.
Separation of tars and oils was reasonably good as the aqueous wastewater
sample was relatively free of particulate material. The wastewater was sealed
in 55-gallon drums to preserve its chemical integrity and shipped to our
laboratories in Chapel Hill. Upon receipt of the drums, a sample of the
virgin wastewater was collected, under an argon atmosphere, for chemical
analysis and for various aquatic and health effects bioassays. The remaining
contents of the drum were re-sealed and stored under an argon atmosphere in
order to avoid exposure of the wastewater to oxygen and to minimize the loss
of volatile constituents of the wastewater.
Table 1 presents the chemical characteristics of the virgin Holston
wastewater as it was received. Two different shipments were received, and
the characteristics of each of the batches are shown. Batch 2, the second
shipment, is stronger than Batch 1, particularly with respect to COD and
ammonia. The composition of the simulated coal conversion wastewater used in
our earlier studies (1, 2, 3) is shown for comparison. The concentrations of
phenols, TOC, and COD in the simulated wastewater are comparable to those in
Batch 1 of the Holston wastewater; the ammonia concentration is appreciably
lower.
TABLE 1. CHARACTERISTICS OF VIRGIN HOLSTON WASTEWATER
Concentration, mg/1*
Parameter
TOC
COD
BODr
4-AAP Phenols
CN~
SCN
NH , as N
pHJ
Batch 1
5,450
14,800
8,000
2,000
4.1
600
3,770
8.0
Batch 2
7,090
25,000
-
2,320
21.7
950
7,260
8.04
Simulated Coal
Conversion Wastewater
4,640
14,300
7,070
2,240
-
-
1,000
7.1
*Except pH
187
-------
Batch 1 of the raw Holston wastewater was also analyzed for selected
trace metals and polynuclear aromatic hydrocarbons (PAH). These analyses
were performed on samples taken several weeks after the drum was first
opened so that a significant amount of suspended material was found in the
aged wastewater. Accordingly, both the aqueous and solid phases were
analyzed. Table 2 shows the concentrations of these selected priority pollu-
tants in the raw wastewater. With the exception of zinc which was present
at a concentration of 1.3 mg/1, the trace metals were found at concentrations
less than 0.2 mg/1. The concentrations of each of the PAH were less than
0.1 mg/1; the high value reported for pyrene is questionable.
TABLE 2. CONCENTRATIONS OF TRACE METALS AND POLYNUCLEAR
AROMATIC HYDROCARBONS IN RAW HOLSTON WASTEWATER*
Concentrations, mg/1
Dissolved Suspended
Total
Metals
Cr
Cu
Mn
Zn
Pb
0.032
0.056
0.020
0.828
0.080
0.016
0.144
0.104
0.496
0.056
0.048
0.200
0.124
1.324
0.136
PAH
Naphthalene
Fluorene
Phenanthrene
Anthracene
Pyrene
0.024
0.008
<0.012
0.048
0.528**
<0.036
<0.016
<0.048
<0.044
<0.056
0.024-0.060
0.008-0.024
< 0.060
0.048-0.092
0.528-0.584**
* Batch 1
** Questionable
Table 3 shows the toxicity of Batch 1 of the virgin Holston wastewater to
Daphnia, fathead minnows, and the Chinese hamster ovary (CHO) mammalian cell
system. The toxicities are relatively comparable for each of the bioassay
systems, with LC50s on the order of 0.1%, i.e. 0.1 ml of wastewater diluted
in 100 ml of clean water will cause 50% lethality of each of the bioassay
indicators. Again, for purposes of comparison, the toxicity of the full-
188
-------
strength simulated wastewater is also shown in Table 3.. From a toxicity
standpoint, the Holston wastewater is approximately four to five times
stronger (more toxic) than the simulated wastewater with which we previously
worked.
TABLE 3. TOXICITY OF VIRGIN HOLSTON WASTEWATER
A. Virgin Holston Wastewater*
LC50, %
Aquatic Toxicity
24-hr.
48-hr.
72-hr.
96-hr.
Daphnia
Fathead Minnow
0.28
0.11
0.11
0.10
0.09
0.09
Mammalian Cytotoxicity
LC50, %
CHO Monoclonal Assay
0.12
B. Simulated Coal Conversion Wastewater
LC50, %
Aquatic Toxicity
24-hr.
48-hr.
72-hr.
96-hr.
Daphnia
Fathead Minnow
0.41
0.5
0.21
0.5
0.19
0.49
0.49
Mammalian Cytotoxicity
CHO Monoclonal Assay
LC50, %
0.48
*Batch 1
The wastewater was diluted to 25% of full-strength, supplemented with
phosphate, and subjected to aerobic biological treatment in a 22.5-liter
completely-mixed activated sludge reactor. The reactor was operated at a
20-day solids residence time and a 10-day hydraulic retention time. No other
pre-treatment was provided. Table 4 shows the chemical quality of the
reactor effluent compared to the diluted raw feed. Both batches of wastewater
appear to be treated relatively effectively, with TOC removals of approxi-
mately 66% and 62% for batches 1 and 2, respectively, and COD removal
averaging 63% and 62%, respectively. The average effluent TOCs and CODs are
respectively 510 and 1650 mg/1 for batch 1 and 629 and 2145 mg/1 for batch 2.
The differences presumably are due to the fact that batch 2 is appreciably
stronger than batch 1.
In both cases, substantial removal of phenols (as measured by the
4-aminoantipyrene wet chemical procedure) occurred. The residual concen-
tration of phenols was frequently below 1 mg/1. HPLC analysis showed that
phenol itself was usually on the order of 0.1-0.2 mg/1 in the reactor
effluent.
189
-------
TABLE 4. QUALITY OF BIOLOGICALLY-TREATED- HOLSTON WASTEWATER
vo
o
Concentration, mg/1**
Batch 1
Parameter
TOC
COD
BOD5
4-AAP Phenols
CN~
SCN"
NH , as N
pH
Reactor
Influent
1510
4490
1700
526
1.0
173
882
7.48
Reactor
Effluent
510 (±81)
1650 (±209)
26
1.2-3.3
1.0
162-193
874
7.45
Batch 2
Reactor
Influent
2000
5580
—
498
—
253
1810
7.0
Reactor
Effluent
629 (±67)
2145 (±470)
—
0.2-3.6
—
443-483
1890
7.0
Simulated Coal
Conversion Wastewater*
Reactor
Influent
1410
3326
1770
560
—
—
252
7.1
Reactor
Effluent
204 (±51)
511 (±121)
22
0.14-2.6
—
—
212
6.9
^'Activated sludge; 10-day HRT, 20-day SRT
**Except pH
-------
No biological nitrification was observed, with effluent ammonia concen-
trations being similar to the influent ammonia concentration. No thiocyanate
removal was apparent, although our results indicate an apparent increase in
SCN for batch 2. Such a production of SCN has not been reported previously,
yet we have measured this increase consistently, and have verified our
analytical results using step addition procedures. Thiocyanate was measured
using the spectrophotometric dithiocyanatopyridine chloroform extraction
procedure (5).
Table 4 also shows the quality of the biologically-treated simulated
coal conversion wastewater under parallel treatment conditions, i.e. diluted
to 25% of full-strength and treated by an activated sludge system with a
solids retention time of 20 days and a hydraulic retention time of 10 days.
Treatment of the simulated wastewater was more effective, providing an 86%
reduction in TOC and an 85% reduction in COD. The effluent TOC and COD
concentrations are approximately 1/2 to 1/3 of those in the biologically-
treated Holston effluent.
Table 5 shows the toxicity of the raw and biologically-treated Holston
wastewater. The "raw" LC50s refer to the 25% diluted Holston wastewater
corresponding to the influent to the biological reactors. It is apparent
that there is a significant reduction in aquatic toxicity to the Daphnia and
fathead minnows, and in the CHO mammalian cytotoxicity; 5 to 15-fold reduc-
tions in toxicity result from the biological treatment of the diluted Holston
wastewater, using these assay systems. Because of the variability in effluent
quality and in order to provide toxicity data for both batches of the raw
wastewater, the bioassays were performed several times, as indicated by the
dates in Table 5. (The reactor feed was switched from batch 1 to batch 2
in early April, 1980.) The LC50 values seem to be fairly consistent
irrespective of this variability in gross chemical quality.
A comparison between the toxicity of the biologically-treated Holston
wastewater and the biologically-treated simulated coal conversion wastewater
(see Table 5) shows that the Holston effluent is appreciably more toxic to
the three bioassay systems tested. Hence, despite the effectiveness of
biological treatment in removing TOC and COD and in reducing the toxicity of
the Holston wastewater, the biologically-treated effluent is still of unac-
ceptable quality for discharge to the aquatic environment. The residual TOC
and COD are still appreciable, as are the NH_ and SCN concentrations. The
toxicity of the effluent is also still substantial, suggesting that additional,
i.e. post-biological, treatment is appropriate.
A 25% dilution of the virgin Holston wastewater was assayed for muta-
genicity using the Ames test. Preliminary screening experiments showed
that TA98 (a strain of Salmonella which tests for frameshift mutagenic
activity) to be the most sensitive strain for this wastewater. With
metabolic activation (the incorporation of the S-9 rat liver homogenate into
the test system), TA98 gave a positive mutagenic response at all sample
volumes tested up to 2.5 ml. The highest reversion ratio of 3.4 occurred
for 1.5 ml of the wastewater sample.
191
-------
TABLE 5. TOXICITY OF RAW* AND BIOLOGICALLY-TREATED** HOLSTON WASTEWATER
A. Holston Wastewater
AQUATIC TOXICITY
Daphnia
Influent
Influent
Effluent
Effluent
Effluent
Effluent
Fathead Minnow
Influent
Influent
Effluent
Effluent
Date
8/26/80
6/23/81
10/1/80
10/27/80
2/16/81
6/29/81
8/12/80
5/15/81
10/1/80
10/15/80
24-hr.
1.12
1.8
6.5
4.5
5.6
4.8
0.44
1.1
11
9.6
LC50, %
48-hr.
0.44
0.76
4.6
4.1
3.5
3.3
0.40
1.0
6.8
5.9
72-hr. 96-hr.
—
0.53 0.49
3.9
3.9
3.2
2.5
0.38 0.38
1.0 1.0
5.7 5.2
4.9
MAMMALIAN CYTOTOXICITY
CHO Monoclonal
Influent
Influent
Effluent
Effluent
Effluent
Assay
B. Simulated
Date
8/26/80
6/23/81
9/26/80
6/28/81
7/14/81
LC50, %
0.48
0.52
4.90
3.64
7.01
Coal Conversion Wastewater
AQUATIC TOXICITY
Daphnia
Influent
Effluent
Fathead Minnow
Influent
Effluent
24-hr.
1.65
57
2.0
Ind.***
48-hr.
0.85
49
2.0
Ind . ***
LC50, %
72-hr.
0.7
38
2.0
Ind.***
96-hr.
<0.42
—
1.9
Ind.***
MAMMALIAN CYTOTOXICITY
CHO Monoclonal
Influent
Effluent
Assay
LC50, %
1.9
15.7
*25% diluted Holston Wastewater
**10-day HRT, 20-day SRT activated sludge system
***Greater than 50% of the test organisms survived at concentrations up to
50% of the effluent.
192
-------
Following biological treatment, no frameshift mutagenic activity, with
or without metabolic activation, was found at sample volumes up to 2.0 ml
using tester strains TA98 and TA1537. Additionally, no base-pair substi-
tution mutagenic activity, using tester strain TA100 with or without
metabolic activation, was found at sample volumes up to 2.0 ml of the treated
wastewater. Apparently, mutagenic activity in the Holston wastewater was
reduced to undetectable levels by biological treatment.
TABLE 6. MOLECULAR WEIGHT DISTRIBUTION OF RESIDUAL ORGANIC
CARBON IN FILTERED ACTIVATED SLUDGE EFFLUENT
Molecular Weight TOG, mg/1
<500 390
500 to 30,00 70
>30,000 200
TOTAL 660
It is also worth noting that the filtered (0.45 ym) effluent following
biological treatment contains a significant amount of high molecular weight
organic material as shown in Table 6. The molecular weight distribution was
measured by ultrafiltration techniques, using two different membranes with
nominal molecular weight cut-offs of 500 and 30,000. Of the 660 mg/1 of TOG,
approximately 30% or 200 mg/1 (on a carbon basis) consisted of organics with
a molecular weight greater than 30,000. Sixty percent, or 390 mg/1, of the
TOC consisted of organics of molecular weight less than 500. The remainder
of the TOC consisted of compounds with a molecular weight in the 500-30,000
range. The fact that approximately 40% of the residual TOC following
activated sludge biological treatment is comprised of compounds with a
molecular weight greater than 500 implies that the residual TOC may cause
problems if the biologically-treated effluent is to be recycled for use in
a cooling tower. It is conceivable that these high molecular weight
compounds will tend to adsorb to heat transfer surfaces in the tower,
thereby fouling the tower and interfering with its operation. The amena-
bility of this high molecular weight organic material to various post-
biological treatment processes should be examined.
POST-TREATMENT OF BIOLOGICALLY-TREATED HOLSTON WASTEWATER
Filtered effluent from the biological reactors treating diluted Holston
wastewater was subjected to a variety of physical-chemical treatment steps
consisting of chemical coagulation and precipitation, ozonation, activated
carbon adsorption, and ammonia stripping. The effectiveness of these post-
biological treatment processes was assessed through measurements of TOC,
COD, NH3, SCN~, and residual Daphnia and CHO toxicity.
193
-------
Table 7 shows the results of coagulation and precipitation experiments
on the filtered biologically-treated Holston wastewater. Alum (aluminum
sulfate) and ferric chloride are standard water supply and wastewater treat-
ment coagulants and have been shown (6) to effectively remove high molecular
weight humic substances from water. Nevertheless, the application of these
coagulants,even at extreme doses of up to 500 mg/1, resulted in no apparent
floe formation. The chemicals were added to the wastewater, and the water
was rapid-mixed to disperse the chemical, slow-mixed to allow for floccu-
lation, and allowed to stand quiescently to provide for settling of any floe
or precipitate. The fact that aluminum hydroxide or ferric hydroxide wasn't
produced suggests that a substantial concentration of metal-complexing
organics are still present in the biologically-treated wastewater.
TABLE 7. COAGULATION OF BIOLOGICALLY-TREATED* HOLSTON WASTEWATER
ALUM
0-500 mg/1 at pH 6.5 - no floe formed, no precipitation
FERRIC CHLORIDE
0-500 mg/1 at pH 6.0 and pH 8.0 - no floe formed, no precipitation
LIME
Dose, mg/1
0
720
2640
3360
5280
PH
7.0
8.5
9.3
9.6
11.6
TOG, mg/1
640
475
460
455
450
H2S04
0 6.9 600
6.5 3.0 570
10 2.5 480
25 2.0 425
60 1.5 420
BETZ 1190 CATIONIC POLYMER
Dose, mg/1 TOG, mg/1
0 640
200 460
400 410
1000 500
5000 1260
'''Filtered activated sludge effluent, 10-day HRT, 20-day SRT
194
-------
The addition of lime (CaO) raised the pH of the water and, after
allowing for settling, resulted in the removal of some of the TOG. Approxi-
mately 25% of the TOG was removed by the addition of 720 mg/1 of lime which
raised the pH to 8.5. Little improvement was achieved with higher doses of
lime.
Sulfuric acid caused precipitation of some of the residual organics by
decreasing the pH of the wastewater. High molecular weight humic substances
tend to precipitate under such acidic conditions. Approximately 30% of the
TOG was removed when the pH was reduced to 2.0. Little precipitation of
TOG was obtained until the pH of the water was decreased to below pH 3.
The addition of a cationic polyelectrolyte, BETZ 1190, a high charge
density, relatively moderate molecular weight polymer, brought about some
coagulation of TOG, but again at rather substantial doses. Edzwald (7) has
shown that such cationic polymers are effective coagulants of high molecular
weight humic substances. The optimal dosage range appeared to be between
200 and 1000 mg/1, with 35% removal of TOG occurring at a dose of 400 mg/1
of the polymer. Apparently, little improvement in the quality of the waste-
water can be obtained through coagulation or acid or base treatment, even
at very high chemical doses.
Table 8 presents the results of an experiment in which the biologically-
treated wastewater was treated further in an ozone contact column. A mix-
ture of ozone and oxygen was bubbled through a sample of wastewater, and
aliquots were removed at various times and analyzed. The pH decreased
substantially during the course of ozonation, presumably due to the conver-
sion of many of the organic impurities to organic acids and C02. Thiocyanate
was oxidized almost completely by the ozone. Total organic carbon decreased
as a result of ozonation, while the COD was decreased to an even greater
degree. The relative decreases in TOG and COD suggest that many of the
organic compounds were converted to organic acids and aldehydes in which
the organic carbon is in a higher oxidation state than in the parent com-
pound, while only a portion of the organic compounds were oxidized com-
pletely to C02. The ozone consumption, which was calculated by measuring
the difference between the applied ozone in the feed gas and the ozone
concentration in the off-gas, is relatively small compared to the change
in COD and SCN~ concentrations, suggesting that some of the organics were
removed by the application of ozone. The initial removal of NH3 was
probably through air-stripping; further ammonia removal was inhibited as
a result of the acidic conditions (low pH) which were generated.
Table 9 shows the results of treating filtered biologically-treated
Holston wastewater with activated carbon. Pulverized Nuchar WV-G (Westvaco
Chemical Co) was used as the adsorbent. The studies were carried out as
batch equilibrium experiments in which various doses of carbon were added
to the wastewater, and the suspension was mixed for 4 hours to reach
equilibrium. Upon equilibration, the activated carbon was removed by
centrifugation and filtration, and the residual TOG was measured. Table 9
shows that the extent of TOC removal increased with increasing doses of
195
-------
TABLE 8. OZONATION OF BIOLOGICALLY-TREATED* HOLSTON WASTEWATER
Ozonation
Time
min.
0
10
30
60
Applied
Ozone Dose
mg/1
0
455
1365
2730
Ozone
Consumption
mg/1
0
450
910
1140
PH
6.77
3.36
2.65
2.59
TOC
mg/1
645
566
520
491
COD
mg/1
2777
1801
1431
1299
SCN-
mg/1
428
106
18
11
NH3
mg/1
2608
1904
1890
1820
*Activated sludge, 10-day HRT, 20-day SRT
TABLE 9. ADSORPTION OF BIOLOGICALLY-TREATED* HOLSTON WASTEWATER
Activated
Carbon**
Dose
mg/1
0
800
1000
1800
2000
3500
TOC COD BOD
mg/1 mg/1 mg/1
662 1480 48
490 1030 10
440
380
354 684 8
283
Daphnia Toxicity
LC50, %
24-hr 48-hr 96-hr
5.6 3.5 3.2
_
_
-
5.6 4.5 2.7
_
*Filtered activated sludge effluent, 10-day HRT, 20-day SRT
**Powdered Westvaco Nuchar WV-G activated carbon
196
-------
activated carbon and that approximately 50% of the TOG was removed with an
activated carbon dose of 2000 mg/1. However, Table 9 also shows that
despite the TOG and COD removals achieved by activated carbon adsorption,
such treatment had little impact on the toxicity of the wastewater to
Daphnia. The LC50s of the carbon-treated samples are essentially the same
as those of the biologically-treated effluent with no carbon treatment.
This may be a result of the high ammonia concentration of the samples, i.e.
the toxicity of the treated wastewater may be due to the approximately 2000
mg/1 of ammonia-nitrogen which is still in the wastewater even after the
activated sludge and activated carbon treatment.
In order to test this hypothesis, samples of the biologically-treated
Holston wastewater were treated with NaOH to raise their pH to approximately
11, air-stripped to release NH3, neutralized to pH 7 with HC1, and subse-
quently treated with activated carbon as described above. Table 10 shows
that while biological treatment of the diluted Holston wastewater reduced its
toxicity to Daphnia and CHO cells by factors of approximately 3 and 13,
respectively, reducing the ammonia concentration from 2000 to 110 mg/1 (a
95% reduction) resulted in an additional 3- to 6-fold reduction in toxicity.
The reason for the apparent increase in TOG which accompanied the
ammonia-stripping step is not known; it may have been due to (a) absorption
of organics from the laboratory air that was used to strip the ammonia,
although an activated carbon plug was used in the air line to trap any
organic contaminants in the air, or (b) to the hydrolysis of some of the
high molecular weight residual organics at the elevated pH which
makes the organic carbon more amenable to detection by the analytical proce-
dure used to measure TOG. The latter involves a high temperature (950°C)
combustion of the organic carbon by oxygen, and measurement of the C02
released. Some of the high molecular weight organic carbon in the sample
prior to ammonia-stripping may not have been oxidized completely to C02 and
therefore may have escaped detection.
Table 10 shows that subsequent treatment of the ammonia-stripped
biologically-treated Holston wastewater with 500 and 3600 mg/1 of activated
carbon reduced the TOG by 23% and 53%, respectively, but had no effect on
the toxicity of the wastewater to Daphnia. However, the toxicity of the
treated wastewater to the CHO cells was reduced to such a degree by
activated carbon that more than 50% of the cells survived at all of the
wastewater concentrations tested. While these activated carbon doses are
relatively extreme, they do illustrate the impact of additional TOG removal
on the toxicity of the wastewater.
In view of the reduction in toxicity resulting from ammonia stripping
and the improvement in the gross chemical quality of the wastewater
following ozonatieu (see Table 8), filtered biologically-treated wastewater
was ammonia-stripped in the same manner as discussed above, and then
subjected to ozonation. In this case, the ammonia-stripped wastewater was
buffered with respect to pH in order to promote broad-based non-selective
oxidation of the residual organics (8). Table 11 shows analytical results
197
-------
parallel to those shown in Table 8: substantial reduction of COD, some
removal of TOG, essentially complete elimination of thiocyanate, and no
oxidation of the residual ammonia. It appears, however, that cyanide is
produced from the oxidation of thiocyanate and, while some of the cyanide
is oxidized further by ozone, a significant concentration of cyanide remains
in solution even after 60 minutes of ozonation. Correspondingly, the
ozonated samples are more toxic to Daphnia and to the microbial seed used
in the BOD measurements. In the former case, a quantitative determination
of the 24-hr. LC-50 could not be made but it was observed that the 24-hr.
LC-50 for the ozonated samples was less than 5% compared to a 24-hr. LC-50
of more than 15% for the ammonia-stripped, biologically-treated wastewater
prior to ozonation. In the latter case, the BOD could not be measured
using more than a 6% dilution of the wastewater; dilutions greater than 6%
were toxic to the microbial seed. While some thiocyanate ozonation
studies have already been conducted (9), additional studies are required to
determine the relative oxidation kinetics of SCN~ and CN~ and to ascertain
whether the observed increase in toxicity following ozonation is due to the
generation of cyanide or to other toxic products of the ozonation reaction .
TABLE 10. ADSORPTION OF AMMONIA-STRIPPED BIOLOGICALLY-TREATED
HOLSTON WASTEWATER
S amp le
TOG
mg/1
NH3
mg/1
as N
Daphnia Toxicity
LC50, %
24-hr 48-hr 72-hr
CHO Cyto-
toxicity
96-hr LC50, %
Holston Feed
(25% strength)
1800 1970
Biologically-treated
Effluent 600
NH3~stripped
Effluent
Activated Carbon-
treated NH_-stripped
Effluent
705
1950
110
1.8 0.76 0.53
4.2 1.9 1.5
17.3 11.8 8.3
*Powdered Westvaco WV-G activated carbon
**Greater than 50% survival at concentrations up to 45%
***Greater than 50% survival at concentrations up to 75%
0.49 0.52
1.45 7.01
8.3 19.4
500
3600
mg/1
mg/1
AC*
AC*
540
330
110
110
18.8
23
11.
11.
3
8
10
8
.4
7.7
7.5
Indet**
Indet***
198
-------
TABLE 11. RESULTS OF OZONATION STUDIES ON AMMONIA-STRIPPED,
BIOLOGICALLY-TREATED HOLSTON WASTEWATER
Parameter
Time of Ozonation, min.
0 20 60
Ozone dose, mg/1
Ozone consumption, mg/1
pH
TOC, mg/1
COD, mg/1
BOD, mg/1
NH,, mg/1 as N
NO.T , mg/1 as N
SCN-, mg/1
CN-, mg/1
0
0
7.10
803
2503
115
146
8.0
607
3.3
900
845
6.68
744
1798
45*
147
4.6
87
152
2700
1505
6.60
676
1499
65*
153
10.0
5
128
Daphnia Toxicity
24-hr LC50, %
< 5
< 5
* Toxic at 6% concentration
SOLVENT-EXTRACTION OF HOLSTON WASTEWATER
In order to evaluate the impact of solvent-extraction of phenols on
the biological treatability of the Holston wastewater, a large volume of the
virgin Holston wastewater (i.e. a fresh sample from a newly-opened barrel of
the wastewater) was extracted with n-butyl acetate. Three extractions,
with a solvent-wastewater ratio of 1 to 10, were employed, and the residual
butyl acetate in the aqueous phase was eliminated by air-stripping. The pH
of the wastewater was raised to approximately 11 with NaOH and the sample
was air-stripped to release NH3- After re-adjustment of the pH to 7 with
HC1, the wastewater was supplemented with phosphate and fed without any
dilution to an activated sludge reactor operated at a 20-day sludge age and
a 10-day hydraulic residence time.
Table 12 gives the results available to date. The solvent-extraction
step reduced the concentration of phenols to 8.0 mg/1 and resulted in TOC
and COD removals of 68% and 67%, respectively. These removals were
accompanied by a 6 to 7-fold reduction in Daphnia and CHO toxicity. Ammonia-
stripping of the solvent-extracted wastewater to a level of 84 mg/1 of
199
-------
ammonia resulted in an additional 6-fold reduction in toxicity to Daphnia.
(Again, it should be noted that both TOG and COD appear to have increased as
a result of pH adjustment and ammonia-stripping. Hydrolysis of high
molecular weight organics or absorption of organics from the laboratory air
are, again, possible explanations for this apparent increase.) The results
of the biological treatment studies are not available at the time of this
writing.
TABLE 12. RESULTS OF SOLVENT-EXTRACTION STUDIES
Parameter
Virgin Holston
Wastewater*
Solvent-Extracted
Holston Wastewater**
NH3~stripped,***
Solvent-Extracted
Wastewater
TOC, mg/1
COD, mg/1
Phenols, mg/1
NH3, mg/1 as N
SON", mg/1
Daphnia Toxicity
24-hr LC50, %
48-hr LC50, %
96-hr LC50, %
CHO Cytotoxicity
LC50, %
7490
24,500
2200
7290
445
0.076
0.050
0.038
0.055
2390
8200
8.0
7200
0.44
0.24
0.23
0.4
2860
10,100
-
84
2.6
1.55
1.38
*Batch 2, full-strength
**n-Butyl acetate; -1/10 solvent/water ratio, 3X; air-stripped to eliminate
butyl acetate
***pH adjustment with NaOH, air-stripped to expel NH^, pH re-adjustment with
HC1
CONCLUSIONS
A comparison has been made between a real coal gasification wastewater
from a fixed-bed atmospheric gasifier and a simulated coal conversion waste-
water. The simulated wastewater was similar with respect to the concentra-
tions of TOC, COD, and phenols, but the real wastewater had an appreciably
higher ammonia content. In addition, the real wastewater was approximately
4 to 5 times more toxic than the simulated wastewater, based on Daphnia,
fish, and CHO bioassays.
200
-------
The real wastewater was biologically-treatable when diluted to quarter-
strength. Treatment in an activated sludge reactor with a 20-day sludge age
and a 10-day hydraulic residence time resulted in residual concentrations of
phenols generally below 1 mg/1, TOG removals of approximately 65%, and
COD removals of approximately 63%. The effluent TOG and COD concentrations,
however, were approximately 2 to 3 times higher than those in the effluent
from an activated sludge reactor treating the simulated coal conversion
wastewater under parallel operating conditions. Additionally, while the
toxicity of the real coal conversion wastewater to Daphnia, fish, and
mammalian cells was reduced appreciably by biological treatment and the
mutagenicity of the wastewater was reduced to undetectable levels, the
effluent was significantly more toxic than the biologically-treated,
simulated wastewater effluent.
A significant portion of the residual TOG (approximately 30%) in the
filtered activated sludge effluent following treatment of the real waste-
water consists of organic compounds with a molecular weight greater than
30,000. If the effluent is to be re-used and concentrated in a cooling
tower, the presence of this relatively large amount of high molecular
weight material may have an adverse impact on the operation of the cooling
towers.
Post-biological treatment involving ammonia-stripping and activated
carbon adsorption significantly alleviated the mammalian cytotoxicity of
the real wastewater; such treatment had no effect on the toxicity of the
wastewater to Daphnia. Ozonation improved the gross chemical quality of
the wastewater, but had an adverse impact on Daphnia toxicity.
ACKNOWLEDGEMENTS
The authors would like to thank Randy Goodman, Larry Day, Steve Shoaf,
Chen-yu Yen, Dave Reckhow, Joe Janeczek, Joel Storrow, Wen-long Shu, Rich
Gullick, Rick Chan, Matt Matthews, Penny Hill, and Donna Volney for their
experimental and analytical contributions to this research effort. The
financial support of the US Environmental Protection Agency, and the
specific assistance of our project officer, Dr. N. Dean Smith of the
Indus-trial Environmental Research Laboratory, Research Triangle Park, is
also gratefully acknowledged.
201
-------
REFERENCES
1. Singer, P. C. et al. , "Evaluation of Coal Conversion Wastewater
Treatability," in Symposium Proceedings: Environmental Aspects of
Fuel Conversion Technology, IV, Report No. EPA-600/7-79-217,
U.S. Environmental Protection Agency, Washington, D.C. (1979).
2. Singer, P. C. et al., Treatability and Assessment of Coal Conversion
Wastewaters: Phase 1, Report No. EPA-600/7-79-248, U.S. Environmental
Protection Agency, Washington, D.C. (1979).
3. Singer, P.C. et al., "Effect of Sludge Age on the Biological
Treatability of a Synthetic Coal Conversion Wastewater," in
Symposium Proceedings: Environmental Aspects of Fuel Conversion
Technology, V, U.S. Environmental Protection Agency (in press).
4. Collins, R. V., K. W. Lee, and D. S. Lewis, "Comparison of Coal
Conversion Wastewaters," in Symposium Proceedings: Environmental
Aspects of Fuel Conversion Technology, V, U.S. Environmental
Protection Agency (in press).
5. Luthy, R. G. , Manual of Methods: Preservation and Analysis of Coal
Gasification Wastewaters, DOE FE-2496-8 U.S. Dept_. of Energy,
(1977).
6. Babcock, D. B. and P. C. Singer, "Chlorination and Coagulation of
Humic and Fulvic Acids," Journ. Amer. Water Works Assn., Vol. 71,
no. 3, 149-152 (1979).
7. Edzwald, J. K., J. D. Haaf, and J. W. Boak, "Polymer Coagulation of
Humic Acid Waters," Journ. Amer. Soc. Civil Eng., Env. Eng. Div.,
103:EE6:989 (1977).
8. Hoigne, J. and H. Bader, "Ozone Initiated Oxidations of Solutes in
Wastewater: A Reaction Kinetic Approach," Progress in Water
Technology, Vol. 10, Nos. 5/6, p. 657-71, Pergamon Press (1978).
9. Kemker, G. H., S. J. Roadcap, and R. Sitchin, The Ozonolysis of
Aqueous Solutions of Phenol and Thiocyanate, ORNL/MIT-318, Dept.
of Energy (1980).
202
-------
TREATMENT OF FOSSIL FUEL DERIVED WASTEWATERS WITH
POWDERED ACTIVATED CARBON/ACTIVATED SLUDGE TECHNOLOGY
By: R.B. Ely, C.L. Berndt
Zimpro Inc.
Rothschild, WI 54474
ABSTRACT
The treatment of high strength fuel conversion wastewaters by
conventional biological treatment processes may be operationally
troublesome and only marginally effective from the standpoint of
treatment system stability and performance. The addition of
powdered activated carbon to the activated sludge process not only
greatly improves product water quality but also provides cost
savings compared to more conventional waste treatment and carbon
regeneration processes.
This paper describes the powdered carbon/activated sludge
wastewater treatment process, discusses the advantages of powdered
carbon addition including performance obtained on fossil fuel
derived wastewaters, and presents cost comparison data for
wastewater treatment and spent carbon regeneration.
INTRODUCTION
Development of the synthetic fuels production industry is
contingent in part on successful treatment of the production
wastewaters since environmental regulations for treated wastewater
discharges are likely to be very stringent and effluent reuse will
be necessary in many facilities. Efficient, reliable waste
treatment is of critical concern due to the constituents present
in most synfuels wastes and the variability anticipated. These
concerns have spurred investigation of powdered activated carbon
addition to the activated sludge wastewater treatment process for
improved treatment performance* and improved organics removals**
among others.
* Luthy, R.G., Stamoudis, V.C., and Campbell, J.R., "Removal
of Organic Contaminants from Coal Conversion Condensates."
Presented at the 54th Annual WPCF Conference, Detroit,
Michigan (October, 1981).
** Wei, I.W., and Chen, J.C.Y., "Fate of Organics in the Treatment
of Oil Shale Retort Water." Presented at the 54th Annual WPCF
Conference, Detroit, Michigan (October, 1981).
203
-------
The addition of powdered carbon to the activated sludge
process, termed PACT*, provides enhanced treatment performance and
reliability over that found in pure biological treatment systems.
The addition of PAC provides improved COD removals and permits
nitrification of the synfuels wastewater.
When Wet Air Carbon Regeneration is applied to the PACT
process, the process is called the Wastewater Reclamation System
(WRS) and is hereafter referred to as such.
Application of Wet Oxidation to synthetic fuels wastes, for
spent carbon regeneration of solids wasted from the Wastewater
Reclamation System and for oxidation of concentrated production
wastes, enables economical disposal of concentrated, difficult to
treat wastes and provides cost-effective spent carbon
regeneration.
TREATMENT CONCEPTS
The addition of powdered activated carbon (PAC) to the
activated sludge process combines simultaneously the advantages of
physical adsorption and biological stabilization in the same
"biophysical" treatment system. The presence of the active
adsorbent (PAC) provides removal of non-biodegradable, adsorbable
organics in the waste stream—organics which would otherwise
escape untreated from a pure biological system. Furthermore, the
high concentration of activated carbon in the treatment system
ensures maintaining reasonable treatment even if biological upset
should occur.
The benefits of adsorption and biodegradation are exploited
by combining both methods of treatment in a single operation. The
combined effects are illustrated in Figure 1 which shows carbon
adsorption isotherms of two wastes, A and B. Waste A (solid line)
is treatable by carbon adsorption as indicated by the shallow
slope of the isotherm. Waste B is not readily treatable by
activated carbon as indicated by the steep slope of the isotherm.
Waste B apparently contains organic constituents which are not
readily adsorbable.
* PACT is a registered servicemark of DuPont.
204
-------
10.0
1.0
O.I
0.01
n i 111 +-
CARBON ADSORPTION TREATMENT
BIOPHYSICAL TREATMENT
I 10 IOO 1000
Equilibrium COD Concentration (mg/L)
Figure I. PHYSICAL AND BIOPHYSICAL ISOTHERMS
The dashed lines show isotherms for the same two wastes when
both biological treatment and adsorption are simultaneously
imposed. Waste B is now treatable as indicated by the similarity
of the two isotherms. In addition, considerably higher organic
loading rates are obtained resulting in a smaller wastewater
treatment system. The activated sludge has aided the activated
carbon in removing organic constituents which are not readily
adsorbable. For synfuels wastes, most of the particularly
obnoxious fossil fuel related components such as multi-phenolics
are adsorbable and are effectively removed in biophysical
treatment. Performance of the Wastewater Reclamation System on
synfuels wastewaters indicate that greater than 95 percent removal
of COD can be obtained with only very low COD residuals remaining
following treatment.
In addition to enhanced performance and increased organic
loading rates, PAC addition to activated sludge adsorbs toxic or
inhibitory components enabling the micro-organisms to function
efficiently. This is important since synfuels wastewaters
frequently contain toxic components in sufficient concentration to
inhibit metabolic rates and nitrification. Further, carbon acts
as a toxic sink to dampen organic fluctuations resulting from
production process variations or upset.
205
-------
The presence of PAC also provides a catalytic or perhaps best
termed an alleo-catalytic effect on biological treatment. The
active adsorbent concentrates on the PAC surfaces the
extra-cellular enzymes needed for organics assimilation in
addition to adsorption of waste organics and molecular oxygen.
This concentration effect serves to catalyze the biological
mechanisms.
Perhaps more important is that contaminants that are slow to
degrade will be held by the activated carbon in the treatment
system for the solids residence time, not the much shorter
hydraulic residence time which would be the case where carbon is
not present. Thus, additional organics are removed biologically
which would otherwise have to be treated by granular carbon, ion
exchange or ozonation.
The foregoing arguments explain the superior performance seen
in powdered carbon/activated sludge systems on a micro-basis.
However, there are some important design considerations that
contribute to the success of the process. Of primary importance
is the settleability of the sludge. The carbon nucleus of the
floe particles serves as a weighting agent. The sludge can be
readily settled and compacted and therefore carried at very high
levels in the aeration basins. The Wastewater Reclamation System
will typically operate at 15-25,000 mg/1 mixed liquor suspended
solids whereas a conventional activated sludge system is typically
2,000-4,000 mg/1. Though a major fraction of the WRS mixed liquor
is PAC, volatile biological solids levels easily exceed
conventional activated sludge systems and may approach 7.000 mg/1
in normal operation.
THE WASTEWATER RECLAMATION SYSTEM
The Wastewater Reclamation System has been or will be used in
numerous applications and will treat a wide variety of wastewaters
including night soil, combined domestic and textile wastes,
nitrification of domestic and industrial wastes, pharmaceutical
wastes and organic chemicals wastes. A list of WRS applications
and the waste treated are shown in Table 1.
206
-------
TABLE 1. WRS INSTALLATIONS
Location
Rothschild, WI
Kimitsu, Japan
Oga, Japan
Vernon , CT
Senroku, Japan
Oizumi, Japan
Medina, OH
Burlington, NC
(East Plant)
Mt. Holly, NJ
Kalamazoo, MI
Burlington, NC
(South Plant)
East St. Louis, IL
Ibaragi, Japan
El Paso, TX
Bedford Heights, OH
North Olmsted, OH
Size
3785 m3/d
500 KL/d
1200 Kl/d
24600 nH/d
1400 KL/d
800 KL/d
37850 m3/d
47300 rrrVd
18425 m3/d
204400 nH/d
35960m3/d
102200 m3/d
1520 KL/d
37850 m3/d
11350 nH/d
26500 m3/d
Wastewater
Domestic
Night Soil
Night Soil
Domestic/Textile
Night Soil
Night Soil
Domestic
Domestic/Textile
Domestic/Textile
Domestic/
Pharmaceutical
Domestic/Textile
Domestic/Organic
Chemicals
Night Soil
Domestic
Domestic/
Industrial
Domestic/
Industrial
Operation
1972-73
Demonstration
1975
1977
1979
1980
1980
1981
1.981
1981
Under
Construction
Under
Construction
Under
Construction
Under Design
Under Design
Under Design
Under Design
The WRS flow scheme is presented in Figure 2. Typical major
process components include aeration, clarification and optional
effluent filtration. Auxiliary process components include dry
carbon storage and liquid polymer addition.
When using WRS, the wastewater is aerated in the presence of
a high concentration of powdered activated carbon (PAC), from
4,000 to 12,000 mg/L, depending on the influent wastewater
characteristics and effluent quality required. The powdered
carbon not only acts as an adsorbent, but also as a weighting
agent, enhancing MLSS settling and enabling higher concentrations
of volatile biological solids to be maintained under aeration.
Thickened clarifier underflow solids, at concentrations typically
ranging from 3.0 to 5.0 percent are recycled to the inlet of the
aeration basin. Due to the high concentrations of PAC and
biological solids maintained in the WRS, a high degree of reliable
treatment is obtained.
The excess secondary sludge from the WRS is wasted from the
aeration tank or clarifier to a gravity thickener. The thickened
207
-------
Figure 2. WASTEWATER RECLAMATION SYSTEM
HEAT
REALTOR EXCHANGER
BOILER "
CARBON REGENERATION SYSTEM
GENERAL PROCESS DIAGRAM
WASTEWATER RECLAMATION SYSTEM
T"^
1 ZIIVIF>RO
underflow solids, at a concentration of 6.0 percent suspended
solids or greater, is pumped to the regeneration unit heat
exchangers by the high pressure pump at a pressure of
approximately 50 kg/cm (800 psig).
Compressed air is added to the carbon slurry flow prior to
the heat exchangers. The combined slurry and air mixture passes
through the heat exchangers where its temperature is raised prior
to entering the reactor. In the reactor, the volatile biological
solids and sorbed organics contained in the carbon slurry are 'wet
oxidized.'
Since a net heat gain (temperature rise) occurs during the
wet oxidation reactions, autothermal (thermally self-sustaining)
operation is obtained. The hot regenerated slurry is then passed
through the heat exchangers to recover the produced heat. The
cooled regenerated slurry flows to the pressure reducing station
and returned to the wastewater flow via a distribution diffusor in
the scrubbing channel. Though a nearly complete oxidation
(85-95%) of chemical oxygen demand occurs during regeneration, a
small amount of low molecular weight residual organics remain
which are returned directly to the treatment system for biological
stabilization. Since these organics are readily biodegradable and
comprised of weak acids, separate sidestream treatment of the
regeneration recycle stream is not required.
208
-------
Control of the mixed liquor suspended ash concentration is
provided by regeneration reactor blowdown. Inerts accumulate at
the reactor bottom and are vented from the reactor during
steady-state operation and disposed. Since these materials are
inert and 'wet oxidized1 during regeneration, disposal as a
non-hazardous material is generally acceptable.
A steam generator is included in the regeneration system to
provide start-up steam requirements.
Wet air regeneration losses of volatile PAC are less than 5
percent of throughput, substantially less than PAC oxidation
losses in conventional thermal regeneration processes.
Regeneration losses include both those resulting from oxidation
losses and reactor inerts blowdown to disposal. Powdered carbon
losses of 1 to 5 percent and autothermal regeneration system
operation have been confirmed in the full scale operations at
Kimitsu, Japan* and Vernon, CT.**
ADVANTAGES TO SYNFUELS WASTE TREATMENT
Major advantages of the Wastewater Reclamatiom System to the
treatment of fossil fuel derived wastewaters are the excellent
product water quality obtained and the reliable treatment process
operation and stability that is ensured with PAC addition.
Treatment process stability is of major significance to the
synfuels facility since biological treatment difficulties
resulting in upset conditions will likely curtail fuel production,
will result in post-biological treatment difficulties in reuse
applications, and will result in failure to meet discharge
requirements where direct effluent discharge is practiced.
Since the treatment system effluent quality is a major
consideration in most synfuels applications, for both effluent
reuse in the facility and for direct discharge, optimum
performance is extremely important. Residuals (COD, ammonia) are
of concern in terms of fouling reverse osmosis membranes,
evaporator tubes and cracking and carbonizing in boilers and
superheaters. Ammonia generally presents corrosion problems in
cooling water systems and boilers. Organic priority pollutants
present in coal derived wastewaters (see Table 2) represent a
potential health hazard in the plant and must be effectively
removed before direct discharge.
* Meidl, J.A. ; Berndt, C.L. and Nomoto, K., "Experience with
Full Scale Wet Oxidation of Spent Carbon from the 'PACT'
Process." Presented at the 51st Annual Conference of the
WPCF, Anaheim, CA, (October, 1978).
** Pitkat, C.A. and Berndt, C.L., "Textile Waste Treatment at a
Municipal PACT Facility." Presented at the 35th Purdue
Conference, Purdue University, West Lafayette, IN, (May,
1980).
209
-------
Enhanced organics (BOD,-, COD) removals and biological
nitrification are obtained when powdered carbon is added to
activated sludge. In numerous treatability demonstrations,
improved organic removals of chemical wastes were obtained*,
efficient nitrification was obtained in the WRS whereas
biological treatment was unsuccessful due to the presence of
Pharmaceuticals** and nitrification of toxic wastes was possible
in a two stage WRS mode***.
TABLE 2. LEVEL OF ORGANIC PRIORITY POLLUTANTS DETECTED IN
EPA SCREENING PROGRAM
No.
1
39
81
80
3
84
55
64
11
74
76
49
75
59
72
78
79
Priority Pollutant
No
. of
Name Samples
* Acenapthene
* Fluoranthene
* Phenanthrene
* Fluorene
Acrylonitrile
* Pyrene
* Naphthalene
Pentachlorophenol
1,1, 1-trichloroethane
* 3, 4-benzofluoranthene
* Chrysene
Trichlorofluoromethane
* Benzo(k) fluoranthene
2, 4-dinitrophenol
* Benzo(a)anthracene
* Anthracene
* Benzo(ghi)perylene
18
21
36
25
9
26
59
35
93
4
27
27
9
16
23
35
8
Mean
Minimum
Value
(mg/L)
216.
147.
130.
80.
65.
61.
43.
37.
26.
24.
24.
22.
22.
17.
15.
15.
14.
8
7
2
2
7
4
4
5
0
8
3
4
1
8
8
1
0
(mg/L)
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
013
011
010
011
043
010
010
012
010
010
010
011
011
011
010
010
013
Maximum
Value
(mg/L)
3,000.
1,400.
3,200.
1,400.
330.
1,100.
1,200.
680.
1,300.
99.
440.
290.
99.
230.
180.
510.
84.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
* Coal Based.
#*
Sago, W.L. and Foresman, M.R., "Joint Municipal/Industrial
Wastewater Treatment - Metro East St. Louis, Illinois."
Presented at the 53rd Annual Conference of the WPCF, Las
Vegas, Nevada (September, 1980).
Sampayo, F.F. and Hollopeter, D.C., "The Influence of
Industrial Waste on Nitrification." Presented at the 33rd
Purdue Conference, Purdue University, West Lafayette, IN
(May, 1978).
*** Frohlich, G., Ely, R.B. and Vollstedt, T.J., "Performance of
a Biophysical Treatment Process on a High Strength
Industrial Waste." Presented at the 31st Purdue Conference,
Purdue University, West Lafayette, IN (May, 1976).
210
-------
140
100
80-
60
40
20
INFLUENT
ACTIVATED SLUDGE EFFLUENT
WRS EFFLUENT
I 2 5 10 20 30 40 50 60 70 80 90 95 98 99
FREQUENCY OF OCCURRENCE
Figure 3. FREQUENCY PLOT OF AMMONIA REMOVAL
Recent performance on fossil fuel derived wastewaters
wherein nitrification was required show that complete
nitrification is readily obtained with PAC addition but is not
obtained in a pure biological activated sludge system even at a
long solids residence time and hydraulic detention time. A
comparison of nitrification performance of activated sludge and
WRS is shown in Figure 3. Both processes were operated in a
single stage mode at an SRT of 35 days, however, the activated
sludge pilot plant hydraulic detention time exceeded 50 hours
more than two times the WRS.
A performance comparison of priority pollutant removals
from conventional activated sludge and the powdered activated
carbon/activated sludge process is shown in Table 3.* Improved
priority pollutant removals were obtained with PAC addition.
Similar results, an approximate one-third greater priority
pollutant removal with PAC enhanced sludge, has been
demonstrated for shale oil retort wastewaters.**
* Hutton, D.G., "Removal of Priority Pollutants with a
Combined Powdered Activated Carbon - Activated Sludge Process."
Presented at the 179th National AIChE Meeting, Houston, TX
(March, 1980).
** Wei, op. cit.
211
-------
TABLE 3. EFFLUENT PRIORITY POLLUTANT COMPARISON
% Removal
Compound
Feed
Concentration
ppb
Activated
Sludge
Powdered Carbon/
Activated Sludge
Benzene
Chlorobenzene 3
Chloroethane
Chloroform
Methyl Chloride
Tetrachloroethylene
1 , 2-Dichlorobenzene
2,4-Dinitrotoluene 1
2,6-Dinitrotoluene 1
Nitrobenzene
1,2, 4-Trichlorobenzene
2, 4-Dichlorophenol
2, 4-Dinitrophenol
4-Nitrophenol 1
81
,660
667
72
138
33
18
,000
,100
330
210
19
140
,100
98.5
99.1
99.8
96.7
98.5
99.5
90.6
31.0
14.0
94.5
99.9
0
39.0
25.0
99.6
99.8
99.9
96.9
99.7
99.5
99.0
90.0
95.0
99.9
99.9
93.0
99.0
97.0
Resiliency to potential toxic upsets due to production
process malfunctions is illustrated in Figure 4. Consistent WRS
performance is maintained in the two stage system though the
total phenol concentration reached 2000 mg/L in the coal
gasification liquor feed for a 1.0 hour duration. This was
preceeded by a 0.5 hour period at 1000 mg/L total phenol, to
simulate actual shock phenol levels occurring in a process
malfunction. The results of Figure 4 show that the WRS effluent
NHn-N and total phenol levels remained low though no WRS
operational adjustments were made to compensate for the shock
loading. Consistent results continued beyond that shown in
Figure 4. The effluent NH,-N levels returned to less than 0.5
mg/1 following the stress tests. Improved response would be
expected with SRT changes during stress conditions.
212
-------
120
NH3-N
INFLUENT
EFFLUENT
SPIKE TO
20OOmg/L
/ \
PHENOL
. 1
\—
I 23456789 IO
12 13 14 15 16 17 18 19 20
DAY
Figure 4. RESPONSE TO PHENOL SHOCK LOADING
PERFORMANCE
Initial WRS treatability investigations were conducted on
coke oven gas flushing liquors — quite similar to synfuels
wastewaters. A performance comparison from studies of activated
sludge and WRS treating these high strength liquors is shown in
Table 4. The results indicate good performance for both
processes, however, ammonia conversion to nitrate nitrogen was
not obtained in the activated sludge treatment. Nitrification
was obtained in the single stage WRS despite the high phenol
concentration (468 mg/L total phenol) and the high waste COD
level.
TABLE 4. PERFORMANCE COMPARISON:
COKE OVEN GAS FLUSHING LIQUORS
Activated Sludge
Influent Effluent
WRS
Influent
Effluent
BOD5, mg/1
COD, mg/1
NH3-N, mg/1
SCN, mg/1
Phenol, mg/1
Cyanide, mg/1
650
1329
600
130
150+
—
10
436
731*
3.5
<1
—
1050
2359
13
279
468
7
4
289
<1
<2
<1
1.2
*SCN is biologically converted to NH-,_N, there is no
nitrification.
213
-------
Subsequent treatment demonstrations confirmed nitrification
of coke oven flushing liquors at solids residence times as low
as 7 days at approximately 25°C-* Wastewater characteristics
were similar to the WRS influent data of Table 4.
More recently, Zimpro Inc. has conducted treatability
studies on fossil fuel derived wastewaters. Laboratory scale
treatment of coal gasification wastes were performed.
Additional process wastes were added to duplicate expected full
scale plant waste characteristics. Both single and two stage
activated carbon/activated sludge systems were operated for
organics removal. Spent carbon regeneration was provided.
Since nitrification was not required, pH control and alkalinity
supplement were not provided. The raw waste pH was slightly
less than 5 while the mixed liquor and effluent pH levels were
approximately 6.5.
Performance results, shown in Table 5, indicate good
organic removals for both single and two stage systems.
TABLE 5. WRS PERFORMANCE: COAL GASIFICATION WASTEWATER
Single Stage Two StagG
Influent Effluent Influent Effluent
BOD5, mg/1
COD, mg/1
TKN, mg/1
NH3-N, mg/1
Phenol, mg/1
Cyanide, mg/1
699
1580
148
104
6.9
10.0
<12
110
110
88
<0.9
0. 17
708
1560
152
103
6.7
10.8
<12
94
116
93
<1.0
0. 14
Similar laboratory scale treatment of a coal gasification
wastewater in a nitrification mode with intermittent NaOH
supplement for pH control provided the results shown in Table 6.
The pH adjustment maintained a minimum pH of 6.5. The
wastewater in this study contained a higher volatile acids
fraction than the previous gasifier waste resulting in a higher
BOD/COD ratio. In addition to nearly complete nitrification in
both single and two stage treatment modes, substantial
denitrification is demonstrated.
Bauer, G.L., Hardie, M.G. and Vollstedt, T.J., "Biophysical
Treatment of Coke Plant Wastewaters." Presented at the 35th
Purdue Conference, Purdue University. West Lafayette, IN (May
1980).
214
-------
TABLE 6. WRS PERFORMANCE: COAL GASIFICATION WASTEWATER
Single Stage Two Stage
Influent Effluent Influent Effluent
BODt-, mg/1
COD, mg/1
TKN, mg/1
NH3-N, mg/1
Phenol , mg/1
Cyanide, mg/1
1344
2270
99
70
2.6
7.5
<6
45
5.5
<1.1
<0.1
0.08
1344
2270
99
70
2.6
7.5
<4
53
5.7
<1.0
<0.1
0.11
An extensive design study on a larger scale pilot basis was
conducted on the gasifier wastewater of Table 6. Performance
results over the 6 month study period were excellent with a
neglible effluent NH^-N concentration from the two stage WRS.
TREATMENT COST COMPARISON
Cost comparisons (Table 7) of WRS and more conventional
biological treatment processes, sponsored by the EPA, show the
WRS is approximately cost equivalent to conventional activated
sludge and activated sludge designed for nitrification.*
Considerable cost savings is obtained employing WRS in-lieu of
activated sludge followed by granular carbon adsorption.
TABLE 7. TREATMENT COST COMPARISON
Cost, $/1000 Gallons
Process 5 mgd 10 mgd 25 mgd
WRS* 0.52 0.40 0.30
Activated Sludge
Conventional 0.49 0.38 0.29
Single Stage
Nitrification 0.51 0.41 0.31
Two Stage Nitrification 0.59 0.46 0.35
Granular Carbon System
@ 1500 Ib carbon/MG 0.73 0.58 0.46
* Designed to nitrify.
Gulp, G.L. and Shuckrow, A.J., "Appraisal of PAC Processes
for Municipal Wastewater Treatment." Environmental
Protection Technology Series, EPA-600/2-77-156, Contract No.
68-03-2211 (September, 1977).
215
-------
Investigation of secondary treatment options for Lurgi
process coal gasification liquors including the powdered
activated carbon/activated sludge process and other applicable
wastewater treatment processes showed the PAC/activated sludge
process employing Wet Air. Regeneration the most cost effective
treatment option (see Table 8).* The annual operating cost of
the PAC/activated sludge process is comparable to conventional
biological treatment with land application of waste sludges,
while considerable capital cost savings are obtained. The net
energy requirements of the PAC treatment system is also
equivalent to conventional activated sludge.
TABLE 8. TREATMENT COST COMPARISON FOR LURGI PROCESS WASTES*
Treatment
Process
Annual
Capital Operating
Cost Cost
Net Energy
Requirements,
KWH/yr
PAC/Activated Sludge
Wet Air
Regeneration 5,788,000
Multiple Hearth
Regeneration 6,761,000
Activated Sludge
Incineration of
1,764,000
2,460,000
18,066,920
23,030,900
Sludges
Land Application
of Sludges
9,862,000
6,769,000
2,347,000
1,799,000
26,115,400
18,401,500
* Based on 242 MM SCF PD SNG
The results of Table 8 show considerable spent carbon
regeneration cost savings with Wet Air Regeneration over that
obtained with multiple hearth regeneration.
Castaldl, F.J., "Application of Combined Powdered
Carbon/Activated Sludge Treatment to Lurgi Process Coal
Gasification Wastewaters." Application of Adsorption to
Wastewater Treatment, Vanderbilt University, Nashville, TN
(February. 1981).
216
-------
CONCLUSIONS
The addition of powdered activated carbon to the activated
sludge process, including spent carbon reactivation by Wet Air
Regeneration, provides improved treatment performance and
ensures stable reliable operation. The addition of PAC provides
further treatment benefits including resistance to shock loading
and wastewater toxicity and permits nitrification of synfuels
wastewaters.
Performance of the Wastewater Reclamation System on coke
oven gas flushing liquors and coal gasification process liquors
is excellent. Both organic treatment and nitrification of these
wastewaters were demonstrated.
Cost evaluations of the Wastewater Reclamation System on
coal gasification wastewaters show the WRS cost effective
compared to conventional biological treatment with land
application of residuals. Wet Air Regeneration was shown more
economical than multiple hearth regeneration for spent powdered
carbon regeneration.
217
-------
LAND TREATMENT OF COAL CONVERSION WASTEWATERS
*
by: R.C. Sims and M.R. Overcash
North Carolina State University
Raleigh, N.C. 2765O
ABSTRACT
This research project investigated the treatment potential of soil
systems for polynuclear aromatic compounds (PNAs) present in aqueous wastes
from coal conversion processes. A protocol for obtaining the soil assim-
ilative capacities for mutagenic and recalcitrant PNA compounds was developed
and, for a subset of compounds, data were obtained to describe: (1) rates of
transformation, including degradation, detoxication, and possible intoxica-
tion; (21 effect of PNA structure on transformation rate; (3) effect of
engineering management options, including nutrient addition, analog enrich-
ment, surfactant addition, and pH adjustment on transformation rates; and
(4) soil acclimation to PNAs.
A three-step protocol including: (1) incubation, (2) identification, and
(3) determination of mutagenic potential involves interfacing high perfor-
mance liquid chromatography (HPLC) for compound and metabolite identification
with the Ames Salmonella typhimurium/mammalian microsome mutagenicity assay
for determining genotoxic potential of PNA compounds and transformation
products in soil. Identification (HPLC) and mutation (Ames assay) were
quantified.
INTRODUCTION
This research has investigated the use of land treatment for the poly-
nuclear aromatic (PNA) class of compounds present in wastes from the coal
gasification industry. Land treatment has been demonstrated to be a cost-
effective environmentally safe technology for a multitude of industrial
wastes. PNA compounds have been identified as byproducts in the synfuel
industry, and are of critical environmental concern due to the following
specific characteristics: (1) chronic health effects (carcinogenicity), (2)
microbial recalcitrance, (3) high bioaccumulation potential, and (4) low
removal efficiencies in traditional wastewater treatment processes (Herbes
et al., 1976). Therefore, a preliminary feasibility assessment regarding
the application of land treatment technology for coal conversion wastes in
general and for hazardous components in particular was undertaken.
Address after January, 1982: Department Civil and Environmental Engineering,
Utah State University, Logan, Utah 84322
218
-------
In a comprehensive review of the literature Sims and Overcash (1981)
summarized the behavior and fate of PNA compounds in terrestrial systems,
including soils and vegetation. The potential for effective treatment and
safe ultimate disposal of PNA compounds is significant with regard to land
application of coal gasification wastes.
Land application is defined for the purpose of this study as the inti-
mate mixing or dispersion of wastes and the soil-plant system with the objec-
tive of microbial stabilization, adsorption, immobilization, selective dis-
persion, or crop recovery leading to an environmentally acceptable assimi-
lation of the waste. In this case coal conversion wastes are applied in thin
layers over land areas to provide intensive waste constituent interaction
with the soil, with substantial soil zones between waste and relevant surface
and ground waters, for the purpose of environmentally acceptable assimila-
tion.
Land application of synfuel wastes is also based on a constraint of non-
degradation of land. That is, the waste when considered on a constituent-
by—constituent basis shall be applied to the plant-soil system at such rates
or over such time spanai that no land is irreversible removed from some other
potential usage (agriculture, development, forestation, etc) (Overcash and
Pal, 1979).
There are four major stages in the design of a total waste management
system for coal conversion wastes. These stages, shown in Figure 1 are:
I. the determination of the land limiting constituent (LLC) or that
parameter or class of parameters requiring the largest land area
for assimilation;
II. the design evaluation of all required components for the land
application system and the cost analysis based on different
amounts of the LLC;
III. the selection and cost analysis of pretreatment or in-plant
alternatives for reducing the total level of the LLC;
IV. the economic balance between the cost of the total land receiver
and the cost of pretreatment processes such that the sum total
system costs is a minimum (Overcash and Pal, 1979).
The first stage of the design methodology is the most difficult. The
assimilative capacities for PNAs have not been established. An objective of
this research project has been to obtain the information necessary to com-
plete stage one for the coal gasification industry. This has required a de-
tailed literature review of coal gasification waste characterization and
plant-soil assimilative capacities for waste constituents, and laboratory
studies to determine the soil assimilative capacities for relevant PNAs.
219
-------
WASTE CHARACTERIZATION
RELATIVE TO LAND SYSTEM
DETERMINATION OF LLC REMOVAL
EFFICIENCY FOR ALTERNATIVE
PROCESSES
APPROXIMATE LAND
APPLICATION RATE
STAGE I
COST ANALYSIS FOR
PRETREATMENT PER UNIT
AMOUNT OF LLC
DETERMINATION OF AREA
REQUIREMENTS FOR
WASTE CONSTITUENT
LAND LIMITING CONSTITUENT
(LLC)
STAGE III
DESIGN OF SYSTEM
COMPONENTS
STAGE II
COST ANALYSIS FOR
LAND SYSTEM PER
UNIT AMOUNT OF LLC
COST COMPARISON OF
PRETREATMENT AND LAND
SYSTEMS - MINIMIZATION
OF TOTAL COSTS
STAGE IV
I 'J
Figure 1. Stages of Unified Methodology for Design of Pretreatment-Land Application Systems.
-------
APPROACH
Stage one of the four-stage methodology required the following infor-
mation: (1) waste characterization on a constituent-by-constituent basis;
(2) determination of the plant-soil assimilative capacity for each waste
constituent or component; (3) determination of the land area requirements
for each waste component or class of components; and (4) determination of the
land limiting constituent (LLC), which is that constituent or class of con-
stituents requiring the largest land area for safe treatment and ultimate
disposal.
The waste characterization for each constituent is expressed as kg/unit
time, while the assimilative capacity is expressed as kg of parameter/unit
area/unit time. The ratio of waste generation to assimilative capacity is
the area (hectares or acres) required for the environmentally acceptable
waste application to the terrestrial system. Ranking the required land areas
indicates one or more constituents as requiring the greatest land size, and
this constituent or class of constituents is defined as the LLC. Using the
LLC area guarantees that other waste constituents are applied at environ-
mentally acceptable rates.
WASTE CHARACTERIZATION AND GENERATION
A review of the literature was conducted to obtain information concern-
ing wastewater characterization in the coal gasification industry. Due to
the experimental and developing nature of "the state of the art" of coal
gasification, it has been impossible to obtain comprehensive information
concerning:(l)coal mass flow rates, (2) water mass flow rates, and (3) con-
centrations of inorganic and organic species including toxic organic com-
pounds (PNAs) in one assessment document or one coal gasification facility.
Waste constituents were identified, quantified, and waste generation
rates (kg/yr) were calculated with information obtained for an expected
typical full scale Lurgi coal gasification facility. That is, wastewater
was characterized for a full scale Lurgi facility: 250 x 10 SCFD of medium
to high BID synthetic natural gas (SNG), operating at a coal feed rate of
2245 x 10 Ib/hr and a condensate flow rate of 1897 x 10 Ib/hr, and using
North Dakota Lignite coal. Because of the dearth of information concerning
PNA concentrations and mass flows in the literature surveyed, several cal-
culational procedures were necessary to derive expected concentrations of
PNAs in the wastewater addressed. Expected concentrations of PNAs were
based on other waste constituents present. The result is a preliminary waste
characterization including constituent identification and waste generation
rates for over 9O individual constituents.
ASSIMILATIVE CAPACITIES AND LAND AREA REQUIREMENTS
Much information already exists with regard to the terrestrial assimila-
tion capacities for several soil types for a multitude of organic and inor-
ganic constituents identified in coal gasification wastewaters (Overcash and
Pal, 1979; and Sims and Overcash, 1980). Information is especially abundant
221
-------
with regard to inorganic species identified. A large body of information
exists indicating that PNAs are notf-generally biomagnified in vegetation
and crops (Sims and Overcash, 1981). However, information concerning the
soil assimilation of PNA constituents in coal gasification wastewaters
needed to be developed for several PNA compounds.
With the information obtained an initial land limiting constituent
analysis was conducted which did not take into account the PNA class of com-
pounds. The LLC analysis for coal gasification wastewater identified
cadmium as the constituent requiring the greatest land area (750 ha) for
land treatment.
To determine the soil assimilative capacities (SACs) for PNAs in coal
gasification wastes, it is necessary to determine realistic concentrations
of PNAs that would result from the land application of a typical coal gas-
ification waste. The land area determined in the initial LLC analysis
provided the basis for calculating the resultant PNA concentrations in soil
for each PNA compound. Waste generation for each PNA was calculated by
multiplying the PNA concentration by the volumetric flow rate to obtain
mass/time (kg/yr). The calculated waste generation was divided by the land
area determined in the LLC analysis (750 ha) to obtain the resultant soil
PNA concentration (mg/kg).
The effect of the presence of PNA compounds in coal gasification
wastewater on land area requirements can be evaluated by experimentally de-
termining the soil assimilative capacities. An evaluation of the SACs could
determine whether an individual PNA compound or the class of PNAs required
more or less land area for treatment than cadmium. With this information
a design for land application for the safe treatment and disposal of hazard-
ous and toxic components as well as other constituents in coal gasification
wastes is assured by using the LLC approach.
EXPERIMENTAL DESIGN
Experiments to determine the SACs of PNA compounds present in coal
gasification wastewaters were designed to obtain the following specific in-
formation: (1) rates of PNA transformation; (2) effect of PNA structure on
transformation rate? (3) effect of engineering management options on trans-
formation rate; (4) soil acclimation to PNAs; and (5) toxicity and geno—
toxic potential of soil-treated PNAs.
SELECTION OF PNA COMPOUNDS FOR STUDY
A subset of the total number of PNAs identified in coal gasification
wastewaters was selected based on the following criteria: (1) genotoxicity,
(2) molecular recalcitrance, (3) priority pollutant status, and (4) lack of
information concerning fate and behavior in the environment. Soil concen-
trations for the PNAs considered based on the LLC analysis conducted are
shown in Table 1. Waste PNA concentration, mass generation, and soil con-
centration resulting from land application using the LLC constraint (750 ha)
are included. For those PNAs not quantified in the literature, the highest
222
-------
concentration on the list was used (O.57 mg/kg in soil).
TABLE 1. PNA COMPOUNDS AND SOIL CONCENTRATIONS
PNA Compound
Acenaphthylene
Dibenzofuran
Acridine
Anthracene
Benzo(b)fluoranthene
Benzo ( k )f luoranthene
Benz(a)pyrene
Indeno (1,2, 3-cd ) py rene
Condensate
Concentration
(mg/1)
0.114
—
_
0.082
O.066
0.034
0.072
—
Waste
Generation
(kg/yr)
855
—
_
615
495
255
540
— •
Soil
Concentration
(mg/kg)
0.57
0.57
0.57
0.41
0.33
0.17
0.36
0.57
PNA COMPOUND APPLICATION
Each PNA compound was investigated as a separate solution applied on
separate soil reactors. Compounds were applied to the soil in small vol-
umes of solvent (methylene chloride), and were mixed thoroughly with the
soil to simulate soil incorporation of applied wastes, and to obtain an even
distribution of the PNA compound throughout the soil at the desired concen-
tration. Triplicate reactors were used for each PNA.
SOIL TYPE
Norfolk fine sandy loam is a common soil type, typical of the coastal
plain, used in land application systems in North Carolina. The Norfolk
series is a member of the fine—loamy, siliceous, thermic family of Typic
PaleuduIts.
ENVIRONMENTAL CONTROL CHAMBER
PNA compounds, at the indicated concentrations in 2OOg and 20OOg soil
in glass beakers, were incubated in an environmentally controlled chamber.
Environmental parameters that were controlled included temperature (25 C),
light exposure (dark to prevent photodegradation), and soil moisture (6O-
80% field capacity). Soil moisture was adjusted to 80% of field capacity
with water. Water was added when the soil moisture capacity decreased to
60% of field capacity to simulate field conditions of wetting and drying.
ANALYTICAL METHODS
Soil Extraction and Sample Preparation
The extraction procedure used for extracting PNAs from soil was based on
the high performance liquid chromatography procedure for analysis of PNA
223
-------
compounds in water samples (Federal Register, 1979). Soil moisture was
adjusted to 80% field capacity prior to extraction. Methylene chloride
(25O ml) was added to 20Og soil. The solvent-soil mixture was homogenized
for two minutes with a Tekmar Tissumizer. The supernatant was decanted
from the soil reactor and filtered through anhydrous sodium sulfate. The
filtrate was concentrated to a final volume of 3-5 ml.
Reverse Phaee-UV HPLC Analysis
A Waters HPLC was utilized with acetonitrile-water as the mobile phase,
and a C-18 Perkin Elmer Reverse Phase column was used as the stationary
phase. PNA compounds were identified with a UV detector at a wavelength of
254 nm.
Sample Fractionation for Ames Assay
Soil extracts were fractionated using a C-8 preparative Lobar size
A prepacked column. Polarity classes of degradation products were collected
in acetonitrile-water, evaporated, and redissolved in dimethyl sulfoxide
for the Ames assay.
BIODEGRADATION DETERMINATION
Kinetic parameters of interest with respect to biodegradation include
half-life (ty in days), rate of transformation (r in kg PIMA/ha-day), and
the rate constant (k in day" ). These kinetic parameters are directly re-
lated to the soil assimilative capacities for PNAs.
GENOTOXICITY
Polarity classes of soil PNA degradation products were tested with
the Ames assay (Ames et al., 1975). This assay is widely used for the
detection of potential carcinogenicity and mutagenicity of environmental
chemicals. Toxicity and mutagenicity were determined and dose-response
curves were developed. The assay was conducted with and without microsomal
activation. Strain TA-98, which detects frameshift mutations, and strain
TA-100, which detects base pair substitution mutations, were used.
ENGINEERING MANAGEMENT OPTIONS
Engineering management options, including analog enrichment, nutrient
addition, surfactant addition, and pH adjustment are tools which the
environmental engineer may use to stimulate biological activity and to
increase the rate of biodegradation of recalcitrant compounds. With
application of industrial wastes with low levels of substrate organics and
PNAs, the level of microbial activity would be expected to be similar to
that of the native soil. Engineering management options for organic con-
stituents are potential accelerators of microbial activity. Since the
soil assimilative capacity is directly related to the kinetics of degrada-
tion, increasing microbial activity may lead directly to increasing the
soil assimilative capacities for PNA compounds in coal gasification wastes.
224
-------
The effect of each engineering management option was tested separately.
Addition of an analog-substrate (carbon and energy source) has been shown to
increase general microbial activity and growth. Addition of the nutrients
nitrogen and phosphorus, which influence the growth of microorganisms, to
microbial cultures stabilizing compounds deficient in these elements has been
demonstrated to increase the rate of stabilization. Since surfactants can be
utilized to increase cell membrane permeability, surfactants may be useful in
increasing the solubility and cell membrane permeability of PNAs with a
resultant greater oxidation and degradation in the soil environment. These
three amendments were investigated for their direct effect on PNA degradation
kinetics.
Although increasing soil pH from less than seven to neutral generally
increases microbial activity, it may be especially important in soil systems
to encourage bacterial growth and competition vis-a-vis fungi. Major dif-
ferences with respect to microbial oxidation pathways of aromatic hydro-
carbons between bacteria and fungi are believed to exist with fungi, pre-
moninant at low soil pH, possibly metabolizing PNAs to more genotoxie pro-
ducts /than with bacteria (Dagley, 1975; Cerniglia et al., 1979). Since pH
has a significant effect on soil bacterial/fungal proportions, pH may be
an important engineering tool to direct the pathway of PNA degradation
through a series of detoxication reactions. Soil pH was adjusted with
calcium carbonate solutions to 7.0 in triplicate soil reactos.
Two PNA compounds chosen for intensive study with amendments were
anthracene and benz(a)pyrene. Anthracene is a three ring aromatic hydrocar*-
bon which serves as a carbon and energy source for microorganisms and is
weakly carcinogenic. Benz(a)pyrene is a five ring aromatic hydrocarbon
that has not been demonstrated to be a carbon and energy source, but is
believed to be degraded through cometabolic processes. Benz(a)pyrene
is a powerful carcinogen.
SOIL ACCLIMATION
Acclimation of the soil to each PNA compound was investigated by
spiking the soil at zero, three, and six months. Rates of degradation were
monitored for each time increment and were compared through time.
STATISTICAL ANALYSIS
Kinetic data were subjected to analysis of variance, and when sig-
nificant differences at the five percent level were found among PNA compounds
Duncan's New Multiple Range Test was employed to separate means. The
statistical procedures were performed using standard package programs of
Statistical Analysis Systems-76 (Barr et al., 1976).
225
-------
RESULTS AND DISCUSSION
RATES OF PNA TRANSFORMATION
Table 2 presents results from the laboratory study for kinetic param-
eters for the biodegradation of PNAs during a 90-day incubation period.
Results represent the average of triplicate reactors. Half-lives range
from a low of 18 days for acenaphthylene and dibenzofuran to indeno(l,2,3-
cd)pyrene which exhibited no measurable loss with time.
TABLE 2. LABORATORY DETERMINED KINETIC DATA FOR PNA DEGRADATION
PNA Compound Number of Initial
Rings
Acenaphthylene
Dibeneofuran
Acridine
Anthracene
Benzo(b)-
fluoranthene
Benzo(k)-
fluoranthene
Benz(a)pyrene
Indeno(l,2,3-cd)-
pyrene
3
3
3
3
5
5
5
6
Concentration
(mg/kg soil)
0.57
0.57
0.57
0.41
O.33
0.17
0.36
0.57
Half-life
Rate
Rate
Constant (mg
(todays)
*
18A
18 B
34^
98
B
89B
R
80c**
(k.day *) kg-day)
O.O39
O.O39
0.007
0.017
0.007
0.008
O.OO9
**
0.022
0.022
0.004
0.007
0.002
0.001
0.003
**
Values represent means of three replicates. Means followed by the same
letter are not significantly different at the O.05 level.
**
No decrease in Indeno(l,2,3—cd)pyrene could be detected by HPLC.
Kinetic parametersfor other PNAs identified in coal gasification wastes
which were not included in the laboratory study are given in Table 3. The
information for these compounds was obtained in a comprehensive review of
the literature (Sims and Overcash, 1981).
TABLE 3.
LITERATURE VALUES FOR KINETIC DATA FOR
PNA DEGRADATION
PNA Compound
Naphthalene
Indole
Fluorene
Fluoranthene
Phenanthrene
Number of
Rings
2
2
3
4
3
Initial
Concentration
(mg/kg soil)
7.0
500
0.9
16.5
2.1
Half-life
(todays)
0.12
1.0
39
143
26
Rate
Constant
(k.day"1)
5.78
O.693
0.018
O.OO5
0.027
-Rate
(mg
kg-kJay)
40.4
364.5
O.0i6
0.080
0.056
226
-------
Results for PNA degradation kinetics from the laboratory study and
from the literature review indicate that most PNAs addressed have reasonable,
finite half-lives in soil pystems.
EFFECT OF PNA STRUCTURE ON TRANSFORMATION RATES
Table 2 summarizes the results of the statistical analysis of PNA
compounds by structure and half-life. Arranging PNAs by number of rings
and half-life indicates that there are three distinct, statistically sig-
nificantly different groups of PNA compounds. The general trend is for
lower ring compounds to exhibit faster degradation kinetics i.e., there is
an inverse relationship between the number of rings (PNA size) and
half-life.
These results are consistent with the findings of other researchers
for aquatic and soil systems for other PNAs. Information developed here
adds to the list of quantitative data available for environmental engineers
concerned with the design of land treatment systems for the coal gasifica-
tion industry.
EFFECT OF ENGINEERING MANAGEMENT OPTIONS ON TRANSFORMATION KINETICS
Results showing the effects of environmental management options on
biodegradation kinetics for the PNAs studied in the laboratory are given
in Tables 4 and 5.
TABLE 4. EFFECT OF AMENDMENTS ON ANTHRACENE DEGRADATION
Amendment
None
Nutrients
PH
Surfactant
Analog Enrichment
Half-life
(todays)
*
41A
41A
45A
A
38
Rate
Constant
(k,day )
0.017
0.017
0.017
0.015
O.018
Rate
(mg
kg-day)
0.68
0.69
0.69
O.63
0.75
Values represent means of three replicates. Means followed by the same
letter are not significantly different at the 0.05 level.
The degradation of anthracene, a three ring PNA compound which serves
as a substrate (carbon and energy source) for soil microorganisms, does not
appear to be influenced by the engineering management options used in
this studv. Statistical analysis of the laboratory data indicate that the
relatively•short half-life for anthracene with no amendment addition is not
statistically different from the half-lives for anthracene treatment with
any of the amendments.
227
-------
TABLE 5. EFFECT OF AMENDMENTS ON BENZ(a)PYRENE DEGRADATION
Amendment
None
Nutrients
pH
Surfactant
Analog Enrichment
Half- life
(tv,days)
" *
90A
81
64A
87B
64
Rate
Constant
(k,day )
O.OO77
0.0082
0.0108
O.O080
0.0108
Rate
(mg
kg-day)
0.28
0.31
0.39
0.29
0.39
Values represent means of three replicates. Means followed by the same
letter are not significantly different at the 0.05 level.
The degradation of benz(a)pyrene, a five ring PNA compound, is con-
sidered to be cometabolized i.e., cannot serve as a source of carbon and
energy for the growth of microorganisms, does appear to be influenced by
the engineering management options used in this study. Statistical analysis
of the data indicated statistically significant differences among the
treatments. The amendments which effected a significant decrease in the
half-life of B(a)P included analog enrichment and pH adjustment.
This information has direct implications for the design of land
treatment systems for coal gasification wastes. The data suggest that it
may be possible to influence the degradation rates of recalcitrant and
hazardous organic compounds through engineering management options.
SOIL ACCLIMATION TO PNAs
Results for the acclimation of soil systems to PNA compounds are still
being analyzed. Extent of acclimation appears to vary among the PNA com-
pounds. Indeno(l,2,3-cd)pyrene showed the greatest acclimation from no
measurable degradation to 360 days half-life to 201 days half-life for
3,6, and 9 months incubation respectively. More information must await
additional data collection and statistical analysis.
TOXICITY AND GENOTOXIC POTENTIAL OF SOIL-INCUBATED PNAs
PNA parent compound, benz(a)pyrene and degradation products collected
as polarity classes were not found to be toxic at concentrations from lOug/
plate to 500 ug/plate to be Salmonella typhimurium strains TA-98 and TA-100
used in this study,
Results for mutagenesis testing for B(a)P and degradation products are
presented in Table 6. Data are given for the soil control, and at six
months of incubation of B(a)P in soil.
228
-------
TABLE 6. MUTAGENICITY OF SOIL-INCUBATED BENZta)PYRENE
Sample
Time
(months)
Mutagenic Ratio at 500 ug/plate
Without Activation
With Activation
TA-98
TA-100
TA-98
TA-100
Soil Control
Parent Compound
(B(a)P)
Polar Class
Fraction
Nonpolar Cl§s,s
Fraction
6
6
6
6
1.81
1.37
1.18
1.28
1.20
1.09
1.26
1.15
1.84
7.82
2.96
3.72
1.72
3.16
1.50
2.19
Mutaqenic Ratio is defined as a number of revertants with sample divided
by the number of revertants without sample. A test compound or sample
is considered negative if the mutagenic ratio is less than 2.0
**
The Nonpolar class fraction was that fraction collected in preparative
high performance liquid chromatography which appeared after the parent
compound (B(a)P) for an elution gradient proceeding from more polar
fractions to less polar fraction with increasing run time.
Results indicate that neither parent compound nor degradation products
are mutagenic without mammalian microsomal activation. This is well known
for B(a)P, but is not known for soil metabolites of B(a)P.
Results also indicate that the mutagenic potential of degradation
products of soil incubated B(a)P are much less than the parent compound.
The highest mutagenic potential (3.72) is associated with the Nonpolar
class fraction.
This information suggests that after six months of soil incubation,
the products of biodegradation of B(a)P are much less mutagenic than the
parent compound. A detoxication pathway is therefore indicated for
B(a)P biodegradation in soil.
CONCLUSIONS
Results from this preliminary study indicate that land application
technology for fossil fuel wastes is promising. A protocol has been
established for obtaining the soil assimilative capacities for recalcitrant
and mutagenic PNA compounds, and for etetermining genotoxic potential of
parent compounds and metabolites in soil. With the significant cost benefit
for land treatment and the demonstrated potential to actually decompose
recalcitrant and hazardous organics, it would apppear reasonable to proceed
to further evaluations. Using this protocol a more detailed design based on
specific waste characterization and site-specific analyses would follow
for a particular sunfuel facility.
229
-------
ACKNOWLEDGMENTS
The authors would like to thank Dr. Dean Smith of the Industrial
Environmental Research Laboratory (Research Triangle Park) of the US EPA
for allowing us the opportunity to present the results of our work at
North Carolina State University. We also would like to thank Dr. Charles
Sparacino and Mr. Thomas Hughes of the Research Triangle Institute
(Research Triangle Park) for their invaluable assistance in the analysis
and toxicity testing of the samples studied.
REFERENCES
1. Ames, B.N., J. Mccann, and E. Yamasaki. 1975. Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian-microsome
mutagenicity test. Mutation Research 31: 347-354.
2. Barr, A. J., J. H. Goodnight, J. P. Sail, and J. T. Helwig, 1976.
A User's Guide to SAS-76. SAS Institute, INC., Raleigh, N. C. 329 pp.
3. Cerniglia, C. E., and D. T. Gibson. 1979. Oxidation of benzo(a)pyrene
by the filamentous fungus Cunninghamella elegans. Journal of Biological
Chemistry 254(23): 12174-12180.
4. Dagley, S. 1975. Microbial degradation of organic compounds in the
biosphere. Scientific American 63: 681-689.
5. Federal Register. 1979. Polynuclear aromatic hydrocarbons- method 610.
44(233): 69514-69517.
6. Herbes, S. E., G. R. Southworth, and C. W. Gehra. 1976. Organic con-
taminants in aqueous coal conversion effluents: environmental consequen-
ces and research priorities. Trace Substances in the Environment- A
Symposium. D. D. Hemphill (ed). University of Missouri, Columbia, MO
7. Overcash, M. R., and D. Pal. 1979. Design of land treatment systems
for industrial wastes-theory and practice. Ann Arbor Science.
8. Sims, R. C., and M. R. Overcash. 1980. TERRETOX- Catalogue of informa-
tion on the behavior of organic chemicals in terrestrial systems. In
partial fulfillment for the requirements for the Ph.D. degree. Publica-
tion in preparation.
9. Sims, R. C., and M. R. Overcash. 1981. Fate of PNAs in soil-plant
systems. In preparation for submission to Residue Reviews^ Springer-
Verlog New York Inc.
230
-------
Session III: AIR-RELATED ENVIRONMENTAL CONSIDERATIONS
Chairman: Theodore G. Brna
U.S. Environmental Protection Agency
Research Triangle Park, NC
231
-------
REMOVAL OF ACID GASES AND OTHER CONTAMINATES FROM COAL GAS t
USING REFRIGERATED METHANOL
by
J. K. Ferrell, R. M. Kelly, R. W. Rousseau,
and R. M. Felder
ABSTRACT
The steam-oxygen gasification of a New Mexico subbituminous coal was
carried out in a pilot-scale fluidized bed gasifier. Gas cleaning was
accomplished by a hot cyclone, a water quench-venturi scrubber, filters, and
an acid gas removal system using refrigerated methanol as the solvent.
Results of both gasification and gas cleaning are described. Refrigerated
methanol proved to be effective in cleaning the gasifier make gas, however,
the presence of several reduced sulfur species and hydrocarbons was detected
in the absorber, flash tank, and stripper exit gas streams over a wide range
of operating conditions. While a variety of simple aromatics accumulated in
the recirculating methanol, essentially no polynuclear aromatic compounds
were detected. Most polynuclear aromatic compounds were evidently removed in
the gas quenching process.
232
-------
INTRODUCTION
As a part of a continuing research program on the environmental aspects
of fuel conversion, the U. S. Environmental Protection Agency has sponsored
a research project on coal gasification at North Carolina State University in
the Department of Chemical Engineering. The facility used for this research
is a small coal gasification-gas cleaning pilot plant. The overall objective
of the project is to characterize the gaseous and condensed phase emissions
from the gasification-gas cleaning process, and to determine how emission
rates of various pollutants depend on adjustable process parameters.
A complete description of the facility and operating procedures is given
by Ferrell et al., Vol I, (1980), and in abbreviated form by Felder et al.
(1980). A schematic diagram of the Gasifier, the Acid Gas Removal System
(AGRS), and other major components is shown in Figure 1.
In an initial series of runs on the gasifier, a pretreated Western
Kentucky No. 11 coal was gasified with steam and oxygen. The results of
this work are given by Ferrell et al., Vol II, (1981), and were presented at
the EPA Symposium on Environmental Aspects of Fuel Conversion Technology V,
held in St. Louis, Mo., September, 1980.
The second major study carried out on the facility was the steam-oxygen
gasification of a New Mexico subbituminous coal (from the Navaho mine of the
Utah International Co.) using refrigerated methanol as the AGRS solvent.
This paper presents a brief summary of the gasifier operation using this
coal, shows examples of analyses of some of the gasifier effluent streams,
and presents a summary of the results of the operation of the AGRS using the
gasifier make gas as feed.
SUMMARY OF GASIFIER OPERATION
The fluidized bed gasifier and raw gas cleaning system (cyclone, venturi
scrubber, filters and heat exchanger) used for these studies was originally
designed for the gasification of a devolatilized coal char with a very low
volatile matter content. Extensive modification of the upper part of the
gasifier, the venturi scrubber system, and the heat exchanger was required
for operation with the high volatile matter New Mexico coal. Table 1 shows
an analysis of the char and coal used in studies to date. After
modification, the system functioned well in providing a clean, dry gas to the
acid gas removal system.
All of the experimental work so far has been carried out with the solid
coal particles fed into the reactor several feet above the top of the
fluidized bed. The particles are thus in contact with the hot product gases
for several seconds before mixing into the fluidized bed, a mode of operation
233
-------
Syn Gaa
Filter
< enturl
crubber
Acid Gas
Dehydrator
Sour-Gaa JL
O—| |
Heat
Exchanger
Heat
Exchanger
I L_
Plant Water
Circulation
Pump
Solvent Pump
S = Sample Port
Figure 1. Pilot Plant Facility
-------
that tends to maximize the production of tars and other organic liquids from
the coal. It is an excellent mode of operation for our present purpose since
it produces relatively high concentrations of environmentally important
elements and compounds.
TABLE 1
COAL AND CHAR ANALYSIS
Coal Char New Mexico Coal
Proximate Analysis
Fixed Carbon 86.0 42.0
Volatile Matter 2.4 35.4
Moisture 0.9 10.5
Ash 10.7 22.6
Ultimate Analysis
Carbon
Hydrogen
Oxygen
Nitrogen
Sulfur
Ash
83.8
0.6
2.2
0.1
2.6
10.7
52.5
4.8
18.3
1.2
0.6
22.6
A total of 15 gasifier runs were made covering a range of reactor
parameters. For this series of runs, the average temperature of the
fluidized bed was varied from about 1600 F to 1800 F, and the molar steam to
carbon ratio was varied from about 1.0 to 2.0. The coal feed rate and the
reactor pressure were kept nearly constant. Several of the first reactor
runs were made with mixtures of coal and char, but all integrated runs
reported on later were made with 100% coal.
At the lower temperatures the production of methane and of tars and
other hydrocarbons is maximized. As the temperature is increased, the make
gas rate increases, the production of methane and other hydrocarbons
decreases, and the concentration of C02 increases. As an example, conditions
and mass balances for run GO-76 are shown in Table 2.
GASIFIER MODELING RESULTS
To aid in the formulation of gasifier performance correlations, a simple
model has been developed which considers the gasification process to occur in
three stages: instantaneous devolatilization of coal in a zone above the
fluidized bed, instantaneous combustion of carbon at the bottom of the bed,
and steam-carbon gasification and water gas shift reaction in a single
perfectly mixed isothermal stage. The model is significant in and of itself,
235
-------
Table 2
mmmmmmmtmttmmmmmt
* *
t NCSU DEPARTMENT OF CHEMICAL ENGINEERING *
* *
* FLUIDIZED BED COAL GASIFICATION REACTOR *
* *
tttmmtmmtmmmttmttmmmt
RUN 60-76 4-28-81 13130-16!15
REACTOR SPECIFICATIONS
PRESSURE = 100,6 PSIG ( 794,9 KPA)
TEMPERATURE - 1711,6 DEG.F ( 933,1 DEG.C)
BED HEIGHT = 38,0 IN, <0,97 METERS)
BED DIAMETER - 6,0 IN, (0,152 METERS)
ESTIMATED BED VOIDAGE = 0,80
SOLIDS HOLDUP = 10,4 LB ( 4.7 KG)
FEED RATES AND RATIOS
COAL =50,86 LB/HR (23,07 KG/HR)
STEAM 57,37 LB/HR (26,02 KG/HR)
OXYGEN = 12,62 LB/HR (5,72 KG/HR)
NITROGEN 6,74 LB/HR (3,06 KG/HR)
PURGE N2 = 10,33 LB/HR ( 4.69 KG/HR)
STEAM/CARBON = 1,54 MOLES STEAM/MOLE C
02/CARBON = 0,19 MOLES 02/MOLE C
N2/02 0,61 MOLES N2/MOLE 02
COAL
GASES
TOTAL INPUT
ELEMENTAL MATERIAL BALANCES t FLOWS IN LB/HR
MASS C H 0 N S
50,9 24.81
87,1 0,00
137,9 24,81
2,01
6.42
8,43
11,53
63,56
75,09
0,51 0,422
17,07 0,000
17,58 0,422
CHAR
DUST
GASES
WASTEWATER
TOTAL OUTPUT
13,3
0,1
121,6
0,0
135,0
6,10
0,03
18,38
0,00
24,51
0,12
0,00
8,44
0,00
8,56
0.00
0,00
76,97
0,00
76,98
0,07
0.00
17.52
0,00
17,59
0.074
0.000
0,263
0.000
0,137
Z RECOVERY
97,91! 98,82 101,62 102,52 100,02 80,02
OUTPUT VARIABLES
CARBON CONVERSION (PERCENT) 74,1
DRY MAKE GAS FLOU RATE (SCFM) 19,2
HEATING VALUE OF SWEET GAS (BTU/SCF) 373.7
EFFLUENT FLOW RATES (LB/HR)
CO 15,65
H2 2,06
CH4 3,40
C02 33,42
N2 17,52
H2S 0,271
COS 0,015
236
-------
but its particular importance to the project is that it enables the
specification of gasifier conditions required to produce a feed to the acid
gas removal system with a predetermined flow rate and composition.
In a previous report (Ferrell et al., 1981), the structure of the model
was presented, and the ability of the model to correlate data on the
gasification of a devolatilized bituminous coal was demonstrated. The model
was subsequently extended to include the evolution of volatile gases in the
pyrolysis stage of the gasification process, and used to fit the data from
the present series of runs with the New Mexico subbituminous coal. The model
takes as input the average reactor bed temperature and pressure, the bed
dimensions, feed rates of coal, steam, oxygen, and nitrogen, solids holdup in
the bed, and ultimate analysis of the feed coal, and calculates carbon
conversion and make gas flow rate and composition. A complete description of
the model in its present form will be given in an EPA report now in
preparation. Plots of model predictions vs measured values of carbon
conversion and dry make gas flow rate are shown in Figures 2-3. The
reasonably close proximity of most points to the 45 degree line is gratifying
in view of the simplicity of the model. The proximity of the points
corresponding to the "best" runs (from the standpoint of satisfying mass
balances) is even more satisfying.
The model also does a good job of correlating data on the evolution of
individual species. Figure 4 shows predicted versus measured values of the
rate of production of CO from the gasifier. Similar plots have been obtained
for the production of H_ and C0~. The good correspondence seen in these
plots suggests that the model can be used to predict the composition of the
gasifier make gas for a specified set of reactor conditions, and also to
study the effects of individual reactor variables on yield.
AGRS OPERATION AND RESULTS
As previously mentioned, top feeding coal into the gasifier allows a
substantial amount of devolatilization to take place before the coal enters
the fluidized bed. While most commercial fluidized bed gasifiers will use a
deep-bed injection method of feeding coal into the fluidized bed, it was
decided not to modify our system in order to maximize the formation of tars,
oils, and other hydrocarbons and to provide a more complete test of the AGRS.
It should also be noted that the relatively simple acid gas removal
system used in this study lacks the complexity of the selective systems found
in many physical absorption processes. These systems, which use more than
one absorber and stripper, and often several flash tanks, separate sulfur
gases from carbon dioxide before further processing of the acid gas. This is
done to concentrate the sulfur gases before they are fed to a sulfur recovery
unit, and to recover the C02 or vent the C02-rich stream to the atmosphere.
While the AGRS used in this study could have been modified to emulate an
237
-------
c
o
-o
-------
Dry Make Gas Flow Rate (SCFM)
O Element mass balance
worse than 8%
<2) All Element mass balances
within 8%
• All element nass balances
within 6%
24
o
22
•o
O)
o.
cu
-a
o
18
16
14
14
16
18
20
22
24
Figure 3.
Experimental
Predicted vs. Experimental Dry Make Gas Flow Rate,
Gasification of New Mexico Coal
239
-------
c
O
-o
01
n_
0)
•o
o
22
20
18
16
14
12
10 ~
8 -
CO Production Rate (Ib/hr)
O Element mass balance
worse than 8%
® All element mass balances
within 8%
^ All element mass balances
within 6%
O
12
14 16
Experimental
18
20
22
Figure 4. Predicted vs. Experimental Production Rate of CO
from Gasification of New Mexico Coal
240
-------
existing selective absorption process, it was decided that data obtained from
a relatively simple but well-characterized system would be of more use than
data obtained from a fairly complex system, similar but not identical, to
existing commercial systems. Through judicious use of computer simulation
and engineering calculations, the data obtained from our system should be
extrapolatable to more industrially significant situations.
EXPERIMENTAL PROGRAM
In designing the experimental program to be used in these studies, the
use of a full factorial experimental design was not believed to be necessary.
The program was designed to cover the broadest range of operating conditions
possible for the system of New Mexico coal and refrigerated methanol.
Effects of variations in important process variables were examined by
comparing all runs to a base case. Although this approach is not exhaustive,
it provides a framework with which to examine the environmental consequences
of acid gas removal with methanol. In addition, the work done in this study
will be useful in developing experimental programs for other coals and acid
gas removal solvents to be studied in our facility.
Table 3 shows the operating conditions used for the nine runs made in
this part of the study. Also shown are inlet and outlet gas concentrations
for the major acid gases in the absorber.
DISCUSSION
From an environmental perspective, operation of the acid gas removal
system in a coal gasification process becomes important when harmful
compounds or pollutants may be discharged to the atmosphere. Although there
are a wide variety of extremely toxic materials released from coal during
gasification, as long as they remain within the gas cleaning system or are
properly processed, they pose little problem. However, while these harmful
materials are seldom purposely discharged to the atmosphere from the acid gas
removal system under normal operating conditions, several of the AGRS gas
streams are fed to downstream processes. There, inability to handle toxic
compounds and pollutants may result in their discharge to the atmosphere. It
is therefore important to know what compounds enter the AGRS, and how they
distribute in the system under various processing conditions.
Of the runs shown in Table 3, run GO-76, AMI-57 will be used to
illustrate AGRS performance results. Gas analyses from the six different
locations shown in Figure 1 are given in Table 4.
241
-------
TABLE 3
OPERATING CONDITIONS FOR COAL GASIFICATION RUNS
Run
Number
AMI-
GO-
43
68B
44
69B
45
70
47
71B
52
72
53
73B
57
76
59
78
60
79
Absorber
Pressure Atm
Packing Height Ft
Inlet Liquid Flow
Inlet Solvent Temp
Inlet Gas Flow
Inlet Gas Temp F
H^S in ppm
H_S out ppm
COS in ppm
COS out ppm
C02 in %
C02 out %
Flash Tank
Pressure Atm
18.0 18.0 31.6 31.6 31.6 18.0 31.6 18.0 24.8
7.1 7.1 7.1 7.1 21.3 7.1 7.1 7.1 7.1
63.6 63.5 129.3 130.5 127.9 127.5 61.7 127.3 100.8
-36.1 0.8 -36.3 -34.9 -35.7 -5.4 -36.2 -3.5 -21.0
18.1 17.0 17.1 16.8 16.8 17.9 15.4 14.9 16.0
40.1 47.6 22.3 36.3 38.1 42.5 50.4 44.2 56.4
2950 2900
220 280
119 112
12 10
20 22
1
2550 4682 3023 1710 2868 3180 2139
260
79
5
21
3
151
133
7
22
105
67
6
23
172
60
7
22
48
76
1
22
190
81
4
23
260
84
19
22
4.2 10.9 11.0 11.0 11.0 10.8 11.0 10.8 10.7
Stripper
Pressure Atm 1.7 1.7
Packing Height Ft 21.3 21.3
Stripping N Flow 1.3 1.3
Inlet Gas Flow 75.0 75.0
Inlet Solvent Temp F 8.4 14.6
Overall Mass Balance Closure
Gasifier
AGRS
98.0 96.6
103.8 102.3
1.7
21.3
1.3
75.0
-5.6
95.3
103.0
1.7
21.3
1.3
75.0
48.0
1.7
21.3
1.3
75.0
48.0
1.7
21.3
1.3
75.0
48.1
1.7
21.3
1.3
75.0
48.1
1.7
21.3
1.3
75.0
48.1
1.7
21.3
1.3
75.0
37.6
103.5 98.6 97.8
104.8 101.4 103.1
97.9 98.3 100.6
99.2 101.8 95.5
All Flows in Ib-mole/hr-ft'
The sample train sample is taken downstream from the cyclone separator
and is the closest sampling point to the gasifier. Unreacted steam in the
gas is first condensed and removed before the sample is taken. The PCS tank
sample is taken after the gas quenching step but before the dehydrating
towers and sour gas compressor. The sour gas sample is taken after the PCS
tank and after the gas has been dehydrated, compressed, and cooled to remove
the heat of compression. The high levels of several hydrocarbon and sulfur
species in the sour gas sample may be attributed to the presence of
condensate in the gas sampling lines. A trap located near this sampling
station accumulated small amounts of a condensed hydrocarbon phase which was
analyzed by GC/MS after run AMI-60. It is thought that this sample provides
qualitative information on the variety of trace compounds entering the AGRS.
242
-------
When the system is operating at steady state, the compositions of the
sample train, PCS tank and sour gas samples will be nearly the same. This is
some indication of the quality of the run. More detailed descriptions of the
sampling and analytical procedures can be found in Ferrell et al. (1981).
TABLE 4
GAS ANALYSIS SUMMARY FOR AMI-57/GO-76
Species
H
c69
C-H,
S2s6
cSs
d
CO
Benzene
Toluene
Ethyl Benz.
Xylenes ^
Thioph,ene
CH-SH ^
CJE^SH
C^2 *
Propylene
Propane
Butane ^
Methanol
Sample
Train
31.60
23.51
0.52
0.72
0.250
0.0078
19.36
6.56
17.29
0.087
0.031
0.0016
0.0080
44
16
TRACE
TRACE
1505
208
185
PCS
Tank
31.11
23.91
0.53
0.72
0.284
0.0076
19.61
6.46
17.47
0.097
0.034
0.0017
0.0094
44
29
3
1521
198
150
Sour
Gas
31.29
21.98
0.56
0.76
0.287
0.0076
19.93
6.57
17.92
0.234
0.534
0.0450
0.1557
127
28
8
TRACE
1811
253
143
Sweet
Gas
42.38
0.0242
0.0164
0.0048
0.0001
26.79
7.54
23.35
TRACE
0.0054
TRACE
TRACE
107
301
54
Flash
Gas
15.58
25.99
1.28
1.92
0.090
0.0041
19.27
14.20
21.55
0.0031
0.0033
5
TRACE
995
172
91
Acid
Gas
0.00
64.74
1.54
2.13
0.66
0.027
23.06
2.36
1.80
0.15
0.030
TRACE
TRACE
4640
2203
71
3.68
Parts per Million (volume)
Estimated
Acid Gas Removal
The primary function of the AGRS is to remove C0« and sulfur compounds
from the gases produced during coal gasification. When using refrigerated
methanol, the absorber also acts as an excellent trap for any other compound
which condenses or disolves in the methanol at absorber conditions. Table 3
shows the concentrations of H2S, COS and C02 for the nine coal gasification
runs. Using AMI-47 as the base case, the effect of process conditions on
acid gas removal can be seen.
Because the acid gas content of the solvent entering the absorber has a
pronounced effect on removal efficiencies, ineffective solvent regeneration
in the stripper can be a problem. In run AMI-45, the stripper was operated
243
-------
at -5.6°F rather than at 48°F as in AMI-47, Comparing the results from
AMI-47 and AMI-45, the former shows a significantly better acid gas removal
efficiency as a result of the higher operating temperature in the stripper.
The effect of packing height on removal efficiency can be seen by
comparing AMI-47 with AMI-52. In AMI-52, 14.2 feet of additional packing was
used with only a marginal improvement in the outlet H~S levels. The acid gas
removal efficiencies for the two runs are almost the same.
From the results of these three runs and the other runs in Table 3, it
appears that for the range of conditions studied here, the most significant
factor in high acid gas removal efficiencies is stripping efficiency. With
the use of more extreme operating conditions and "cleaner" methanol fed to
the absorber, the levels of C0_, COS and H~S in the sweet gas can be reduced
to acceptable levels. This is a particularly important point in the case of
COS removal which poses problems for many coal gas cleaning systems. From
the data collected in this study, it appears that refrigerated methanol is
effective in removing COS and no unusual solubility characteristics were
evident at moderate pressures and low liquid temperatures.
Trace Sulfur Compounds
There are also several other sulfur compounds besides H_S and COS
present in the gas fed to the AGRS which must be removed. Table 4 shows the
distribution of several of these compounds in the AGRS. While there is some
scatter in the analyses for methyl mercaptan, thiophene, CS_, and ethyl
mercaptan/dimethyl sulfide, it appears that in most runs they are removed to
very low levels in the absorber.
A point of potential environmental significance is that while these
compounds are removed to low levels, they are not completely accounted for in
the flash and acid gas streams. This can be seen for methyl mercaptan and
thiophene, which are present in relatively high levels in the feed gas.
These compounds will accumulate in the recirculatory solvent and most likely
eventually leave the system in one of three exit streams: sweet gas, flash
gas, or acid gas. Because most sulfur recovery systems cannot treat
mercaptans and thiophene, they will present emission problems if some
additional method of treating these gases is not used. This can be A
significant problem because the total sulfur from mercaptans, organic
sulfides, CS-, and thiophene is approximately half of the total sulfur
associated with COS. If these compounds appear with the sweet gas, they are
likely to affect adversely downstream methanation catalysts. The presence of
these compounds in the sweet gas stream is also a problem if the gas is to be
burned for immediate use because the sulfur in these compounds will be
converted to S0_.
In examining the results from all runs, there appears to be some pattern
of trace sulfur species distribution. The increase in stripper temperature
from the low levels of AMI-45 to 48°F for AMI-47 resulted in substantially
greater amounts of mercaptan and thiophene in the acid gas stream. The
244
-------
results from AMI-44, which represents the "worst case" for absorber
performance, show an increase in mercaptan levels in the sweet gas.
Apparently, the conditions used in the absorber for this run were not
sufficient to remove the mercaptans to low levels. CS_ seems to distribute
to all exit streams in most of the runs despite the differences in process
conditions.
Perhaps the most significant finding here is that over a wide range of
processing conditions, the presence of at least small amounts of several
different sulfur species is to be expected in all AGRS exit streams, and
provision must be made for handling the associated problems.
Aliphatic Hydrocarbons
As the amount of volatile matter present in a particular coal increases,
the production of aliphatic, aromatic, and polynuclear aromatic compounds
produced during gasification also increases. Over the range of conditions
studied here, the most significant point to be made about the distribution of
aliphatic hydrocarbons is their presence in significant quantities in the
flash and acid gases. Although flashing of the methanol down to atmospheric
pressure prior to stripping would release most of the hydrocarbons, the
CO^-rich flash gas would still contain substantial amounts of several
hydrocarbon species. This stream would require further processing before it
could be vented.
In run AMI-57, in which the gasifier was operated at a lower temperature
to increase the production of hydrocarbons, the aliphatics (excluding
methane) make up almost 4.5% of the acid gas stream and 3.5% of the flash gas
stream. While staging the flashing operations may result in a better
distribution of these compounds, the total product from the flashing and
stripping operations must be either recovered as product, fed to a sulfur
recovery unit, or vented to the atmosphere. Since it is unlikely that all of
the aliphatic hydrocarbons will appear in the sweet gas stream, as evidenced
by the data collected here, additional treatment will be necessary to prevent
their eventual appearance in a vent stream.
There appears to be no unusual pattern of distribution of aliphatic
hydrocarbons in the AGRS. The lighter hydrocarbons— methane, ethylene, and
ethane— seem to distribute as would be indicated from an examination of
their pure-component solubilities in methanol. The magnitude of their
solubilities, however, are greater than would be expected from Henry's law,
especially at the high pressures used in the absorber. This is evident from
the lower than predicted levels of ethane and ethylene in the sweet gas in
several of the runs.
245
-------
Aromatic Hydrocarbons
Because large amounts of aromatic hydrocarbons are produced during coal
gasification, the potential for environmental problems is great. These
compounds, which range from benzene to polynuclear species of many forms,
must be prevented from escaping from the gas cleaning process and their
distribution throughout the gas cleaning system is of great concern.
Table 3 summarizes the information obtained from a single run. The
simpler aromatics, benzene, toluene, and xylene, typically make up 0.1% (by
volume) of the gas stream entering the AGRS.
Analyses performed for selected runs indicate that significant
quantities of these compounds are found in the solvent leaving the stripper.
Results from two runs are reported in Table 5. These compounds will build up
in the solvent to the point of saturation. If the solvent is not effectively
purged of these compounds periodically, they will begin to appear in several
of the process streams.
TABLE 5
METHANOL ANALYSIS FOR STRIPPER EXIT
ALL ANALYSES REPORTED IN PPM (VOLUME)
AMI-44/GO-69BAMI-57/GO-76
Benzene
Toluene
Ethylbenzene
Xylenes
190
200
30
70
157
196
87
203
Methanol Analysis
In order to identify the various hydrocarbon species that accumulate in
the methanol, samples of the methanol leaving the stripper were taken for
several runs. These samples were then analyzed by gas chromatography/mass
spectrometry.
Initial samples taken of the stripped methanol were analyzed by the
Research Triangle Institute, Research Triangle Park, North Carolina. The
results from AMI-44/GO-69B and AMI-57/GO-76 are shown in Figures 5 and 6 and
in Tables 6 and 7. The gasifier conditions for AMI-57/GO-76 were designed to
result in the production of larger amounts of heavy organics and tar than the
other runs made in this study. The spectra from Figures 5 and 6 show that
this result was achieved. The presence of several siloxanes and phthalates
were probably related to some contamination of the sample during processing.
246
-------
OQ
I-!
CD
Ol
01
H-
CD
O
CO
rt-
i-l
H-
T3
Tl
n>
w
fD
rt
ff
P
l-i
i
H
I
B»nz«n*
Tolucn*
o
a
• toe
Atoj,r..,.niAfc^
IKO f*OO t»OO
I | I I
2 MO
Tim* (••<:)
O
O
I
w
-------
OQ
c
l-i
01
3OOOO
ho
-P-
00
W
H-
cn
o
Hi
fl)
rt
y
PI
3
O
X)
fl>
i-i
w
2OOOO-
100OO
20SO
21OO
2160
2200
2260
23OO
236O
24OO
T
2460
Tbm (t«c)
-------
Results from these runs indicate that most of the compounds accumulating
in the methanol are simple aromatics, primarily substituted benzenes. A few
Cj~ and C,, isomers were identified, indicating that napthalene is probably
present but at trace levels. The presence of trace amounts of C,, and C
isomers were found in AMI-57 but they could not be better identified.
may be polynuclear aromatics but they were present in very small amounts
relative to the simpler aromatics.
These
Later in the sampling program, samples from AMI-60/GO-79 were analyzed
by the GC/MS facility at North Carolina State University. The results from
these analyses are shown in Figures 7 and 8. These Figures show the mass
spectra for the stripped methanol before and after the run. Although
compound identification was not performed for these analyses, comparison of
the two spectra shows the relative changes in the levels of hydrocarbons.
This methanol had been used for several previous runs and had accumulated
significant quantities of a variety of organics. The spectra for the sample
taken at the end of the run show that the locations of most peaks have not
changed but the relative sizes of several peaks have. This indicates that
these hydrocarbons were in fact, accumulating and will continue to do so
until they saturate the solvent.
TABLE 6
STRIPPER EXIT METHANOL FOR AMI-44/GO-69B
Peak Number from Figure 5
1. C02 12.
2. butene isomer 13.
3. pentene isomer 14.
4. 2-methyl-2-butene 15.
5. cyclopentadiene 16.
6. cyclopentene 17-
7. C/-H,2 isomer 19.
8. C6H14 isomer 20.
9. c6Hi0 isomer 21.
10. methyl cyclopentane 22.
11. methyl cyclopentadiene 23.
benzene
C..R., isomer
C..H, , isomer
. /, lo
toluene
CgH,, isomer
CgH,g isomer
ethyl benzene
xylene (M,P)
styrene
xylene (0)
CnH,Q isomer
V Jo
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
CqH-0 isomer
C1()H20 isomer
C10H22 lsomer
^10^22 i-somer
undecane
C,iH^- isomer
Sat a hydrocarbon
phthalate
Sat'd hydrocarbon
Samples of liquid condensing in the knockout tank downstream from the
sour gas compressor were collected and analyzed by GC/MS. This condensate
contains most of the heavier hydrocarbons fed to the AGRS. Results of these
analyses are presented in Figure 9 and Table 8, and show that the compounds
identified are very similar to those found in the stripped methanol from
AMI-44 and AMI-57. Again, mostly simple aromatics were found. No
polynuclear aromatics were present, which supports the findings of the
earlier analyses.
249
-------
H-
e
(D
•^J
CHEH.ENG.30-2SOU2/'MIN 6-15 81
TH=IS A.-n=3 IOM ov-iaix.
5025
7697
1-0
Ul
13
fD
I-i
w
X
H-
td
(D
O
I-!
(D
ft)
CTQ
hi
ft)
O
3
O
i-i
I
ON
O
14 21 28 35 42 49 56 63 70 77 34 91 93 105 112 119
-------
09
II
(U
00
UL SfiMPLE*ll STREXIT BKGD . SUB> *2043
33-485HMU 32-250«2-'MIN FRH5025
5826
Ln
rr
1-1
(D
l-i
H-
rt
0
O
to
PJ
6065
i
ON
O
14
;l 23 35 42 49 56 63 ?Q 77 84 91 93 195 112
-------
H-
OQ
i-j
(D
ro
Ln
ro
O
n
s"
en
o
3
O
n
o
I
i-{
fD
W
CO
O
o
o
I
rt
n
o
o
&.
(D
cn
UL COMPRESSOR KNOCKOUT SPLIT 206:1 OV-101 10M
33-48SflMU TH=? fl-D=3 -20-
5031
67517
n
3
I
12
I
16
I I T I 1
2& 24 23 32 36
M
I
I
40
I
44
I 1 I 1 I
43 52 56 60 64
-------
TABLE 7
AMI-57/GO-76 STRIPPER EXIT METHANOL
1.
2.
3.
4.
5.
6.
7.
8.
9.
10
11
12
13
14
15
16
17
18
19
20
sat'd hydrocarbon
co2
C,Hg isomer
tetramethylsilane
trichlorof luro-
me thane
C-H,n isomer
unknown
Freon 113
cyclopentadiene
C,H,2 isomer
C,H, , isomer
C/-H, n isomer
benzene
. CyH, , isomer
C-H, , isomer
C-,H, , isomer
r7u16 .
C-jH, 2 isomer
C-jH. „ isomer
7 12
C-H, 2 isomer
unknown
hydrocarbon
Peak
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
COMPRESSOR
Number from Figure 6
toluene 42. C3 alkyl benzene
methyl thiophene 43. C~ alkyl benzene
isomer 44. C, H00 isomer
10 JL2.
CoH-i f isomer 45. C,^H«« isomer
CgH,, isomer 46. C, alkyl benzene
C0E, , isomer 47. CinH00 isomer
o 1 o 1022
CgH,, isomer 48. CinH20 :"-somer
(trace) 49. unknown hydrocarbon
CgH14 isomer 50. C^IQ
(trace)
hexamethyl 51. C-Hg isomer
cyclotrisiloxane
CgH2Q isomer 52. alkyl benzene isomer
CqH,o isomer 53. C,1H24 isomer
ethyl benzene 54. ColL^D isomer
xylene (M,P) 55. ^i^-jU isomer
styrene 56. ^Q"inn isomer
xylene (0) 57. unknown siloxane
CgHjn isomer 58. unknown siloxane
CqELfl isomer 59. unknown siloxane
C, alkyl 60. C,,!!™ isomer
J« X H" J \J
benzene
CinH00 61. C./H-n isomer
10.22 14 30
isomer
unknown 62. unknown
hydrocarbon
unknown 63. C1CH00 isomer
1 j j 2.
hydrocarbon
^11^24 i-somer
TABLE 8
KNOCKOUT SAMPLE FROM AMI-60/GO-79
PEAK NUMBER FROM FIGURE 9
1 . 1-pentene
2. hydrocarbon
3 . benzene
4. hydrocarbon
5. Toluene
6. cyclo C4-C5
7. hydrocarbon
8. ethyl benzene
9. dimethyl benzene
10. substituted benzene
11. Co hydrocarbon
12. C_ hydrocarbon
13. propyl or ethyl methyl substituted benzene
14. propyl or ethyl methyl substituted benzene
15. 1-decene
16. 2-propyl benzene
17. l-ethyl-4-methyl benzene
253
-------
Results from these analyses indicate that very little, if any,
polynuclear aromatic compounds were present in the gas fed to the AGRS. This
is a particularly important finding. Analyses of the water used to quench
the gasifier product gas stream showed that a substantial amount of
polynuclear aromatics were present. Evidently, scrubbing of the raw product
gas with water effectively removes these compounds.
Although polynuclear aromatics are removed by the quenching process,
substantial amounts of simpler aromatics will be present in the sour gas fed
to the AGRS. The use of cold traps may remove some of these compounds but
provision must be made to prevent their release to the atmosphere through
vent streams or through the sulfur recovery unit. The accumulation of these
compounds in the methanol further complicates the problem because of the
increased likelihood of their distribution to a number of process streams.
Achievine efficient solvent regeneration is, therefore, a key step in
avoiding environmental problems.
SUMMARY
A cyclone, a cold water quench scrubber, and a refrigerated methanol
absorber have been used to clean the make gas from the steam-oxygen
gasification of a New Mexico subbituminous coal in a pilot-scale fluidized
bed ractor. A model developed for the gasifier provides the capability of
predicting the make gas amount and composition as a function of gasifier
operating conditions. The methanol functioned effectively for acid gas
removal. Removal of C02, COS, and H2S to sufficiently low levels was
achieved with proper choice of operating conditions and effective solvent
regeneration.
The presence of several trace sulfur compounds—mercaptans, thiophenes,
organic sulfides, and CS~—complicates the gas cleaning process because these
compounds were found to distribute among all exit streams from the AGRS.
Since no provision is made to specifically treat these forms of sulfur, the
possibility of their emission into the atmosphere exists and must be dealt
with to avoid significant environmental problems.
A wide variety of aliphatic and aromatic hydrocarbons are present in the
gas stream fed to the AGRS. The aliphatic hydrocarbons, ranging from methane
to butane, cover a wide range of solubilities. Their presence in all AGRS
streams must be anticipated to prevent their emission to the atmosphere.
While a wide range of simple aromatics were identified in the gas stream
fed to the AGRS, essentially no polynuclear aromatic compounds were found.
Apparently, the water quenching process effectively removes these compounds
from the gasifier product gas. However, significant quantities of simple
aromatics were found to accumulate in the recirculating methanol, indicating
a potential for their eventual discharge to the atmosphere. Provision must
be made to periodically purge the solvent of these compounds and/or remove
them prior to the AGRS through cold traps.
254
-------
REFERENCES
1. Ferrell, J. K., R. M. Felder, R. W. Rousseau, J. C. McCue, R.
M. Kelly, and W. E. Willis, "Coal Gasification/Gas Cleanup Test
Facility: Vol I. Description and Operation", EPA-600/7-80-046a,
(1980).
2. Felder, R. M., R. M. Kelly, J. K. Ferrell, and R. W.
Rousseau, "How Clean Gas is Made from Coal", Env. Science and
Tech., Vol 14, 658, (1980).
3. Ferrell, J. K., , R. M. Felder, R. W. Rousseau, S. Ganesan, R.
M. Kelly, J. C. McCue, and M. J. Purdy, "Coal Gasification/Gas
Cleanup Test Facility: Vol II. Environmental Assesment of
Operation with Devolatilized Bituminous Coal and Chilled Methanol",
EPA, (1981).
255
-------
ADVANCED TECHNIQUES FOR FLUE GAS DESULFURIZATIONf
by: Charles C. Masser, Theodore G. Brna,
and Michael A. Maxwell
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
In 1979 the combustion of sulfur-bearing fuels accounted for
more than 80 percent of the SC>2 emissions in the United States. These
emissions can be controlled to a degree by burning low-sulfur fuels
or by pretreating the fuel to lower its sulfur content. Currently the
most widely-practiced technological control involves scrubbing the
combustion flue gases to remove the SC^. Flue gas desulfurization
systems can be categorized as nonregenerable or "throwaway" and
regenerable or producing a saleable product. Several systems in
each category will be discussed as to their advantages and disad-
vantages. In addition, several recent developments regarding waste
disposal and enhanced SC^ removal will be presented.
256
-------
INTRODUCTION
Sulfur dioxide (S02) is one of a number of sulfur-containing pollutants
found in the atmosphere. It enters the air primarily from the combustion
of coal and oil, but also from various other industrial processes. The
combustion of sulfur-containing fuels accounts for more than 80 percent
of the S02 emissions in the United States^1>2/. These emissions can be
controlled to a degree by burning low-sulfur fuels or by pretreating the
fuel to lower its sulfur content. Currently the most effective control
involves scrubbing the combustion flue gases to remove S02 by flue gas
desulfurization (FGD) technology.
This paper briefly discusses methods for controlling these emissions,
related waste disposal, and process costs. Since most FGD technology
has been developed in relation to coal-fired steam electric generating
plants, the major emphasis will be advanced FGD systems for these plants.
Such systems, however, are being used at industrial sites and may be
adaptable to commercial gasification facilities. Sulfur oxides (SOX)
are pollutants of major environmental concern, and their formation in
power generation units of commercial medium- and high-Btu gasification
plants may be at levels requiring control. Several proposed gasification
plants include FGD systems on coal-fired power generation units for
these plants.
While SOX are not the major pollutant in raw product gases from coal
gasifiers, commercial processes for cleaning these gases or upgrading
the quality of the synthetic gas produced may produce SOX at levels
requiring control to meet air quality constraints. For example, the
production of sulfur from sulfur compounds in raw product gases can
lead to tail gases containing reduced levels of sulfur compounds and
other combustible gases. Combustion of the tail gases along with coal
in the power-producing component of a medium- or high-Btu gasification
plant may generate SOX at levels requiring FGD, but would reduce fuel
requirements through combustion of the tail gases. Thus the integrated
control of SOX in flue gases from the incineration of pollutant-bearing
tail gases and the combustion of coal in connection with power or steam
production may be more cost effective than treating separate pollutant
streams.
FGD systems are classified into two categories: nonregenerable or "throw-
away'' systems and regenerable systems which produce a saleable product.
They may be further classified into wet and dry FGD systems, the distinction
being that saturated (with water) and unsaturated flue gas, respectively,
result from the gas cleaning process. A brief discussion of these
categories of advanced FGD systems will now be presented.
257
-------
NONREGENERABLE FGD SYSTEMS
Presently, nonregenerable FGD systems can be classified into two types,
wet and dry. Each type of process will be discussed.
Most commercial wet FGD systems, that are either operating or planned for
use in utility applications, are lime or limestone based systems. The
major driving force for using these "throwaway" systems rather than
regenerable FGD technology is one of economics. Wet limestone systems
are slightly more economical than wet lime systems because of the cost
and energy requirements associated with calcination of the limestone to
produce lime. This cost difference is expected to increase with rising
energy costs. Although these calcium-based systems are in wide use,
their performance to date has been limited by reagent reactivity which
results in low soluble alkalinity, relatively higher liquid-to-gas (L/G)
ratio requirements, and larger reaction tanks than other FGD processes.
Wet Lime/Limestone^^
Wet lime/limestone FGD processes (Figure 1) employ a scrubbing slurry of
lime or limestone to remove SC>2 • As a side benefit, these processes can
also be designed to remove fly ash and chlorides simultaneously. Because
lime/limestone processes are nonregenerable, they produce large quantities
of waste solids. This characteristic could place them at a disadvantage
compared with regenerable processes where disposal costs are high.
Regenerable processes, however, still require disposal of waste fly ash
and chlorides by environmentally acceptable methods, and these waste
products can amount to more than 50 percent (high ash fuels) of the
volume of solid waste produced by lime/limestone processes.
Lime/limestone systems are usually less complex than regenerable systems,
and they generally cost less to install and operate than other wet FGD
processes. Consequently, lime/limestone FGD processes are the most
widely used wet FGD systems in operation.
Lime/limestone FGD processes consist of four steps:
1. Feed material processing.
2. Absorption.
3. Solids precipitation.
4. Solids concentration and disposal.
Flue gas enters the absorber (Figure 1) where it contacts the circulating
scrubbing slurry containing calcium ions from dissolved lime or limestone.
SC>2, fly ash, and chlorides contained in the flue gas are removed by the
258
-------
Figure 1. Typical Lime/Lime stone FGD Process
259
-------
circulating slurry. Alkaline species in the liquor neutralize the
absorbed S02, promoting the formation of ions of sulfite and sulfate.
Water droplets are removed from the cleaned flue gas by mist eliminators
as they leave the absorber. The clean, water-saturated, flue gas is
reheated, if necessary, to counter material corrosion and/or plume
dispersion problems and then is exhausted through the stack to the
atmosphere.
The scrubbing slurry, which may be supersaturated with solids of calcium
sulfite and calcium sulfate, flows to an effluent holding tank or pre-
cipitation vessel. In the holding tank, fresh makeup lime or limestone
is added, and reaction products are precipitated. One effluent stream
from the holding tank is recycled to the absorber; another is bled off
for concentration and disposal of waste solids.
Solids in the bleed stream may be concentrated in a thickener, filter,
or centrifuge, or may be sent directly to a holding/settling pond.
Clarified process water is returned to the system. Concentrated solids
may be disposed of in ponds or used for landfill and may or may not be
chemically stabilized. It is also possible to convert the solids to
gypsum for use in portland cement or wallboard manufacture. This,
however, requires the additional process operations of oxidation and
purification.
Wet Lime/Limestone with Adipic Acid Enhancement^'
IERL-RTP has sponsored extensive laboratory, pilot, prototype, and
commercial scale studies on an adipic acid enhanced wet limestone system
which shows significant improvement in operation over non-adipic acid
enhanced wet limestone systems. Addition of relatively small quantities
of adipic acid (approximately 1500 ppm), in either conventional or
forced oxidation limestone FGD systems, provides these important benefits:
• Significantly enhanced S02 removal efficiency in
either conventional or forced oxidation modes (compared
with additives such as MgO which may be of little
benefit in forced oxidation systems).
• Increased limestone utilization; hence, decreased
waste solids disposal requirements and improved
scrubber reliability.
• Lower projected capital and operating costs than
conventional limestone FGD systems.
• Not adversely affected by chloride as is the lime-
stone/MgO process; thus, it is especially attractive
for closed-loop operation.
• Less expensive and less energy intensive limestone
rather than lime is used.
260
-------
Adipic acid is a weak dicarboxylic acid which buffers the slurry pH and
thus enhances the SC>2 solubility and limestone dissolution rate. An
important advantage of adipic acid as compared with other scrubber
additives is that its ability to improve S02 removal is not affected by
chlorides. Adipic acid significantly enhances SC>2 removal over a range
of operating conditions (with scrubber slurry pH's of 5.0 to 5.5).
Operation at the lower end of the typical pH range increases limestone
utilization and may reduce scaling and mist eliminator fouling. Improved
limestone utilization in turn reduces both the amount of limestone
required and the quantity of solid wastes produced by the FGD system.
These improvements represent an estimated 6 percent reduction in capital
investment and 7 percent reduction in operating costs.
Figure 2 shows enhanced S02 removal due to adipic acid addition. For
example, a scrubber feed with a pH of 5.5 and an adipic acid concentration
of approximately 1600 ppm resulted in 95 percent S02 removal, as opposed
to 70 percent S02 removal at the same operating conditions without
adipic acid. Adipic acid addition also improved limestone utilization,
increasing it to over 90 percent.
Research and development activities have focused on adipic acid to
enhance the performance of the limestone FGD process. However, other
organic acids will also enhance the process. One of the most intriguing
alternatives is dibasic acid (DBA) material which is a by-product of
the adipic acid manufacturing process. This material has been tested at
lERL-RTP's pilot plant and at a full scale scrubber installation. The
results show that the DBA material enhances the performance of the
scrubbers similar to pure adipic acid. Since the DBA material is projected
to cost only one-third to half the cost of adipic acid, this alternative
is particularly attractive.
The testing to date has found that the adipic acid or DBA additive
reduces energy requirements of the process over conventional limestone
scrubbing processes. This results primarily from operating at a lower
L/G and from reducing the solid waste handling requirements. The environ-
mental impact of the enhanced process is about the same as conventional
limestone scrubbing, except that the solid waste loading is less from
the enhanced process. There is no significant difference in the toxicity
of the wastes from the two processes.
Wet Lime/Limestone with Forced Oxidation(5)
A major advancement in the wet lime/limestone scrubbing process is the
stabilization of the waste material by forced oxidation (e.g., air
sparging into slurry hold tank). In the past, a disadvantage of lime/
limestone scrubbing processes has been the large volume of waste solids
produced. This waste slurry, consisting of predominantly calcium
sulfite, could only be dewatered to about 50 to 60 percent solids, thus
261
-------
NO
o>
ho
100
90
80
fa
PM
W
I 70
«M
•o
CO
60
50
pH5.5
500 1000
ADIPIC ACID CONCENTRATION, ppm
1500
2000
Figure 2. Effect of Adipic Acid Concentration and pH on S02 Removal Efficiency
in Limestone Wet
-------
producing a material which may be unsuitable for landfill. In the past,
the primary utility practice involved the use of lined ponds for con-
tainment in order to prevent contamination of ground and surface waters.
Although less expensive than other disposal options compatible with
landfill, such as chemical fixation and fly ash blending, ponding of
this material represented as much as 20 to 25 percent of the overall
scrubbing process costs. Furthermore, the large land areas required for
these disposal ponds were difficult to reclaim for other productive use
due to the poor mechanical stability and load bearing strength of the
waste material. One solution to the situation is the forced oxidation
of the calcium sulfite produced to calcium sulfate (gypsum), a material
easily dewatered to greater than 80 percent solids. Since, in the
United States, by-product gypsum may be unable to compete with the
widely available natural gypsum, the incentive has been to develop
simplified low-cost forced-oxidation procedures directed primarily
toward improving waste solids handling and disposal properties while
minimizing waste disposal costs.
Wet ponding, landfilling, and mine disposal are three current means of
disposal. Ponds can be designed based on diking or excavation and can
even be engineered on slopes. A special case of wet ponding is FGD
gypsum stacking. Gypsum slurry from the forced oxidation system is
piped to a pond and allowed to settle, and the supernate recycled.
Periodically the gypsum is dredged and stacked around the embankment.
For disposal in a landfill, dewatered wastes are transmitted to the
disposal site where they are spread on the ground to a thickness of
about 0.3 to 1 meter. Compaction by heavy equipment follows, and a
layering process proceeds at the site. A disposal method that is
receiving increased attention is mine disposal, particularly in the
West. Surface coal mines are the most likely candidates for FGD waste
disposal. Coal mines offer the greatest capacity for disposal, and they
frequently have direct transportation (e.g., rail) connections tied to
power plants. In fact, many new coal-fired power plants are '"mine-
mouth" (located within a few kilometers of the mine), and the mine
provides a dedicated coal supply. Since the amount of FGD wastes
produced is considerably less than the amount of coal burned, such mines
usually would have the capacity for disposal throughout the life of the
power plant.
Surface mines have basically three options for the disposal of FGD
wastes:
• In the working pit, following coal extraction, and
prior to return of overburden.
• In the spoil banks, after return of overburden, but
prior to reclamation.
• Mixed with, or "sandwiched" between, layers of overburden.
263
-------
The latter two options appear to be more environmentally sound and are
expected to show strong growth in the future.
Dual Alkali(6)
The dual alkali FGD process consists of four basic steps:
1. Flue gas pretreatment (optional).
2. SC>2 absorption.
3. Absorbent regeneration.
4. Solid/liquid separation and solids dewatering.
Figure 3 illustrates the process flow for a typical dual alkali FGD
system.
During pretreatment, flue gas from the boiler can be routed through an
electrostatic precipitator (ESP) to remove particles (fly ash) upstream
of the absorber. Pretreatment can also involve wet scrubbing, alone or
in series with the ESP, for particle and chloride removal. Pretreatment
is not always necessary in dual alkali FGD; its use depends on site-
specific conditions such as fuel characteristics and cost considerations.
The flue gas then flows to an absorber and is brought in contact with a
recirculating solution containing an equilibrium mixture of sodium
sulfite (Na2S03), sodium bisulfite (NaHS03), sodium hydroxide (NaOH),
sodium carbonate (Na2CC>3), and sodium bicarbonate (NaHC03). SC>2 is
absorbed by this solution and reacts with the alkali in solution to form
soluble sulfur salts.
Desulfurized flue gas leaves the absorber, is reheated if necessary,
and is exhausted through the stack to the atmosphere. A portion of the
circulating absorbent solution is routed to the absorbent regeneration
system to be reacted with lime, to precipitate the absorbed SOX as:
• Calcium sulfite hemihydrate (CaSC>3' 1/2H20) -
• Gypsum (CaSC^-Zl^O) (only in dilute dual alkali systems).
• A mixed crystal of hydrated calcium sulfite/sulfate.
The precipitation reaction also regenerates soluble alkali for recycle
to the absorber.
264
-------
Wash water
Makeup water
Figure 3. Dual Alkali FGD Process
(6)
265
-------
The precipitated SOX salts are separated from the scrubbing liquor and
concentrated for disposal in the solid/liquid separation and solids
dewatering steps. The solids settle out of the slurry in a clarifier-
thickener; they are dewatered further in a vacuum filter or centrifuge
and are washed to recover sodium salts before disposal. The clear
liquor overflow from the clarifier-thickener is combined with makeup
soda ash solution and returned to the absorption system.
Spray Drying(7)
Nonregenerable spray drying processes are presently the only commercially
applied dry FGD processes in the United States. Other dry FGD processes
under development include dry injection and combustion of coal/alkali-
fuel mixtures. Several factors, including increases in coal use and the
1979 new source performance standards (NSPS) for utility boilers, have
promoted increased research and development and commercial application
of the dry FGD technology.
Interest in spray drying FGD has primarily been spurred by the potential
cost savings dry FGD offers over conventional wet FGD, particularly for
low-sulfur coal (less than 1.5 percent sulfur) applications. In addition
to the production of a dry waste, advantages of spray drying FGD over
wet FGD systems include potentially lower initial capital investment,
lower operating costs for up to moderate fuel sulfur content (possibly
3 percent), and less process complexity, which may lead to greater
system reliability.
The major disadvantage of spray drying FGD relative to wet FGD systems
is the higher absorbent cost, which results from the higher priced
absorbent (lime versus limestone) and the higher stoichiometric ratios
necessary. The applicability of spray drying FGD for high-sulfur coal-
fired installations is limited by the lack of data on the 862 removal
capability and the higher costs of this technology.
The spray drying FGD process (Figure 4) consists of three steps:
1. Absorbent preparation.
2. S02 absorption drying.
3. Solids collection.
For economic comparison of wet and dry FGD systems, waste disposal cost
should also be included. In this regard, waste disposal would be the
last step of each of the process steps for FGD systems.
266
-------
c
m
4>
XI
i-i
O
i-!
>Flue gas
CW
o
o
O
n
ro
en
en
1
1
1
Combustion !
air '
preheat er
V
/'
)t or wara\
;as bypass^"
1
1
1
.^
F
. fci
1 "
Dilution «w I
Spray dryer
water~~^M
S\^
/ N
Absorbent
feed
'
*^^
•
w^ Bv
V V
\ Cleaned
/ flue gas
/ f
Solids
collection
device
k
W
N^l
Collected
solids
Fresh
absorbent
holding tank
Recycle
solids
slurry tank
4*
_,\ «««te
/ solids
To disposal
-------
Flue gas exiting the combustion air preheater is contacted with an
alkaline solution or slurry in a spray dryer. In the spray dryer, the
flue gas passes through a contacting chamber, and the solution or slurry
is sprayed into the chamber with a rotary or nozzle atomizer. The heat
of the flue gas dries the atomized droplets, while the droplets absorb
SC>2 from the flue gas. The 502 reacts with the alkaline reagent to
form solid-phase sulfite and sulfate salts.
Most of these solids (and any fly ash present) are carried from the
dryer in the exiting flue gas. The rest fall to a hopper at the bottom
of the dryer. In contrast to wet FGD systems, the flue gas is not
saturated with moisture after the absorption step but is within 11 to
28°C (20 to 50°F) of the saturation temperature^8).
The solution or slurry is pumped to the dryer from an absorbent holding
tank. Fresh absorbent and dilution water are added to this tank as
needed. (In some systems, dilution water for temperature control in the
spray dryer is added to the absorbent feed just upstream of the spray
dryer.) Recycle solids from the spray dryer hopper or downstream solids
collection equipment contain unreacted absorbent and may be used to
supplement the fresh absorbent feed. Recycle solids are either slurried
separately and added to the absorbent feed just upstream of the spray
dryer or are added directly to the fresh absorbent holding tank.
Flue gas may be reheated after it leaves the spray dryer to prevent
condensation in downstream solids collection equipment. Reheating may
be desirable and economically justifiable: SC>2 removal in the spray
dryer is greatly improved as the flue gas temperature approaches the
saturation temperature of the gas at the spray dryer exit. Reheat may
be accomplished by mixing the flue gas from the spray dryer with either
hot flue gas from upstream of the combustion air preheater or warm flue
gas from upstream of the spray dryer. Other methods of reheat could be
used such as heating air and injecting it into the cleaned flue gas,
heating part of the cleaned flue gas and re-injecting it into the
remainder of the gas stream, or heating all the treated flue gas in a
surface heat exchanger upstream of the particle collection device.
Also note that using dirty flue gas (either warm or hot) for reheating
means that higher S02 removal is required of the spray dryer to meet
given emission limits. The reheated flue gas then flows to the solids
collection device where the dry solids (which consist of reaction
products, unreacted absorbent, and fly ash) are collected. A fabric
filter (baghouse) is the most common solids collection device, but
electrostatic precipitators (ESPs) are also used. When a baghouse is
used, significant absorption of S02 may occur during the solids collection
step. Absorbent in the solids collected on the surface of the bags
reacts with S02 remaining in the flue gas, and the desulfurized flue gas
is exhausted to the atmosphere through a stack.
268
-------
While sodium compounds may serve as absorbents, most nonregenerable
spray dryer FGD applications use lime because of its lower cost and the
more stable wastes produced through its use'9). Since dry waste solids
are generated in dry FGD processes, their disposal is typically by
landfill. With sodium compounds (e.g., soda ash) as absorbents, the
high water solubility of the resulting sodium salt wastes could require
expensive lined landfills to control leaching into the ground water.
REGENERABLE FGD SYSTEMS
There are two primary regenerable wet FGD systems currently available
for commercial applications: Magnesium Oxide and Wellman-Lord. Two
other processes, the aqueous carbonate and citrate processes, are under
development. These systems are capable of removing 90 to 95 percent of
the flue gas
The principal advantages of regenerable FGD systems over nonregenerable
systems are the economic advantages gained from the reduction of waste
disposal problems and the sale of recovered by-products. Wastewater
streams are collected and can be neutralized by standard treatment
systems, and most of the spent solution can be recirculated to the
process. Solid waste loads are also considerably reduced. On the other
hand, regenerable systems are more complex than current nonregenerable
systems and generally involve higher capital investment and higher
operating costs.
Wellman-Lordd1)
The Wellman-Lord process consists of four basic steps:
1. Flue gas pretreatment.
2. S02 absorption.
3. Purge treatment.
4. Sodium sulfite regeneration.
A fifth step, the processing of S02 into by-product sulfur, is not a
part of the Wellman-Lord process but is generally associated with Wellman-
Lord installations. Figure 5 illustrates the process flow for a typical
Wellman-Lord system installed on a coal- or oil-fired boiler.
Boiler flue gas is pretreated by contact with water, usually in a venturi
scrubber. This step cools and saturates the gas, absorbs corrosive
chlorides, and removes some of the particles remaining in the gas after
upstream particle removal efforts.
-------
•«2S04
cake to
dicpotal
or tale
SuHuror
•oHuncaod
Figure 5. Typical Wellman-Lord Process^ ^
270
-------
The flue gas then flows to an absorber where it is contacted with a
sodium sulfite (^2803) solution. The SC>2 in the flue gas reacts with
the N32S03 to produce sodium bisulfite (NaHS03). In a side reaction,
some sodium sulfate (Na2SC>4) is formed by direct oxidation of Na2S03.
Desulfurized flue gas leaves the absorber, is reheated to improve plume
buoyancy and to vaporize any liquid droplets present, if necessary, and
is exhausted through the stack to the atmosphere. If reheat is not
used, then protective linings in stacks and acid-corrosion-resistant
material in ducts are generally used in wet FGD applications. The
effluent from the absorption tower, rich in NaHSC>3 and also containing
some Na2SC>3 and Na2SC>4, is split into two streams. Approximately
15 percent of the effluent is routed to a purge treatment for sulfate
removal. The remaining 85 percent goes to a regeneration process.
The purge stream is cooled in a chiller and a mixture of Na2SC>4 and
N32S03 is crystallized out of the solution. This crystalline mixture is
removed from the process and dried for sale or disposal.
Regeneration is accomplished in an evaporator where the remainder of the
SOX absorber effluent is heated to convert NaHSC>3 to Na2SC>3 and to drive
off SC>2 . The regenerated Na2S03 crystallizes and then is redissolved
and recycled to the absorber. Sodium lost during the process, primarily
from the purge operation, is replenished by adding sodium carbonate
to the feed dissolving tank.
The fifth step, SC>2 processing, uses the S02 by-product from the Wellman-
Lord process. The output of the Wellman-Lord process is a gas stream of
about 85 percent 502', the remainder is mostly water vapor. This concen-
trated SC>2 stream may be dried and marketed without further processing,
reduced to elemental sulfur, or oxidized and reacted with water to form
sulfuric acid
Magnesium Oxide ( 12)
The magnesium oxide (MgO or Mag-Ox) FGD process consists of four major
processing steps:
1. Flue gas pretreatment.
2. SC>2 absorption.
3. Solids separation and drying.
4. Regeneration.
SC>2 processing may be considered a fifth step because it is often
associated with the MgO FGD process. Figure 6 illustrates the process
flow for a typical MgO FGD system.
271
-------
Recycled
absorber
liquor
Solids
separation
and drying
Sulfurlc acid
or sulfur
Figure 6. Magnesium Oxide FGD Process with Regeneration
and SC> Conversion(12)
272
-------
In the first step, water scrubbing cools and saturates the boiler flue
gas and removes fly ash and chlorides upstream of the absorber. While
flue gas from oil-fired boilers generally does not require pretreatment
by quenching, this step is necessary in coal-fired applications.
In the absorber (Step 2), S(>2 is removed from the flue gas by contact
with a recirculating slurry of magnesium oxide (MgO), magnesium sulfite
(MgS03), and magnesium sulfate (MgS04). Flue gas SC>2 is absorbed by
this slurry and reacts with MgO to form MgSC>3, some of which reacts with
oxygen (02) present in the flue gas to form MgS04. Additional MgS04 is
formed when flue gas sulfur trioxide (803) reacts with MgO.
Desulfurized flue gas leaves the absorber, is reheated if necessary, and
is exhausted through the stack. The scrubbing liquor is continuously
recycled to the absorber after a continuous bleed stream has been with-
drawn from the recirculation loop for solids separation and regeneration
processing. Fresh MgO slurry is added to the recirculation loop to
replace the scrubbing liquor removed by the bleed stream.
In the third step, the bleed stream is routed to a centrifuge where it
is concentrated to 60 percent solids by weight, and the mother liquor is
recycled to the absorber recirculation loop. The concentrated solids
flow to a dryer where surface moisture and most of the water of hydration
are removed, producing a dry powder of MgS03, MgS04, unreacted MgO, and
inert materials.
Calcination of the dry powder in the regeneration processing stage
(Step 4) converts MgS03 and MgS04 to MgO, which is recycled to the
absorber recirculation loop. MgS04 is reduced with coke during the
calcination process. Calcination also produces an S02~rich by-product
stream that may be processed further to form sulfuric acid or elemental
sulfur.
Thus, the MgO FGD process not only regenerates the essential absorbent,
MgO, but also produces S02 at concentrations practical for conversion to
sulfuric acid or elemental sulfur.
.Aqueous Carbonate^^)
The aqueous carbonate process can be divided into four major operations:
flue gas handling, S02 absorption and product collection, absorbent
regeneration, and sulfur production. Figure 7 is a process flow diagram
of a typical aqueous carbonate system.
Flue gas is contacted with sodium carbonate solution in a spray dryer,
and S02 is absorbed. The solid sodium sulfite and sodium sulfate formed
is collected in cyclones and an ESP. The flue gas, which is only partially
quenched, is emitted to the atmosphere without reheating. Overall
reactions in the spray dryer are:
273
-------
ID fan
Clean gas
to stack
•» ft*
r
Partlculate
revoval
Incinerator
:
"w
0
1
a '
S
•H i
1|
8 1
CN
Sulfur
production
(Claue Plant)
i
i —
H2S rich
1
r
S02
scrubbing
(spray dryer)
H20
__-
f
i
p
Carbonatlon 1
Ha2S rich
solution
Gas plus
entrained solids
Reducer off-gas
H20
f
Quench &
filtration
1 i
Product
(cyclone or ESP)
~~f t
-? c*>
O O
CO (J
y *
o Air
^
f
Reduction
Coke
Elemental
sulfur
Ash & Coke
Figure 7. Aqueous Carbonate FGD Process
(13)
-------
S02 + Na2C03 -> Na2S03 + C02 7 (D
Na2S03 + 1/2 02 -> Na2S04 (2)
In the regeneration area, spent absorbent is melted, mixed with coal or
petroleum coke, and sparged with air in the reducer vessel. The following
reactions occur:
Na2S03 + 2 C + 1/2 02 -> Na2S + 2 C02 (3)
3 C + 02 -> Na2S + 3 C02 (4)
A portion of the smelt is continuously withdrawn to a quench tank and
dissolved in water. The "green liquor" thus produced is clarified,
filtered, and contacted with the cooled reducer off-gas in a series of
tray towers to regenerate sodium carbonate and evolve hydrogen sulfide
by the following reactions:
Na2S + 2 C02 + 2 H20 -> H2S + 2 NaHCC>3 (5)
2 NaHC03 -> Na2CC>3 + C02 + H20 (6)
The regenerated liquor is filtered and recycled to the spray dryer/absorber.
Solids, mainly composed of ash and miscellaneous impurities, from the
green liquor filters and from the regenerated liquor filters are disposed
of.
Sulfur production involves the conversion of H2S to elemental sulfur in
a three-stage Glaus unit. The tail gas from the unit is incinerated and
recycled to the spray dryer/absorber.
Currently no dry FGD systems use regeneration, although those using
Na2C03 in spray dryers duplicate the absorption step of the aqueous
carbonate process.
ECONOMIC COMPARISONS
Cost estimates for most wet FGD processes are readily obtained from the
wealth of design and operating experience for these processes. Cost
projections for the only dry FGD (spray dryer) process yet commercialized
are based on pilot- and demonstration-scale tests and vendor estimates.
As TVA has performed comprehensive studies for EPA on the economics of
FGD, costs from these studies are the bases for the comparisons reported
here.
275
-------
The state-of-the-art wet FGD system is considered to be the non-
regenerable limestone spray tower process. Other wet FGD systems are
compared with this system, the costs for each system having been deter-
mined using the same design and economic premises'!^ »15,16)t Since the
costs of the processes ranked in Table 1 are being updated to conform to
the 1979 NSPS for utility boilers and revised design and economic
premises, no numerical values are listed. Table 1 ranks these processes
in order from lowest to highest costs.
A comparison of dry FGD (spray dryer) costs with wet limestone FGD
(spray tower) costs is given in Table 2^9). xhe costs include particulate
matter removal and waste disposal because particulate matter removal is
an inherent part of the dry FGD process. While wet FGD costs are founded
on extensive design and operating experience, only limited pilot- and
demonstration-scale data are available for dry FGD. Since the same
estimation basis and methods were used for each process evaluated, the
accuracy for comparison is reasonable for preliminary studies.
Table 2 shows that the capital investment and first year revenue require-
ments for dry FGD are less than those for wet FGD for all cases compared.
For high sulfur coal (3.5 percent sulfur), however, the lifetime cost
for wet limestone FGD is about 2 percent less than for lime spray
drying. While the capital investment advantage for the lime spray dryer
over the wet limestone process ranges from about 14 to 30 percent, the
annual revenue requirement advantage for the spray dryer over the wet
limestone process falls from about 28 to 2 percent for increasing sulfur
content of the eastern coals. This is attributed to the higher unit
cost and the higher stoichiometric ratio for the lime system relative to
the limestone system. By increasing the sulfur content of the coal from
0.7 to 3.5 percent, the absorbent costs increase about 10-fold and
represent about 27 percent of the first year revenue requirements for
the lime spray dryer, while corresponding values for the wet limestone
FGD system are 7-fold and 3 percent.
Table 2 also indicates that the lime spray dryer would be the economic
choice over soda ash spray drying for low-sulfur western coal. Because
of the higher unit cost of soda ash relative to lime, the expected
sources for soda ash being in the West, and high disposal costs for
sodium salt wastes, the economic advantage of the lime over the soda
ash spray dryer is expected to be even greater at eastern sites.
As noted earlier, the addition of adipic acid to the wet limestone FGD
process improves both S02 removal and limestone utilization. Both of
these improvements lead to decreased lifetime cost, and the quantification
of their effects is underway. Limited pilot plant testing also showed
adipic acid to improve both S02 removal and sorbent utilization when
either lime or limestone are used in spray drying with partial recycling
of waste solids(17). However, the low reactivity of limestone at the
usual flue gas conditions in spray drying apparently limits S02 removal
276
-------
TABLE 1. COMPARATIVE ECONOMICS OF ALTERNATE WET FGD PROCESSES IN
ORDER OF INCREASING COSTS<14>15>16)
Lifetime
Cost
Limestone
Dual alkali
Lime
Aqueous carbonate
Magnesium oxide
WeiIman-Lord/acid
Wellman-Lord/RESOX
Citrate
Capital
Investment
Lime
Limestone
Dual alkali
Aqueous carbonate
Wellman-Lord/acid
Magnesium oxide
Wellman-Lord/RESOX
Citrate
First Year
Annual Revenue
Requirements
Limestone
Dual alkali
Lime
Aqueous carbonate
Magnesium oxide
Wellman-Lord/acid
Wellman-Lord/RESOX
Citrate
Waste or
By-product
Waste
Waste
Waste
By-product
By-product
By-product
By-product
By-product
NOTES: 1. Credit is taken for the sale of by-products for the last
five processes listed.
2. Particulate matter removal and waste disposal costs are
not included for any process listed.
3. Design and economic premises are: new 500-MWe midwestern
plant firing eastern bituminous coal (3.5% S, 12% ash,
5833 kcal/kg (10,500 Btu/lb), 30-year plant life, and
S02 emissions meeting 1971 NSPS).
277
-------
TABLE 2. COMPARATIVE ECONOMICS FOR DRY (SPRAY DRYER) AND WET LIMESTONE
FGD PROCESS FOR SEVERAL COALS AND SULFUR CONTENTS^7»9)
Process
Lime spray dryer
Limestone wet FGD
Soda ash spray
dryer
Lime spray dryer
Limestone wet FGD
Fuel
Lignite
0.9% S
Subbituminous
(western) coal
0.7% S
Lifetime
Cost, $106
860.8
1069.5
844.4
774.7
885.5
Capital
Investment
$106
82.6
107.4
79.4
77.1
88.1
First Year
Annual Revenue
Requirements, $lp6
20.9
26.3
20.4
19.0
21.7
Lime spray dryer
Limestone wet FGD
Bituminous 757.1
(eastern) coal 936.4
0.7% S
75.3
92.6
18.6
23.9
Lime spray dryer
Limestone wet FGD
Bituminous 1413.3
(eastern) coal 1355.8
3.5% S
100.1
121.9
31.9
32.4
BASIS: A new 500 MWe plant is assumed to be located in Wyoming,
Colorado, Nebraska, North Dakota, or South Dakota when lignite
or low sulfur western coal is the fuel. A midwestern plant
site (Kentucky, Illinois, or Indiana) is used for the eastern
coal estimates. TVA design and economic premises were applied
with capital investment expressed in 1982 dollars. Investment
costs include those for control of S02 emissions and disposal
of scrubber waste via landfilling 1.6 km (1 mi) from the plant
site. The plant has an operating life of 165,000 hours over a
30-year period (equivalent to full load over 5500 hr/yr). The
boiler heat rate is 2394 kcal/kWh (9500 Btu/kWh) for coal and
2948 kcal/kWh (11700 Btu/kWh) for lignite. Revenue requirements
are in 1984 dollars, while the total evaluated cost is based
on a fixed charge rate of 14.7 percent and a levelized operation
and maintenance factor of 1.886, which account for inflation
and the cost of money over the plant life.
FUEL DATA:
Lignite
Western Coal
Eastern Coal
Eastern Coal
Heating
kcal/kg
3667
5390
6501
6501
As Fired
Value Ash
Btu/lb %
6600 7.2
9700 9.7
11700 15.1
11700 15.1
Moisture
%
36.3
16.0
4.0
4.0
Dry Basis
Sulfur
%
0.9
0.7
0.7
3.5
278
-------
to about 30 to 35 percent. While the apparent improved performance
using adipic acid in the wet FGD system may exceed that in the spray
dryer, the impact this additive has on lifetime cost may be greater for
the dry FGD system because of the relative effect of sorbent cost on
revenue requirements (as noted earlier). Consequently, further work is
needed to definitize the cost effect of adipic acid additive on both of
these FGD processes.
SUMMARY
Wet FGD processes can effectively and reliably control S02 emissions
from coal-fired boilers. Among the available wet processes, the non-
regenerable (throwaway) wet limestone system predominates the power
plant applications because of its cost advantage. A recent improvement
to the wet lime stone/lime FGD process is the use of forced oxidation to
produce a more suitable waste product (gypsum) for landfill disposal.
Currently, the addition of adipic acid to limestone is demonstrating
improved SC>2 removal (consistently above 90 percent) and sorbent utiliza-
tion (over 95 percent) and appears to offer significant performance
improvement and cost savings.
Dry FGD has recently emerged as a potentially more economical and reliable
option for low-to-moderate-sulfur coal applications. Its viability
remains to be demonstrated in full-scale applications: the first utility
system is slated for operation this year. The lime-based spray dryer
appears to offer capital investment savings due to its simpler design,
but it requires lime which is more expensive than limestone. The applica-
tion of dry FGD to high-sulfur coal may be enhanced by using adipic acid
in the lime spray dryer method.
Regenerable FGD processes offer sulfur or sulfuric acid as by-products.
The Wellman-Lord process is being used at several power plants and a
100-MWe demonstration of the aqueous carbonate process is in the construction
phase. Generally, the higher lifetime costs and the markets for these
by-products have not encouraged widespread selection of regenerable FGD
processes.
Wastes from nonregenerable FGD processes are classed as nonhazardous,
with disposal by landfill becoming the general practice. Forced oxidation
has improved the disposal characteristics of wastes from the wet limestone
FGD process, making these wastes more suited to landfill disposal. Dry
wastes from lime spray dryers are well-suited to landfill disposal, but
the use of sodium compounds in dry FGD may require lined landfills to
limit the leaching of sodium salts into ground water.
279
-------
REFERENCES
1. National Air Pollutant Emission Estimates, 1970-1980.
U.S. Environmental Protection Agency, Monitoring and Data Analysis
Division, Research Triangle Park, NC 27711. Publication No. EPA-
450/4-81-010, March 1981.
2. Hunt, W.F. and E.D. Tillis, 1980 Ambient Assessment-Air Portion.
U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, Research Triangle Park, NC 27711.
Publication No. EPA-450/4-81-014, NTIS No. PB 81-178469, February
1981.
3. Sulfur Oxides Control Technology Series: Flue Gas Desulfurization,
Lime/Limestone Process. U.S. Environmental Protection Agency,
Office of Research and Development, Research Triangle Park, NC
27711. Publication No. EPA-625/8-81-006, April 1981.
4. Chang, J., and J.D. Mobley, "Overview of EPA's Testing of the
Adipic Acid Enhanced Limestone FGD Process," presented at the EPA
Industrial Briefing on the Adipic Acid Enhanced Limestone FGD
Process, Springfield, MO, July 15, 1981.
5. Jones, J.W. and C.J. Santhanam, "Flue Gas Cleaning Waste Manage-
ment in the United States," presented at the Third ECE Conference
on Desulfurization of Fuels and Combustion Gases, Salzburg, Austria,
May 18-22, 1981.
6. Sulfur Oxide Control Technology Series: Flue Gas Desulfurization,
Dual Alkali Process. U.S. Environmental Protection Agency, Office
of Research and Development, Research Triangle Park, NC 27711.
Publication No. EPA-625/8-80-004, October 1980.
7. Brna, T.G., "Dry Flue Gas Desulfurization in the United States,"
presented at the ECE Conference on Desulfurization of Fuels and
Combustion Gases, Salzburg, Austria, May 18-22, 1981.
8. Gibson, E.D., M.A. Palazzolo, and M.E. Kelly, Draft Summary Report,
Sulfur Oxides Control Technology Series: Flue Gas Desulfurization -
Spray Drying Process, September 1981, prepared by Radian Corporation
under EPA Contract No. 68-02-3171, Task No. 37.
9. Burnett, T.A. and K.D. Anderson, Technical Review of Dry FGD Systems
and Economic Evaluation of Spray Dryer FGD Systems. Publication
No. EPA-600/7-81-014, TVA EDT-127, NTIS No. PB 81-206476, February
1981.
280
-------
10. Smith, M. et al., EPA Utility FGD Survey: April-June 1981.
U.S. Environmental Protection Agency, Industrial Environmental
Research Laboratory, Research Triangle Park, NC 27711. Publication
No. EPA-600/7-81-012d, August 1981.
11. Sulfur Oxides Control Technology Series: Flue Gas Desulfurization,
Wellman-Lord Process. U.S. Environmental Protection Agency, Office
of Research and Development, Research Triangle Park, NC 27711.
Publication No. EPA-625/8-79-001, February 1979.
12. Sulfur Oxides Control Technology Series: Flue Gas Desulfurization,
Magnesium Oxide Process. U.S. Environmental Protection Agency,
Office of Research and Development, Research Triangle Park, NC
27711. Publication No. EPA-625/8-81-005, April 1981.
13. Kaplan, N. and M.A. Maxwell, "Regenerable Flue Gas Desulfurization
Systems in the United States," presented at the Third ECE Conference
on Desulfurization of Fuels and Combustion Gases, Salzburg, Austria,
May 18-22, 1981.
14. Byrd, J.R. , K.D. Anderson, S.V. Tomlinson, and R.L. Torstrick,
Definitive SOX Control Process Evaluations: Aqueous Carbonate and
Wellman-Lord (Acid, Allied Chemical, and Resox) FGD Technologies.
Publication No. EPA-600/7-81-099 , TVA EDT-121, June 1981.
15. Tomlinson, S.V., P.M. Kennedy, F.A. Sudhoff, and R.L. Torstrick,
Definitive SOX Control Process Evaluations: Limestone, Double
Alkali, and Citrate FGD Processes. Publication No. EPA-600/7-79-
177, TVA ECDP B-4, NT1S No. PB 80-105828, August 1979.
16. Anderson, K.D., J.W. Barrier, W.E. O'Brien, and S.V. Tomlinson,
Definitive SOX Control Process Evaluations: Limestone, Lime, and
Magnesia FGD Processes. Publication No. EPA-600/7-80-001, TVA
EDCP B-7, NTIS No. PB 80-196314, January 1980.
17. Parsons, E.L., V. Boscak, T.G. Brna, and R.L. Ostop, S02 Removal by
Dry Injection and Spray Absorption Techniques, presented at the
U.S. EPA's Third Symposium on the Transfer and Utilization of
Particulate Control Technology, Orlando, FL, March 1981.
281
-------
HEALTH AND ENVIRONMENTAL STUDIES OF COAL GASIFICATION
PROCESS STREAMS AND EFFLUENTS
by: C.A. Reilly, Jr., A.S. Boparai, S. Bourne, R.D. Flotard,
D.A. Haugen, R.E. Jones, F.R. Kirchner, T. Matsushita,
M.J. Peak, V.C. Stamoudis, J.R. Stetter, and K.E. Wilzbach
Synfuels Environmental Research Program
Argonne National Laboratory
Argonne, Illinois 60439
ABSTRACT
The Synfuels Environmental Research Program at Argonne National
Laboratory is investigating the impact of high-BTU coal gasification on
health and the environment. Activities include a toxicologic and chemical
characterization of process streams in the gasifier and pretreater sections
of the HYGAS coal gasification pilot plant, and process streams and work-
place air from the Grand Forks Energy Technology Center's slagging fixed-
bed gasifier facility. Cellular assays for mutagenicity, cytotoxicity, and
functional impairment are performed to determine relative toxicity.
Various acute and chronic whole animal toxicological evaluations, including
skin tumorigenesis, are performed for streams found to contain potential
toxicants (e.g. oils and tars). The chemical characteristics of vapor phase
and airborne particulate-associated organics, as well as biologically active
materials isolated from process streams, are investigated by physical and
chemical fractionation of the samples, with biological monitoring and
detailed GC and GC/MS analyses of the fractions. Present data indicate that
toxicants are present, but their levels of activity are relatively low. As
a result of these studies, we tentatively conclude that with appropriate
control technology and industrial hygiene procedures there appear to be no
serious health or environmental problems associated with coal gasification.
INTRODUCTION
A thorough evaluation of the potential impact on human health and the
environment is a prerequisite to implementation of new fossil fuel conver-
sion technologies. The Department of Energy (DOE), through its Offices of
Energy Research and Fossil Energy, sponsors several research programs to
provide this evaluation. Argonne National Laboratory has the major respon-
sibility in the area of high-BTU coal gasification, and thus has the
obligation to develop:
282
-------
• A comprehensive health and environmental data base for coal gasifi-
cation
• A reliable assessment of the risks associated with large-scale
coal gasification
Projects within Argonne's Synfuels Environmental Research Program
include a completed study of the HYGAS pilot plant at the Institute of
Gas Technology (IGT) Energy Development Center (Chicago, IL) and ongoing
studies of the slagging fixed-bed gasifier at the Grand Forks Energy
Technology Center (GFETC) (Grand Forks, ND), and a bench scale gasifier
at the Carnegie-Mellon Institute of Research (Pittsburgh, PA). This report
discusses the HYGAS pilot plant studies and gives initial results from
GFETC. Experiments with the Carnegie-Mellon gasifier samples were only
recently initiated and will not be presented.
The Argonne program employs an integrated multidisciplinary approach
to sample characterization. Sample preparation and fractionation activi-
ties, toxicological characterization, and identification of specific
chemical components required interaction and collaboration among personnel
in three Argonne Divisions:
Analytical Chemistry Laboratory (Chemical Engineering Division)
Biological and Medical Research Division
Energy and Environmental Systems Division
BACKGROUND
The end product of high-BTU coal gasification is a substitute natural
gas (SNG), essentially indistinguishable from natural gas, and therefore
presents no new health or environmental issues. However, most gasification
processes also produce by-product oils and tars known to contain a variety
of noxious chemicals, including carcinogens. These oils and tars are
extremely complex mixtures and are at present chemically illdefined.
Although these potentially toxic by-products are readily removed from the
product gas and can be completely consumed on site, potential routes of
human exposure do remain - namely, direct contact with solids and liquids,
or inhalation of fugitive vapor, aerosol and particulate emissions. This
program emphasizes the toxicological and chemical characterization of the
organic components of process streams. This report summarizes our activi-
ties to date at IGT and GFETC. Additional more detailed information has
appeared elsewhere (1-5).
PROCESS DESCRIPTIONS
HYGAS
This advanced process for high-BTU coal gasification was developed
by IGT with support from the Gas Research Institute and DOE. An 80 ton
per day pilot-plant gasifier has operated for several years on a variety
of coal types. A detailed description of the process, the operating con-
ditions during sampling, and the samples that were collected is included
283
-------
in a recent report presented at the Second International Gas Research
Conference (6).
Emphasis in this report is placed on the pilot plant process streams
deemed to be of environmental importance in commercial facilities. Sam-
ples from the low temperature reactor (LTR) provide a reasonable approxi-
mation of the organic compounds formed during coal gasification. Not all
of these organic compounds leave the gasifier, but those that do enter
the recycle oil. The recycle oil system, which accounts for the largest
mass flow of organics within a plant, has the greatest potential for leaks
and fugitive emissions. It is also the most significant environmentally,
since it will represent a steady-state composition of high molecular
weight organic compounds leaving the gasifier. Organics formed in the
coal pretreatment process also could be a major source of fugitive emission
in the commercial plant. In the pilot plant these organic compounds are
found in the pretreater quench water. Finally, organic compounds present
in the gasifier quench water, if not removed in the water treatment system,
could enter the atmosphere through cooling tower water desorption.
Although spent char is a significant plant discharge stream, it would be
incinerated in a commercial plant and thus is not considered to be environ-
mentally important.
SLAGGING FIXED-BED GASIFIER
The gasifier operated by GFETC is a 25 ton per day slagging fixed-bed
pressurized gasifier which differs from other fixed-bed gasifiers in that a
lower steam:oxygen ratio is used for the gasification reaction, and
operating temperatures are high enough to melt the ash for discharge as a
molten slag. This gives the slagging process several advantages, inclu-
ding higher throughput, lower steam consumption, and lower wastewater
production. A detailed description of the process was presented at the
1981 Lignite Symposium (7).
Waste effluent streams from a fixed-bed gasification process consist
of (a) gaseous contaminants (chiefly H2S, C02, and light hydrocarbons),
which are cleaned from the product gas by commercially available processes
to meet end use requirements; and (b) solid and liquid effluent streams
consisting of slag, slag quench water, and the gas liquor, composed of
condensed tar, oil, water, and coal dust entrained in the product gas.
Additional solid wastes may also be generated by wastewater treatment pro-
cesses. Emphasis in our studies has been placed on what is considered to
be potentially the most noxious gasifier waste stream, the gas liquor
obtained from the condensation of tars, oils, and water in the spray
cooler. While the GFETC gasifier is slightly different from a commercial
design, the chemical nature of the tar and oil should be more dependent on
the nature of the feed coal than of gasifier design.
METHODOLOGY
SAMPLING
All process stream samples were collected by the plant operators
284
-------
under conditions of steady operation. Following collection, the samples
were refrigerated (4°C) and shipped to Argonne. At Argonne a locked
refrigerated storage vault was used. Detailed records were systematically
maintained to document sample preparation, transfer to and from storage,
fractionation, and subsequent distribution.
The HYGAS samples studied were from the final four HYGAS test runs,
all of which used a Western Kentucky feed coal. Samples from GFETC came
from a test run that used North Dakota Indian Head lignite. Samples were
collected from the tar-oil separator and consisted of process oil, tar,
and water.
Samples of airborne particulates and fugitive organic vapors were
collected by Argonne personnel in collaboration with GFETC process and
environmental engineers at various locations in the gasifier building.
Vapor samples were collected on XAD-2 resin at level 2 (steam injection
and control room) and level 7 (lockhopper) during a run with Indian Head
lignite. The resin was extracted with methylene chloride and the organic
compounds present in the extract were identified by GC/MS and quantified
by GC. Procedures and techniques have been reported in an earlier study
concerning sampling at the HYGAS facility (5).
Particulate sampling was conducted during shutdown operations and
operation of the gasifier following an aborted start up. A size-fraction-
ated particle sample collected with a Sierra high-volume impactor was
subjected to analysis by GC/MS and scanning electron microscopy; aerosols
collected with an Anderson low-volume impactor were subjected to flameless
atomic absorption analysis.
TOXICOLOGY
Three cellular tests were used to establish relative toxicities of
all process samples. This battery of procedures was required to provide
the variety of toxicological end points and systems necessary to evaluate
the broad chemical spectrum of compounds present in the sample materials.
The Ames Salmonella plate incorporation assay was used to determine muta-
genicity. Strain TA98 was used exclusively because it was found to be the
most sensitive of the five commonly used strains. Metabolic activation of
samples with rat liver S9 enzymes was essential for expression of muta-
genicity. Mouse myeloma cells were used to measure both genotoxicity (by
sister chromatid exchange) and cytotoxicity (by growth inhibition).
Finally, two additional measures of cytotoxicity, a gross measure (cell
death) and a subtle measure (loss of normal cell function), were obtained
using the rabbit alveolar macrophage (RAM) assay. The functional loss was
evaluated by determining the inhibition of normal phagocytic activity in
these cells.
In addition, whole animal toxicological assays were performed on
HYGAS recycle oil. The assays included measurement of the effects of both
acute and chronic dermal exposures and acute ocular exposures. Dermal
effects were studied in SKH hairless mice (carcinogenicity), albino guinea
pigs (hypersensitivity), and New Zealand albino rabbits (acute effects).
285
-------
Ocular tests were performed in New Zealand albino rabbits.
BIO-DIRECTED CHEMICAL CHARACTERIZATION
'Chemical characterization of samples was always performed in support
of toxicological determinations. Table 1 gives the procedures employed.
Many of the process samples were heterogeneous and not suited for direct
testing in the cellular assay systems. Materials were physically sepa-
rated and the organic components were extracted from the aqueous and solid
phases. The resulting extracts and oils were then fractionated on the
basis of volatility- In our experience only the nonvolatile organic (NVO)
fractions containing components boiling at greater than % 200°C exhibited
mutagenic activity and thus were the materials entered in the toxicity
screening tests. Materials found to show significant toxicity in the test
screen were further fractionated on the basis of acidity and polarity.
Mutagenic activity in the various fractions was monitored with the Ames
Salmonella assay. Chemical fractionation procedures included both liquid/
liquid partitioning and high efficiency column chromatography. GC/MS was
used for identification of the components in a given fraction and fused
silica capillary column GC was used for quantification (1).
TABLE 1. BIO-DIRECTED CHEMICAL CHARACTERIZATION
Physical Fractionation
Phase
Volatility
Chemical Fractionation
Acidity
Polarity
Biomonitoring (Ames assay)
Chemical Testing (Nitrous acid)
Compound Identification (GC/MS)
Compound Quantification (GC)
RESULTS
Information is available for scores of samples and fractions of HYGAS
materials and is rapidly accumulating for the GFETC gasifier. Presentation
of HYGAS results is restricted to average toxicities observed in process
streams deemed to be significant, either by the degree of toxicity or the
potential for human exposure. GFETC data, being preliminary in nature,
is restricted to process tars and oils, which are the putative major toxi-
cants .
The streams that are discussed, and their approximate mass flow rates
relative to feed coal are shown in Table 2.
286
-------
TABLE 2. SIGNIFICANT PILOT PLANT STREAMS
Stream Mass Flow
(coal = 1)
HYGAS
Recycle Oil 3
Pretreater Quench Water 3
Gasifier Quench Water 1
Low Temperature Reactor Gas 1.4
GFETC Gasifier
Tars and Oils 0.05
Water 0.6
It is important to emphasize that the mutagenicities of HYGAS process
samples were too low to measure directly and the toxicity data presented
applies only to the NVO fraction. These fractions always constitute a very
small part of the process stream (0.1 to 4 weight percent of the sample).
Process stream toxicities (calculated as the product of NVO toxicity and
its weight fraction) are accordingly quite low. Tars and oils from the
GFETC gasifier have relatively greater weight percent of NVO's: approxi-
mately 90% NVO for tars and 50% for oils but only 5% of the coal is con-
verted to tars. Because these tars and oils are not recycled, total
toxicity of the raw product gas stream remains relatively low.
Results observed in the Ames Salmonella Assay of NVO fractions of
HYGAS samples (Table 3) show that the greatest specific mutagenic activity
is in the low temperature reactor condensate. Its specific mutagenicity is
about 15% of the known carcinogen benzo(a)pyrene (BaP), but material balance
studies have shown that not all of this mutagenicity leaves the gasifier.
The most important HYGAS stream, the recycle oil, has an average NVO
mutagenic activity less than 3% of BaP. Low or insignificant specific
mutagenicity is observed with quench waters from the pretreater and gasi-
fier and no mutagenicity could be detected in the spent char. Extrapo-
lated specific mutagenicity of all process streams is less than 1
revertant/yg.
287
-------
TABLE 3. MUTAGENICITY OF HYGAS PROCESS STREAMS
(AMES SALMONELLA ASSAY)
Sample
Recycle Oil
Pretreater Quench Water
Gasifier Quench Water
Low Temperature Reactor
Condensate
Spent Char
Benzo(a)pyrene
Mutagenicity, rev/yg
NVO Fraction Process Stream
7
2.4
0.7
35
260
0.17
0.007
0.0005
0.37
neg.
Cytotoxicity and genotoxicity measurements in mouse myeloma cells
support the general conclusion that on a process stream basis toxicity
is low. However, untreated quench waters contain significant toxicity
(Table 4). They are the most toxic sample type, being threefold more
genotoxic than LTR condensates or recycle oil. This test does not require
metabolic activation but when the LTR condensate is activated with rat
liver S9 enzymes, genotoxicity increases tenfold and approximates the
relative activity of LTR condensate to BaP seen in the Ames Assay.
TABLE 4. TOXICITY OF HYGAS PROCESS STREAMS
(MOUSE MYELOMA CELLS)
Sample
Cytotoxicity* Genotoxicity**
L/g L/g
Recycle Oil
Pretreater Quench Water
Gasifier Quench Water
Low Temperature Reactor
Condensate
16.5
19.3
55.7
12.3
15.5
31.4
47.3
17.7
Methyl Methane Sulfonate 36.0
Low Temperature Reactor 20.8
Condensate with Activation
370
175
Benzo(a)pyrene with Activation 112
1230
*The reciprocal of the NVO concentration for 50% growth
inhibition.
**The reciprocal of the NVO concentration for a twofold
increase in sister chromatid exchanges.
288
-------
In the RAM assay HYGAS materials are approximately equivalent within
a given end point (Table 5). There is however a clear indication of subtle
toxicity (functional loss) at concentrations of HYGAS materials signifi-
cantly lower than those required for cell killing.
TABLE 5. TOXICITY OF HYGAS PROCESS STREAMS
(RAM ASSAY)
Functional'
Cytotoxicity* Impairment**
Sample L/g NVO L/g NVO
Recycle Oil
Gasifier Quench Water
Pretreater Quench Water
Low Temperature Reactor
Condensate
Vanadium Oxide
7.1
4.1
6.1
7.4
153
9.8
10.8
8.1
11.4
208
*The reciprocal of the NVO concentration for 50% cell
killing.
**The reciprocal of the NVO concentration causing a 50%
reduction in the phagocytic activity of viable cells.
Ocular toxicity tests in rabbits have demonstrated that recycle oil
NVO is a severe irritant according to National Academy of Science criteria
(NAS publication 1138, 1977). We observed inflammatory reactions, corneal
ulcers, and panus that persisted for 21 days. Likewise, rabbit skin expo-
sure results in mild to severe inflammatory reactions with some skin
necrosis. Marked skin hypersensitivity is detected in guinea pigs. We
found that raw recycle oil is a mouse skin carcinogen, inducing tumors in
SKH hairless mice following chronic exposure (weekly 150 pi doses). The
tumorigenic response is considerably less than that for BaP (.03 yg/week);
however, 105 .yg of the recyle oil NVO approximates the BaP tumor response.
Tumor response is based on gross observation, but histologically confirmed
squamous cell carcinomas of the skin have been observed, some with metas-
tatic nodules. Thus while it is clear that recycle oil is toxic, it is
important to emphasize that the potential for human exposure is limited.
The nature of the stream and normal industrial hygiene protocols should
make hazards associated with recycle oil totally manageable.
Table 6 gives preliminary toxicological evaluations of GFETC tars
and oils. Results are for unfractionated samples and should not be
confused with the toxicity of NVO fractions. Mutagenicity is insigni-
ficant in oils but present at a level 5% that of BaP in tars. Samples
are cytotoxic and genotoxic but again not at high levels. The toxicity
of untreated process water while relatively low is significant when the
large volume of this stream is considered (Table 2).
289
-------
TABLE 6. TOXICITY PRESENT IN THE GRAND FORKS ENERGY TECHNO-
LOGY CENTER'S SLAGGING FIXED-BED GASIFIER PROCESS STREAMS
Tar Oil Water Reference*
Standards
Mutagenicity , rev/ g
Cytotoxicity*, L/g
Genotoxicity*, L/g
Functional loss*, L/g
4
11
16
14
<1
11
12
0
0.7
1.4
9
260
36
370
208
*As in assay of HYGAS samples, see Tables 3-5.
AIR MONITORING
The organic vapor concentrations we observed in the GFETC gasifier
facility ranged from approximately 1-500 yg/m for individual components
with an overall organic vapor concentration of from 2-3.5 mg/m . One- to
three-ring aromatic hydrocarbons accounted for the bulk (50-70%) of the
material with aliphatic hydrocarbons (30-40%) accounting for most of the
remainder. Phenols and heterocyclic compounds each accounted for about 2%
of the total organic fraction. The concentrations of individual compounds
(e.g., benzene) were well below TLV limits in all cases.
The bulk of the particulate-associated trace organic material was
associated with particles having an aerodynamic size of <3.1 ym. The
particulate phase organics contained significantly higher proportions of
aliphatic and phenolic compounds than the vapor phase samples and this is
consistent with published data for ambient air. Particle morphologies
resembled those of lignite fly ash from combustion, and particle types
included smooth spheres, vesicular spheres, agglomerated masses and
crystalline fragments. Trace element size distributions were bimodal and
resembled those for ambient air. Lead particle sizes were predominantly
submicron, while particles of Al, Fe, and other crustal species were
mostly of supermicron size. Aluminum-based aerosol enrichment factors
calculated from Indian Head lignite showed that the composition of the
aerosol resembled that of the coal, with the exception of modest enrich-
ment of Mg, Na, As, and Pb in the submicron size range. Aerosol enrich-
ment factors based on the earth's crustal composition were somewhat greater
than those based on coal composition for several elements, suggesting
potential errors in using crustal enrichment data to investigate chemical
fractionation during aerosol formation.
CHARACTERIZATION OF TOXIC COMPOUNDS
Because toxicity was confined to nonvolatile materials, it was pos-
sible to fractionate and concentrate without evaporative loss of toxic
material. As previously mentioned, the components of recycle oil were
fractionated by acidity and polarity. Separation by polarity was accom-
290
-------
increasing proportions of benzene, accounted for half of the sample weight
but were only slightly mutagenic despite the fact that they contained
polyaromatic hydrocarbons such as BaP. Virtually all of the mutagenicity
recovered was contained in the fourth most polar fraction, (eluted with
methanol).
In the initial pH fractionations, methylene chloride samples were
partitioned between an aqueous acid or base. The base fraction contained
more than 70% of the recovered mutagenicity. Although the base fraction
contained only 4% of the initial weight, its specific mutagenicity was
more than tenfold higher than that of the original sample. The neutral
fraction contained the remainder of the recovered mutagenicity with the
acid fraction (exclusively phenolic) being nonmutagenic. The neutral
and acidic fractions accounted for 50% and 30% of the initial weight,
respectively.
Chemical analysis showed that the base fraction contained azaarenes
(AA) and primary aromatic amines (PAA) in the ratio of 4:1. Since mem-
bers of both classes of compounds are known to be toxic, further analy-
ses were performed to determine which components were mutagenic. The
loss of mutagenicity following mild nitrous acid treatment (which modi-
fies PAA but not AA) suggests that the PAA are responsible for the
mutagenicity. More conclusive evidence was obtained by applying a new
procedure for cation exchange high performance liquid chromatography (8).
this procedure separates PAA (weaker bases) from AA and resolves members
of each class having pKa values differing by less than 0.2 pKa units.
Chromatography of the LTR base fraction revealed that most of the muta-
genicity is concentrated in a few fractions containing 2- to 4-ring PAA
as demonstrated by GC/MS. Azaarenes which elute in the later fractions
contain little of the mutagenicity.
CONCLUSIONS
The results of this study on the toxicological and chemical charac-
terization of two high-BTU coal gasification pilot plants demonstrate
that while toxicants are present, they are a minor component of the
process streams. The toxicity is largely confined to the nonvolatile
components of the by-product tars and oils. These materials will not
leave the commercial plant site because they are ultimately consumed in
the process. On-site emissions can be controlled through appropriate
control technology. Occupational exposures can be minimized through
effective industrial hygiene procedures. These considerations allow
the general assessment that no apparent serious health or environmental
problems are associated with coal gasification.
ACKNOWLEDGMENT
We wish to express our appreciation to: Paul Duhamel and Mayo
Carrington, the DOE program managers, for their guidance; to Rich
Biljetina and Lou Anastasia of IGT and Warrick Willson of GFETC for
their helpful assistance in obtaining process samples; to Cliff Davidson
and Suresh Santhanan for their aid in particulate sampling and atomic
291
-------
absorption analysis; to the principal investigators staff, for their
valuable contributions; and to M. Rosenthal for her helpful suggestions in
the preparation of this manuscript. This work was supported by the U.S.
Department of Energy under contract No. W-31-109-ENG-38.
REFERENCES CITED
1. V.C. Stamoudis, S. Bourne, D.A. Haugen, M.J. Peak, C.A. Reilly,
Jr., J.R. Stetter, and K. Wilzbach, "Chemical and Biological
Characterization of High-BTU Gasification (The HYGAS Process)
I. Chemical Characterization of Mutagenic Fractions," Proce-
edings of the 20th Annual Hanford Life Sciences Symposium, 1980
in press.
2. D.A. Haugen, M.J. Peak, and C.A. Reilly, Jr., "Chemical and
Biological Characterization of High-BTU Gasification (The
HYGAS Process) II. Nitrous Acid Treatment for Detection of
Mutagenic Primary Aromatic Amines: Non-Specific Reactions,"
Proceedings of the 20th Annual Hanford Life Sciences Symposium,
1980, in press.
3. S. Bourne, A. Jirka, and M.J. Peak, "Chemical and Biological
Characterization of High-BTU Gasification (The HYGAS Process)
III. Mass and Mutagenicity Balances in Chemical Fractions,"
Proceedings of the 20th Annual Hanford Life Sciences Symposium,
1980, in press.
4. C.A. Reilly, Jr., M.J. Peak, T. Matsushita, F.R. Kirchner, and
D.A. Haugen, "Chemical and Biological Characterization of
High-BTU Gasification (The HYGAS Process) IV. Biological Acti-
vity," Proceedings of the 20th Annual Hanford Life Sciences
Symposium, 1980, in press.
5. R.D. Flotard, J.R. Stetter, and V.C. Stamoudis, "Workplace Air
Sampling at Coal-Conversion Facilities," Proceedings of the
20th Annual Hanford Life Sciences Symposium, 1980, in press.
6. K. E. Wilzbach and C. A. Reilly, Jr. "Chemical and Toxicological
Characterization of HYGAS Pilot Plant Streams," Proceedings of the
2nd International Gas Research Conference, in press.
7. W. G. Willson, L. E. Paulson, R. S. Majkrzak, W. B. Hauserman,
and R. G. Luthy. "Slagging Fixed-Bed Gasification of Lignite,"
Proceedings of the 1981 Lignite Symposium, in press.
8. D. A. Haugen, V. C. Stamoudis, M. J. Peak, and K. M. Suhrbier.
Isolation of mutagenic aromatic amines from a coal conversion
oil by cation exchange Chromatography, Anal. Chem., 54 (in
press).
292
-------
GASEOUS FUGITIVE EMISSIONS FROM SYNFUELS PRODUCTION - SOURCES AND CONTROLS*
by: R.L. Honerkamp
Radian Corporation
8501 MoPac Blvd.
Austin, TX 78766
ABSTRACT
Fugitive emissions are generally defined as emissions that are not re-
leased through an enclosure such as a duct or vent pipe. This definition
includes sources of fugitive particulate emissions and sources of gaseous
fugitive emissions. In this paper, the potential sources and control options
for gaseous fugitive emissions from synfuels production facilities are des-
cribed. Gaseous fugitive emissions are caused by process fluid leakage from
seals (valves, pumps, flanges), process fluid purges (sampling, equipment
cleaning), and secondary emission sources (drains, wastewater systems, cool-
ing towers). The majority of sources of fugitive emissions in the U.S. are
currently found in petroleum production and refining facilities, organic
chemical manufacturing plants, and coke by-product plants. Synfuels produc-
tion facilities will also have fugitive emission sources.
Fugitive emission regulations have been applied to California petroleum
refineries for several years and U.S. New Source Performance Standards (NSPS)
are currently under development for several industries. These regulations
are based on the need to reduce emissions of volatile organic compounds
(VOC), because VOC are photochemical ozone precursors. Some fugitive emis-
sions also need to be controlled because compounds released in the emissions
may be harmful. U.S. National Emission Standards for Hazardous Air
Pollutants (NESHAP) have been developed for controlling fugitive emissions of
vinyl chloride and benzene. Fugitive emissions from synfuels production
facilities may require control because they contribute to atmospheric ozone
formation and/or because the emissions contain harmful compounds. The nature
of potentially harmful compounds will be dependent on variables such as the
type of process, feedstock characteristics, and operating parameters.
Fugitive emission controls can be categorized as either work practices
or engineering controls. Work practices include leak detection/leak repair
programs and "housekeeping" practices. Leak detection and repair programs
involve periodic testing to locate significant leaks and subsequent repairs
to reduce or eliminate the leakage. Housekeeping practices would include
procedures to minimize process fluid spills and to expedite spill cleanup.
Engineering controls are generally equipment substitution strategies. For
example, closed loop sampling connections eliminate process fluid purge emis-
sions, and double mechanical pump seals can be operated to minimize seal
293
-------
emission potential. These types of equipment could be substituted for equip-
ment with a greater potential to leak process fluids.
Because the emission sources (pumps, valves, flanges, etc.) in synfuels
plants will be similar to those in existing U.S. industries, emission control
techniques used in existing industries will also be applicable to synfuels
facilities. The experience that has been gained in applying fugitive emis-
sion controls will be valuable in developing emission control strategies for
synfuels plants.
INTRODUCTION
Gaseous fugitive emissions are the result of process fluid leakage and
process fluid purges. Secondary emission sources such as cooling towers and
wastewater systems may also be classified as fugitive emission sources. In
contrast to process emissions, which are released through ducts or vent
pipes, fugitive emissions are released from numerous discrete sources such as
valves, pumps, and flanges located throughout a process unit. Process
emissions are amenable to application of emission control devices that either
recover or destroy the emissions conveyed to the device by a duct or pipe,
but fugitive emission sources require a completely different type of emission
control strategy. The purpose of this paper is to identify the potential
sources of gaseous fugitive emissions in synfuels production facilities and
to discuss the types of controls that can be applied to reduce emissions from
those sources.
Fugitive emissions have received rapidly increasing attention in the
last five years and there are several reasons for this increased awareness of
a need to control fugitive emissions. Because process emissions are released
through an enclosed pipe or duct, it is fairly straightforward to convey
these emissions to a control device. Furthermore, process emission sources
generally contribute a much larger portion of the total emissions compared to
fugitive emission sources. For these reasons, process emission sources have
been selected for application of controls first. As more controls are ap-
plied to process sources, fugitive emissions become a significant contributor
to the remaining controllable emissions from a process unit. In addition,
fugitive emission sources may be the major contributor to the total emissions
of specific compounds that require control.
The compounds released from fugitive emission sources may require con-
trol because they are volatile organic compounds (VOC), which have been
linked to photochemical atmospheric ozone formation. Other compounds such as
benzene and hydrogen sulfide may require control because the compounds them-
selves are health hazards. Regulations to control fugitive emissions have
been applied to the petroleum industry in California for VOC control for sev-
eral years, and other states are developing regulations. Federal New Source
Performance Standards (NSPS) are currently being developed for VOC control in
several industries. National Emission Standards for Hazardous Air Pollutants
(NESHAP) have been developed for fugitive emissions of benzene and vinyl
chloride. These regulatory activities have resulted in increased awareness
of the need to learn more about the sources and controls of fugitive emis-
sions.
294
-------
SOURCES OF FUGITIVE EMISSIONS
Fugitive emissions are caused by process fluid leakage, process fluid
purging and atmospheric exposure of process fluids by secondary emission
sources. Any type of process equipment that is capable of allowing process
fluids to come in contact with the atmosphere is a potential fugitive emis-
sion source. These sources of fugitive emissions are shown in Table 1.
PROCESS FLUID LEAKAGE
Sources of process fluid leakage include valve packings, seals on pumps,
compressors and agitators, flange gaskets, relief valve seats, and valve
seats on open-ended lines. An open-ended line is defined as a process valve
that is installed with one side of the valve in contact with process fluid
and the other side in contact with the atmosphere, such as purge valves,
drain valves and vent valves.
All process valves, except relief valves and check valves, are activated
by a valve stem which may have either a rotational or linear motion, depen-
ding on the specific design. The moveable stem requires a sealing element to
resist fluid leakage while permitting movement of the stem. In most valves,
this seal is achieved with a packing compression gland. Other valves may
have elastomeric 0-rings or grease-filled lantern rings to prevent leakage of
process fluid. Although these types of seals are satisfactory for preventing
gross leakage of process fluids, they can allow a significant amount of fugi-
tive leakage. Corrosive or toxic process fluids may require the use of a
valve with a diaphragm or flexible bellows to isolate the stem sealing ele-
ment from the process fluid, and these valves would also provide increased
resistance to fugitive leakage.
Packed seals on pumps, compressors, and agitators are similar to packed
seals on valves. Because the shafts on these devices rotate constantly, per-
iodic adjustment of the packing is required. Mechanical seals consist of
stationary and rotating elements that are machined to a very close tolerance.
The mechanical contact of the two elements resists fluid leakage. As with
packed seals, leaks can occur where the shaft protrudes through the seal.
Double mechanical seals and oil film seals have a barrier fluid system that
resists seal leakage. However, seal leakage can be entrained in the oil sys-
tem and can be released to the atmosphere by degassing from the oil
reservoir.
Leaks from flange gaskets can be caused by loose bolts, improper speci-
fication of materials, thermal stresses, and deterioration of the gasket mat-
erial. Although they are the most numerous type of fugitive emission source,
flanges contribute a small portion of total fugitive emissions.
Relief valves are designed to open at a predetermined pressure in order
to protect process equipment from damage due to overpressure. The discharge
that occurs when these valves open is considered a process emission. Fugi-
tive emissions from relief valves are the result of leakage through the valve
when it is closed. This leakage can be caused by improper reseating after an
295
-------
TABLE 1. FUGITIVE EMISSION SOURCES
Source Type
Location of Emission Release
Sources of Fluid Leakage
Valves
Pumps and Agitators
Packed seals & single mechanical
seals
Double mechanical seals
Compressors
Packed seals & single mechanical
seals
Double mechanical seals & oil
film seals
Flanges
Relief Valves
Open-ended lines (valves)
Sources of Process Fluid Purges
Sampling Operations
Equipment Emptying Operations
Secondary Emission Sources
Cooling Towers
Wastewater Systems
Stem/body junction
Shaft/case junction
Shaft/case junction; oil reservoir
degassing vent
Shaft/case junction
Shaft/case junction; oil reservoir
degassing vent
Face/gasket junction
Disc/seat junction
Valve disc/seat junction
Purge/atmosphere contact
Purge/atmosphere contact
Cooling tower plume
Drains, open sewers or canals,
collecting basins, separators,
aeration ponds.
296
-------
overpressure, deterioration of the valve .seat, and operation of the process
at or near the set pressure that causes the valve to open.
Open-ended lines are found on valves used for draining equipment, for
purging or venting equipment, and for obtaining process fluid samples. A
faulty valve seat or incomplete closure of the valve would allow process
fluid to leak to the atmosphere through the open-ended line.
PROCESS FLUID PURGES
The sources of process fluid purges are sampling operations and equip-
ment draining and venting operations. In order to obtain a representative
sample of process fluid, the sample line is purged with the process fluid.
If this purge is allowed to contact the atmosphere, fugitive emissions may be
created. Process fluids are also purged from equipment prior to removing the
equipment from service for inspection, repair, replacement, etc. Atmospheric
contact with these purges can also result in fugitive emissions.
SECONDARY EMISSION SOURCES
Cooling towers and wastewater systems are considered as secondary
sources of fugitive emissions because they are not the initial source of the
process fluids. Process fluids may enter a cooling tower water system due to
leakage in heat exchangers or from the use of contaminated process water as
cooling tower make-up water. As the contaminated water is circulated through
the cooling towers, process fluid components are stripped from the water and
are released to the atmosphere with the evaporated cooling water.
Wastewater systems consist of drains, collection basins, canals, separa-
tors, and water treatment facilities. Because these systems have numerous
locations where the wastewater contacts the atmosphere and the wastewater is
frequently in a state of turbulent mixing, process fluids in the wastewater
can readily become atmospheric fugitive emissions.
SEVERITY OF FUGITIVE EMISSIONS
The degree of environmental severity associated with fugitive emissions
is dependent on two variables: 1) species emitted and 2) the total emission
rate. The relative importance of these variables depends on the type of
environmental impact that is being evaluated. With respect to the impact of
fugitive emissions on atmospheric ozone formation, the total emission rate of
VOC is the most significant consideration. Impacts on industrial hygiene
would be dependent on the types of substances emitted and the proximity of
emission release points to workers. Some of the less volatile process fluids
may also accummulate over a period of time, and workers may come into contact
with harmful species at any time after the initial release of the process
fluid.
The contribution of fugitive emission sources must also be accounted for
in considering the need for Prevention of Significant Deterioration (PSD)
review due to emissions in excess of De Minimis levels. The provisions of 40
297
-------
CFR 51-52 allow exemption from PSD review if the annual controlled pollutant
mass emission rate from a plant is less than the established De Minimis
value.
Emission control cost effectiveness is closely related to the total
number of sources and the emission factor for each type of source. It is
less expensive to control a few sources with high emission rates than to
control many sources with low emission rates, although the total uncon-
trolled emission contribution of the two groups may be equal.
ASSESSMENT OF POTENTIALLY HARMFUL SPECIES
One of the first criteria that can be used in evaluating pollutants of
concern in the process stream is to identify compounds covered by existing
regulations. These would include "criteria pollutants" covered by National
Ambient Air Quality Standards (NAAQS, 40 CFR 50) and "regulated pollutants"
covered by National Emission Standards for Hazardous Air Pollutants (NESHAP,
40 CFR 61). Some compounds of potential environmental concern may be present
in synfuels streams, but have not been the subject of specific regulations.
Such cases arise when there are insufficient or inconclusive data available
for the promulgation of enforceable regulations. One method of identifying
those compounds that need additional evaluation is to compare their
Multimedia Environmental Goals (MEG's).
Multimedia Environmental Goals (MEG's) have been established for over
650 chemical substances and physical agents (e.g., noise, heat). These goals
(in air, water, and solid waste streams) are the maximum discharge concentra-
tions (DMEG's) arid maximum ambient concentrations (AMEG's) which will avoid
potentially hazardous risks for public health or the ecology. These goals
are intended to be used in prioritizing research efforts, not in establishing
discharge limits. Most of the MEG's are derived using models that translate
toxicological data (threshold limit values, water quality criteria, carcino-
gen test results, etc.), dispersion assumptions, and federal standards or
criteria into discharge and ambient level goals. In many cases the models
translate data from one medium to goals for another medium. Despite their
obvious limitations, the MEG's do provide a method (and often the only
method) of identifying pollutants of potential concern.
In order to prioritize the need for concern about specific substances,
their MEG values can be compared. In Table 2, maximum ambient concentrations
(AMEG's) are shown for several species that could be present in synfuels
process streams. The AMEG values are in micrograms of pollutant per cubic
meter of ambient air; however, the actual numerical values are not directly
applicable to fugitive emission sources. These AMEG values are shown here in
order to give an indication of what ambient concentrations may be of concern
for some species which may be present in synfuels plant fugitive emissions.
Except for the benzene and vinyl chloride NESHAP, fugitive emission
regulations have been developed in the U.S. because of the need to reduce VOC
emissions. Synfuels facilities will also have sources of fugitive VOC emis-
sions and the presence of harmful species in these fugitive emissions will
298
-------
TABLE 2. HARMFUL SPECIES POTENTIALLY PRESENT
IN SYNFUELS PROCESS STREAMS
Pollutant AMEG*
Sulfur Compounds
Hydrogen sulfide 36
Carbonyl sulfide 800
Carbon disulfide 143
Mercaptans 2.4
Thiosulfates
Nitrogen Compounds
Ammonia 43
Cyanides 26
Thiocyanates -
Nitrates, Nitrites
Organic Compounds
Carbon monoxide 10,000
Benzene and other aromatics 7.1
Polynuclear aromatics (PNA's) 0.00005 to 119
Phenols 24 to 45
Organometallic compounds -
Methanol 619
Trace Metals
Antimony 1.2
Arsenic 0.005
Beryllium 0.01
Cadmium 0.02
Lead 0.36
Mercury 0.01
Nickel 0.035
Selenium 0.03
* Ambient Multimedia Environmental Goal (yg/m^) from References 1, 2, and
3.
299
-------
depend on the type of process, feedstock characteristics, and operating con-
ditions. In the first phase of an environmental assessment program, the
major process discharge streams are characterized. Except for the fugitive
emission sources that contain (and therefore have the potential to emit) the
species in these discharge streams, very little of this characterization is
applicable to fugitive emissions. Because fugitive emissions are composed of
the process fluid, it is necessary to know what harmful species are in these
process fluids. Process stream characterization data for synfuels processes
are frequently unavailable, especially for developing technologies. This
information may be deemed proprietary by the process developer because the
purpose of most of these process stream analyses is to determine the effects
of process variables on the yield of the primary reaction products, not to
characterize potentially harmful species in the streams.
Although limited test data are available, it is possible to estimate the
harmful species that are likely to be present in a process stream. In Table
2, several general categories of harmful substances that could be found in
synfuels processes are shown. This list does not include all possible harm-
ful compounds, and each type of process would have a different distribution
of harmful species. Some general conclusions can be reached when comparing
different types of synfuels processes. For example, a process that produces
organic liquids either as primary products or as by-products is more likely
to have some streams containing phenols, aromatics, and polynuclear aromatics
(PNA's) compared to a process that produces only a gaseous product primarily
composed of carbon monoxide and hydrogen. Since trace metals are introduced
into the synfuels processing facility in the coal feedstock, analysis of the
feedstock would show which elements need to be considered for a particular
feedstock.
ASSESSMENT OF EMISSION RATES
The average emission rate for a particular type of source is called the
emission factor. The total number of sources multiplied by the emission fac-
tor equals the total emission rate from that type of source. For assessment
of the severity of VOC emissions, the total emission rate is the main factor
to consider. Other factors become significant when evaluating VOC emission
control strategies, as discussed in the section on fugitive emission
controls.
The total emission rate of harmful species is also important, but the
emission rate of individual sources may also be significant. The effect on
the environment beyond the boundaries of the synfuels plant is primarily
determined by the total emission rate from the plant, but industrial hygiene
considerations within the plant boundaries are also dependent on individual
emission factors for sources. For example, if a particular type of source
has a very high emission factor but there are very few sources present, the
emission contribution may be a small fraction of the total hazardous fugitive
emissions from the plant. However, if plant workers are frequently required
to be in close proximity to this type of source, additional emission controls
or protective equipment for workers may be needed.
300
-------
Each synfuels process will have unique characteristics, and therefore
process-specific (or even site-specific) evaluation of the severity of fugi-
tive emissions will be necessary. Extensive fugitive emission testing has
been conducted in petroleum production and refining facilities, coke by-
product plants and organic chemical plants in the U.S., and a limited fugi-
tive emission test has been performed at a European coal gasification plant
(Ref. 4). The results of these tests show that emission factors for the same
type of source (valve, pump, flange) can vary over several orders of magni-
tude for different types of processes, and significant variations exist for
the same type of process at different locations. Because of this variabil-
ity, it is difficult to estimate emissions from one type of process based on
data obtained from a different process. The primary benefit that these tests
results provide for synfuels processes pertains to development of emission
control strategies as described in the section on fugitive emission controls.
Because of the two types of tests that can be performed, there are two
types of results that are generated in fugitive emission testing; leak
screening and leak rate measurement. Leak screening consists of a method to
identify the relative magnitude of leakage from fugitive emission sources.
Leak rate measurement involves enclosure of a leaking source and measurement
of the pollutant mass emission rate from the source. These two types of
testing results can be combined to develop emission factors. In Table 3,
emission factors are shown for fugitive emission sources in several indus-
tries. These emission factors have units of kilograms per day per source.
Therefore, the total emissions from a particular type of source can be
estimated by multiplying the number of sources by the emission factor.
In Table 3, the 95 percent confidence intervals for the emission factors
are also shown. It is important to consider these confidence intervals since
they indicate that the true emission factor is expected to be found within
these confidence intervals 95 percent of the time. If confidence intervals
(for different sources or processes) overlap, it is not possible to state
that the true emission factors for the different sources or processes are
significantly different.
In addition to their contribution to emissions of harmful substances,
fugitive emissions also need to be included in De Minimus calculations. If
the total controlled emission rate from a plant exceeds the De Minimus level
for a particulate pollutant, PSD review is required. De Minimus levels for
several pollutants are shown in Table 4 in metric tons (Mg) per year. As an
example of how fugitive emissions might contribute to De Minimus levels, the
total hydrocarbon and carbon monoxide emission rates estimated for the Kosovo
coal gasification plant are shown in Table 5. Hydrocarbon emissions of 5.72
kg/day would be 2.1 metric tons/year, which is approximately 6 percent of the
De Minimus level of 36 tons/year of volatile organic compounds.
FUGITIVE EMISSION CONTROLS
Fugitive emission controls can be categorized as either work practices
or engineering controls. Work practices are specific work activities whose
objective is to prevent emissions, to reduce the potential for emissions, to
301
-------
TABLE 3. COMPARISON OF EMISSION FACTORS
LO
O
ro
Source Type
Service
Valves
Hydrogen
Gas/Vapor
Light Liquid
Heavy Liquid
Aqueous
Puraps
Light Liquid
Heavy Liquid
Aqueous
Compressors
Hydrogen
Gas/Vapor
Flanges
Open-ended lines
Relief Valves
Drains
Emission Factor (95% confidence Interval) for
Various "types of Industrial Processes (kg/day-source)
Kosovo coal Organic Chemicals Manufacturing**
gasification plant* Vinyl Acetate Cumene Ethylene
.0019(. 0002, .016) .05(.011,.33) .12(.033, . 54) .26(.087, .76)
( .0026(.0008,.01) .0033(.001,.022) .061( .022, . 22) . 22(.076, .65)
.0012(Neg*****,.12) - - -
i.005(.009,.03) .047(.001,1.1) .57(.011,2. 9) .75(.065,8. 7)
.0012(Neg,.0021) - - -
,00036(Neg,.02) - -
-
-
- - -
Refineries***
.20(.076,.49)
.64(.33,1.2)
.26(.18,.39)
.0054(. 0022, .016)
2.7(1.7,4.0)
.50(.21,1.2)
1.2(.54,2.5)
15(7.2,32)
.006K. 0022, .027)
.054(.017,.17)
2.07(.76,5.3)
.76(.25,2.2)
Natural
Gas Plants****
j.48(.2,l)
l-5(.5,4)
4.9(.7,30)
.026(.01,.05)
•53(.2,1)
4. 5(. 1,100)
-
Total hydrocarbons from Reference
** Nonraethane hydrocarbons from Reference 5
*** Nonmethane hydrocarbons from Reference 6
**** Total hydrocarbons from Reference 7
*****NegllgIble = <3 x 10~5 kg/day-source
-------
TABLE 4. DE MINIMIS LEVELS TRIGGERING PSD REVIEW
Pollutant
De Minimis Level*
metric tons/yr
Carbon monoxide
Nitrogen oxides
Sulfur dioxide
Ozone
Lead
Asbestos
Beryllium
Mercury
Fluorides
Sulfuric acid mist
Hydrogen sulfide (H2S)
Total reduced sulfur (including
91
36
36
36 (as volatile
organic compounds)
0.5
0.006
0.0004
0.09
2.7
9
9
* 40 CFR 51-52.
303
-------
TABLE 5. TOTAL EMISSION RATES FROM KOSOVO FUGITIVE EMISSION SOURCES (from Ref .
O
.p-
Source Type/Service
(Emission Factor -
kg/day-source)
Valves/gas
(0.0019)
Valves/hydrocarbon
liquid
(0.0026)
Val ves/aqueoua
(0.0012)
Flanges/gas
(0.00024)
Flanges/hydrocarbon
liquid
(0.00048)
Flanges/aqueous
(< 3 x 10-5)
Pumps/hydrocarbon
liquid
(0.005)
Pumps /aqueous
(0.0012)
TOTAL
Carbon Monoxide
Emissions
Valves/gas
(0.0012)
Process Unit
Rectlsol Phenosolvan Tar Separation Gas Cooling Gasification Total
Number Number Number Number Number AA Number vv
Sources Emissions Sources Emissions Sources Emissions Sources Emissions Sources Emissions Sources Emissions
169 0.32 37 0.07 33 0.06 85 0.16 99 0.19 423 O.BO
544 1.41 112 0.29 176 0.46 68 0.18 0 0 900 2.34
7o 0.09 209 0.25 56 0.07 0 0 140 0.17 483 0.58
558 0.13 133 0.03 151 0.04 263 0.06 355 0.09 1460 0.35
1459 0.70 494 0.24 524 0.25 292 0.14 0 0 2769 1.33
208 <0.006 916 <0.03 165 <0.005 0 0 273 <0.008 1562 <0.05
33 0.17 8 0.04 6 0.03 0 00 0 47 0.24
2 <.003 7 0.01 8 0.04 0 06 0.01 23 0.03
2.83 - 0.96 - .92 - O.S4 - 0.46 - 5.72
169 0.20 37 0.04 33 0.04 85 0.10 99 0.12 423 0.51
* .isrd on two of six gaslficrs In operation.
** DC* not Include emissions from open-ended lines, sampling purges, or wastewater systems.
-------
identify emitting sources and to mitigate emissions from these sources.
Engineering controls include equipment design, equipment operation, and
equipment specification procedures that either reduce emission potential of
sources or capture and control emissions from the sources. The effectiveness
of fugitive emission controls can be dependent on many variables, and in some
cases effectiveness cannot be assessed before applying the controls.
WORK PRACTICES
Work practices that can prevent or reduce the potential for fugitive
emissions are sometimes called "housekeeping" practices. The procedures are
implemented in all types of industrial plants in order to reduce safety and
fire hazards, and they can also be applied to reduce fugitive emissions.
Specific procedures regarding process fluid spills and spill cleanup can be
used to minimize fugitive emissions from these sources. Fugitive emissions
from equipment draining, purging and venting operations can be minimized by
specifying procedures that prevent or reduce the emissions. Atmospheric
contact with these process fluids may occur at the point of discharge from
process equipment or in the wastewater systems. Process fluids that are
drained, purged or vented from process equipment can be collected for
recycle, disposal, or pollutant destruction instead of allowing the process
fluids to become atmospheric emissions. Work practices that identify emit-
ting sources and apply emission reduction techniques are generally referred
to as leak detection and repair programs. Fugitive emission tests have con-
sistently shown that a large fraction of total emissions are contributed by a
small fraction of the total number of sources. Therefore, periodic repair or
replacement of all sources would be a very inefficient approach to fugitive
emission control. Leak detection methods provide a way to identify which
sources are contributing the bulk of emissions and therefore warrant emission
reduction efforts. Leak detection and repair programs can be applied to
these sources of process fluid leakage: valves, pumps, compressors, agita-
tors, flanges, relief valves, and open-ended lines.
Leak detection methods include individual component surveys, area (walk-
through) surveys, and fixed-point monitors. They are described in this order
because the first method is also included as part of the other methods. In
the individual component survey, every fugitive emission source (pump, valve,
compressor , etc.) is checked for evidence of process fluid leakage at regu-
lar intervals (monthly, quarterly, yearly, etc.). The method used to detect
leakage may involve sensory examination, soap bubbles spraying, or instrument
techniques. Liquid leaks, especially pump seal failures, can be readily
detected visually, but the liquid leak may be water or other unimportant com-
pounds. High pressure leaks may be audible, and leakage of odorous compounds
can sometimes be detected by smell. These sensory techniques are only useful
for identifying very large leaks.
An individual component survey using soap bubbles involves spraying a
soap solution on the area of potential leakage and observing any bubble for-
mation caused by a gaseous leak. This technique is fairly rapid and
inexpensive, but it is not applicable to moving shafts, hot sources (above
305
-------
100°C), cold sources (below 0°C), or sources where leaks of compounds other
than pollutants could give a false indication of leakage.
Instrument techniques require the use of some type of portable pollutant
detector. The probe of the detector is traversed around the potential leak
areas, and an increase in the detected pollutant concentration identifies the
leak. Various types of detectors can be used for an instrument survey. An
appropriate "action level" or leak definition is chosen, and all sources that
exceed this level are repaired or replaced in order to reduce the leakage
from the source. In the development of fugitive emission regulations, the
most commonly selected "action level" has been 10,000 ppmv. In Table 6, the
percent of sources that would be expected to exceed this action level and
r.equire repair is shown for several types of sources in different industries.
The results of a leak detection survey show which types of sources have the
most significant leaks.
A walk-through survey involves periodic leak detection by using a por-
table pollutant detector for measurement of ambient pollutant levels in the
process unit. Areas that are found to have elevated pollutant concentrations
are then subjected to individual component surveys in order to locate the
leakage sources for repair. Fixed-point monitors have permanent pollutant
detectors operating throughout the process unit. If elevated pollutant
levels are detected, individual component checks are used to find the sources
needing repair.
Once a source has been identified as a leak requiring repair, appropri-
ate action is taken to reduce or eliminate the leakage. Repair methods vary,
depending on the type of source, and source replacement is also a repair
option. Most pumps have spares that can be operated while the pump is out of
service for repair. Many compressors do not have spares, and if the seal
repair required a shutdown of the process unit, temporary emissions due to
the shutdown could exceed the emissions from the seal if it was not repaired
until the next scheduled shutdown. Leaks from packed seals or pumps, com-
pressors, agitators, and valves may be reduced by simple tightening of the
packing. Mechanical seals require removal from the equipment for repair or
replacement. Grease injection in some types of valves may reduce leakage.
Leaks from open-ended lines can be reduced by closing the valve seat more
completely.
Leak detection and repair for cooling towers would require the use of
periodic or continuous monitoring of pollutant concentrations in the cooling
water. Elevated concentrations would indicate leakage, but individual pro-
cess equipment such as heat exchangers would be difficult to pinpoint as the
source of the leak.
ENGINEERING CONTROLS
Engineering controls involve the use of equipment that can capture and
control emissions, or that prevents emissions. Each type of source requires
assessment of operating conditions and constraints in order to determine
which types of engineering controls are applicable.
306
-------
TABLE 6. COMPARISON OF PERCENT OF SOURCES LEAKING
Source Type
Service
Valves
Hydrogen
Gas/Vapor
Light Liquid
Heavy Liquid
Pumps
Light Liquid
Heavy Liquid
Compressors
Hydrogen
Gas /Vapor
Flanges
Open-ended lines
Relief valves
Drains
Percent of Sources Leaking*
Kosovo Coal Organic Chemical
Gasification Plant** Manufacturing***
14
11.4
1 6.5
0.4
12 8.8
' 2.1
9.1
0.5 2.1
4 3.9
3.2
3.8
for Various Types of
Refineries****
20.8
12.6
11.4
0.2
24.0
3.8
44.6
57.0
0.54
7.7
8.6
4.7
Industrial Processes
Natural
Gas Plants*****
J16.4
29.7
52.8
3.1
11.9
17.5
17.0
Coke By— product
Plants******
3.23
20.5
g.g*******
0.0
-
-
* Leaking defined as a screening value >10,000 ppmv
** from Reference 4
*** from Reference 5
**** from Reference 6
***** from Reference 7
****** from References 8, 9, and 10
*******Exhausters
-------
Engineering controls for pumps include sealless pumps, double mechanical
seals, and closed vent systems. Sealless pumps such as diaphragm pumps or
"canned" pumps do not have a shaft/case junction that is exposed to process
fluid. Therefore the potential to emit is eliminated, although these pumps
have operating limitations that prevent universal application. Double mech-
anical seals consist of two mechanical sealing elements with a barrier fluid
in a chamber between the seals. This barrier fluid system can be operated to
purge into the process fluid to prevent leaks or to dissolve any seal leakage
in the barrier fluid. Leakage dissolved in barrier fluids can be emitted by
degassing from the barrier fluid reservoir. Closed vent systems can be used
to transport pump seal leakage to a control device such as a combustion
source or vapor recovery system. Closed vents can be connected to the oil
reservoir degassing vent or to an enclosure fitted to the pump case in order
to contain seal leakage.
Engineering controls for compressors are similar to those for pumps. In
addition to double mechanical seals, some types of compressor seals may also
have oil reservoir degassing vents that can be connected to closed vent sys-
tems. Many reciprocating compressors have closed vent systems to transport
seal leakage to a safe release point. These vents can also be connected to
control devices. Engineering controls for agitators are similar to those for
pumps and compressors.
Fugitive emissions from relief valves can be controlled with rupture
discs, resilient seat relief valves, and closed vent systems. A rupture disc
upstream of the relief valve will prevent leakage through the valve seat, but
the disc must be replaced after each overpressure release. Resilient seat
relief valves may have superior ability to re-seat after overpressure re-
lease, compared to rigid seat relief valves, but no test data are available
to verify this advantage. Closed vent systems are frequently used to trans-
port relief valve discharges to recovery or disposal systems. These closed
vent systems would also convey any fugitive leakage to the control device.
Leaks from open-ended lines are the result of leakage through a valve
seat to the atmosphere via the open-ended line. These leaks can be mini-
mized by installing a cap, plug, blind flange or another valve to the open-
ended line. These devices would be opened only when the lines were put into
service for draining or purging.
i
Process fluids that are purged from sampling connections can be con-
trolled by using closed loop sampling systems. The closed loop system is
operated such that the process fluid purge is either returned to the process
or is collected in a closed vessel for eventual recycle or disposal.
Engineering controls for valves provide an internal barrier to prevent
contact of process fluid with the valve stem. Diaphragm valves and bellows
sealed valves have a moveable internal seal to resist leakage. Operating
constraints limit the applicability of these types of controls.
308
-------
Engineering controls for wastewater systems are primarily aimed at iso-
lating the contaminated wastewater from the atmosphere. Drains with liquid
traps prevent atmospheric contact with the drain system vapor space. Covers
for wastewater separators and transport systems also reduce atmospheric con-
tact. A closed vent system connected to the wastewater system vapor space
would provide the best control potential, but would be difficult to apply.
CONTROL EFFECTIVENESS
Control effectiveness for fugitive emission controls is dependent on
many variables. Some of those variables cannot be estimated before actually
applying the control method. It is possible to estimate the maximum achiev-
able control effectiveness, although actual effectiveness will probably be
lower. Table 7 shows the estimated control effectiveness for work practice
and engineering controls.
Leak detection and repair programs are subject to many variables. The
frequency of inspection, leak definition, interval between leak detection and
repair, repair effectiveness, occurrence rate, and recurrence rate are all
related to the overall effectiveness of leak detection and repair programs.
The costs and benefits of changing these variables will determine the optimum
control strategy for each process unit. Limited repair studies for valves in
the organic chemicals industry indicate that a 70 percent emission reduction
can be achieved by using "directed maintenance". Directed maintenance
requires use of the pollutant detection instrument during repair in order to
determine the success of repair immediately. This method has been shown to
be much more effective compared with undirected maintenance, where the repair
is completed and then the source is re-checked for evidence of leakage.
The effectiveness of double mechanical seals can approach 100 percent if
the barrier fluid is at higher pressure than the process fluid. Closed vent
systems are dependent on the control efficiency of the device to which emis-
sions are transported.
Rupture discs provide 100 percent control, but must be replaced after
overpressure release or deterioration of the disc. Closed vents are connec-
ted to control devices capable of handling overpressure relief discharges.
Depending on the turn down capability of the control device, effectiveness
can range from 60 to 90 percent.
Because the controls for open-ended lines, sampling connections, and
valves essentially eliminate the source of emissions, control effectiveness
approaches 100 percent. The achievable control effectiveness for wastewater
systems is difficult to estimate because each system is different and all
systems are complex with numerous potential emission points.
CONCLUSIONS AND RECOMMENDATIONS
Synfuels production facilities will have the same types of fugitive
emission sources that are currently found in U.S. petroleum production and
refining facilities, organic chemical plants, and coke by-product plants.
309
-------
TABLE 7. MAXIMUM POTENTIAL EFFECTIVENESS OF FUGITIVE EMISSION CONTROLS
Source Type Control Effectiveness
Control Method (percent reduction)
Valves, pumps, compressors, agitators
flanges, open-ended lines, relief valves
Leak detection/directed maintenance 70*
Pumps, compressors, agitators
Double mechanical seals 100
Sealless equipment 100
Closed vent systems 90
Relief valves
Rupture discs 100
Resilient seat valves -**
Closed vent systems 60-90
Open-ended lines
Caps, plugs, blinds, valves 100
Sampling connections
Closed loop sampling 100
Valves
Diaphragm/bellows seal 100
Wastewater systems
Trapped drains -**
Covered systems -**
Closed vent system 90-100
* Based on test data for valves in organic chemical industry, Reference 11.
**Effectiveness not estimated
310
-------
The magnitude and severity of fugitive emissions from synfuels facili-
ties will be dependent on various factors, some of them process- or site-
specific. Fugitive emission testing in U.S. industries has provided a basis
for developing test strategies and control techniques for synfuels facili-
ties.
Because the sources of fugitive emissions are the same, emission control
techniques identified for U.S. industies should also be applicable to syn-
fuels facilities. The applicability and control effectiveness of these con-
trols will also require a case-by-case analysis for each facility. The
developing nature of this industry in the U.S. provides an excellent oppor-
tunity to develop and evaluate fugitive emission controls throughout the
development of a process from design to commercialization.
The most significant recommendations that can be made regarding fugitive
emissions in the U.S. synfuels industry are related to control strategy
development. As each technology develops, the process of providing fugitive
emission assessment and control can also develop.
In the design phase of a process, streams that will require fugitive
emission control can be identified. Design changes to minimize the number of
sources or to make the sources accessible for inspection and repair can be
initiated. As the process moves into the pilot plant stage, fugitive emis-
sion testing can be applied to estimate the severity of the problem and to
identify areas where special emission control efforts are needed. During
pilot plant operation, different types of engineering controls can be eval-
uated, especially for sources in severe service due to temperature, abrasive
fluids, hazardous compounds, etc. Pilot plant experience can be valuable for
evaluating seal lifetimes, repairability of sources, and other factors that
determine the most cost-effective fugitive emission control strategy.
Evaluation of fugitive emission controls throughout the development of
synfuels process should result in a well-defined, effective control strategy
that will be implemented upon start-up of full-scale facilities. Proper
assessment of hazards and cost effectiveness of controls will prevent delays
in obtaining approval of emission controls for operating permits.
311
-------
REFERENCES
1. Cleland, J.G., and G.L. Kingsbury. Multimedia Environmental Goals for
Environmental Assessment, Volume II MEG Charts and Background Informa-
tion. EPA-600/7-77-136b. Research Triangle Institute, Research
Triangle Park, NC November 1977.
2. Environmental Protection Agency. Multimedia Environmental Goals for
Environmental Assessment; Volume III. MEG Charts and Background
Information Summaries (Categories 1-12). EPA-600/7-79-176a. August
1979.
3. Kingsbury, G.L., R.C. Sims, and J.B. White. Multimedia Environmental
Goals for Environmental Assessment; Volume IV. MEG Charts and Back-
ground Information Summaries (Categories 13-26). EPA-600/7-79-176b.
August 1979.
4. Radian Corporation. Fugitive Emission Testing at the Kosovo Coal
Gasification Plant, Draft. 81-219-075-02. August 1981.
5. Langley, G.J., S.M. Dennis, L.P. Provost, and J.F. Ward. Analysis of
SOCMI VOC Fugitive Emissions Data. EPA-600/2-81-111. Radian
Corporation, Austin, TX, June 1981.
6. Radian Corporation. Assessment of Atmospheric Emissions from Petroleum
Refining, Appendix B: Detailed Results. 80-219-144-33. April 1980.
7. DuBose, D.A., J.I. Steinmetz, and G.E. Harris. Emission Factors and
Leak Frequencies for Fittings in Gas Plants, Draft Final Report.
80-FOL-l. Radian Corporation, Austin, TX, September 1981.
8. Wiesenborn, D.P., J.I. Steinmetz, and G.E. Harris. Emissions Test
Report Fugitive Emissions Testing at the Bethlehem Steel Bethlehem
Plant. EMB No. 80-BYC-9. Radian Corporation, Austin, TX, September
1981.
9. Wiesenborn, D.P., J.I. Steinmetz, and G.E. Harris. Emission Test Report
Fugitive Emissions Testing at the Wheeling-Pittsburgh Steel Monessen
Plant. EMB No. 80-BYC-ll. Radian Corporation, Austin, TX, September
1981.
10. Wiesenborn, D.P., J.I. Steinmetz, and G.E. Harris. Emission Test Report
Fugitive Emission Testing at the Republic Steel Gadsden Plant. EMB No.
80-BYC-10. Radian Corporation, Austin, TX, September 1981.
11. Langley, G.J., and R.G. Wetherold. Evaluation of Maintenance for
Fugitive VOC Emissions Control. EPA-600/52-81-080. Radian Corporation,
Austin, TX, May, 1981.
312
-------
CONTROL SYSTEMS FOR AIR EMISSIONS FROM COAL GASIFICATION
By: Sid Thomson
Fluor Corporation
Irvine, CA 92730
ABSTRACT
This paper discusses control systems somewhat unique to coal conver-
sion processes. The main subjects covered will be the control of emissions
resulting from both the loading of gasification reactors and from the
removal of acid gas from the raw process gas. Alternate control systems
will be identified and difficulties in establishing Best Available Control
Technology (BACT) will be addressed.
GASIFICATION REACTORS
Gasification reactors consist of primarily two types with regard to
coal feeding: continuous and intermittent.
With continuous feeding, a coal slurry is usually the feedstock. Coal
gasifiers utilizing slurry feeds are fed under pressure, thereby eliminat-
ing the need for a coal lock hopper. Air emissions from the gasifier
feeding operation are eliminated, since the process occurs in a totally
enclosed system.
With intermittent feeding, dry coal is usually the feedstock. This
type of feeding requires the use of a coal lock hopper (Figure 1). The
various operations required in intermittent dry coal feeding produce emis-
sions that necessitate control to mitigate their environmental impact.
The sequence of the coal lock hopper operation consists of loading,
isolating, pressurizing, unloading, isolating, depressurizing, and restart-
ing the cycle. This cycle operates continuously, even though the coal is
fed into the gasifier intermittently. To demonstrate this operation and
the resultant emissions, assume that Step 1 begins when the coal dump has
been completed and the bottom valve has been closed. At this point, the
coal lock hopper is filled with reactor gas at reactor pressure. The next
step in the cycle is the depressurizing of the coal lock hopper. These
gases can be accumulated in a low-pressure vessel from which they may be
transferred by compression to the product gas or fuel gas systems (Figure
2). When the vented gas is utilized as fuel gas, it must be treated to
remove sulfur compounds before or after combustion due to its high sulfur
content.
313
-------
To Control
System
To
Atmosphere
Repressuring
Figure 1. Coal Lock Hopper
314
-------
FM
Lock Hopper
Gas Storage
Alternate I
Combustion
Process Gas
Gas
Treating
Fuel Gas
Alternate II
Flue Gas
Treating
Alternate III
Atmosphere
Figure 2. Coal Lock Hopper Vent Gas Treating
315
-------
An evacuation step is essential since the coal lock hopper can only be
depressurized to slightly above atmospheric pressure. This step depletes
the amount of gas remaining in the coal lock hopper and ensures that gas
does not flow from the hopper when the upper coal feed valve is opened.
Evacuation is continued throughout the coal loading operation to ensure
that no explosive mixture occurs as the coal is introduced.
Three methods of air emission control have been proposed for handling
the evacuated material. The first, and most popular, method is direct
venting of the gas through an evacuation jet, since this stream would
contain very little total contaminants. A second method is routing the
discharge of the evacuation jet through a scrubber for removal of the
contaminants (Figure 3). A loss of evacuation jet motive force can cause
an explosive mixture to occur in the system, thereby creating an explosion
hazard. Care must be take to prevent the risk of creating this hazard when
evacuating gas from the coal lock hopper. The protection against this
hazard creates expenditures which are difficult to justify due to the small
amount of contaminants prevented from entering the atmosphere. The third
method is pressurizing the coal lock hopper with inert gas and maintaining
the pressure in the hopper above reactor pressure during the reactor coal
feed cycle. This method necessitates an extensive system to compress the
gas and introduce it into the coal lock hopper as required. Additional raw
gas feed processing is needed to remove the recycled inert gas required to
guarantee that reactor gas does not diffuse into the coal lock hopper.
Clean Gas
Evacuated
Gas
Solvent
Stream
Air
or
Other Inert
Motive
Force
Figure 3. Treatment of Evacuated Gas
Rich Solvent
316
-------
Table 1 compares the differences in emissions between the control
system utilizing raw gas pressurizing and the control system utilizing
inert gas pressurizing.
TABLE 1. EMISSIONS FROM COAL LOCK HOPPER
PRESSURIZING MEDIUM RAW GAS
INERT GAS
(1)
COAL TYPE:
PLANT SIZE:
CONTROL SYSTEM:
NO. 6 ILLINOIS
270 BILLION BTU/DAY
UNCONTROLLED 98% RECOVERY
NOT STATED
250 BILLION BTU/DAY
RECOVERY
EMISSIONS:
SULFUR
(T/D)
C2 + HYDROCARBONS
METHANE
1
3
21
,930
,160
,160
35
70
429
400 PSIG^250
25(2)
4,000UJ
PSIG
(1)
EPA 450/2-78-012 Guideline Series.
Listed as hydrocarbons.
RAW GAS TREATING
Competitive gas treating processes for H2S and C02 removal from the
raw gas stream are shown in Table 2.
TABLE 2. ACID GAS REMOVAL PROCESSES
Physical Absorption Processes
Rectisol
Purisol
Selexol
Fluor Solvent
Estasolvan
Chemical Absorption Processes
• MEA
Fluor Econamine
• Benfield
Solvent Used
Methanol
N-methyl-2-pyrrolidone
Dimethyl ether of polyethylene glycol
Propylene Carbonate
Tri-n-butyl phosphate
Solvent Used
Monoethanolamine
Diglycolamine
Potassium Carbonate Solution
The three processes receiving the most attention in the treating of
raw gas from coal gasification are Rectisol, Purisol, and Selexol (Figure
4). Rectisol has the advantage in that it uses a methanol solvent which is
317
-------
Clean Gas Stream Out
Acid Gas Rich
Gas Stream In
Absorber
Cooling
Heat
Depressuring
Acid Gas Rich
Solvent
Acid Gas Out
Stripper
Lean Solvent
Figure 4. Typical Gas Treating Process
318
-------
manufactured in plants that produce synthol liquids and methanol. Results
of research continuously being conducted improve the performance of exist-
ing processes and are used to derive new processes for the removal of acid
gas. This research work may change the favorability of the processes.
These acid gas removal processes can be operated in two modes: selec-
tive and nonselective. In the selective mode, the acid gas is removed in
two streams. One stream of C02 is highly concentrated with H2S and the
second stream contains small amounts of H2S in the C02. The selective
operating mode is accomplished by the use of either two absorption steps
and two stripping steps or in one absorption step with two stripping steps.
Unfortunately, at different operating conditions, none of the solvents
removes all of the H2S without a large amount of the C02 also being re-
moved. For this reason, numerous processing operations must be considered
for removal of sulfur compounds to prevent their escape to the atmosphere.
Figures 5 through 14 demonstrate ten methods of removing sulfur compounds
from the acid gas streams based on selective and nonselective modes of
operation for the Rectisol Process. By utilizing the Selexol and Purisol
Processes, 20 additional processing operations can be drawn and by shifting
the processes into different positions, a number of other operations can be
devised.
Process Gas
Raw Gas
Rectisol
Acid
Gas
Stretford
T
CO2+COS, CS2
Hydrocarbons
Sulfur
Figure 5. Raw Gas Treating Alternate I
319
-------
CO
K>
O
Process Gass
Raw Gas
Rectisol
Acid
Gas
Figure 6. Raw Gas Treating Alternate II
Process Gas
Raw Gas
Rectisol
Acid
Gas
CO2 Hydrocarbons
Adip
Acid
Gas
Claus
Sulfur
Tail
Gas Reducing Gas
f
Beavon/
Stretford
CO2
Sulfur
CO2 Hydrocarbons
Adip
Claus
Sulfur
H2S
Recycle
Tail Reducing Gas-
Gas
I
Scot
CO2
Figure 7. Raw Gas Treating Alternate III
-------
Process Gas
Raw Gas
Rectisol
Acid
Gas
Combustion
SO2
CO2
Wellman
Lord
CO2 _
Figure 8. Raw Gas Treating Alternate IV
H2SO«
u>
ro
Process Gas
Raw Gas
I
Rectisol
Acid
Gas
Combustion
SO2
CO2
Takahax
1
CO2
Sulfur
Figure 9. Raw Gas Treating Alternate V
Process Gas
Raw Gas
CO2 Gas
1
Selective
Rectisol
Acid
Gas
Combustion
Flue
Gas
FGD
CO2
I
Sludge
Figure 10. Raw Gas Treating Alternate VI
-------
Process Gas
HC & CO2
Raw Gas
Rectisol
Acid
Gas
Adip
Acid
Gas ,
S
Claus
L
O2
Wellman
Lord
CO2
Sulfur
Figure 11. Raw Gas Treating Alternate VII
Process Gas
Raw Gas
Rectisol
Acid
Gas
Combustion
Reducing Gas _
SO2
IFP
CO2
Sulfur
Figure 12. Raw Gas Treating Alternate VIII
-------
Process Gas
Raw Gas
Rectisol
Acid
Gas
CO2
I
Adip
Acid
Gas
Claus
Sulfur
! l 1
FGD
T-J
I Sludge
Figure 13. Raw Gas Treating Alternate IX
Process Gas
Raw Gas
CO2 Gas
Selective
Rectisol
Acid
Gas
Claus
\
Tail
Gas
Sulfur
Reducing Gas
Beavon/
Stretford
CO2
Reducing
Scot
—I
Wellman
Lord
Figure 14. Raw Gas Treating Alternate X
-------
Some prescreening must be done before designs and estimates proceed to
perform BACT analysis in a reasonable period of time. The first prescreen-
ing step is the elimination of the processing operations which will not
meet the emission regulatory requirements of New Source Performance
Standards (NSPS) or Prevention of Significant Deterioration (PSD).
It is under the PSD regulations that modeling of the air dispersion
characteristics of the plant site are required to estimate the amount of
allowable emissions. Once this estimate is determined, those process
operations which will not comply with these regulations can be eliminated
from consideration.
The next prescreening step is the elimination of those processing
operations that have been determined unable to meet the cost-effective
demands on previous studies. Following this step, the remaining processes
are reviewed to determine if they have special requirements which cannot be
satisfied (e.g., availability of the required solvent, difficulty in ob-
taining equipment, excess delivery time for custom-made equipment, etc.).
Finally, a review is conducted regarding the commercial applicability of
the remaining processes to determine whether they have been proven in pilot
plant, semicommercial, or commercial operations. A cost estimate is made
for the two or three remaining process operations resulting from the pre-
screening steps. The most cost-effective operation that satisfies the
regulatory requirements is then selected.
BEST AVAILABLE CONTROL TECHNOLOGY (BACT)
There have been numerous studies made for the Environmental Protection
Agency (EPA) and Department of Energy (DOE) to determine the best control
scheme for given conditions or plant sites. A list of these studies and
the selected acid gas removal and treatment schemes follows:
1. EPA 650/2-74-009-b, June 1974, "Evaluation of Pollution
Control in Fossil Fuel Conversion Processes; Gasifica-
tion: Section 1: Synthane Process," by Esso Research and
Engineering Company. The Benfield Process was selected for
acid gas removal with the Stretford Process for sulfur
recovery. An economic evaluation of the scheme was not
indicated. Selection is assumed to be based on engineering
judgment.
2. EPA 650/2-74-009-c, July 1974, "Evaluation of Pollution
Control in Fossil Fuel Conversion Processes; Gasification;
Section I: Lurgi Process," by Exxon Research and Engineer-
ing. Acid gas treatment was mainly based on the Steams-
Roger design for the El Paso Natural Gas Company. Rectisol
with Stretford Process was selected. The selection was
apparently the result of economic studies conducted by
Stearns-Roger for El Paso.
324
-------
3. EPA 650/2-74-009-b, December 1974, "Evaluation of Pollution
Control in Fossil Fuel Conversion Processes, Gasification;
Section I: C02 Acceptor Process," by Exxon Research and
Engineering Company. The study states "consideration should
be given to using an absorption/oxidation process such as
Stretford, Takahax, IFP, etc., on the raw gas directly."
4. EPA 650/2-74-009-g, May 1975, "Evaluation of Pollution
Control in Fossil Fuel Conversion Process; Gasification:
Section 5. BI-GAS Process," by Exxon Research and Engineer-
ing. Benfield with Claus and tail gas recovery was selected
for acid gas removal. This study did not include an econom-
ic evaluation. Selection was assumed to be based on engi-
neering judgment.
5. EPA 650/2-74-009-j, September 1975, "Evaluation of Pollution
Control in Fossil Fuel Conversion Processes; Gasification:
Section 8. Winkler Process," by Exxon Research and
Engineering Company. Benfield utilizing the selective mode
of operation for acid gas removal was employed. A Claus
Sulfur Plant with a Tail Gas Unit was selected for sulfur
removal from acid gas. No economic evaluation was indi-
cated.
6. EPA 650/2-74-072, July 1974, "Sasol-Type Process for Gaso-
line Methanol, SNG, and Low-Btu Gas from Coal," by M. W.
Kellogg Co. Nonselective Rectisol plus Stretford Processes
for removal of acid gas were utilized. No economic evalua-
tion was indicated. Selection was assumed to be based on
engineering judgment.
7. EPA 600/2-76-101, April 1976, "Evaluation of Pollution
Control in Fossil Fuel Conversion Processes: Final Report,"
by Exxon Research and Engineering Company. For acid gas
removal units, the study states: "Each case must be ex-
amined individually, not only to choose the best type of
acid gas removal process for the particular application, but
also as "to what modification to choose for the best type."
8. EPA-450/2-78-012, March 1978, "Guideline Series Control of
Emissions from Lurgi Coal Gasification Plants," by the
Environmental Protection Agency. The cost study compared:
(1) Selective Rectisol, Stretford Unit on lean H2S stream,
Claus Plant followed by Tail Gas Incinerator on H2S-rich
stream, (2) Nonselective Rectisol, Stretford Unit and Tail
Gas Incinerator, and (3) Selective Rectisol, Stretford Unit
and Tail Gas Incinerator on lean gas stream, Claus Plant
with Tail Gas Incinerator and tail gas scrubbing on H2S-rich
gas stream. A cost analysis indicated that the Nonselective
Rectisol Process with a Stretford Unit was the most accept-
able alternative from cost standpoint with comparable sulfur
recovery efficiency.
325
-------
9. DOE No. FE-2240-50, August 1978, "Sulfur Recovery in a Coal
Gasification Plant," by C. F. Braun. Five different pro-
cessing schemes were evaluated for both western and eastern
coals. The study indicated that the Nonselective Selexol
with Stretford Process and FMC Double-Alkali for boiler gas
treating were the best selections for western (low-sulfur)
coal. Selective Selexol with the Glaus Plant and FMC
Double-Alkali for the Boiler and tail gas treating were
found most favorable for eastern coal. This study had one
significant qualification: "Due to the large number of
available alternatives and the limited number of cases that
have been considered, the conclusions are only tentative."
10. DOE PNL 3140, September 1979, "Assessment of Environmental
Control Technologies for Koppers-Totzek, Winkler and Texaco
Coal Gasification Systems," by Pacific Northwest Laboratory.
Acid gas removal discussions were general in nature.
11. ORNL-5722, August 1981, "The Impact of Environmental Control
Costs on an Indirect Coal Liquefaction Process," by Oak
Ridge National Laboratory/Fluor E & C, Inc., Houston, Texas.
Six different cases were evaluated based on different strin-
gency control and plant sizes. Case 4 contained the most
stringent controls and an evaluation of methods of Boiler
Flue Gas Emission Control. Nonselective Rectisol with a
Flue Gas Desulfurization Unit was selected for the less
stringent cases. Nonselective Rectisol with the Stretford
Process was utilized for the most stringent control.
The various studies discussed indicate that selection of the most
favorable process for acid gas removal and control of sulfur emissions is
dependent upon the gasification process selected and the site location.
Coal gasification plants that utilize a coal-fired boiler for steam
and power production may find it advantageous to integrate the boiler plant
flue gas treating with the acid gas treating. This integration provides
additional alternate schemes for consideration. Sulfur concentrations in
boiler plant flue gases are low when compared to sulfur concentrations in
the raw gas and acid gas streams. A sulfur removal efficiency of greater
than 90 percent from boiler flue gases on a continuous basis places an
excessive burden on the state of the art for some of the FGD processes.
Table 3 illustrates the difference in sulfur concentrations of flue gas and
Lurgi acid gas streams when processing Illinois No. 6 coal. Efficiency of
removal is dependent on inlet flue gas concentration. This factor must be
considered when integrating the acid gas treating system. In some in-
stances, the acid gas stream routed to the FGD Unit may not have sufficient
concentration to justify FGD treatment.
326
-------
TABLE 3. ILLINOIS NO. 6 COAL - SULFUR CONTENT COMPARISON
(CALCULATED FROM ESTIMATED YIELD DATA)
Gas Volume %
Lurgi Raw Gas
1% as H2S
Acid Gas
3.23% as H2S
Boiler Flue Gas
0.21% as S02
The cost of gas produced in a coal gasification plant is not competi-
tive with the current cost of natural gas. Nonjustifiable expenditures
resulting from delays in obtaining permits and from unnecessary environ-
mental control systems create even more of a negative cost impact. Since
synfuels plants are experiencing difficulty in meeting return on investment
requirements essential for financing, every effort must be made to elimi-
nate expenditures caused by unnecessary regulatory requirements.
BACT determinations and PSD regulations often create delays which
outweigh their benefits. The regulations are burdening for both the regu-
lator and those being regulated. Arriving at an agreement on a BACT deter-
mination, containing numerous options, creates never-ending arguments. On
a case-by-case basis, these arguments become extremely burdensome for both
the regulatory agency and the permittee. Allocation of PSD increments to
satisfy all permittees is an assignment given to our regulators even though
it is doubtful that Solomon, the wise man, could find a satisfactory solu-
tion to this problem.
Suggestions for better solutions to environmental regulations are as
numerous as the process operations available for acid gas removal and
treatment. Unfortunately, each solution is usually self-serving for those
proposing the suggestion and does not consider the adverse effects on
others. It is extremely difficult to arrive at a solution that is benefi-
cial to the majority, since an active minority is often a controlling
element in our political arena.
Industry, regulators and environmentalists must cease their role as
adversaries and become partners in establishing regulations that provide
maximum benefit to the majority. Since very few people can say their
interests lie entirely in one direction, it should not be so difficult to
work together for such a worthy goal.
327
-------
Session IV: SOLID WASTE-RELATED ENVIRONMENTAL CONSIDERATIONS
Chairman: David A. Kirchgessner
U.S. Environmental Protection Agency
Research Triangle Park, NC
Cochairman: Kimm W. Crawford
TRW, Inc.
Redondo Beach, CA
328
-------
HEALTH EFFECTS BIOASSAY RESULTS FROM COAL CONVERSION SOLID WASTES*
M. P. Maskarinec, F. W. Larimer, J. L. Epler, C. W. Francis
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
ABSTRACT
To assist EPA and DOE in identifying solid wastes that may pose a poten-
tial hazard to human health and environment, the Oak Ridge National Laboratory
has conducted studies on extracts from solid wastes obtained from various coal
liquefaction and gasification processes. Analytical procedures to chemically
characterize and separate the organic and inorganic constituents were devel-
oped. Various approaches to extraction were compared. Batteries of health
effects and environmental assays were applied to the extracts or fractions
thereof to serve as indicators of chronic hazards. The applicability and com-
patibility of the coupled chemical and biological procedures will be evaluated
with particular emphasis on the Ames mutagenicity test.
INTRODUCTION
Recent examples of improper disposal at various hazardous chemical sites
has dramatically increased the public awareness of the environmental and
health effects associated with the disposal of solid and hazardous wastes (1).
Therefore, increased emphasis has recently been placed on the regulatory
aspects of the transport, treatment, storage and disposal of solid industrial
waste (2) .
At the same time, trends toward increased use of coal reserves in this
country dictates that large volumes of solid wastes will result from various
coal conversion technologies (3f4). These wastes include solids from coal-
cleaning processes, flue-gas disulfurization sludges from ancillary boilers,
spent catalysts, tar and oil sludges, and ash/slags. While the ashes and
slags will constitute the largest volume of waste generated (> 90?), they are
by and large devoid of organic material (5,6). Also, the sorptive capacity of
these materials is usually large (6), and organic matter is not likely to
migrate in the environment by dissolution. Thus, the environmeal and health
consequences of these materials can largely be predicted from studies of
inorganic content and leachability.
In the case of wastewater treatment plant sludges, which will be genera-
ted in considerably smaller but still signficant volumes, the organic content
is likely to be much higher (7). and the leachability of organics from these
solid wastes must be studied with respect to health and environmental effects.
*Research sponsored jointly by the U.S. Environmental Protection Agency (IAG
78-DX-0372) and the Office of Health and Environmental Research, U. S.
Department of Energy under contract W-7405-eng-26 with the Union Carbide
Corporation.
329
-------
This is not to imply that inroganics should be ignored in such wastes, but
only to indicate the presence of a new set of risks.
The steps involved in evaluating the health and environmental effects of
wastewater treatment plant sludges from coal conversion solid wastes include:
1) physical/chemical characterization of the specific wastes; 2) determination
of the environmental mobility of the various chemical constituents of the
waste by evaluation of aqueous extracts intended to simulate specific disposal
scenarios, and 3) preparation of the wastes and aqueous extracts for bioassay.
This work represents a summary of data relevant to these three areas.
MATERIALS AND METHODS
Sample Collection
Samples were collected from operating pilot plants during a steady-state
period Sample 1 was a filtered sludge from a pilot-scale coal liquefaction
wastewater treatment plant, Sample 2 was collected from a coal cleaning plant
and represented the final wastewater treatment plant solid waste. Sample 3
was collected as a centrifuged-residual from a liquefaction wastewater treat-
ment plant.
Generation of Aqueous Extracts of Solid Wastes
Five techniques were used for the generation of aqueous extracts of the
solid wastes. These included the EPA-EP, a distilled water extraction carried
out in a manner identical to the EPA-EP (t^O-EP), a sodium-resin displace-
ment extraction, a citric acid extraction and an upward-flow column extraction
with distilled water. The EPA-EP and the citric acid extractions are intended
to mimic the co-disposal of municipal and industrial solid waste. The distil-
led water EP and sodium displacement techniques are more applicable to the
disposal scenario of 100? industrial waste. The upward flow column extraction
can be used to simulate either scenario depending on the extractant, but is
primarily intended to avoid the artificial solid/solution separations inherent
in the batch extractions, regardless of the extractant used. The variable and
constant factors involved in the extractions are listed in Table 1.
Preparation of Solid Wastes and Extracts from Ames Bioassay
The solid wastes were prepared for the Ames test (8) in two ways. The
solid wastes (50 g) were Soxhlet-extracted for 24 hours using methylene chlor-
ide (9). An aliquot of the Soxhlet extract was concentrated to dryness and
redissolved in 2 ml dimethylsulfoxide. This solution was bioassayed.
In addition, the solid wastes were extracted using a three-step extrac-
tion procedure (10). Briefly, this procedure involves equilibration of the
solid waste with acid, followed by base, followed finally by organic solvent.
Thus, the procedure results in three fractions for bioassay: acids, bases
330
-------
TABLE 1
EXTRACTION PROCEDURES: IDENTIFICATION OF VARIABLE AND CONSTANT LEACHING FACTORS
Variable Factors
OJ
u>
Extraction
1. EP
Initial
Leaching Medium
Distilled deionized
2. Water
3. Na-Resin
4.
5. Column
Distilled deionized
water
Distilled deionized
water with 1 g cal-
culated dry wt
chelex 100/10 g
sample
Mode of
Extraction
Batch:
magnetically
stirred
Batch:
magnetically
stirred
Batch:
magnetically
stirred
pH
Adjustment
Adjust to pH 5
with 0.5 N.
acetic acid -
maximum limit of
2 meq/g sample
None
Adjust to pH 7
with 0.1 N HC1
Treatment of Leachate Solution
for Extract Analysis
Pressure filtered through 0.4
ym nuclepore filter
Pressure filtered through 0.
ym nuclepore filter
Pressure filtered through 0.
ym nuclepore filter
0.5 M citrate buffer Batch:
rotary
extractor
Distilled deionized Column:
water upward flow
Factor
1. Sample particle size
2. Extraction temperature
3. Extraction time:
None
None
Constant Factors
Condition Used
< 9.5 mm
room temperature
Pressure filtered through 0.4
ym nuclepore filter
Leachate from column directly
passed through XAD-2 resin
Batch mode
Column
4. Number of leachings on same
sample
5. Effective solid:solution ratio
24-hour
until effective solid:solution ratio is reached
1
1:20
-------
and neutrals. An aliquot of easch fraction was concentrated to dryness and
redissolved in dimethylsulfoxide for the Ames test.
The aqueous extracts were prepared as follows: a 500 ml aliquot was
adjusted to pH 6.8 using phosphate buffer and to conductivity 20 mS using
sodium chloride. The adjusted extract was passed through a column containing
4 ml XAD-2 resin. The resin was eluted with 20 ml acetone. The acetone was
concentrated to dryness and the residue taken up in 2 ml dimethylsulfoxide.
In the case of the column extraction, the XAD-2 was located directly above the
column. This XAD-2 was extracted in a manner identical to that used in the
batch extractions. An aliquot of the acetone was evaporated to dryness and
taken up in 2 ml dimethylsulfoxide.
Analysis of Wastes and Extracts
All extract and fractions described above were characterized using gas
chrornatography and combined gas chromatography/mass spectrometry. GC was done
on a Hewlett-Packard Model 5736-A gas chromatograph equipped with a flame ion-
ization detector and a H-P Model 3390 integrator. A twenty-five meter fused
silica capillary column (J&W Scientific) was used. GC/MS was done on a H-P
Model 5985-A GC/MS/DS equipped with a similar column.
When possible, the solid waste extracts were applied to a preweighed
filter pad; the solvent was evaporated and the pad reweighed. The difference
was used as a crude indication of the mass of material present.
Ames Mutagenicity Test
The general methodology for the Salmonella/microsome assay has been
described (11). In screening mode, the assay is restricted to two strains:
TA 100, the hisG base-pair substitution in the uvrB rfa pKM101 background and
TA 98, the hisD frameshift, also carrying uvrB rfa and pKM101. The full range
of metabolic activation was examined, however, using microsomal preparations
from both phenobarbital and Arochlor-treated rats.
RESULTS AND DISCUSSION
Analysis of Solid Wastes and Extracts
The results of the characterization work on the solid wastes and extracts
are reported elsewhere (7), however, some general comments are appropriate
here. In terms of the wastes themselves, considerably more organic material
was extracted using the three-step procedure than was extracted by the Soxhlet
extraction. This is true in terms of total mass as well as in terms of the
levels of individual compounds. Qualitatively, the three wastes were similar.
All contained a variety of compounds, although the neutral fraction was
responsible for much of the organic content. All contained aromatic hydro-
carbons and aromatic heterocycles (including nitrogen, oxygen, and sulfur
332
-------
containing species). In addition, all contained significant quantities of
volatile organic compounds. Therefore, any assessment of the health and envi-
ronmental effects of these materials must consider potential inhalation and
air-quality problems. The levels of organics were highest from Sample 3
followed by Sample 1, and finally Sample 2.
The characterization of the aqueous extracts revealed the following gene-
ral trends. The levels of organic materials in the extracts were more closely
related to the technique used for extraction than to the extraction medium.
For example, the extraction of volatile organics appeared to be superior in
the citrate buffer extraction. However, this extraction is carried out in a
closed system. Use of distilled water in the closed extractor produced com-
parable levels of volatile organics. Conversely, the extraction of phenol and
the cresols did not appear to be relatd to the pH of the extractant. While no
one batch extraction procedure was consistently superior in terms of extract-
ing organics, when the organic content of the solid waste was high (e.g.,
Sample 3) all procedures were comparable. When the organic content of the
solid wastes was low, the distilled water-EP appeared to be the most effective
batch extraction technique.
The column extraction consistently extracted higher levels of organic
compounds than did any of the batch extractions. This is due partly to the
fact that no filtration is required, but also partly due to more aggressive
displacement of organic compounds. This is particularly true when considering
nonpolar compounds.
Ames Bioassay Results
The Ames Salmonella mutagenesis bioassay is widely recognized as an indi-
cator of bacterial mutagenesis. It may also be an indicator of potential
mammalian carcinogenesis. The test has the advantages of being relatively
inexpensive, short-term, and simple to perform. The test is primarily sensi-
tive to organic rnutagens; thus, the characterization work described earlier is
directly applicable to the Ames test.
The bioasay results from Samples 1 and 2 are shown in Tables 2 and 3.
The extracts of Sample 1 were all extremely toxic. There was an indication of
mutagenic activity in all but the most toxic extracts. The most active
extract was the acid fraction, showing a non-linear dose-response in TA 98
(with phenobarbital activation) giving a peak mutation induction 15-fold over
the untreated control (Figure 1). The extracts of Sample 2 were non-muta-
genic, and only the Soxhlet extract and the acid fraction showed significant
non-specific toxicity. The extracts of Sample 3 were extremely toxic; even at
a 10-fold dilution these samples were too toxic for assay.
The results of the Ames test on the aqueous extracts of Sample 1 are
shown in Table 4. Again, all exhibited some degree of toxicity. Those
extracts which were not too toxic to test displayed mutagenic activity. The
aqueous extracts of Sample 2 were not active in either strain. Extracts from
Sample 3 were diluted 10-fold and the results are shown in Table 5. All were
mutagenic including the EP extract, which displayed a linear dose-response
(Figure 2), even after dilution.
333
-------
TABLE 2
MUTAGENICITY OF SOLID WASTE SAMPLE 1
Revertants/plate
TA 98
TA 100
OJ
-P-
75
50
25
10
0
75
50
25
10
0
50
0
Soxhlet
Acid
Base
Neutral
Soxhlet
Acid
Base
Neutral
PHENOBARBITAL ACTIVATION
T*
93
94
97
22
T
609
556
391
39
T
T
T
T
22
T
60
61
52
42
T
T
T
T
199
T
670
645
536
167
T
T
T
T
91
T
T
T
T
106
AROCHLOR ACTIVATION
T
T
70
70
22
T
21
T
T
156
202
48
T
40
T
T
T
T
42
T
27
T
T
T
54
35
NO ACTIVATION
T
44
T
T
T
T
192
150
105
T
T
T
229
126
T
159
T
T
T
T
153
T
153
T
T
T
T
123
T
99
-------
TABLE 3
MUTAGENICITY OF SOLID WASTE SAMPLE 2
Revertants/piate
u>
75
50
25
10
0
75
50
25
10
0
50
0
TA 98
Soxhlet
Acid
Base
Neutral
Soxhlet
TA 100
Acid Base
Neutral
PHENOBARBITAL ACTIVATION
41
33
41
47
17
65
58
60
52
47
T
53
41
71
47
57
43
35
32
19
T T
T T
117 227
112 220
104 173
187
187
150
216
173
162
165
163
135
174
AROCHLOR ACTIVATION
T
T
T
47
27
T
29
T
T
T
39
28
T
35
46
52
46
59
28
26
35
54
40
30
32
20
NO ACTIVATION
33
28
T
T
T
107
98
T
118
T 207
T 160
T 138
23 1 205
182 182
T 204
192 192
173
160
151
163
98
92
103
-------
UJ
H
CO
f-
H
ce
UJ
>
UJ
cc
769
668
see
499
309
299
iee
e
8 18 28 38 48 58 68
CONG dil/PLATE)
Figure 1. Mutagenicity of Acid Fraction of Sample 1
(TA 98, Phenobarbital Activation)
336
-------
OJ
OJ
75
50
25
10
0
75
50
25
10
0
50
0
TABLE 4
MUTAGENICITY OF SOLID WASTE EXTRACTS FROM SAMPLE 1
Revertants/plate
TA 98 TA 100
EP
HpO-EP
Citrate
Na-Resin
Column
EP
HpO-EP
Citrate
Na-Resin
Column
PHENOBARBITAL ACTIVATION
T
T
T
44
17
T
T
58
41
52
T
T
T
45
22
T
T
35
33
16
T
T
94
78
42
T
T
T
156
109
T
T
T
T
110
T
T
T
T
154
T
T
T
117
111
T
T
T
149
112
AROCHLOR ACTIVATION
T
T
T
31
13
T
7
T
T
T
42
43
T
42
T
T
T
T
28
T
58
T
T
T
19
8
NO
T
17
T
T
119
48
31
ACTIVATION
T
42
T
T
T
T
96
T
94
T
T
T
T
91
109
76
T
T
T
T
167
T
110
T
T
T
108
94
T
71
T
T
T
117
108
T
111
-------
Cone (yl/plate)
TABLE 5
MUTAGENICITY OF BATCH EXTRACTS FROM SAMPLE 3
Revertants/plate
TA 98
EP H?0-EP Citrate Na-Resin
PHENOBARBITAL ACTIVATION
TA 100
EP HpO-EP Citrate
Na-Resin
U)
U>
00
7.5
5.0
2.5
1.0
0
7.5
5.0
2.5
1.0
0
5.0
0
98
59
53
29
20
7
4
74
52
27
29
18
9
7
71
57
54
49
42
80
86
48
39
40
42
30
T
45
46
70
22
AROCHLOR ACTIVATION
T
T
110
97
29
NO ACTIVATION
27
22
T
T
117
93
85
T
71
136
119
80
163
213
169
152
150
119
165
125
123
106
T
T
T
78
65
T
T
78
70
74
156
165
214
176
135
151
109
182
143
110
T
40
T
30
95
95
120
85
-------
UJ
E-H
cu
-V
c/)
H
cc
UJ
>•
UJ
cc
128
iee
80
68
48
28
8
8
246
CONG (id/PLATE)
8
Figure 2. Mutagenicity of EP Extract of Sample 3
(TA 98, Phenobarbital Activation)
339
-------
SUMMARY
The organic content of three solid wastes, representing coal conversion
wastewater treatment plant sludges, were compared. State-of-the-art analyt-
ical techniques coupled with the Ames mutagenesis bioassay were used. A
three-step fractionation/isolation scheme improved the bioassay results by
isolating toxicity in the "acid" fraction. In addition, the wastes were
extracted using five different environmental mobility tests. The extracts
were analyzed and assayed (Ames test). In general, the results of the Ames
Bioassay parallelled the results of the analytical characterization.
REFERENCES
1. Me Doug all, W. J. , Fusco, R. A., and O'Brien, R. P. "Containment and
Treatment of the Love Canal Landfill Leachate," Paper presented at the
52nd Annual Conference of the Water Pollution Control Federation,
Houston, Texas, October 1979.
2. U.S. Government Resource Conservation and Recovery Act of 1976 Public Law
94-580, 94th Congress, 1976.
3. Griffin, R. H., Schuller, R. M., Soloway, J. J. , Shimp, N. F. , Childers,
W. F., and Shiley, R. H. "Chemical and Biological Characterization of
Leachates from Coal Solid Wastes," Environmental Geology Notes 89, Illi-
nois State Geological Survey. Urbana, Illinois, November, 1980, 98 pp.
4. Holt, N. A., McDaniel, J. E., and O'Shea, T. P. "Environmental Test
Results from Coal Gasification Pilot Plants," Fifth Symposium on "Envi-
ronmental Aspects of Fuel Conversion Technology," St. Louis, Missouri,
September 16-19, 1980.
5. Francis, C. W., Boegly, W. J., Jr., Turner, R. R., and Davis, E. C. "Coal
Conversion Solid Waste Disposal," American Society of Civil Engineers
Conference: "Energy in the Man-Built Environment, August 3-5, 1981,
Vail, Colorado.
6. Browman, M. G. , and Maskarinec, M. P. "Environmental Aspects of Organics
in Coal Conversion Solid Wastes," Environ. Sci. Eng. (Submitted).
7. Maskarinec, M. P., Brown, D. K. , Brazell, R. S., and Harvey, R. W.
"Analysis of Solid Wastes and Associated Leachates from Coal Conversion
Facilities," Anal. Chim. Acta (Submitted).
8. Ames, B. N., McCann, J., Yamasaki, E. "Methods for Detecting Carcinogens
and Mutagens with Salmonella/Mammalian Microsomes Mutagenicity Test,"
Mutat. Res. 3J_, 347-366, 1975.
9. "Test Methods for Evaluating Solid Wastes," USEPA SW-846, 1979.
10. Maskarinec, M. P., and Harvey, R. W. "Screening of Sludges and Solid
Wastes from Organic Compounds," Int. J. Environ. Anal. Chem. (In Press).
11. Epler, J. L., et al. "Toxicity of Leachates," EPA 600/2-80-057, 1980.
340
-------
A COMPARISON OF RCRA LEACHATES OF SOLID WASTES FROM COAL-FIRED UTILITIES t
AND LOW- AND MEDIUM-BTU GASIFICATION PROCESSES
by: Michael R. Fuchs, Donnie L. Heinrich, Larry J. Holcombe, Kishore
T. Ajmera
Radian Corporation
8501 Mopac Boulevard
Austin, Texas 78766
ABSTRACT
EPA has promulgated regulations which temporarily exclude utility
wastes, including fly ash and bottom ash from coal-fired generating
stations, from Subtitle C of Resource Conservation and Recovery Act (RCRA)
regulations. EPA, using broad interpretation of amendments to the act, has
also excluded coal gasification solid wastes from Subtitle C regulations and
these wastes are listed as non-hazardous pending further data evaluation.
This paper presents comparative results of RCRA leachates of the solid
wastes from two low-BTU gasification processes and coal-fired utility solid
wastes. The three facilities from which solid wastes were obtained used the
same lignite feedstock. Also presented are comparable RCRA leachate results
of solid wastes from a medium-BTU gasification process and a coal-fired
power plant, both fueled with identical lignite feedstocks. The results
indicate that solid wastes from coal-fired utilities and the solid wastes
generated directly by low- and medium-BTU gasification processes are
non-hazardous according to RCRA protocol and limits.
INTRODUCTION
On May 19, 1980, EPA promulgated regulations implementing Subtitle C of
the Resource Conservation and Recovery Act (RCRA). The regulations define
solid and hazardous wastes and establish criteria for handling and disposal
of hazardous wastes. Excluded from Subtitle C regulations were fossil fuel
combustion wastes which were then the subject of pending Congressional legi-
slation. The Solid Waste Disposal Act of October 21, 1980 mandated the
exclusion of fossil fuel combustion wastes from Subtitle C regulations and
includes specifically fly ash waste and bottom ash waste of coal combustion
processes (Reference 1).
The exclusion of coal combustion solid wastes from Subtitle C regula-
tions is temporary. A revision of the exclusion of these wastes may be
enacted pending the assessment of the environmental effect of these wastes
by EPA.
Excluded from Subtitle C regulations by amendments to RCRA in November,
1980 were "solid waste from the extraction, beneficiation, and processing
341
-------
of ores and minerals". Under the interpretation that coal is a "mineral or
ore", EPA has excluded solid wastes generated directly by coal gasification
processes from Subtitle C regulations by considering gasification to be
"extraction, beneficiation, and processing" of the "mineral or ore" i.e.,
coal. The exclusion from Subtitle C regulations of gasification solid
wastes applies only to solid wastes produced directly by the gasification
process. Wastes generated refining or upgrading the product are not
excluded. This broad interpretation of the ruling will remain in effect
until EPA has had an opportunity to evaluate the scope of specific exclu-
sions (Reference 2).
Although the exclusion from Subtitle C regulations is temporary for
both coal-fired power plants and coal gasification solid wastes, coal gasi-
fication is a fledgling industry in this nation, with limited data available
to assess the environmental implications of the disposal of solid wastes
generated by the various processes. On the other hand, utilities produce a
substantial portion of the nation's electricity at coal-fired power plants.
The characteristics of coal-fired power plant solid wastes are much more
defined and recognized.
This paper presents data which provide an opportunity to evaluate com-
parative results of RCRA leachates of solid wastes from fossil fuel combus-
tion and solid wastes from coal gasification processes. The data includes
RCRA leachate results of the solid wastes from two low-BTU gasification
processes and the solid wastes from a coal-fired power plant, the feed-
stocks of the three processes from the same mine. Other data presented are
RCRA leachate results of a medium-BTU gasification solid waste and solid
wastes from a coal-fired power plant, both facilities having the same lig-
nite feedstock, but not the same as the low-BTU gasification feedstocks.
RCRA leachate results on the solid wastes from coal-fired power plants
and gasification processes provide data developed from solid wastes produced
under similar conditions. Coal-fired power plant bottom ash and coal gasi-
fication gasifier ash are subjected to the very hot temperatures associated
with each process and are primarily fused ash having a coarse texture. One
of the major differences of the two processes is that solid wastes from
coal-fired power plants are generated in an oxidizing atmosphere while solid
wastes from gasification processes are generated in a reducing atmosphere.
Gasifier ash passes through both reducing and oxidizing zones within the
gasifier. Wet samples of bottom and gasifier ashes are generally collected,
either after sluicing to disposal ponds (bottom ash) or through the water
pressure seal of the gasifier (gasifier ash). Precipitator ash and cyclone
dust are finer particulate matter entrained in the combustion effluents or
product gas. Precipitator ash is collected dry, but may be sluiced to dis-
posal ponds. Cyclone dust samples are most often retrieved dry, but may be
collected wet from water quench systems. The solid wastes from gasification
processes (gasifier ash and cyclone dust) have considerable concentrations
(approximately 20-50%) of carbon, while coal-fired power plant solid wastes
have quite low concentrations (< 1%) of carbon.
342
-------
TECHNICAL APPROACH
As required by RCRA, EPA has established five categories to define the
characteristics of hazardous waste. The five characteristics are:
o General - a solid waste is a hazardous waste if it exhibits any of
the characteristics of hazardous waste.
o Ignitability - ignitable wastes have a low flash point, or are
liable to cause fires or are oxidizers.
o Corrosivity - corrosive wastes have a pH of less than or equal to
2 or greater than or equal to 12 or corrode steel at a specified
rate.
o Reactivity - reactive wastes react violently, generate toxic fumes
or are explosive.
o Extraction Procedure (EP) Toxicity - an extraction procedure is
specified and maximum concentrations of contaminants listed; the
waste is hazardous if the concentration of any contaminant in the
leachate is equal to or greater than the listed contaminant level.
The extraction procedure has been designed to identify wastes which
would leach hazardous concentrations of toxic constituents into ground-
waters under conditions of improper management. The characteristic of EP
toxicity contaminants are presented in Tables 1, 5, and 7 along with the
maximum allowable concentrations of each. The list of contaminants includes
eight elements, four pesticides, and two herbicides.
The conclusions presented in this paper are based upon the charac-
teristic of EP toxicity and primarily upon the inorganic element contami-
nants. The contaminants listed in the EP toxicity characteristic are the
toxic contaminants listed in the National Interim Primary Drinking Water
Standards (NIPDWS). The maximum concentration levels of the EP toxicity
contaminants are ten times the concentrations specified in the NIPDWS.
Radian Corporation is presently under contract to the EPA to conduct an
"Environmental Assessment of Low/Medium-BTU Gasification Technology". As
part of this program, Radian has conducted source test and evaluations at
commercial and pilot scale low- and medium-BTU gasification facilities.
Included in the source test and evaluations, solid wastes of the gasifica-
tion processes (gasifier ash and cyclone dust) have been subjected to the
RCRA extraction procedure with subsequent analyses of the leachates for the
eight RCRA elemental contaminants.
Two low-BTU gasification facilities tested were a Wellman-Galusha gasi-
fier located at the U.S. Bureau of Mines, Twin Cities Metallurgy Research
Center, Ft. Snelling site, in Minneapolis, Minnesota, and the Riley Gas
Producer located at the Riley Research Center in Worcester, Massachusetts.
During testing, both units were operating with North Dakota Indianhead
lignite. Both gasifiers are air-blown, atmospheric pressure units. The
343
-------
major difference between the two gasifiers is that the Wellman-Galusha is a
thick fixed-bed (app. 4 feet) design while the Riley Gas Producer is a thin
fixed-bed (app. 2 feet) design. At the Ft. Snelling site (Welltnan-Galusha),
the cyclone dust is water quenched after removal from the product gas.
To evaluate RCRA results of coal gasification and coal-fired power
plant solid wastes generated from facilities operating with the same feed-
stock, Radian, with the aid of American Natural Service Company, identified
a coal-fired power plant using the Indianhead lignite as the feedstock.
United Power association, headquartered in Elk River, Minnesota, operates a
mine-mouth power plant firing Indianhead lignite in Stanton, North Dakota.
Radian received samples of lignite, bottom ash, and electrostatic precipi-
tator ash from the Stanton Plant. As with the solid wastes of the gasifica-
tion processes, the coal-fired power plant solid wastes (bottom ash and pre-
cipitator ash) were subjected to the RCRA extraction procedure and the
leachates analyzed for the eight RCRA elemental contaminants.
Also as part of the EPA program, Radian has conducted source test and
evaluations at a commercial Lurgi-based coal gasification facility located
in the Kosovo region of Yugoslavia. A coal-fired power plant also operates
at the plant site and utilizes the same coal feedstock as the gasification
facility. During the site testing, samples of the gasifier ash from the
Lurgi gasifiers and bottom ash and precipitater ash from the power plant
were collected. The three samples were subjected to the RCRA extraction
procedure and the leachates analyzed for the eight RCRA elemental
contaminants.
To allow an evaluation of the similarities or dissimilarities of the
feedstocks of the two low-BTU gasification processes, the lignite collected
at each facility was analyzed for proximate and ultimate parameters and the
eight RCRA element contaminants. The solid wastes from these processes were
analyzed for the eight RCRA element contaminants to assess the relationship
between RCRA leachate concentrations of the RCRA element contaminants to the
concentrations of these elements in the solid. Proximate and ultimate
analyses were performed on the solid wastes from the two low-BTU gasifica-
tion processes and the Stanton Plant to review the similarities of the solid
wastes with respect to major components.
RESULTS
LOW-BTU GASIFICATION
As discussed earlier, Radian has performed source test and evaluations
at several low-BTU coal gasification facilities. The results of two facili-
ties presented in this paper have been taken from source test and evaluation
programs performed at the two facilities (Reference 3 and 4). The Riley Gas
Producer STER will be finalized in November of this year. The United Power
Association power plant data and the medium-BTU gasification data have been
generated independently of the above two projects, but have been funded by
EPA.
344
-------
Table 1 presents the RCRA leachate results for the eight RCRA elemental
contaminants of the solid wastes from the two low-BTU gasification processes
and the Indianhead lignite-fired power plant. Also presented are maximum
levels of the eight contaminant elements, any of which exceeded in the RCRA
leachate of the solid wastes characterize the solid waste as hazardous and
regulated under Subtitle C.
The concentrations of the eight RCRA elemental contaminants of the
solid wastes from the two low-BTU gasification facilities and from the
coal-fired power plant are presented in Table 2.
Concentrations of the eight RCRA elemental contaminants and proximate
and ultimate analytical results of the lignite feedstocks from the two low-
BTU gasification facilities and the coal-fired power plant are presented in
Table 3.
Presented in Table 4 are proximate and ultimate analytical results of
the solid wastes from the three processes with Indianhead lignite as the
feedstock to allow a comparison of wastes generated by the two low-BTU
gasification processes and the coal-fired power plant.
Table 5 presents the analyses of RCRA leachates of the Riley Gas Pro-
ducer gasifier ash and cyclone dust for the contaminant pesticides and
herbicides. These are the only RCRA leachates which were analyzed for these
contaminants.
The percent of the total element of each solid waste leached by the
RCRA extraction procedure from the solid wastes from the two low-BTU gasi-
fication processes and from the coal-fired power plant operating on Indian-
head lignite is presented in Table 6.
MEDIUM-BTU GASIFICATION
Table 7 presents the analytical results for the eight RCRA protocol
elements of the RCRA leachates of the Lurgi gasifier ash and the bottom ash
and precipitator ash from a coal-fired power plant located at the plant site
and using the same coal as feedstock. The RCRA elemental contaminant maxi-
mum levels are also presented.
The elemental concentrations of the eight RCRA protocol elements in the
solid wastes from the gasification facility and coal-fired power plant at
the Kosovo site are presented in Table 8.
Table 9 presents the percent of the total element of each solid waste
leached by the RCRA extraction procedure from the solid wastes from the
medium-BTU gasification process and from the coal-fired power plant at the
Kosovo plant.
345
-------
TABLE 1. RCRA LEACHATE RESULTS OF LOW-BTU GASIFICATION
SOLID WASTES AND COAL-FIRED POWER PLANT
SOLID WASTES HAVING SAME LIGNITE FEEDSTOCK
Wei Iman-Galusha
(Ft.
Snelling)
Gasifier Cyclone
Contaminant
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Ash
ug/L
19
1,000**
<7**
1**
7**
<0.6
14
NA
Dust
ug/L
33
1,000**
NA
1**
8**
<0.3
6
NA
Riley Gas
Gasifier
Ash
ug/L
33
680
<0.5
<1
<2
<0.5
6
<0.5
Producer
Cyclone
Dust
ug/L
4
390
<0.5
<1
<2
<0.5
2
<0.5
United Power Assn.
Bottom
Ash
ug/L
8
300
40
<200
300
0.3
<4
25
Stanton Plant
Precipitator
Ash
ug/L
58
920
50
<200
400
0.4
87
30
RCRA
Contaminant
Maximum
Concentrat ion*
ug/L
5,000
100,000
1,000
5,000
5,000
200
1,000
5,000
*Reference 1
**Analysis by spark source mass spectroscopy
NA - not analyzed.
- all other analyses by atomic absorption spectroscopy
-------
TABLE 2. ELEMENTAL CONCENTRATIONS OF LOW-BTU GASIFICATION SOLID WASTES
AND COAL-FIRED POWER PLANT SOLID WASTES*
u>
J^
Wei Iman-Galushat
(Ft. Snelling)
Element
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Gasifier
Ash
ug/g
30**
1900***
0.8
21***
5
1.7**
15**
0.6
Cyclone
Dust
ug/g
63**
630
2
<6.2***
8
<3.2**
14**
2
Riley Gas Producertt
Gasifier
Ash
ug/g
58**
3300***
<3
<0.2***
7
0.0005**
1000
2
26
17
NA
4
<0.8
Precipitator
Ash
ug/g
53
>1000
2
12
18
NA
10
<0.4
*Analysis by spark source mass spectroscopy
**Analysis by atomic absorption spectroscopy
***Analysis by inductively coupled plasma emission spectrooscopy
NA-Not analyzed
tReference 3
ttReference 4
-------
TABLE 3. RCRA ELEMENTAL CONTAMINANTS AND PROXIMATE AND ULTIMATE
ANALYTICAL RESULTS, INDIANHEAD LIGNITE FEEDSTOCKS*
Wellman-Galushat
(Ft. Snelling)
Lignite
Riley Gas
Producertt
Lignite
United Power Assn.
Stanton Plant
Lignite
Proximate Analysis
% Ash 10.91
% Volatile 41.93
% Fixed Carbon 47.16
BTU/lb 10475
% Sulfur 0.61
12.09
42.40
45.51
10630
1.10
9.15
39.69
51.16
10923
1.04
Ultimate Analysis
% Carbon
% Hydrogen
% Nitrogen
% Chlorine
% Sulfur
% Ash
% Oxygen (diff.)
% H20
62.69
4.60
0.91
0.03
0.61
10.91
20.25
63.32
4.31
1.02
0.002
1.10
12.09
17.94
32.8
66.27
4.41
0.75
0.00
1.04
9.15
18.38
32.9
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
6.5***
630**
0.4**
10**
2**
0.4***
1***
1**
23***
430****
3**
3. 2****
2**
0.15***
<0.5***
<4**
11**
>1000**
0.2**
0.2**
1**
NA
<0.1**
0.1**
*dry basis
**analysis by spark source mass spectroscopy
***analysis by atomic absorption spectroscopy
****analysis by inductively coupled plasma emission spectroscopy
NA-not analyzed
TReference 3
TtReference 4
348
-------
TABLE 4. PROXIMATE AND ULTIMATE ANALYTICAL RESULTS OF LOW-BTU GASIFICATION SOLID WASTES
AND COAL-FIRED POWER PLANT SOLID WASTES OPERATING ON INDIANHEAD LIGNITE*
Wei Iman-Galusha**
(Ft. Snelling)
Proximate Analysis
% Ash
% Volatile
% Fixed Carbon
BTU/lb
% Sulfur
Ultimate Analysis
% Carbon
% Hydrogen
% Nitrogen
% Chlorine
% Sulfur
% Ash
% Oxygen (diff.)
Gasif ier
Ash
74.41
6.56
19.03
3747
1.09
24.64
0.41
0.17
0.02
1.09
74.41
0.60
Cyclone
Dust
18.97
24.41
56.62
10717
1.51
67.18
2.32
0.93
0.03
1.51
18.97
9.06
Riley Gas
Producer***
Gasif ier
Ash
62.54
10.59
26.87
4993
1.64
35.36
0.61
0.46
0.003
1.64
62.54
0
Cyclone
Dust
39.93
15.00
45.08
8478
1.67
53.69
1.52
0.82
0
1.67
39.93
2.39
United Power Assn.
Stanton Plant
Bottom
Ash
98.68
3.92
-2.60
0
0.01
0.96
0.04
0.07
0
0.01
98.68
0.24
Precipitator
Ash
99.57
3.98
-3.65
0
0
0.55
0.07
0.04
0
0
99.67
-0.33
*dry basis
**Reference 3
***Reference 4
-------
TABLE 5. RCRA PESTICIDE/HERBICIDE CONTAMINANTS RESULTS OF LOW-BTU
GASIFICATION SOLID WASTES LEACHATES
Contaminant
Endrin
Lindane
Methoxychlor
Toxaphene
2,4-n
2,4,5-TP Silvex
Riley Gas
Gasifier
Ash
(ug/L)
BDL
BDL
BDL
BDL
BDL
BDL
Producer*
Cyclone
Dust
(ug/L)
BDL
BDL
BDL
BDL
BDL
BDL
RCRA
Contaminant
Maximum
Concentration**
(ug/L)
20
400
10,000
500
10,000
1,000
BDL - Below detection limit
Detection Limits:
Endrin <2 ug/L
Lindane <0. 2 ug/L
Methoxychlor <2 ug/L
Toxaphene <100 ug/L
2,4-D <0.8 ug/L
2,4,5-TP Silvex <0.3 ug/L
^Reference 4
**Reference 1
350
-------
UJ
Ln
TABLE 6. LOW-BTU GASIFICATION AND COAL-FIRED POWER PLANT
RCRA LEACHATES OF SOLID WASTES
PERCENT OF TOTAL ELEMENT LEACHED
Wellman-Galusha
(Ft. Snelling)
Contaminant
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Gasifier
Ash
1.3
1.1
<17.5*
0.1
2.8
<0.7*
1.9
NA
Cyclone
Dust
1.0
0.9
NA
>0.3**
2.0
***
0.9
NA
Riley Gas
Gasifier
Ash
1.1
0.4
***
***
<0.6*
<2000*
>12**
***
Producer
Cyclone
Dust
13.3
0.5
***
<0.05*
<0.1*
<50
3.3
***
United Power Assn.
Stanton Plant
Bottom
Ash
0.84
>0.6****
40
<33.3*
35.3
NA
<2*
>62.5**
Precipitator
Ash
2.6
<1.8****
50
<33.3*
44.4
NA
17.4
>150**
NA-not analyzed
*Leachate concentration below lower detection limit.
**Solid concentration below lower detection limit.
***A11 results below lower detection limit.
****Solid concentration above upper detection limit.
-------
TABLE 7. RCRA LEACHATE RESULTS OF MEDIUM-BTU GASIFICATION SOLID WASTE AND
COAL-FIRED POWER PLANT SOLID WASTES HAVING SAME LIGNITE FEEDSTOCK*
Contaminant
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Lurgi
Gasifier
Ash
ug/L
21
1200
<0.5
330
140
<0.2
<4
<1
Power Plant
Bottom Ash
ug/L
9
510
<0.5
130
47
<0.2
<4
<1
Power Plant
Precipitator Ash
ug/L
<3
410
<0.5
140
250
<0.2
<4
<1
RCRA
Contaminant
Maximum
Concentration**
ug/L
5,000
100,000
1,000
5,000
5,000
200
1,000
5.000
^Analysis by atomic absorption spectroscopy
**Reference 1
TABLE 8. ELEMENTAL CONCENTRATIONS OF MEDIUM-BTU GASIFICATION SOLID WASTE
AND COAL-FIRED POWER PLANT SOLID WASTES*
Lurgi
Gasifier Power Plant
Ash Bottom Ash
Element ug/g ug/g
Power Plant
Precipitator Ash
ug/g
Arsenic
Barium
Cadmum
Chromium
Lead
Mercury
Selenium
Silver
<5.7
970
<0.8
100
<8
NA
<6
<0.1
<5.7
280
<0.8
69
<8
NA
<6
<0.1
<5.7
560
<0. 8
86
<8
NA
<6
<0. 1
^Analysis by inductively coupled plasma emissions spectroscopy.
NA-not analyzed
352
-------
TABLE 9. MEDIUM-BTU GASIFICATION AND COAL-FIRED POWER PLANT
RCRA LEACHATES OF SOLID WASTES
PERCENT OF TOTAL ELEMENT LEACHED
Contaminant
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Lurgi
Gasifier Ash
>7.4*
2.5
**
6.6
>35*
NA
**
**
Power Plant
Bottom Ash
>3.2*
3.6
**
3.8
>12*
NA
**
**
Power Plant
Precipitator Ash
**
1.5
**
3.2
>62*
NA
A*
**
NA-not analyzed
*Solid concentration below detection limit.
**All results below detection limit.
353
-------
CONCLUSIONS
Tables 10 and 11 present the percent of the RCRA elemental contaminants
maximum concentration represented by the elemental concentrations in the
RCRA leachates of the solid wastes from the coal-fired power plants and gas-
ification processes. There are no values over ten percent, and only five
values greater than or equal to five percent. The values above five of the
percent of the RCRA contaminants maximum level represented by RCRA leachate
concentrations are:
Stanton Plant - Precipitator Ash - Lead 8%
Stanton Plant - Precipitator Ash - Selenium 8.7%
Stanton Plant - Bottom Ash - Lead 6%
Kosovo Power Plant - Precipitator Ash - Lead 5%
Lurgi Gasifier - Gasifier Ash - Chromium 6.6%
The data indicates that the solid wastes tested from the
Wellman-Galusha Gasifier,
Riley Gas Producer,
Stanton Plant (power plant),
Lurgi Gasifier, and
Kosovo Power Plant
processes should be listed as non-hazardous according to the EP toxicity
characteristic. Lead in coal-fired power plant precipitator ashes appears
to be the single elemental contaminant which contributes most significantly
to the toxicity of the RCRA leachates. This may be explained by the theory
that lead, being a volatile element, is most probably vaporized during com-
bustion of the coal and condenses upon the precipitator ash as the flue
gases cool, thereby enriching the lead concentration in the precipitator
ash.
The concentrations of the pesticide and herbicide contaminants in the
RCRA leachates of the gasifier ash and cyclone dust from the Riley Gas Pro-
ducer were not detected by the instrumental analytical method. This data
indicates that no pesticides or herbicides, either generated by the process
or present in the lignite feedstock, are emitted in gasification solid
wastes.
One of the goals of this paper is to present RCRA leachate results
developed on solid wastes of coal gasification processes and coal-fired
power plants that were using the same feedstock. Table 3 presented proxi-
mate and ultimate results and elemental concentrations of the Indianhead
lignite collected at the two low-BTU gasifiers and the coal-fired power
plant. The proximate and ultimate data indicate that the feedstocks at the
three facilities were quite similar. However, the elemental concentrations
indicate considerable variability in the three feedstocks with respect to
the eight RCRA elemental contaminants.
354
-------
(jG
01
TABLE 10. LOW-RTU GASIFICATION AND COAL-FIRED POWER PLANT
RCRA LEACHATES OF SOLID WASTES
PERCENT OF RCRA CONTAMINANT MAXIMUM CONCENTRATION
Wei Iman-Galusha
(Ft. Snelling)
Contaminant
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Gasifier
Ash
0.38
1.0
<0.7
0.02
0.14
<0.3
1.4
NA
Cyclone
Dust
0.66
1.0
NA
0.02
0.16
<0.15
0.6
NA
Riley Gas
Gasifier
Ash
0.66
0.68
<0.05
<0.02
<0.04
<0.25
0.6
<0.01
Producer
Cyclone
Dust
0.08
0.39
<0.05
<0.02
<0.04
<0.25
0.2
<0.01
United Power Assn.
Stanton Plant
Bottom
Ash
0.16
0.30
0.4
<4
6
0.15
<0.4
0.5
Precipitator
Ash
1.4
0.92
0.5
<4
8
0.2
8.7
0.6
NA - not analyzed
-------
TABLE 11. MEDIUM-BID GASIFICATION AND COAL-FIRED POWER PLANT
RCRA LEACHATES OF SOLID WASTES
PERCENT OF RCRA CONTAMINANT MAXIMUM CONCENTRATION
Contaminant
Arsenic
Barium
Cadmiun
Chromium
Lead
Mercury
Selenium
Silver
Lurgi
Gasifier
Ash
0.42
1.2
<0.05
6.6
2.8
<0.1
<0.4
<0.02
Bottom
Ash
0.18
0.51
<0.05
2.6
0.94
<0.01
<0.04
<0.02
Precipitator
Ash
<0.06
0.41
<0.05
2.8
5
<0.1
<0.4
<0.02
The coal feedstocks of the gasification facility and the coal-fired
power plant at the Kosovo site were retrieved from the same stocks, and
parameter variabilities of the coal are applicable to both processes.
The data presented in Tables 6 and 9, percent of total element leached
by the RCRA extraction procedure, is significantly affected by analytical
sensitivities. Of the values not affected by analytical sensitivities, the
highest percentages (17-50%) of elements from the solids leached were for
cadmium, lead, and selenium in the bottom ash and precipitator ash of the
power plant firing Indianhead lignite. The concentrations of these elements
were also the highest values measured in the solid wastes; however, no con-
centration of any of these elements in the RCRA leachates represented as
much as ten percent of the RCRA contaminant maximum level. Only one "per-
cent of total element leached" value (Riley Gas Producer cyclone dust -
13.3%) of the RCRA contaminants for the gasification solid wastes exceeded
ten percent. These results indicate that the majority of elemental contami-
nants present in coal gasification solid wastes and coal-fired power plant
solid wastes are bound in the solids such that the leachability of the
elements is relatively low.
The results of this paper indicate that the solid wastes of specific
coal-fired power plants and coal gasification processes tested warrant
listing as non-hazardous. However, the non-hazardous listing of these
wastes is based upon the characteristic of EP toxicity and primarily upon
the elemental contaminants and does not include a severe evaluation of the
wastes using other pertinent criteria, such as organic constituents or
356
-------
radioactive components, that may need to be evaluated to determine if there
may be a contribution to groundwater contamination. Additional data must be
generated to apply the findings of this paper to the solid wastes generated
by other coal combustion and coal gasification processes and feedstocks to
fully evaluate the status of solid wastes from these industries with regard
to Subtitle C regulations.
ACKNOWLEDGEMENTS
The authors would like to thank
Fred Jones - American Natural Service Company
Mike O'Brien - United Power Association
Maureen Kilpatrick - Radian Corporation
Steve Gibson - Radian Corporation
for their contributions to the investigations associated with this paper.
REFERENCES
1. Environmental Protection Agency. Hazardous Waste Management System:
Identification and Listing of Hazardous Wastes. Federal Register, Vol.
45, No. 98. May 19, 1980.
2. Environmental Protection Agency. Identification and Listing of
Hazardous Waste. Federal Register, Vol. 45, No. 225, November 19,
1980.
3. Kilpatrick, M.P., R.A. Magee, and T.E. Emmel. Environmental
Assessment: Source Test and Evaluation Report — Wellman-Galusha (Ft.
Snelling) Low-BTU Gasification. EPA-600/7-80-097 (NTIS-PB-80-219 330).
Radian Corporation, Austin, Texas, 1980.
4. Source Test and Evaluation program, Riley Gas Producer.
357
-------
CHARACTERIZATION OF SOLID WASTES FROM t
INDIRECT LIQUEFACTION FACILITIES
by
Cora A. Hunter, Kar Y. Yu and Kimm W. Crawford
TRW Environmental Division
Presented at the Sixth Symposium
Environmental Aspects of Fuel Conversion Technology
Denver, Colorado
October 26-30, 1981
ABSTRACT
Gasification ash and slag are the major solid wastes generated in indirect
coal liquefaction facilities. Smaller amounts of spent catalysts and pollution
control sludges may also be generated. There is a limited amount of data on
the hazardous and nonhazardous characteristics of these solid wastes. Leachate
data for gasifier ash and slag from Lurgi, Wellman-Galusha, and Texaco gasifi-
cation have been presented elsewhere. The RCRA leaching characteristics of
quenched gasifier slag and dust from commercial scale Koppers-Totzek gasifica-
tion tests in Greece are presented in this paper. The potential accumulation
of trace elements in the sludges from biological oxidation of Lurgi gasification
condensates are estimated. Koppers-Totzek and Texaco gasification condensates
will contain negligible amounts of organics as compared to the Lurgi gasifica-
tion condensates and will not require biological oxidation. The potential
accumulation of trace elements on high temperature shift catalyst are examined
as a function of degree of gasification and feed coal characteristics.
358
-------
1.0 INTRODUCTION
The Resource Conservation and Recovery Act (RCRA) of 1976 directs the
Environmental Protection Agency to promulgate regulations to insure the proper
disposal of solid wastes for the protection of both human health and the envir-
onment. With the recent emphasis on America's coal resources, indirect coal
liquefaction may soon be providing a portion of America's energy needs. The
The proper disposal of solid wastes generated in the production of liquid
fuels and chemicals from coal will be part of the environmental protection
required under RCRA. EPA has set forth procedures to determine the potential
hazards of solid wastes. Characterization of solid waste streams from indirect
liquefaction facilities is the first step toward assuring proper disposal of
these wastes.
There is a limited amount of data on the hazardous and nonhazardous char-
acteristics of solid wastes from indirect coal liquefaction facilities. The
data are dependent upon the coal used. Leachate data for gasifier ash and slag
from Lurgi, Wellman-Galusha, and Texaco gasification have been presented else-
(1 2)
where. ' ' The RCRA leaching characteristics of quenched gasifier slag and
dust from commercial scale Koppers-Totzek gasification tests in Greece are pre-
sented in this paper. The potential accumulation of trace elements in the
sludges from biological oxidation of Lurgi gasification condensates is estimated.
The potential accumulation of trace elements on high temperature shift catalyst
is examined as a function of degree of gasification and feed coal characteristics.
2.0 INDIRECT LIQUEFACTION PROCESSES AND SOLID WASTES
Indirect liquefaction combines coal gasification technologies with catalytic
synthesis technologies to produce a range of liquid fuels and chemicals. Figure
1 indicates the basic sequence of process steps necessary for indirect lique-
faction. The raw coal is prepared to gasifier feed specifications and gasified
(gasification technologies currently in use or under development include the
Lurgi, Koppers-Totzek, and Texaco processes). The raw product gas is quenched
and upgraded for synthesis by dust removal, shift conversion, and acid gas
(e.g., C02 and H2S) removal. The purified synthesis gas is catalytically con-
verted into crude liquid products which can either be used directly as fuels or
further refined (synthesis processes currently in use or under development include
Fischer-Tropsch, Methanol, and Mobil M gasoline synthesis). Not shown in Figure
1 are the units necessary for on-site steam and power generation, boiler flue
gas desulfurization (FGD), oxygen production, raw water treatment, and process
cooling. 359
-------
j
. TAIL GASES
U)
ON
O
TRANSIENT AND 1
WASTE GASES ACID GASES 1
t t 1
~^
SULFUR
RECOVERY
1 1
1 |_ ^. BY-PRODUCT
1 ~ SULFUR
I 1 1 H2S-RICH 1
ACID 1 cwT
STE.M | | | OASES ' >• SVJivsT
T 1 II
L+
RAW 1 fc COAL k ri~irirn hi COOLING AND k SHIFT fc ACID GAS
COAL ( W PREPARATION •" ""•"' ILH ^ DUST REMOVAL ^^ CONVERSION W REMOVAL
^"^^
1 1 1 1
TRACE
SULFUR
REMOVAL
_^ PRODUCT
*" SYNTHESIS
1 i 1
''"*-' 1 1 | I CONDENSATES | I |
OXYGEN ' 'CONDENSATES . 'AND STILL
1 1 | 1 BOTTOMS
1 1 1
. PRODUCTS
SPENT
CATALYST
GASIFIER
ASH/SLAG
TARS
(LURGI)
SPENT GUARD
MATERIAL
. SPENT
I CATALYST
SYNTHESIS
CONDENSATES I
TREATED WATER/
• CONCENTRATED
BRINE
INORGANIC BIOLOGICAL
SLUDGES OXIDATION
(K-T AND SLUDGES
TEXACO) (LURGI)
Figure 1. Types of Wastes Generated in Indirect Coal Liquefaction Facilities
-------
The major solid waste streams from indirect liquefaction facilities in-
clude quenched gasifier ash and slag, gasifier dust, heavy tars and oils, boiler
bottom and fly ash, flue gas desulfurization sludge, biological treatment sludges,
and sulfur (if not sold as a by-product). Other solid waste streams include
spent catalysts, spent sulfur guard, raw water treatment sludges, and chemical
precipitation sludges. Leachable trace elements are pollutants of potential
concern in all of the solid waste streams. With the exception of the biologi-
cal oxidation sludges, all of the solid waste streams are inorganic based. The
key solid waste streams addressed in this paper are gasifier slag, gasifier dust,
biological oxidation sludges, and spent catalysts.
The dry ash Lurgi gasifier operates at temperatures below coal ash fusion
temperatures (1815 to 1930°C), while Koppers-Totzek and Texaco gasifiers operate
at higher temperatures (2100 to 2600°C). A portion of the coal ash will leave
the K-T and Texaco gasifiers as dust entrained in the raw gas stream while the
remaining coal ash exits as molten slag from the bottom. Gasifier ash, slag,
and dust will consist mainly of nonvolatile and unreacted portions (primarily
mineral matter) of the feed coal. Toxic trace elements and substances derived
from the parent coal are potential pollutants of concern. Gasifier ash and slag
are ordinarily quenched with process water for cooling and/or transportation
purposes, and thus will contain substances found in the quench water. Gasifier
dust may contain substances found in the wash water.
Biological oxidation sludges result from biological wastewater treatment
processes used to treat gasification and synthesis condensates. Nonbiodegrad-
able toxic organic compounds and trace elements derived from gasification and
synthesis condensates are the potential pollutants of concern. Koppers-Totzek
and Texaco gasification condensates will contain negligible amount of organics
as compared to the Lurgi gasification condensates due to the higher combustion
temperatures in the Koppers-Totzek and Texaco gasifiers. Lurgi gasification
condensates will contain large amounts of dissolved and suspended organics rang-
ing from simple phenols to complex organic acids. Condensates from the Fischer-
Tropsch, Methanol, or Mobil M gasoline synthesis section of integrated indirect
liquefaction facilities will also contain high loadings of soluble organic pol-
lutants (e.g., alcohols, ketones, organic acids).
There are several types of catalysts which may be used in indirect lique-
faction facilities. Shift catalysts include cobalt-molybdate, copper/zinc,
and iron chrome based catalysts. Copper/zinc based catalysts are used for
Methanol synthesis. Fischer-Tropsch synthesis catalysts are iron based with
361
-------
transition elements as promoters. Zeolites are used for Mobil M gasoline synthe-
sis. Methanation catalysts are nickel based. Certain catalysts are known to con-
tain toxic consituents (e.g., methanation catalysts are nickel-based). High
temperature shift catalysts may accumulate toxic constituents through prolonged
contact with raw coal gases.
3.0 RCRA HAZARDOUS WASTE CLASSIFICATION CRITERIA
The current Federal hazardous waste regulations define the testing proce-
dures and thresholds which cause a solid waste to be classified as hazardous.
A solid waste is considered hazardous if it meets test criteria for ignitability,
corrosivity, reactivity, or toxicity. EPA can also list wastes as hazardous if
the waste has been found to be fatal to humans in low doses or toxic as indicated
by the LD5Q or LC50 levels. Solid wastes containing any of the EPA-specified
hazardous constituents* may also be listed as hazardous after taking into con-
sideration some intrinsic factors such as concentration of the constituents in
the waste, persistence of the constituent, quantity of wastes, the nature of the
toxicity presented by the constituent, and other appropriate factors.
The toxic characteristics of solid wastes are measured by the RCRA Extrac-
tion Procedure (EP) Toxicity Test, which is designed to roughly approximate
the extraction of soluble material with rain water. The solid is extracted with
a sixteen-fold excess of leaching solution at a pH of 5.0 for a 24-hour time
period at room temperature. Following the extraction period, the sample is
diluted to an aqueous volume of 20 times the sample weight and then filtered
to separate the liquid and solid phases. The extract is then analyzed for eight
trace elements (arsenic, barium, cadmium, chromium, lead, mercury, selenium, and
silver) and other identified hazardous constituents which are listed in the Ex-
traction Procedure. ' The RCRA standards for these eight trace elements are
100 times the Primary Drinking Water Standards.
4.0 CHARACTERISTICS OF INDIRECT LIQUEFACTION WASTES
No indirect coal liquefaction solid wastes are listed as hazardous wastes
at the present time. There is insufficient information available at present to
*There are more than 350 specified hazardous constituents including cyanides,
nickel, vanadium pentoxide, phenols, naphthylamines, etc. (see 40 CFR 261,
May 19, 1980).
362
-------
determine the hazardous or nonhazardous characteristics of many of the wastes
according to RCRA criteria. Some of these wastes are known to contain certain
identified constituents of concern (e.g., methanation catalysts are known to
contain nickel).
4.1 GASIFIER SLAG AND DUST
The RCRA 1 eachate characteristics of quenched gasifier slag from commercial -
scale Koppers-Totzek gasification of Greek Lignite and Illinois #6 coals in
Ptolemais, Greece are shown in Table 1. ' Although the quenched gasifier slag
samples were collected under various gasifier operating conditions, the RCRA
leachate trace element concentrations are quite uniform. When compared to the
RCRA Standard (100 times the primary drinking water standards), none of the
samples analyzed would be classified as hazardous. In fact, most of the RCRA
leachate trace element concentrations are less than 10 times the primary drink-
ing water standards (selenium concentrations may actually be less than 10 times
the primary drinking water standard, but analytical sensitivity is limited in
these data). Neutral pH leachate tests on these samples resulted in uniform
leachate trace element concentrations similar to those found for the RCRA leach-
tfl\
ates. ' The leachate characteristics of the Koppers-Totzek gasifier slag are
similar to those presented by other investigators for other coal gasification
ashes.'1'2'5'
As with K-T slag, dust from Koppers-Totzek gasification of Greek lignite
coal would not be classified as hazardous from a trace element standpoint based
(n\
on data in Table 2.v ; Most of the RCRA leachate trace element concentrations
are less than 10 times the primary drinking water standards (selenium concen-
trations may actually be less than 10 times the primary drinking water standard,
but were not detected as such in these tests). Neutral pH leachate tests on
these samples resulted in fairly uniform leachate trace element concentrations
(4)
with minor differences between the RCRA leachate and neutral pH leachate.
There is little difference in the leachate characteristics of gasifier slag and
dust disposed of in settling ponds.
The commercial-scale Koppers-Totzek gasification tests with Illinois #6
coal employed a cyclone for dry collection of gasifier dust samples, since the
wet sludge from clarification of wash water associated with Illinois #6 coal
could not be isolated from that of Greek lignite. In conventional plant designs,
the dust is removed from the raw gas in a washer cooler system and this dust
would exit the system as solids suspended in the wash water. Some of the toxic
363
-------
TABLE 1. RCRA LEACHATE CHARACTERISTICS OF QUENCHED KOPPERS-TOTZEK GASIFIER SLAG (GREEK LIGNITE
AND ILLINOIS #6 COAL)(4)
Leachate Trace Element Concentration (mg/1)
Trace Element
Ag
As
Ba
Cd
Cr
Hg
Pb
Se
Greek Lignite Coal*
<0.01
<0.4
0.1
<0.007
<0.04
<0.0002
<0.05
<0.4
Illinois #6 Coal*
<0.01
<0.4
<0.03
<0.007
<0.04
<0.0002
<0.09
<0.4
RCRA Standard:
5
5
100
1
5
0.2
5
1
*
Average value obtained from two samples
Average value obtained from nine samples
*100 times the Primary Drinking Water Standards
-------
TABLE 2. RCRA LEACHATE CHARACTERISTICS OF KOPPERS-TOTZEK GASIFIER DUST
DISPOSED IN SETTLING PONDS (GREEK LIGNITE COAL)(4)
Leachate Trace Element Concentration (mg/1)
Trace Element Greek Lignite Coal* RCRA Standard"1"
Ag <0.01 5
As <0.2 5
Ba 0.38 TOO
Cd <0.007 1
Cr <0.04 5
Hg <0.0002 0.2
Pb <0.05 5
Se <0.4 1
*
Average value obtained from four samples
times the Primary Drinking Water Standards
365
-------
components in the gasifier dust would become solubilized in the wash water.
Also, toxic components in the wash water could be introduced into the wet dust.
The leachate characteristics of dry dust samples from Koppers-Totzek gasification
(4)
of Illinois #6 coal shown in Table 3X ' are thus a conservative estimate of the
leachate characteristics of dust that would be collected in washer cooler systems
neglecting the addition of any toxic components that might come from the wash
water since some leaching will occur as a result of contact with wash water. The
levels of silver, barium, chromium, mercury, and lead are well below the RCRA
Standard for classification as a hazardous waste. The arsenic, cadmium, and selen-
ium concentrations are also below the RCRA Standard, but the margin of safety is
lower. The neutral pH leachate characteristics are fairly similar to the RCRA
leachate characteristics, except for barium and cadmium, which are more readily
leached under neutral pH conditions. Although there is no RCRA Standard for
(4)
boron, its RCRA leachate concentration of 2.2 mg per liter exceeds the irri-
gation water quality standard of 0.75 mg per liter. Thus, Teachability of boron
may be an important water quality concern at specific disposal sites even though
this element is not considered to be toxic to man or higher animals. It should
also be mentioned that the leaching characteristics of the K-T dust do not dif-
fer significantly from that of the parent Illinois #6 coal itself.
All available data indicate that gasification ash/slag and dust would be
classified as nonhazardous based on the RCRA Extraction Procedure requirements.^ ' '
However, it is possible that some of these wastes could be hazardous
RCRA if process wastewaters containing Teachable toxic substances are used to
quench the raw gas or ash.
Leachable trace elements are not the only basis upon which gasifier slag and
dust may be listed or classified as hazardous. In the case of both K-T and
Texaco gasification dust, leachates may contain substances such as cyanides,
sulfides, thiocyanates and ammonia derived from the aqueous condensates or wash
waters which have been in contact with dust. Only limited data are currently
available regarding the presence and Teachability of any such constituents in
the "wet" dust from the subject processes.
4.2 BIOLOGICAL OXIDATION SLUDGES
Biological treatment of gasification and synthesis wastewaters is envision-
ed for many proposed synthetic fuel facilities in the U.S. especially those based
upon Lurgi gasification. In these facilities, biological sludges would be gen-
erated as a waste from the treatment process. Although there are very limited
366
-------
TABLE 3. RCRA LEACHATE CHARACTERISTICS OF DRY DUST SAMPLES FROM
KOPPERS-TOTZEK GASIFICATION (ILLINOIS #6 COAL) (4)
Leachate Trace Element Concentration (mg/1)
Trace Element
Ag
As
Ba
Cd
Cr
Hg
Pb
Se
Cyclone Dust*
<0.01
0.35
<0.02
<0.007
<0.02
<0.0002
<0.15
0.6
RCRA Standard"1"
5
5
100
1
5
0.2
5
1
*
One sample
f100 times the Primary Drinking Water Standards
367
-------
leachate and bioassay data available at present on the characteristics of such
sludges, the presence of potentially toxic organics (e.g., aromatic amines) and/
or trace elements (e.g., Hg, Cd) in the raw wastewaters would suggest that the
sludges could be hazardous.
It is possible to estimate the amounts of various trace elements which may
accumulate in Lurgi gasification condensate biological oxidation sludges since
a limited amount of data are available on the trace element composition of Lurgi
gasification condensates. ' ' The accumulation of trace elements in the bio-
logical oxidation sludges can be estimated from removal efficiencies achieved
for biological treatment of industrial and municipal wastewaters.' ' ' ' The
maximum trace element concentrations Teachable from Lurgi gasification conden-
sate biological oxidation sludges are estimated in Table 4, assuming that all of
the accumulated material is Teachable. As indicated in the table, the maxium
leachate trace element concentrations may exceed 100 times the Primary Drinking
Water Standards. Although barium is not listed in Table 4, it should not be a
problem due to its low concentration in the Lurgi raw gas liquor and relatively
high RCRA standard concentration. The Lurgi gasification condensate concentra-
tions and biological oxidation removal efficiencies are summarized in the Appen-
dix.
Incineration of biological oxidation sludge has also been proposed for Lurgi
facilities to destroy the toxic organics in the waste. However, the incineration
residue may also be hazardous due to Teachable trace elements, as indicated by
calculations in Table 4. The trace element concentrations could be increased
by a factor of three or more due to incineration. For these calculations, in-
cineration is assumed to result in a 70 percent reduction in waste quantity
(on a dry basis). All of the trace elements present in the biological oxidation
sludge are also assumed to accumulate in the incineration residue and to be
Teachable.
4.3 SPENT CATALYSTS
There is insufficient information available at present to determine the
hazardous or nonhazardous characteristics of spent catalysts from indirect
liquefaction processes. Due to the proprietary nature of most catalysts, there
is little data publicly available on their specific compositions. Some catalysts
are known to contain certain identified hazardous constituents (e.g., methanation
catalysts are nickel-based). However, many catalysts are presumed to not con-
tain any hazardous constituents (e.g., Mobil M gasoline synthesis and Claus
368
-------
TABLE 4. PREDICTED RCRA LEACHATE CHARACTERISTICS OF SLUDGE FROM BIOLOGICAL
OXIDATION OF LURGI GASIFICATION CONDENSATES, WORST CASE
Predicted Maximum Leachate Trace Element Concentration (mg/1)
Biological Oxidation RCRA
Incineration Residue Standard*
18.2 5
1.8 5
13.8 1
181.8 5
7.9 0.2
18.2 5
9.7 1
19.2 5
2.9 5
14.6 1
576.1 5
99.9 0.2
576.1 5
76.8 1
24.0 5
179.8 5
2.7 1
3.6 5
4.7 0.2
24.0 5
383.6 1
Trace Element
Rosebud Coal
Ag
As
Cd
Cr
Hg
Pb
Se
Illinois #6 Coal
Ag
As
Cd
Cr
Hg
Pb
Se
Dunn Lignite Coal
Ag
As
Cd
Cr
Hg
Pb
Se
Sludge
5.5
0.5
4.1
54.5
2.4
5.5
2.9
5.8
0.9
4.4
172.8
30.0
172.8
23.0
7.2
53.9
0.8
1.1
1.4
7.2
115.1
100 times the Primary Drinking Water Standards
369
-------
catalysts which are zeolite- and alumina-based catalysts, respectively).
Although fresh catalysts may not contain any toxic constituents, they may
accumulate such constituents through prolonged contact with the coal gases. In
particular, potentially volatile trace elements originally present in the feed
coal (e.g., As, Cd, Cr, Hg, Pb, and Se) may accumulate in the sulfur tolerant
shift catalyst over time, since this type of catalyst is directly exposed to
n 21 ^}
hot raw gas and is known to have an affinity for various trace elements/ ' '
There is essentially no leachate data available in the public domain on any of
the catalysts used in indirect liquefaction processes.
The potential accumulation of various trace elements on the shift catalyst
could be estimated based on the trace element composition of the raw coal gas,
however, sufficient data are not available. Some data are available to allow
indirect calculation of the degree of gasification of several trace elements in
various gasifiers, although a wide range of values can be derived depending on
which set of data are used. ' ' For purposes of this paper, therefore, the
accumulation of trace elements on shift catalyst have been estimated as a func-
tion of degree of gasification and feed coal characteristics. Table 5 summar-
izes the trace element contents of American coals.
Assuming all of the gaseous trace elements are deposited on by the shift
catalyst and are subsequently Teachable, the time required for shift catalyst
to become hazardous due to trace element deposition can be estimated. Figures
2 to 4 show the results as a function of trace element concentration in the coal
and percent of the trace element gasified. Shift catalysts is estimated to
become hazardous within twelve hours under the worst case (i.e., 100 percent
gasification) for coals with the mean concentrations of the trace elements shown
in Table 5. The nonhazardous lifetime would be increased to about 3 months when
only one percent gasification of the trace elements occurs. Minimum trace ele-
ment levels found in American coals would still result in a hazardous catalyst
within a week if 100 percent of the trace elements are gasified. The nonhazardous
lifetime of the shift catalyst would be increased to about 3 years when only one
percent gasification of the trace elements occurs.
Although there are large uncertainties in the exact levels of various ele-
ments which would accumulate on the shift catalyst, the calculations presented
indicate a reasonable potential for the spent catalyst to become hazardous. Also,
shift catalyst may be affected by trace elements in terms of activity. Elements
370
-------
TABLE 5. TRACE ELEMENT CONTENTS OF AMERICAN COALS^14*15^
Concentration in Coal (ppm)
Trace Element Mean Minimum Maximum
As 16.4 0.5 357
Cd 1.8 0.02 100
Cr 15.3 <0.5 70
Hg 0.17 0.01 3.3
Pb 21.2 <0.7 283
Se 3.6 <0.10 150
371
-------
N>
a
i-
LU
Ul
O
UL
O
O
flC
iu
ft.
MEAN LEAD CONCENTRATION
IN COAL
\ \
MEAN ARSENIC CONCENTRATION
MEAN CHROMIUM CONCENTRATION
IN COAL
1 10 102
TRACE ELEMENT CONCENTRATION IN COAL (PPM)
Figure 2. Predicted Shift Catalyst Lifetime Required to Reach RCRA Leachate
Standards for As, Cr, and Pb (5 mg/l) *
(BASED ON 2 KG RAW COAL GAS PER HOUR PER KG CATALYST)
-------
102
10
UJ
ui
UJ
U>
UJ
u
tr
10-
MEAN CADMIUM CONCENTRATION IN COAL
\ \
MEAN SELENIUM CONCENTRATION IN COAL
10-2
icr1 1 10
TRACE ELEMENT CONCENTRATION IN COAL (PPM)
102
103
Figure 3. Predicted Shift Catalyst Lifetime Required to Reach RCRA Leachate
Standards for Cd and Se (1 mg/l) *
* (BASED ON 2 KG RAW COAL GAS PER HOUR PER KG CATALYST)
-------
MEAN MERCURY CONCENTRATION
IN COAL
1CT1 1
TRACE ELEMENT CONCENTRATION IN COAL (PPM)
Figure 4. Predicted Shift Catalyst Lifetime Required to Reach RCRA Leachate
Standard for Hg (0.2 mg/l) *
* (BASED ON 2 KG RAW COAL GAS PER HOUR PER KG CATALYST)
374
-------
such as As, Pb, Cr, Hg, and Cd are likely to be catalyst poisons at some levels,
and hence catalyst life could actually be shorter than that found in non-coal
applications due to deactivation by trace element accumulation. Process design-
ers should be aware of the potential for catalyst deactivation by coal derived
trace elements. Finally, it should be mentioned that many catalysts contain
metals of commercial value and hence may not have to be viewed as wastes if these
metals are reclaimed.
5.0 CONCLUSIONS
(1) Commercial-scale Koppers-Totzek gasification tests with Greek lignite
and Illinois #6 coals in Ptolemais, Greece indicate that quenched
gasifier slag and dry or wet dust would not be classified as hazard-
ous based upon RCRA leachate criteria for trace elements. However
process or wastewaters used to cool or quench solids may introduce
toxic constituents.
(2) Calculations indicate that maximum trace element concentrations
Teachable from Lurgi gasification condensate biological oxidation
sludges may exceed 100 times the Primary Drinking Water Standards.
Although incineration of the biological oxidation sludge is expected
to destroy the toxic organics in the sludge, the incineration resi-
due may still be hazardous.
(3) Certain spent catalysts (e.g., nickel based methanol or methanation
catalysts) are expected to be inherently hazardous. High tempera-
ture shift catalysts may become hazardous due to accumulation of
leachable trace elements through prolonged contact with coal gases.
6.0 RECOMMENDATIONS
(1) Additional RCRA leachate data for gasifier ash and slag produced
by various gasification technologies using several coals would be
helpful to verify the nonhazardous characteristics of the ash and
slag. The presence of toxic organic or inorganic compounds in ash
quenched with process wastewater could be indicated by both chemi-
cal analyses and bioassay testing of solids and/or leachates.
(2) RCRA leachate data should be collected to determine the hazardous
or nonhazardous characteristics of biological oxidation sludges
from wastewater treatment. Performance of bioassay tests would
375
-------
provide information on the trace elements and the nonbiodegrad-
able but toxic organics that might be present in these sludges.
(3) Obtaining RCRA leachate and bioassay data on fresh catalysts
would allow determination of the hazardous and nonhazardous
characteristics of the basic catalyst materials. RCRA leachate
and bioassay data on spent catalysts would provide insight into
the potential accumulation of trace elements or toxic organics
through contact with coal derived gases.
376
-------
APPENDIX
EXPECTED CHARACTERISTICS OF LURGI GASIFICATION CONDENSATES
Trace Elements (mg/£)
Ag
As
Ba
Cd
Cr
Hg
Pb
Se
Raw Gas Liquor
Production Rate
(1000 kg/hr)
Biological Oxidation
Sludge Production
Rate (kg/hr)
Roseb
0.3
0.06
<0.01
0.3
3
0.1
0.3
0.1
304
1900
Illinois #6
CoaUS)
0.1
0.03
<0.1
<0.1
8
0.4
3
0.25
507
1000
Dunn Lignite
Coal (7.8)
<0.2
3
0.03
<0.03
<0.03
0.2
2
441
1400
Design Basis - 2.5 x 10^ kcal/day energy output from Methanol Synthesis
BIOLOGICAL OXIDATION REMOVAL EFFICIENCIES
Trace Element
Ag
As
Cd
Cr
Hg
Pb
Se
Percentage
Removal
50
25
38
50
65
50
80
Reference
(9)
(10)
(10)
(11)
(11)
(11)
(11)
377
-------
REFERENCES
1. Richter, G.N. and W.G. Sch1inger, Environmental Assessment of the Texaco
Coal Gasification Process, Hydrocarbon Processing, October 1980, p. 66.
2. Yu, K.Y. and G.M. Crawford, Characterization of Coal Gasification Ash
Leachates Using the RCRA Extraction Procedure. Paper presented at the
Symposium of Environmental Aspects of Fuel Conversion Technology-V,
St. Louis, Missouri, September 16-19, 1980.
3. Federal Register, Vol. 45, No. 98, Part III, May 19, 1980.
4. Data gnerated as part of commercial-seale gasification tests with U.S.
coal in Ptolemais, Greece, TRW, 1981.
5. Boston, C.R. and W.J. Boegly, Jr., Leachate Studies on Coal and Coal
Conversion Wastes, NTIS CONF-790571-1, 1979.
6. Trials of American Coals in a Lurgi Gasifier at Westfield, Scotland,
Woodall-Duckham, Ltd., Sussex, England, ERDA R&D Report No. 105, 1974.
7. Bromel , M.C. and J.R. Fluker, Biotreating and Chemistry of Wastewaters
from the South African Coal, Oil and Gas Corporation (SASOL) Coal
Gasification Plant, North Dakota State University, Fargo, North Dakota,
December 1976.
8. Somerville, M.H., J.L. Elder, et al., An Environmental Assessment of a
250 MMSCFD Dry Ash Lurgi Coal Gasification Facility in Dunn County, North
Dakota, University of North Dakota Engineering Experiment Station,
Bulletin No. 76-12-EES-01 Volumes I-IV, December 1976.
9. Cohen, Jesse M., Municipal Environmental Research Laboratory, Trace
Metal Removal by Wastewater Treatment, Technology Transfer (January
1977). Cincinnati, OH, EPA, 1977.
10. Esmond, Steven E. and Albert C. Petrasek, Jr., Trace Metal Removal, Ind.
Water Eng. Vol. II, No. 3, May-June 1974, p. 14-17.
11. Zemansky, Gilbert M., Removal of Trace Metals During Conventional Water
Treatment, J. Am. Water Works Assoc., Vol. 66, No. 10, Part 1, October
1974, p. 606-609.
12. Telephone memorandum, Ib Dybkjaer, Haldor Topsoe Inc., June 1981.
13. Colony, 1974, An Environmental Impact Analysis for a Shale Oil Complex
at Parachute Creek, Colorado: V. 1.
378
-------
14. Swanson, V.E., J.H. Medlin, J.R. Hatch, S.L. Coleman, C.H. Wood, Jr.,
S.D. Woodruff, and R.T. Hildebrand, Collection, Chemical Analysis, and
Evaluation of Coal Samples in 1975, U.S. Geological Survey Open File
Report OFR 76-468, 1976.
15. Gloskoter, H.J., R.R. Ruch, et al., Trace Elements in Coal: Occurrence
and Distribution, Illinois State Geological Survey, Urbana, Illinois,
ISGS Circulation No. 499, 162 p.
379
-------
ASH/SLAG RESIDUALS AND WASTEWATER TREATMENT
PLANT SLUDGES FROM SYNFUELS FACILITIES:
CHARACTERIZATIONS AND IMPLICATIONS FOR DISPOSAL
by: Ronald D. Neufeld, Associate Professor of Civil Engineering
Georg Keleti, Associate Professor Graduate School of Public Health
J. Bern, C. Moretti, S. Wallach, H. Erdogen, Graduate Students
University of Pittsburgh
Pittsburgh, PA 15261
ABSTRACT
The purpose of this paper is to present an overview of research
conducted at the University of Pittsburgh in the area of coal conversion ash
and slag. Residuals were obtained from the GFETC "slagging Lurgi type"
gasifier (two different runs), METC (Wellman-Galusha type) pressurized
gasifier, DOE-Chapman gasifier fly ash, and two H-Coal vacuum bottoms
residuals. A first screen bioassay of SRC-II Fort Lewis coal liquefaction
residuals and sludges is also presented. In addition, research has been
conducted at the University of Pittsburgh for the past few years in the area
of developing a stable pretreatment and biological treatment facility for the
processing of phenolic type coal gasification wastewaters. During the
processing of wastewaters, sludges are produced which are assessed for
toxicity, mutagenicity and overall disposability characteristics.
OVERVIEW OF RESULTS
It may be expected that a commercial sized coal conversion facility will
produce a variety of solid waste residuals. The wastes generated that may
exert the greatest influence on residuals management are: (1,2)
1. COAL PREPARATION PLANT RESIDUALS-to include coal refuse, coal dust and
wastewater from the tailing pond
2. COAL GASIFICATION PLANT AREA WASTES-to include residual ash, slag and
quench waters
3. STEAM AND POWER PLANT GENERATION WASTES-to include residual ash flue gas
desulfurization sludge
4. RAW WATER TREATMENT AREA-to include sludge from solids in the raw water
source
5. WASTEWATER TREATMENT PLANT AREA SLUDGES-to include lime sludge, organic
sludge, waste biological sludges and oil and tar residuals
380
-------
6. TAR SEPARATION AREA
7. PHENOL REMOVAL AREA-to include filter backwash and sludges containing
phenolics when solid extraction processes are used
8. SULFUR REMOVAL AREA SLUDGES-to include elemental or product sulfur if
nonsaleable or stored on site
9. TAILGAS TREATMENT AREA SLUDGES-to include residual sludge materials.
Land disposal of solid residuals is an economically logical choice for
an overall management scheme of commercial scale synthetic fuel facility
solid wastes. Such land disposal, however, must be done in an environmentally
and "RCRA" acceptable manner. Bern et al.(l) have outlined management
alternatives that are available to owners of commercial sized synthetic fuels
facilities.
Neufeld et al.(2) have reported on chemical and biological properties
of coal conversion ash residuals derived from U.S. DOE sponsored large scale
coal gasification and direct liquefaction facilities. Characterizations of
such solid wastes include proximate analysis, development of natural particle
sized distributions, and heavy metal analysis of leachates from each sized
fraction. This work showed that the smaller sized fractions yield much
greater quantities of heavy metals in derived leachates. In no case did
resulting leachates using the EPA "EP" procedures and ASTM-A distilled
de-ionized water leaching procedures yield concentrations in excess of one
hundred times the concentration of primary drinking water heavy metals; a
value above which wastes are determined to be "hazardous". In addition, no
coal conversion waste ash or slag residual gave positive result in Ames
testing. On the other hand, evidence of Daphnia toxicity was observed in
some coal conversion derived leachates.
Wastewater treatment sludges were generated as part of our study of
METC gasifier effluent control technology development. Wastewater treatment
plant sludges that were studied include lime sludges developed from pH
adjustment prior to ammonia stripping, organic sludges developed by
filtration and precipitation prior to biological oxidation, and biological
sludges from the treatment of fixed bed coal gasification wastewater.
Leachates from such sludges are shown to be toxic to Daphnia magna while
negative results were observed in Ames testing.
DESCRIPTION OF COAL CONVERSION SOLID RESIDUALS
CHAPMAN (WILPUTTE) GASIFIER
The Chapman fixed bed dry bottom gasification facility at Kings Port,
Tennessee produces a low BTU product gas used for combustion fuel. Gas
cleaning and purification operations involve cyclone removal of dry
particulates and aqueous gas quenching. Solid wastes coming from this
facility are gasifier and cyclone ash with cyclone ash being used in our
experimental procedures. It should be noted that cyclone ash differs
inherently from the more familiar coal combustion facility fly ash in that
381
-------
gasifier particulates have gone through a reducing zone as compared to coal
combustion fly ash particulates which go through an oxidizing zone.
GRAND FORKS ENERGY TECHNOLOGY CENTER
The Grand Forks Energy Technology Center has a "Lurgi type" oxygen
blown slagging gasifier. Two samples were obtained from this facility, the
first being a brown colored slag from run //R-52 using Indian-Head lignite
coal, the second being a black colored slag obtained from run //RA-93.
MORGANTOWN COAL CONVERSION FACILITY
The Morgantown Energy Technology Center gasifier is a pressurized
"Wellman-Galusha" type of system. The solid waste material obtained was
bottom ash/slag from the gasifier when operated using a bituminous coal and
was composed of principally large (2") particle sizes.
H-COAL LIQUEFACTION WASTES
The H-Coal process is a direct liquefaction facility developed by
Hydrocarbon Research Incorporated. The two H-Coal solid waste samples
obtained were both vacuum still bottoms from the direct liquefaction step.
One sample was generated from Illinois coal when the system was operated in
the "syncrude" mode, while the second sample was generated from Illinois
coal when operated in the "fuel oil" mode. These samples will be referred
to as "H-Coal #3" and "H-Coal #4". Both H-Coal samples were irregularly
shaped black "chunks" of materials with a majority of chunks larger than
three inches by three inches.
WASTEWATER SLUDGE SAMPLES
Figure 1 is a flow diagram of the research pretreatment and biological
treatment steps associated with the processing of METC coal gasification
wastewaters as conducted at the University of Pittsburgh. Wastewater was
provided to the University from the hot gas quench of the stirred fixed bed
gas producer located and operated by METC. The goal of the treatment
processes at the University of Pittsburgh were to develop a linkage of
operations that could effectively treat coal conversion wastewaters in a
stable fashion. Stability for the bioreactor was defined as occurring when
at least three sludge ages had passed.
Table 1 is a characterization of three different shipments of METC
wastewaters. The first shipment represents a "nontypical sample" produced
at least one year earlier to our testing while the second sample was "more
typical" being produced within several months of our evaluations. Sludges
for this study are generated from the "typical" wastewater sample.
382
-------
rt>
1
»»
pH = !0.5
1 '
T
C
)
-ILTER
Air
T=110°F
FREE AMMONIA STRIPPER
H2S04
Lime tm = 30 min
Lime Sludge
>pH*K).5
Tx25°C
©
I
P
y
C
t)
H
m - 5min
•1- 3-6
Sludge
NH,
STRIPPER (FIXED LEG)
T=I40°F
COOLER
1
BIO
REACTO
FEED
No OH
NUTRIENTS
to "B1
>-
i
V
C
D
_
k
Sludge Recycle
^_
TERTIARY
TREATMENT
for BAT
^ TO DISCHARGE
OR RE-USE
BIOLOGICAL REACTOR Excess Sludge
Figure 1. METC Gasifier Wastewater Process Flow Diagram
TABLE 1. METC WASTEWATER CHARACTERIZATION
ITEM
PHENOL
COD
TOC
TIC
TOTAL RESIDUE
FIXED RESIDUE
VOLATILE RESIDUE
FREON SOL. OIL &
GREASE
ACETONE SOL. OIL &
GREASE
pH
ALKALINITY (pH 1.5)
SCN
NH3
PHENOL/TOC RATIO
SHIPMENT #1
(AGED SAMPLE)
CONCENTRATION
970 MG/L
53,021 MG/L
11,102 MG/L
30 MG/L
72,331 MG/L
1,331 MG/L
71,000 MG/L
356 MG/L
1,633 MG/L
7.5
2,100 MG/L
11,000 MG/L
.069
SHIPMENT n
RUN 91
CONCENTRATION
2,375 MG/L
13,350 MG/L
5,371 MG/L
212 MG/L
1,319 MG/L
113.2 MG/L
1,205.8 MG/L
1,195.0 MG/L
106.3 MG/L
8.0
23,750 MG/L
372 MG/L
3,200 MG/L
.11
SHIPMENT 13
RUN 95
CONCENTRATION
3,750 MG/L
12,750 MG/L
5,390 MG/L
1,350 MG/L
1,120 MG/L
630 MG/L
3,790 MG/L
8.8
21,855 MG/L
7,000 MG/L
.70
383
-------
The pretreatment train used to treat coal conversion waters, and to
develop sludges within our laboratory represent a linkage of steps as follows:
Step 1-Free Ammonia Leg
This is accomplished in a laboratory via aerating a 15 to 20 gallon
batch of wastewater at a temperature of 60°C. Such aeration liberates
noticeable quantities of I^S, volatile organics, free ammonia, and results
in a reduction in the overall alkalinity of solution, thus minimizing lime
requirements for the fixed leg.
Step 2-Lime Addition
Lime (as CaOH) is added to the wastewater in sufficient quantities to
bring the pH to a range of 10 to 11.
Step 3-Filtration to Remove Lime Sludge
A large Buchner funnel with coarse grade filter paper is utilized to
remove precipitated lime. The resultant sludge is brown in color, and
contains organic materials. This sludge, referred to as "lime sludge", was
subsequently dried and leached in accordance with the EPA "EP" and ASTM-A
extraction procedures and tested for heavy metal content and toxicity to
Daphnia magna.
Step 4-Fixed Leg Ammonia Stripping
Ammonia is stripped at 140°F and pH 10% batchwise in a 15 gallon
stripper to simulate commercial scale fixed leg ammonia stripping. The
wastewater is kept in the ammonia stripper until the total ammonia in
solution reaches about 100 mg/1. The wastewater is then removed from the
ammonia stripper and placed into a large glass jar where it is subsequently
air cooled.
Step 5-Filtration
After ammonia stripping, the wastewater is pH adjusted using sulfuric
acid. Polymerization of trace organics appears to take place in the stripper
thus resulting in an organic sludge formation which is filtered out prior to
subsequent biological oxidation. Our approach is to remove the maximum
quantity of organics possible prior to biological oxidation via judicious pH
adjustment, flocculation and filtration. This sludge, called an "alum sludge"
(due to the addition of alum to promote coagulation/floculation) was also
tested in this study for leachate evaluations using Daphnia magna and atomic
adsorption spectrosocopy.
Step 6-Biological Reaction Phase
Pretreated wastewater is diluted as desired and fed on a continuous
basis to completely mixed activated sludge type bio-reactors with hydraulic
detention times of 1.0 days and sludge ages in the range of 20 days. During
one of our studies, a maximum of 60% wastewater diluted with tap water was
384
-------
utilized. Phase II of our present study is an attempt to minimize dilution
water requirement in the biological reaction phase. Biological sludges
harvested from the activated sludge reactors are being subjected to Ames
testing, Daphnia toxicity testing and extensive chemical evaluations during
the current phase of study.
PHYSICAL/CHEMICAL ANALYSIS OF ASH/SLAG SOLID WASTE RESIDUALS
All samples, with the exception of the lime and alum sludges from the
wastewater treatment train, were subjected to particle size distribution
analysis without altering the nature of samples. The philosophy of this
approach was to more properly reflect that which would be placed in landfill
systems; thus, crushing and grinding were not done. H-Coal samples were
subjected to crushing and grinding because of the rather large chunks of
materials received.
All sieving was conducted with U.S. standard sieves #'s 10, 20, 40, 60,
100 and 200 for sufficient duration to collect enough sample of each size
fraction as required for leaching tests.
Samples of each of the mesh sizes were subjected to leaching via the
ASTM-A leaching procedure and current EPA-EP leaching test and a self
designed "University of Pittsburgh" procedure using pH=2 HNO-j. A portion
of the leachates were segregated for heavy metal AA analysis, and Daphnia
magna evaluations.
DAPHNIA TOXICITY EVALUATIONS
Acute toxicity testing was conducted on generated sludge using Daphnia
magna standarized procedures outlined in Standard Methods, and in draft ASTM
procedures. It should be noted, however, that all samples of leachates were
adjusted to a pH between 7.4 and 7.6 before being subjected to the Daphnia
magna testing. The philosophy of our approach is not to evaluate the
toxicity of H+ and OH~, but rather than to evaluate the toxicity of
constituents contained in the leachates. Figure 2 is a typical plot of data
showing conductivity of GFETC lignite slag leachates as a function of
particle size of solid waste. As may be seen from this figure, smaller
particle sizes tend to leach greater quantities of dissolved materials than
larger particles. The differences in conductivity value from one test to
another is a function of water to solid ratio and additives specific to each
leaching procedure.
Table 2 is a summary of GFETC solid waste heavy metal constituents in
leachates as a function of the leaching test procedures also showing smaller
particle sizes leaching greater quantities of specific key metals. For
comparison purposes, table 3 list results of leaching tests using "H-Coal
#3" solid waste samples. Similar data was developed for the H-Coal #4, METC
and Chapman leachates, as was done to the GFETC leachates. Table 4 is a
summary of the compositional results of leaching of lime and alum sludges
produced from the treatment of METC gasification wastewaters.
385
-------
Figure 2
1 1 1 1 1 1
120000- a
100000- ^^"^^
80000 - ^^v.
A ^V^
~ 60000 - ^V^ -
2. A - Pitt
>. - a- EPA-EP
« 16000 -
1 ft
o 12000- Y
8000 - N^a
4000 - ^ ^ -
0 i i i i ' i
s
X
0 0.1 02 03 0.4 05 0.6 0.7
3 (mm)
Conductivity Data - GFETC 1IO. 1
(Gasifier Bottom Slag from RA-52 Using Lignite Coal)
TABLE 2 . RESULTS OF LEACHING TESTS - GFETC
(GASIFIER BOTTOM SLAG FROM RA-52 UAING LIGNITE COAL
Test
Procedure
ASTM-A
ASTM-A
ASTM-A
ASTM-A
ASTM-A
EPA-EP
EPA-EP
EPA-EP
EPA-EP
EPA-EP
Pin
pin
pin
pin
PITT
Metal Concentrations (all units mg/1)
Mesh
Size pH Ag Be Ca Cd Cr Cu Fe Mq Mn Pb Zn
20-40 9.62 0 0 196 0 00 0 1.7 0 0 0.1
40-60 9.25 0 0 220 0 00 0 1.8 0 0 0.1
60-100 9.01 0 0 300 0 0 0 0 1.8 0 0 0.3
100-200 8.80 0 0 380 0 00 0 2.1 0 0 0.4
<200 8.58 0 0 580 0 0 0 0 2.2 0 0 0.7
20-40 4.97 0 0 344 0 00 84 75 0.8 0 0.3
40-60 4.98 0 0 540 0 00 96 90 1.6 0 0.3
60-100 5.44 0 0 800 0 0 0 165 100 1.8 0 0.4
100-200 5.66 0 0 1140 0 0 0 78 105 2.9 0 0.4
<200 5.76 0 0 1440 0 0 0 93 105 5.3 0 0.9
20-40 2.03 0 0 2870 0 0 0 420 1200 2.9 0 0.7
40-60 1.70 0 0.5 4800 0 0 0 540 1980 4.4 0 1.3
60-100 1.64 0 0.5 5400 0 2.5 0 680 2200 5.0 1.5 1.4
100-200 1.69 0 0.8 6400 0 3.7 0 1040 2940 8.0 1.5 2.3
<200 1.74 0 0.5 3920 0 3.7 52.8 2150 1700 10.5 1.5 39.0
386
-------
TABLE 3. RESULTS OF LEACHING TESTS - H-COAL #3
(LIQUEFACTION VACUUM STILL BOTTOMS FROM SYNCRUDE MODE USING ILLINOIS COAL)
Metal Concentrations (all units mg/1 except Hg - ug/1)
Test
Procedure
ASTM-A
ASTM-A
ASTM-A
ASTM-A
ASTM-A
EPA-EP
EPA-EP
EPA-EP
EPA-EP
EPA-EP
PITT
PITT
PITT
PITT
pin
Mesh
Size
20-40
40-60
60-100
100-200
<200
20-40
40-60
60-100
100-200
<200
20-40
40-60
60-100
100-200
<200
DH
11.03
11.22
11.33
11.48
11.49
4.96
4.90
4.88
4.87
.91
.72
.74
.76
.77
1.71
TABLE 4. RESULTS
(FROM
Aq
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Be
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Ca
130
160
190
300
370
150
220
320
400
420
184
270
340
430
410
OF LEACHING
TREATMENT OF
Concentrations
Material
LIME
LIME
ALUM
ALUM
Test
Procedure
ASTM-A
EPA-EP
ASTM-A
EPA-EP
Cd Cr
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
TESTS
METC
Cu Fe
0 0
0 0
0 0
0 0
0 0
0 1.1
0 2.0
0 3.4
0 4.8
0 4.8
0.9 7.0
1.3 10.0
1.3 15.0
1.3 33.0
1.3 32.0
Hg
0.6
0.3
0.6
0.3
0.3
0
0
0
0.3
0
0.3
0
0.3
0.3
0.3
Mg
0.2
0.3
0.4
0.4
0.3
1.0
1.2
1.7
2.0
2.0
1.4
1.7
1.9
4.8
5.5
Mn Pb
0 0
0 0
0 0
0 0
0.2 0
1.0 0
1.4 0
2.3 0
2.9 0
2.9 0
1.2 0
1.6 0
2.3 0
3.1 0
2.9 0
Zn
0
0.1
0.1
0.3
0.6
0
0
0.1
0.3
0.4
0.3
0.3
0.4
0.4
0.4
- LIME AND ALUM SLUDGES
WASTEWATER
(all units mg/1) *
Conductivity TOC
pH
11.84
7.29
7.75
6.92
Ca
640
1620
340
1800
Cd
0
0.1
0
0
Cu
0
0
0
0.3
Fe
1.4
0.8
7.0
0.8
MB
0.1
26
176
26
Zn
0
0.2
0
0.4
(urahos)
4800
6400
5300
7100
(rag/1)
1590
2630
1960
2800
Phenol
(mg/1)
720
155
1550
455
*Ag, Be, Cr, Mn, Pb below detectable limits
Figure 3 is a plot of Daphnia toxicity information for leachates
derived from the EPA-"EP" extraction procedures utilizing the smallest and
largest particle sizes of GFETC solid waste residuals. The 48 hour LC-50
values for the largest particle size is 8.9% dilution, while for the
smallest particle size, is 7.0% dilution. In a summary of LC-50 data on
table 5, ASTM-A distilled water leaching protocol always showed LC-50 values
on the order of 100% dilution with lower values for the EPA-EP test. It may
thus be concluded that the inherent nature of the EPA-"EP" procedure using
acetic acid causes Daphnia toxicity, thus raising questions as to the
validity of the application of Daphnia toxicity testing to leachates
produced in accord with the EPA approach. Evaluation, however, of Daphnia
toxicity to acetic acid reagents as used in the EPA extraction procedure
test where no solid wastes are leached (after neutralization) show the LC-50
387
-------
value to be 85% in concentration and thus do not explain the total toxicity
of EPA leachates of solid wastes to Daphnia. It was noted from this and a
series of similar tests, that results using EPA extraction procedure
protocols appear far more toxic to Daphnia in all cases than results using the
ASTM-A "distilled water" approach.
100
10
O-porttota tin • 20-4O
D-partfete tin • <200
10
2O 30 4O 50 60 70 80
90 95
98
Percentage Dead
Figure 3. Daphnia Toxicity Data From GFETC NO. 1 EPA-EP LEACHATES
TABLE 5. RESULTS OF DAPHNIA MAGNA TOXICITY TESTING
Katerlal
CFA
CFA
CFA
CFA
H-COAL 14
H-COAL 14
H-COAL »4
H-COAL »4
H-COAL 13
H-COAL 13
H-COAL 13
H-COAL 13
GFETC 11
6FETC 11
GFETC 11
GFETC fl
GFETC 12
GFETC 12
METC
METC
METC
METC
LIKE SLUDGE
LIME SLUDGE
ALUM SLUDGE
ALUM SLUDGE
ACID CONTROL
Test
Procedure
ASTM-A
ASTM-A
EPA-EP
EPA-EP
ASTM-A
ASTM-A
EPA-EP
EPA-EP
ASTM-A
ASTM-A
EPA-EP
EPA-EP
ASTM-A
ASTM-A
EPA-EP
EPA-EP
ASTM-A
EPA-EP
ASTM-A
ASTM-A
EPA-EP
EPA-EP
ASTM-A
EPA-EP
ASTM-A
EPA-EP
—
Mesh
Size
20-40
<200
20-40
<200
20-40
<200
20-40
<200
20-40
<200
20-40
<200
20-40
<200
20-40
<200
< 60
< 60
> 10
< 60
> 10
< 60
..
..
..
..
~
48hr LC50
t
>100
>100
2.9
4.B
>100
>100
2.0
l.B
>100
>100
22
23
>100
>100
8.9
7.0
>100
11.2
>100
>100
0.28
0.12
1.4
4.3
0.74
1.6
85
9SI Confidence
UCL
..
..
3.4
5.8
._
2.5
2.3
..
27
29
__
11.1
8.0
12.6
0.37
0.17
1.6
5.2
0.82
1.7
95
Limits
LCL
..
2.4
3.9
_*
1.6
1.4
..
.„
18
18
_.
7.1
6.1
9.9
_.
0.20
0.09
1.2
3.6
0.67
1.4
76
388
-------
Table 5 also indicates that Daphnia toxicity to wastewater treatment
sludge leachates are far more toxic than leachates produced from gasification
bottom ash or slag residuals. As one illustration, figure 4 is a plot of
Daphnia toxicity from leachates produced from lime sludges generated from the
waste treatment scheme when processing METC fixed bed coal conversion
wastewaters. As can be seen, despite the numerical difference of results
when leaching via the EPA or the ASTM-A approach, toxicities of waste
treatment plant sludges are considerably greater than toxicities of ash/slag
leachates. It should be noted, however, that in no case did primary drinking
water heavy metals exceed 100 times drinking water standards; thus implying
that such sludges are not to be considered as hazardous in a "RCRA" context.
IO.O
1
u 1.0
O.I
O- ASTM-A
D- EPA-EP
10 20 30 40 50 60 70
Percentage Dead
80
90 95 98
Figure 4. Daphni.3 Toxicity Data From Lime Sludge Leachates
CORRELATION OF DAPHNIA TOXICITY DATA WITH CHEMICAL COMPOSITION DATA
Biesinger (3) published Daphnia magna 48 hour LC-50 data for various
primary and secondary EPA drinking water metals. Figure 5 is our plot of
Biesinger's data illustrating that an empirical relationship exists for most
heavy metals with EPA drinking water standards. It should be noted that
drinking water standards are not based upon Daphnia magna toxicity data, and
the correlation illustrated on figure 5 is quite emperical. The outlier
points for copper and zinc represent the extreme toxicity of these metals to
Daphnia magna, and lack of such toxicities to mammals and humans in specific,
389
-------
1000
10.0
1.0 r
0.1
0.01
0 Bo
OPb
ZnD
DCu -
. OHg
OOQI* ' ' ' 'I ' iiil i i i il i tii
0001 0.01 O.t 1.0 10.0
EPA Drinking Water Standard (mg/D
Figure 5. Correlation Between Daphnia Toxicity and Drinking Water Standards
In an attempt to correlate LC-50 data with metal characterization data
of leachates, measured metal concentrations were weighted to account for the
fact that some substances are more toxic to Daphnia magna than are other
metals. The weighting procedure used was based upon the following equation:
m = (Ca/Ca0)+(Cd/Cd0)+(Cu/Cu0)+(Fe/Fe0)+(Mg/Mg0)
+(Mn/Mn0)+(Pb/Pb0)+(Zn/Zn0)
The numerator for each metal is the metal concentration for leachates as
measured, and the denominator represents data published by Biesinger (3).
The resultant equation, for application to data developed in this research is:
XM = (Ca/52)+(Cu/0.01)+(Fe/9.6)+(Mg/140)
+(Mn/9.8)+(Zn/0.1)
Figure 6 is a correlation of Daphnia toxicity LC-50 values as measured
in the course of this research with the measured weighted metal concentrations
(ZM) for leachates generated from coal conversion ash and slag residuals.
The correlation with trace metals did not hold for wastewater treatment
plant sludges, however, as shown on figure 7, LC-50 values for Daphnia
toxicity are correlated with phenol concentrations measured in the ASTM-A
and EPA "EP" leaching protocols of generated lime and alum wastewater
treatment sludges.
390
-------
i
1000.0
100.0
10.0
1.0
O.I
-CF»
-H-Co.l II---
-K-Co«l «J---
-H-Co«l 14--
-H-Coil 14--
-tFITC II---
-CFETC II---
-eF[TC 12---
10---«TC
I1---MTC
02
20-40
<200
20-40
<200
20-40
.200
20-40
.200
.60
06
10
100
1000
EM
DOOO
Figure 6. Relationship Between Daphnia Toxicity and Weighted Metal
Concentrations of Ash and Bottom Leachates
IQO
1.0
0.1
4A
2—Lin Sludg«—EPA-EP
3—Aim Sludge—ASTM-A
4—AluB Sludge—EPA-EP
. .1
too 1000 loooo
Phenol Concentration (mg/l)
Figure 7. Relationship Between Daphnia LC50 Data and
Phenol Concentration of Sludge Leachates
391
-------
AMES TESTING FOR MUTAGENICITY POTENTIALS
SRC-II solid wastes generated at the Ft. Lewis pilot plant were
provided by the Pittsburgh and Midway Coal Company under the guidance of the
U.S. Department of Energy, Pittsburgh Energy Technology Center. The samples
received were two shipments of vacuum bottoms from the vacuum flash drum
of the direct liquefaction step, and wastewater sludges consisting of alum
sludge from the pre-biological (flotation) step, waste activated biological
sludge, and digested activated biological sludge. These samples were
subjected to simple chemical screening analysis and Ames testing for
potential mutagenicity. Both whole materials and liquid phases filtered from
whole materials (for clarifier and digester biosludges) were evaluated for
mutagenicity. In addition, a serial organic extraction protocol was
developed using hexane, toluene, methylene chloride, and acetonitrile to gain
a qualitative assessment of the polarity and chemical nature of leached
substance causing mutagenicity. For clarifier and digester biosludges, both
liquid and solid phases of filtered sludges were analyzed for Ames
mutagenicity. It is interesting to note that in all cases, the filtrate of
sludge sanples showed negative Ames results while the whole sample and
retained filtered solid samples showed positive results. Samples of the
dried sludge were processed by sequential organic extraction as outlined
above to generate four additional extracts and residue for testing. Five
tester strains, TA98, TA100, TA1535, TA1537 and TA1538 were employed in the
tests with and without microsomal activation. Routine sterility and toxicity
checks were made during the course of the run. It was found that none of the
mutagenicity tests which used TA1535 without S9 and TA1537 with S9 resulted
in positive plates in early phases of the investigation, therefore, TA1535
was eliminated from all tests and TA1537 were not used when microsomal
activation was applied.
The standard criteria used to define a positive result in the Ames
bioassay for mutagenic activity include (a), a two-fold or greater increase
in the number of revertants exposed to the test material compared to
respontaneous revertant rates; (b), repeatability i.e...a confirmation of the
positive result by running the test again after a two week period; (c), for
compounds of low mutagenicity, a reproducible dose response rate.
Ames test results were uniformly negative where microsomal activation
was not included in the test procedures. The most sensitive tester strain
showing the greatest number of revertants compared to the spontaneous
revertant rate is shown to be strain TA98 with S9, a result which agrees
with other investigations. All other tester strains showed marginal positive
results.
Toxicity to the tester organisms by test materials was encountered in
all of the individual bioassays with alum sludges showing perhaps the
highest toxicity. It is hypothesized that this may be due to either
organics, or to the fine alum "slime" particles which interfered with growth
of revertants.
The highest mutagenicity activity observed (revertants per mg) were
exhibited by the vacuum bottoms solid wastes. Vacuum bottoms solid wastes
392
-------
contain organics that are not highly water soluble and thus the probability
of release in a landfill is small. Clearly, mutagenic substances exist on
these solid waste residuals as evidenced by conducting Ames testing of whole
materials dissolved in organic solvents, but aqueous leachates show no
mutagenic activity. Philosophical questions are raised leading to a need for
policy delineation by EPA as to the acceptability of disposing of such
materials in hazardous or conventional waste landfills.
Table 6 is a summary of Ames test results using tester strains TA98
with S9 activation. It should be noted that the average number of
revertants for the control is 46 with standard deviation of 16. This
table shows the whole vacuum bottoms residual gave rise to 13.2 times the
number of revertants on a negative control plate (spontaneous revertants)
for sample number 1936, while for vacuum bottoms sample number 2277 (sample
numbers provided by the Pittsburgh and Midway Coal Mining Company), a ratio
of 21.4 times spontaneous revertants were observed. As can be seen from
this table, from a column of data using whole substances, the vacuum
bottoms from the SRC-II Ft. Lewis facility show far more mutagenicity than
do the other solid residuals measured. All solid residuals, as observed
from the Ft. Lewis facility, give rise to positive mutagenicity potentials
as determined by criteria of values being greater than 2 times the
spontaneous revertants indicating positive Ames tests.
TABLE 6. AMES TESTING OF SRC-II FORT LEWIS SOLID WASTE RESIDUALS
AND EXTRACT FRACTIONS
TEST SAMPLE
VACUUM BOTTOMS
SAMPLE
NUMBER
1936
2277
ORGANIC SOLVENT EXTRACT FRACTIONS
METHYLENE
WHOLE*
rev/mg
2875
11095
(R)
13.2
21.4
HEXANE
rev/mg
11600
2567
(R)
16.6
15.6
TOLUENE
rev/mg
3695
12200
(R)
39.4
44.0
CHLORIDE
rev/mg
2365
5552
(R)
13.6
13.8
ACETONITRILE
rev/mg
20330
9578
(R)
25.5
19.9
RESIDUE
rev/mg
560
NM
(R)
2.5
2.0
ALUM SLUDGE
CLARIFIER SLUDGE
DIGESTER SLUDGE
2280
1937
2278
1938
2279
7705
4050
932
530
1273
3.0
5.8
3.8
6.2
2.8
NM
1268
NM
236
NM
NM
3.9
NM
2.6
NM
NM
1290
NM
984
2803
NM
3.9
NM
2.8
2.4
1987
1349
960
NM
868
5.1
5.0
2.4
NM
2.5
1295
14720
NM
NM
NM
4.5
31.0
NM
NM
NM
NM
NM
NM
NM
247
NM
NM
NM
NM
2.5
NOTE:
* "Specific Mutagenic Activity" for Vacuum Bottoms in revertants/mg for all other whole materials (sludges)
in revertants/ml.
NM Not Mutagenic
R Ratio of revertants on test plate (spontaneous + induced)/spontaneous revertants on control plate
393
-------
Work is continuing at the University of Pittsburgh in the area of
evaluating mutagenicity potentials of wastewater sludges and their leachates
when treating GFETC wastewaters, and the evaluations of methodologies of
changing the pretreatment and biological treatment step to minimize such
mutagenicity and toxicity potentials.
RESEARCH IN COMPUTER MODELING OF LEACHATE CONCENTRATIONS
AT LANDFILL BOUNDARIES
A series of five simultaneous differential equations utilizing concepts
of diffusivity, film diffusion, intraparticle diffusion and liquid-solid
equilibria has been developed for the prediction of leachate compositions
at the boundary line from a landfill containing coal gasification solid
waste residuals. The model is based upon deterministic concepts and simple
equilibrium and diffusion data and was calibrated in the lab using GFETC #1
slag residuals. The system of simultaneous differential equation has been
solved using numerical computational methods.
This model has been extrapolated from lab scale to predict concentration
profiles of a commercial scale landfill (600 meters x 600 meters x 6 meters
deep) filled with coal conversion solid wastes. Under the assumption of
unidirectional flows and small fluid velocity, profiles of concentration
with duration at the landfill boundary were computed. As an example of the
results of computer modeling, figure 8 is a plot of predicted concentration
in leachates versus time for a coal conversion solid waste landfill where
particle sizes are in range of 20 to 40 mesh at flooded conditions with
indicated groundwater velocities through the landfill site.
345
TIME (YEAR)
Figure 8. Calculated Calcium Concentration Profiles for a Landfill
394
-------
By understanding the concepts and implications of intraparticle
diffusion, figure 9 was developed for conditions of discontinuous flow. The
breaks in the curve indicate periods of "no rain" or dry conditions. This
figure illustrates the concept of "a first flush phenomena" by showing that
under flooded conditions, steady state mass transfer from the solid phase
to the liquid phase occurs, and is predictable. Under dry conditions,
however, concentration of leachable pollutants at the particle surface
increases with time to become uniform throughout due to intraparticle
diffusion. The first flush phenomena, as shown on figure 9, predicts an
increase in aqueous concentration over that which would be ordinarily
expected under the flooding conditions shown in figure 8. The overall area
under all the curves of figure 9 is proportional to the total quantity of
leachable substances produced. As can be seen, commensing of flooding causes
high excursions in aqueous concentration; this concept we call a "first
flush phenomena". This model may predict difficulties in compliance with
concentration restrictions as outlined by RCRA regulations for ash/slag
landfills under conditions of "first flush" or discontinuous flows.
Additional details are available in a dissertation by Erdogan (4).
H
z
u
u
z
o
CJ
2000
1500
1000
500
Fluid velocity:0.005 cm/min
Particle size :20-40 Mesh
TIME (YEAR)
Figure 9. Predicted Concentration Profile for a Landfill with Discontinuous
Flow
395
-------
ACKNOWLEDGEMENTS
The Senior Author would like to thank the Grand Forks Energy Technology
Center for solid waste samples and for support of Mr. S. Wallach who
conducted Daphnia toxicity evaluations, Mr. H. Erdogan who conducted
mathematical modeling of solid waste residuals, Mr. J. Bern who conducted
Ames toxicity testing and developed concepts for disposal of solid waste
residuals. The Morgantown Energy Technology Center provided wastewater and
supported Mr. C. Moretti, Graduate Student, who conducted treatability
evaluations of METC gasifier wastewater and developed the wastewater sludges
used in this research. The Pittsburgh Energy Technology Center provided
supported testing and provided samples of SRC-II solid waste residuals for
Ames test evaluation as conducted under the supervision of Dr. Keleti at the
Graduate School of Public Health.
REFERENCES
(1) Bern, J., Neufeld, R.D., Shapiro, M. , "Solid Waste Management of Coal
Management of Coal Conversion Residuals from a Commercial Sized
Facility: Environmental Engineering Aspects" Final report under
contract //DE-AC22-79PC20023 available from DOE as DOE/ET/20023-5,
November 30, 1980.
(2) Neufeld, R.D., Wallach, S.H., Erdogan, H., Bern, J., "Chemical and
Biological Properties of Coal Conversion Solid Wastes" Technical Annual
Report to the Grand Forks Energy Technology Center, U.S. DOE available
as DOE/ET/10061-1.
(3) Biesinger, K.E. and Christenson, G.M., "Effects of Various Metals on
Survival, Growth, Reproduction and Metabolism of Daphnia magna,
Journal of the Fisheries Research Board of Canada, Vol.29, No.12 (1972),
pp. 1691-1700.
(4) Erdogan, H., (1981) Dissertation submitted to University of Pittsburgh,
"Mathematical Modeling of Leaching with Intraparticle and External Film
Diffusion as Rate Controlling Mechanisms: Application to Coal Conversion
Solid Waste".
396
-------
UPDATE ON EPA'S REGULATORY VIEWS ON COAL CONVERSION SOLID WASTESt
by: Yvonne M. Garbe
Office of Solid Waste
U.S. Environmental Protection Agency
Washington, DC 20460
ABSTRACT
The Resource Conservation and Recovery Act of 1976 (RCRA) charges EPA
with the responsibility for establishing a program for the management of
hazardous solid wastes. This paper summarizes current and anticipated RCRA
regulations affecting the synfuels industry. Included in the various RCRA
issues pertaining to the synfuels industry is a discussion of the RCRA
mining exemption. An overview is given of the Office of Solid Waste's
planned research activities to support, future synfuels solid waste regula-
tions .
(Only the abstract is published herein.)
397
-------
Session V: MULTIMEDIA ENVIRONMENTAL CONSIDERATIONS
Chairman: T. Kelly Janes
U.S. Environmental Protection Agency
Research Triangle Park, NC
Cochairman: John T. Dale
U.S. Environmental Protection Agency, Region VIII
Denver, CO
398
-------
A PERMITTER'S VIEW OF SYNFUEL COMMERCIALIZATIONT
by: George L. Harlow
Air and Hazardous Materials Division
U.S. Environmental Protection Agency
Region IV
Atlanta, GA 30365
ABSTRACT
The Environmental Protection Agency has responsibility for the issu-
ance of permits to synfuel plants for the control of various liquid, gas-
eous, and solid waste streams. These permits comprise the Prevention of
Significant Deterioration (PSD) under the Clean Air Act of 1977, the
National Pollutant Discharge Elimination System (MPDES) and the Section 404
Dredge and Fill permits under the Clean Water Act of 1977 and the hazardous
waste permits under the Resource Conservation and Recovery Act (RCRA) of
1976.
Since there will likely not be federal regulations established by EPA
setting standards on requirements for the first generation synfuel plants,
the environmental permits will have to be individually negotiated, case by
case, with each applicant using best engineering practice. This places an
unusual burden upon the permit writer who will be negotiating with the
discharger from an uninformed and defenseless position. In order to over-
come this burden and to avoid long, time-consuming delays in the permit
process, the company should disclose in its application for permit exactly
what steps will be taken to control air emissions, water discharges and
hazardous wastes.
(Only the abstract is published herein.)
399
-------
COMPARISON OF ENVIRONMENTAL DESIGN ASPECTS t
OF SOME LURGI-BASED SYNFUELS PLANTS
Milton R. Beychok, Consulting Engineer
William J. Rhodes, EPA/IERL-RTP
INTRODUCTION
A number of commercial-scale projects have been proposed in the United States for the
production of gas and liquid synfuels from coal. Many of these proposed projects are planning
to use Lurgi coal gasifiers and related Lurgi technology such as the Rectisol gas purification
process and the Phenosolvan process for recovering phenols from coal gasification wastewaters.
These projects represent several different architectural and engineering contractors and, therefore,
probably different design philosophies and preferences. As a result, a comparison of how each
contractor handled some of the environmental concerns would indicate a segment of industry's
views on plant configurations and control alternatives.
Table 1 identifies 14 Lurgi-based synfuels projects which are currently being proposed,
studied, or underway in the United States. In terms of their design progress, their environmental
permitting status, and their investment financing arrangements, the most advanced project among
those listed in Table 1 appears to be the Great Plains Gasification Associates' project in North
Dakota. Some of the other projects have completed fairly detailed feasibility studies and have
prepared environmental impact studies as well as environmental permitting applications. How-
ever, none of the other projects appear to be as well advanced as the Great Plains project in
North Dakota.
Process design information has been obtained for five of the projects listed in Table 1J ~9,
and this paper describes and compares the key environmental design aspects and features of these
five projects:
• The Great Plains Gasification Associates' project in North Dakota
(initiated by the American Natural Gas Service Company).
• The Hampshire Energy Company's project in Wyoming.
• The Nokota Company's project in North Dakota (initiated by the
Natural Gas Pipeline Company of America).
• The Tenneco project in Montana (known as the Beach-Wibaux project).
• The WyCoalGas, Inc. project in Wyoming (a subsidiary of Panhandle
Eastern Pipe Line Company).
400
-------
TABLE 1. PROPOSED COAL GASIFICATION PROJECTS PLANNING TO USE LURGI GASIFIERS
PROJECT SPONSOR
Tri-State Synfuels
Louisiana Gasification Associates
Crow Indians
Tenneco Coal Gasification
Texas Eastern Synfuels
Great Plains Gasification Associates
Nokota Company
North Dakota Synfuels Group
Exxon USA
Transco Energy Company
Ohio Valley Synthetic Fuels
Hampshire Energy Company
Lake Desmet Synfuels
WyCoalGas, Inc
LOCATION
Kentucky
Louisiana
Montana
Montana
New Mexico
N. Dakota
N. Dakota
N. Dakota
Texas
Texas
W. Virginia
Wyoming
Wyoming
Wyoming
TYPE OF
GASIFIERS
Lurgi
Lurgi
Lurgi
Lurgi
Lurgi
Lurgi
Lurgi
Lurgi
Lurgi
Lurgi
BGC/Lurgi c
and Texaco
Lurgi and
KBWa
Lurgi
Lurgi
TOTAL COAL,
T/D (Mg/D)
28,000 (25,400)
n.a.
12,000(10,900)
41,000 (37,200)
30,000 (27,200)
28,700 (26,000) b
42,000 (38,100)
20,000(18,100)
42,000(38,100)
n.a.
50,000 (45,400) d
15,000(13,600)
38,000 (34,500)
32,600 (29,600)
A&E
CONTRACTOR
Fluor
n.a.
Fluor
Fluor
Bechtel
Lummus
Fluor
Stone &
Webster
n.a.
n.a.
Foster
Wheeler
Fluor
n.a.
Bechtel
PRIMARY
PRODUCTS
High-Btu SNG,
gasoline, and
chemicals
Synthesis gas
High-Btu SNG
High-Btu SNG
High-Btu SNG and
methanol
High-Btu SNG
High-Btu SNG and
methanol
High-Btu SNG and
methanol
Synthesis gas
Medium-Btu gas
High-Btu SNG and
methanol
Gasoline, propane,
and butanes
High-Btu SNG and
methanol
High-Btu SNG
a Koppers and Babcock-Wilcox entrained gasifiers.
b To be built in 2 phases, each for 1-
c British Gas Corporation and Lurgi
** To hf« hnilt in 3 cihasps: chase 1 =
4 ,350 T/D (13,000
slagging gasifiers.
5. 000 T/D (4500 l\
Mg/D) of coal.
yia/D) coal, ohase 2 = 25
.000 T/D (22.700Ma/D) (
=oal. ohase 3 = 50.000 T
/D (4B4DO Ma/m
n.a. Not available, or not yet selected
-------
The gasifier feedstock coals for the five projects are summarized below, on a "run-of-mine"
basis:
PROJECT
Great Plains
Hampshire
Nokota
Tenneco
WyCoalGas
COAL RANK
Lignite
Subbituminous
Lignite
Lignite
Subbituminous
AMOUNT
T/D (Mg/D)
28,670 (26,000)
15,000(13,600)
28,350 (25,700)
33,000 (29,900)
22,820 (20,700)
HEATING VALUE,
Btu/lb (kJ/kg)
7,185(16,710)
8,075 (18,780)
6,985 (16,250)
7,020(16,330)
8,450(19,650)
SULFUR
%Wt
1.01
0.33
0.85
0.82
0.32
These coal amounts refer specifically to the gasifier feedstock coal, whereas the amounts
given in Table 1 include any coal burned in boilers to generate plant steam as well as any
coal fines returned to the mine or sent elsewhere.
OVERALL PROCESS DESIGNS
The five coal-to-synfuels plant designs described in this paper use a number of in-
dividual process steps, arranged in various configurations. The major process steps are
briefly described below:
Lurgi gasification — Coal, steam, and oxygen are reacted and result in a crude gas containing
hydrogen, carbon monoxide, carbon dioxide, methane, excess steam, hydrogen sulfide,
ammonia, and various byproducts and impurities. The crude gas is washed and
cooled, condensing out a "gas liquor" containing water, tars, oil, phenols, and
ammonia.
Shift conversion — Part of the carbon monoxide in the crude gas is "shifted" (i.e., converted
to carbon dioxide and hydrogen), so as to provide the ratio of hydrogen to carbon
monoxide needed for the subsequent synthesis of methanol or methane. The shifted
gas is then further cooled, condensing out additional gas liquor.
402
-------
Gas purification — The acid gases hydrogen sulfide and carbon dioxide are removed from
the shifted gas by absorption in a solvent, using the Rectisol process and, in one case,
the Selexol process. The shifted and purified gas is then routed to the subsequent
synthesis step to produce either methanol or methane. The absorbed gases are stripped
from the absorption solvent and recovered as acid gas streams. Those which are rich in
hydrogen sulfide are processed further for conversion into sulfur.
Methanol synthesis — The hydrogen and carbon monoxide in the purified gas are reacted
in the presence of a specific synthesis catalyst to form methanol. The methanol
synthesis step also generates a purge gas stream, which may be further processed for
conversion into methane and/or to provide a source of hydrogen for hydrotreating of
Lurgi byproduct naphtha. The methanol produced may be sold as a product or may be
processed further for conversion into gasoline.
Methanation — The hydrogen and carbon monoxide in the purified gas (from shift con-
version and gas purification), or in the methanol synthesis purge gas, are reacted in
the presence of a specific methanation catalyst to form methane. Methane is the
principal constituent of the product SNG (substitute natural gas).
Gas liquor cleanup — Tars and oils are separated from the gas liquor and recovered. Next,
the bulk of the phenols in the gas liquor are removed by the Phenosolvan process 10,
which uses extraction by a selective solvent. Ammonia is then stripped from the
dephenolized gas liquor and recovered as a byproduct. The further treatment of
the residual wastewater (stripped and dephenolized gas liquor) is described later in
this paper.
Partial oxidation — Liquid hydrocarbon byproducts (such as the Lurgi tars, oils, naphtha,
and phenols) may be reacted with steam and oxygen to result in a crude gas con-
taining hydrogen, carbon monoxide, carbon dioxide, excess steam, hydrogen sulfide,
ammonia, and a very small amount of methane and other impurities. The subsequent
processing of the partial oxidation crude gas is very similar to that described herein for
the Lurgi crude gas.
Gasoline production — The MTG (methanol to gasoline) process first catalytically converts
methanol to a mixture of methanol, dimethyl ether, and water vapor. The methanol
and dimethyl ether are then catalytically converted to form hydrocarbons in the
403
-------
gasoline boiling range (C4 to C1Q). The hydrocarbons are fractionated into stabilized
gasoline, LPG, and butanes. Part of the butanes are alkylated to form additional
high-octane gasoline. The Lurgi naphtha is desulfurized in a catalytic hydrotreater to
provide an additional gasoline component. Thus, the product gasoline includes stabi-
lized MTG gasoline, alkylate, and hydrotreated Lurgi naphtha.
KBW gasification — Lurgi gasifiers require a sized coal in the range of 0.25 - 1.50 in.
(0.64 — 3.8 cm). Thus, the coal fines produced from crushing and sizing of run-of-
mine coal could be used as boiler fuel, disposed of in the mine or elsewhere, or gasified
in some other type of gasifier. Entrained bed gasifiers, such as the KBW gasifiers (see
Table 1), may be used to react the coal fines with steam and oxygen to produce a
crude gas containing hydrogen, carbon monoxide, carbon dioxide, excess steam,
hydrogen sulfide, ammonia, and a very small amount of methane and other impurities.
The subsequent processing of the crude gas is very similar to that described herein for
the Lurgi crude gas.
Sulfur recovery — It is beyond the scope of this paper to describe the many different
processes that could be used for converting hydrogen-sulfide-rich acid gases into
recovered sulfur. However, since four of the five coal-to-synfuels plant designs dis-
cussed in this paper plan to use the Stretford process, that process is described briefly
herein.
The Stretford process involves liquid-phase oxidation of hydrogen sulfide in an aque-
ous solution of sodium vanadate and anthraquinone disulfonic acid (ADA). The
hydrogen sulfide is absorbed and oxidized to sulfur, which is subsequently removed as
a froth by flotation and purified by centrifuging followed by melting. The Stretford
process can be designed to remove essentially all of the hydrogen sulfide in the feed-
stock gas and convert it into byproduct sulfur. However, the Stretford process
accomplishes little, if any, removal and conversion of organic sulfur compounds such
as carbonyl sulfide (COS), carbon disulfide (CS2), and mercaptans (RSH), all of
which are present in varying amounts in the gasification crude gases.
404
-------
The flow diagram in Figure 1 presents the overall process design for the Hampshire
project. Some key points of this design are:
• The products and byproducts are gasoline, LPG, butanes, sulfur, ammonia, and
carbon dioxide.
• Coal fines are gasified in KBW gasifiers, eliminating the need to burn any coal fines.
• Plant steam and power are supplied by burning methanol synthesis purge gas in gas
turbines and generating steam by recovering heat from the turbine exhaust flue
gases.
• Lurgi byproduct tars, oils, and phenols are gasified via partial oxidation.
• The combined Lurgi, KBW, and partial oxidation crude gases are purified in a
selective Rectisol unit.
• Sulfur recovery utilizes the Adip, Glaus, and Scot processes.
The flow diagram in Figure 2 presents the overall process design for the Tenneco
project. Some key points of this design are:
• The products and byproducts are SNG, sulfur, and ammonia.
• Coal fines are burned in steam-generating boilers. Electrostatic precipitators (ESPs)
followed by wet limestone scrubbers provide flue gas particulate removal and flue
gas desulfurization.
• Lurgi tars, oils, naphtha, and phenols are gasified via partial oxidation.
• The Lurgi crude gas is purified in a non-selective Rectisol unit, and the partial
oxidation crude gas is purified in a selective Selexol unit.
• Sulfur recovery utilizes the Stretford process.
The flow diagram in Figure 3 presents the overall process design for the Nokota pro-
ject. Some key points of this design are:
• The products and byproducts are methanol, SNG, phenols, oil, naphtha, sulfur, and
ammonia. Excess coal fines will be either a byproduct or waste.
• Coal fines and Lurgi tars are burned in steam-generating boilers. Dry scrubbing
followed by baghouses provide flue gas desulfurization and flue gas particulate
removal.
• The Lurgi crude gas is purified in a selective Rectisol unit.
• Sulfur recovery utilizes the Stretford process.
405
-------
Sized
coal
Oxygen
Coal
fines
Oxygen
Steam
Gas 1 i quor
LURGI
GASIFIERS
KBW
GASIFIERS
t
GAS LIQUOR
CLEANfJP
GaS
SHIFT
CONVERSION &
GAS COOLING
COa'rich gas from Rect sol
• i-
to pipeline
Byproduct
ammonia
Tars, oils,
phenols
^Wastewater
to treating
PARTIAL
OXIDATION
i_
Oxygen
Carbon
di oxi de a
GAS PURIFICATION
&
SULFUR RECOVERY
naphtha
t
Byproduct
su 1 fur
METHANOL
SYNTHESIS
Methanol
Plant steam
and power
1
GASTURB.NES
& WASTE
HEAT BOILERS
HYDROGEN
SEPARATION
Purge
gas
Hydrogen
LPG and butanes
GASOLINE
PRODUCTION
, . . .
Lurgi naphtha
3 r
Hydrogen
HYDRO-
TREATING
FIGURE 1. HAMPSHIRE ENERGY COMPANY (WYOMING) PROJECT
-------
(0
3 Sized
coal
GAS LIQUOR
CLEANUP
. Byproduct
ammonia
Tars, oils,
phenols
^Wastewater
to treating
JC.
CL
SHIFT
CONVERSION &
GAS COOLING
COS
HYDROLYSIS
Carbon
dioxide a
GAS
PURIFICATION
(SELEXOL)
Lurgi naphtha
Gas 1i quor
LURGI
GASIFIERS
Oxygen
Steam
SHIFT
CONVERSION &
GAS COOLING
o
<.
Acid gas
SULFUR
RECOVERY
Carbon
dioxi de
Byproduct
sulfur
METHANATION
GAS
PURIFICATION
(RECTISOL)
METHANATION
SNG
a i
es
BOILER
PLANT
and p
ower
FLUE GAS
SCRUBBING
J L
g
i
as
L
C02~rich gas to atmosphere
Incinerated tail gas from
sulfur recovery
FIGURE 2. TENNECO COAL GASIFICATION (MONTANA) PROJECT
-------
Ul
CJ
LURGI
GASIFIERS
Oxygen-
Steam-
Byproduct
ammonia
GAS LIQUOR
CLEANUP
Tars
to boiler
Wastewater
to treating
Carbon
d i ox i de a
Byproduct phenols
Byproduct oil
Fue1 o i 1
to vent gas
incinerator
SHIFT
CONVERSION &
GAS COOLING
GAS PURIFICATION
&
SULFUR RECOVERY
METHANOL
SYNTHESIS
(U
Ol
Byproduct
naphtha
Byproduct
sulfur
(D
Ol
METHANATI ON
Coal
fines
Excess
fines "
Tars-
Pi ant steam
and power
Stack
gas
BOILER
PLANT
FLUE GAS
SCRUBBING
t
Incinerated tail gas from
sulfur recovery and C02~rich
gas from Rectisol
Methanol
SNG
FIGURE 3. NOKOTA COMPANY (NORTH DAKOTA) PROJECT
-------
The flow diagram in Figure 4 presents the overall process design for the WyCoalGas
project. Some key points of this design are:
• The products and byproducts are SNG, sulfur, and ammonia. Excess coal fines will
be either a byproduct or a waste.
• Coal fines are burned in steam-generating boilers. ESPs followed by wet limestone
scrubbers provide flue gas particulate removal and desulfurization.
• Lurgi tars, oils, naphtha, and phenols are gasified via partial oxidation.
• The combined Lurgi and partial oxidation crude gases are purified in a non-
selective Rectisol unit.
• Sulfur recovery utilizes the Stretford process.
The flow diagram in Figure 5 presents the overall process design for the Great Plains
project. Some key points of this design are:
• The products and byproducts are SNG, sulfur, ammonia, and all of the coal fines.
• Lurgi tars, oils, naphtha, and phenols are burned as fuel in steam generating boilers.
Particulates are removed from the tar-fired superheater's flue gas by an ESP
• Lurgi naphtha and phenols are also burned as fuel in wastewater incinerator.
• The Lurgi crude gas is purified in a non-selective Rectisol unit.
• Sulfur recovery utilizes the Stretford process.
• A small amount of methanol is produced for Rectisol absorbent makeup.
As an overall commentary on the five plant designs, it is of interest to note the
following:
• Three of the five designs use partial oxidation to gasify the Lurgi liquids (tars, oils,
naphtha, and phenols) for on-site use.
• Four of the five designs utilize the Stretford sulfur recovery process.
• Two of the designs use selective Rectisol for acid gas removal, two use non-selective
Rectisol, and one uses both a selective Selexol unit and a non-selective Rectisol
unit.
409
-------
ro
ta
S i zed
coal
GAS LIQUOR
CLEANUP
Byproduct
ammonia
Tars, oils,
phenols
Wastewater
to treating
LURGI
GASIFIERS
Oxygen
Steam
Coal
Fines
Gas liquor
SHIFT
CONVERSION &
GAS COOLING
PARTIAL
OX I DAT I ON
Oxygen
-C
Q-
Carbon
dioxi de a
GAS PURIFICATION
&
SULFUR RECOVERY
METHANATION
SNG
Byproduct
sulfur
Excess
fines
-—
BOILER
PLANT
Plant
and f
^
steam
>ower
FLUE GAS
SCRUBBING
St;
9 =
Catalytica1ly incinerated
tail gas from sulfur
recovery
FIGURE 4. WYCOALGAS, INC. (WYOMING) PROJECT
-------
o
3
-------
• All three of the designs which burn coal to generate steam include flue gas desul-
furization. The one design which burns liquids to generate steam does not include
flue gas desulfurization.
• One design gasifies the coal fines and generates steam and power by burning purge
gas in gas turbines.
• There is a broad diversity of products, byproducts, and process configurations
among the five designs.
SULFUR EMISSIONS CONTROL
Figure 6 presents flow diagrams of the sulfur emissions control systems in each of the
five coal-to-synfuels designs. In examining these systems, certain process characteristics
should be kept in mind:
Selective acid gas removal processes (either Rectisol or Selexol) are those which
produce (a) carbon-dioxide-rich offgas from which most of the hydrogen sulfide has
been removed and (b) an acid gas stream (often called the hydrogen-sulfide-rich
stream) which is also carbon-dioxide-rich but contains most of the hydrogen sulfide
removed from the shifted, gasifier product crude gas.
Non-selective acid gas removal processes (either Rectisol or Selexol) are those which
produce a single acid gas stream containing all of the carbon dioxide and all of the
hydrogen sulfide removed from the shifted, gasifier product crude gas.
The carbon-dioxide-rich offgas and the acid gas streams, produced by either selective
or non-selective Rectisol or Selexol processes, contain hydrocarbon gases. Indepen-
dent of any sulfur emissions control considerations, the carbon-dioxide-rich offgas
and acid gas streams could be controlled (e.g. by incineration), to reduce the emissions
of hydrocarbons".
The designs indicate that essentially all of the hydrogen sulfide fed to the Stretford
process is converted into byproduct sulfur, but little (if any) organic sulfur is convert-
ed into byproduct sulfur. Thus, the residual tail gas from a Stretford process might
be incinerated for two reasons: (a) to control the emissions of hydrocarbons as
discussed above and (b) to convert organic sulfur to sulfur dioxide.
An Adip unit concentrates a hydrogen-sulfide-containing acid gas by removing hydro-
carbons and some carbon dioxide from the acid gas. About 94 — 98 percent of the
hydrogen sulfide in the acid gas can then be converted into byproduct sulfur in a
Glaus unit. A Scot unit converts the residual sulfur compounds in a Glaus unit tail
gas into hydrogen sulfide, which is then recovered and recycled to the Glaus unit.
The only sulfur species remaining in the Scot unit tail gas in any potentially signifi-
cant amount (200 — 500 ppmv) is hydrogen sulfide, and the tail gas is usually inciner-
ated to convert the hydrogen sulfide into sulfur dioxide.
412
-------
HAMPSHIRE
ENERGY
COMPANY
TREATED
FEED
GAS *
r,t
«
SELECTIVE
RECTISOL
>CO, TO PIPELINE
ACID GAS
TO
FUE
GAS
I
INCINERATOR
ADIP
UNIT
RECYCLE H2S
9
CLAUS
UNIT
I
*
SCOT
UNIT
SULFUR
NOKOTA COMPANY
TO'
INCI
TREATED
GAS c°i OFFGAS
t J
FEED „ SELECTIVE STHETFOHD G
"AS RECTISOL ( UNIT
1
SULFUR
FINES BOILER FLUE GAS *
TARS PLANT SCRUBBING
HERMAL
VERATOR
o,S
GAS
CaSO.,
SOLIDS
WYCOALGAS, INC.
FEED
GAS *
COAL ^
TREATED
GAS
t
SELECTIVE
RECTISOL
BOILER
TO CATALYTIC
ACID
GAS
STRETFORD
UNIT
FLUE GAS
NCINE
RATOR
^"''GAS
SLUDGE
TENNECO COAL GASIFICATION
TREATED
t
FEED .. SELECTIVE TO ATMOS
GAS ' SELEXOL
,, GAS
FEED NON- 1
GAS
FINES * PLANT
TO THERMAL
PHEHE *
STRETFORD
UNIT
I
SULFUR
. STACK
* SCRUBBING CaSO,
* si unfiF
GREAT PLAINS GASIFICATION ASSOCIATES
TARS, OILS,
PHENOLS AND
NAPHTHA
FIGURE 6. SULFUR EMISSIONS CONTROL IN FIVE OF THE PROPOSED
COAL GASIFICATION PROJECTS
413
-------
Depressuring the coal lockhoppers on the Lurgi gasifiers, each time they are loaded
with feedstock coal, requires the venting of gas from the lockhoppers. That gas con-
tains hydrocarbons and acid gases, and it may be desirable to recover and/or inciner-
ate the gas.
It should be noted that the Nokota and Tenneco designs incinerate Stretford tail gas in
fuel-fired incinerators, the WyCoalGas design catalytically incinerates the Stretford tail gas,
and the Great Plains design incinerates the Stretford tail gas in the boiler fireboxes. The
Scot tail gas in the Hampshire design is incinerated in a gas-fired incinerator.
Table 2 summarizes the sulfur balances for the gasification process units for the five
designs (excluding sulfur derived from any burning of coal fines). As a percentage of the
sulfur in the gasified coal, the sulfur discharges for the five designs range from 2.8 to 5.3
percent. In terms of equivalent sulfur dioxide, the discharges for the five designs range from
0.02 to 0.15 Ib per million Btu (8.6 to 65 ng/J) of gasified coal. Also note that the sulfur
allocated to the gasifier ash in three of the designs ranges from about 3 to 7 percent of the
sulfur in the gasified coal, which is within the usual range of assumption. However, one of
the designs allocates 0.1 percent of the coal sulfur to the gasifier ash, and another of the
designs allocates 13 percent of the coal sulfur to the gasifier ash. It is not known if special
circumstances or data are available to support these assumptions.
WATER USAGE AND WASTEWATER TREATMENT
As shown in Table 3, the intake and usage of raw water ranges for the five coal-to-
synfuels designs from 1.00 to 1.80 tons (1.00 to 1.80 Mg) of water per ton (Mg) of gasified
coal, and the average is 1.26 tons (1.26 Mg) per ton (Mg) of gasified coal. Using that average,
the gasification of 28,000 tons (25,400 Mg) of coal per day requires about 5,900 gpm
(1340 m3/hr) of water intake, which is equivalent to about 9,500 acre-ft (11.7 km3) of
water per year.
414
-------
TABLE 2. GASIFICATION SULFUR BALANCES AND DISCHARGES IN FIVE OF THE PROPOSED PROJECTS
(Excluding sulfur derived from any coal fired in steam generators)
Ul
HAMPSHIRE
PROJECT
COAL GASIFIED, 109 Btu/D (TJ/D)
SULFUR INPUT, T/D (Mg/D):
In gasified coal
SULFUR OUTPUTS, T/D (Mg/D):
As byproduct sulfur
In CaS04 b
In Na2S04c
In gasifierash
In product SNG
In product methanol
In liquid products and
byproducts
Sulfur discharges
SULFUR DISCHARGES AS:
Percent of sulfur input
lbofS02/(106Btuof
gasified coal) (ng/J)
242
49.5
44.0
na
na
3.5
na
na
nil
47.5
2.0
49.5
4.0
0.03
(255)
(44.9)
(39.9)
(3.2)
(43.1)
(1.8)
(44.9)
(12.9)
WYCOALGAS
PROJECT
386
73.0
66.1
nil
na
4.6
nil
na
na
70.7
2.3
73.0
3.2
0.02
(407)
(66.2)
(60.0)
(4.2)
(64.2)
(2.0)
(66.2)
(8.6)
NOKOTA
PROJECT
396
244.0
223.7
10.2
na
0.3
nil
nil
3.0
237.2
6.8
244.0
2.8
0.07
(418)
(221)
(202.9)
(9.2)
(0.27)
(2.7)
(215.1)
(6.2)
(221.3)
(30.1)
TENNECO
PROJECT
462
271.3
249.1
5.0
na
9.2
nil
na
na
263.3
8.0
271.3
2.9
0.07
(487)
(246.1)
(225.9)
(4.5)
(8.3)
(238.7)
(7.3)
(246.0)
(30.1)
GREAT PLAINS
PROJECT
412
290.5
231.4
na
4.1
38.8
nil
na
0.9
275.2
15.3
290.5
5.3
0.15
(435) a
(263.5)
(209.9)
(3.7)
(35.2)
(0.82)
(249.6)
(13.9)
(263.5)
(64.5)
a Total for ultimate full-size plant.
^ Gasification sulfur outputs allocated to CaS04 occur only for those plants burning tars, oils, naphtha, phenols, lock gas, etc. in boiler plants
equipped with flue gas desulfurizing scrubbers.
c Recovered from Stretford unit waste liquid.
na Not applicable
nil Essentially zero
-------
TABLE 3. RAW WATER USAGE AND GAS LIQUOR TREATMENT IN FIVE OF THE PROPOSED PROJECTS
HAMPSHIRE
PROJECT
RAW WATER INTAKE:
Gal./min(m3/hr) 2,500 (570)
Tons/ton (Mg/Mg) of gasified coal 1 .00 (1 .00)
GAS LIQUOR TREATMENT AND
FLOW SEQUENCE:
Phennk pytrgrtinn a •
H?S and NH-, removal b •
Bioloqical oxidation •
Evaporation c
To coolinq tower as makeup water •
Cooling tower blowdown:
Evaporation ^ f
Incineration f
Disposal of residual:
To gasifier ash quenching •
To ash handling
To evaporation pond •
WYCOALGAS NOKOTA TENNECO
PROJECT PROJECT PROJECT
3,860(880) 6,000(1,360) 6,800(1,540)
1.01 (1.01) 1.27 (1.27) 1.24 (1.24)
• •
• •
• •
• •
• • f
f
e •
e •
e
GREAT PLAINS
PROJECT
8,600(1,950)
1.80 (1.80)
•
•
•
•
•
e
e
e
a Via Phenosolvan process
b Via stripping process (such as Phosam process)
c Evaporation ahead of cooling tower, with condensate used as cooling tower makeup.
Blowdown from cooling tower returns to the evaporation unit.
d Evaporation of cooling tower blowdown, with condensate reused inplant.
e The residual disposal is not made clear in the available references.
f Cooling tower blowdown sent directly to residual disposal.
-------
The contaminated gas liquor generated by Lurgi coal gasification constitutes the major
wastewater stream in a coal gasification plant. The quantity of dephenolized, stripped gas
liquor for three of the five designs is:
tons of gas liquor
Project gpm (m3/hr) per ton of gasified coal (Mg/Mg)
Great Plains 4,700(1,070) 1.0(1.0)
Hampshire 1,700(390) 0.7 (0.7)
WyCoalGas 3,130(710) 0.8(0.8)
The gas liquor treatment sequence for the five designs is also presented in Table 3. It
is of interest to note that:
• All five designs use the Phenosolvan process for extracting the bulk of the phenols
from the gas liquor.
• All five of the designs use a stripping process to remove hydrogen sulfide and to
recover byproduct ammonia from the gas liquor. Three of the designs plan to use
the Phosam stripping process and one of the designs plans to use the Chemi-Linz/
Lurgi (CLL) stripping process.
• Four of the designs further treat the stripped liquor via biological oxidation prior to
using the treated wastewater as cooling tower makeup. One of the designs uses the
stripped liquor as cooling tower makeup without prior biological treatment.
• Three of the designs evaporate the cooling tower blowdown to recover water for
inplant reuse. One of those three designs evaporates the stripped liquor and the
cooling tower blowdown to obtain the cooling tower makeup.
• One of the designs incinerates the concentrate from evaporation of the cooling
tower blowdown.
It is also of interest to note that a Lurgi author12 recommends that the treatment sequence
be: phenol extraction, stripping, biological oxidation, activated carbon adsorption, and ion
exchange. The recommended treatment is stated to be needed prior to using the treated
water as cooling tower makeup.
417
-------
GAS VENTING FROM COAL LOCKHOPPERS
As discussed earlier herein, technologies such as incineration or recovery/reuse are
available for gases vented from the Lurgi gasifier coal lockhoppers. Table 4 summarizes
how that venting is handled in three of the designs. The venting of gasifiers during shut-
down and start-up is also summarized in Table 4.
418
-------
TABLE 4. HANDLING OF COAL LOCK VENTING AND OF GASIFIER START-UP/SHUTDOWN
VENTING IN THREE OF THE PROPOSED PROJECTS
COAL LOCK VENTING
GREAT PLAINS
PROJECT
WYCOALGAS
PROJECT
NOKOTA
PROJECT
High pressure gas
Low pressure gas
VO
Vented through venturi
scrubber to Stretford
unit for desulfurizing
and then routed to the
boilers for incineration.
Vented through venturi
scrubber and recorrv
pressed to join the high
pressure vent gas routed
to the Stretford unit and
then to the boilers.
Vented to a gas holder and
recompressed back into
the process gas.
Displaced by slipstream
of cooled process gas and
vented to gas holder for
recompression back into
the process gas.
The available reference
documents are unclear
on this point.
Displaced by slipstream
of carbon dioxide offgas
from the Rectisol unit
and vented to the
boiler fireboxes for
incineration.
Exhaust gas
GASIFIER START-UP AND
SHUTDOWN VENTING:
Crude process gas
Evacuated by ejector,
using motive air, and
vented to atmosphere.
Vented to start-up
incinerator.
Evacuated by fans and
incinerated in the boiler
fireboxes.
Vented to plant flare
for incineration.
Evacuated by ejector,
using motive air, and
vented to the atmo-
sphere.
It has been assumed that
the plant's vent gas incin-
erator would also handle
this service
-------
REFERENCES:
(1) "Coal To Gasoline Plant," brochure describing Hampshire Energy Company's proposed
project at Gillette, Wyoming, provided by Hampshire Energy Company by transmittal
dated August 1981.
(2) Meeting and discussion with Hampshire Energy Company personnel in August 1981.
(3) Flow diagrams and fact sheet for Tenneco's proposed project in Montana, provided by
Tenneco by transmittal dated August 1981.
(4) "Prevention Of Significant Air Quality Deterioration Permit Application," February
1980, submitted by Nokota Company to the North Dakota Department of Health.
Provided by Dames and Moore by transmittal dated August 1981.
(5) "Air Quality Permit Application For A Proposed Coal Gasification Plant," June 1981,
submitted by WyCoalGas, Inc. to the Wyoming Department of Environmental Quality.
(6) Flow diagram and process description for WyCoalGas's proposed project in Wyoming,
provided by the Panhandle Eastern Pipe Line Company by transmittal dated August
1981.
(7) "Plant Sulfur Disposition, " WyCoalGas flow diagram provided to the EPA's NAPCTAC
committee in 1977.
(8) "Request For Amendments To The Permit To Construct For The ANG Coal Gasification
Plant (Great Plains Gasification Associates), Mercer County, North Dakota," submitted
by ANG Coal Gasification Company to the North Dakota State Department Of Health,
February 1979.
(9) "Final Environmental Impact Statement, Great Plains Gasification Project, Mercer
County, North Dakota," U.S. Department of Energy, August 1980.
(10) Beychok, M.R., "Coal Gasification And The Phenosolvan Process," Division of Fuel
Chemistry, 168th National ACS Meeting, Atlantic City, September 1974.
(11) Beychok, M.R., "Sulfur Emission Controls For A Coal Gasification Plant," EPA-600/
2-76-149 (NTIS PB257-182), Symposium Proceedings: Environmental Aspects Of Fuel
Conversion Technology, II, (December 1975, Hollywood, Florida), June 1976.
(12) Rolke, D., "Treatment Of Gas Liquor From Coal Gasification Plants," Lurgi Information
No. 7, Vol. 6, July 1981.
420
-------
Session VI:
Chairman:
PRODUCT-RELATED CONSIDERATIONS
Robert P. Hangebrauck
U.S. Environmental Protection Agency
Research Triangle Park, NC
Cochairperson: Minn Triet-Lethi
U.S. Environmental Protection Agency
Washington, DC
421
-------
RISK ASSESSMENT OF SYNFUEL TECHNOLOGYt
A. Alan' Moghissi
U.S. EPA
Washington, DC 20460
(No paper or abstract available.)
422
-------
PREMANUFACTURE REVIEW OF SYNFUELS UNDER TSCAt
Matthew Hale, Jr. and Carl Mazza
Office of Toxic Substances
U.S. Environmental Protection Agency
Washington, D.C. 20460
ABSTRACT
The Toxic Substances Control Act (TSCA) requires manu-
facturers to notify EPA at least 90 days before they produce a
new chemical substance for commercial purposes. Once notified,
EPA has 90 days, extendable for good cause to 180 days, to review
the chemical. During the review period, the Agency can act to
prohibit or limit the manufacture, processing, or use of a new
chemical substance where it finds that the information available
on the substance is insufficient for a reasoned evaluation of its
risks and that (1) the chemical may present an unreasonable risk
to human health or the environment or (2) significant human or
environmental exposure can. reasonably be expected. Certain
synthetic fuel products (including certain byproducts and
intermediates) may be new chemical substances under TSCA and
therefore potentially subject to premanufacture notice require-
ments. This paper outlines TSCA premanufacture notification
requirements; it describes how "new" chemical substances are
defined; and it discusses the types of data that might be
provided to EPA with a premanufacture notice on a synfuel.
INTRODUCTION
The Toxic Substances Control Act (TSCA) of 1976 was the
first Federal statute addressing commercial chemicals through all
phases of their life cycle — manufacture, processing,
distribution in commerce, use, and disposal -- rather than
specific uses of chemicals or particular media in which they
might be found. A key feature of the act, which Congress passed
in response to highly publicized incidents involving chemicals
like PCB's, vinyl chloride, and BCME, was its focus on
prevention. By giving EPA authority to require testing on
suspected chemicals and by requiring it to review new chemicals
before mariuf ac ture, Congress hoped to make it possible for the
Agency to act against unreasonable risks before actual harm to
423
-------
human health and the environment occurred, rather than to address
hazardous situations only after the damage had been done.
TSCA's premanufacture notice (PMN) requirements for new
chemicals epitomize this preventive approach. Under §5,
companies must notify EPA 90 days before they produce a new
chemical, giving EPA the opportunity to review the chemical
before exposure occurs. Synthetic fuels developers, because they
will be manufacturing new fuels and related products, may in some
cases be subject to these requirements. We recognize that this
possibility has raised considerable concern in the synfuels
indus try.
Because of this concern, the EPA Office of Toxic Substances,
which is responsible for administering TSCA, is committed to
working with industry to clarify TSCA requirements and to ensure
that premanufacture notice requirements do not unnecessarily
delay the development of synfuels. Toward this end, we have met
with several trade organizations and private companies to address
both general and specific concerns, and we are cooperating with
other offices in EPA and other government agencies to avoid
duplication and to ensure a consistent approach. In carrying out
our responsibilities under TSCA, we will be careful to avoid
constructing artificial barriers to development -- that is, those
that do not contribute to results with substantial health or
environmental benefits.
i
In the remainder of this paper, we discuss in more detail
TSCA's premanufacture notice requirements, the applicability of
these requirements to synthetic fuels (in particular, coal-based
fuels), and the types of data that manufacturers might develop in
preparing a PMN.
SECTION 5 PREMANUFACTURE NOTICE REQUIREMENTS
Section 5(a) of TSCA requires companies to notify EPA at
least 90 days before beginning to manufacture or import a "new
chemical substance" for commercial purposes. As explained later,
new chemical substances are defined under the Act as substances
not listed on EPA's Chemical Substance Inventory, a compilation
of chemicals in commercial production first published in 1979.
Once notified, EPA has 90 days, extendable for good cause to 180
days, to review the potential risks likely to be posed by the new
sub s tance .
During the review period, EPA can act under §5(e) to
prohibit or limit the manufacturing, processing, distribution,
use, or disposal of the substance, pending the development of
data, if it finds that the information available on the substance
424
-------
is insufficient for a reasoned evaluation of its health and
environmental effects and that (1) the chemical may present an
unreasonable risk, to human health or the environment or (2)
significant human or environmental exposure can reasonably be
expected. If the Agency finds that the substance will present an
unreasonable risk., it can regulate it under §5(f). When EPA does
not take action under §5(e) or §5(f) during the review period,
manufacture or import can begin without restrictions. After
commercial manufacture begins, the substance is added to the
inventory. At that point, the substance is no longer "new," and
other manufacturers are free to produce it without submitting a
PMN.
Section 5(d) of TSCA specifies the information that must be
included in a PMN. In general, manufacturers must provide known
or "reasonably ascertainable" information on chemical identity,
anticipated production volume, categories of use, byproducts,
workplace exposure, and manner or methods of disposal.* They are
also required to provide test data that they have already
developed and to describe any other information on health and
safety they know or can "reasonably ascertain." However, TSCA —
unlike laws regulating the introduction of pesticides or drugs
into commerce — imposes no mandatory testing requirements for
new chemicals.
The key to EPA's review of new chemicals under TSCA is the
concept of "unreasonable risk." The Agency has not developed any
general criteria for determining "unreasonable risk," because the
finding depends too much on the specific situation. The Agency's
approach to determining unreasonable risk, however, is
illustrated in Figure 1. Potential toxicity (including
ecotoxicity) and exposure define the risks a substance presents
under specific circumstances of manufacture, processing,
distribution, use, or disposal. To determine whether these risks
are "reasonable," the Agency balances them against the benefits
to be derived from the product, the cost of measures necessary to
reduce risks, the availability of substitutes, and the
comparative risks posed by products they may replace in the
market.
Some of the information submitted in a premanufacture notice
may be confidential, including highly sensitive business
information. The Office of Toxic Substances routinely handles
such information under TSCA, and it has established elaborate
procedures (including serious penalties) to prevent its
disclo sure.
425
-------
NJ
Toxic
Hazard
Exposure _
Unreasonable
Risk?
Figure 1. Determination of Unreasonable Risk Under TSCA
-------
Several points about the review of new chemicals under TSCA
§5 that are often misunderstood, particularly by companies more
familiar with EPA. permitting offices, should by now be clear.
First, §5 imposes a notification requirement; it does not set up
a licensing or registration program. To limit or halt
production, EPA must take positive action against a chemical,
based on certain specific findings. Otherwise, the chemical can
enter commerce unregulated. Consequently, the Agency has in no
sense "approved" a chemical that it has not regulated under §5.
Second, TSCA does not impose testing requirements on manu-
facturers of new chemicals. Instead, the manufacturer of a new
chemical has the responsibility to determine what level of
testing, if any, is appropriate for a chemical, given its
composition and projected uses. Finally, the "unreasonable risk"
standard of TSCA incorporates the principle that the risks of a
chemical can only be evaluated meaningfully within the context of
the benefits derived from it and the costs of regulation. EPA's
goal under TSCA is to balance these considerations rather than to
reduce risk to some absolute "acceptable" level or to impose some
other standard, such as best available technology.
To date, EPA has reviewed more than 800 new chemicals under
the premanufacture review program. All these have been general
industrial chemicals, such as intermediates, dyes, photographic
chemicals, and lubricant additives. None has been a synthetic
fuel. Therefore, it is difficult to make observations on
synfuels and PMN requirements based on the history of the PMN
program to date. Several special features of synfuels will
distinguish them from new chemicals previously reviewed in the
PMN program and raise particular issues for the PMN review
process and for companies submitting notices. These features
include:
o The national interest in alternate fuels development and
energy independence
o The tremendous investments in synfuels development before
commerc ialization
o The staged development of synfuel projects, which may
include process and product changes in the course of
commercialization, and therefore may complicate the task
of characterizing the product and its toxicity
o The large production volume projected for synfuels, the
potential for exposure to some commercial fuel products,
and the presence of potentially toxic substances in some
synfuel products
427
-------
o The difficulties involved in evaluating health and
environmental effects of complex, multicomponent
substances like synfuels
o Public concern about potential hazards from synfuels
Because of the importance of synfuels projects and the money
committed to them, developers subject to PMN requirements are
encouraged to consult EPA well before PMN submission to ensure
that they are developing information sufficient for a reasoned
evaluation of risfc. In this way, EPA and industry can ensure
that the PMN process will not unecessarily delay the
commercialization of a product.
PMN REQUIREMENTS AND SYNFUELS
Some synfuels developers -- and companies refining new
synfuels -- may be intending to make products that would be "new
chemical substances" subject to TSCA premanufacture notice
requirements. OTS is now reviewing the applicability of §5
requirements to synfuels (for example, how the research and
development exemption of §5(h)(3) should apply to projects of
this scale), and it is developing a consistent approach to
defining and characterizing synfuel products, so that industry
can readily determine whether a specific product is new.
Although this wor< is not yet completed, it is possible at this
point to provide developers some preliminary guidance on the
Office of Toxic Substances' current thinking on premanufacture
notice requirements for synfuels. For more specific guidance, we
recommend that individual developers consult the Office.
WHEN IS A PMN REQUIRED?
The PMN Requirement Is Substance-Specific
"Chemical substances" have a special definition under
TSCA — the term covers both discrete chemical compounds (e.g.,
benzene or sodium chloride) and complex substances produced by
chemical reaction (e.g., coal tar or slag), including refined
products (e.g., petroleum distillates). However, the term
excludes "mixtures" that could be produced for commercial
purposes by combining substances without a chemical reaction.
Complex materials such as typical coal liquids are not considered
"mixtures" under TSCA, but rather are chemical substances,
because they could not practicably be made by mixing their
cons ti tuent s.
TSCA premanufacture notice requirements apply to such
"substances" if they are new. In this respect, these
428
-------
requirements differ from permitting requirements, which apply to
facilities rather than chemicals. A new facility producing
"existing" substances would not be subject to PMN requirements.
On the other hand, a single facility is likely to produce several
products, any or all of which might be "new" and therefore
subject to PMN.
"New" Chemical Substances Are Chemicals Not Listed on the TSCA
Chemical Substance Inventory
Under §8(b) of TSCA, EPA has compiled and keeps current an
inventory of chemical substances in commerce, first published in
June 1979. Chemicals listed on the inventory are considered
"existing" substances, not subject to PMN notice requirements.
When chemicals complete PMN review and enter commercial
production, they are added to the inventory. Therefore, manu-
facturers may determine whether their substances are new by
consulting this list or, where questions of product identifi-
cation are difficult, by asking the Office of Toxic Substances
whether the substances are listed.
The Substance Must Be Manufactured or Imported "For Commercial
Pur po ses"
TSCA §5 requirements specifically apply to chemical
substances manufactured "for commercial purposes." This includes
intermediates and other chemicals consumed entirely on the site
at which they are manufactured. As a result, intermediate
streams used in making new synthetic fuels may be subject to PMN
requirements, even if they are never sold or distributed in
commerce.
Research and Development Chemicals Are Exempt From PMN
Requirements
Chemicals manufactured "only in small quantities" solely for
research and development are specifically exempted from PMN
requirements by §5(h)(3) of the Act. Activities falling within
the category of R&D include the evaluation of the physical,
chemical, production, and performance characteristics of a
substance. Thus, pilot plant operations designed to assess
manufacturing or refining processes, test burns to evaluate fuel
efficiency or emissions, and other product characterizations are
possible without a PMN. These evaluations may be carried out by
people other than the manufacturer, including potential
industrial customers. Furthermore, the sale of a product to a
potential customer who will use it only for R&D does not remove
the product from the category of R&D. EPA has not placed a
specific volume limit on the R&D exemption, but rather has stated
429
-------
that "only in small quantities" means only in quantities no
greater than reasonably necessary for R&D (see 40 CFR 710.2(y)).
For synfuels, because of the nature of R&D activities, "small
quantities" may be large compared to production volumes for
typical industrial chemicals.
Nonisolated Intermediates Are Exempt
Chemical intermediates not intentionally removed from the
equipment in which they are manufactured are exempt from PMN
requirements. (See 40 CFR 710.4(d)(8).) As a result, non-
isolated process streams in a synfuels plant are not subject to
these requirements. However, if the intermediate stream is
removed from the plant equipment -- including for storage — it
may be subject.
Some Commercial Byproducts Are Exempt
The inventory reporting rules exempt from PMN requirements
byproducts that have commercial value only to organizations who
(1) burn them as fuel, (2) dispose of them as waste, including in
a landfill or for enriching soil, and (3) extract component
chemical substances from them. (See CFR 710.4(d)(2) . ) Under
this provision, certain byproduct streams burned for process heat
as an alternative to disposal may be exempt from PMN require
ments -- for example, phenols produced as a byproduct of coal
gasification would not be subject to PMN if incinerated.
HOW CHEMICAL SUBSTANCES ARE DEFINED
When EPA compiled the initial TSCA inventory, it faced a
number of complex issues related to chemical identification and
nomenclature. The resolution of these issues, reflected in the
way products were reported for the inventory and how they are
listed, now defines the Agency's approach to defining products
for PMN purposes.
For single-component substances that can be characterized by
a molecular formula -- like ammonia, benzene, and methanol -- the
problem of identification was simple. These products are listed
on the inventory under their chemical names; manufacturers of the
substances, therefore, are not subject to PMN requirements,
regardless of how the substances are made and what levels of
impurities they contain.
Complex reaction products -- for example, materials produced
in coking coal or refining petroleum — presented a very
different problem. Here, the products could not be defined by a
single structure or an unambiguous chemical name. Instead, these
430
-------
products were defined by source material and process of
manufacture rather than by compositional data alone. An attempt
was made to define product categories broadly enough so that
limited variations in source (e.g., substitution of one grade of
coal or one petroleum crude for another) or slight changes in
process did not create a new product, but at the same time
narrowly enough so that substances within a product category
could be expected to be similar in composition and biological
activity.
This approach can best be illustrated by the listing of
refined petroleum products on the inventory. For example, the
inventory entry "light hydrocracked distillate (petroleum)" is
defined as "a complex combination of hydrocarbons from the
distillation of the products of a hydrocracking process. It
consists primarily of saturated hydrocarbons having carbon
numbers predominantly in the range of C,Q through Cig, and
boiling in the range of 160"C to 320"C. This description, it
can be seen, identifies the source material (petroleum), the
process of manufacture (hydrocracking and distillation), and
composition (CiQ-Cig saturated hydrocarbons) as well as a
physical property (boiling range) that roughly correlates with
chemical composition. Any hydrocarbon product that met these
criteria would be considered the same product for inventory
purposes and therefore would not be subject to PMN. Comparable
products from a different source material or manufactured by a
different process, however, would be different chemical
substances under the inventory rules. (Other examples of
petroleum products are given in Table 1.)
This discussion should make it clear that, for TSCA
inventory purposes, coal-derived synthetic fuels are, a priori,
different chemical substances from petroleum-based fuels. An
inventory entry for petroleum naphtha, for example, would not
cover a naphtha derived from coal, even if the general
composition and the boiling range of the products were similar,
because their source materials are different. In the same way, a
naphtha derived from coal gasification is not comparable to a
naphtha derived from the solvent-refining of coal, because of the
clear difference in process. Therefore, a PMN might be required
for a naphtha produced in a high-Btu coal gasification operation
even though petroleum naphtha and coal naphtha produced by
pyrolysis were listed on the inventory. (However, we recognize
that at some point in refining coal, oil shale, and petroleum
products becomes so similar that source should no longer be a
factor iu product definition.)
The Office of Toxic Substances is now working to develop
product definitions for synfuels comparable to definitions
431
-------
TABLE 1. SAMPLE PETROLEUM PRODUCT DEFINITIONS ON THE
TSCA CHEMICAL SUBSTANCE INVENTORY (FROM TOXIC SUBSTANCES CONTROL ACT)
CRUDE OIL DISTILLATION STREAMS
Light Straight Run Naphtha (Petroleum) [*64741-46-4]
A complex combination of hydrocarbons produced by distillation of
crude oil. It consists predominantly of aliphatic hydrocarbons having
carbon numbers predominantly in the range of C4 through C10 and boiling in
the range of approximately minus 20°C to 180°C (-4°F to 356°F).
Heavy Straight Run Naphtha (Petroleum) ["64741-41-9]
A complex combination of hydrocarbons produced by distillation of
crude oil. It consists of hydrocarbons having carbon numbers predomi-
nantly in the range of C6 through C12 and boiling in the range of ap-
proximately 65°C to 230°C (149°F to 446°F).
Straight Run Kerosine (Petroleum) ["8008-20-8]
A complex combination of hydrocarbons produced by the distillation of
crude oil. It consists of hydrocarbons having carbon numbers predomi-
nantly in the range of C9 through C16 and boiling in the range of ap-
proximately 150°C to 290°C (320°F to 554°F).
Straight Run Middle Distillate (Petroleum) ["64741-44-2]
A complex combination of hydrocarbons produced by the distillation of
crude oil. It consists of hydrocarbons having carbon numbers predomi-
nantly in the range of C1:L through C2o and boiling in the range of 205°C
to 345°C (401°F to 653°F).
Straight Run Gas Oil (Petroleum) ["64741-43-1]
A complex combination of hydrocarbons produced by the distillation of
crude oil. It consists of hydrocarbons having carbon numbers predomi-
nantly in the range of C1X through C2s and boiling in the range of ap-
proximately 205°C to 400°C (401°F to 752°F).
OTHER PRODUCTS
Light Hydrocracked Distillate (Petroleum) [-64741-77-1]
A complex combination of hydrocarbons from the distillation of the
products from a hydrocracking process. It consists predominantly of
saturated hydrocarbons having carbon numbers predominantly in the range of
C10 through C18, and boiling in the range of approximately 160°C to 320°C
(320°F to 608°F).
432
-------
Hydrotreated Light Distillate (Petroleum) ["64742-47-8]
A complex combination of hydrocarbons obtained by treating a petro-
leum fraction with hydrogen in the presence of a catalyst. It consists of
hydrocarbons having carbon numbers predominantly in the range of CQ
through C16 and boiling in the range of approximately 150°C to 290°C
(302°F to 554°F).
Light Catalytic Cracked Distillate (Petroleum) [*64741-59-9]
A complex combination of hydrocarbons produced by the distillation of
products from a catalytic cracking process. It consists of hydrocarbons
having carbon numbers predominantly in the range of C9 through C2s and
boiling in the range of approximately 150°C to 400°C (302°F to 752°F). It
contains a relatively large proportion of bicyclic aromatic hydrocarbons.
Crude Phenolic Compounds (Petroleum) ["64743-03-9]
A complex combination of organic compounds, predominantly phenol,
cresols, xylenols and other alkylated phenols obtained primarily from
cracked naphtha or distillate streams by alkaline extraction.
Vacuum Residuum (Petroleum) [^64741-56-6]
A complex residuum from the vacuum distillation of the residuum from
atmospheric distillation of crude oil. It consists of hydrocarbons having
carbon numbers predominantly greater than C34 and boiling above approxi-
mately 495°C (923°F).
433
-------
already developed for petroleum products. The goal is to divide
the spectrum of potential synfuel products into generic chemical
substance categories that will be unambiguous both to industry
and EPA and that will reflect likely compositional differences.
This scheme would define how new synfuels would be listed on the
inventory. As a result, it would make it possible to determine
which products should be considered equivalent to existing
products for PMN purposes and might be subject to PMN
requirements. Specifically, OTS is addressing questions like:
Are products from different coal liquefaction processes (e.g.,
SRC-I1, EDS, and H-coal) likely to be sufficiently similar in
composition and biological activity to justify their treatment as
the same substance for inventory and PMN purposes? (In other
words, if an SRC-II liquid is entered on the inventory, would a
comparable BUS or H-coal product automatically become an existing
substance not subject to PMN?) At what point in the refining
process should synfuel products be considered essentially
equivalent to comparable petroleum products and therefore not be
subject to PMN requirements? In this work, OTS has solicited
information and advice from the American Petroleum Institute, and
it is in contact with the National Council on Synthetic Fuels
Production. The Office is also willing to meet with other
organizations or individual companies who have an interest in
these questions.
Until this work is completed, it is difficult to provide
definitive answers to questions about whether one synfuel product
should be considered equivalent to another for PMN purposes, or
how many synfuels are likely to be subject to PMN requirements.
It is possible, nevertheless, to provide some guidance on
requirements for certain specific products. For example:
1. Sulfur, ammonia, and carbon dioxide produced in the
gasification or liquefaction of coal are existing
substances and thus are not subject to PMN requirements.
2. Methanol produced from coal is equivalent to methanol
listed on the inventory and thus is not subject to PMN.
However, indirect coal liquids are not on the inventory
and therefore may be subject to PMN requirements if
manufactured for commercial purposes.
3. Substitute natural gas produced by coal gasification is
predominately methane, which is listed on the inventory,
and therefore is not subject to PMN.
4. SRC I wash solvent, SRC 1 mineral residue, and SRC
naphtha, which are listed on the inventory, are not
subject to PMN requirements. Other SRC products reported
434
-------
for the inventory are under review by OTS, as described
below.
Because of the complexity of product definitions, we encourage
synfuel developers to consult OTS on the applicability of PMN
requirements to their products.
STATUS OF SYNFUELS REPORTED FOR THE INVENTORY
Under the previous administration, OTS began a review of 20
synfuel products (9 SRC products and 11 oil shale products)
reported for the inventory to determine if they should have been
included on the list. The key question was whether the products
had been manufactured for purposes other than research and
development during the period when the inventory was compiled.
OTS has determined that certain of the products -- including the
SRC products listed under item number 4 above — had in fact been
properly listed. For the other products, OTS decided that it did
not have enough information to make a determination, and it asked
for further information from the companies that had reported them
for the inventory. We anticipate that the Agency will decide the
status of these products in the near future.
PMN SUBMISSIONS
As previously discussed, TSCA requires PMN submitters to
provide certain information on chemical identity and exposure,
but it does not require manufacturers of new chemical substances
to develop health and safety data specifically for their
notices.* However, §5(e) gives EPA the authority to delay the
commercial production of a new chemical in the absence of data
necessary for a reasoned evaluation of the chemical's health and
environmental effects — if the substance "may present an
unreasonable risk" or that there will be "significant or
substantial exposure" to it.
The nature of this §5(e) authority, and EPA's interpretation
of it, has raised concern among some prospective synthetic fuel
producers. Several companies have asked OTS to identify the data
For general guidance on EPA's interpretation of premanufacture
notice requirements, see Toxic Substances Control:
Premanufacturing Notification Requirements and Review
Procedures; Statement of Interim Policy (44 FR 28564, May 15,
1979) and Toxic Substances Premanufacture Notification
Requirements and Review Procedures: Statement of Revised
Interim Policy (45 FR 74378, November 7, 1980).
435
-------
it would consider sufficient for a "reasoned evaluation" of a
particular synfuel and to comment on the appropriateness of
specific test plans. To address such questions, the Agency has
established a Synfuels/Toxics Workgroup, managed by OTS, which
can provide guidance to individual producers and will facilitate
the review of PMN's on synfuels. Synfuels developers are
encouraged to discuss questions concerning data development and
methods of controlling risks with this group before submitting a
PMN.
Because there are no testing requirements under TSCA for new
chemicals, EPA has not developed prescriptive guidelines for data
development on synfuels. In addition, it is difficult to define
a single approach for different products because, among other
reasons, the specific composition of a product and the conditions
of its production and use will influence how much and what types
of information might be appropriate. However, we believe that
the following general principles are applicable to any program
evaluating risks from synthetic fuels:
o Data should be appropriate to what is known about
chemical composition and exposure. For example, if
exposure is limited, limited data may be sufficient for a
reasoned evaluation.
o Full characterization of risks before a synfuel is
manufactured commercially may in some cases be
infeasible. Although the amount of data available may be
limited early in commercial development, concern for risk
posed by a substance would be limited by the fact that
exposure and production volume are relatively low.
However, as a substance grows in the market, more data
might in some cases be appropriate.
o New synthetic fuels should be evaluated in comparison to
the petroleum products they would replace to provide a
perspective on the risks they might present. If
replacing petroleum products by a synfuel will not lead
to an increase in risk, risks from the new synfuel should
generally be considered reasonable.
In the remainder of this presentation, we will describe in
somewhat more detail the kinds of thinking that typically goes
into a risk evaluation and that lies behind OTS1 general approach
to assessing data provided in a PMN.
436
-------
CHEMICAL CHARACTERIZATION
In many cases, the chemical composition of a product —
including the extent to which it contains minor constituents of
known or suspected toxicity, like aromatic amines, heterocyclic
nitrogen compounds, and PNA's — can serve as an important guide
in determining what data are appropriate for evaluating risics.
For example, a chemical analysis of a gasoline derived from
indirect liquefaction might show that it was less aromatic and
more aliphatic than typical petroleum gasolines, and contained a
considerably lower level of toxicologically significant
constituents. This could provide a rationale for limiting the
extent of toxicity testing. At the same time, extensive testing
of substances known to be highly hazardous may be redundant. For
example, if a coal-derived residual fuel contained significant
quantities of known or suspected carcinogens, the premanufacture
review of this substance would focus on potential exposure and
the manner of uses to establish that risks are adequately
controlied.
EXPOSURE ASSESSMENT
Conditions of exposure are also an important factor in
deciding what health and environmental-effects data would be
appropriate to evaluate risks posed by a specific substance.
Typically, exposure assessments address direct exposure to
humans, indirect exposure to humans from environmental release,
and exposure to the environment during all phases of a
substance's life cycle — manufacture, handling, distribution,
storage, and end use. Anticipated production volume for
different uses, potential targets of exposure, and magnitude of
exposure are also factors that often guide data development. We
recognize that there is no simple formula for translating such
considerations into a testing strategy. However, in reviewing
PMN's on new chemicals, the Agency evaluates the data presented
in the light of exposure-related considerations.
It is possible to illustrate in general terms how different
exposure scenarios might influence data development. The
following uses, for example, would on the whole reflect
increasing levels of direct human exposure: industrial boiler
fuel, diesel transport fuel, and consumer gasoline. EPA's review
of health and environmental effects data on products within each
of these categories would consider the different levels and modes
of exposure — where exposure is likely to be higher, data should
provide greater certainty that a substance does not present an
unreasonable risk. As a second example, tentative or preliminary
data might be reasonable for products made in early-term plants,
if the products would be used in a restricted or controlled
manner.
437
-------
EXISTING DATA
The Department of Energy, EPA, private companies, and other
organizations have developed a considerable amount of information
relevant to risks that may be posed by new synfuels. This
includes data on the toxicity of comparable petroleum and synfuel
products, exposure information on different fuel uses, infor-
mation on the use of specific toxicity tests for complex mixtures
like synfuels, toxicity data on chemicals likely to be found in a
synfuel, and similar information. PMN submitters should consider
the implications of this information in determining what and how
much data they should develop.
rt
TOXICITY AND ENVIRONMENTAL EFFECTS TESTING
As we stated before, TSCA does not require the testing of
new chemicals. In addition, because the review of risks posed by
a new synfuel will depend on the specific product and its
projected uses, it is impossible to develop prescriptive
guidelines for health and environmental effects data. Instead,
synfuels developers are encouraged to discuss their products and
testing plans with the Office of Toxic Substances before PMN
s ubmi s s io n .
OTS recognizes that the scale and scheduling of many synfuel
projects are likely to make it difficult for developers to
provide final health and environmental-effects data sufficient
for evaluating risks associated with a full-scale commercial
operation at the time they submit a premanufacture notice. For
example, if a manufacturer is conducting long-term tests, results
might not be available at the time of notice submission. In
addition, products are likely to change in scaleup or as a result
of process changes; in some cases, pilot-plant material available
for toxicity testing may not be comparable to products later made
in a commercial plant. Thus, if tests are being conducted on
early-stage products, the relevance of the results of these tests
to an evaluation of the potential effects of final commercial
products should be considered. In such circumstances, technical
judgment can. be used to evaluate whether the final product is
likely to present more or less of a problem than the tested
mater ial.
EPA understands that it is often common for the development
of data to proceed as technology develops and commercial samples
become available. It is possible, of course, that a commercial
substance might later prove to be more hazardous than initially
believed, or that subsequent data might show that the substance
would present an unreasonable risk under certain circumstances.
Companies developing a new product would typically have contin-
438
-------
gency plans for controlling exposure in this situation. These
plans, for example, might call for restricting uses; imposing
engineering controls; upgrading the product; changing the process
or product slate; or similar measures. In reviewing PMN's on new
synfuels, EPA. will consider all these factors in assessing the
reasonability of risk.
By early consultation with EPA about PMN-related issues,
synfuels developers can ensure that PMN requirements do not
unnecessarily delay the commercialization of their products, and
that any differences on appropriate data development are resolved
before formal PMN submission.
CONCLUSION
The evaluation of risks posed by synthetic fuels raises a
number of complex issues. We cannot expect to achieve perfect
certainty in this area, nor can we hope to eliminate all risk.
Instead, EPA1s standard under TSCA is "unreasonable" risk, which
takes into account potential benefits, availability -of.
substitutes, and risks posed by comparable products in society.
Under §5, EPA has the responsibility to review new chemicals
according to this standard before they enter commercial
production. However, we recognize the unique issues raised by
the preiaanufacture review of synfuels. Where PMN requirements
apply to new synfuels, the Agency will work with developers to
ensure that these requirements do not unduly impede technological
innovation, while protecting health and the environment from
unreasonable risk.
439
-------
METHANOL AS A CLEAN MAJOR FUEL
by: Paul W. Spaite
Cincinnati, OH 45213
ABSTRACT
The objective of this investigation of methanol as a major
fuel was to provide perspective for officials of the U.S. Envi-
ronmental Protection Agency's Industrial Environmental Research
Laboratory at Research Triangle Park regarding possibilities for
commercialization and the environmental implications associated
with wide use of methanol as a substitute for petroleum-derived
fuels.
It is recognized that the future of methanol fuel will ulti-
mately be determined by economics. To gain widespread acceptance,
methanol will have to be cheaper than competitive fuels after all
advantages and disadvantages have.been considered. No attempt is
made here, however, to assess the competitiveness of methanol
fuels at present prices for crude oil or to project the price at
which they could be competitive. Such evaluations would be far
beyond the scope of the study. Instead, the methanol fuels are
considered relative to other fuels that might be used if an
effort is launched to apply available technology to displacement
of petroleum fuels as soon as possible. The major factors con-
sidered are:
1) Potential environmental consequences of introducing
methanol.
2) Status of development of methanol fuel technology.
3) Cost and efficiency of synfuel processes.
440
-------
4) Potential markets.
5) Prospects for commercialization of methanol fuels.
The intent is to develop an overview perspective by identi-
fying all important factors in each category and presenting
enough quantitative data to permit relative comparisons, without
excessive detail.
BACKGROUND
At present there is concern over the rate of progress in
development of advanced coal conversion processes for a synthetic
fuels industry. One of the principal impediments is the infla-
tion associated with a cost-spiral driven by continuing increases
in the cost of oil and other fuels, including coal.
Because of the inflationary trend, many believe that plants
that could be built now to use available technology will be
cheaper to operate than plants built later to use improved
processes that might come onstream in a few years. Also there is
a continuing concern over America's continuing dependence on
foreign oil. These factors have combined to create interest in
utilizing immediately applicable coal conversion technology.
The only proven coal conversion technology is indirect liq-
uefaction; that is, the conversion of coal to synthesis gas and
subsequent conversion of this gas to liquid fuel. The proven
routes for coal conversion include (1) the Fischer-Tropsch
process, which converts synthesis gas directly to gasoline and
other byproducts, and (2) a number of catalytic processes, which
convert synthesis gas to methanol. Although the Fischer-Tropsch
process has the advantage of producing gasoline directly, it has
the disadvantage of producing many coproducts and byproducts,
which must be marketed. Methanol may be used directly, as a
premium fuel, in some applications, but may have to undergo
subsequent conversion to gasoline, at some added cost, for use as
a transportation fuel.
441
-------
If a decision is made to build synthetic fuel plants with
presently available technology, the Fischer-Tropsch process and
methanol fuel processes will likely be used. The Fischer-Tropsch
products are essentially the same as petroleum-derived fuels, so
that their introduction into commerce would not require signifi-
cant adjustment. In contrast, the introduction of methanol as a
major fuel would require significant adjustment.
442
-------
METHANOL AS FUEL: ENVIRONMENTAL IMPLICATIONS
Although some testing has been carried out to evaluate the
use of methanol as a major fuel for automobiles and stationary
sources, work to evaluate the potential environmental effects has
not been extensive. Whereas some properties of methanol make it
attractive as a fuel, others present problems. Experimental work
to date has been encouraging, but many questions remain unanswered,
Following are some of the more important environmental considera-
tions.
1) Methanol has a lower flame temperature than petroleum-
derived products. It also has wide limits of combustibility.
These properties combine to make either automobiles or
stationary sources that are designed for methanol fuels
relatively lower emitters of nitrogen oxides.
2) Methanol combustion is essentially particulate-free. No
carbon-to-carbon bonds are present to promote soot formation,
which is associated with burning of petroleum-derived fuels.
3) Because sulfur in the feedstocks for methanol is removed
in processing, combustion of methanol generates virtually no
sulfur emissions.
4) Because of its high octane rating, methanol can be used
in motor vehicles without additives, eliminating the emis-
sions associated with additives to petroleum-derived fuels.
5) Methanol's low heat content (about half that of gasoline
on a volumetric basis) necessitates the use of twice the
volume and over twice the weight of fuel when it is substi-
tuted for gasoline or distillate oil.
6) Some methanol properties such as corrosivity, toxicity,
and explosivity call for careful consideration. Although
they have not caused problems in the closely controlled
situations where methanol has been used as a commercial
chemical, they must be given careful attention if it is
widely used as a major fuel.
443
-------
7) Other environmental considerations that have not been
evaluated are the reactivity, persistence, and sensory
detectability of methanol in the environment. These factors
could be of great importance for a chemical with potential
for release in large amounts to the environment, as illus-
trated by the experiences with oil spills. The high solu-
bility of methanol in water suggests that spills of methanol
would not persist as oil spills do. On the other hand, the
contamination of lakes or major rivers with a toxic material
that disburses into water could cause fish kills and also
could produce water contamination that would not be readily
detected without special precautions.
The most extensive body of experimental work on methanol as
a fuel has dealt with its use as a gasoline substitute. Most at-
tention has been given to methanol-gasoline mixtures, but consid-
eration has also been given to the use of 100 percent methanol
fuel for automobiles. Although it has been established that
methanol could be substituted for gasoline, there is considerable
controversy over advantages and disadvantages of doing so. Some
researchers expect that methanol will give higher efficiency,
improved performance, and reduced pollution. Others claim the
2 3
opposite on all or some of these points. ' It is generally
accepted, however, that the use of methanol in engines designed
to take advantage of its high octane and unusual combustion
characteristics would give performance as good as, or superior to
that of gasoline on an equivalent Btu basis.
Experimental work with methanol as a fuel for use by sta-
tionary sources has been encouraging. Tests in which methanol
fuel was fired in a utility boiler designed to burn natural gas
4
or distillate oil showed methanol to be a superior fuel.
Concentrations of pollutants in the combustion gases were very
low (no particulates, no sulfur oxides, and low nitrogen oxides).
Also, the methanol fuel burned efficiently with a stable flame,
and carbon previously deposited by oil burning was burned off of
heat transfer surfaces with a resultant improvement in heat
transfer. Tests of methanol fuels in commercial combustion
turbines were also promising. Performance was excellent, and
nitrogen oxide emissions were lower than those produced by firing
444
-------
natural gas. Studies of methanol as a turbine fuel for combined-
cycle plants were also promising, and it has been suggested that
such plants could be designed to be virtually pollution free.
Consideration of methanol as a fuel for nonutility station-
ary sources led to the conclusion that it could replace distillate
oil in home heating and would give increased efficiency. This
study also concluded that methanol fuels could replace gas or
distillate oil in commercial and industrial applications if due
consideration is given to potential problems associated with its
toxicity and flammability.
In summary, past work indicates that methanol has potential
for wide use as a high-quality environmentally attractive fuel.
The studies also show clearly, however, that its use as a fuel
will require special measures for environmental protection.
445
-------
STATUS OF DEVELOPMENT FOR METHANOL FUEL PRODUCTION PROCESSES
All of the technology necessary to produce methanol for fuel
use is proven. At present chemical-grade methanol is produced in
amounts estimated at 30,000 ton/day. Most is produced from
synthesis gas made from natural gas. The largest plant in opera-
tion today is a 2500-ton/day single-train plant, which has been
operational for 10 years. Plants twice this large are now con-
sidered feasible. It is claimed that because of reduced quality
requirements and improvements in technology.- a 5000-ton/day plant
for production of fuel-grade methanol would be only slightly
larger than the operating plant producing 2500 ton/day. It is
further suggested that methanol fuel plants should consist of 5
trains of 5000 ton/day each in capacity.
Technology for production of synthesis gas from coal is also
being applied widely outside of the United States. Lurgi and
Koppers-Totzek coal gasifiers are the most discussed for use in
commercial production of liquid fuel from coal. Both types have
a long history of application in service of the general type
required for production of methanol fuels, and both have been
incorporated in planned installations.
The development of the Mobil-M process, which is said to
convert methanol to gasoline with an efficiency of 95 percent,
may be the key to avoidance of distribution and handling problems
that might otherwise impede the application of methanol fuel
Q
technology. The process was announced in 1976. Since then a 4-
bbl/day pilot plant has been operated. Economic comparisons with
commercially established Fischer-Tropsch units are claimed to
show that the Mobil process is the most promising route from coal
9
to gasoline. Construction of a plant to convert methane-derived
446
-------
methanol to 12,500 bbl/day of gasoline is expected to begin in
late 1981 in New Zealand. The plant, to be completed in the mid-
1980 's, will supply an estimated 1/3 of that country's transpor-
tation fuel.
Although all major components for production of methanol
fuel from coal are proven technology, no plant has yet been
built. Construction of such a plant would involve making the
connection between coal gasifiers producing synthesis gas and
methanol plants for the first time. Also, economy of scale would
require the design of methanol trains larger than any yet built.
And coal would be gasified on a scale unprecedented except in
South Africa, where the "Sasol I" plant employing Fischer-Tropsch
technology has operated since 1955. This plant employs thirteen
gasifiers, each 12 feet in diameter. Proposed plants will be
even larger. Sasol II, which came on stream recently, employs
36 gasifiers. The problem associated with adaptation of
processes and large scale operation should not present serious
technical problems, but any element of risk has potential for
making investors cautious about investing in multi-billion dollar
plants.
447
-------
COST AND EFFICIENCY OF METHANOL FUEL PROCESSES
The attractiveness of methanol fuels over fuels from alter-
native processes will depend primarily on cost. The thermal
efficiency of the conversion process will be an important factor
in the final production cost. Comparisons of both cost and
efficiency of alternative production routes are complicated by
the dependence of both on the quality of feed materials and the
markets for potential products and coproducts. This is illustrat-
ed in Table 1, which shows a comparison of plants employing Lurgi
gasification for production of methanol, Mobil M-gasoline, and
Fischer-Tropsch synthesis, with and without, coproduction of
SNG. The column for efficiency shows the percentage of the in-
put Btu that comes out as product. The last column shows invest-
ment cost in dollars per million Btu output per year. The lower
efficiency and higher cost shown where SNG is not a product reflect
losses associated with conversion of methane formed in gasifica-
tion to synthesis gas for conversion to additional liquid product.
Table 1. Efficiency of Investment Cost Indirect Coal Liquefaction
Efficiency, Investment Cost,
$/106 Btu/yr
Methanol
Methanol
from Syn Gas
Methanol
Methanol
- Mobil M
Gasoline
Gasoline
Fischer-Tropsch
Gasoline
Gasoline
+ SNG
+ SNG
+ diesel
+ diesel + SNG
50.8
60.4
48.7
58.2
35.7
50.8
28.2
21.8
34.3
24.0
45.3
25.2
448
-------
The cost of production of liquid fuels is frequently given
in dollars per million Btu in all products. Because this ap-
proach fails to account for differences in the value of the end
products, however, it can give a distorted perspective of the
potential for a given technology to satisfy present needs. Also,
costs are often compared without due consideration of uncertain-
ties attributable to stage of development. One recent study,
however, generated data that give some feeling for the importance
of these uncertainties in comparison of technologies. Data
from that report are shown in Table 2. The confidence index in
Column 1 has two components: a letter indicating stage of de-
velopment and a number indicating the estimated reliability of
the cost. The energy cost is based on the total energy value for
all products. The "reference price" is based on Btu outputs,
adjusted downward in proportion to their value relative to
gasoline for all products that are less valuable.
Data such as these must be considered approximations subject
to variation not relating to the skill or objectivity of the
estimators. They do, however, highlight several important points
that are creating pressure to use presently available technology
as a basis for beginning the development of a synthetic fuels
industry:
1) Fischer-Tropsch and methanol fuels are more costly than
new processes are expected to be. The estimated costs,
however, are more reliable (as indicated by the confidence
index) than those for the four developmental processes.
2) The cost advantages of developmental processes are not
great. Unforeseen circumstances or inflation during the
developmental period could cause them to be more expensive
than plants that could be built now.
3) When credits are applied for quality of product, the
relative economics change significantly. The net result is
that methanol shows the lowest reference price and a con-
fidence index better than that for any other process except
Fischer-Tropsch.
449
-------
TABLE 2. COST COMPARISON FOR ALTERNATIVE PROCESSES FOR
PRODUCTION OF LIQUID FUELS FROM COAL12
Fischer-Tropsch
Methanol
Mobil M-Gasoline
Exxon donor solvent
H-coal
SRC II
Confidence
index*
A-2
A-2
C-3
C-3
C-2
B-4
Energy cost, Reference price,
$/lQ6 Btu $/1Q6 Btu
4.99
4.32
4.84
3.96
3.58
3.62
5.52
4.54
4.91
5.40
4.81
5.59
* Confidence index factors:
Process development
D - Exploratory stage - not beyond
simple bench tests
C - Development stage - operated on
small integrated scale only
B - Pre-commercial - successful
pilot plant operation
A - Complete - process demonstrated
sufficiently to insure commercial
success
Economic reliability
Screening estimate, very
approximate
Incomplete definition for
estimates used
4
3
2 - Firm basis for values developed
1 - Values considered to be satis-
factory for commercial venture
It is not intended to suggest that these data indicate
superiority of any given process. Many situation-specific fac-
tors (type of coal, markets served, transportation modes availa^
ble) will influence process selection for commercial projects.
The results do, however, illustrate the potential advantages of
applying available technology now.
450
-------
POTENTIAL MARKETS FOR METHANOL FUEL
Methanol fuels have been demonstrated in a variety of
applications:
1. Fuel for motor vehicles, alone, or in combination with
gasoline.
2. Fuel for electric utilities, to be burned as supplemental
fuel in coal-fired boilers and in combustion turbines.
3. Fuel to replace distillate oil and residual oil being
burned in boilers and furnaces for space heat in the
residential and commercial sectors.
4. Fuel to replace distillate oil for industrial boilers
and direct-fired processes.
FUEL FOR MOTOR VEHICLES
Opinions differ on the ease with which the methanol could be
introduced as fuel for motor vehicles. Many believe that methanol
could be utilized, with adaptation of the engines, in all types
of motor vehicles. Also, many believe that a fuel consisting of
up to 10 percent methanol in gasoline could be used in gasoline
engines with only minor changes in present practices. Even at
the 10 percent level, the market would be significant. Further,
even if it is determined that the use of methanol pure or at
higher concentrations in gasoline,-will require time-consuming
adjustments, the feasibility of converting methanol to gasoline
with the Mobil-M process could open the way for substituting
synthetic fuels for unlimited amounts of our gasoline consumption.
Gasoline consumption in 1980 was 2409 x 10 bbl (12.66 x
10 Btu)-* Ten percent of this total is equivalent to over 60
million tons of methanol. This demand alone would consume the
*
All fuel consumption data taken'from Reference 13.
451
-------
output of eight 25,000-ton/day plants* of the type that has been
suggested as an optimum size.
FUEL FOR ELECTRIC UTILITY BOILERS
Utilities currently burn a substantial amount of both dis-
tillate oil and residual oil; the distillate is used mostly as a
supplemental fuel for startup and for flame stability in coal-
fired boilers or in oil-fired combustion turbines. Residual oil
is burned as a base fuel in large boilers. Methanol has been
demonstrated to be applicable as a substitute for both types of
fuel and has been used to fire utility boilers. The 1980 con-
sumption of distillate by electric utilities was 39 x 10 bbl
(0.22 x 10 Btu) and their consumption of residual oil was 438
x 10 bbl (2.75 x 10 Btu). Replacement of the distillate with
methanol would represent a valuable use as a premium fuel and
would consume about 10 x 10 tons per year of methanol at present
levels of consumption.
Although methanol could be substituted for residual oil as a
base fuel, this probably would not be the best application of a
premium fuel in light of other possible uses. Substitution for
the portion of residual oil that is imported would operate to
reduce dependence on foreign oil. But with refineries worldwide
necessarily continuing to produce residual oil (as they will for
some years), outlets will be needed. Utilities and industrial
combustion may be the most effective way to utilize the residual
oil, especially that fraction produced in the United States,
which is the dominant portion of that used in this country.
FUEL FOR RESIDENTIAL AND COMMERCIAL SPACE HEAT
The residential and commercial sectors consume large amounts
of distillate and residual oil, which is used almost exclusively
for space heat and could beneficially be replaced by methanol.
Substitution for residual oil in these sectors would offer advan-
tages in that the more complex equipment for burning heavy oil in
*
Assumed to be operated at 90 to 95 percent of capacity.
452
-------
commercial establishments could be eliminated, air pollution re-
duced, and dependence on foreign oil reduced. Consumption
levels in the residential and commercial sectors in 1980 were
distillate, 353 x 106 bbl (2.06 x 1015 Btu), and residual, 86
x 10 bbl (0.54 x 10 Btu). This is equivalent to 150 x 106
tons of methanol at present levels of consumption.
FUEL FOR INDUSTRIAL BOILERS AND DIRECT-FIRED PROCESSES
Methanol also appears to be a satisfactory substitute for
distillate oil in industrial boilers. Distillate oil burned in
the industrial sector goes both into boilers and into direct-
fired processes such as dryers and kilns. Even though direct-
fired processes are highly heterogeneous, it seems reasonable to
assume that methanol could be used in almost any situation where
distillate is direct-fired. For reasons discussed in connection
with utility boilers, the industrial combustion of residual oil
is not included as a potential market for methanol fuel, even
though it could be used in such applications.
The industrial consumption of distillate oil in 1980 was 257
x 106 bbl (1.50 x 1015 Btu), the equivalent of 75 x 106 tons of
methanol.
Table 3 shows a summary of the major applications in which
methanol appears to be substitutable.
TABLE 3. SUMMARY OF METHANOL-SUBSTITUTABLE OIL CONSUMPTION
(1980)
Methanol Oil
Consumption, equivalent, equivalent,
1015 Btu 106 tons 106 bbl
Distillate oil, utility sector 0.22 11 39
Distillate oil, res/comm sectors 2.06 103 353
Residual oil, res/comm sectors 0.54 27 86
Distillate oil, industrial sector 1.50 75 257
Motor gasoline (10%) 1.27 64 241
5.59 280 976
453
-------
The total consumption shown in Table 3 amounts to over 15
percent of the total U.S. oil consumption of 34.3 x 10 Btu in
1980. This figure would be considerably larger if it were assumed
that methanol converted to gasoline with the Mobil-M process
could be substituted for the entire gasoline consumption of 12.66
x 10 Btu. Also, amounts for consumption of diesel fuel (2.33
x 10 Btu in 1979) are not included, even thought it is said to
be replaceable with methanol with appropriate engine modifications.
Replacement of the oil products indicated in Table 3 with
methanol would require building about thirty-five 25,000-ton/day
plants at a cost of about $100 billion. In terms of oil consump-
tion, this comes to a little under 3 million barrels per day, or
about 50 percent of our imports. An additional 65 to 70 plants
costing around $175 to 200 billion would be required to produce
gasoline in amounts equal to 1980 consumption.*
*
Plant sizes assumed and costs estimated are from Reference 7.
454
-------
PROSPECTS FOR COMMERCIALIZATION OF METHANOL AS FUEL
It is widely accepted that nontechnical problems such as
lack of assured markets, unclear policies in regulatory agencies,
potential siting difficulties, and related social, economic, and
institutional problems are the main barriers to commercialization
of methanol fuel or other fuels produced by presently available
technologies. Growing pressure for the use of present technology
to replace petroleum-derived fuels should alleviate these prob-
lems. If it does, the prospects for methanol fuels will depend
primarily on advantages they offer over competitive fuels. The
following is a discussion of methanol relative to the other fuels
that might be produced by present technology to compete, directly
or indirectly, with methanol fuels in replacement of petroleum-
derived liquid fuels. These are the principal options:
1. Natural gas.
2. Low- or medium-Btu gas made from solid fossil fuels
with existing technology.
3. Gasoline derived directly from synthesis gas from coal
using Fischer-Tropsch technology.
4. Gasoline produced by subsequent processing of methanol,
derived from fossil fuels, using the Mobile-M process.
5. Ethanol produced by fermentation of agricultural crops.
6. Shale oil.
It might be argued that synthetic natural gas (SNG) and
fuels produced from direct liquefaction should be considered
455
-------
along with those listed above. They are not, however, because
these technologies are not equivalent to the others in terms of
stage of development or potential application. Although one SNG
plant is reported under construction, this plant will produce
supplemental fuel for existing natural gas distribution systems
and will not be in direct competition with the fuels being con-
sidered. Moreover, the facts do not indicate that direct lique-
faction technologies are presently utilizable in the same sense
as those used for the above fuels.
METHANOL VERSUS NATURAL GAS
Methanol and natural gas both have potential for replacement
of petroleum-derived fuels. Gas can be used directly or as a
feedstock for production of methanol. Whether or not natural gas
should be used in either way depends on the adequacy of supplies
for other critical uses. Until recently the expanded use of
natural gas would have been impossible because of short supplies.
Since passage of the Natural Gas Policy Act of 1978, which pro-
vides for progressive deregulation of natural gas prices, drill-
ing has been increased so that supplies have increased. Although
the proven reserves for the lower 48 states were only 195 tril-
lion cubic feed (Tcf) at the end of 1979 (a 10-year supply at
1980 rates of consumption), the total remaining conventional gas
14
resources have been estimated to be 563 to 1219 Tcf. The
higher figure is the most recent estimate. In addition, natural
gas is known to be recoverable from "unconventional" domestic
sources, which include geopressure zones, Western "tight sands",
methane from coal seams, and Devonian shales underlying Appa-
lachia. ' Estimates of recoverable natural gas from these
resources were recently summarized; these data are presented in
Table 4. The wide range of values reflects our present poor
understanding of the character of the resources.
456
-------
TABLE 4. ESTIMATED UNCONVENTIONAL GAS RESOURCES FOR THE UNITED STATES^
Estimated total Recoverable
resource in place, resources,
Resource Km3 (Tcf) Km3 (Tcf)
Western tight 1,400-17,000 710-8,860
gas sands (49-600) (25-313)
Eastern devonian 2,100-20,000 280-14,300
gas shales (74-706) (10-505)
Methane from 2,000-24,000 450-13,800
coal seams (71-847) (16-487)
Geopressured 85,000-1,400,000 4,200-57,000
methane (3,000-49,420) (148-2.012)
90,500-1,461,000 5,640-93,960
(3,794-51,573) (199-3,317)
In recent months natural gas advocates have argued for "the
natural gas option" as a worldwide approach to reducing dependence
on oil. They point out that proven worldwide reserves of conven-
tional gas are 2200 Tcf. Estimated remaining undiscovered re-
serves are said to be 7500 Tcf, giving a total resource that is
believed adequate for 50 years even if the present annual world-
wide consumption rate of 50 Tcf is doubled. Even if one
accepts a lower estimate made in 1975 of 6000 Tcf for total re-
T 8
coverable conventional reserves, the world supplies seem impres-
sive. Utilization of the worldwide gas supplies will, however,
require capture of the gas and transport to remote demand points.
Some propose that this be accomplished with pipelines and ships
transporting liquid natural gas (LNG). Others suggest that where
pipelines must be over 5000 miles long or ship transport exceeds
3000 miles, conversion to methanol for shipment is more economical,
In addition, the methanol advocates cite the advantages of liquid
fuels in markets such as transportation fuels, where natural gas
is not widely applicable.
In summary, it appears that natural gas may become increas-
ingly important as a direct substitute for petroleum. At the
same time, it also seems appropriate to consider conversion of
457
-------
substantial quantities of natural gas to methanol by present
technology to produce direct substitutes for some of the liquid
fuels that we are now consuming in amounts equivalent to about 34
x 10 Btu per year. These fuels are now produced partly from
domestic oil supplies and partly from about 17 x 10 Btu of
imported oil. The magnitude of these numbers is illustrated by
comparison with the natural gas consumption for recent years of 20
Tcf/yr, which represents approximately 20 x 10 Btu. No single
approach will provide more than a partial solution. Even if the
use of natural gas is greatly expanded, there might still be a
role for methanol fuels.
METHANOL VERSUS LOW- AND MEDIUM-Btu GAS FROM COAL
Low- and medium-Btu gas can be produced with existing tech-
nology and used on-site. Medium-Btu gas, which can be moved by
pipeline for short distances, can be produced for use in plants
within about 100 miles. Hence, where coal is available near a
point of demand, there may be little incentive to produce methanol
from coal-derived gas rather than burn the gas directly. Sup-
plies of solid fuel in remote locations, however, might be profit-
ably gasified, converted to methanol, and shipped to distant
demand points. This is especially true of low-grade fuels, which
are expensive to ship (on a Btu basis) and are more effectively
gasified than high-grade coal. Several such plants are being
19
designed to utilize lignite in the United States. Peat, which
has little value as fuel except on-site, has also been suggested
to be an excellent gasification feedstock. One report indicates
that 11,000 and 37,000 square miles of peat bogs with thicknesses
of 5 to 25 ft are located in the U.S. and Canada, respectively.
The data suggest that the U.S. supply might be equivalent to 6.5
billion tons that could yield about 2.0 billion tons of methanol
f- 0 C]
or 80 x 10 ton/yr for 25 years. This annual amount is over
12 percent of our total gasoline consumption in 1980.
458
-------
METHANOL VERSUS GASOLINE FROM COAL (FISCHER-TROPSCH)
Production of gasoline from coal by the Fischer-Tropsch
process might be an attractive alternative for production of non-
imported liquid fuels. This technology has been used for many
years in South Africa and is being greatly expanded in new capac-
ity. The process, however, produces a wide variety of products
for which markets must be available. Further, the quality of the
fuel as produced is low relative to methanol fuel or Mobil-M gas-
oline. Additional processing is required to produce high-octane
gasoline. Also, the Fischer-Tropsch process appears to be rela-
tively lower in efficiency and higher in cost, as discussed
earlier, when the value of the products is considered. The
process does, however, produce a significant amount of gasoline
directly, and unless the Mobil-M process is successful, it will
be the only currently available option for doing so.
METHANOL FUEL VERSUS GASOLINE FROM METHANOL (Mobil M-Gasoline)
It may be debatable whether the Mobil-M process can be
considered available technology, since no full-scale process is
in operaton. It is, however, further along in development than
other processes in that a commercial plant is to be built. Some
consider that processing of methanol in an additional step, as
this process does, is unnecessary because methanol is claimed to
be usable in amounts of 10 percent or more with gasoline in motor
vehicles of conventional design and to be usable pure in motor
vehicles of modified designs. Others argue that this is an
oversimplification, claiming that certain properties of methanol,
including its corrosiveness, toxicity, and affinity for water,
constitute problems that would require time-consuming modifica-
tions of present practices if methanol is to be widely used in
motor vehicles. The Mobil M-Gasoline process in claimed to have
95.5 percent thermal efficiency in conversion, and is said to add
21
only 5£ per gallon to the cost of output fuel. If this per-
formance is attainable, the technology could be quite useful in
459
-------
attaining faster penetration for coal-derived fuels in the trans-
portation fuel market.
METHANOL VERSUS ETHANOL FROM FERMENTATION OF CROPS
Ethanol from fermentation of crops is being used as motor
fuel both in the United States and abroad. Problems and advan-
tages associated with its use are in many ways similar to those
associated with the use of methanol. Ethanol is, however, sub-
ject to certain unique limitations, primarily associated with
availability of raw materials. Thus, even though ethanol produc-
tion is a useful technology, it may be more limited in applica-
bility than that for methanol fuels, in the long run.
Ethanol plants are expected to be relatively small so that
they can be located near raw material supplies (such as corn) and
near outlets for byproduct animal feed, the sale of which is es-
sential to process economics. Also they effectively remove land
from food production at a time when there is already concern over
the rate at which farm land is being lost to other uses. Experi-
ence to date suggests that ethanol will play a role in replacement
of petroleum fuels but is not likely to be a dominant contributor.
METHANOL VERSUS FUEL FROM OIL SHALE
Fuels from shale oil, like M-Gasoline, have not been pro-'
duced commercially, but plans have been made for commercial
plants. There is a considerable body of pilot plant data to
support the scaleup of oil shale processes. The technical risk
for commercial plants appears to be minimal. Further, oil shale
deposits are very extensive and could supply our oil needs for
hundreds of years. Because of economic uncertainties, however,
developers are reluctant to make firm committments without such
incentives as guaranteed markets. Hence, prospects are poor for
near-term production of large amounts of synfuel from oil shale.
Also, crude feedstocks from oil shale are of low quality compared
with methanol. Thus, it appears that markets for methanol fuel
should exist even if shale oil ventures are highly successful.
460
-------
CONCLUSIONS
Methanol fuel technology appears to be very cost-competitive
with other technologies that could be applied in a synthetic
fuels industry today. Although the projected cost of methanol
fuels is somewhat higher than today's prices for distillate oil
and gasoline, methanol fuel plants built now could prove to be
highly profitable at prices that may prevail when they come on
stream.
The "clean burning" characteristics of methanol make it
potentially attractive from the standpoint of combustion system
design and control of environmental impacts associated with its
use. Also, methanol is easily transportable and could be pro-
duced from abundant supplies of low-grade fossil fuels located in
regions of the United States remote from points of demand for
premium fuels. Hence, technology for production of methanol
could be applied to utilize energy supplies that would otherwise
be of limited usefulness.
Methanol fuels seem to be an attractive alternative to
premium fuels in several critical applications that are expected
to grow in importance. One of the most important involves re-
placement of gas and distillate oil fired in turbines used by
utilities for peaking, in combined cycles, or "repowering" to
increase the capacity of existing power plants.
The use of methanol fuel technology to convert natural gas
to liquid fuels as a short-term solution for oil shortages should
be given serious consideration. Markets in which methanol fuels
could be substituted are large and represent a significant por-
tion of our current oil imports. The amounts of natural gas that
461
-------
could be produced over the next 20 years are highly controver-
sial. The optimistic estimates suggest that allocation of sig-
nificant quantities to production of liquid fuels could be
helpful in solution of short-term problems.
A thorough study of possibilities for the use of methanol
fuels on a wide scale is needed. Such a study should begin with
analysis of gaps in the available information, which has been
developed in piecemeal studies conducted over the past 10 to 15
years. This full-scale analysis should lead to definitive con-
clusions with respect to the policies to be adopted in future
energy programs.
462
-------
REFERENCES
1. Pefley, R.A., et al. Characterization and Research Investi-
gations of Methanol and Methyl Fuels. University of Santa
Clara for U.S. Environmental Protection Agency. EPA Grant
No. R803548-01, August 1977.
2. Wagner, T.O., et al. Practicality of Alcohols as Motor
Fuel. Society of Automotive Engineers Technical Paper
Series, 0148-7191/79/0226-0429, March 1979.
3. Baratz, Bernard, et al. Survey of Alcohol Fuel Technology.
Mitre Corp. for National Science Foundation. NTIS No.
PB256 007, November 1975.
4. Duhl, R.W. Methanol as Boiler Fuel. Chemical Engineering
Progress, July 1976.
5. Seglem, C.E. Performance of Combined Cycle Power Plants.
Presented at 13th Middle Atlantic Regional Meeting of the
American Chemical Society, March 1979.
6. Hayden, A.S.C. Utilization of Methanol in Stationary Source
Combustion. Canadian Combustion Research Laboratory,
Ottowa, Canada, November 1977.
7. Othmer, Donald F. Methanol: The Versatile Fuel and Chemical
Raw Material. Polytechnic Institute of New York, Brooklyn,
New York, March 1980.
8. Harney, Brian M., and G. Alex Mills. Coal to Gasoline via
Syngas, Hydrocarbon Processing, February 1980.
9. Kuo, J.C.W., and M. Schreiner. Status of the Mobil Process
for Converting Methanol to High Quality Gasoline. Presented
at the 5th Annual Conference on Commercialization of Coal
Gasification, Liquefaction and Conversion to Electricity,
Pittsburg, Pennsylvania, August 1978.
10. Anastai, J.L. Sasol: South Africas Oil from Coal Story.
TRW, Inc., for U.S. Environmental Protection Agency,
EPA-600/8-80-002, January 1980.
463
-------
11. Rudolph, Paul F.H. Synfuels from Coal: How and at What
Cost? Presented at 7th Energy Technology Conference,
Washington, D.C., U.S.A.
12. Rogers, K.A., and R.F. Hill. Coal Conversion Comparisons.
The Engineering Societies Commission on Energy, Inc., for
U.S. Department of Energy, Contract No. EF-77-C-01-2468,
July 1979.
13. Energy Information Administration Annual Report to Congress.
U.S. Department of Energy, April, 1981.
14. The Role of Natural Liquefied Gas in a Worldwide Gas Energy
Option. American Gas Association Monthly, April 1980.
15. Hodgson, Bryan. Natural Gas: The Search Goes On. National
Geographic, November 1978.
16. Rosenberg, Robert B., and John C. Sharer. Natural Gas from
Geopressured Zones. Oil and Gas Journal, April 28, 1980.
17. McCormack, Wm. T., Jr. AGA Study Assesses World Natural Gas
Supply. Oil and Gas Journal, February 13, 1978.
18. Greatest Gas Potential is in Middle East and Russia. Oil
and Gas Journal, May 26, 1975.
19. Methanol-A Synthetic Liquid Fuel. Mechanical Engineering,
June 1980.
20. Barr, Wm. J., and Frank A. Parker. The Introduction of
Methanol as a New Fuel into the United States Economy.
American Energy Research Company, McLean, Virginia, for
Foundation for Ocean Research. March 1976.
21. Mersel, S.L., et al. Gasoline from Methanol in One Step.
Chemtech, February 1976.
464
-------
CONVERSION AND EQUIVALENCY FACTORS
1 bbl (barrel) = 42 gallons
1 bbl gasoline = 5.4 x 10 Btu
1 bbl methanol = 2.7 x 106 Btu
1 ton methanol = 20 x 106 Btu
1 ton methanol = 7.4 bbl methanol and is equivalent to 3.7 bbl
gasoline
1 Tcf (trillion cubic feet) of natural gas =10 Btu
1 Km (cubic kilometer) of natural gas = 35.3 x 10 cf (cubic
feet)
3 12
1 Km of natural gas = 35.3 x 10 Btu
Density of gasoline = 5.8 Ib/gal
Density of methanol = 6.6 Ib/gal
A 25,000 ton/day methanol plant produces 8.2 x 10 ton/yr
which is equivalent to 30.3 x 10 bbl of gasoline.
Motor gasoline consumption for the U.S. was 2,409 x 10 bbl
in 1980. This is equivalent to 12.66 x 10 Btu. This amounts
to 6.3 x 106 bbl/day.
Oil imports for 1980 were 6.8 x 10 bbl/day. This included
refined petroleum products amounting to 1.6 x 10 bbl/day (3.2
x 10 Btu/yr) and crude oil amounting to 5.2 x 10 bbl/day (10.4
x 1015 Btu/yr).
Natural gas consumption in the United States in 1980 was
21.5 Tcf, which is equivalent to 21.5 x 10 Btu or 10.7 x 10
bbl/day of crude oil.
465
-------
METHANOL AS AN ALTERNATIVE TRANSPORTATION FUELt
by: Richard Rykowski, Dwight Atkinson, Daniel Reiser,
John McGuckin, David Fletcher, Jeff Alson, and
Murray Rosenfeld
Emission Control Technology Division
U.S. Environmental Protection Agency
2565 Plymouth Road
Ann Arbor, MI 48105
ABSTRACT
Over the remaining years of this century synthetic fuels will
play a key role in the nation's drive for energy independence.
Although self-reliance is indeed a desirable goal, many people believe
it cannot be achieved without significant compromises in environmental
quality. This may not be the case. One synfuel, methanol, could be
used to replace both gasoline and diesel fuel and yield environmental
benefits. This paper compares methanol with synthetic fuels from
other coal liquefaction processes in terms of the environmental and
economic consequences of their use.
INTRODUCTION
Several factors must be addressed when considering the viability
of an alternative motor fuel. These can broadly be grouped into two
categories, environmental and economic. Each of these categories
would include the production, distribution, and in-use aspects of the
fuel in question. In the report that follows, we have attempted to
address these issues for several alternative automotive fuels, espe-
cially methanol, which could be produced from coal. In addition to
methanol from indirect liquefaction, fuels from the following tech-
nologies were examined: the Mobil Methanol to Gasoline (MTG) indirect
liquefaction process, and the Exxon Donor Solvent (EDS), H-Coal, and
Solvent Refined Coal (SRC-II) direct liquefaction processes.
Of the subjects examined below, the environmental analyses of
production and distribution are the most general since the least
amount of information was available in these areas. Although more
detail is provided in other sections, the preliminary nature of the
entire report should be emphasized. More work is needed before final
conclusions can be stated with confidence.
466
-------
ENVIRONMENTAL
PRODUCTION AND DISTRIBUTION
It should first be recognized that coal itself contains many
diverse elements and compounds in addition to hydrogen and carbon,
such as organic nitro-compounds, organic and inorganic sulfur, and
trace metals, such as lead, arsenic, etc. The conversion of coal to
other fuels offers a number of opportunities for these pollutants to
reach the environment in harmful ways, regardless of the particular
conversion process used.
One potential advantage of processes which gasify coal, such as
those leading to methanol or gasoline (via methanol), is that the
gasification itself places most of the potentially harmful elements
and compounds into forms which can be removed relatively easily. For
example, minerals and heavy metals are removed from the gasifier as
slag which cools to a solid. While the high concentration of metals,
etc. requires careful disposal, this disposal may not be as difficult
as that connected with coal liquefaction. With direct coal liquefac-
tion, these compounds are entrained in the heavy organic liquid and
must be separated from the liquid phase later in the process. This
solid-liquid separation is very difficult (basic research is still
underway in this area[l]) and the separation from a solid cannot be
made as completely as the separation from a gas. Inevitably, some
liquid will end up with the solid waste and some heavy metals will be
left in the crude fuels. Thus, not only may the solid waste disposal
problem be worsened by the addition of complex, polycyclic organic
material to the waste, but the fuel itself still contains more
minerals and heavy metals.
One factor which may mitigate or eliminate this problem for most
direct liquefaction processes is the high probability that most of the
heaviest liquid fraction will be gasified to produce hydrogen.[2,3]
If this is done, most of the minerals and heavy metals can be removed
from the gas fairly early, since this heavy liquid fraction should
contain most of the coal's impurities. Thus, the full extent of this
disadvantage may depend primarily on the fraction of the impurities
which can be removed via gasification and the fraction which must be
removed directly from the liquid itself.
Another potential advantage of gasification over direct liquefac-
tion is the fact that all of the organic nitrogen and sulfur is broken
down to simple compounds like ammonia and hydrogen sulfide. These are
relatively easy to separate from the carbon monoxide and hydrogen
which make up the major part of the synthesis gas. Also, since the
carbon monoxide and hydrogen must be essentially free of nitrogen and
sulfur before reacting over the catalyst to form methanol, there is an
economic incentive to remove these two elements. Although the nitro-
gen which is not removed prior to the catalyst will be removed by the
467
-------
catalyst itself, slowly deactivating it, any unremoved sulfur would
rapidly deactivate the catalyst.
Coal liquefaction, on the other hand, inherently leaves most of
the sulfur and nitrogen in the liquid phase, bound with the organics.
The most effective technique to remove these compounds is hydrogena-
tion, which also is used to upgrade the fuel. However, hydrogenation
is expensive, because of the large amounts of hydrogen consumed, and
will likely be limited to only the degree that is necessary to market
the fuel. [4] If the fuel is upgraded to gasoline or high quality No.
2 fuel oil, most of the sulfur and nitrogen will be removed and there
should not be any significant problems. However, that portion of the
synthetic crude which may be burned with little or no refinement could
contain relatively high levels of these elements and represents more
of an environmental hazard than gasification products.
The remaining distinct difference between the environmental
effects of coal gasification and coal liquefaction processes (prior to
end-use) is in exposure to the fuel itself, after production and in
distribution. While coal liquids are for the most part hydrocarbons
and, as such, are similar to petroleum, they are more aromatic and
contain significant quantities of polycyclic and heterocyclic organic
compounds. Some of these compounds are definitely mutagenic in bio-
assays and many have produced tumors in animals. Thus, while the non-
carcinogenic health effects of these materials would be more similar
to those of crude petroleum, they would definitely have the potential
to be more carcinogenic. There is also some evidence that much of
this bioactivity can be removed by moderate to severe levels of hydro-
genation which would occur if high grade products were produced.
Thus, again the potental hazard is dependent upon the degree of hydro-
genation given the products.
Indirect liquefaction products, on the other hand, do not appear
to exhibit mutagenicity or carcinogenicity. Methanol is neither muta-
genic nor carcinogenic and early tests run on M-gasoline have shown it
to be nonmutagenic, similar to petroleum-derived gasoline. Therefore,
either of these two products offers some degree of benefit over direct
liquefaction products. It is possible, however, that methanol pro-
duced from coal may contain impurities and that such impurities may
affect exhaust products when used. Research needs to be done in this
area, also.
Methanol, of course, is highly toxic in heavy exposures, leading
to blindness or death. Much of its notoriety in this area is due to
people confusing it with ethanol and drinking it in large quantities.
Hydrocarbon fuels, while also toxic, do not suffer from this confusion
and are not often taken internally. With proper education of the
public, confusion between methanol and ethanol should be minimized.
However, more work is still needed in this area also.
468
-------
The final point which deserves mention here is the difference
between the effect of an oil spill and a methanol spill. The effects
of oil spills are well known; oil films stretching for miles, ruined
beaches, surface fires, etc. The effects of a methanol spill are
expected to be quite different, primarily because methanol is soluble
in water. While high levels of methanol are toxic to fish and fauna,
a methanol spill would quickly disperse to nontoxic concentrations
and, particularly in water, leave little trace of its presence after-
ward. [5] Sea life should be able to migrate back quickly and plant
life should begin to grow back quickly, though complete renewal would
take the time necessary for new plants to grow back. Also, if a
methanol fire does start, it can be effectively dispersed with water,
which is not possible with an oil fire. However, methanol flames can
be invisible, making them more difficult to avoid.
The various relative environmental aspects of synthetic fuels
production and use mentioned above are those which appear to stand out
at this time. More work, however, is still needed in most areas.
Although natural gas to methanol plants exist and have led to much
experience in handling methanol, questions related to methanol produc-
tion from coal are not known with absolute certainty since such large
scale facilities do not currently exist. Similarly, no real life
experience of the effects of the production of synthetic crudes
exists, nor of their use. Given these caveats and the need for fur-
ther research, however, the indirect liquefaction route to yield
methanol or gasoline (from methanol) appears to have some potential
environmental advantages over direct liquefaction processes.
VEHICLE USE
The data presented below were obtained from tests of actual
methanol engines. However, it should be noted that these data were
taken using engines which were only roughly converted to use of meth-
anol; fully optimized engines would be expected to show further
improvements in fuel efficiency and emissions.
The worst problem concerning methanol's actual use centers around
its low vapor pressure and high heat of vaporization. These proper-
ties make it difficult to start a neat methanol engine in cold wea-
ther. [6] Also, methanol has a very low cetane number of approximately
3, which means that it is very difficult to ignite in a compres-
sion-ignition engine (e.g., a diesel). Problems associated with
materials compatibility and lubrication also exist, but these problems
already appear to be solvable with existing technology, requiring only
that the auto designer know that methanol is going to be the engine
fuel.
Various techniques are already being tested which will improve
the cold-starting capability of gasoline engines operating on meth-
469
-------
anal, such as better mechanical fuel atomization, electrical fuel pre-
heating, and the blending of volatile, low boiling point components
into the methanol. Methanol's ignition problems are more serious in
diesel engines, but several possible solutions are being investigated,
such as glow plugs and spark ignition. Brazil already has an experi-
mental methanol-fueled diesel running on the road which uses rela-
tively inexpensive glow plugs as ignition aids and M.A.N. in Germany
has designed a diesel bus engine with spark ignition which runs on
methanol.[7,7a]
As will be seen later in the section on fuel consumption in the
economics section, the fuel properties of methanol which lead to these
difficulties also lead to many advantages, such as increased thermal
efficiency relative to gasoline engines. Past experience with both
gasoline and diesel engines has shown that the disadvantages of a fuel
can usually be overcome to allow exploitation of the advantages,
particularly when the advantages are as large as they appear to be for
methanol.
Methanol engines promise improved emission characteristics over
gasoline and diesel engines in a number of areas. Especially impor-
tant are low emissions of nitrogen oxides (NOx) and an absence of
emissions of particulate matter, heavy organics and sulfur-bearing
compounds. One possible side benefit of methanol use could be that
precious metal catalysts might not be needed for emissions control.
Because methanol fuel will contain no sulfur, phosphorus, lead, or
other metals, base metal catalysts (e.g., nickel, copper, etc.) may
suffice. One likely negative impact of methanol engines would be an
increase in engine-out aldehyde emissions, particularly formaldehyde.
Catalytic converters, however, would be expected to reduce aldehyde
emissions to acceptable levels. The available data supporting these
effects are discussed below.
A search of the literature shows a general consensus that meth-
anol engines produce approximately one-half of the NOx emissions of
gasoline engines at similar operating conditions, with individual
studies showing reductions between 30 percent and 65 per-
cent . [8,9,10,11,12] One of the major engine design changes expected
with methanol engines is the use of higher compression ratios to
increase engine efficiency. Experiments have confirmed the theoreti-
cal expectation that these higher compression ratios, with no other
design changes, will increase NOx emissions considerably due to the
higher combustion temperatures.[13,14J However, with high compression
ratios, less spark timing advance is needed. Retarding spark timing is
known to reduce both NOx emissions and engine efficiency. Fortuna-
tely, it has been shown that the combination of a much larger compres-
sion ratio with a few degrees of spark timing retard can both increase
thermal efficiency and decrease NOx emissions.[14] This raises the
possibility of methanol vehicles being able to meet the current 1.0
gram per mile NOx emission standard without the need for a NOx reduc-
tion catalyst.
470
-------
Use of methanol in a diesel engine should also reduce NOx emis-
sions by the same degree as that described above. Diesel engines have
higher peak combustion temperatures and the effect of a cooler-burning
fuel should actually be even more apparent in a diesel than in a gaso-
line engine. Unfortunately, no data to confirm this is yet available
from a diesel engine running on pure methanol. However, emission
tests have been performed on a dual-fuel diesel, where a small amount
of diesel fuel is injected to initiate combustion of the methanol.
These tests have shown NOx emission reductions as high as 50 per-
cent. [15,16]
These lower NOx emissions would aid many areas of the country in
attaining the ambient standard for N0£ in the future. (Most areas
are currently under compliance with the N0£ ambient air quality
standard, but many are projected to exceed it in the future as NOx
emissions continue to rise.) Lower NOx emissions would also help
alleviate the acid rain problem, though the majority of this problem
appears to be due to stationary source emissions. Finally, the use of
methanol would aj.so provide a method for heavy-duty engines to reduce
NOx emissions closer to the congressionally-mandated level without
giving up any of the fuel economy advantage of diesels, as will be
seen later.
The lack of hard data on diesels operating on pure methanol indi-
cated above will also be evident below as other aspects of meth-
anol-fueled diesel engines are discussed. The basic reason for this
lack of data is that until recently methanol has not been seriously
considered to be an acceptable fuel for a diesel engine because of its
very low cetane number. For many years, studies examining methanol as
an engine fuel concentrated on gasoline-type engines (fuel inducted
with combustion air). However, as the more recent studies are indi-
cating, it appears possible to burn methanol in a diesel accompanied
with some kind of ignition assist and, therefore, utilize the effi-
ciency of the diesel concept.
In addition to the positive effect on NOx emissions, use of meth-
anol engines should provide even greater benefits with respect to
emissions of particulate matter and heavy organics from diesels. Gaso-
line engines operated on unleaded fuel emit only small quantities of
particulate matter, composed primarily of sulfate particles. Thus, any
improvement in particulate emissions from switching to methanol from
gasoline would be small.
However, diesel engines emit large quantities of particulate mat-
ter consisting of solid carbonaceous particles (soot) and liquid
aerosols. The former are generally formed when the injected fuel
droplets are incompletely combusted, leaving carbon particles. These
solid particles can then serve as nuclei for more harmful organic
species to adsorb onto and as "vehicles" for such compounds to reach
(and possibly lodge in) the deep regions of the lung. Although reduc-
tions in diesel engine particulate have been reported, particulate
471
-------
matter seems to be an inherent pollutant in diesel-fueled compression
ignition engines .
Methanol, on the other hand, is a "light" fuel relative to diesel
fuel and should produce far less carbonaceous particles, as do other
hydrocarbon fuels "lighter" than diesel fuel. In addition, since metha-
nol does not contain inorganic materials like sulfur or lead, there should
not be any other types of solid particulate formed. Accordingly, with pure
methanol there would be no nuclei for liquid aerosols to adsorb onto and
total particulate emissions would be expected to be near zero. [17] This
is certain to be the case with a well designed methanol-fueled spark-igni-
tion engine. [18] Unfortunately, however, we know of no studies which
have measured particulate from compression ignition engines burning neat
methanol. Several studies (all of which used a small amount of diesel pilot
fuel) have reported much lower smoke levels, both in single-cylinder tests
and in a 6-cyclinder, turbocharged, direct-injected engine. [7,15,19] There
seems to be little question, however, that neat methanol combustion in com-
pression ignition engines would result in very low (and possibly zero) par-
ticulate emissions. This would result in a very important environmental
advantage compared to diesel fuel combustion.
As mentioned earlier, formaldehyde emissions from methanol engines
are of some concern since there is some evidence that formaldehyde is carci-
nogenic . Formaldehyde is an intermediate specie in methanol oxidation and
would be expeced to be emitted from methanol engines in greater quantities
than either diesel or gasoline engines . Many studies have shown total alde-
hyde emissions (mostly formaldehyde) from methanol engines to be two to ten
times greater than aldehyde emissions from gasoline engines. [20,21,22,23]
At the same time, catalytic converters have been shown to be effective
in removing approximately 90 percent of exhaust aldehydes. [9,10,23,24]
Much research has been performed regarding the parameters which influence
aldehyde formation in gasoline engines, with low exhaust temperatures and
high oxygen concentrations identified as leading to higher formaldehyde
formation rates, and this knowledge should facilitate aldehyde control in
future engine designs. [22,25] Aldehyde emissions from methanol combustion
in diesel engines are also expected to be greater than from diesel fuel
combustion.
The last benefit of methanol engines to be discussed concerns sulfur
emissions. Because of the way methanol is produced it contains essentially
no sulfur. And, if there is no sulfur in the fuel, no emissions of sulfur-
bearing compounds, such as sulfur dioxide, sulfuric acid, or hydrogen sulfide,
can occur. This is a slight improvement over gasoline emissions, since gaso-
line does have a small amount of sulfur in it. Catalyst-equipped gasoline
engines currently emit between 0.005 and 0.03 grams per mile of sulfate and
this would disappear with the use of methanol, even if catalysts were still
used.
(Rev. 5/25/82)
472
-------
The improvement over the diesel, however, would be more pro-
nounced. Diesel fuel currently contains 0.2-0.5 percent sulfur by
weight. This translates into about 0.25 grams per mile of elemental
sulfur from diesel trucks (0.5 grams per mile of sulfur dioxide, or
0.75 grams per mile of sulfate, equivalent). Diesel cars emit about
one-fifth this amount. Since the sulfur level in diesel fuel is
expected to rise in the future, these emission levels would also rise
in the future. With the use of methanol these emissions would dis-
appear altogether.
ENVIRONMENTAL SUMMARY
Although coal contains many substances which could be environmen-
tally damaging, it appears that indirect liquefaction processes, meth-
anol and Mobil MTG, can facilitate their removal easier than is pos-
sible through direct liquefaction routes such as EDS, SRC-II and
H-Coal. Further, since indirect liquefaction necessitates the removal
of all sulfur before the fuel is synthesized, the use of relatively
cheap base metal catalysts (as opposed to noble metals currently in
use) on automobiles is a possibility.
Neither methanol nor Mobil M-gasoline appear to exhibit mutageni-
city or carcinogenicity- It should be remembered, however, that com-
mercial coal-to-methanol plants are not yet available so the influence
of possible impurities is not yet known. Direct coal liquefaction
products are more aromatic and contain significant quantities of poly-
cyclic and heterocyclic organic compounds, some of which are muta-
genic. There is some evidence, however, that much of this bioactivity
can be removed by moderate to severe levels of hydrogenation. More
work needs to be done in these areas before definitive conclusions can
be reached.
The effects of a methanol spill are expected to be quite dif-
ferent from that of the classical oil spill since methanol is soluble
in water. Although high levels of methanol are toxic to fish, a meth-
anol spill should quickly disperse to nontoxic levels.
Methanol engines promise emission benefits over both gasoline and
diesel engines. Lower emissions of nitrogen oxides, and the virtual
absence of particulate matter, heavy organics and sulfur bearing com-
pounds from vehicle exhaust are promising. A possible detriment of
methanol engines is that they emit higher amounts of aldehydes, prin-
cipally formaldehyde which is carcinogenic. Catalytic converters,
however, have been shown to be effective in removing 90 percent of
exhaust aldehydes. As was the case with the environmental conse-
quences of synfuel production, more work needs to be done in the vehi-
cle-use area as well.
473
-------
ECONOMIC
We have analyzed a large number of studies in order to estimate
the costs associated with the production and use of synthetic fuels.
A superficial review of their conclusions quickly revealed a wide
variety of conclusions and recommendations. One reason for this is
that the economic bases used by the various studies often differ,
affecting costs by as much as 100 percent. Another reason is that
each study uses the best information available at the time of the
study. Since the product mixes, efficiencies and costs of many of
these processes, especially the direct liquefaction processes, change
frequently as more is understood about the process, studies performed
even 2 or 3 years ago cannot be compared to the latest studies.
Thus, we - have attempted to go back in each instance to the ori-
ginal engineering studies to assess the viability of the cost esti-
mates. We also have compared the available designs of each process to
ascertain which are out-dated or based on now inaccurate assumptions.
After doing this, the projects were placed on the same economic basis
and adjusted for plant size.
While the difficulties and apparent discrepancies described above
primarily involve the costs of producing synthetic fuels, the overall
economic picture involves more. The entire process of producing syn-
thetic fuels and using them in motor vehicles will be broken down into
three areas. The first area consists of the production of a usable
liquid fuel from raw materials. The second area consists of distri-
bution of this fuel. Finally, the third area includes the use of these
fuels in motor vehicles. All costs will be presented in 1981 dol-
lars. It should be noted that the general approach followed in this
section is from a long-term perspective. That is, we have not identi-
fied any detailed costs associated with the implementation of methanol
as a "new" transportation fuel.
PRODUCTION COSTS
Determining the economics of the production of usable synthetic
liquid fuels is probably the most difficult of the three areas to be
examined. The engineering and financial bases that have been chosen
are shown in Tables 1 and 2. As shown in Table 1, two different sets
of financial parameters were chosen. These were selected from a sur-
vey of recent studies[26,27,28,29, 30,31] done on coal liquefaction
processes and represent two extreme cases for capital charge. The low
capital charge rate and accompanying parameters were chosen from the
ESCOE report [26] while the high capital charge data were taken from
the Chevron study.[28] The important factors yielding these two CCRs
are also shown in Table 1.
Table 2 shows the remaining input factors. All plants were nor-
malized to 50,000 fuel oil equivalent barrels per calendar day
474
-------
TABLE 1. COMMON FINANCIAL PARAMETERS
Financial Parameters Low Cost Case[26] High Cost Case[28]
11.5
Capital Charge Rate,
Percent
Debt/Equity Ratio
Discounted Cash Flow
Rate of Return on In-
vestment, Percent
Project Life, Yrs.
Construction Period, Yrs.
Investment Schedule,
%/Yr.
Plant Start Up Ratios
Debt Interest, Nominal
Rate, Percent
40/60
Not Available
20
4
9/25/36/30
50, 90, 100...
10
Investment Tax Credit, % 9
Depreciation Method Sum of Year's Digits
Tax Life, Yrs. 15
30
0/100
15
20
4
10/15/25/50
50/100
10
Sum of Year's Digits
13
475
-------
TABLE 2. PROCESS COST INPUTS AND OTHER
FACTORS COMMON TO ALL STUDIES
Cost Inputs and Other Factors
Product Yield
Coal
a) Bituminous
b) Subbituminous
c) Lignite
Operating Costs
a) Utilities
b) Working Capital Interest
c) Fuel Cost
Scaling Factors
a) Capital Costs
b) Labor Costs
c) Maintenance, Taxes,
Insurance, General
d) Coal, Catalysts and
Chemicals, Utilities,
Fuel, Natural Gas
Value
50,000 FOEB/CD
$27.50/ton
$17.00/ton
$10.00/ton
$0.035/kw-HR
6% of working
capital per year.
$35/bbl
0.75
0.20
Same percentage
of plant invest-
ment as specified
by each individ-
ual study.
Amount varies
directly propor-
tional to plant
size.
By-Product Credit
a) Sulfur
b) Ammonia
c) Phenol
Contingency factor
Inflation Rate
a) 1976
b) 1977
c) 1978
d) 1979
e) 1980
Real Cost Increases (%/year)
a) Fuel Oil
b) Natural Gas
c) Coal
$50/ton
$180/ton
$112.6/bbl
15%
5%
6%
7%
9%
9%
2%
2%
0%
476
-------
(FOEB/CD)(one FOB equals 5.9 mBtu, higher heating value). The costs
selected for bituminous, subbituminous and lignite coals are respec-
tively $27.50, $17.00, and $10.00 per ton. Because capital costs do
not usually vary in direct proportion to plant size, a scaling factor
(an exponent) is normally used to modify the ratio of plant sizes (by
yield). The scaling factor used here was 0.75, which is an average of
factors found from various studies.[29,31, 32,33] To adjust labor and
supervision costs a scaling factor of 0.2 was used.[26,32] The rest
of the operating costs were assumed to vary directly with plant size.
The inflation rate for adjusting the costs of studies to $1981 was
based on the Chemical Engineering plant cost index.
The product mix expected from each of the various synfuel pro-
cesses being investigated can be found in Table 3. In order to put
the discussion on costs into a more meaningful perspective, several
points should be kept in mind. First, indirect liquefaction processes
can yield a product mix which is either essentially 100 percent trans-
portation fuel or a 50-50 mix of transportation fuel and SNG. The
latter appears to be more efficient and economical for either methanol
or MTG-gasoline production, but the cost of producing essentially 100
percent transportation fuel will be used here since the nation's
energy shortfalls are primarily in the transportation area. Second,
the product mix from direct liquefaction processes depends largely on
the degree of refining applied. Each of the direct liquefaction pro-
cedures yields some SNG or LPG which can be sold without further pro-
cessing, while the remainder of the products in most cases must be
refined before marketing. This refining adds to the product's cost.
Third, the mixes reported in Table 3 were taken from available refin-
ing reports. The SRC-II study was based on maximizing gasoline pro-
duction while the EDS and H-Coal studies also considered No. 2 fuel
oil production. Fourth, none of the synfuel processes being examined
produce residual oil or diesel fuel. Residual oil could of course be
obtained by the direct liquefaction routes simply by applying less
refining. However, products from direct liquefaction plants appear to
be too high in aromatics to allow economical production of diesel fuel.
Turning once again to Table 3, it can be seen that capital costs
range from $2.04 billion to $3.3 billion. The methanol plants tend to
have the lowest capital costs ($2.0-2.5 billion), while that of the
EDS process is in the same range near the high end. Using the incre-
mental cost of the MTG process, a gasoline-from-coal plant would cost
between $2.6 billion and $3.1 billion. The H-Coal and SRC-II pro-
cesses are next at $3.3 billion. (The capital costs do not include
refinery costs since it is unlikely that new refineries would be
built.)
A product value approach was utilized to estimate costs for indi-
vidual products. This technique assumes that the future prices of
particular fuels will maintain a certain relationship, based on rela-
tive demand. All prices are normalized relative to a reference pro-
477
-------
TABLE 3. PRODUCT AND CAPITAL COSTS OF SELECTED
COAL LIQUEFACTION PROCESSES (1981 DOLLARS)
Refined
Product
Cost ($/mBtu)
Process
Direct Liquefaction
EDS (Bituminous)
H-Coal (Bituminous)
SRC-II (Bituminous)
Indirect
Liquefaction
Texaco (Bituminous)
Koppers (Bitum. )
Lurgi (Subbit.)
Modified Winkler
(Lignite)
Lurgi Mobil MTG
(Subbit.)
Product Mix
32.7% Reg. Gasoline
14.0% Prem. Gasoline
25.6% No. 2 Fuel Oil
9.6% LPG
18.1% SNG
50.7% Reg. Gasoline
11.0% Prem. Gasoline
20.1% No. 2 Fuel Oil
18.2% LPG
64.7% Gasoline
12.1% LPG
23.2% SNG
100% MeOH**
100% MeOH**
47.9% MeOH**
49.7% SNG
2.4% Gasoline
100% MeOH**
41.2% Reg. Gasoline
53.3% SNG
5.5% LPG
11.5%
CCR
10.11
10.87
8.29
7.78
8.09
8.41
9.04
6.90
6.48
9.87
7.60
7.90
5.90-
6.16
6.97
5.82
6.03
7.54
5.25
7.54
6.03
5.80
Capital
Cost*
30% (Billions
CCR of Dollars)
16.57
17.81
13.59
12.76
13.26
16.13
17.34
13.23
12.42
19.06
14.68
15.24
9.80-
10.00
11.73
10.02
10.55
13.19
9.12
13.19
10.55
10.16
2.50
3.30
3.30
2.06
2.51
2.32
2.04
2.92
Mobil MTG
Incremental Cost
85-90% Reg. Gasoline
10-15% LPG
1.72 3.17
0.6
* Capital costs are instantaneous costs. Capital costs do not
include refinery capital costs.
** MeOH = 95-98% methanol, 1-3% water, and the remainder higher
alcohols.
-------
duct, which here was chosen to be gasoline. In this report, a rela-
tionship between various fuels similar to that reported in the ICF
report was used and is as follows:
1. If the cost of unleaded regular gasoline is $G/mBtu,
2. The cost of No. 2 fuel oil is (0.82)(G)/mBtu, and
3. The cost of LPG is (0.77)(G)/mBtu.[29]
Since unleaded premium gasoline is produced in some cases (EDS
and H-Coal), a relationship between this fuel and regular gasoline is
also necessary. Unfortunately, a history of the relationship between
these two fuels was not readily available. The cost ratio of leaded
premium to leaded regular gasoline was used instead. This relation-
ship indicated a cost ratio of 1.075.[34] This product cost relation-
ship was then applied to premium and regular unleaded gasoline.
The cost for SNG was assumed to be (0.8)(G). This value was
obtained by averaging those for No. 2 fuel oil and LPG since SNG
should share markets with each, especially No. 2 fuel oil.
The product costs, along with capital costs discussed earlier,
are shown in Table 3. As can be seen, they follow a similar pattern
as capital costs, though not exactly. Speaking first of the low cost
scenario, methanol is the cheapest product, ranging from $5.25-^6.97
per million Btu (mBtu) for fully commercial gasifiers and i>5.90-$6.16
per mBtu for the near commercial Texaco gasifier. Gasoline via the
Mobil MTG process would be $1.72 per mBtu more, or $6.97-^7.69 per
mBtu using fully commercial gasifiers and $7.62-$7.84 per mBtu with
the Texaco gasifier. H-Coal gasoline costs slightly more at $8.41 per
mBtu, while SRC-II gasoline is projected to cost $9.87 per mBtu.
Finally, EDS gasoline is projected to cost the most of the automotive
products at $10.11 per mBtu.
A similar order holds for the higher cost scenario. In this
case, SRC-II has replaced EDS as the process yielding the highest cost
product. This is primarily due to the higher capital costs involved
for SRC-II. It should also be noted that the absolute difference
between methanol costs and the cost of gasoline from the other pro-
cesses increases because the capital cost of the methanol plant is
lower. The same is true for MTG gasoline in most cases. A large
change occurs in the difference between EDS and H-Coal process costs.
While the EDS costs were 20 percent higher using the low CCR, they are
less than 3 percent higher using the high CCR.
Using all the studies which are publicly available, it would
generally appear that the indirect coal liquefaction processes can
produce usable fuel cheaper than the direct liquefaction technologies.
479
-------
DISTRIBUTION COSTS
Since distribution systems already exist for gasoline, the econo-
mics in this area would, of course, favor the continued use of this
fuel over the introduction of methanol. In addition, gasoline also has
the advantage of possessing a higher energy density: 115,400 Btu/gal
for gasoline compared with 56,560 Btu/gal for methanol. Thus, because
transportation costs depend primarily on volume, gasoline would neces-
sarily be less expensive to transport per Btu.
The costs of distributing a fuel can most easily be divided into
three areas; 1) distribution from refinery or plantgate (if no refin-
ing is required) to the regional distributor, 2) distribution from the
regional distributor to the retailer, and 3) distribution by the
retailer (i.e., the gas station). These three aspects of distribution
will be discussed below.
More detail could of course be added to this analysis to improve
the resulting estimates but such information has not yet been assi-
milated. However, the general conclusions reached below should not
change substantially.
To simplify the presentation here, long-range distribution is
approximated by that of pipeline transport to a distance of roughly
650 miles.[29] It should be noted that if pipelines are needed to
connect coal fields (where synfuel plants are likely to be located)
with major markets, then the total costs will be roughly the same
whether the plant produces methanol or synthetic gasoline. This is
evident since the pipeline must be built in either case and the con-
struction and operating costs increase only slightly with a doubling
in size. Further, right-of-way and engineering costs should not
change at all with capacity in this range.
In the case of distributing methanol, the ,total amount of energy
distributed would only be about 80 percent that of gasoline due to
vehicle efficiency improvements which will be discussed later. How-
ever, a gallon of methanol only contains half the energy contained in
a gallon of gasoline, so 60 percent more volume of methanol would need
to be transported than that of gasoline.
To determine the potential range of the cost of transporting
methanol, two bracketing assumptions can be made. One, the cost of
transport per volume of fuel can be assumed to remain constant. Two,
total distribution costs can be assumed to remain constant. With the
first assumption, the estimated cost for gasoline transportation is
$0.22 per mBtu.[29] Methanol transportation would cost twice this
amount or $0.44 per mBtu. Using the second assumption, where total
costs remain constant, the cost for methanol would be $0.27 per mBtu,
since only 80 percent as much energy is being transported. Thus, the
cost of long-range distribution of methanol is $0.27-0.44 per mBtu.
480
-------
The costs involved with a switch to methanol will be more related
to the increase in volumetric capacity than differences in chemical
properties. Pipelines and pumps are almost entirely made from steel
or brass, with which methanol is compatible. Rubber seals on pumps may
need to be replaced with more durable rubber compounds, but this
should be a minor cost.
As mentioned earlier, the next step of local distribution con-
sists of storing fuel at the regional distributor and transporting it
to the retailers. This distribution is primarily done by tanker truck
and is estimated to cost just over $0.05 per gallon of gasoline, or
$0.46 per mBtu. If one conservatively assumes that the cost per vol-
ume remains constant, the $0.46 per mBtu cost for gasoline would
translate into a $0.92 per mBtu cost for methanol.
Here the cost of conversion to methanol should be very small,
even negligible. The only change required should be new rubber seals
and hoses, if they were not already made from a material compatible
with methanol.
The costs of retailing fuel (the last step) are more like that of
long-range distribution than local distribution. The costs of retail-
ing are primarily fixed costs, such as land or rent. Retailing dif-
fers from both long-range and local distribution, however, in that
fuel energy is the critical marketing factor, not volume.
Typical retailer mark-ups are estimated to be in the range of
$0.05-0.18 per gallon of gasoline.[35] However, since the lower
'mark-ups are usually associated with the high-volume stations, the
average mark-up per gallon of gasoline sold in the U.S. should be
somewhere between $0.09-0.11, or $0.76-0.95 per mBtu. For methanol,
the cost would lie between this range and 25 percent more since the
total amount of energy distributed would be 20 percent less. Thus,
the cost of retailing methanol would be $0.76-1.19 per mBtu.
In deriving these retail costs, no attempt was made to account
for any additional costs the retailer would bear when methanol is
first introduced. For example, he will have to make some monetary
allowance for the initial small volume of customers. The retailers in
some instances will also incur costs associated with installing new
tanks if the existing ones are incompatible or unavailable due to
large demands for the specific fuels they contain. The abovementioned
retailing costs should therefore be considered as long-term costs,
after the methanol market stabilizes.
The total cost of distributing methanol and gasoline can now be
calculated by simply combining the costs presented in the last three
sections. Methanol would cost $1.95-2.55 per mBtu to distribute;
gasoline would cost $1.44-1.63 per mBtu. Gasoline has a significant
advantage over methanol in terms of percentage (26-36 percent lower),
but the absolute difference is only $0.51-0.92 per mBtu.
481
-------
IN-USE COSTS
In order to determine in-use costs associated with methanol, it
is necessary to know its fuel efficiency characteristics. There is
general agreement among researchers that methanol is a more energy
efficient vehicle fuel than gasoline. There are at least two theore-
tical reasons why this is so. One, methanol's lower flame temperature
reduces the amount of heat transfer from the combustion chamber to the
vehicle coolant system. Two, its high heat of vaporization acts as an
internal coolant and reduces the mixture temperature during the com-
pression stroke. These characteristics are realized in experiments
without having to make any major design changes in current gasoline
engines. Studies have shown these inherent properties of methanol to
increase the energy efficiency of a passenger vehicle by 3 to 10 per-
cent with a middle range of about 5 percent.[9,12,13]
Other properties of neat methanol combustion allow even greater
efficiency improvements. Its wider flammability limits and higher
flame speeds relative to gasoline allow methanol to be combusted at
leaner conditions while still providing good engine performance. This
lean burning capability decreases the peak flame temperature even
further and allows more complete combustion, improving energy effi-
ciency. Early testing on a single-cylinder engine yielded estimated
energy efficiency improvements of 10 percent due to leaning of the
methanol mixture as compared to gasoline tests.[36j Subsequent vehi-
cle testing has shown relative efficiency improvements of lean meth-
anol combustion of 6 to 14 percent.[8,9] Given these results, it
would appear that methanol's lean burning capability yields approxi-
mately a 10 percent efficiency improvement over and above the 3-10
percent improvement mentioned above. Of course, stratified charge
engines have been developed to allow leaner combustion of gasoline as
well, and this efficiency advantage of methanol would be lessened with
respect to a stratified charge engine.
Methanol's higher octane number also allows the usage of higher
compression ratios with correspondingly higher thermal efficiencies.
Early single-cylinder testing have estimated the thermal energy effi-
ciency improvements of the higher compression ratios to be in the
range of 16 to 20 percent.[14,36] Unfortunately, little vehicle data
exist to confirm these figures, but it must be expected that improve-
ments of at least 10 to 15 percent are likely.
Adding up the possible improvements indicates that methanol
engines may well be 25 to 30 percent more energy efficient than their
gasoline counterparts when the methanol engine is designed specifi-
cally for methanol.
However, since such methanol engines are not available for mass
distribution today, this section will use a more conservative fuel
efficiency advantage for methanol engines over their gasoline counter-
482
-------
parts of 20 percent. Using a fuel economy of 30 miles per gallon for
the average gasoline-fueled vehicle, this average vehicle would
require about 0.0038 mBtu per mile to operate. A methanol-fueled
vehicle would be expected to use at least 20 percent less energy or
about 0.0030 mBtu per mile.
Using 12,000 miles per year and the average delivered fuel costs,
calculated by combining production and distribution costs, the annual
fuel savings relative to gasoline produced via indirect liquefaction
(Mobil MTG process) were determined (see Table 4). These savings
include two separate effects. One, they include the effect of dif-
ferences in at-the-pump fuel costs. Two, they also include the effect
of methanol engines being more fuel efficient than gasoline engines.
For consistency, all fuels were assumed to be derived from bituminous
coal.
Following this procedure and using the lowest fuel cost (based on
the low CCR) and the highest fuel cost (based on a 30 percent CCR),
methanol would produce a savings of $131-240 per year compared to
gasoline from the Mobil MTG process. Direct liquefaction gasoline
would cost an extra $36-410 per year over MTG gasoline, because of its
potentially higher at-the-pump cost.
To this fuel savings must be added any difference in engine or
vehicle cost. While a methanol-fueled diesel engine may be developed
with a fuel efficiency advantage comparable to that of a standard
diesel, the conservative 20 percent efficiency advantage over the
gasoline engine should be attainable with engines similar to the gaso-
line engine in terms of both design and cost. While a larger fuel
tank and a special cold start system may increase costs, savings
should be attained with respect to emission control, particularly if
NOx reduction catalysts are no longer needed and if base metal oxida-
tion catalysts can be used instead of platinum and paladium. Thus,
whether a methanol engine will cost more or less than a gasoline
engine in the long run is still an open question at this time. It
would be rather safe to project, however, that any potential extra
cost would not override the kind of fuel efficiency benefit described
earlier.
ECONOMICS SUMMARY
The results of the past three sections are shown in Table 4. As
can be seen when the results are combined, methanol compares favorably
to the other fuels. With respect to synthetic gasoline, methanol
appears to cost less at the plant gate. This is true whether the low
CCR is used or the high CCR. Higher distribution costs lower the dif-
ference, but even after distribution, methanol appears to still hold
some advantage. This advantage is $1.21- $2.25 per mBtu over MTG
gasoline and $2.00-$6.41 per mBtu over direct liquefaction gasoline.
For vehicles driven 12,000 miles per year and achieving 30 miles per
483
-------
TABLE 4. SYNTHETIC FUEL COSTS ($ per mBtu)*
Indirect Coal
Liquefaction
Methanol
Gasoline
Direct Coal
Liquefaction
Gasoline
Production
Plantgate
Cost
Distribution
Long-Range
Local
Retail
Cost at Pump
5.90-11.73
7.62-14.90
8.41-19.06
0.27-0.44
0.92
0.76-1.19
0.22
0.46
0.76-0.95
0.22
0.46
0.76-0.95
7.85-14.28
9.06-16.53
9.85-20.69
ANNUAL FUEL SAVINGS (RELATIVE TO GASOLINE
AT $9.06-16.53 per MBtu)**
$131-240 $0 $-(36-189)
ADDED ENGINE COST OVER GASOLINE ENGINE
000
* Range of plantgate cost is the lowest cost using the low CCR
and the highest cost using the high CCR for bituminous feed-
stocks.
** Includes effect of increased engine efficiences and dif-
ferences in at-the-pump fuel costs.
484
-------
gallon (gasoline), methanol would save $131-$240 per year over MTG
gasoline and $167-$429 per year over direct liquefaction gasoline if
allowances are made for the increased efficiency of methanol engines.
Without including the improved engine efficiency, annual savings would
be $55-$103 relative to MTG gasoline and $91-292 over direct liquefac-
tion gasoline.
It should be stated that no comparison" was made between methanol
and diesel fuel since none of the coal conversion processes examined
produces diesel fuel of sufficient quality for today's diesel
engines. All of these economic results are of course subject to the
qualifications which have been stated previously; the primary ones
being that the detail of the engineering designs could not be compared
across processes, and that cost estimates reflect different points of
development for different synfuels.
CONCLUDING STATEMENT
Looking back over the topics addressed in this paper, it can be
concluded that at this point in time methanol appears to have environ-
mental and economic advantages over other synthetic transportation
fuels derived from coal. The ultimate viability of this conclusion
depends on a number of key events or findings. One, a cost-competi-
tive methanol engine must be able to meet the driveability needs of
most of the U.S. (e.g., cold-starting in nearly all climates). Two,
aldehyde emissions must be controllable at low cost. Three, no other
unique and uncontrollable environmental problems of methanol use or
production are discovered. Four, the production and distribution cost
comparisons made here must hold up against future scrutiny.
The probability of these events occurring can only be estimated
by a review of the support for each presented in this study. At this
time, we believe the evidence available suggests that the benefits of
methanol outweigh its costs.
485
-------
REFERENCES
1. O'Leary, J.R. and G.C. Rappe, "Scale-Up of an SRC Deashing
Process," Chemical Engineering Progress, Vol. 77, No. 5, May 1981,
pp. 67-72.
2. "EDS Coal Liquefaction Process Development, Phase V, EDS
Commercial Plant Study Design Update/Illinois Coal," FE-2893- 61,
March 1981.
3. Schmid, B.K. and D.H. Jackson, "The SRC-II Process,"
(Pittsburg and Midway Coal Mining) presented at discussion meeting on
New Coal Chemistry, Organized by the Royal Society, London, England,
May 21-22, 1980.
4. "Catalogue of Synthetic Fuels Projects in the U.S.," Energy
Policy Division, U.S. EPA, April 1981.
5. D'Elisen, Prof. P.N., "Biological Effects of Methanol
Spills into Marine, Estuarine, and Freshwater Habitats," Presented at
the International Symposium on Alcohol Fuel Technology, Methanol and
Ethanol, Wolfsburg, FRG, November 21-23, 1977, CONF - 771175.
6. "Methanol Fuels in Automobiles — Experiences at Volks-
wagenwerk AG and Conclusions for Europe," Dr. Ing. W. Bernhardt,
Volkswagenwerk AG, Wolfsburg, Germany.
7. "B-39, Use of Glow-Plugs in Order to Obtain Multifuel
Capability of Diesel Engines," Institute Maua de Tecnologia, Fourth
International Symposium on Alcohol Fuels Technology, October 5-8, 1980.
8. "Methanol as a Motor Fuel or a Gasoline Blending Component,
"J.C. Ingamells and R.H. Lindquist, SAE 750123.
9. "Vehicle Evaluation of Neat Methanol - Compromises Among
Exhaust Emissions, Fuel Economy and Driveability," Norman D. Brinkman,
Energy Research, Vol. 3, 243-274, 1979.
10. "The Influence of Engine Parameters on the Aldehyde Emis-
sions of a Methanol Operated Four-Stroke Otto Cycle Engine," Franz F.
Pischinger and Klaus Kramer, Paper 11-25, Third International Sympo-
sium on Alcohol Fuels Technology, May 29-31, 1979, published by DOE in
April 1980.
11. "Research and Development - Alcohol Fuel Usage in Auto-
mobiles," University of Santa Clara, DOE Automotive Technology
Development Contractor Coordination Meeting, November 13, 1980.
12. "A Motor Vehicle Powerplant for Ethanol and Methanol Opera-
tion," H. Menrad, Paper II-26, Third International Symposium on
486
-------
Alcohol Fuels Technology, May 29-31, 1979, published by DOE In April
1980.
13. "Development of a Pure Methanol Fuel Car," Holger Menrad,
Wenpo Lee, and Winfried Bernhardt, SAE 770790.
14. "Effect of Compression Ratio on Exhaust Emissions and Per-
formance of a Methanol-Fueled Single-Cylinder Engine," Norman D.
Brinkman, SAE 770791.
15. "A New Way of Direct Injection of Methanol in a Diesel
Engine," Franz F. Pischinger and Cornells Havenith, Paper 11-28, Third
International Symposium on Alcohol Fuels Technology, May 29-31, 1979,
published by DOE in April 1980.
16. "Alternative Diesel Engine Fuels: An Experimental Investi-
gation of Methanol, Ethanol, Methane, and Ammonia in a D.I. Diesel
Engine with Pilot Injection," Klaus Bro and Peter Sunn Pedersen, SAE
770794.
17. "Alcohols in Diesel Engines - A Review," Henry Adelman,
SAE790956.
18. "The Utilization of Alcohol in Light-Duty Diesel Engines,"
Ricardo Consulting Engineers, Ltd., for EPA, May 28, 1981, EPA-460/3-
81-010.
19. "The Utilization of Different Fuels in a Diesel Engine with
Two Separate Injection Systems," P.S. Berg, E. Holmer, and B.I.
Bertilsson, Paper 11-29, Third Symposium on Alcohol Fuels Technology,
May 29-31, 1979, published by DOE in April 1980.
20. Hilden, David L. and Fred B. Parks, " A single-Cylinder
Engine Study of Methanol Fuel-Emphasis on Organic Emissions," SAE
760378.
21. "Driving Cycle Economy, Emissions, and Photochemical Reac-
tivity Using Alcohol Fuels and Gasoline," Richard Bechtold and J.
Barrett Pullman, SAE 800260.
22. Browning, L. H. and R. K. Pefley, "An Analytical Study of
Aldehyde Formation During the Exhaust Smoke of a MethanolFueled SI
Engine," Paper B-62, Fourth International Symposium on Alcohol Fuels
Technology, Oct. 5-8, 1980.
23. Baisley, W.H. and C.F. Edwards,"Emission and Wear Charac-
teristics of an Alcohol Fueled Fleet Using Feedback Carburetion and
Three-Way Catalysts, B-61, Fourth International Symposium on Alcohol
Fuels Technology, Oct. 5-8, 1980.
487
-------
24. "Alcohol Engine Emissions - Emphasis on Unregulated Com-
pounds," M. Matsuno et al. , Paper 111-64, Third International Sympo-
sium on Alcohol Fuels Technology, May 29-31, 1979, published by DOE in
April 1980.
25. "Methanol and Formaldehyde Kinetics in the Exhaust System
of a Methanol Fueled Spark Ignition Engine," Kenichito and Toshiaki
Yaro, Paper B-65, Fourth International Symposium on Alcohol Fuels
Technology, Oct 5-8, 1980.
26. K.A. Rogers, R.F. Hill, "Coal Conversion Comparison,"
ESCOE, DOE FE-2468-51, July 1979, pg. 59.
27- "Methanol From Coal, An Adaptation From the Past," E.E.
Bailey, (Davy McKee), presented at The Sixth Annual International Con-
ference; Coal Gasification, Liquefaction and Conversion to Electri-
city, University of Pittsburgh, 1979.
28. Sullivan and Frankin, "Refining and Upgrading of Synfuels
from Coal and Oil Shales by Advanced Catalytic Processes," March 1980,
Chevron Research Co., for DOE, FE-2315-47.
29. "Methanol from Coal: Prospects and Performance as a Fuel
and a Feedstock," ICF, Inc., for the National Alcohol Fuels Commis-
sion, December 1980.
30. "Economic Feasibility Study, Fuel Grade Methanol from Coal
for Office of Commercialization of the Energy Research and Development
Administration," McGeorge, Arthur, Dupont Company, for U.S. ERDA
TID-27606.
31. "Methanol Use Options Study," (Draft) DHR, Inc. for DOE,
December, 1980; Contract No. DE-ACOI-79 PE-70027.
32. Peters, Max S. and Timmerhaus, Klaus D., Plant Design and
Economics for Chemical Engineers, McGraw-Hill Co., 2nd Ed., 1968.
33. Kermode, R. I., A. F. Micholson, D. F. Holmes, and M. E.
Jones, Jr., "The Potential for Methanol from Coal: Kentucky's Per-
spective on Costs and Markets," Div. of Technology Assessment, Ken-
tucky Center for Energy Research, Lexington, KY, March 1979.
34. Monthly Energy Review. U.S. DOE, DOE/EIA-0035 (81/04),
April 1981.
35. Ayling, John, personal communication, 2/11/81, Lundberg
Survey Inc., North Hollywood, California.
36. Most, W. J., and J. P. Longwell, "Single-Cylinder Engine
Evaluation of Methanol Improved Energy Economy and Reduced NOx," SAE
750119.
488
-------
PROJECT SUMMARY
A COMPENDIUM OF SYNFUEL END USE TESTING PROGRAMSt
By:
Masood Ghassemi, Sandra Quinlivan, and Michael Haro
Environmental Division
Energy Development Group of TRW, Inc.
One Space Park, Redondo Beach, CA 90278
ABSTRACT
A "Compendium of Synfuel End Use Testing Programs", which provides
information on major recently-completed, current and planned synfuel end use
testing projects, has been developed. The compendium is intended to promote
flow of information among various synfuel testing programs, thereby reducing
chances for duplication of effort and enabling design and implementation of
cost-effective and systematic approaches to the collection of appropriate
environmental data in conjunction with ongoing and planned performance
testing projects. It is EPA's intention to update this compendium to include
results from current and future testing programs.
Projects described in the compendium involve testing of shale-derived
fuels, SRC-II middle distillates, EDS fuel oils, H-coal liquids and methanol-
indolene mixtures in various equipment such as utility boilers, steam genera-
tors, diesel engines (lab-scale and full-scale), auto engines, and various
other combustors. Published reports on various testing efforts and discus-
sions with test sponsors/contractors are the sources of data for the com-
pendium.
Based on the data presented in this compendium, the thrust of the synfuel
testing program which has been carried out to date has been to assess equip-
ment performance and fuel handling characteristics. Where some emissions
monitoring has been conducted, such efforts have been limited in scope and
have primarily emphasized measurement of criteria pollutants (NOX, SOX, par-
ticulates, etc.). Essentially no data have been collected on emissions of
non-criteria/non-regulated pollutants.
INTRODUCTION AND OBJECTIVES OF THE COMPENDIUM
A recently-completed synfuel utilization background study identified a
great need for better coordination among various agencies involved in synfuel
*M. Ghassemi and R. Iyer, "Environmental Aspects of Synfuel Utilization", EPA
Report No. EPA-600/7-81-025, March 1981. (Note: For a summary of this
report, see Environmental Science and Technology, Volume 15, No. 8, August
1981, pp. 866-873.)
489
-------
end use testing programs so as to promote more systematic approaches to the
collection of environmental data in connection with such programs and to
reduce chances for duplication of effort. Per recommendation of the back-
ground study, a compendium of synfuel end use testing programs has been
developed as an information source on major recently completed, ongoing, and
planned synfuel end use testing programs. The dissemination of the document
among agencies/organizations engaged in various aspects of synfuel produc-
tion, testing, utilization, and regulation, coupled with holding regular
symposia/workshops on synfuel utilization and end use testing, should greatly
enhance coordination and flow of information among various programs and, in
the long run, contribute to the goal of more rapid establishment of an envir-
onmentally acceptable commercial synfuel industry in the U.S. EPA plans to
periodically update this compendium to include results from current and future
testing programs.
DATA BASE USED AND DATA PRESENTATION
Information presented on the synfuel testing programs has been obtained
from published documents and via telephone calls and/or interviews with
organizations involved in the testing programs. The key individuals/agencies
providing most of the reports and data used in the compendium are listed in
Table 1.
A separate "data sheet" has been devoted to each project covered in this
compendium to permit periodic updating of the document to include additional
projects and incorporation of further results from ongoing studies. The data
sheets are grouped into four categories, covering projects for which the key
sponsors/participants are Electric Power Research Institute (EPRI), Depart-
ment of Defense (DOD). Department of Energy (DOE), and Miscellaneous agencies
(e.g., EPA). Data sheets are presented for a total of 44 projects, of which
7 are in the EPRI-sponsored category, 15 in the DOD category, 13 in the DOE
category, and 9 in the Miscellaneous category.
Where data have been available, each data sheet provides the following
information on a test project: type of fuel tested (both synfuel and the
reference petrofuel, where indicated), test equipment used, test site, test
objectives, sponsoring agency, contractor, test conditions, environmental
monitoring, project status, summary of results, and references (where a report
or reports have been published on a project).
A summary of the data contained in the data sheets is presented in
Table 2. Tables 3 and 4 present brief descriptions of some of the recently
initiated and tentatively planned synfuel testing programs. Two examples of
the data sheets are presented.
OVERVIEW OF SYNFUEL TESTING PROGRAMS
Based on the data presented in the test program data sheets and summar-
ized in Table 2, and on the discussions which have been held with a number of
synfuel developers, trade associations and potential major users of synfuels,
490
-------
TABLE 1. LIST OF ORGANIZATIONS/INDIVIDUALS PROVIDING INFORMATION
USED IN THE DEVELOPMENT OF THE COMPENDIUM
Electric Power Research Institute
3412 Hillvlew Drive
Palo Alto. CA 94303
Mr. Al Dolbec
Air Force Wright Aeronautical
Laboratory, Aero Propulsion Laboratory
Wright-Patterson AFB/POSF
Dayton, Ohio 45433
Mr. Charles Delaney
Navy Air Propulsion Center
P. 0. Box 7176
Trenton, NJ 08628
Mr. C. J. Nowack
David W. Taylor Naval Ship R&D Center
Code 2705
Annapolis, MD 21402
Mr. Carl A. Hershner
Army Mobility Equipment Research and
Command Center - Attn: DRDME-GL
Ft. Belvoir, VA 22060
Mr. F. Schaekel
U.S. A1r Force HQ AFESC/RDV
Tyndall AFB
Tyndall, FL 32403
Major J. Tom Slankas
DOE, Bartlesville Energy
Technology Center
P. 0. Box 1398
Bartlesville, OK 74003
Mr. Dan Gurney
DOE, Conservation and Solar Energy Div.
Washington, D.C.
Mr. Gene Ecklund
DOE, Office of Coal Utilization
Fossil Energy Research Center
Germantown, MD
Mr. John Fairbanks
OOE, Laramie Energy Technology Center
P. 0. Box 3395
Laramie, WY 82071
Or. R. Poulson
DOE, Pittsburgh Energy Technology
Center, Analytical Chemistry Division
Pittsburgh, CA
Mr. Curt White
National Aeronautics and Space
Administration, Lewis Research Center
21000 Brook Park Drive
Cleveland, OH 44135
Mr. Rick Niedzwiecki
EPA, Special Studies Branch
Industrial Environmental Research Lab.
Research Triangle Park, N.C. 27711
Mr. G. Blair Martin
EPA, Motor Vehicle Emission Laboratory
2625 Plymouth Road
Ann Arbor, MI
Mr. Robert Garbe
EPA, Combustion Research Branch
Industrial Environmental Research Lab.
Research Triangle Park, N.C. 27711
Mr. G. Blair Martin
EPA, Office of Environmental
Engineering and Technology
Industrial Environmental Research Lab.
Research Triangle Park, N.C. 27711
Mr. W. S. Lanier
EPA, Mobile Sources Laboratory
Research Triangle Park, N.C.
Mr. Frank Black
Southwest Research Institute
Automotive Research Division
6220 Culebra Road
San Antonio, TX 78284
Mr. Charles T. Hare
Southwest Research Institute
Mobile Energy Division
6220 Culebra Road
San Antonio, TX 78284
Mr. John A. Russell
U.S. Department of Transportation
Systems Center
Kendall Square
Cambridge, MA 02142
Mr. Joe Sturm
U.S. Department of Energy and
Coordinating Research Council
Atlanta, GA
Mr. Al Zingle
Carson Associates for
Bank of America
4117 Robertson Boulevard
Alexandria, VA 22309
Mr. Gavin McGurdy
Energy and Environmental Research
Corporation
8001 Irvine Boulevard
Santa Ana, CA 92705
Mr. Dave Pershing
Ford Motor Company
Scientific Research Laboratory
Dearborn, MI
Vulcan Cincinnati, Inc.
Cincinnati, OH
Mr. R. W. Duhl
-------
TABLE 2. SYNFUELS-COMBUSTION SYSTEM COMBINATIONS TESTED AND EMISSIONS MONITORED
Test No. Agency Synfuel
1 EPRI SRC-II fuel oil
2 EPRI SRC-II fuel oil
H-Coal
EDS oil
3 EPRI SRC-II fuel oil
4 EPRI SRC-II, H-Coal
Reference Fuel
No. 6 fuel oil
No. 6 and No. 2
fuel oils
No. 2 and No. 5
fuel oil
No. 2 diesel
fuel
Combustion System
Tangentially-fired
utility boiler
Scaled-down
utility boiler
Babcock & Wilcox
package boiler
Three catalytic
reactors
Emissions Monitored
NOX, CO, THC, S03,
POM, particulates,
particle size, par-
ticulate composition
NO, CO?, CO, S02,
SOs, THC, smoke,
particulates, par-
ticle size
NOX, CO, C02, S02,
hydrocarbons, 02,
and dust
N0x and CO
General Conclusions
t No adverse boiler performance effects
with SRC-II fuel.
• HOX emissions nominally 70% higher than
No. 6 fuel.
• Higher fuel nitrogen content of SRC-II
fuels produced higher NO emissions than
reference fuels.
t NO emissions from H-Coal and EOS liquids
were lower than SRC-II.
t No unique differences in combustion or
emission characteristics of SRC-II fuel
blends.
i NOx emissions consistent with fuel nitro-
gen content.
• Combustion performance of SRC-II fuel oil
was similar to No. 2 and No. 5 fuel oils.
• Coal-derived liquids can be burned cata-
tytically but SRC-II, and to a lesser
EPRI
EPRI
Hydrogenated
shale oil and
various liquid
fuels for SRC-I,
H-Coal, EDS,
and SRC-II
Solvent refined
coal
No. 2 distillate
fuel
Full-scale and
sub-scale turbine
combustors
NOX, CO, UHC, par-
ticulates, and
smoke
Bituminous coal
Utility boiler
EPRI
Jet-A fuel,
natural gas,
methanol
Two utility gas
turbines
NOX, S02, C02, par-
ticulates, particu-
late composition
NOX, CO, S02, THC,
POM, sulfates, par-
ticulates, aldehydes,
opacity
degree H-Coal, appeared to degrade reactor
performance significantly as evidenced by
higher CO emissions.
• NOX emissions were
nitrogen content.
• A selected number of coal liquids and
shale oil fuels can be used in current
turbines.
« Emission levels of CO, UHC, and particu-
lates for synfuels were about the same as
for No. 2 fuel - not significant.
• Significant quantities of F8N are con-
verted to NOX causing emissions higher
than EPA limits.
• The boiler stayed much cleaner with SRC
than with coal, producing an equivalent
boiler efficiency as coal at full load.
• The quantity of SRC flyash was 10 to 15"
of that of coal flyash with no bottom ash
accumulation from SRC.
• Particulates,S02 and NOX emissions from
SRC were all under EPA limits.
t Methanol is a suitable fuel for gas tur-
bines; turbine performance and NOX and
particulate emissions are improved over
the other fuels.
(Continued)
-------
TABLE 2. (Continued)
Test No. Agency
Synfuel
Reference Fuel
Combustion System Emissions Monitored
General Conclusions
8 DOD Shale-derived
JP-5 and blends
with petroleum
JP-5
Petroleum JP-5 DOD helicopter NOX, CO, C02, and
engine: Allison THC
T63-A-5A turbo-
shaft
• NOX emissions increased with increasing
fuel nitrogen content; conversion effi-
ciency was about 45%.
• No significant effects were noted on en-
gine performance or CO, C0;>, and THC
emissions due to the presence of high
levels of fuel bound nitrogen.
DOG
Shale-derived DFM
10 DOD
JP-5 from oil
shale, coal, and
tar sands
UJ
11 DOD
12 DOD
13-15 DOD
Shale fuel oil
Shale-derived
diesel fuel
16 DOO
Oil shale-derived
JP-5 fuel
Petroleum diesel
fuel
(MIL-F-16884G)
Jet-A. JP-5,
diesel marine
fuel, leaded
gasoline, and
blends of the
above
Petroleum diesel
fuel marine
(DFM)
Petroleum
distillate
Shale-derived DFM Petroleum DFM
Petroleum-
derived JP-5
fuel
U.S. Navy LM2500
turbine engine
Two high tempera-
ture/pressure
research combustors
Steam generator
diesel engine
Lab-scale diesel
engine
3 different types
of prototype steam
generators
DOD helicopter
engine: Allison
T62-A-5A turbo-
shaft
NOX, CO, THC, and
smoke
NOX, CO, UHC, and
smoke
Particulates and
particulate compo-
sition
NOX, THC, and smoke
NOX, S02, CO, C02,
THC, 02, and smoke
NOX, CO, and THC
• Combustor and engine operating character-
istics were identical when using marine
diesel or DFM shale oil; thus, DFM shale
oil would be suitable for use in LM2500
engines.
• NOX emissions followed fuel nitrogen con-
tent; CO and THC levels were essentially
the same for both fuels.
• In all performance areas, the synfuels
correlated in the same manner as petro-
leum-derived fuels except for NOX emis-
sions from the shale oil fuel.
• Smoke formation was dependent on hydrogen
content; combustion efficiency, CO, and
UHC depend more on higher boiling point
components than fuel viscosity.
• No significant differences between parti-
culate emission products measured in the
study from the combustion of DFM or shale
fuel oil.
• There was no significant difference in
performance or emissions with the shale-
derived fuel.
• There were no significant differences in
measured pollutant emissions resulting
from the combustion of petroleum DFM or
shale-derived DFM on the CVA-60, DDG-15,
and the FF-1040 boilers. In each case,
S02, NOX, and smoke were below levels set
by EPA.
• Performance, CO, and THC emissions were
equivalent for both fuels.
• NOX emissions followed fuel nitrogen con-
tent.
(Continued)
-------
TABLE 2. (Continued)
Test No.
17
18
19-21
22
23
24
25
26
Agency
DOD
DOD
DOD
DOD
DOE
DOE
DOE
DOE
Synfuel
Unifined kerosene
derived from tar
sands
Distillate, avia-
tion, turbine, and
diesel fuels de-
rived from coal, tar
sands and oil shale
*
*
SRC-II middle
distillate
SRC-II middle
distillate
SRC-tl middle
distillate
SRC-II middle
distillate
Reference Fuel
Petroleum-
derived JP-5
fuel
Various petro-
leum-derived
fuels
13 petroleum de-
rived fuels: JP-4,
JP-8, diesel No.
2 & various blends
12 petroleum-
derived fuels:
JP-4, JP-8, and
various blends
Low quality resi-
dual oil, and
petroleum refe-
rence distil late
fuel
Petroleum
distillate
Low qual ity
residual oil and
distillate fuel
Low qual ity
residual oil,
petroleum refe-
rence distillate
oil , and natural
gas
Combustion System Emissions Monitored
DOD helicopter NOX, CO, and UHC
engine: Allison
T63-A-5A turbo-
shaft
Wide variety of Various pollutants
Army power-plant
systems
General Electric NOX, CO, UHC, and
F101 turbofan, J79- smoke
17C turbojet, and
J79 turbojet engines
TF41 turbofan com- NOX, CO, UHC, and
bustor smoke
Combustor sized NOX, CO, COj, THC,
for use with in- and smoke
dustrial gas
turbine
Various combustor NOX, smoke
concepts
Seven combustors NOX, smoke, CO, un-
of varying .designs burned HC
for use in utility
gas turbine engines
Combustors for use NOX, CO, THC, smoke
in utility gas
turbine engines
General Conclusions
• Unifined Kerosene was a satisfactory sub-
stitute for petroleum JP-5 fuel.
• NOX emissions were slightly higher when
using unifined kerosene than with JP-5.
• Product quality of many synfuels tested
and other results are described in indivi-
dual abstracts.
• In all three engines, fuel hydrogen content
strongly affected smoke and NOX emissions.
NOx emissions were also highly dependent
upon combustor operating conditions.
• All pollutant emissions measured were
highly dependent upon operating condi-
tions. CO and smoke levels were also
strongly affected by hydrogen and aroma-
tic content of fuels.
• The combustor was able to achieve low NOX
with all fuels.
• CO and smoke varied directly with rich
zone equivalence ratio and inversely with
lean zone equivalence ratio.
t Values of NOX were reduced for the smaller
diameter quench zone and increased for
larger diameter quench zone.
• Rich-lean burn stage combustion system can
meet EPA emission standards.
• A lean-lean combustor has potential for
achieving ultra-low NOX emissions with
distillate, residual or other fuels ccn-
taining up to 0.25% (wt.) fuel nitrogen.
CO and smoke met program goals from this
combustor also.
• Lean-lean combustor NOX emission levels
were higher than emission goals using SRC-
II fuel. CO emissions remained low using
SRC-II fuel, while no smoke was detectable
and UHC levels were negligible throughout
these tests.
• Rich-lean combustor NOX emissions appeared
to reach a minimum below the NOX emission
goal for rich primary zone condition.
(Continued)
-------
TABLE 2. (Continued)
Test No. Agency
Synfuel
Reference Fuel
Combustion System Emissions Monitored
General Conclusions
DOt
SRC-11 middle
distillate
Low quality
residual oil.
petroleum refe-
rence distillate
oil
Experimental con-
bus tor for use
with utility gas
turbine engines
NQX, CO. UHC, snoke
SRC-11 middle and
heavy distillate.
fuel oils & three
blends of the
above
No. 2 and No. 6
petroleum-based
fuel oils
A 2Q-hp Johnston,
fire-tube boiler
NOV SO?, CO, HC
and polynuclear
aromatic hydro-
carbons
• Five coitustors have been found adequate
for further development: rich-lean diffu-
sion flame venturl quench, burner ceramic
lined pipe lean burner, nultlannular swirl
burner, Rolls-Royce combustor, and lean
catalytic combustor. These meet NOX
emission limits set by EPA with petroleum
distillate and/or residual oils.
• SRC-fl fuel NOX emissions Here close to
meeting CPA limits In only two combustors:
rich-lean diffusion and ceramic lined pipe
lean burners.
• The levels of NOX and SO? produced were pro-
portional to the amount of nitrogen and
Sulfur In the fuel.
• There appear to be two sources of trace or-
ganlcs 1n the exhaust gases: small amounts
of the fuel itself not burned during combus-
tion, and the products of combustion. For
the petroleum fuels, n-alkanes and polynuclear
aromatic hydrocarbons are seen In the exhaust
gas; for the SRC-II fuels, the alkanes are
absent or present at very low levels, and
polynuclear aromatic hydrocarbons not seen in
the petroleum exhaust gases are present.
29 DOE *
30 DOE *
31 DOE *
32-34 DOE *
35 OOE *
36 Vulcan *
Cincinnati
37 Ford *
Motor
Co.
Indolene and 101
methanol/901
Indolene
Unleaded gasoline
and methane I/
Indolene mixtures
10% methanol/901
gasoline blends
Ethanol.
methanol, and
gasoline blends
Indolene. indo-
lene/methanol
blends and
ethanol/indolene
blends
No. 6 residual
oil, natural gas,
and met ha no I
He t ha no 1,
Indonele, and
blends
Two light duty
vehicles
Auto engines (10)
Auto engines (7)
Fleet vehicles
Pontiac 4-cylinder
modified engine
Small scale boiler
test stand and a
49 HW utility
hni lar-
DOI ICr
Ford 400 CID engine
and 1975 Ford LTD
with 400 CID engine
Evaporative emissions
(hydrocarbons and
methanol )
NOX, CO. THC. alde-
hydes, and methanol
NOX. CO, and eva-
porative emissions
(HC and methanol )
Evaporative and
tailpipe hydrocarbon
emissions
Total aldehydes and
specific organlcs
NO., CO. and
aldehydes
Total hydrocarbons
and specific
organlcs
• Using methanol 101 blend Increased evapo-
rative emissions by 1301 for short term
use and 2201 for long term use.
• Aldehyde, methanol, and hydrocarbon emis-
sions Increased with higher concentration
of methanol in the fuel.
• CO was reduced by the addition of methanol
to the base fuel.
• Data show consistent reduction In CO
emissions with use of UK t Hanoi blends.
emissions with methanol blends.
• 751 increase in evaporative emissions with
methanol blends over a straight gasoline.
• fitissions were lower for vehicles fueled
with gas oho 1 but data was inadequate to
conclude a significant difference.
• Total aldehydes Increased 251 in going from
Indolene to ethanol/lndolene and methanol/
indolene blends.
• Formaldehyde Is the largest component of
the total aldehydes (up to 90 mole percent
of the total).
• In the utility boiler, methanol NOX levels
were 7-141 of those measured during resi-
dual oil combustion.
than 100 ppm and generally less than those
observed for the residual oil.
t Aldehyde emissions during methanol com-
bustion were generally less than 1 ppm.
• Methanol/lndolene blends gave significantly
higher hydrocarbon and aroodttc emissions
than Indolene without a catalyst, but only
slightly higher emissions with a catalyst.
-------
TABLE 2. (Continued)
Test No. Agency
Synfuel
Reference Fuel
Combustion System Emissions Monitored
General Conclusions
38
Shale-derived DFM No. 2 dlesel
fuel
VW Rabbit engine NO , CO. THC, parti-
cutates, Ames test
on participates
39
40
Bank of
America
EPA
42
EPA
EPA
*
Shale-derived DFM
SRC- II middle
distillate fuel
oil and shale-
derived residual
oil
*
Hethanol/gasol Ine
blends
No. 2 fuel, and
No. 2 fuel with
0.51 nitrogen
No. 2 fuel oil
and Indonesian/
Malaysian
residual oil
Residual and
distillate oils.
natural gas.
propane, Isopro-
panol, methano)
Fleet vehicles
Two configurations
of a ful l-scale
prototype (25-HW
engine-size) gas
turbine combustor
utilizing a Rich
burn/Quick Quench
combustor concept
Prototype full-scale
(25-HW engine-size)
Rich Burn/Quick
Quench gas turbine
with two combustor
configurations
Experimental wall
furnace and proto-
type Industrial
boiler
NO, CO, unburned
hydrocarbons
NO,. CO, unburned
hydrocarbons
NOX, CO, unburned
hydrocarbon, and
smoke
NO,, NO. CO, HC.
and aldehydes.
43
EPA
EPA
No. 5 residua)
oil, natural gas.
and methanol
Indolene and
ethanol blends
Industrial water- NO
tube and fire-tube
boilers
Two light duty
vehicles
NOw, CO, THC.
ethanol, and evapo-
rative emissions
t HC and CO emissions were found to be lower
and NO- levels higher for the shale-
derived fuel as compared to the petroleum-
derived fuel. Particulate emissions were
similar for both fuels.
• Mutagenlc activity of the organlcs from
the participate matter was similar Tor the
two fuels.
• Blends of 2 to IB: methanol decrease emis-
sions of CO and unburned hydrocarbons and
result in Improved mileage In new cars.
• Certain blends result in operating cost
decreases of U/mlle.
• Both combustor configurations met program
emissions goals using both reference
fuels and synfuel.
• Unburned HC emissions from one combustor
ranged from 0.9 to 7.3 ppmv for No. 2
fuel; 1.1 to 21.8 ppm for No. 2 fuel
with 0.5% nitrogen; and 1.3 to 15.3 ppmv
for shale-derived OFM at 15* 02-
• All emissions exhaust goals met.
• Relationship demonstrated between primary
zone residence time and attainable NOH
emission concentrations.
• NO emission levels for the five fuels were
as follows: distillate oil •> propane >
Isopropanol • alcohol mixture > methanol.
• Although there was considerable scatter In
the data, aldehyde concentrations were
around 10 ppm for methanol.
• NO emissions for all fuels decreased with
Increasing fraction of flue gas reclrcula-
tion.
• CO and hydrocarbon emissions were always
below 50 ppm and smoke was not observed
for any fuel.
• Flue gas reclrculatlon was capable of re-
ducing NOX emissions during methanol com-
bustion.
• Methano] NOX emissions were significantly
lower than during restdual oil combustion
and were also less than during natural
gas combustion.
• The addition of ethanol to Indolene re-
duced tailpipe emissions of THC and CO.
but Increased NOX.
• Use of gasohol Increased evaporative
emissions substantially.
Because of the unavailability of syflfuels, the fuels used in some of these programs were not "true" synfuels (e.g., metnanol-derived from natural
gas was used Instead of coal-derived melhanol). These studies, however, are included In this report because they were conducted to show what
night be expected from the combustion of actual synfuels In the Indicated combustion systems.
-------
TABLE 3. ON-GOING SYNFUEL TESTING PROGRAMS
Sponsoring Agency
EPA. Motor Vehicle
Emission Laboratory
Test Fuels
Shale-derived diesel fuel
and SRC- 11 fuel versus
National Average Baseline
Diesel Fuel.
Tin* Period
1981 to —
Project Description
Volkswagen Rabbit dieset engine testing. Lmissions
monitored to include part iculdtes. NOX, CO/CO^.
hydrocarbons, and aldehydes.
Hob1l-M gasoline.
DOE. Bartlesville
Energy Technology Center;
Contractor/test site:
A. General Electric.
Erie, PA
SHC-11 middle distillate
and oil shale distillate
Transaraerlca
Del aval.
Oakland, CA
SRC-II middle distillate
A.D. Little
Be Jolt. Wl
Energy and Envi-
ronmental Research
Springfield, OH
SRC-11 middle distillate
Shale-derived distillate
oil and Exxon Donor Solvent
coal-derived liquids
1981 Oldsmobile 350 and other engines testing. Emissions
monitored include particulates, NOX, CO/CO?, and
hydrocarbons.
1981 Testing of r,E £01-8. 8-cylfnder "V" configuration,
6344 cu. In. standing diesel engine for electric
power, rail and marine applications. Parameters
being evaluated include: starting ability, injec-
tion timing, fuel rate variation effects and inter-
nal engine temperatures. Emissions monitored include
0-., C0/C02, HO,. SO?. HC. H?S04 and participates.
1981 Testing of Delaval DSft 46. 6-cyJfnder in-Une confi-
guration, 28,600 cu. In. standing diesel engine for
electric power, compressor and marine applications.
Performance parameters being evaluated include
starting ability, precocnbustion chamber effects.
Ignition delay, and other engine parameters- To
date, the engine has been operated at full load using
a pre-mixed blend of 601 SRC-1I liquid and 40X diesel
oil which had been Injected Into the combustion
chamber with no modification of the engine, followed
by increasing oroportlons of SRC-II liquid up to
IQOt. Emissions monitored Include 0?. CO/CO?, NOX.
SOX. THC. and smoke.
1981 Fairbank-Horse 38 to 8-1/8, 6-cylinder opposed piston
design, 3J08 cu. in. standing diese? engine for elec-
tric power and marine applications, compressors and
pumos being tested. Parameters being evaluated in-
clude effects of load variations, combustion pressure
vs. time, and engine delay. Emissions monitored In-
clude C0/C02, NO, N02, S02. S04. HC, PAH, partlcu-
lates and oxidants.
1981 Testing of Superior 6-cylinder in line configuration
turbo-chargetJ 4120 cu. in. standing diesel engine for
use in compressors, pumping and electrical power gen-
eration. The purpose of the tests is to compare
engine performance parameters during synfuel and con-
ventional fuel combustion. Tests with shale-derived
distillate oil and a baseline No. 2 diesel fuel in-
clude SASS train sampling for PAH and participate.
Other emissions monitored include CO. HC, NOX, and
smoke. Tests with Exxon Donor Solvent liquids will
probably include the above procedures and also pilot
Injection and pre-injection starting tests.
(Continued)
-------
TABLE 3. (Continued)
00
Sponsoring Agency Test Fuels Time Period Project Description
E. Acurex Shale oil residuals. 1981 Testing of A.P.F. Allen BSC l^B 6-cylinder. in-line
Shoreham-by-the-Sea, configuration, 5101 cu. in. standing diesel engine
England for nanne, pumping, compressor dnd electric power
applications. Tests include injection, starting,
combustion duration and steadiness. Emissions moni-
tored include CO/C02. NOX. N02, THC, and smoke.
DOE, Conservation and Various shale- and 1978-1981 Auto engine dynamometer testing being conducted at
Solar Energy Division coal-derived fuels. SwRl. Particulates, NOX. CO/CO?, hydrocarbons, and
aldehydes being monitored.
SRC-II distillates and 1981 to --- Slurry/fuel project involving diesel engine testing.
shale-derived JP-5 and Particulates, NOX, and other emissions being moni-
DFM mixed with powdered tored.
carbon, sawdust, or other
cellulosic material.
Coal-derived methanol 1981 to --- Testing in 1,000 fleet vehicles; program currently
and gasohol. constrained for lack of fuel samples.
DOE, Office of Coal SRC-ll and shale-derived 1980 to --- Medium speed diesel engine testing conducted by
Utilization fuels. StMT-Pielstich. Paris; Baumester Wain, Copenhagen;
Grandi Motori Trieste, Trieste; and Selzer of
Swi tzerland.
SRC-II middle distillates, 1980 to --- Program conducted at Norwegian Technical Institute
•*~T a 2.9 to 1 blend of SRC-II in various ships.
;— middle and heavy distil-
late, and shale-derived
fuels.
SRC-ll middle distillate. 1981 to --- Continuation of low NOX fuel combustor concept pro-
gram (see TTH 32-36). Several combustors to be
tested by Uestinghouse; staged combustor to be
tested at several operating loads at Detroit Diesel
Allison; testing of 5 combustors planned at GE.
DOE, Pittsburgh Energy Blomass Fuel, H-Loal, 1981 to Continuation of small scale combustion of synthetic
Technology Center Exxon Donor Solvent, October 1982 fuels program (see Test 28). A 20-hp firetube
and shale fuel oils. boiler is to be tested with the above synfuels using
No. 2 and No. 6 fuel oils as a baseline. The pur-
pose of the program is to assess the possible envi-
ronmental impact of substituting synfuels for
petroleum in jollity and industrial boilers.
Department of Coal- and shale-derived 1981 to 1982 Testing of a recently-designed and constructed one
Transportation and diesel fuel. cylinder diesel engine, including collection of
Rutgers University particulates and other combustion products.
Sandla Laboratories Petroleum-derived synfuel 1981 to — Testing being conducted in single cylinder dfesel
simulation fuels, with systems and auto/truck engines from Cumins Engine
higher hydrocarbon/aroma- Co tmphasis on measurement of flame fronts and
tic content than conven- other engine/burn parameters. Limited emissions
tlonal fuels. monitoring performed.
Bank of America Methanol/gasoline blends. 1980 to — Testing being conducted in blends ranging from 2 to
187 methanol in fleet vehicles, with emphasis on
blends of 2 and 41:. CO, NO, and unburned hydro-
carbons being monitored.
-------
TABLE 4. TENTATIVE SYNFUEL TESTING PROGRAMS
Sponsoring Agency
Army, HERADCOM,
ft. Belvolr, VA
Fuels to be Tested Time Period
Diesel fuels and other 1982 to ---
synfuels (high aromatic
content fuels, low
lubricity fuels).
Project Description
Development of accelerated fuel qualifi-
cation test procedures, including matrix
of specific Army equipment components
and candidate fuels; project is part of
Army Alternative Fuels Program.
Air Force/Navy/EPA
(Under the direction
of Capt. H. Cewell,
USAF Civil Engineering
and Services Center,
Tindall AFB)
Navy Air Propulsion
Test Center
(NAVSSES),
Trenton, NJ
Shale-derived JP-4,
JP-5, and JP-8.
Various shale-derived
fuels.
Late 1981 Collection of particulates from various
engine combustion tests for toxicity and
biological effects testing.
AF Wright Aeronautical Various shale-derived
Lab, Aero Propulsion fuels.
Laboratory, Wright-
Patterson AFB,
Cincinnati
EPRI
EPA, Motor Vehicle
Emission Laboratory
Various liquid and solid
synfuels, including
shale-derived heavy and
middle residuals, and
methanol
EDS and H-coal liquids
Pending
receipt of
synfuel
samples
1982-1983
1981-1986
EPA, Industrial
Environmental
Research Labora-
tory, RTP, N.C.
SRC-11 fuel
Coal-derived middle
and heavy distillates;
shale-derived No. 2
fuel oil, methanol
(technical grade); and
petroleum reference
fuels
Testing of synfuels in various test
burners and aviation equipment.
Engine augmenter tests and whole engine
tests on 3 engines; emissions monitoring
for NOX, CO/C02, and hydrocarbons.
Testing of synfuels in various diesel
engines, turbines, and boilers; limited
emissions monitoring for SOX, NOX,
CO/CO?, 02 and/or particulates.
Late 1981 to Large standing diesel engines and a GE
September research engine, using same contractors
1982 as on-going programs (see Table 3).
Particulate matter to be collected.
1982 Electronically controlled internal com-
bustion engine at UTC, East Hartford,
CT. Limited emissions monitoring.
1981-1982 Comparative synfuel/petrofuel combustion
(Phase I) testing in a 2.5 MM Btu/hr packaged
boiler and In a 165 kw stationary diesel
to identify conditions leading to major
differences in emissions, and to deve-
lop recommendations for comparative
testing.
-------
the following are some general observations on the status, nature, and thrust
of the synfuel testing programs:
• Since the primary use of synfuel products is expected to be as
combustion fuels, nearly all synfuel end use testing programs
have involved evaluation of fuel suitability for use in combus-
tion systems (auto engines, industrial/utility boilers, turbines,
etc.) .
• Reflecting the developmental status of the synfuel technologies,
the thrust of the synfuel testing programs which have been carried
out to date has been to assess equipment performance and fuel
handling characteristics. Where some emissions monitoring has
been conducted, such monitoring efforts have been limited in scope
and have primarily emphasized measurements of gross parameters
such as particulates, NOX, SOX, etc., emissions. The limited
scope of the monitoring programs has also been in part due to:
(a) an absence of a clear definition of the specific environmental
data which would be required on synfuel products by regulatory
agencies (e.g., by EPA's Office of Pesticides and Toxic Substances
in connection with the Premanufacturing Notification Section of
the Toxic Substances Control Act); and (b) lack of a standard pro-
tocol for testing for environmental data acquisition.
• Most of the synfuel end use testing programs have been, or are
being, conducted/funded by DOD, EPRI, and DOE. The programs of
these organizations have, respectively, emphasized use of shale oil
products in military aviation and ship equipment; use of coal
liquids in boilers; and testing of methanol and methanol-gasoline
blends in auto engines and use of coal and shale-derived fuels in
stationary diesel engines.
• Many synfuel developers appear to have in-house synfuel testing
programs; the emphasis of these programs is primarily on synfuel
characterization and not on end use testing. The data generated
in these programs are generally considered company proprietary
and are not published.
• Nearly all the refined shale oil products which have been used
in combustion testing to date have been from the refining of
the 100,000 barrels of Paraho shale oil at Sohio's Toledo (Ohio)
refinery. Since this refining operation apparently did not
involve the use of typical unit operations which would be
employed in commercial refining of shale oil, the refined products
from this operation are not considered to be representative of
products from any future commercial refining of the shale -oil.
• To date the synfuel testing effort has been severely curtailed
by lack of adequate quantities of fuel for testing. Some of
the planned testing programs will utilize shale oil products
from the forthcoming refining of 50,000 barrels of shale oil
by Union Oil for the Defense Fuel Supply Center.
500
-------
• Synfuel products (especially the shale-derived materials) which
will be marketed in the future will most likely be blends and not
100 percent pure products. The use of 100 percent pure products
in the initial synfuel testing programs has been justified on
grounds that it would simulate a possible "extreme/worst" case
condition (at least from the standpoint of emissions and their
environmental implications).
• Although the performance testing is continuing, the limited data
which have been gathered to date indicate that the tested synfuels
are generally comparable to petrofuels and do not present any
unique problems from the standpoint of fuel handling and combustion
characteristics. Potential problems with long-term fuel storage
stability (observed with certain shale- and petroleum-derived
middle distillates) and durability and material compatibility
problems (e.g., possible increase in the engine wear with methanol
use) are under investigation.
• The very limited data which have been collected on the emission of
criteria pollutants (particulates, NOX, SOX, etc.) indicate that,
except for a higher emission of NOX with synfuels having a higher
content of fuel-bound nitrogen, the emissions of such criteria
pollutants are similar for both synfuel and their petrofuel coun-
terparts. For most synfuels, however, no data have been collected
on emissions of non-criteria pollutants such as polycyclic organic
matter (POM's), primary aromatic amines, nitropyrenes and other
organics. There is also very limited data on overall trace element
composition of emissions.
ACKNOWLEDGMENTS
This work has been performed under EPA Contract No. 68-02-3174, Work
Assignment Nos. 18 and 72, for the Industrial Environmental Research Labora-
tory, Office of Environmental Engineering and Technology, Research Triangle
Park, N.C. Gratitude is expressed to the EPA Project Officer, Mr. Joe
McSorley, for suggesting the subject study and for his advice and guidance
during the course of the effort.
The synfuel end use testing compendium is based on information and
documents provided to the study by individuals/organizations engaged in syn-
fuel characterization and end use testing; gratitude is expressed to the
supporting individuals/organizations, particularly those in Table 1, to whom
the project is deeply indebted.
501
-------
EXAMPLE DATA SHEET NO. 1
COMBUSTTON AND EMISSION
CHARACTERISTICS OF COAL-DERIVED LIQUID
1. FUELS TESTED
Synfuels: SRC-II fuel (5 ratios of medium and heavy boiling range com-
ponents); H-Coal (syncrude mode of operation, full-range distillate);
EDS (full-range distillate).
Reference fuel: No. 6 and No. 2 petroleum-derived fuels.
2. TEST EQUIPMENT
An 80-HP firetube boiler system extensively modified to simulate a util-
ity boiler including an indirectly fired air preheater, a scaled-down
utility boiler burner, radiation shields to increase the thermal envir-
onment in the combustion chamber, and capabilities to implement staged
combustion.
3. TEST SITE
KVB Combustion Research Laboratory, Tustin, California.
4. TEST OBJECTIVES
• Develop an understanding of the effect of compositional variations
of a particular coal liquid and the resulting effects on the imple-
mentation of combustion modifications for pollutant emission reduc-
tions;
• Establish an understanding of the difference in the combustion and
emission characteristics of coal liquids produced from various pro-
cesses—specifically the SRC-II Process, the Exxon Donor Solvent
Process, and the H-Coal Process;
• Establish a standard test method, using a small-scale facility, to
predict the response to changes in operation of smoking tendency,
CO, and NOx. This will be used to differentiate various fuel pro-
perties and the performance of each fuel in a large variety of com-
mercial boilers.
5. SPONSORING AGENCY
Electric Power Research Institute (EPRI)
Power Generation Program
Advanced Power Systems Division
Palo Alto, California
502
-------
EPRI Project Manager: W.C. Rovesti
Telephone No: 415-855-2519
6. CONTRACTOR
KVB Inc.
Irvine, California
Principal investigators: L.J. Muzio, J.K. Arand
Telephone No. 714-641-6200
7. TEST CONDITIONS
A systematic set of experiments was conducted which investigated the
following variables: excess air with single stage combustion, burner
stoichiometry with two-staged combustion, firing rate, air preheat
temperature, fuel temperature (viscosity), and atomizer (mechanical,
steam) .
8. ENVIRONMENTAL MONITORING
CO, NO, S02 , 50$, unburned hydrocarbons, smoke number, particu-
late size distribution.
9. PROJECT STATUS
Completed.
10. RESULTS
Emissions from the various synfuels combustion tests in this program are
summarized in Table A. A brief description of other emission test
results are shown below.
SRC II
Particle size data indicate that SRC-II fuel blends produced finer-
size-distribution particulate than No. 6 oil, the exception being SRC-II
heavy distillate component under single-stage combustion. Measured S02
emissions were consistent with the fuel sulfur content, with nearly all
fuel sulfur emitted as S02. An 803 concentration of 2 ppm for heavy
distillate component was the only SRC-II test detecting this pollutant.
Reference fuel No. 6 oil burn test also emitted 2 ppm 803. Unburned
hydrocarbon concentrations measured for SRC-II combustion tests ranged
from 1 to 14 ppm.
H-Coal
Average particle size of particulate matter proved to be less than
0.4 microns. Measured S02 emissions were consistent with fuel sulfur
content in that the S02 emissions were the lowest of all synfuels tested.
S03 was not detected. Unburned hydrocarbon emissions ranged from 1 to
4 ppm.
503
-------
EDS
Two particle sizing tests showed the average particle size to be
less than 0.4 microns. Measured SC^ emissions were consistent with
the fuel sulfur content. EDS flue gas samples showed no detectable
levels of 503. Measured unburned hydrocarbon emissions were 1 and 2 ppm.
11. REFERENCE
Muzio, L.J. and J.K. Arand. Combustion and Emission Characteristics of
Coal-Derived Liquid Fuels. EPRI AP-1878, Electric Power Research Insti-
tute, Palo Alto, Calif., 1981.
504
-------
TABLE A. SUMMARY OF EMISSIONS
Fuel Type
No. 6 oil
SRC-II 5.75/1
SRC- I I Medium
Distillate
SRC-II 2.9/1
SRC-II 0.4/1
SRC-II Heavy
Distillate
SRC-II Heavy
Distillate (210°F
Fuel Temperature)
H-Coal
EDS fuel
S
Fuel Ash n
Content 2
lb/10b Btu %
0.
0.
0.
0.
0.
0.
0.
0.
0.
0045
0017
0012
0041
018
034
034
0095
0045
3.7
3.8
4.0
3.3
3.4
3.3
3.8
-
2.8
2.8
ingle-Stage
Part.
lb/106
Btu
0.024
0.014
0.011
0.012
0.031
0.029
0.037
-
0.022
0.022
NO
ppm &
3% 02
270
400
476
361
509
381
392
—
247
259
Two-Stage (Low 0?)
°2
%
3.6
3.2
3.1
2.9
3.3
3.5
3.2
3.1
3.2
Part.
lb/106
Btu
0.037
0.022
0.017
0.015
0.039
0.184
0.065
0.037
0.0184
NO
ppm @
3% 0?
199
303
307
308
279
249
339
226
270
Two-Stage (High 0?)
°2
%
-
4.9
4.2
4.5
4.7
4.6
4.95
5.15
Part.
lb/106
Btu
-
0.020
0.012
0.017
0.039
0.090
0.034
0.0154
NO
ppm @
3% 0?
-
382
342
371
375
269
202
216
-------
EXAMPLE DATA SHEET NO. 2
EFFECT OF FUEL BOUND NITROGEN ON OXIDES OF
NITROGEN EMISSIONS FROM A GAS TURBINE ENGINE
FUELS TESTED
Synfuel: JP-5 type fuel derived from crude shale oil.
Reference fuel: JP-5 derived from petroleum.
2. TEST EQUIPMENT
Allison T63-A-5A turboshaft engine (free turbine type used in Army OH-58A
and Navy TF-57A helicopters).
3. TEST SITE
Naval Air Propulsion Test Center
Trenton, New Jersey
4. TEST OBJECTIVES
• Confirm the presence of high levels of NC^ in engine exhaust;
• Obtain information on conversion efficiency of fuel bound nitrogen
into NOx;
• Assess the impacts of high nitrogen fuel on meeting pollution control
regulations.
5. SPONSORING AGENCY
Deputy Chief of Naval Material (Development)
Department of the Navy
Washington, D.C. 20361
Project Officer: L. Maggitti
Telephone No: 202-545-6700
6. CONTRACTOR
Naval Air Propulsion Center
Fuels and Fluid Systems Division, PE71
Trenton, New Jersey 08628
Authors: A.F. Klarman, A.J. Rollo
Telephone No.: 609-896-5841
506
-------
7. TEST CONDITIONS
The T63-A-5A engine was installed in a sea level test cell using a three-
point mounting system. A flywheel and an Industrial Engineering Water
Brake, Type 400, were connected to the engine gearbox assembly at the
forward power output pad to absorb the engine power. The brake reaction
was measured by a Baldwin load cell. All parameters to determine the
engine starting and steady-state performance with the fuels were measured
using standard test cell instrumentation. Engine performance data is
contained in the reference report.
Fuels of varying nitrogen content were tested in a T63-A-5A engine to
measure their effects on exhaust gas emissions. Five test fuels varying
in fuel bound nitrogen content from 3 yg (nitrogen)/g (fuel) to 902 yg
(nitrogen)/g (fuel) were evaluated. The nitrogen content in the fuel was
adjusted by mixing a JP-5 type fuel derived from shale oil (902 yg (ni-
trogen) /g (fuel)) and regular petroleum JP-5 fuel (3 yg (nitrogen)/g
(fuel)).
8. ENVIRONMENTAL MONITORING
Hydrocarbons, carbon dioxide, carbon monoxide, and nitrogen oxides.
9. PROJECT STATUS
Project report completed November 1977. This is part of an ongoing Naval
program to evaluate fuel products derived from alternate sources.
10. RESULTS
Table B shows the results of the exhaust gas measurements performed
during the test program. Additional results include the following:
• NOX emissions for the same engine power rating increased with
increasing fuel nitrogen content.
• The conversion efficiency of fuel bound nitrogen to NO and NOx
was approximately 45 percent for the test data in which the NO
and NOjj values could be accurately measured.
• No significant effects were noted on engine performance or carbon
monoxide (CO) and unburned hydrocarbons (HC) emissions due to the
presence of high levels of fuel bound nitrogen.
• The use of shale derived JP-5 fuel with a high nitrogen content
will make it more difficult to meet the EPA NOX standards for
aircraft gas turbine engines.
11. REFERENCE
Klarman, A.F. and A.J. Rollo. "Effect of Fuel Bound Nitrogen on Oxides
of Nitrogen Emission From a Gas Turbine Engine", Naval Air Propulsion
Center, Trenton, New Jersey, NAPC-PE-1, November 1977, 32 pp.
507
-------
TABLE B. EMISSION DATA SUMMARY
O
O>
fuel Engine
Nitrogen Power
A4/9 fuel Rate
IDLE
1 60% MR
MIL
IDLE
47 60% HR
MIL
IDLE
267 60% HR
MIL
IDLE
SIS 6O% HR
MIL
IDLE
9O2 6O% HR
MIL
00,
1.98
_
1.01
2.08
2.41
3.03
2.08
2.43
3.03
2.10
2.43
3.01
2.10
2.41
3.03
PC*
1035
_
140
985
430
130
1005
380
140
950
445
130
992
460
135
CO
9/»
0.714
_
0.227
O.692
O.482
0.207
0.698
0.438
0.224
0.688
0.482
0.210
0.710
0.500
0.218
B/kg fuel
99.2
.
9.25
90.5
35.0
8.6O
92.3
31.0
9.26
86.7
3«.2
8.60
90.4
37.4
8.91
Pirn
6.7
-
21.9
7.7
12.7
24.1
9.1
16.5
27.6
11.6
17.8
11.6
14.9
22.1
15.9
HO
g/s ?
0.00495
-
•
0.0416
0.00579
0.0152
0.0415
0.00677
0.0204
0.0471
0.00900
0.02O6
0.0547
O.O114
0.0257
0.0621
NO, (a* HOj)
I/kg fuel
0.688
-
1.69
0.758
1.11
1.72
0.89S
1.44
1.96
1.11
1.55
2.24
1.45
1.92
2.55
Pt» g/c g/kg fuel
6.7
-
21.9
7.1
11.1
24.1
9.4
16.7
27.6
12.1
18.4
11.6
16.0
22.5
16.1
0.00690
_
0.0617
0.00887
0.0241
0.0615
0.01O8
0.0115
0.0726
0.0146
0.0327
0.0818
0.0188
0.0401
0.0962
1.06
.
2.59
1.16
1.75
2.64
1.42
2.24
3.00
1.85
2.47
1.-44
2.19
1.01
1.95
PP«
157
-
5.6
111
18.1
8.4
114
14.5
11.1
109.6
18.6
8.7
116
18.2
8.4
HC
g/« 9/kg fuel
0.0501
_
0.00422
0.0427
0.00952
0.00621
0.0412
0.00775
0.00825
0.0368
O.O093S
0.00652
0.0185
O.OO918
0.00629
6.99
-
0.172
5.59
0.692
0.258
5.71
0.549
0.141
4.65
0.7O2
0.267
4.91
0.687
O.2S8
(calculated)
O.O0979
-
0.0146
0.0105
0.0119
0.0146
0.0105
0.0119
0.0146
0.0106
0.0119
0.0146
0.0106
O.O119
0.0146
-------
COMPARATIVE TESTING OF EMISSIONS FROM COMBUSTIONf
OF SYNTHETIC AND PETROLEUM FUELS
by: W. Gene Tucker and Joseph A. McSorley
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
EMISSIONS FROM COMBUSTION OF SYNTHETIC FUELS
There are two basic reasons to investigate the emissions from the combus-
tion of coal- and shale-derived synthetic fuels:
o The physical and chemical characteristics of these synfuels
will probably be different from the petroleum-based analogs
that they will replace or supplement (e.g., by blending);
therefore, the emissions from their burning are likely to
be different.
o The types and numbers of combustors in which synfuels might
be used are very large; therefore, the potential for exposure
to their emissions is very great.
These two reasons argue for research and development now, before exten-
sive commercialization of synfuels, on procedures that can be used to test
emissions from representative combustors burning prototype synfuels, and pe-
trofuels that they may replace. Once developed, such procedures can be used
to determine which synfuel/combustor combinations should be avoided, and which
combinations will result in "clean" emissions (perhaps cleaner than from pres-
ent combustion of petroleum-based fuels, or from future combustion of lower-
grade petrofuels).
CHARACTERISTICS OF SYNTHETIC FUELS
Both physical and chemical characteristics of fuels can affect combustion
emissions. Physical properties of solid and liquid fuels such as particle
size, density, viscosity, and surface tension affect the rate at which the
fuel volatizes to a combustible (gaseous) state. Many of the solid and liquid
products available to date from U.S. synfuel pilot plants have physical prop-
erties that tend to make them volatize less easily than the coal and petroleum-
based fuels they may replace.
Generally, the chemical properties of pilot-scale synfuels produced to
date have also been of concern relative to petroleum analogs, mainly because
of their greater concentration of high-molecular-weight organics. A consider-
able and growing literature exists on the content of aromatic and substituted
aromatic components of coal- and shale-based synthetics (e.g., reference 1).
There are, however, many process options for producing clean synfuels such as
methanol, or refining crude products to specifications meeting or exceeding
those for current petroleum fuels.
There is, therefore, a trade off between cleaning the synthetic product
before combustion and burning the fuel cleanly. Aside from consideration of
509
-------
fuel handling and distribution concerns, the degree of need for a clean fuel
will depend on the combustion application.
THE POTENTIAL POPULATION OF SYNFUELED COMBUSTORS
Emissions are greatly affected by type of combustor and how well it is
being operated. Light oil, wood, and even methane can lead to undesirable.
emissions if they are burned improperly. Aside from the tens of millions of
mobile internal-combustion engines that are candidates for synthetic fuels
(or blends with petrofuels), many stationary units in this country are pres-
ently fired with oil:
o Thousands of large utility boilers.
o Hundreds of thousands of industrial and commercial boilers.
o Hundreds of thousands of stationary diesel engines.
o Millions of commercial and residential furnaces.
There are certainly examples of both "clean" and "dirty" burning units in
each of the above categories. Generally speaking, however, the amount of at-
tention given to the operation of the units decreases from top to bottom of
the list. Typical combustion efficiency of units in each of the four categor-
ies probably follows the same order.
On the other hand, fuels burned in residential and commercial furnaces
are generally lighter and cleaner than those in diesels, which in turn are
lighter and cleaner than those burned in industrial boilers. Overall, large
utility boilers most frequently burn the heaviest fuels of all.
This apparent inverse relationship between attention to operation and
cleanliness of fuel leads us to suspect that the primary categories of con-
cern among stationary sources might be the middle two — industrial/commercial
boilers and stationary diesels. Also, a recently completed study on synfuels
uses (reference 2) tends to indicate that these two categories are likely to
be among the first stationary sources to use synfuels in commercial quantities.
EPA therefore initiated a research and development program early this
year to develop a set of engineering procedures for comparative testing of
emissions from combustion of coal- and shale-based liquid synthetics and
petroleum-based analogs. It is designed to be a multi-phased program with
several iterations of procedure development, followed by combustion tests to
hone the procedures. The following sections of this paper describe the cur-
rent status of the initial work (Phase I) of this program.
EXISTING EMISSIONS DATA
Data on emissions from combustion of synfuels are very limited. Data on
combustion products from oil burning, especially organics, are also limited.
Whereas emissions of inorganics are fairly predictable as oxidation products
510
-------
of fuel constituents, organic products of incomplete combustion are a differ-
ent story. The possibilities are virtually limitless and much more difficult
to predict; carefully collected empirical data are needed.
Because changes in emissions are of greater concern than absolute emis-
sion rates when switching to synthetics, data of greatest value will be com-
parative data on emissions from a synfuel and its petroleum analog(s), burned
in an appropriate combustor at representative operating conditions. One rea-
son is the oft-stated observation that emissions from combustion of currently
burned petroleum fuels constitute an accepted baseline. Another reason is
that physical, biological, and chemical characteristics of synthetic fuels
(and their emissions) will be evolving as the synfuel industry evolves. It
will therefore be important to continually combustion-test emerging synthetic
fuels to understand the best environmental and economic balance between clean-
ing these fuels and burning them cleanly.
PROCEDURES FOR COMPARATIVE TESTING OF EMISSIONS
After several months of Phase I of the EPA program, a very preliminary
set of procedures has been developed that addresses personnel safety, combus-
tor operation, emissions sampling, and sample analysis. An overview of cur-
rent thoughts on each of these aspects follows.
PERSONNEL SAFETY
Because of the hazardous nature of some of the fuels, samples, and resi-
dues being handled, precautions are being taken to protect technicians, super-
visory personnel, and observers. The materials requiring greatest attention
are spills of synthetic and heavy petroleum fuels, residues from cleaning the
combustor, and the collected samples of combustion products. During combus-
tion runs and combustor cleaning operations, specified disposable protective
clothing and cartridge respirators must be used. Personnel involved in sample
handling, preparation, and analysis are required to follow standard precaution-
ary laboratory procedures.
COMBUSTOR OPERATION
The following considerations are especially important for development of
procedures for comparative testing of synfuel combustion emissions:
o A combustor that is representative of intended uses must be used.
This will generally preclude use of laboratory-scale burners, and
will often require combustors with substantial fuel feed rates.
o Large quantities of the synfuels to be tested will often not be
available. This will dictate relatively short combustion runs.
o Run-to-run cross-contamination of internal combustor surfaces is
a potential problem that may confuse emission measurement results.
Some method of equipment cleaning between runs needs to be devel-
oped.
511
-------
With these three factors, plus other considerations derived from a knowl-
edge of how combustors and their operation affect emissions, a preliminary set
of procedures, summarized below, has been established:
1. Clean combustor surfaces. This step will consist of brushing and
vacuuming accessible internal surfaces to remove loose deposits
from the previous run. This step also applies to the dilution
tunnel discussed later.
2. Burn No. 2 oil. A typical No. 2 fuel oil, available in sufficient
quantity to be used as a reference fuel for all runs, will be used
for approximately 1 hour to bring the combustor to steady operation
and "recondition" the internal surfaces.
3. Burn test fuel. The fuel supply will be switched to the synthetic
or petroleum fuel to be tested. Each test burn is expected to last
2 to 6 hours.
4. Shut down the combustor, and repeat cycle. Eventually, it is hoped
to be able to complete one run (steps 1-3) per day or four runs per
week. It remains to be seen, of course, whether stable operation
and repeatable results can be obtained in sizable combustors with
such short turnaround time.
The test fuel firing rate will be set at 80% load and the excess air ad-
justed to achieve 10% opacity or less in the stack gas from the boiler (excess
air will generally be in the range 5%-10%). This opacity setting represents
energy-efficient operation for oil-fired boilers. It also represents marginal
performance from a particle emissions standpoint. Differences between fuels
in emissions potential will therefore tend to be accentuated at this setting,
which should expedite screening for potential problems.
With some of the cleaner fuels, an opacity as high as 10% may not be at-
tainable. In such situations, an excess air setting of about 5% is planned.
If, for some of the heavier fuels, an opacity as low as 10% cannot be main-
tained at a reasonable excess air setting, control at about 35% excess air is
planned.
The diesel engine will be operated at its continuous load setting of
165 kW (80% load). It will be operated at approximately 85% excess air, which
is typical for such combustors, and the opacity measured but left to vary from
fuel to fuel.
FUEL AND EMISSIONS SAMPLING
As shown on Figure 1, five types of samples are being taken during Phase
I of the program. They are, briefly:
1. Fuel samples. Grab samples are taken from fuel storage (most of the
fuels for Phase I of the program are stored in drums).
512
-------
Filtered
To Stack
u>
1 Fuel 1 Fuel
| storage |
1
1
1
1
1
1
1
•
Cornbustor
1
1
1
Air
Stack Gas T"
I
•
Dilution
Tunnel
Filter
i
- 1 Modlfled 1
P| Method5|
i— — — — -- j^.— 1
-1
/ Paniculate \
Samples
(•v250°F)
Figure 1. Sampling - Comparative Emissions Test Procedure (Phase I)
-------
2. Continuous monitoring. Stack gas measurements of 0«, CO, C0?, NO,
NO SO,,, total hydrocarbons, and opacity are made continuously
while trie conditioning No. 2 fuel and the test fuels are being
burned.
3. Particulate samples. Particles in the stack gas will be sampled by
a modified Method 5 train (Figure 2 and Reference 3). Particles
will be collected in a fiberglass filter at approximately 125°-150°C
(250°-300°F) over a 2- to 3-hour period during each run of a test
fuel.
4. Vapor samples. Stack gas vapors that pass through the filter of the
modified Method 5 train will be cooled to approximately 15°C (60°F)
and collected on XAD-2 sorbent material. Vapor samples will be
collected over the same periods as the particulate samples.
5. "Ambient" samples. A portion of the stack gas from the combustor
will be mixed with filtered air in a dilution tunnel (air-to-stack
gas ratio of approximately 10:1). A large (50-cm square) Teflon-
coated fiberglass filter at the end of the dilution tunnel will
collect particles during the full length of each run of a test fuel.
The dilution tunnel is included in the preliminary procedures for two
reasons: (a) by simulating atmospheric dilution/cooling conditions near the
exit of the stack, it provides a sample more representative of ambient par-
ticles than the ones collected in the stack, and (b) it is an inexpensive way
to collect relatively large samples for both chemical and biological testing.
FUEL AND EMISSIONS ANALYSES
Figure 3 summarizes the physical, chemical, and biological analyses being
done on the samples of fuels, stack gas particles, stack gas vapors, and sim-
ulated "ambient" particles from the dilution tunnel. The primary details of
the preliminary analytical procedures follow.
1. Fuel specifications. Standard ASTM procedures are being used to
measure the fuel parameters of most common interest to people who
purchase or burn fuels. The following measurements are also made
for each fuel: inorganic screening by spark source mass spectrom-
etry (SSMS), gas chromatography-mass spectrometry (GC-MS)for quali-
tative organic screening, spot test for polycyclic aromatic hydro-
carbons (PAHs), and boiling point analyses for organics between
100°-300°C and >300°C.
2. Inorganics. Elemental constituents in the fuels will be semi-quan-
titatively screened by SSMS. Elements selected from the fuel
screening will be analyzed in the stack gas and "ambient" particles
by atomic absorption (AA).
3. Organics. The objective for analysis of organics, as for inorganics,
is to screen for major compositional differences between samples
from synthetic fuels and from their petroleum analogs. The battery
514
-------
Stack
Exhaust
.
(Particles)
Cooling
Probe
A
Stack
Gases
Pump
Absorber
(Organic Gases)
Implnger Train
(Inorganic Gases)
Figure 2. Modified EPA Method 5 Train
515
-------
INDUSTRIAL
BOILER
15 TEST RUNS
o 9 PETRO RUNS
o 5 SYNFUEL RUNS
0 1 METHANOL RUN
1
STATIONARY
DIESEL
ENGINE
8 TEST RUNS
0 4 PETRO RUNS
o A SYNFUEL RUNS
FUELS ANALYSIS MODIFIED METHOD 5 SAMPLES DILUTION TUNNEL SAMPLES
INORGANIC ANALYSIS SCREENING
BY SMSS SEMI-QUANTITATIVE
(AMBIENT SIMULATION)
INORGANIC ANALYSES OF
SELECTED ELEMENTS (AA)
1
SAMPLE PREPARATION
FOR CHEMISTRY
INORGANIC ANALYSIS OF I
SELECTED ELEMENTS (AA)|
|
SAMPLE PREPARATION I
FOR CHEMISTRY 1
TOTAL ORGANIC QUANTITATION
INTO 100" 300° C BOILING
POINT AND >300° C BOILING
POINT
PAH SPOT TEST
SEMI-QUANTITATIVE
GC-MS SCREEN
QUALITATIVE
TOTAL ORGANIC
QUANTITATIVE ANALYSIS
SAMPLE PREPARATION
(CHEMISTRY)
SAMPLE PREPARATION AND 1
BIOTESTS (AMES TEST) I
|
TOTAL ORGANIC QUANTITATIVE!
ANALYSIS |
PAH SPOT TEST 1
SEMI-QUANTITATIVE I
ANALYSIS
GC/MS SCREEN
QUALITATIVE ID
GC/FID QUANTITATIVE ANALYSIS
COMPOUNDS /CLASSES
PAH SPOT TEST
SEMI-QUANTITATIVE
GC/MS SCREEN
QUALITATIVE ID
GC/FID QUANTITATION FOR I
SPECIFIC CLASSES OR COMPOUNDS 1
GC/FID QUANTITATION FOR 1
SPECIFIC CLASSES/COMPOUNDS 1
T I I
QUANTITATION OF NON
GC - ABLE COMPOUNDS
QUANTITATION OF NONI
GC - ABLE COMPOUNDS!
1
QUANTITATION OF NON I
GC - ABLE COMPOUNDS!
1
-REPORT AND RECOMMENDATIONSl
REPORT AND RECOMMENDATIONSl
[ REPORT AMD RECOMMENDATIONS 1
Bioassay Screening - Baaed on 200 mg participate sample (20 mg organic extractablea).
Figure 3. Analyses - Comparative Emissions Test Procedure (Phase I)
516
-------
of techniques includes (a) quantitation of total organics; (b) the
"spot" test (reference 4) for PAHs; (c) a qualitative screening by
GC-MS, to obtain a very rough "fingerprint" of the organic emissions;
and (d) quantitation of the non-gas-chromatographable portion. Sam-
ples that are compositionally distinctive, based on the above tests,
will be further analyzed by gas chromatography-flame ionization
detection (GC-FID) to obtain semi-quantitative information on major
classes or compounds present.
All three emissions samples — stack gas particles, stack gas vapors,
and "ambient" particles — will undergo this battery of tests. The
fuels will be analyzed similarly.
4. Bioassays. Comparative biological screening of emission samples in
Phase I of the program will be limited to short-term bacterial muta-
genicity tests of the type originally developed by Ames. The mini-
mum desirable sample quantity for these tests is 20 mg of organic
extractables. If the extractables constitute 10% of the total
weight of particles collected, a minimum of 200 mg of particulate
catch will be required for the bioassays alone. This amount of
sample can only be obtained on the filter at the end of the dilution
tunnel, with current procedures. In fact, several of the planned
runs with relatively clean fuels are not expected to produce suffi-
cient sample for biological screening. Runs with sufficient sample
will be tested using the Salmonella typhimurium strain TA98, reverse
mutation assay. Each test will be run at 5 to 7 dose levels, both
with and without metabolic activation. Any testing beyond this
simple assay, such as assays on fractions of samples, will be done
only as screening indicates a need and as sample material allows.
The need for more extensive biological testing (e.g., additional mu-
tagenesis assays or carcinogenesis assays) in future phases of the
program will be determined largely from the results of Phase I.
COMBUSTION EMISSIONS TESTING
A series of comparative combustion emissions tests has been planned as
part of Phase I of the program, to evaluate the soundness and practicality of
the preliminary testing procedures. The following sections describe the com-
bustors to be used, the fuels to be burned, and the schedule for the remainder
of Phase I.
COMBUSTORS
Table 1 lists the characteristics of the two combustors being used at
EPA's combustion research facility at the Research Triangle Park, NC. The
package boiler (so-called because units of this size can be shop-fabricated
and delivered to the site as a "package," rather than being erected at the
site) represents small-to-medium-sized fire-tube boilers used in industry and
commercial establishments. In addition to its normal dual-fuel burner, it can
be (and has been, in past experiments) equipped with a "low-NO " burner which
promotes staged combustion and lower emissions of NO . In Phase I, the con-
ventional burner will be used; in subsequent phases, the effectiveness of the
517
-------
Table 1. Combustors Being Used in Comparative
Emissions Tests (Phase I)
PACKAGE BOILER
o North American scotch marine boiler
o Typical of a broad range of small-to-medium industrial and
commercial boilers
o Capacity: 2.5 x 10 kJ/hr fuel rate (2.5 x 10 Btu/hr fuel rate;
2,000 Ib/hr steam)
o Operating rate: 80% of capacity; approximately 50 liters per hour
(13 gal./hr) of fuel
o Dual-fuel burner (heavy oil and gas)
o Outside dimensions: 1.4 meters (4-1/2 ft) diameter, 3 meters (10 ft)
long
STATIONARY DIESEL
o Caterpillar Model D334
o Typical of medium-sized industrial stationary engines
o Capacity: 205 kW (generator output)
o Operating rate: 80% of capacity; approximately 53 liters per hour
(14 gal./hr) of fuel
518
-------
new low-NO^ burner design may be tested on synfuels.
The stationary diesel represents medium-sized industrial and commercial
engines used for backup power generation, pumping applications, and powering
various other mechanical equipment. Both combustors will be operated as
described in the previous section on "Combustor Operation."
Future phases of the program are expected to repeat tests with these
combustors for various load and operating settings. In addition, tests may
be run with the low-NOx burner to determine its effect on synfuel combustion
emissions. Another possibility is a series of tests on residential furnaces.
FUELS
The fuels used in Phase I testing were chosen to cover a broad range of
petroleum and synthetic products. This is mainly to check the applicability
of the test procedures. A secondary purpose is to obtain information on major
differences in emissions among fuels. It is important to understand that,
whereas the coal- and shale-based synthetics being used are typical of those
currently available in the U.S. in barrel quantities, they may not be at all
typical of synthetics that are eventually marketed for use in industrial boil-
ers and stationary diesels. Therefore, whereas the results from Phase I will
be useful in refining the test procedures and planning for Phase II testing,
they are not intended for use in environmental assessment of synfuel combus-
tion.
Table 2 lists the fuels being combusted. Additional descriptions follow.
1. Petroleum fuels. Seven petroleum fuels will be tested — six in the
package boiler, and three in the diesel, with two of the seven burn-
ed in both units. Four of the fuels will be heavy (No. 6 grade) ,
with sulfur contents ranging from 1 to 3%, nitrogen 0.04 to 0.7%,
and ash 0.05 to 0.3%. The other three fuels are lighter (No. 2
grade) with sulfur contents of 0.02 to 0.5%, nitrogen 0.04 to 0.1%,
and <0.1% ash. All seven fuels were obtained from east coast dis-
tributors .
2. Coal-derived distillates. Three different coal-derived synthetics
will be tested. An SRC-II heavy distillate from the Ft. Lewis Sol-
vent Refined Coal pilot plant and an EDS middle distillate from the
Exxon Donor Solvent pilot plant in Baytown, Texas, will be burned
in the package boiler. The EDS middle distillate and an SRC-II
middle distillate will be burned in the stationary diesel.
3. Shale-derived fuel. Refined product (light No. 2) from the Sohio
refinery run of Paraho shale oil will be burned in both the package
boiler and diesel. This oil has been heavily hydrotreated, and
appears to be one of the cleanest fuels to be burned in Phase I.
Future phases of the program are planned to repeat burns with these
fuels, other petroleum fuels, other synthetic fuels as they become available,
and blends of synthetics and petrofuels.
519
-------
Table 2. Fuels Being Used in Comparative
Emissions Tests (Phase I)
PACKAGE BOILER RUNS
o 4 Heavy Petroleum Fuels 6
o 2 Light Petroleum Fuels 3
o 1 Coal-Derived Middle Distillate 2
o 1 Coal-Derived Heavy Distillate 2
o 1 Shale-Derived No. 2 Fuel 1
o 1 Methanol 1
15
STATIONARY DIESEL
o 3 Light Petroleum Fuels 4
o 2 Coal-Derived Middle Distillates 3
o 1 Shale-Derived No. 2 Fuel !_
8
Two of the light (No. 2) petroleum fuels, one of the coal-
derived middle distillates, and the shale-derived No. 2
fuel are identical for both combustors.
520
-------
SCHEDULE
The series of combustion emission tests just described will be conducted
during November and December of this year. The bulk of the samples will be
analyzed from January through March 1982. Data will be compiled and distrib-
uted to program participants and fuel suppliers during early spring. A work-
shop for discussion of data interpretations, test procedure revisions, and
plans for Phase II of the program is planned for June 1982. The workshop will
bring together EPA and DOE participants, fuel suppliers, and selected addition-
al experts in combustion, analytical chemistry, and data analysis.
SUMMARY
As coal- and shale-derived synthetic fuels begin to enter the market in
the 1980s, questions will arise regarding the nature of the emissions from
their combustion. A program was recently initiated by EPA to develop engin-
eering procedures for measuring emissions so that concerned parties (EPA,
synfuel developers, synfuel users, and others) can address such questions.
The basic approach that has been taken is to compare emissions from syn-
fuels burning to emissions from the burning of petroleum-derived fuels that
will be displaced, in combustors that are representative of expected synfuel
applications. An important objective of the program is to devise testing
procedures that are as simple and inexpensive as possible, but that highlight
important differences in emissions from synfuels and petrofuels, where they
exist.
The program for development of procedures for making such tests will be
multiphased, over a several-year period. Preliminary procedures have been
developed for liquid-fueled industrial boilers and stationary diesels. Com-
bustion testing is now underway to check the feasibility and practicality of
the procedures. The procedures and data from the first-phase results will
be reviewed at a workshop by program participants and additional experts.
Subsequent phases of the program will focus on refinement of the procedures
and expansion of their applicability to other fuel/combustor combinations.
* * *
Acknowledgements
Many people in the Environmental Protection Agency have been involved in
the planning of this program, and many more are expected to contribute as the
program matures. The authors especially appreciate the contributions to date
of: Michael C. Osborne and Jack H. Wasser of the Combustion Research Branch,
Raymond G. Merrill of the Technical Support Staff, and Robert P. Hangebrauck
of the Energy Assessment and Control Division, all of the Industrial Environ-
mental Research Laboratory, RTP; Joellen Lewtas of the Health Effects Research
Laboratory, RTP; and Donald Barnes and Carl Mazza of the Office of Toxic Sub-
stances.
521
-------
References
1. "A Critical Review of the Mutagenic and Other Genotoxic Effects of
Direct Coal Liquefaction," ORNL-5721, July 1981.
2. "Environmental Aspects of Synfuel Utilization," EPA-600/7-81-025
(NTIS PB 81-175937), TRW, Inc., Redondo Beach, CA, March 1981.
3. "Sampling and Analysis Methods for Hazardous Waste Incineration,"
Draft Report, Arthur D. Little, Inc., Cambridge, MA, under EPA
Contract 68-02-3111, July 1981.
4. "Sensitized Fluorescence for the Detection of Polycyclic Aromatic
Hydrocarbons," EPA-600/7-78-182 (NTIS PB 287-181), Arthur D. Little,
Inc., Cambridge, MA, September 1978.
522
-------
UNPRESENTED PAPERS
523
-------
PROBLEMS ASSOCIATED WITH THE ANALYSIS OF SYNFUELS T
PRODUCT, PROCESS, AND WASTEWATER STREAMS
by: H. C. Higman, D. K. Rohrbaugh,
R. H. Colleton, and R. A. Auel
Hittman Associates, Inc.
Columbia, MD 21045
ABSTRACT
Hittman Associates, Inc., as part of an environmental assessment of
coal liquefaction technology sponsored by the U.S. Environmental Protec-
tion Agency (EPA), performed various analyses on samples from the Solvent
Refined Coal II (SRC-II) plant in Ft. Lewis, Washington, and the Exxon
Donor Solvent Plant (EDS) plant in Baytown, Texas. This paper describes
several of the problems encountered in these analyses and methods taken to
mitigate them. Recommendations are made on approaches for avoiding such
problems.
INTRODUCTION
Though there exist standard methodologies for the analysis of pure
organic extracts, the complex nature of the product and process streams
from synfuels plants cause specific problems which are often complicated
by the analytical requirements of the particular study. This paper ad-
dresses several problems encountered in the analyses of samples from the
SRC-II pilot plant in Ft. Lewis, Washington, and the EDS facility in
Baytown, Texas: (1) the analysis of products and effluents to determine
process variability over a finite time period; (2) the analysis of vola-
tile organic compounds from heavily loaded sample matrices; (3) the anal-
ysis of phenolics from heavily loaded phenolic streams; and (4) the analy-
sis of sulfides, cyanide, thiosulfates, and thiocyanates from heavily
loaded aqueous streams.
PROBLEM AREAS
REPRODUCIBILITY IN PROCESS VARIABILITY STUDIES
It is necessary to use a consistent approach in analyzing samples for
process variability. For the Ft. Lewis effort, two sets of two samples of
a heavy distillate stream were obtained over a 3-day period. This proce-
dure provided a set of 12 discrete samples for studying process vari-
ability with built-in controls for sampling and analytical variability.
524
-------
The heavy distillate stream is a very complex sample matrix contain-
ing several hundred discrete components which range in concentration from
the parts-per-billion level to several percent of the overall mixture.
Analytical options included: (1) several types of fractionation procedures
such as the Level 1 Assessment protocol, which yields seven discrete
fractions from a silicic acid column; (2) fractionation by Florisil chroma-
tography followed by chemical and further column separation to achieve
separations by class for sulfur, nitrogen, and polycyclic aromatic hydro-
carbons; and (3) analysis of the gross mixture for major constituents. To
eliminate variations which would be introduced by fractionation and con-
centration procedures, it was decided to analyze the gross mixtures and to
use the data obtained to define process, sampling, and analytical
variability.
With mixtures of this complexity, capillary gas chromatography pro-
vides the most effective separation. It can be coupled with mass spec-
trometry to obtain as much qualitative information as possible about the
major constituents of the mixture. Figure 1 shows the region of the total
ion reconstructed chromatogram of a typical sample from the heavy distil-
late series. The broad peaks are not a function of poor chromatography
but, rather, indicate co-eluting components. Several of the major com-
ponents are identified.
** SPECTRUM DISPLRY/EDIT **
PV-B DIL 1:186 (6-9-81)
3UL Sf«1. + 1UL Die INJECTED
FRN 5317
1ST SC/PG: 228
x= i.ee Y= i.ee
TI
14
I1?
ife
22
Figure 1. Total Ion Reconstructed Chromatogram of Heavy Distillate Sample
from SRC Pilot Plant.
525
-------
Figure 2 shows four separate total ion chromatograms of different
samples taken the same day. The run-to-run reproducibility of these
samples is good and can be used effectively for comparison purposes.
13 14 IS 16 17 18 19 28 21 22 23 24 25 26 27 28 29 38 31
Figure 2. Comparison of Process Variability Samples.
Figure 3 shows selected extracted ion chromatographic profiles (EICP)
of the molecular ions for naphthothiophene, dibenzothiophene, phenanthrene/
anthracene, and a series of C? biphenyls and acenaphthenes. Using EICP, it
is possible to discriminate between species in the mixture, while components
co-elute when gas chromatography is used alone.
Figure 4 shows another example of co-eluting species. In this case,
pyrene and fluoranthene co-elute with substituted phenanthrene/anthracene
components and with a series of C9 carbazoles.
526
-------
** SPECTRUM DISPLflY/EDIT **
PV-B DIL 1:188 (6-9-81)
3UL SflH. + 1UL Die INJECTED
FRN 5317
1ST SC/P9: 232
X= 1.88 Y= 1.88
82.
184.«
178.(
1 - C_ biphenyls
2 - C acenaphthenes
naphthothiophene
f*~COL)rtP) dibenzothiophene
phehanthrene/
anthracene
Figure 3. Co-eluting Components in the Anthracene Region
** SPECTRUM DISPLflY/EDIT **
PV-B DIL 1:188 (6-9-81)
3UL SflH. + 1UL D18 INJECTED
FRN 5317
1ST SC/PG: 484
X= 1.88 Y= 1.88
C, phenanthrenes/
anthracenes
Figure 4. Co-eluting components in the Pyrene Region
527
-------
By using the extracted ion chromatographic profiles for various
components and analyzing the mass spectral data, it is possible to iden-
tify a discrete set of components. The areas of peaks from the extracted
ion chromatographic profiles can be integrated to generate quantitative
data for specific components from the gross sample. The most commonly
used calculation involves obtaining integrated peak areas from known
amounts of standards and comparing these peak areas to those of a known
amount of internal standard. From these data, it is possible to calculate
the relative response factor (RRF). The amount of specific components
present can then be determined using the equation below. For gas chroma-
tography, this is an excellent and extremely reproducible calculation
because the detectors used are very stable.
RRF = amount d!0 x Area Std.
amount Std. Area d,,.
Amount X = Areax x 1/RRF x D
Area d,0
Where: Std. = compound of interest
din = internal standard
D = dilution factor
When used in conjunction with mass spectrometry, this method yields
acceptable reproducibility. However, mass spectrometry is not as stable a
detector and is more sensitive to changes in relative concentration of the
sample and internal standard than is gas chromatography. This means that
RRF values must be calculated very frequently if acceptable quantitative
data are to be obtained for repetitive studies. Additionally, the overall
sensitivity of the mass spectrometer can change dramatically over short
periods of time.
The RRF method was used to generate the data in Table 1. Several
representative concentrations for components found in the heavy distil-
lates are shown. Samples A and B are one pair and C and D are a second
pair. A and B were taken at the same time, C and D at different times,
during the same day. These data were obtained as part of the process
variability study and are, we feel, representative in light of the com-
plexity of the sample matrix.
The values in Table 2 were derived by using the data obtained for
standards normally used in the generation of RRF values. In this case,
these data were used to calculate a working standard curve by least
squares linear regression (LSLR) analysis. A second set of least squares
linear regression lines were calculated using the data but normalizing the
output of the internal standard to a set figure and adjusting the areas of
the known compounds to reflect this normalization. All of the lines used
had a correlation coefficient exceeding the 95 percent confidence level.
528
-------
Generally, the values obtained for analytical pairs are in better agree-
ment for the normalized lines.
TABLE 1. COMPARISON OF VALUES OBTAINED BY RRF CALCULATION (g/kg)
Sample
Component
Fluorene
Carbazole
Dibenzothiophene
Phenanthrene/ Anthracene
Pyrene
Chrysene
A
2.17
5.67
8.46
20.02
11.66
1.09
B
2.36
6.67
8.96
22.64
13.68
1.57
Sample
C D
2.12
7.55
9.20
24.46
13.28
1.73
1.50
5.68
8.36
20.88
11.37
1.21
Benzo(a)Pyrene/Benzo(e)Pyrene
[B(a)P + B(e)P]
.83
.56
.85
.38
TABLE 2. COMPARISON OF SAMPLES BY LSLR AND NORMALIZED LSLR
LSLR LSLR-N LSLR LSLR-N
Component
Fluorene
Carbazole
Dibenzothiophene
Phenanthrene/
Anthracene
Pyrene
Chrysene
B(a)P + B(e)P
A
2
4
7
16
13
1
.04
.95
.27
.99
.27
.32
.40
B
1
4
6
15
12
1
.81
.68
.20
.41
.47
.49
.49
A
1.74
4.48
6.67
15.27
14.8
1.09
.31
B
1
5
7
17
17
1
.89
.27
.06
.27
.39
.58
.47
C
1.23
4.47
5.42
13.15
9.58
1.28
.52
D
1.07
3.40
4.90
11.96
8.69
1.04
.34
C
1.44
5.59
7.25
16.56
15.08
1.44
.48
D
1.21
4.48
6.59
15.93
14.42
1.22
.34
Table 3 shows a comparison of the three methods used for quantitation
of these selected compounds. The values shown represent a percentage
variation from an average value for the paired samples. Using the LSLR
529
-------
data does little to alter the measurements obtained by RRF values when the
data are very close, as in the A-B pair for fluorene and in the dibenzo-
thiophene pairs. However, for data which show large variations by the RRF
method, such as chrysene and B(a)P+B(e)P, the data obtained by LSLR analy-
sis are more precise. Overall, for the full set of reported components
(40-50 in all), the normalized LSLR data sets proved to be more precise.
It is our experience that the use of normalized least square linear regres-
sion analysis for measurements requiring higher precision than a normal
screening technique is a useful alternative to standard procedures. The
data from screening analyses could also be calculated by this type of
analysis to afford more consistent results.
TABLE 3. COMPARISON OF METHODS BY PERCENTAGE DEVIATION FROM AVERAGE
METHOD/SAMPLE
Component
Fluorene
Carbazole
Dibenzothiophene
Phenanthrene/
Anthracene
Pyrene
Chrysene
B(a)P + B(e)P
Fluorene 1.
Carbazole 5.
Dibenzo- 8.
thiophene
Phenan- 20.
threne/
Anthracene
Pyrene 11.
Chrysene 1.
B(a)P +
B(e)P
AB
4.4
8.1
2.8
6.13
8
18.0
24
RRF
50 - 2.
67-7.
36-9.
02 -24.
37 - 13
09 - 1.
38 - .
RRF
CD
17.1
14.0
4.7
7.9
7.78
17.6
39
INTERVAL
Absolute
Differ-
ence
36 .86
55 1.88
20 .84
46 4.44 1
.68 2.31
73 .65
85 .47
LSLR
AB CD
5.6 7.0
2.9 13.0
7.8 5.0
5.1 4.8
3.1 4.9
5.6 10.0
11 20
BETWEEN COMPONENTS
Absolute
Differ-
LSLR ence
1.07 - 2.04 .97
3.40-4.95 1.55
4.90 - 7.27 2.37
1.96 16.99 5.03
8.69 - 13.27 4.58
1.04-1.49 .45
.34 - .56 .22
LSLR-N
AB
3.8
8.1
2.9
6.1
8.0
6.4
9.5
LSLR-N
1,21 - 1.
4.48 - 5.
6.59 - 7.
15.93 - 17
14.42 - 17
1.09 - 1
.34 -
CD
8.3
11.0
4.77
1.9
2.2
10
17
Absolute
Differ-
ence
84 .63
77 1.29
06 .47
.44 1.51
.39 2.97
.59 .50
.59 .25
530
-------
PURGE AND TRAP ANALYSIS
The next area of difficulty is the analysis of heavily loaded process
streams for volatile organics. The major problem associated with these
analyses is the limited capacity of the adsorbing material to efficiently
adsorb the levels of organic materials present in process streams. The
adsorbent used in our synfuels studies was Tenax resin with a capacity to
adsorb 1 to 2 percent of the weight of Tenax in the trap. The actual
amount of Tenax is generally 0.5 gm or less. A 0.5 to 1.0 percent loading
of this material would be equivalent to 2.5 to 5.0 rag capacity. Above
this level, the Tenax will hold more material but with a dramatic de-
crease in trapping efficiency. Table 4 shows data obtained from a heavily
loaded process stream from the EDS plant at Baytown, Texas. Samples B,
through B,- are the same sample purged at different levels of concentration.
Samples A through F are different samples from the same stream.
TABLE 4. COMPARISON OF PURGE AND TRAP DATA
BY VOLUME PURGED
Quantity Quantity
Purged Calculated Observed Load Actual Load
Sample (mis) (mg/1) on Tenax/(mg) on System (mg)
Bl
B2
B3
B4
B5
A
B
C
D
E
F
0.25
1.00
2.50
5.00
10.00
1.00
1.00
1.00
5.00
10.00
10.00
2890
2769
1700
997
460
2216
2805
2642
1255
523
508
.7
2.7
4.3
4.5
4.6
2.6
2.8
2.6
6.3
5.2
5.1
.7
2.7
6.8
13.5
27.0
2.6
2.8
2.6
13.0
26.0
26.0
As indicated in Table 4, adsorbing capacity has no real effect for
the samples run at 1.0 mis. Above this level, however, the data are
affected significantly. It is important to note that although the overall
content of purgeable materials is extremely high, they are not readily
amenable to analysis by direct aqueous injection because it does not
effectively separate the many components present in these samples. Using
the mass spectrometer as a detector and injecting even several (Jl of
531
-------
aqueous material, there are nanogram to low microgram quantities of many
of the components. At levels of 2 g/1, only 2 |Jg/|Jl are being analyzed by
direct injection.
Table 5 shows the results obtained from a naphtha sample, which should
contain a very high percentage of purgeable materials. This sample was run
as a pure organic. The samples were diluted 1:10 with methanol. Two sam-
ples were run with 100 pi of the diluate injected into 10 mis of H?0. Two
samples were run with 50 [Jl of the diluate injected into a 10 ml aqueous
matrix.
TABLE 5. COMPARISON OF PURGE VALUES FOR NAPHTHA
Sample
Naphtha A
Naphtha B
Naphtha C
Naphtha D
Analyzed
(Ml)
10
10
5
5
Observed
Concentration
(g/D
382
297
584
576
Observed Loading
Tenax (mg/1)
3.82
2.97
2.92
2.86
Actual Tenax
Loading (mg/1)
5.80
5.80
2.92
2.86
The values at a 50 (Jl injection of a 1:10 aliquot are reasonable for
this type of stream, indicating that approximately 75 percent of the com-
ponents in the naphtha streams are purgeables.
We recommend that internal process streams be run at a volume of 0.25
to 0.50 mis of sample for purge and trap analysis except for those streams
for which there is little or no chance that purgeables are present. A
second run can then be made after calculation of an effective column loading.
Also, we recommend that volatile organic materials such as naphtha streams
be run at levels of no more than 50 |Jl of a 1:10 diluate. Analytical
parameters for streams with higher boiling ranges are based on volatile
content. Industry literature can be used as a guide for estimating the
quantity of samples to be analyzed. For process streams, we have found
that sour water streams, including streams from pumps and drums as well as
gas scrubber streams, should be regarded as heavily loaded process streams
for volatiles.
PHENOLICS ANALYSIS
A third problem in analyzing process streams is the high level of phe-
nolic materials present in sour water streams. Toxic gases evolve during
acidification of the aqueous samples. Also, the high levels of phenolics
present in extracts tend to create problems when concentrated for analysis
by gas chromatography (GC) and gas chromatography/mass spectometry (GC/MS).
Very low recovery rates (20 to 40 percent) have been reported in many cases.
The more volatile phenols are extremely difficult to recover quantitatively.
For this reason, EPA has recommended a colorimetric method for the analysis
532
-------
of steam-distillable phenol. The method is fairly specific for phenol; how-
ever, there are other major phenolics present in these streams, such as
cresols, xylenols, and trimethyl phenols. These phenols are not quantita-
tively measured by this method. We have analyzed these streams by direct
aqueous injections as well as colorimetric and GC/MS methods.
Table 6 shows a comparison of the results obtained by GC/MS, colori-
metric, and direct injection analysis for two streams heavily loaded with
phenolics. In both cases, the GC/MS gave lower values than the direct in-
jection GC method. For heavily loaded samples, the GC method has the
advantage of allowing analysis in situ with a minimum of sample handling.
The only possible alteration to this method would be to carefully acidify
to a slightly acid pH (6.0 to 7.0) any samples which were very basic to
assure that all phenolics have been analyzed.
TABLE 6. PHENOLICS BY COMPARATIVE METHODS
(mg/1)
METHOD
Sample
Sample A
Sample B
GC/MS
3500
4300
Colorimetric
2439
16024
Direct Injection
11321
13131
The method we used was a slight modification of Standard Method 510E from
the 14th edition of Standard Methods for the Examination of Water and Waste-
water. 1
ANALYSIS OF THIOCYANATES AND RELATED SPECIES
There are also problems associated with the analysis of sulfide,
cyanide, thiosulfate, and thiocyanate (SCN ) in streams which have a high
sulfide and/or H«S content. These samples were analyzed in accordance
with procedures defined in Manual of Methods: Preservation and Analysis
of Coal Gasification Wastewater.2 We found that the precipitation of sul-
fide from heavily loaded streams as lead sulfide is not easily accomplished
and that the precipitations required several days. In addition, cyanide
can easily be lost by occlusion during the precipitation. The amounts of
lead sulfide were so great that this occlusion is a very real problem.
Dilution of the original samples prior to precipitation was not possible
because the lower limits of the required analytical range preclude dilution.
Another consideration in the cyanide analysis is the equilibrium shown below
It appears that the equilibrium is being forced to the left as sulfur is
removed from the system.
Pb+2 + 2S~2 + CN~=S=:SCN~ + PbS i
533
-------
Theoretically, the analysis of SCN should then become an important param-
eter. However, the same problems occur in this analysis since sulfides
must initially be precipitated and there are high levels of lead sulfide
present. To date, no adequate solution to this problem has been developed.
CONCLUSION
There is currently no all-encompassing methodology available for the
analysis of process, product, and wastewater streams from synthetic fuels
plants. Each type of stream and each individual process stream must be
handled under conditions which will optimize the value and validity of the
data obtained. In our current studies, we are attempting to modify exist-
ing procedures, as appropriate, to provide the most effective analytical
approach. In particular, we are correlating GC/MS and GC data by utiliz-
ing the qualitative data obtained from GC/MS as a guide, then using
capillary gas chromatographic data as the eventual quantitative tool. By
incorporating the specificity of the mass spectral data, we are better
able to quantitate unresolved gas chromatographic peaks. We are also
assessing alternatives to existing methodologies of volatile organics
analysis to obtain a more consistent approach to the problem of heavily
loaded process streams. Finally, we are attempting to modify the precipita-
tion procedures for sulfides to adapt a method which is viable for heavily
loaded process streams.
REFERENCES
1. Standard Methods for the Examination of Water and Wastewater, 14th
edition. Prepared and published jointly by the American Public
Health Association, the American Water Works Association, and the
Water Pollution Control Federation, 1975.
2. Luthy, R.G. Manual of Methods: Preservation and Analysis for Coal
Gasification Wastewaters. Prepared for the U.S. Department of Energy,
by the Environmental Studies Institute, Carnegie, Mellon University,
Pittsburgh, PA, July 1978.
534
-------
SOLVENT EXTRACTION PROCESSING
FOR COAL CONVERSION WASTEUATERS
James R. Campbell and Richard G. Luthy
Department of Civil Engineering
Carnegie-Mellon University
Pittsburgh, Pennsylvania 15213
and
Manuel J. T. Corrondo
Department of Environmental Sciences
New University of Lisbon
Lisboa, Portugal
ABSTRACT
This paper outlines experimental and modeling techniques that are being
used to evaluate solvent extraction processing of coal conversion wastewaters.
The project includes characterization of organic contaminants in slagging fixed-
bed gasification process wastewater, as well as screening studies to evaluate
removal efficiencies for these contaminants. Experiments are also in progress
to measure distribution coefficients for several solvent types with phenol and
representative base- and neutral-fraction aromatic solutes. These experiments
are being performed with both clean water and wastewater systems. Results from
these experiments are being evaluated in light of three techniques for estima-
ting distribution coefficients: Modified regular solution theory as used in
chemical engineering processing, expanded solubility parameter approach as used
in liquid-liquid chromatography, and estimation of octanol-water partition co-
efficient as employed in environmental science. This paper reviews results
obtained to date and explains direction for work during the coming year.
INTRODUCTION
Solvent extraction is a candidate treatment process for reducing organic
contaminants from coal conversion wastewater. Solvent extraction is especially
attractive for treatment of highly contaminated streams where the cost of treat-
ment may be compensated, in whole or in part, by the value of recovered material
and by reduction of loadings on downstream wastewater processing units. Solvent
extraction may also eliminate the need for additional physicochemical wastewater
treatment steps.
Experimental work now in progress is aimed at defining solvent extraction
535
-------
processing characteristics of a slagging fixed-bed coal gasification waste-
water. This work includes tests with both wastewater and clean water systems
as well as theoretical considerations aimed at development of a model to aid
prediction of the fate of organic contaminants during solvent extraction treat-
ment.
SOLVENT EXTRACTION PROCESSING
A typical solvent extraction system is illustrated in Figure 1 (Earhart,
et al., 1977), where it is shown that the process is comprised of three basic
unit operations: (1) an extractor where wastewater and solvent are mixed
and separated, (2) a solvent regenerator where solvent is separated from
extracted solutes for reuse, and (3) a solvent recovery step where residual
solvent is removed from the treated wastewater.
Solvent extractors may be classified as either stagewise contactors,
such as mixer-settlers, or as differential contact extractors, such as a packed
column or rotary disk contactor. Solvent regeneration is usually accomplished
by distillation, and solvent recovery may be achieved by either stripping or
secondary solvent extraction.
Numerous solvents are available for use in solvent extraction systems.
Solvents which have been employed for processing phenolic streams include:
light aromatic oil mixtures, tricresyl phosphate, n-butyl acetate (NBA), di-
isopropylether (DIPE), and methylisobutyl ketone (MIBK). When choosing an
extraction solvent, two of the most important considerations are high solute
distribution coefficient and low aqueous solubility. While aqueous solubility
data are available for most solvents, solvent-solute distribution coefficient
data are available for relatively few compounds, notably phenol and its der-
ivatives. It is especially noteworthy that there is essentially no solute
distribution coefficient data for the variety of base- and neutral-fraction
solutes which may exist in a coal conversion wastewater.
The solute distribution coefficient may be defined as either the equilib
ram ratio of solute mass concentration (C, mg/1) in solvent and water nhases
(Kn')j or as the ratio of solute mole fraction activity coefficients (Y) in
each phase (KD);
= Cs/Cw
i
KD = Yw/Ys
where the subscript s refers to the solvent phase and w to the water phase.
Mass concentration and mole fraction activity distribution coefficients are
related by the ratio of water and solvent molar volumes;
Usually solute distribution coefficient data must be determined from lab-
oratory testing, and this may be a costly and time consuming task. However,
536
-------
PROCESS
WASTEWATER
LOADED
EXTRACTOR
SOLVENT
SOLVENT
REGENERATION
VIA DISTILLATION
[ RECOVERED ORGANIC
f CONTAMINANTS
SOLVENT RECYCLE
SOLVENT
RECOVERY
VIA STRIPPING
CLEANED
PROCESS
WASTEWATER
FIG.I TYPICAL SOLVENT EXTRACTION PROCESS WITH
A VOLATILE SOLVENT
537
-------
time and expense can be saved by using thermodynamic models to estimate solute
mole fraction activity distribution coefficients (K.,). Evaluation of the
accuracy and applicability of thermodynamic models is included in the theoreti-
cal aspects of this project.
PROJECT OBJECTIVES
Work now in progress entails both experimental and theoretical aspects of
modeling the fate of organic compounds during wastewater treatment, with emphasis
on solvent extraction. Specific objectives of the project are outlined below:
1) Characterize organic contaminates in slagging fixed-bed
gasification process wastewater before and after several
steps of bench-scale treatment. This work will include
screening studies with several solvents to evaluate the
efficiency of solvent extraction for removal of phenolics,
as well as for removal of base and neutral fraction aro-
matics. In addition, data from these tests will be used
to observe organic specie removal during ammonia stripping
and biological oxidation.
2) Review the literature for assessment of models to predict
distribution coefficients and compile a listing of the avail-
able experimental data on distribution coefficients for waste-
water contaminants.
3) Perform experiments to measure distribution coefficients for
phenol and representative base and neutral fraction aromatic
compounds in both clean water and wastewater systems. The
effect of solute concentration on the value of the distribution
coefficient will also be evaluated.
4) Assess the applicability and accuracy of distribution coefficient
models by comparing model results with experimental data found in
this study and experimental data reported in the literature.
RESULTS TO DATE
Several study objectives have been completed or are in the process of being
completed. Screening studies have been performed with several solvents to assess
their suitability for extraction of phenolic solutes from raw slagging fixed-
bed coal gasification wastewater. This wastewater was generated from gasifi-
cation of a lignite coal.
As a result of these screening studies, MIBK was selected for execution
of bench-scale treatment tests incorporating solvent extraction, ammonia strip-
ping and biological oxidation. As a part of this study, organic contaminants
were characterized in collaboration with Argonne National Laboratory to assess
removal of acid, base and neutral fraction solutes following solvent extraction-
538
-------
ammonia stripping and biological oxidation. Results of this work were presented
at the 54th Annual Conference of the Water Pollution Control Federation (Luthy,
et al., 1981). A summary of this work is provided in Tables 1 and 2.
Table I shows average solvent extraction treatment characteristics for
MIBK extracted wastewater. MIBK is particularly effective for removal of phe-
nol ics (KD~100), and the data show that in the process of reducing phenolics
to about 5 mgA there is also substantial reduction of TOC, COD and BOD. Bio-
logical oxidation was evaluated by both activated sludge (AS) and powdered
activated carbon-activated sludge (PAC-AS) treatment with 5000 mg/£ PAC. Both
AS and PAC-AS showed good removal of the pollutants shown in Table 1. The bio-
logical oxidation studies showed that solvent extracted wastewater was easier
to treat via AS in comparison with wastewater not pretreated for reduction of
phenolics. Solvent extraction eliminated the requirement for dilution prior
to AS and also reduced wastewater foaming during biological treatment. Solvent
extraction also resulted in lower mass loading of residual organic material
(eg. color, TOC, and COD) in the biological reactor effluent.
GC/MS analysis of acid, base, and neutral fraction organics were performed
on raw condensate, solvent extracted-ammonia stripped wastewater, and AS and
PAC-AS effluents. HPLC analysis were performed after each treatment step for
detection of eleven polycyclic aromatic hydrocarbons. Table 2, which provides
a summary of the analytical results, shows that no organic contaminants could
be detected in the acid, base, and neutral fraction suspended phase samples of
MIBK extracted wastewater. Analvsis of solvent extracted-ammonia stripped
aqueous phase acid fraction samples showed the presence of residual phenol, cre-
sols, and other acid fraction compounds, while base and neutral fraction species
showed mainly low levels of relatively few compounds. Analysis of AS and PAC-AS
treated water showed excellent reduction of those few organic compounds which
remained after extraction and stripping. These results showed that solvent ex-
traction for reduction of phenolIcs offers several wastewater processing advan-
tages for treatment of coal conversion condensates.
This work was followed by a preliminary investigation of thermodynamic
models for the prediction of solute distribution coefficients between water
and an organic solvent for phenol and other aromatic solutes. The results
of this work were presented at the Symposium on Water Management and Pollution
Control for Coal Gasification and Liquefaction, sponsored by the Division of
Environmental Chemistry at the 182nd ACS National Meeting in August (Campbell
and Luthy, 1981). This work showed that most of the experimental solvent ex-
traction studies reported in the literature have focused on phenolic compounds.
No distribution coefficient data were found for base and neutral fraction sol-
utes with solvents normally used for phenol recovery. Furthermore, no dis-
tribution coefficient data was found for tests using actual coal refinery waste-
waters.
A review of the chemical engineering and liquid-liquid chromatography
literature revealed that solvent extraction models which are used in these
disciplines are based on developments evolving from regular solution theory.
These concepts have been applied to several solute-solvent systems, and it was
found that some empiricism is necessary for estimation of certain thermodynamic
539
-------
TABLE 1. SLAGGING FIXED-BED WASTEWATER
TREATMENT CHARACTERISTICS1
Parameter
mg.'l unless
noted
Raw
Wastewater
RA-52
Solvent
Extracted
MIBK
Ammonia
Stripped
Activated
Sludge
PAC/AS
TOC
COD
BOD
Phenolics
Org-N
NH -N
NO, -N
sen*
CN
CM101
,- amn _.
Freon Ext
Alk las CaCO )
Cond
(/imhc/cm)
Color
(Pt-Co units)
1 1.100
32.000
26.000
5.500
115
6.300
<5
120
1.8
0.1
410
20.700
20.000
500
1.950
3.900
2.900
5
51
4.400
<5
1 10
1.5
16,300
18,600
500
1.380
2.980
1.820
3
33
30
<5
105
1.5
0.1
10
850
1,490
700
580
1.340
32
0.1
10
84
40
4
1.4
0.1
<5
175
2.230
500
385
640
30
<0.02
4
20
100
<0.5
1.3
<0.1
<5
72
2,200
<20
Luthy, Stamoudis, Campbell, and Harrison, 1981 Analyses of solvent
extracted samples for TOC, COD and BOD were performed after gentle heating to expel
residual MIBK.
540
-------
Table 2
Concentration (iig/1) of Representative Organic
Compounds Identified in Slagging Fixed Bed
Quench Water at Various Stages of Treatment
Peak
Number
Raw Water
Compound Name
SS
FW
MIBK-Stripped
SS
FW
MIBK-AS
FW
MIBK-AS/PAC
FW
Acid Fractions
20
50
00
110
260
370
Phenol
Methyphenol
Methyphenol
Cj-Phenol
llydroxindan
1-Naphthol
15,100
6,300
16,100
8,890
310
250
3,080,000
427.000
494,000
155,000
3.820
1,150
NT
NT
NT
NT
NT
NT
500
100
80
50
40
40
10
1
1
NT
NT
NT
3
1
1
NT
NT
NT
Base Fraction
10
15
20
110
3 DO
730
Pyrtdine
Toluene
Methylpyridine
Aniline
Azanaphthalene
Azafluorene
3.9
NT
12
27
150
20
14,530
77
7 120
1 f 1 (- \J
6,500
500
15
NT
NT
NT
NT
NT
NT
NT
NT
1.0
NT
NT
NT
NT
NT
NT
NT
NT
NT
0.1
NT
NT
NT
NT
NT
Neutral Fraction
32
35
/O
125
180
330
300
'160
710
790
810
Methylcyclohexane, or
C2-cyclopentene
Benzoni trile
Indene
C3-Thiazole
Naphthalene
Indole
Biphenyl
Acenaphthene
Plienanthrene
Fluoranthene
Pyrene
NO
NO
2300
NO
16,300
150
1100
2180
4680
1510
830
22,750
2,900
910
ND
26,600
5,000
NO
ND
ND
ND
NO
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
4230
NT
NT
31
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
1.8
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
2.0
NT
NT
NT
NT
NT
NT
NT
SS. "Suspended Solids; FW, Filtered Water; ND. Not Determinable; NT. Not Dectected.
-------
parameters. Despite this limitation, our analysis has shown that for MIBK
and water, neutral fraction aromatic solutes are predicted to have distribution
coefficients substantially greater than that for phenol. Thus, in the process
of reducing phenolics from relatively high concentrations to comparatively low
concentrations, it is expected that neutral fraction solutes would be reduced
by 'even-greater'proportions.
CURRENT AND FUTURE WORK
Recent experimental work has been directed towards measuring distribution
coefficients in both clean water and wastewater systems. This work has examined
three solutes: phenol as a representative acid fraction solute, and aniline
and pyridine as representative base fraction solutes. These compounds comprise
the predominate parent chemical species for compounds previously identified in
each of these fractions. It is planned to measure distribution coefficients
for benzene with several solvents, as benzene is the parent specie for neutral
fraction solutes.
Five solvents have been incorporated in these tests, methylisobutyl ketone
(MIBK), di-isopropylether (DIPE), n-butyl acetate (NBA), toluene, and tetradecane
These compounds are representative of major classes of organic solvents. MIBK
is reportedly used in an extraction process licensed by the Chem-Pro Equipment
Corp. (Greminger, et al., 1980), while DIPE is employed in the Lurgi Phenol-
solvan process. NBA shows a relatively high distribution coefficient for
phenol, and it has been proposed for use in dual-solvent extraction systems
(Earheart, et al., 1977). Toluene is a component of coal-devired light oil,
which was widely used at one time for extraction of phenolics from coke plant
ammonia liquor. Tetradecane was included in this study for comparison purposes
because it is an alkane, and because crude oil or related compounds are some-
times involved in petroleum refining operations for extraction of phenol from
water or for washing phenol from refinery products.
Solvent extraction tests were performed with these solvents and solutes, in
single and multiple solute clean water systems as well as in actual wastewater,
to investigate potential synergistic/antagonistic effects. The effect of solute
concentration was also investigated. The results of this work are being sum-
marized in the form of a technical paper.
It is planned to execute another treatment study using slagging fixed-bed
wastewater generated from gasification of another type of coal. Results ob-
tained from this work would be used to verify the previous results, as well as
to assess processing differences for a different water-solvent system. The
tests would also include detailed wastewater characterization at different
levels of phenol removal. These analyses would provide information on whether
various contaminants are removed concomitantly in proportion to their respective
distribution coefficients.
The latter tests are important from an economic point of view. Goldstein
(1981) notes that single stage extraction is less costly than multiple-stage
extraction, and that partial phenol recovery may be economically attractive.
An 80 percent recovery of 5,500 rng/1 phenol ics is reported as supplying enough
energy in the recovered material to run the extraction process. Goldstein also
recommends solvent extraction if phenolic levels are high and if BOD concentra-
tions are greater than 4,000 to 6,000 mg/1.
542
-------
ADVANTAGES OF SOLVENT EXTRACTION
The economic issues regarding cost of solvent extraction versus reduced
cost of additional wastewater treatment are not easily evaluated. However, it
is clear that solvent extraction of phenolic condensates is advantageous for
numerous reasons. Some of these reasons are outlined below.
(a) Solvent extraction removes most base and neutral fraction
solutes. This is significant because many of the toxic or
hazardous organic contaminants in coal conversion wastewaters
are found in these fractions.
(b) Recovered material may be combusted for heat value, and this
heat may be used to drive the extraction process. Properly
designed combusters would destroy hazardous organic compounds.
(c) Solvent extraction would remove hazardous organics, and this
would reduce or eliminate problems with disposal of hazardous
organic sludges formed as a result of wastewater treatment and
reuse.
(d) Solvent extraction would reduce or eliminate problems with carry
over of volatile aromatic hydrocarbons during sour water treat-
ment.
(e) Extraction removes creosotes, and thus it is likely that most
"tar acids" would be removed. This may be particularly important
in treatment of coal liquefaction wastewater, where it is be-
lieved that acid treatment is required prior to biological ox-
idation to precipitate tar acids (Drummond, et al., 1981). Solvent
extraction may eliminate the need for this step.
(f) Solvent extraction can eliminate the need for dilution prior to
biological oxidation. It has been found in various studies that
dilution is required when biological oxidation is employed for
treatment of heavily contaminated gasification or liquefaction
process condensates (Luthy, 1981; Drummond et al., 1981). Our
recent work (Luthy, et al., 1981) has shown that dilution was
not necessary when treating solvent extracted coal gasification
condensate.
(g) Pretreatment by solvent extraction results in lower mass loading
of residual organic material (i.e. TOC, COD, and color) in bio-
logical reactor effluent. Also, foaming was not a problem when
solvent extracted wastewater was subjected to activated sludge
treatment.
(h) Since solvent extraction pretreatment can eliminate the need for
dilution water as well as result in lower loadings of residual
organics, it should benefit any additional treatment required
prior to wastewater reuse.
543
-------
REFERENCES
Campbell, James R., and R.G. Luthy, "Estimation of Distribution Coefficients
for Wastewater Aromatic Solutes", paper presented at Symposium on Water
Management and Pollution Control for Coal Gasification and Liquefaction,
Division of Environmental Chemistry, 182nd ASC National Meeting, New York,
NY, August, 1981.
Drummond, C.J., R.P. Noceti, and Jack G. Walters, "Treatment of Solvent Refined
Coal (SRC-I) Wastewater: A Laboratory Evaluation," paper presented at
the 91st National Meeting of the American Institute of Chemical Engineers,
Detroit, MI, August, 1981.
Earhart, J.P., K.W. Won, J.M. Prausnitz, and C.J. King, "Recovery of Organic
Pollutants Via Solvent Extraction," Chemical and Engineering Progress,
pp 67-73, May 1977.
Goldstein, D.J., "Wastewater Treatment in Coal Conversion" paper presented at
the 54th Annual Conference, Water Pollution Control Federation, Detroit,
MI, October, 1981.
Luthy, R.G., "Treatment of Coal Coking and Coal Gasification Wastewater,"
J. Water Pollution Control Federation. Vol. 53, No. 3, pp 325-339, 1981.
Luthy, R.G., V.C. Stamoudis, J.R, Campbell, and W. Harrison, "Removal of Organic
Contaminants from Coal Conversion Compensates," paper presented at the
54th Annual Conference, Water Pollution Control Federation, Detroit, MI,
October, 1981. Submitted to J. Water Pollution Control Federation.
544
-------
APPENDIX
Attendees
545
-------
Allen
Altschuler
Anastasia
Anderson
Anderson
Atkinson
Ayer
Bailly
Beavers
Bee
Beekley
Bell
Beltran
Berkey
Beychok
Bixel
Blake
Bolton
Bombaugh
Bonkoski
Bosman
Bradshaw
Brna
Brown
Brown
Burns
Burton
Carley
Casamassima
Chartrand
Christopher
Cooper
Corbett
Cowser
Crandell
Crawford
Crawford
Dale
D' Alonzo
Dennis
Dickert
Dounoucos
Drabkin
Drummond
Dubberly
Dunkle
Duprey
El-Ashry
Elms
Ely
Engel
Englick
Erickson
Doreen C.
Morris H.
Loui s
Joseph M.
Rick
R. Dwight
Franklin A
F. David
R. W.
Robert W.
Pamela
Linda R.
Rafael G.
Ed
Milton R.
John C.
John H.
Graham
Karl J.
W. A.
M.
W. N.
Theodore G
Charles W.
Richard D.
E. A.
Ellison S.
Robert
Sal
Dean.
J.
Elliot
William E.
Kenneth E.
Chris
Jay
Kimm W.
John T.
John
Patrick
Charles T.
Angelo
Marvin
Charles J.
Agnes K.
D. M.
Robert L.
Mohamed T.
Bruce R.
Robert B.
Paul
Jack
Ron
145 Vista Avenue
401 M Street, S. W. (RD-681)
4201 W. 36th Street
One Gateway Center
P. 0. Box 2511
2565 Plymouth Road
P. 0. Box 12194
1626 Cole Boulevard
P. 0. Box 5888
20030 Century Boulevard
3401 Lagrande Boulevard
Natural Resources Bldg., P&C Staff
A.A. 4976
127 Alleyne Drive
63 Oak Tree Lane
P. 0. Box 1026
Fifty Beale Street
8150 Leesburg Pike, Suite 600
8501 Mo-Pac Blvd., P. 0. Box 9948
77 Havemeyer
P. 0. Box 19
P. 0. Box 5888
IERL, MD-61
3800 Frederica St.
1820 Dolley Madison Boulevard
P. 0. Box 1620
3344 Federal, Office of Fossil Energy
2420 Alcott Street
P. 0. Box 2180
15740 Shady Grove Road
P. 0. Box 2521
1860 Lincoln Street
8501 Mo-Pac Blvd., P. 0. Box 9948
Building 1505
2420 Alcott .Street
1200 Missouri Avenue
One Space Park Drive, R4/1136
1860 Lincoln Street (8AH-A)
115 Gibraltar Road
2200 Churchill Road
Independence Mall West
P. 0. Box 43, Bldg.. 5, Room 425
2100 H Street, N.W., Suite 607
P. 0. Box 10940
650 South Cherry Street, Suite 400
115 Gibraltar Road
1860 Lincoln Street (8AH)
12075 E. 45th Avenue
9210 South East 68th Street
2244 Walnut Grove Ave., Rm. 492
133 S. Van Gordon St., Ste. 206C
Pasadena
Washington
Chicago
Pittsburgh
Houston
Ann Arbor
Research Triangle Park
Golden
Denver
Germantown
Sacramento
Norris
Bogota D.E.
Pittsburgh
I rvine
Princeton
San Francisco
Vienna
Austin
Stamford
9700 MA Groningen
Denver
Research Triangle Park
Owensboro
McLean
La Jolla
Washington
Denver
Houston
Gaithersburg
Houston
Denver
Austin
Oak Ridge
Denver
Bismarck
Redondo Beach
Denver
Horsham
Springfield
Philadelphia
Schenectady
Washington
Pittsburgh
Denver
Horsham
Denver
Norris
Denver
Mercer Island
Rosemrad
Denver
Mi ssoul a
CA
DC
IL
PA
TX
MI
NC
CO
co
MD
CA
TN
COLOMBIA
PA
CA
NJ
CA
VA
TX
CT
NETHERLANDS
CO
NC
KY
VA
CA
DC
CO
TX
MD
TX
CO
TX
TN
CO
ND
CA
CO
PA
IL
PA
NY
DC
PA
CO
PA
CO
TN
CO
WA
CA
CO
MT
91106
20460
60632
15222
77001
48105
27709
80401
80217
20874
95823
37828
15215
92715
08540
94105
22180
78766
06904
80217
27711
42301
22102
92038
20461
80211
77001
20877
77001
80295
78766
37830
80211
58505
90278
80295
19044
62706
19105
12301
20586
15236
80222
19044
80295
37828
80239
98040
91770
80228
59812
Aerovironment Inc.
U.S. EPA
Institute of Gas Technology
PPG Industries, Inc.
Tennessee Gas Transmi ss ion Corp.
U.S. EPA
Research Triangle Institute
Dames & Moore
Stearns-Roger Engineering Corp.
The Aerospace Corporation
Radian Corporation
Tennessee Valley Authority
University of Los Andes
Synco Consultants Inc.
Consulting Engineer
Mobil Research & Development Corp.
Bechtel Group, Inc.
International Gas Consultancy Ltd.
Radian Corporation
Dorr-Oliver
N.V. Nederlandse Gasunie
Stearns-Roger Engineering Corp.
U.S. EPA
Texas Gas Transmission Corp.
The MITRE Corporation
Systems, Science and Software
U.S. Department of Energy
McLaughlin Industrial Waste Engineers
Exxon Company, U.S.A.
Bechtel Power Corporation
Tri-State Synfuels Company
U.S. EPA, Region VIII
Radian Corporation
Oak Ridge National Laboratory
McLaughlin Industrial Waste Engineers
ND State Dept. of Health
TRW, Inc.
U.S. EPA, Region VIII
IU Conversion Systems, Inc.
Illinois EPA
Rohm and Haas Company
General EJ ectric Company
U.S. Synthetic Fuels Corp.
U.S. Department of Energy
Pace Co. Consultants and Engineers
IU Conversion Systems, Inc.
U.S. EPA, Region VIII
Tennessee Valley Authority
Black & Veatch
Zimpro, Inc.
Southern Cali forn i a Edison Company
FMC Corporation
University of Montana
-------
Felder Richard M. Dept. of Chemical Engineering
Felix W. Dale 329 Building, 300 Area
Ferraro Paul 4210 E. llth Avenue
Ferrell James K. P. 0. Box 5035
Fischer Marshall 1860 Lincoln Street
Fischer William H. P. 0. Box 1498
Freeman Philip G. Box 8213, University Station
French William E. P. 0. Box 880, Collins Ferry Road
Fuchs Michael R. 8501 Mo-Pac Blvd., P. 0. Box 9948
Garbe Yvonne M. 401 M Street, S. W. (WH-565)
Geise Bill 1860 Lincoln Street
Ghassemi Masood One Space Park Drive, R4/2150
Goldgraben Gerald 1820 Dolley Madison Boulevard
Gorshaw Cindy P. 0. Box 2309
Granger Tom Two World Trade Center, 90th Floor
Hafker William R. P. 0. Box 101
Hale, Jr. Matthew 401 M Street, S. W. (TS-794)
Hamanishi Henry P. 0. Box 7808, II M-K Plaza
Hangebrauck Robert P. IERL, MD-61
Harlow George L. 345 Courtland Street
Hayward, Jr. James C. Eritex Building, P. 0. Box 3105
Henschel D. Bruce IERL, MD-61
Higgins Charles P. 0. Box 806
Higman Howard 9190 Red Branch Road
Hoffman Dale A. 134 Union Boulevard
l_n Honefenger Ron L. P. 0. Box 1396
•C* Honerkamp Russell L. 8501 Mo-Pac Blvd., P. 0. Box 9948
^~J Hoskins W. W. 700 S. Ash
Hospodarec Robert W. 3333 Michelson Drive
Howard F. Sidney One Davis Drive
Huff Patricia G. 432 CB Midland Building
Hughes W. David P. 0. Box 68, 695 North 7th
Hull Jeane L. 500 Boyd Building, 18th & Carey Ave.
Janes T. Kelly IERL, MD-61
Jonardi Robert J. 7700 East Arapahoe Road, Suite 330
Jones Fred L. One Woodward Avenue
Jones N. Stuart P. 0. Box 12194
Jones Peter W. 15305 Calle Enrique, Suite D
Juday Nancy L. P. 0. Box 5888
Kalish Robert P. 0. Box 150, Building 3502
Kamperman E. F. P. 0. Box 5440
Kelley Colleen D. P. 0. Box 2100
Kent Raymond L. 9190 Red Branch Road
Kimura S. Gene P. 0. Box 43
King Jeannette W. 2390 S. York
Kingsbury Garrie L. P. 0. Box 12194
Kirchgessner David A. IERL, MD-61
Klein Jerry A. P. 0. Box X
Klein Mor[on J. 10 West 35th Street
Koblitz Jerry H. 7800 E. Union Avenue, Suite 930
Komai Ralph Y. P. 0. Box 10412, 3412 Hillview Avenue
Koump V. Churchil Boro
Kuntz Mike 1860 Lincoln Street
Raleigh NC
Richland WA
Denver CO
Raleigh NC
Denver CO
Reading PA
Grand Forks ND
Morgantown WV
Austin TX
Washington DC
Denver CO
Redondo Beach CA
McLean VA
Englewood CO
New York NY
Florham Park NJ
Washington DC
Boise ID
Research Triangle Park NC
Atlanta GA
Houston TX
Research Triangle Park NC
Franklin TN
Columbia MD
Lakewood CO
Houston TX
Austin TX
Denver CO
Irvine CA
Belmont CA
Cleveland OH
Brighton CO
Cheyenne WY
Research Triangle Park NC
Englewood CO
Detroit MI
Research Triangle Park NC
Morgan Hill CA
Denver CO
Plaquemine LA
Norwalk CT
Denver CO
Columbia MD
Schenectady NY
Denver CO
Research Triangle Park NC
Research Triangle Park NC
Oak Ridge TN
Chicago IL
Denver CO
Palo Alto CA
Pittsburgh PA
Denver CO
27650 North Carolina State University
99352 Battelle Pacific Northwest Laboratory
80220 Colorado Department of Health
27607 North Carolina State University
80295 U.S. EPA, Region VIII
19607 Gilbert/Commonwealth Assoc.
58202 U.S. Department of Energy
26505 U.S. Department of Energy
78766 Radian Corporation
20460 U.S. EPA
80295 U.S. EPA, Region VIII
90278 TRW, Inc.
22102 The MITRE Corporation
80155 Phillips Petroleum Co.
10048 Envirosphere Co.
07932 Exxon Research & Engineering Co.
20460 U.S. EPA
83729 Morrison-Knudsen Company, Inc.
27711 U.S. EPA
30365 U.S. EPA, Region IV
77001 Shell Oil Company
27711 U.S. EPA
37064 MCI/Consulting Engineers Inc.
21045 Hittman Associates, Inc.
80225 U.S. Fish & Wildlife Service
77251 Transco Companies, Inc.
78766 Radian Corporation
80217 Steams-Roger Engineering Corp.
92730 Fluor Engineers & Constructors, Inc.
94002 Lurgi Corp.
44115 Standard Oil of Ohio
80601 Century Refining Co.
82002 Wyoming Office of Industrial Siting
27711 U.S. EPA
80112 Environmental Research & Technology Inc.
48226 American Natural Resources
27709 Research Triangle Institute
95037 Colucci and Associates
80217 Stearns-Roger Engineering Corp.
70764 Dow Chemical USA
06856 UOP/Air Correction Division
80201 Texaco, Inc.
21045 Hittman Associates, Inc.
12301 General Electric Company
80208 Denver Research Institute
27709 Research Triangle Institute
27711 U.S. EPA
37830 Oak Ridge National Laboratory
60076 IIT Research Institute
80237 Espey, Huston & Associates, Inc.
94303 Electric Power Research Institute
15235 Westinghouse Research & Development
80295 U.S. EPA, Region VIII
-------
t_n
-P-
00
Kurey William J. P. 0. Box 295
La fond Timothy J. One Gateway Center
Lagemann Robert C. IERL, MD-61
Larsen Deborah J. Acorn Park
Lethi Minh-Triet 401 M Street, S. W. (PM-221)
Lewis Alan J. 150 Ninth Avenue N.
Lewis Ramilla 2390 S. York
Liao F. Y. P. 0. Box 1026
Lidwin Maria IERL, MD-61
Lienhard H. 666 Kinderkamack Road
Linkhardt Douglas 1860 Lincoln Street
Long Dick 1860 Lincoln Street
Longenberger Fred 1860 Lincoln Street
Lynott William P. 3344 N. Torrey Pines Court #B
Lytle Robert S. 1617 Cole Blvd., Suite 250
Magee Robert A. 8501 Mo-Pac Blvd., P. 0. Box 9948
Martin Joan 1320 17th Street
Maskarinec Michael P. P. 0. Box X
Maxwell Michael A. IERL, MD-61
Mazza Carl 401 M Street, S. W. (TS-794)
McCann Edward W. 5990 S. Syracuse St., Suite 320
Meissner, III Roland E. (Hap) 100 West Walnut
Melcer Henryk 867 Lakeshore Road, P. 0. Box 5050
Menzies William R. 8501 Mo-Pac Blvd., P. 0. Box 9948
Miller Eli Dept. of Environmental Sci. Eng'g
Milliken J. 0. IERL, MD-61
Mirchandani Dilip Two World Trade Center
Miullo Nat 1860 Lincoln Street
Moghissi A. Alar 401 M Street, S. W. (RD-682)
Mohr Volker H. 1140 Avenue of the Americas
Mohr, Jr. Donald H. Department of Chemical Engineering
Moyer Thomas J. 7000 South Broadway
Murphy Andrew P. 0. Box 13109
Murray, Jr. Charles F. One Space Park Drive
Nelson Roger 11455 West 48th Avenue
Neufeld Ronald D. 949 Benedum Hall, Dept. Civil Eng.
Noceti Richard P. P. 0. Box 10940
Notich Mark D. 9151 Rumsey Road
O'Brien Maryln J. 9238 Rawsonville Road
O'Donnell Frank 1828 L Street, N.W., Suite 510
Offen George 485 Clyde Avenue
Orf Dianna 1515 Cleveland Place, Suite 360
Orgeron Paul S. 2150 Shattuck Avenue
Orzynski Stephen J. 1320 17th Street
Parker Gary E. 5660 S. Syracuse Circle, Suite 206
Peel Robert C. 10 Longs Peak Drive, Box 2000
Piepers 0. P. 0. Box 19
Place Barry G. P. 0. Box 512
Ponder Wade H. IERL, MD-61
Poulson Richard E. P. 0. Box 3395, LETC
Princiotta Frank T. IERL, MD-60
Pugsley E. B. P. 0. Box 5888
Raben Irwin A. 130 Sandringham South
Roosevelt
Pittsburgh
Research Triangle Park
Cambridge
Washington
Nashville
Denver
Princeton
Research Triangle Park
River Edge
Denver
Denver
Denver
La Jolla
Golden
Austin
Denver
Oak Ridge
Research Triangle Park
Washington
Englewood
Pasadena
Burlington, Ontario
Austin
Chapel Hill
Research Triangle Park
New York
Denver
Washington
New York
Berkeley
Littleton
Research Triangle Park
Redondo Beach
Wheatridge
Pittsburgh
Pittsburgh
Columbia
Belleville
Washington
Mountain View
Denver
Berkeley
Denver
Englewood
Broomf ield
9700 MA Groningen
Mi Iwaukee
Research Triangle Park
Laramie
Research Triangle Park
Denver
Moraga
UT
PA
NC
MA
DC
TN
CO
NJ
NC
NJ
CO
CO
CO
CA
CO
TX
CO
TN
NC
DC
CO
CA
CANADA
TX
NC
NC
NY
CO
DC
NY
CA
CO
NC
CA
CO
PA
PA
MD
MI
DC
CA
CO
CA
CO
CO
CO
NETHERLANDS
WI
NC
WY
NC
CO
CA
84066
15222
27711
02140
20460
57203
80208
08540
27711
07630
80295
80295
80295
92037
80401
78766
80202
37830
27711
20460
80111
91124
L7R 4A6
78766
27514
27711
10048
80295
20460
10036
94720
80122
27709
90278
80033
15261
15236
21045
48111
20036
94042
80202
94704
80202
80111
80020
53201
27711
82071
27711
80217
94556
U.S. Fish & Wildlife Service
PPG Industries, Inc.
U.S. EPA
Arthur D. Little, Inc.
U.S. EPA
State of Tennessee
Denver Research Institute
Mobil Research & Development Corp.
U.S. EPA
Lurgi Kohle
U.S. EPA, Region VIII
U.S. EPA, Region VIII
U.S. EPA, Region VIII
Energy Resources Co., Inc.
Envirosphere Company
Radian Corporation
Fred C. Hart Associates, Inc.
Oak Ridge National Laboratory
U.S. EPA
U.S. EPA
Hampshire Energy
The Ralph M. Parsons Company
Environmental Protection Service
Radian Corporation
University of North Carolina
U.S. EPA
Envirosphere Co.
U.S. EPA, Region VIII
U.S. EPA
Lotepro Corp.
University of California
Tom Moyer & Associates, Ltd.
Acurex Corp.
TRW, Inc.
Camp Dresser & McKee, Inc.
University of Pittsburgh
U.S. Department of Energy
Hittman Associates, Inc.
State of Michigan
Synfuels Week
Acurex Corp.
Occidental Oil Shale Co.
Tera Corporation
Fred C. Hart Associates, Inc.
Council of Energy Resource Tribes
Rocky Mountain Energy
N.V. Nederlandse Gasunie
Allis Chalmers
U.S. EPA
U.S. Department of Energy
U.S. EPA
Steams-Roger Engineering Corp.
IAR Technology Inc.
-------
Ul
a-
O
T3
O
Rasch
Reeser
Reilly, Jr.
Reveal
Rhodes
Richards
Rigg
Robbins
Rodriguez
Roosmagi
Rosenfeld
Rosenfield
Rowe
Ruppel
Rydell
Schauer
Shankland
Sheesley
Sims
Singer
Smith
Smith
Smith
Soukup
Spaite
Stevens
Stewart
Stoner
Thielen
Thomas
Thomas
Thompson
Thomson
Thornton
Trask
Troscinski
Tucker
Tucker
Vaos •
Voerman
Wallace
Wailine
Weiner
Weinreich
Wetherold
Wetzel
Whipple
White
Wilder
Wilks
Wilson
Witmer
Yee
Yeh
Yen
Yu
Donald
Warner K.
Christopher A.
William S.
William J.
Mike
Jack
Steve M.
Francisco A.
Carl E.
Michael F.
Henry
Gilbert
Thomas C.
Charles D.
J. V.
Bob
David
Ronald C.
Philip C.
G. Richard
N. Dean
Tracey L.
Charles L.
Paul W.
Christopher M.
Gary M.
David W.
Carol J.
Richard L.
William C.
Gerard P.
Sidney
Richard
Jeff
Ed
J. D.
W. Gene
Elizabeth
Garrett
Anna W.
Rob
Larry
Gary
Robert G.
Rolf
Gary
David
Russell
Walter H.
Weston
Fred E.
William C.
Rick
Alan F.
Wen C.
1860 Lincoln Street
5660 S. Syracuse Circle
Building 202, Biological Research
P. 0. Box 10412, 3412
1ERL, MD-61
8500 W. Capitol Drive
1515 Cleveland Place, Suite 360
P. 0. Box 3 (04-107)
A.A. 4976
P. 0. Box 3395, Univi
245 Summer Street
P. 0. Box 2521
P. 0. Box 101
P. 0. Box 10940
1200 Missouri Avenue
P. 0. Box 4320
1860 Lincoln Street
P. 0. Box 3395, LETC
3728 Randell Road
Dept. of Environmental .Sci. Eng'g
363 South Harlan St., Suite 209
IJERL, MD-61
6315 Grand Vista Avenue
4800 Oak Grove Drive, MS 502-310
P. 0. Box 547
2950 South Jamaica Court, //208
8501 Mo-Pac Blvd., P. 0.
P. 0. Box 101
8501 Mo-Pac Blvd., P. 0.
P. 0. Box 2752
3333 Michelson Drive
1660 S. Albion
576 Standard Avenue
2901 Butterfield Road
5790 Old Millington Road
IERL, MD-63
401 M Street, S. W. (PM-208)
1860 Lincoln Street
P. 0. Box 12194
1860 Lincoln Street
9725 E. Hampden
One Woodward Avenue
8501 Mo-Pac Blvd., P. 0.
Molkestrasse 29
P. 0. Box 8361
11455 West 48th Avenue
300 Bear Hill Road
1500 Market Street
1860 Lincoln Street
EP-33, Office of Env. Programs
913 Hamilton Bank Building
P. 0. Box 27125, 830
P. 0. Box 2752
7929 Westpark Drive
Research
Llview Avenue
ite 360
:y Station
:i. Eng'g
ite 209
Power 3555
502-310
, //208
Box 9948
Box 9948
208)
Box 9948
ograms
ng
W. Freeway
Denver
Englewood
Argonne
Palo Alto
Research Triangle Park
Milwaukee
Denver
Houston
Bogota D.E.
Laramie
Boston
Houston
Florham Park
Pittsburgh
Bismarck
Houston
Denver
Laramie
Garner
Chapel Hill
Lakewood
Research Triangle Park
Houston
Rothschild
Cincinnati
Pasadena
Worcester
Aurora
Austin
Florham Park
Austin
Allentown
Irvine
Denver
Richmond
Oak Brook
Memphis
Research Triangle Park
Washington
Denver
Research Triangle Park
Denver
Denver
Detroit
Austin
D-4300 Essen 1
South Charleston
Wheatridge
Waltham
Phi ladelphia
Denver
Washington
Knoxvi 1 le
Houston
Allentown
McLean
CO
CO
IL
CA
NC
WI
CO
TX
COLOMBIA
WY
MA
TX
NJ
PA
ND
TX
CO
WY
NC
NC
CO
NC
TX
WI
OH
CA
MA
CO
TX
NJ
TX
PA
CA
CO
CA
IL
TN
NC
DC
CO
NC
CO
CO
MI
TX
W GERMANY
WV
CO
MA
PA
CO
DC
TN
TX
PA
VA
80295
80111
60439
94303
27711
53222
80202
77001
82071
02107
77001
07932
15236
58505
77210
80295
82071
27529
27514
80226
27711
77001
54474
45213
91109
01613
80014
78766
07932
78766
18001
92730
80222
94572
60521
38127
27711
20460
80295
27709
80295
80231
48226
78766
25303
80033
02154
19102
80295
20545
37902
77027
18001
22102
U.S. EPA, Region VIII
Council of Energy Resource Tribes
Argonne National Laboratory
Electric Power Research Institute
U.S. EPA
Camp Dresser & McKee, Inc.
Occidental Oil Shale Co.
Brown & Root, Inc.
University of Los Andes
Laramie Energy Technology Center
Stone & Webster Engineering Corp.
Texas Eastern
Exxon Research & Engineering Co.
U.S. Department of Energy
ND State Dept. of Health
Shell Oil Company
U.S. EPA, Region VIII
U.S. Department of Energy
North Carolina State University
University of North Carolina
Envirotech Process Equipment
U.S. EPA
Exxon Company, U.S.A.
Zimpro, Inc.
Consultant
Jet Propulsion Laboratory
Riler Stoker Corp.
WATEC Inc.
Radian Corporation
Exxon Research & Engineering Co.
Radian Corporation
International Coal Refining Co.
Fluor Engineers & Constructors, Inc.
Camp Dresser & McKee, Inc.
Chevron Research Company
Nalco Chemical Co.
W. R. Grace & Co.
U.S. EPA
U.S. EPA
U.S. EPA, Region VIII
Research Triangle Institute
U.S. EPA, Region VIII
Rioblanco Oil Shale
ANG Coal Gasification Co.
Radian Corporation
Gesellschaft fur Kohle-Technologie mbH
Union Carbide Corp.
Camp Dresser & McKee, Inc.
EG&G Environmental Consultants
Catalytic, Inc.
U.S. EPA, Region VIII
U.S. Department of Energy
Tennessee Valley Authority
Black, Sivalls & Bryson, Inc.
International Coal Refining Co.
UOP/SDC
------- |