-------
AN ASSESSMENT OF EMISSIONS FROM A HAZARDOUS
WASTE INCINERATION FACILITY
L. J. Staley
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
6. A. Holton, F. R. O'Donnell, and C. A. Little
Health and Safety Research Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
ABSTRACT
The exposure in a nearby population to volatile organic compounds (VOCs) from emis-
sions of a municipal hazardous waste incinerator is determined by measuring the emission
rates and estimating the dispersion of those pollutants over the surrounding community.
Measurements of fugitive VOC emissions from leaky pipe fittings, flanges, pumps, and
valves are made. VOC emissions from the stack and air' pollution control devices are also
measured. An emission assessment is then performed to determine the percentage concentra-
tion and population exposure associated with selected emissions from each source at the
facility. Results indicate that, in this particular case, fugitive and stack emissions
do not contribute significantly to pollutant concentration and population exposure. Fur-
ther, data for specific constitutents of the VOC emissions also show extremely low con-
centrations and exposure, but these results are based on only a partial analysis and ad-
ditional qualitative emission analyses are.required.
INTRODUCTION
Incineration of hazardous wastes pro-
duces stack and nonstack emissions. Until
now, the Environmental Protection Agency
(EPA) has been concentrating on minimizing
stack emissions. EPA has begun to suspect,
however, that stack emissions may not be
the predominant source of air emissions
from hazardous waste incinerators. To bet-
ter characterize all sources of air pollu-
tion from these facilities, EPA has entered
into a three-year interagency agreement
with the Department of Energy at Oak Ridge
National Laboratory (DOE/ORNL) to determine
the relative magnitude and significance of
stack and nonstack emissions, and to ulti-
mately evaluate potential control tech-
niques.
Although .stack testing is not new and
fugitive emissions measurements have been
made at oil refineries and chemical plants,
stack and fugitive emissions measurements
have not been made simultaneously at the
same hazardous waste facility. Since this
is essential for comparison,,EPA decided
to do it at the first available opport-
unity.
This paper presents results of an ex-
periment conducted at the liquid fluid
incinerator operated by Cincinnati's Metro-
politan Sewer District (MSD) in July 1981.
The purpose of the experiment was to deter-
mine for one set of conditions the popula-
tion exposure to hydrocarbons that are
emitted by an incinerator. Hydrocarbon
emission rates from both stack and non-
stack sources were measured. Atmospheric
concentrations were estimated from these
measurements using the Industrial Source
Complex Dispersion (ISC) model (3).
Population exposure to these concentrations
were then calculated. Using toxicity data
for six principal organic hazardous con-
31
-------
stituents (POHCs) contained in the hydro-
carbon emissions, the significance of these
exposures was assessed.
This experiment represents the first
attempt EPA has made to measure both stack
and nonstack emissions simultaneously at
one incinerator. Consequently, it is pre-
mature to draw broad general conclusions
about the relative significance of stack
and fugitive emissions at hazardous waste
incinerators from the work completed thus
far.
Source Term Development
The Cincinnati MSD incinerator is a
two-year old facility with a thermal capa-
city of 3.50 x 1010 cal/hr (139 x 106 Btu/
hr). It consists of a rotary kiln and a
liquid injection furnace in parallel fol-
lowed by a secondary combustion chamber,
quench, .high pressure venturi scrubber, and
packed bed absorber. Nitrogen blanked
tanks, 9.46 x 104 1 (25,000 gallons), allow
for storage of liquid waste which is typi-
cally delivered by tank truck.
Prior to conducting field measurements,
fugitive emission rates were estimated us-
ing emission factors for valves, flanges,
and pumps developed from studies of oil
refineries (8). These calculations indi-
cated that, indeed, fugitive emissions may
greatly exceed stack emissions at 99.99
percent ORE and may be significant at lower
destruction and removal efficiencies.
Results of these calculations are
shown in Figure 1 (7). To gather some pre-
liminary information to determine the vali-
dity of these calculations, EPA monitored
the fugitive emissions at the Cincinnati
incinerator during a trial burn at which
stack testing was conducted.
During the trial burn, approximately
3.41 x 105 1 (90, 000 gallons) of waste
containing chlorinated compounds as well as
alcohols, carbonyl compounds, and CQ - CQ
alkanes from nearby landfills were burned.
Hydrocarbon stack emissions, measured as
methane, were determined using the Dohrmann
DC-50 continuous monitor as a direct flame
ionization detector (1). Fugitive hydro-
carbon emissions were determined by screen-
ing each valve, vent, flange, and pump seal
in the storage, loading dock, and at the
feed inlets to the incinerators using the
Century Systems OVA 108 and OVA 128 port-
able organic vapor analyzers and EPA stand-
ard Reference Method 21 (6). Both instru-
ments were calibrated to methane and in-
dicated hydrocarbon concentrations in ppm
methane. These readings were converted to
emission rates in g/hr using correlations
developed in another study (5).
Modeling Methodology
. The impact of these emissions on the
surrounding population was assessed using
the Industrial Source Complex Dispersion
Model. Together with local meteorological
and popuTat-ioh data supplied by ORNL, the
model can be used to assess the air quality
impact of emissions from area and point
sources associated with an industrial com-
plex. The model assumes Gaussian plume
dispersion and can account for the effects
of elevated terrain surrounding the incin-
erator and building wake effects. Using
meteorological data and incinerator, design,
operation, and emission release data (Table
1), the ISC computer model was run to esti-
mate downwind concentrations at 176 recep-
tors (16 directions and 11 distances) (see
Figure 2). Each receptor point was defined
by the intersection of 1 of 11 rings of
various diameters centered around the in-
cinerator and 1 of 16 radial lines emanat-
ing from the complex center. Each radial
was separated by 22.5° intervals beginning
with due north (0°) and proceeding clock-
wise. The 11 ring diameters were 0.2, 0.3,
0.5, 0.7, -r.0,,'2.0, 3.0, 5.0, 7.0, 10.0,
and 15.0"km.. These distances were chosen
to span!the'greater Cincinnati metropolitan
area.
To account for Cincinnati's hilly ter-
rain,' elevations of each receptor point
were taken from topographic maps. At 91 of
the 176 receptor locations where the eleva-
tion exceeded the stack height, the stack
height elevation less 30.5 cm (1 ft) was
used. Such treatment may be simplified,
but it's conservative in that the concen-
tration calculated at this elevation would
be the maximum experienced at that particu-
lar location.
The concentration calculated by ISC
were than coupled with 1980 census data to
give population exposures (2, 4).
RESULTS
Downwind concentration estimates at
the 176 receptors were calculated for total
32
-------
CO
<
CO
CO
CO
a*
I
UJ
0
3
§ 8
-h—f 1 1 1 h
g .8
s
cvi
S
cn
cn
cn
*
CO
en
en
i
CO
en
UJ
oi
m CM
S2
09
(A U
Z <
e i—
is: to
V)
+*
C8
UJ
a.
US
oa
33
-------
r
M
o
4J
S
S
*rt
o
e
•rl
a
*£*
"
•1-1
i
c
o
.s
o
a>
4J
•"
0)
0)
O
1-1
3
O
w
M-l
o
:=-,
ro
g
3
.
i-H
CJ
t— 1
R)
H
o
^"
o»
•
«
n
t.
o
V)
1
I
3
£
f
«"*
51
i
<
«.
3) •«
si
SI
e>
« j
SI
2s
KS *
•£ (
O«*-
5C
«.
«O
"~
01
a
i
fO
CM
1
i/>
OOO
1 I 1
OOO
XXX
NCMCM
CO CO CO
000
XXX
CO CO CO
«tr «»••*•
CO CO CO
ooo
XXX
CO COCO
r— i—i—
"TTT
ooo
XXX
~~~
CO COCO
(0(0 «0
XXX
COCO CO
in m in
in in in
lit
000
XXX
555
i— r~ r—
CM CM CM
XXX
sss
^s»-
in in m
otot en
\Q Ot CO
coS S
i i i
CM CM CM
in in in
in in in
t— CM CO
i/t
01
0
g
i
•9
CO
i
I
in
i
0
X
s
^
in
X
f,.
^
^-
f
o
X
s
IO
T
o
X
8
CO
in
X
o.
CO
t
o
X
s
in
~0
X
g
iri
1
g
o
1
J
^
J
ia
S
t
£
8
B
^
S
ui £
« ^
1 |
• 0 S
O) T- O
•^ li °
t^" "5 3
0
« S "S
JM- !3 S u
1 § u ^
1 ^ | 1 . C J
<*- Qt t- O -^ • Z
+•> O ^ ZC r-.aQ.i— coaiai
•Sli ' s f ^ s 1
^"^g-SoSSS
EOI >tO4-)i— OO O
^S^Sg^iii
Si-oSfeSSSS
bOU-l— OOI— 3=313=
•Q O *O <8 **-. Oi rf; •*» "r^
34
-------
RECEPTOR GRID
337.5
31S.O
202.5.
22.5
135.0
157.5
FIGURE 2
35
-------
ro
4J
o
o
•*->
03
u
i- O
=3 3Z
(/) t-
O
JC
u c
n) o
O i-
+J
C C
o •
re C
O 3
reo
u. cn
CM
CD
0
O
0
•~
t-
CM
O
10
O
O
rr*
O
CD
§
O
CD
in
cn
o
^.
P;
o
o
CM
J*
•r—
O
10
o
o
o
o
CD
r*.
8
CD
p:
cn
o
vo
CO
CD
0
O
o
CO
in
CM
r—
O
CO
O
o
CO
o
o
o
• AD
0
CD
CD
O
CM
cn
CD
f*^
ID
CO'
CD
8
o
in
co
r—
o
CM
O
CD
CO
O
O
O
O
O
O
CM
CM
cn
O
^.
t^.
CO
CD
O
O
O
•^
JO
o
o
p;
o
o
8
o
vo
o
o
CD
CO
CM
cn
CD
co
cn
o
0
CD
O
0
r—
CO
O
O
O
1**,
0
0
ID
O
CD
O
in
CD
0
o
in
CM
cn
o
_
CM
cn
0
o
o
o
in
36
-------
TABLE 3
LOCATIONS AND VALUES OF MAXIMUM AVERAGE GROUND LEVEL AIR CONCENTRATIONS
POLLUTANT
CHCL3 '
CHCL3
CHCL3
CHO.3
CCL4
CCL4
CCL4
CCL4
C2CL4
C2CL4
C2CL4
C2CL4
HCCP
HCCP
HCCP
HCCP
C2CL6
C2CL6
C2CL6
C2CL6
C6CL6
C6CL6
C6CL6
C6CL6
SOURCE
STACK
STRUCTURE
TANK FARH
ALL SOURCES
STACK
STRUCTURE
TANK FARH
ALL SOURCES
STACK
STRUCTURE
TANK FARH
ALL SOURCES
STACK
STRUCTURE
TANK FARH
ALL SOURCES
STACK
STRUCTURE
TANK FARH
ALL SOURCES
STACK
STRUCTURE
TANK FARH
ALL SOURCES'
CONCENTRATION.
MTPOnnOAIfC PCD
rULKuoKHnb rtK
CUBIC HETER
1.33E-03
5.59E-05
7.32E-04
1.86E-03
1.80E-04
2.27E-05
2.98E-04
4.52E-04
7.82E-04
1.21E-06
1.58E-05
7.93E-04
l,i7E-05
7.31E-08
9.53E-07
1.75E-05
4.54E-05
1.A2E-08
2.12E-07
4.56E-05
4.20E-05
3.21E-09
4.19E-08
4.21E-05
LOCATION OF
DEGREES
FRQH NORTH
348,75
326,25
11.25
348,75
348.75
326,25
11,25
11,25
348,75
326,25
11,25
348.75
348.75
326,25
11,25
348,75
348,75
326.25
11,25
348,75
348,75
326,25
11.25
348.75
CENTROID
METERS
FROM ORIGIN
250,00
250.00
250.00
250,00
250,00
250,00
250,00
250,00
250,00
250,00
250,00
250.00
250,00
250,00
250,00
250,00
250.00
250,00
250,00
250.00
250.00
250.00
250.00
250.00
,.37
-------
TABLE 4
SUMMARY OF EXPOSURES TO 877641. PERSONS AT
CINCINNATI MSD INCINERATOR - POHC RUN
POLLUTANT
CHCL3
CHCL3
CHCL3
CHCL3
ecu
CCL4
CCL4
CCL4
C2CL4
C2CL4
C2CL4
C2CL4
HCCP
HCCP
HCCP
HCCP,
C2CL6
C2CL6
C2CL6
C2CL6
C4CL&
C6CL6-
C6CL6
C6CL6
ALL 1C
ALLHC
ALLHC
ALLHC
SOURCE
STACK
STRUCTURE
TANK FARM
ALL SOURCES
STACK
STRUCTURE
TANK FARM
ALL SOURCES
STACK
STRUCTURE
TANK FARM
ALL SOURCES
STACK
STRUCTURE
TANK FARM
ALL SOURCES
STACK
STRUCTURE
TANK FARM
ALL SOURCES
STACK
STRUCTURE
TANK FARM
ALL SOURCES
STACK
STRUCTURE
TANK FARM
ALL SOURCES
TOTAL EXPOSURE!
PERSON«UG/M«3
3.30EI01
4.33E-01
5.22EIOO
3.87Et01
4.47E+00
1.76E-01
2.13E+00
6.80E+00
1.94EW1
9.35E-03
1.13E-01
1.96E+01
4.16E-01
5.66E-04
4.82E-03
4.23E-01
1.13E+00
1.26E-04
1.51E-03
1.13EWO
1.05E+00
2.49E-05
3.00E-04
1.05E+00
3,60Et04
3.12E+02
3.74E+03
4,01Et04
AVERAGE INDIVIDUAL
EXPOSURE* UG/H»*3
3.75E-05
4.94E-07
5.95E-06
4.41E-05
5.09E-06
2.01E-07
2.42E-06
7.74E-06
2.22E-05
1.07E-08
1.28E-07
2.23E-05
4.74E-07
6.45E-10
7.78E-09
4.82E-07
1.29E-06
1.43E-10
1.73E-09
1.29E-06
1.19E-06
2.83E-11
3.42E-10
1.19E-06
4.10E-02
3.55E-04 ,
4.28E-03
4.57E-02
38
-------
hydrocarbon (HC) releases and for release
values of six specified principal organic
hazardous constituents (POHCs). Concentra-
tions at the 176 receptors were calculated
for each of three sources and total for the
seven pollutants (HC + 6 POHCs). The three
sources were stack emissions, fugitive
emissions from the incinerator structure,
and fugitive emissions from the tank farm
area.
Total Hydrocarbon Air Concentration
For all directions, the maximum con-
centration occurred immediately downwind.
Not surprisingly, the average concentrations
decrease rapidly with distance to less than
2 percent of the maximum at a distance of
12.5 km downwind.
Contribution of Sources to HC Concentration
As one might expect, the relative con-
tribution to total ground level HC by each
of the three release types varies with both
distance and direction from the facility.
Table 2 summarizes the relative contribu-
tion of each of the three sources to total
ground level air concentration of HC as a
function of distance.
POHC Air Concentration
With fugitive and stack data available
for six POHCs, a more detailed emission/ex-
posure assessment is possible Table 3 con-
tains maximum concentration and location
data for six identified POHCs. The largest
POHC concentration is 1.86 x 10->Wj/m3
(3.75 x ID'7 ppm) of chloroform (CHC13)
occurring in the sector segment described
by the centroid location of 250 m from the
incinerator on an angle of 348.75° clock-
wise from due north. This concentration is
very low, as is the concentration of the
most toxic POHC, hexachlorocyclopentadiene
(HCCP), has a maximum concentration of
1.75 x lO-5^ g/m3 (1.54 x 10'9 ppm).
Population Exposure to POHC
Exposures to POHCs are also very low.
Total exposure (person-/4g/m3) ranges from
1.05 for C6C16 to 38.7 for CHC13 while
average individual exposure (person-/tg/m3)
ranges from 1.19 x 1Q-6 for C6C16 to 4.41 x
TO'5 for CHC13. Table 4 also contains the
total exposure and average individual expo-
sure to each POHC from each source. Clear-
ly, most,POHC exposure is due to stack re-
leases, but significant exposure can also
be due to fugitive releases when the POHC
is volatile (CHCL3 and CC14 have the high-
est vapor pressure of the identified POHCs
so their fugitive emission rates are high-
er). For example, over 14 percent of total
exposure to CHC13 is caused by fugitive re-
leases whereas only about 0.03 percent of
total exposure to semi-volatile C5C15 is
caused by fugitive releases.
CONCLUSIONS
Public exposure to HC from this in-
cinerator is low. An assessment of emission
sources shows that from the site 89.8 per-
cent of total exposure to HC is due to
stack emissions, but that for locations
close to the incinerator, fugitive HC
sources may dominate (5). Perhaps, because
they are emitted near the ground, fugitive
emissions have a disproportionately high
effect on close-in population exposure.
The methodology employed to measure
fugitive emissions has limitations in ac-
curacy and composition identification. For
example, actual samples could have been
taken of the gas emitted from fugitive leaks
to verify the chemical composition of the
leak and the leak rate, but shortages of
both time and_ money prevented this. Never-
theless, results of this study are useful.
Not only are future research areas defined,
but also this study is based on measured
data which, with future measurement, can be
used to characterize emissions from hazard-
ous waste incinerators.
The emission assessment used data from
a test burn for a very well run incinerator
which achieved destruction efficiencies
that met or exceeded the 99.99 percent ORE
standard. (In addition, the facility was
new and, therefore, would not be expected to
have as many fugitive emissions as an older
less well maintained facility.) This data
reflects only one day of operation which is
extrapolated to annual operation, a simpli-
fying assumption that may lead to large
source term inaccuracies. The validity of
this assumption needs to be studied. The
possibility of large, short-term releases
due to accidents, spills, and sprays should
also be investigated. If it is assumed that
day to day operation is similar to test
burn conditions, then it seems reasonable
to conclude that normal incinerator opera-
tion at or exceeding the standard levels
produces minimum HC emissions. Conclusions
39
-------
concerning total hydrocarbon exposure re-
quire additional qualitative and quantita-
tive composition analyses.
REFERENCES
1. Ananth, K. P., et al., Midwest Research
Institute, 1981. Work In Progress.
2. Anderson, G. E., 1981. Human Exposure
to Atmospheric Concentration of Se-
lected Chemicals, Attachment B. Sys-
tems Applications, Inc., San Rafael,
California.
3. Bowers, J. F., J. R. Bjorklund, and
C. S. Cheney, 1979. Industrial Source
Complex (ISC) Dispersion Model User's
Guide. EPA-450/4-79-030. H. E. Cramer
Company, Inc., Salt Lake City, Utah.
4. Durfee, R. C., 1981. Work In Progress.
5. Radian Corporation, 1980. Assessment
of Atmosphenic Emissions from Petrole-
um Refining, Volumes 1, 3 and 4. EPA-
68-02-2147, Office of Research and De-
velopment, U.S. Environmental Protec-
tion Agency, Washington, D.C.
6 U.S. Environmental Protection Agency,
1981. EPA Standard Reference Method
21. Federal Register 46(2): 1160-
1161.
7. L. Staley, memo, "Estimation of Fugi-
tive Emissions from Cincinnati's MSD
Incinerator."
8. Emission Factors and Frequency of Leak
Occurrence for Fittingsin Refinery
Process Units, Robert Wetherold and
Lloyd Provost, EPA, February 1979.
40
-------
CHARACTERIZATION OF HAZARDOUS WASTES
GENERATED BY THE PESTICIDE MANUFACTURING INDUSTRY
Bruce A. Tichenor
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Research Triangle Park, N.C. 27711
ABSTRACT
The paper describes a program conducted to characterize waste streams from pesticide
manufacturing processes to determine whether they should be considered as toxic hazardous
wastes under Section 3001 of the Resource Conservation and Recovery Act (RCRA). The
program is being conducted by EPA's Office of Research and Development, Industrial
Environmental Research Laboratory, Research Triangle Parks NC, in support of the Agency's
Office of Solid Waste. The study was initiated by classifying pesticide products and
ranking them by such factors as production volume, toxicity, bioaccumulation, and per-
sistence. The rankings were then used to select those production processes to be
analyzed. For each process selected, an Engineering Analysis is conducted to define
the process chemistry and determine potential hazardous waste streams. These waste
streams are then sampled at an operating pesticide manufacturing plant. The samples
are analyzed for those parameters necessary to make a hazardous waste determination,
such as: chemical composition, corrosivity (pH), reactivity, and EP toxicity. The
analytical results will be combined with the engineering analysis and presented to the
Office of Solid Waste in appropriate background reports.
INTRODUCTION
Purpose
EPA's Office of Research and Develop-
ment j Industrial Environmental Research
Laboratory, Research Triangle Park, NC,
is conducting a study for the Agency's
Office of Solid Waste (OSW) to character-
ize waste streams from pesticide manufac-
turing plants. The purpose of the char-
acterization is to enable OSW to determine
whether specific waste streams should be
listed as toxic hazardous wastes under
Section 3001 of the Resource Conservation
and Recovery Act (RCRA).
The study is being supported by five
contracts: a contract with TRW to perform
engineering, analyses and assist ORD in
overall project coordination and liaison;
three contracts (with TRWj A. D. Little,
and Battelle Columbus Laboratories) to
conduct on-site sampling and laboratory
analyses of pesticide manufacturing waste
streams; and contract with the Research
Triangle Institute to perform QA/QC audits
of the sampling and analysis phases of the
study.
the Study would evaluate all
waste streams within the pesticide indus-
try; however, the magnitude of the indus-
try (92 corporate producers, 141 produc-
tion facilities, 288 pesticide products)*
precluded 100% coverage within the pro-
* Based on 1978 data
41
-------
ject's budget and time frame. Thus,
priorities were established to provide
broad coverage of the pesticide manufactur-
ing industry. It is emphasized that this
selection procedure does not presume that
the waste streams from the processes
selected for evaluation are, in fact,
hazardous, nor is it presumed that other
processes do not produce hazardous wastes
Only a careful technical evaluation can
determine whether a specific waste stream
is hazardous.
INDUSTRY CHARACTERIZATION
As discussed above, it was necessary
to develop priorities for directing evalu-
ations of processes within the pesticide
manufacturing industry. This was accom-
plished in three steps:
1) Classification of Pesticides
2) Ranking of Pesticides
3) Selection of Processes to
be Evaluated
Pesticide Classification
Pesticides can be classified in var-
ious ways, including by:
a) Use, and
b) Chemical Structure
a) Pesticide Use
Classifying pesticides by use results
in the following breakdown:
* Insecticide (I)
• Herbicide (H)
* Fungicide (F)
• Nematicide & Fumigant (N)
• Rodenticide (R)
• Plant Growth Regulator (P)
b) Chemical Structure
Pesticides can be classified by chem-
ical structure into the following
categories:
I. Chlorinated Hydrocarbons
II. Organophosphates
III. Carbamates
IV. Triazines
V. Anilides
VI . Organometallics
VII. General Nitrogenous
VIII. Diene-based
IX. Ureas & Uracils
X. Nitrated Hydrocarbons
XI. Miscellaneous
Since the wastes produced in any
process are a function of process chemis-
try, the classification by chemical struc-
ture was deemed more appropriate than
pesticide use.
Pesticide Ranking
The pesticide processes were ranked
according to the following parameters
[Kelso, et al. (2)]:
• Production Volume
• Acute Toxicity of the Pesticide
• Special Toxicity of the Pesticide
• Acute Toxicity of the Raw Material
• Special Toxicity of the Raw Material
* Persistence of the Pesticide
* Bioaccumulation of the Pesticide
For each of the 288 pesticides iden-
tified, each parameter was scored zero
through four, with zero indicating no
impact and four indicating maximum impact.
The parameter values were then summed to
give a total value for each pesticide
[Harden, et al. (1)].
Selection of Processes
Three criteria were used to select
the pesticide processes to be evaluated:
1) Chemical Classification
For each of the 11 chemical classi-
fications, it was desirable to have
several processes represented.
2) Pesticide Rank
Higher ranked pesticides were given
priority over those with lower scores.
3) Current Production
Many pesticides on the initial list
are no longer produced. Only those in
current production or planned for future
manufacture were considered for evalua-
tion.
42
-------
Using the above three criteria, 22
pesticide manufacturing plants covering
73 pesticide processes were initially
selected for evaluation.
STUDY DESIGN
Each pesticide manufacturing process
selected for evaluation is subjected to a
rigorous analysis involving several
elements:
Engineering Analysis
Using information available in the
literature as well as other information
available through EPA, an engineering
analysis is conducted for each process.
This analysis produces the following
information:
1) Pesticide Properties (e.g., use
and chemical classification,
chemical name and structure,
molecular weight, physical prop-
erties, solubility, and toxicity).
2) Raw Materials.
3) Process Description and Flow
Diagram.
4) Process Chemistry (including
discussions of chemical reac-
tions and side-reactions).
5) Waste Stream Descriptions (includ-
ing potential chemical composi-
tion) .
6) Waste Disposal Methods (based on
current on-site practices).
7) Sampling Recommendations (i.e,
designation of waste streams to
be sampled at an operating
facility).
Pre-Sampling Site Visit
A pre-sampling site visit is made to
each facility to be sampled. The pur-
poses of the visit are:
1) Determine the accuracy and com-
pleteness of the engineering
analysis. (The analysis is sent
to the facility for review prior
to the visit.)
2) Determine accessibility to
sampling locations via an in-
plant evaluation.
3) Answer questions raised by
industry personnel.
4) Determine sampling dates con-
venient to the plant and consis-
tent with production schedules.
Sampling and Analysis Plan
Subsequent to the pre-sampling site
visit, necessary corrections are made to
the engineering analysis, and a sampling
and analysis plan is prepared covering
the following for each process:
• Process Flow Diagram
(revised as necessary)
• Identification of Sampling
Locations
• Specification of:
- Sample Code Numbers
- Sampling Devices and
Containers
- Sample Volumes and Repetitions
- Chain of Custody Procedures
- Safety Requirements
• Specification of Analytical
Procedures
- Defined for Each Sample
- Referenced to Protocol
Manual*
This plan, prepared by the Sampling
and Analysis Contractor, is reviewed by
OSW and ORD. When approved, it is sent
to the pesticide plant prior to the
actual sampling.
On-Site Sampling
Sampling is conducted at the facil-
ity in accordance with the Sampling and
Analysis Plan. The samples are handled
and shipped in accordance with regula-
tions regarding transportation of
hazardous materials.
+ "Sampling and Analysis Procedures for
Pesticide Manufacturing Wastes," pre-
pared for EPA by A. D. Little, Inc.,
under Contract No. 68-02-3111, Task
Directive No. 123.
43
-------
Laboratory Analysis
PROGRAM STATUS
The samples are sent to the contrac-
tor's laboratory and analyzed according
Co the Sampling and Analysis Plan. The
Protocol Manual defines the analytical
procedures; however, modifications to the
procedures may be required depending on
the nature of the wastes. Parameters to
be determined include chemical composition,
corrosivity, reactivity, and EP toxicity.
The range of analytical procedures includes
pH, TOG, SSMS, ICAP/AA, HPLC, GS/FID,
GS/HS, FTIR. The procedures selected for
each sample depend upon the expected
composition.
QA/QC
The study is being conducted under a
Quality Assurance/Quality Control Program
developed and directed by the Technical
Support Staff (TSS) of EPA's Industrial
Environmental Research Laboratory, Research
Triangle Park, NC. Each of the three
sampling and analysis contractors is sub-
ject to QA/QC audits by the Research Tri-
angle Institute, TSS's QA/QC contractor,
including:
• Sampling Audits
• Laboratory Performance Audits
• Data Validation Audits
In addition, each sampling and analy-
sis contractor has prepared and revised a
QA/QC plan which has been approved by
TSS.
The study was initiated in April 1981.
Process/plant selection was completed in
June, at which time contacts were made with
the selected plants and confidentiality
agreements negotiated between the con-
tractors and the manufacturers. The ini-
tial pre-sampling meeting was conducted in
September, and the first samples were
collected in October. As of December,
preliminary engineering analyses had been
completed for 6 plants/22 processes;' pre-
sampling visits made to 4 plants/16 proc-
esses; and sampling and analysis completed
at 2 plants/4 processes.
REFERENCES
1. Harden, J. M., A. J. Kaufman,
S. V, Kulkarni, and J. A. Kezerle.
1981. Pesticide Manufacturing Over^-
view. EPA Contract 68-02-3174, Task
No. 28, U.S. EPA, Industrial Environ^-••
mental Research Laboratory, Research
Triangle Park, N.C. 62 pp.(Unpublished)
2. Kelso, G. L., R. R. Wilkinson, •
J. R. Malone, Jr., and T. L. Ferguson.
1978. Development of Information on
Pesticides Manufacturing for Source
Assessment. EPA-600/2-78-100 (NTIS
No. PB 283051), U.S. EPA, Industrial
Environmental Research Laboratory,
Research Triangle Park, N.C. 413 pp.
PROGRAM OUTPUTS
The study will result in the delivery
to OSW of a background report for each
pesticide process evaluated. These reports
will include engineering analyses, the
results of the sampling and analysis
effort, and, if available, supplementary
information gathered in response to RCRA
Section 3007 questionnaires sent to the
pesticide producers by OSW. The reports
will be used by OSW as input to the pro-
cess of determining whether or not a
specific waste stream should be listed as
a toxic hazardous waste under Section '3001
of RCRA. The reports will also be used
as supporting information for the waste
management studies being conducted by
OSW.
44
-------
OVERVIEW OF INDUSTRY STUDIES PROGRAM
by
Francine Sakin Jacoff
U.S. Environmental Protection Agerjcy
Office of Solid Waste
Washington, D.G. 20460
INTRODUCTION
The Environmental Protection Agency is
conducting studies of selected industries
tto establish an extensive information base
with .regard to the generation and manage-
ment of hazardous waste. The goal of the
Agency is more effective regulation of such
wastes. In a sense, the Industry Studies
will become the basis for the next major
phase of regulations under the Resource
Conservation and Recovery Act of 1976
(R.CRA), as amended.
EPA promulgated phase I of the hazard-
ous waste regulations on May 19, 1980.
This initial phase defined and listed
hazardous wastes and established require-
ments .for generators, transporters, and
treatment, storage, and disposal facili-
ties. Despite the scope of these regula-
tions, they still constitute only the first
major step toward the comprehensive regula-
tion of hazardous waste. EPA promulgated
phase II of the regulations in January and
February of 1981. the phase II regulations
set technical standards for specific types
of hazardous waste facilities. These tech-
nical standards provide the basis for the
issuance of permits to these facilities.
EPA also announced a third phase of
ru-lemaking. This will involve further
.expansion and refinement of the lists of
hazardous wastes and the promulgation of
facility standards tailored to the degree
and type of hazard posed by specific wastes
pr industries. The hazardous wastes Indus-
try Studies will form the major information
b'ase for these tailored regulations.
OBJECTIVES
The objectives of the Industry
Studies program are to:
fulfill the mandate of RCRA Section
3001 which requires that EPA identify
and characterize hazardous waste
streams that should be subject to
RCRA controls:
. develop waste management guidance
and/or regulations tailored to degree
and type of hazard;
« foster recycling/recpvery/treatment
practices as alternatives to land
.disposal; „
. provide the decisionmaking infor-
mation necessary to accomplish an
. integrated approach toward regulation
of industry under RCRA and other.
acts.
APPROACH
The effective development, refinement,
and implementation of the regulatory pro-
gram in the future requires the rapid
establishment of an extensive information
base on the generation, composition,
management and ultimate fate of wastes
produced by industry. To this end, the
Industry Studies program consists of two
related components; a waste character-
ization component and a waste management
assessment component.
45
-------
The data gathered under the waste
characterization component will determine
the waste streams of concern from specific
product/production processes for listing
under 40 CFR Part 261.32 of Section 3001
of RCRA. The waste management component
will assist in the waste listing effort
by characterizing existing practices.
Additionally it will develop an informa-
tional framework from which decision
makers can select strategies for tailoring
of management standards. This framework
will provide an integrated approach to
the type and degree of hazard posed by
particular waste/environment/technology
situations by means of a cross cutting
analysis of various alternatives. The
analysis will provide the necessary
elements to fulfill the requirements of
Executive Order 12291 which calls for a
Regulatory Impact Analysis (RIA) for
proposed regulatory actions.
STATUS
Studies are currently underway in
various industry segments of the organic
chemical industry including:
• Industry Organic Chemicals
• Organic Chemical Products
- Pesticides
- Dyes
- Pigments
46
-------
THE INDUSTRY STUDIES PROGRAM: SYNTHETIC ORGANIC CHEMICALS INDUSTRY
by
Ronald J. Turner and Robert A. Olexsey
U. S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45268
INTRODUCTION
The Resource Conservation and Recovery
Act (RCRA) of 1976, under Section 3001,
requires EPA to identify and list hazardous
wastes. EPA's Office of Solid Waste (OSW)
is currently pursuing studies to determine
the waste streams from organic chemical
production processes which should be listed
as toxic under 40 CFR Part 261.32 of RCRA.
EPA's Office of Research and Development
(ORD) is providing assistance to OSW in
this industry studies program. The industry
segments assigned to ORD include pesticides,
nohchlorinated industrial organic chemicals,
and dyes and pigments. This paper summa-
rize the approach and progress to date on
the study of the synthetic organic chemicals
industry (SOCMI) being conducted by EPA's
Industrial Environmental Research Laboratory
(lERL-Ci).
APPROACH
In the development of: a program to
assess the environmental impact of haz-
ardous wastes from the chemical industry,
it has been necessary to (1) break down
the industry into manageable segments for
study and (2) determine for each segment
which product/processes are potential
sources of hazardous wastes. Table 1
lists the industry breakdown into segments.
It was agreed with OSW that lERL-Cincinnati
would first address the industrial organics
and the organic dyes and organic pigments
segments of the industry; IERL-RTP would
address pesticides. Other industry groups
would be considered as funds become avail-
able .
In view of the magnitude of the indus-
trial organics industry, (roughly 400
individual product/process configurations) ,
it was important to establish an approach
which would reduce the need for a plant
sampling and analysis program to the abso-
lute minimum since S&A is by far the most
expensive phase of any such program. It
was agreed that this approach would comprise
the following steps:
1. Screening of the industrial organic
product/processes to establish a basis
for a priority ranking.
2. Selection of an initial 35 product/
processes for detailed engineering
analysis to determine: 1) if there are
sufficient data for listing the wastes
as hazardous; 2) if there are insuffi-
cient data but there is reason to pur-
sue gathering additional data for
listing purpose; and 3) there is suffi-
cient reason for removing that product/
process from further concern. This
selection was done by lERL-Ci staff
with OSW concurrence.
3. Those product/processes from the 35
which remain of concern would be sub-
jected to site visits or supplemental
questionnaires, and subsequent sampling
and analysis where appropriate as per
the logic diagram described in Figure
1.
Step 1 was carried out by IT Enviro-
science using the data they had developed
in their major study for OAQPS on the
organic chemicals industry. The ITE report
lists product/processes ranked on the basis
of the total quantity of non-aqueous wastes.
It was agreed with OSW that selection of the
initial 35 product/processes for detailed
engineering analysis would be based on that
list. The following additional qualifica-
tions also were specified by OSW:
47
-------
TABLE 1.SUMMARY OF DATA ON THE ORGANIC CHEMICALS PRODUCTS INDUSTRIES^1)
Industry
Basic Petrochemicals
Industrial Organics
Plastics & Resins
Synthetic Fibers
Synthetic Rubber
Plasticizers
Organic Dyes/Pigments
Pesticides
Pharmaceuticals
Surfactants
Specialty Organics
Flavors/Fragrances
Gum/Wood Chemicals
Fats & Oils
Production
Chemicals
(1)
11
398
79
175
17
150
1,000
200
100
ND
ND
ND
ND
ND
Processes
(#)
28
600
20
9
18
3
48
37
25
Companies
' (#)
34
• 260
323
60
71
60
58
104
300
Facilities
(#)
107
544
400
149
141
78
87
115
525
Volume
106 #/yr
67,000
139,000
22,000
6,000
4,000
2,000
350
1,300
1
comprises rough estimates based largely on 1976-7 information contained in
"Industrial Process Profiles for Environmental Use"
ND • No data
48
-------
-------
1. Chlorinated hydrocarbons were not
included since these are covered by a
separate OSW study.
2. Product/processes already covered
adequately by the RCRA hazardous
waste listing were not included.
STATUS
As of the date of this paper, prelimi-
nary engineering analyses have been com-
pleted for most of the 35 initial product/
processes arrangements. (Table 2). These
preliminary engineering analysis reports
describe the industry and the production
processes and identify plants that produce
the subject product. (Table 3). Available
data on waste streams of concern is pre-
sented in the reports.
In addition, a prototype questionnaire
has been prepared for the pesticide irSdustry
segment. Based on this prototype, question-
naires tailored to each industry segment
will be sent to specific plants. This
information gathering is done under the
provisions of Section 3007 of RCRA, which
authorizes EPA to obtain data on processes
and wastes pursuant to the listing process.
The prototype questionnaire is currently
being reviewed by OMB for approval prior
to being sent by EPA to targeted industrial
concerns.
A prioritized list of facilities to
be visited has been prepared for each
product/ process. Site visits are being
scheduled. Table 4 summarizes the current
program status.
The outputs from this study will be a
series of background reports that will
provide information to the Office of
Solid Waste to be used in its decision
process for listing product/processes.
50
-------
TABLE 2.FIRST GROUP OF PRODUCT/PROCESSES RECOMMENDED FOR ENGINEERING ANALYSIS
(TOTAL NON-AQUEOUS WASTE, Mg/yr - ITE)
(1) Propylene oxide via saponification of propylene chlorohydrin (1,550,000 Mg/yr)
(2) Epichlorohydrin via ally! chloride chlorination (179,000 Mg/yr)
(3) Glycerin via allyl chloride, hydrogenation of acrolein, and isomerization of
propylene oxide (113,000 Mg/yr, 5,000 Mg/yr and 4,496 Mg/yr)
(4) Propylene oxide via peroxidation of ethylbenzene (23,000 Mg/yr)
(5) Acrylic acid and acrolein via propylene oxidation, and acrylic esters via direct
esterification of acrylic acid (21,000 Mg/yr, 6,000 Mg/yr and 9,000 Mg/yr)
(6) Caprolactam via cyclohexanone (11,000 Mg/yr)
(7) Adipic acid via cyclohexanol oxidation, and cyclohexanol/cyclohexanone via
cyclohexane oxidation (400,000 Mg/yr)
(8) Crude terephthalic acid via oxidation of p-xylene and dimethyl terephthalate via'
p-xylene oxidation and esterification (213,000 Mg/yr and 192,000 Mg/yr)
(9) Ethylene oxide/ethylene glycol via ethylene oxidation (56,000 Mg/yr and
106,000 Mg/yr)
(10) Acrylamide (all processes (1,700 Mg/yr)
(11) Bisphenol A via phenol/acetone (2,700 Mg/yr)
(12) Oxo-alcohols, including n-butyl alcohol, 2 ethylhexanol (oxo-process) (43,000 Mg/yr)
(13) Butadiene via furfural extraction and other extraction process (28,000 Mg/yr)
(14) Phenol and acetone from cumene (154,000'Mg/yr)
(15) Cumene via benzene (1,100 Mg/yr)
(16) Benzene via toluene hydroalkylation (23,000 Mg/yr)
(17) Acetic acid via butane oxidation (131,000 Mg/yr)
(18) Cyanuric chloride via HCN chlorination (128,000 Mg/yr)
(19) Hexamethylene diamine via adipbnitrile and adiponitrile via butadiene (65,000 Mg/yr)
(20) Ethylamines via ethanol ammoholysis (31,000 Mg/yr)
(21) Ethyl benzene/styrene via benzene (160,000 Mg/yr and 1,900 Mg/yr)
(22) Maleic anhydride via benzene and n-butane (23,000 Mg/yr and 3,000 Mg/yr)
(23) Methyl methacrylate via acetone cyanohydrin (1,000,000 Mg/yr)
(24) Toluene diisocyanate via toluene (116,000 Mg/yr)
(25) Methylene diphenly diisocyanate via aniline condensation and phosgene via
chlorination of carbon monoxide (181,000 Mg/yr and 34,000 Mg/yr)
(26) Phenol via toluene (4,343 Mg/yr)
(27) Cyclohexanol via hydrogenation of xylene (11,486 Mg/yr)
(28) Carbon disulfide (15,436 Mg/yr)
(29) Lead alkyls (726 Mg/yr)
(30) Isophthalic acid via oxidation of xylene (2,996 Mg/yr)
(31) Ethylene from natural gas and heavy liquids (307,812 Mg/yr and 205,208 Mg/yr)
(32) Methyl amine via ammonolysis of methanol (5,866 Mg/yr)
(33) Dimethyl hydrazine (4,485 Mg/yr)
(34) Ethylene diamine via ammonolysis of ethylene dichloride (2,978 Mg/yr)
(35) Acetic anhydride (5,599 Mg/yr)
51
-------
TABLE 3:PRODUCT/PROCESS REPORT CONTENTS
INDUSTRY DESCRIPTION
PROCESS DESCRIPTION
INTERMEDIATE PROCESS STREAMS, PROCESS WASTES,
BY-PRODUCTS
DISCUSSION ON AVAILABLE INDUSTRY INFORMATION
. types and quantities of wastes generated
• identification of known waste stream constituents
• pollution control equipment
• waste management practices
TABLE 4.STATUS OP SYNTHETIC ORGANIC CHEMICALS STUDY (3/82)
Completed
1) Development of Preliminary Engineering Analyses for 35 product/processes
2) Drafted Questionnaires for 35 product/processes
3) Evaluated Industry Responses for 2 product/processes
4) Assigned 10 additional product/processes for Engineering Analysis (12/81)
To Be Scheduled
1) Pre-sampling Plant visits
2) Sampling and Analysis as Required
52
-------
THE INDUSTRY STUDIES PROGRAM: The Organic Dyes and
the Organic Pigments Industry
Robert Olexsey
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Cincinnati, Ohio: 45268
Yvonne M. Garbe
U.S. Environmental Protection Agency
Office of Solid Waste
Washington, D.C. 20460
INTRODUCTION
The Resource Conservation and Recovery
Act (RCRA) of 1976, under Section 3001
requires EPA to identify and list hazardous
wastes. EPA's Office of Solid Waste (OSW)
is currently pursuing studies to determine
which waste streams from particular pro-
duction processes of organic chemicals
should be listed as toxic under 40 CFR •
Part 261.32 of RCRA. EPA's Office of
Research and Development (ORD) is providing
assistance to OSW in this industry studies
program. The industry segments assigned
to ORD include pesticides, non-chlorinated
industrial organic chemicals, dyes and
pigments. This paper summarizes the
approach and progress to date on the study,
of the organic dyes industry and the or-
ganic pigments industry being conducted by
EPA's Industrial Environmental Research
Laboratory in Cincinnati, Ohio (IERL-Ci).
APPROACH
The organic dye and organic pigment
industries are made up of approximately 70
plants which manufacture on the order of
2000 products. Most products are of rela-
tively low volume compared to synthetic
organic chemicals (SOCMI) products, but
there are indications that there may be
toxic waste streams produced from these
batch process operations.
The approach to this waste survey program
was to delineate a group of chemical
classes which included those organic dyes
and pigments which are most likely to have
associated wastes of concern, based on
engineering judgment and existing toxico-
logical information.
The chemical classes which have been
selected include the following:
1. Azo and Azoic dyes and pigments
including the subclass benzidine and
benzidine congener dyes and pigments.
2. Anthraquinone dyes and pigments
3. Stilbene dyes and pigments, including
fluorescent brighteners
4. Sulfur dyes and pigments
5. Phthalocyanine dyes and pigments
6. Poly-aryl methane dyes and pigments
7. Methine and polymethine dyes and
pigments
8. Xanthene dyes and pigments.
It is estimated that 80$ of the total
production volume of organic dyes and
organic pigments is covered in the Azo and
Anthraquinone classes. •
53
-------
Selection of the number of product/
processes for each class will vary with the
intention of choosing a representative
group.
It is very possible that the best
waste candidates for listing under RCRA
will exist as a result of manufacture of
intermediates. Synthesis of intermediates
frequently involve use of chlorinated
hydrocarbon solvents which when reclaimed
could yield hazardous still bottoms.
Other waste concerns involve the many
amine intermediates that are toxic or are
suspected carcinogens. Benzidine from
benzidine dye manufacture was the first
major documented occupational cancer
causing problem in the chemical industry.
A waste which contains significant amounts
of benzidine or its congeners would be
considered potentially hazardous.
CURRENT PROJECT STATUS
The waste survey program for the or-
ganic dye and tb,e organic pigment indus-
tries is being conducted by three con-
tractors. »0ne of the contractors, SRI, is
providing industry background documents for
each of the dye and pigment chemical
classes. The other contractors, TRW and
6CA, are responsible for conducting the
plant surveys and subsequent sampling and
analyses.
TRW and GCA are now in the final
stages of making initial industry contact
by verifying the number and types of dyes
and/or pigments produced at selected
plants. Follow-up contact (e.g., site
visits, questionnaires, and/or sampling and
analyses) should begin early spring 1982.
54
-------
FIGURE 1
Dye and Pigment Plants
Hazardous Waste Characterization Program
Logic Diagram
Select Plants to Cover
Chemical Classes
I
Contractor
Literature search
& prel iminary
engineering analysis
s
X
Project Officer
initiates plant
contact
f
Reject
Plant
Plant Visit
Set up secrecy agreement
Contractor
Preliminary draft. Contact
by telephone to verify current
product/process
Select product/process wastes of concern
Information
inadequate
Subsequent Sampling & Analysis
of Waste
Information
adequate
Listing Background Document
55
-------
INCINERATION TECHNOLOGY FOR SELECTED
SMALL QUANTITY HAZARDOUS WASTE GENERATORS
Victor S. Engleman and D. L. deLesdernier
Science Applications, Inc.
La Jolla, CA
and
Sidney F. Paige
JRB Associates
McLean, VA
ABSTRACT
The purpose of this study is to assess the application of incineration technology for
destruction of small quantities of hazardous wastes generated by hospitals, universities,
research stations, and related industrial sectors. The work includes characterization of
the hazardous waste, geographic locations and nature of activities of generators. Quan- ,
titles and applicability of available incineration technology are assessed for selected
cases. A number of incinerator types were examined in the study and selected hazardous
waste generators were contacted to determine their needs, problems, approaches, and
experience with incineration of toxic and hazardous wastes. The generators included in
this study represent a diversity of waste types as well as quantity and frequency of
generation. The special problems that this diversity provides for the application of
incineration are addressed in the paper.
INTRODUCTION
While small volume generators (SVG's)
of hazardous wastes (with certain except-
ions, those that produce less than 1000 kg/
mo) are exempt from many of the provisions
of RCRA, the problems they face in disposal
of these wastes are similar to those faced
by large volume generators. Since many
SVG's depend for waste disposal on contract
haulers and other outside organizations who
themselves are regulated by RCRA, they
still must comply with many of its require-
ments.
Hospitals, universities and research
organizations represent a particularly
challenging set of problems for hazardous
waste disposal. They generate a variety
of hazardous waste types from pathological
and infectious wastes to inorganic and
radioactive wastes. In addition the waste
mix can vary substantially over a period
of time depending on the nature of activi-
It
ties at the facility.
Increasing numbers of articles have
appeared recently in newspapers concerning
illegal dumping of hospital wastes and
improper disposal .of infectious wastes.
is not always clear from these articles
whether the problem is that these wastes
are placed in a proper landfill but are
properly sterilized, or that they are being
dumped in improper locations. However, the
use of dumps or landfills for wastes of
this type does not represent an ultimate
solution since the wastes have not bee'n
destroyed. On the other 'hand, incineratiohj
when effective, can convert hazardous
materials to innocuous forms. However, in'-
cineration is hot applicable to all wastes
and may not be applicable to specific faci-
lities. The purpose of this study is to
assess the application of incineration tech-
nology for destruction of small quantities
of hazardous wastes generated by hospital's,
universities, and research facilities.
. 56
-------
CHARACTERIZATION OF WASTE
The wastes produced by the three types
of generators under consideration generally
fall into ten categories: infectious wastes,
pathological wastes, carcinogens, waste
Pharmaceuticals, waste solvents and oils,
pesticides, explosives, radioactive wastes,
acids and alkalis, miscellaneous organic and
inorganic chemicals. Infectious wastes are
contaminated with disease-carrying agents
and therefore require special' handling so
that they are not disseminated in the com-
munity. The moisture content and heating
value varies considerably from burnable
solids to infected liquid specimens. Path-
ological wastes, which are sometimes group-
ed with infectious wastes, are not specifi-
cally infectious but are subject to bio-
logical decay and may become a breeding
ground for disease carriers. These wastes
are generally high in moisture content and
require auxiliary fuel for combustion.
Carcinogens are hazardous compounds and,
while many are combustible, they require
special handling to prevent dispersion
before combustion and they must be incin-
erated to a high degree of destruction.
Carcinogens are generated in relatively
small quantities. Waste Pharmaceuticals
which are more likely to be generated by
pharmaceutical' industries than by the sub-
ject SIC's, are generated by hospitals in
small quantities and consist of drug resi-
dues and waste pills and capsules. Con-
trolled substances may be subject to special
accountability. Solvents and oils are
generally flammable liquids consisting of
hydrocarbons, chlorinated hydrocarbons, and
sulfur-containing solvents. The value of
these materials makes them attractive for
recovery and reuse if the waste mixture
does not contain too many components. Waste
pesticides are usually generated in low
volume or consist of spent containers.
These are generally organic materials and
include chlorinated hydrocarbons and phos-
phorus compounds. These are incinerable in
special equipment with flue gas scrubbing
but are generally not recommended for in-
cineration in standard equipment. Explo-
sives likewise are incinerable in special
equipment, but handling and cleanup require-
ments make these generally unattractive for
incineration. Radioactive wastes are con-
trolled by special licensing procedures and
Will not be addressed here. While many low-
level radioactive wastes are incinerable,
the incineration process does not change
the radioactivity of the substances, and
special handling and effluent controls would
be required. Acids and alkalis are general-
ly liquid wastes.that are corrosive and
poisonous. They are most easily handled by
neutralization and are not generally amena-
ble to incineration. A wide variety of
other organic and inorganic substances may
be generated at research facilities depend-
ing on the specific nature of ongoing
research.
GENERATION OF WASTE
The applicability of incineration may
depend not only on the characteristics of
the waste generated but also on the quanti-
ties and frequency of waste generation and,
to a certain extent, on geographic location.
The generation of consistently large quanti-
ties of combustible hazardous waste makes
on-site incineration more attractive pro-
vided local regulations can be met.
Only limited data are available on the
quantity and frequency of waste generation
and the data base is not sufficient to pro-
vide precise quantitative information. How-
ever, it appears that qualitative and direc-
tional information can be derived from the
data. The information below was drawn pri-
marily from Reference 1 with cross-checks
from a number of facility contacts made
under this study. While in some cases the
number of establishments contacted in this
study exceeded the number on which the
statistics in Reference 1 were based, the
information was generally in qualitative
agreement.
For the purposes of this study, small
volume generators are those which produce
less than 1000 kg/mo of hazardous waste.
While RCRA controls certain substances at
lower generation rates, the 1000 kg/mo value
was used to examine the generation data
base. To put things in perspective for the
three types of waste generators under con-
sideration, SIC 28-Chemicals and Allied
Products produces a total If 1500 x 106
kg/mo of hazardous waste. Small volume
generators within SIC 28 (those producing
less than 1000 kg/mo produce a total of 4.5
x 106 kg/mo.
Small volume generators in the subject
SIC codes produce the following quantities
of hazardous waste:
• SIC 806 - Hospitals
0.8 x 106 kg/mo
• SIC 8071 - Medical Laboratories
57
-------
0.3 x 10s kg/mo
• SIC 822 - Colleges, Universities,
Professional Schools, and Junior
Colleges
0.2 x 106 kg/mo
• SIC 892 - Noncommercial Educational,
Scientific, and Research Organiza-
tions
0.05 x 10G kg/mo
• SIC 7391 - Research and Development
Laboratories
0.3 x 106 kg/mo
* SIC 7397 - Commercial Testing
Laboratories
0.3 x 106 kg/mo
It is important to emphasize that the above
numbers are based on limited data but are
probably directionally correct.
Census data were used in Reference 1 to
provide a breakdown by SIC code of the num-
ber of facilities in each EPA region. Using
the assumption that small volume generators
are distributed geographically the same way
as all generators within the SIC code, Table
1 provides an approximate distribution of
small volume generators in the subject SIC
codes.
Using the above numbers for total
quantities of hazardous wastes generated by
small volume generators and the total number
of small volume generators, Table 2 provides
an idea of the average rate of hazardous
waste generation and the relationship of
small volume generators to all generators in
the SIC.
CURRENT WASTE MANAGEMENT PRACTICES
Methods by which each of the waste
categories mentioned in a previous section
are managed by small and large volume gen-
erators will be discussed briefly. For the
most part there are few differences in the
types of management practices used by small
and large generators, but economy of scale
will favor one management technique over
another. In general, economics will deter-
mine the specific choice from among those
available.
Infectious and pathological wastes are
generally treated by one of three methods.
They are usually disinfected by autoclaving,
or chemical treatment, or incineration, and
the residue is contract hauled or disposed
of as normal trash. Waste drugs are gener-
ally incinerated on-site, returned to the
manufacturer, diluted to the sewer or sent
to a drug enforcement agency. Oils and
solvents are generally sent to a recycler,
treated for reuse on-site, drummed and
contract hauled, or burned in a boiler or an
incinerator. Waste pesticide liquids are
generally drummed and contract hauled.
Empty containers are generally rinsed to the
ground or the sewer and the rinsed contain-
ers are disposed in the normal trash. Acids
and alkalis are generally neutralized and
diluted to the sewer.
The wastes that are most amenable to
incineration from the above are infectious
and pathological wastes, some waste pharma-
ceutical s, waste solvents and oils, and mis-
cellaneous 'organic chemicals. Decisions on
the most suitable method for disposal will
be made on site-specific and economic bases.
Other methods that may help reduce the
ultimate amount of hazardous materials dis-
carded have been suggested. Management
controls on procurement policy to discourage
the ordering of surplus chemicals and to
encourage recycling of chemicals may make a
contribution to reducing the amount of waste
but will not completely solve the disposal
problems. (2)
Contract hauling handles many problems
for the waste generator including responsi-
bility for disposal and legal and regulatory
requirements. However, the waste generator
may still be liable if the contract hauler
disposes of the hazardous material improper-
ly. For some wastes contract hauling is the
only viable solution.
Incineration, if applicable and effec-
tive, provides an ultimate solution to the
waste disposal problem. As opposed to land
disposal where the hazardous material still
have the potential to create environmental
problems, incineration converts the hazard-
ous material into an innocuous form. Un-
fortunately, especially for small volume
generators, there are operational, energy
and economic problems that may limit its
usefulness.
APPLICABILITY OF INCINERATOR TECHNOLOGY
Since the current study is still in
progress, it is premature to draw final
conclusions, but a number of incinerator
types have been examined and selected haz-
ardous waste generators have been contacted
to determine their needs, problems,
58
-------
approaches, and experience with incineration
of toxic and hazardous wastes. Some of the
salient points will be summarized in this
section.
Thirty-nine incinerator manufacturers
were contacted to obtain information on the
availability of incinerators for destruction
of hazardous wastes from small volume
generators. Units with capacities as low as
11 kg/hr were found. As waste generators
were contacted about their experience, addi-
tional incinerator manufacturers were iden-
tified. The approaches varied widely from
manufacturer to manufacturer. The most
highly developed hazardous waste incinera-
tion technology is represented by rotary
kiln and liquid injection incinerators.
Fluidized bed and multiple hearth incinera-
tors are also commercially available. A
number of innovative and emerging technolo-
gies, including starved air combustion/
pyrolysis, molten salt incineration, molten
glass incineration and plasma arc pyrolysis
have also been put into use. The assess-
ment of incinerators has not been completed
at the time of writing. Descriptions of
the more common types of incinerators, their
status, wastes handled, advantages and dis-
advantages may be found in EPA's Engineering
Handbook for Hazardous Waste Incineration.
(1)
A recent paper from the National
Institutes of Health (2) reported on a
screening study of five incinerators for
their general suitability for destruction
of hazardous organic chemicals. Five units
were tested 1) a two-chamber refractory-
lined incinerator, 2) a three-chamber, ver-
tically-aligned incinerator, 3) a dual-
chamber, rectangular incinerator, 4) a dual-
chamber, refractory-lined, batch incinera-
tor, 5) a molten salt catalytic incinerator.
An organic chemical tracer was used to
monitor the effectiveness of destruction of
common organic chemicals and a fire retard-
ant chemical was used to monitor the ef-
fectiveness of combustion. In addition,
hydrocarbons, nitrogen oxides, hydrogen
chloride, sulfur oxides, and particulates
were monitored. As indicated in Table 3,
all units demonstrated relatively good per-
formance although Unit 5 had some problems
with hydrocarbons and particulates, possi-
bly caused by short-duration smoke emissions
after charging.
Incineration offers significant ad-
vantages in disposing of wastes consisting
primarily of organic materials with high
heating values. Unlikely candidates for
incineration are heavy metals, high-moisture
content waste, inert material, inorganic
salts, and material with high inorganic con-
tent. A number of small volume generators
already make use of incineration for wastes
not currently regulated by RCRA and it
appears that a number of units show promise
for waste disposal at hospitals, univer-
sities and research facilities. While RCRA
does not require trial burns for small vol-
ume generators disposing of hazardous wastes
by incineration, many want assurance that a
particular incinerator can achieve a des-
truction/removal efficiency of 99.99% with
their mix of wastes.
REFERENCES
1. Ghassemi, M., et^ al_ 1979. Technical
Environmental Impacts of Various
Approaches for Regulating Small Volume
Hazardous Waste Generators. Final
Report, Contract 68-02-2613 Work
Assignment 27 and Contract 68-03-2560
Work Directives T-5012, T-5014, and
T-5015, U.S. Environmental Protection
Agency, Washington, D.C.
2. Rogers, H.W. 1979. Hazardous Wastes-
New Developments. J. American College
Health Association 28: 158-164.
3. Bonner, T. e_tal_ 1980. Engineering
Handbook for Hazardous Waste Inciner-
ation. Draft Final Report, Contract
68-03-2550 Work Directive T1016, U.S.
Environmental Protection Agency,
Cincinnati, OH.
59
-------
CO
CM i—
00
CO
CO
o
I— CTl
ID 00
i— O
CB
oo
CM
CO *d-
O i—
CO
CO
CO
INI
co
60
-------
1
o
1—
LU
LU
LU
§5
•t
O
_J
5:
E:
CO
Q
LU
I—
0
LU
LU
<"
>—
ca
z:
o
P
LU
CD
LU
I—
^c
3
CM
LU
_J
CQ
h-
C
CO
O CO
S- +•>
CO CO
D- CB
3 O
CO l-H
eC i — CO
ca
a> +-> c
-l-> O «r-
co 1—
CO
o
CD
CO
c
CD (/)
O CO
s~ •<-
CO -I-5
o.-1-o
co •!— co
<: o
CO C
CD Hr-
> 0
CO
CO
q- CD
O >
CO
CO
CO -O *-"~^
Di O
C E
CO O ^
O5*^" Ol
CO
«C C
CO
CD
CO
CO -
CO CD -— -
3 > Q
CO £•
re! E cn
O £-: —
CO
.n co
3 CD
•z. >•
CO
la 1-
,[^ f~^
i2.
LU
a
o
CJ
0
1— 1
co
!***• CO CO
CM r^ oo
CM CT> r^»
CT» C7^ CO
o o o
CM r-. CM
r™ " r— •
to 10 to
00 0
x xx
CO OO CM
O O C3
^t* vo r^
i— 1^ O
CO CO CO
IO OO r—
00 CO
IO •!-
CO _J 4->
CO r-i CO
•4-> CO S-
•r- O CO
O. r— •!- >
10 co r-» x> CM «p-
O O O CO CM CI
co re co 2: co ZD
•— r— CM
CM i— US
CM OO •=!-
O O O
m oo CM
CM CM
to j} CQ
O CO S- «* CM
i — O IO i — i —
3 CO
"O CO
r—
cO =3 IO C
0 C _! -r-
O CO r— r>- +J
CM C T- O> Q CT> CO
cn o o oo os oo cu
OO Z CO l»- OS l~- r—
S
CO
S,
CO
4i
01
•r-
cp
CO
-a
E
1
1
10
3
O
CO
o
*
61
-------
££
3 •
CJ =J
•r- O
t^ce"
IOCS
Q.
LU
CJ
UJ
on
LU
to
LU
too.
O 0.
a: o.
O
to
az
co
LU
CQ
o o.
SE CL
s-
o
£-
cu
to
CM
C3
C3
O
CM
O
C5
CVI
in
CM
O
U3
O
C5
CM
m
LO
CO
in
O
CM
CO
O
CD
CO
a •
ca «a-
v
co
CQ CM
V
CM
CO
CO CM
V
r-.
CM
co
CO
I
o
62
-------
HAZARDOUS WASTE CONTROL TECHNOLOGY DATA BASE
Richard L. Holberger
The MITRE Corporation
McLean, VA 22102
Dr. C. C. Lee
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
*
This paper describes the program to develop an automated data base storing technical
information on the thermal destruction of hazardous wastes. The data base will store
detailed design, .operating, and performance data obtained from incineration.facility
permit applications, trial burn reports, research and development projects, and contractor
data gathering efforts. The paper describes the development process itself and presents
some of the details of the system as it will be implemented, including descriptions of the
data content and structure, and input and retrieval features.
INTRODUCTION
Incineration has long been an impor-
tant technique for the destruction of
hazardous waste. Recently, however,
especially since the passage of the
Resource Conservation and Recovery Act
(RCRA), incineration practices have begun
a period of rapid advances in technical
sophistication and complexity. Many new
systems are evolving using a diverse group
of thermal destruction concepts and designs.
As required by RCRA, EPA is developing
a regulatory program addressing the incin-
eration of hazardous wastes. In order to
insure that this program is developed and
implemented with the benefit of the most
current industry experience with advanced
incineration techniques, there is a need
for EPA to obtain and analyze technical
design and operating data from both exist-
ing incineration facilities and research
activities in this area.
As part of its support to the Office
of Solid Waste, the Incineration Research
Branch (IRB) of the Industrial Environmen-
tal Research Laboratory in Cincinnati has
a multi-faceted program underway to develop
such a data base, which will be called the
Hazardous Waste Control Technology Data
Base (HWCTDB). The objectives of this data
base are:
• to provide direct technical support
to^Regional permit and compliance
programs, especially in the evalua-
tion of requests for trial burn
exemptions;
• to form a sound technical basis for
development and promulgation of
future revisions of regulations for
hazardous waste incinerators;
• to define the range of current
engineering practice, particularly
regarding the performance associated
with various design and operating
characteristics;
• to provide a foundation on which to
plan future R&D efforts addressing
thermal destruction of hazardous
wastes.
The data base will store detailed tech-
nical design, operating, and performance
data obtained from incineration facility
permit applications, trial burn reports,
research and development projects, and con-
tractor data gathering efforts. Specific
data content was chosen based on a user
needs study conducted early in 1981
(Barrett, et. al., 1981), and on the infor-
mation requirements for permit evaluation as
63
-------
identified in the "Guidance Manual for
Evaluating Permit Applications for the
Operation of Hazardous Waste Incinerator
Units" (Vogel, et. al., 1981).
This paper describes the program to
develop the data base as well as some of
the details of the system as it will be
implemented. Since the system will not be
completed until early 1982, several months
after preparation of this paper, some of
the details described here may be modified
slightly in the operational system. The
program to develop this data base consists
of three major phases, as described below:
initial development, implementaion, and
system operation.
INITIAL DEVELOPMENT
The initial development phase was
completed by the MITRE Corporation this
summer. It consisted of a user needs sur-
vey to identify user requirements for both
content of and access to the data base, a.
review of existing data bases for potential
implementation of the HWCTDB, and the de-
velopment of an initial list and descrip-
tion of desired data elements.
The user needs study identified six-
teen organizations within EPA and DOE as
potential users of the data base. Discus-
sions with various members of each
organization indicated a need for obtaining
a wide range of data on both administrative
and technical features of incineration
facilities, as well as other types of
facilities destroying hazardous wastes.
Further, a strong preference was expressed
for on-line retrievals and the ability to
sort by many different criteria, including
various waste characteristics and a range
of incineration design and operating
features.
To minimize the time and effort re-
quired for system development and user
training, and to maximize standardization
of EPA's information systems, an effort was
made to identify an existing data base
which could be easily adapted to meet the
needs defined above (Dratch and Keitz,
1981). A review of EPA's System Informa-
tion Directory (US EPA, 1980) revealed
seven potentially applicable data bases.
Further analysis narrowed the consi-
deration to two systems: the Hazardous
Waste Data Management System (HWDMS),
operating for OSW, and the Environmental
Assessment Data Systems (EADS), managed by
ORD. Both utilize Intel Systems Corpora-
tion's System 2000 Data Base Management
System to organize their data. Analysis
of the two systems indicated that each
could be adapted to meet the needs of
HWCTDB. EADS is well suited to storage of.
source characterization and emissions test
data, but would still require special pro-
grams to obtain many of the common sorting
and retrieval capabilities desired by
users. HWDMS is designed to store and
organize information received in applica-
tions for hazardous waste facility permits.
Current plans are to expand the data base-
in phase with OSW's permit program to even-
tually include the data received on Part B
applications.* HWDMS, therefore, already
contains the administrative information
desired by users and would eventually con-
tain much of the technical design and
operating data required on Part B facility
permit applications. Further HWDMS is
already familiar to and used by OSW (who
will be one of the major users of HWCTDB),
has a user-support group in the EPA Region-
al Offices, and has a more versatile data
input system to facilitate data entry and
validation. The planned expansion of HWDMS
would also provide the opportunity to build
the desired capabilities into-the primary
structure of the data base. Therefore, a
recommendation was made to build HWCTDB as
an expansion of HWDMS.
The final portion of the initial de->
velopment phase was the construction of a
preliminary data dictionary describing the
desired data elements and their relation to
each other (Holberger, 1981). A total of
155 data elements, grouped into 22 record
groups (repeating .groups) were identified.
This effort also included specification of
some of the reports and retrievals which
would be requested from the system.
*The application process for obtaining a
hazardous waste facility permit occurs in
two stages. The Part A application re-
quires administrative information and a
general description of the facility. It
is the basis EPA uses for granting interim
status. The Part B application will con-
tain the detailed technical data necessary
for EPA's evaluation of the ability of the
facility to meet the requirements for a
full hazardous waste facility permit.
64
-------
IMPLEMENTATION PHASE
The Implementation Phase consists of
three parts: systems development and
documentation, data collection, and accep-
tance testing (which includes the initial
data loading). Both the systems develop-
ment and data collection portions are
underway as of this writing.
Systems Development and Documentation •
This effort includes the finalization
of the system requirements; system design,
systems programming, documentation and
user training. (Documentation and user
training activities will also extend into
the system operation phase.) The final
system design, the systems programming,
and the software-related portions of the
documentation arid user training will be
prepared by the Computer Sciences Corpora-
tion (CSC), based on the system require-
ments specified by EPA and MITRE. As of
this writing, the data dictionary is being
completed with the addition of edit cri-
teria and specification of required data
for input to the data base; a preliminary
system design is being prepared for final
review; CRT screen formats are being devel-
oped for data entry; and final report
formats and retrieval criteria are being
prepared.
Figure 1 shows the structure and con-
tent of the data base as of this writing.
Final implementation may differ slightly.
At present, there are 159 data elements in
21 records (repeating groups).* The pri-
mary record under which the others are
organized is the CO, or Facility Record.
It contains much of the administrative data
presently existing in HWDMS. Each facility
would have one Facility Record. All other
records would be considered descendants of
the Facility Record, and may occur many
times for each facility. For .example,' a
given facility may burn more than one
waste, and each waste will contain numerous
constituents, some of which will be desig-
nated as Principle Organic Hazardous Con-
stituents (POHC's). Therefore, the name of
each constituent analyzed in the waste, its
concentration in :t.he waste, and a flag in-
dicating its status as a POHC would com-
prise a separate Waste Composition Record
(C2840) descending from the Waste Charac-
terization Record (C1800) organizing the
complete description of that waste. Each
group of records describing a single waste
stream (consisting of one C1800 record and
all its descendants) would be a separate
repeating group descending from the Facili-
ity Record (CO).
As designed, the HWCTDB will store in-
formation on multiple units at each facil-
ity, each with multiple sets of operating
conditions. Separate descriptions of de-
sign and operating conditions will be main-
tained for each chamber of multiple chamber
facilities, and many separate waste streams
may be associated with each chamber. Since
the Waste Characterization Record is not a
descendant of the Operating Conditions
Record Group, data may be stored describing
the destruction of the same waste stream
under many different operating conditions
or designs without repeating the waste
description. The use of comments describ-
ing distinctive characteristics of any
portion of the facility or its operations
is encouraged, with comment numbers design-
ed to indicate the subject of each comment.
The system programming and CSC's in-
ternal testing will be conducted during the
winter of 1981-1982, with a planned turn-
over date in early spring 1982. The pro-
gramming effort will involve actual con-
struction of the data base, programs to
load information from and maintain consis-
tency with HWDMS permit data, programs for
report generation, and the provisions for
a few special retrievals expected to be too
complicated for most users.
Documentation of the data base will
include preparation of a System Management
Summary, System User's Guide,"System
Maintenance and Operations Manual and
Technical User's Guide. The first three
documents will address the system-oriented
procedures for manipulating and maintaining
the data base. The fourth document will
address the non-computer oriented user
community, providing descriptions of
methods to maximize utility of the data
base to support both regulators and re-
searchers. User-training sessions will
proceed in parallel with the documentation.
*In this discussion, a record or repeating
group is defined as a group of related
data elements describing one particular
aspect of a facility and its operations.
Each record and each data element has been
assigned a .reference number prefixed by
the letter C.
65
-------
2u|i B M SSSIISSI||2SS>.B M
c .
— — « — «—
UUUOUUUUUOUUUUUUUUU
RHIT-DATES Record
TIF-RECPT-DATE
-RECPT-DATE
CI600 - HUF-RCRA-PE
C1601 - KUF-IU
CIM2 - HUF-PA
1200 - UUF-SIC-CODES Rccor
C1201 - HUF-SIC-CODE
, sl
^ II
s ??
a ss
i §§
u
1
! s
CIOOO - HWT-ADHIHISTRATIVI
CIOOI - HWT-COKHENT-l
CI002 - HUT-COHHEKT
ciaoo
HUT-UASTE
CHARACTERIZATION Record
SEE FIGURE 1-D
1 «
Is i "
g§s c
• •» H " bl
—
- [-
•J
0
1 2
ill!
"H "*
sl i
Sba II
C2S20
HUT-UASTE
SOURCES Record
SEE FIGURE 1-D
-a
• , 8 o
iff I
U S U vi
ffl
3** s £e u
SO «
—
•-
-
as o ""
l|s a
C3590
HWT-EHISSIONS
Record
SEE FIGURE 1-C
^
il v
ll I
•o
g" V
ai u
ll 1
G S U M
C3460
HWT-HOMITOR
SVSTEH-DESC Record
SEE FIGURE 1-B
C3430
Hwr-HAJOR
COHPOMEHT-DESC Record
SEE FIGURE 1-B
•w
S
* f
C3400
HOT-DESIC
CHAHBER-D
SEE FiniR
1
:3650
lUT-MONITOR
IESULTS Record
'EE FICURE 1-C
V
S?g c
66
-------
-a
o
Jj g
1 Jl »H . = „ =
u
si 2
§ z "
IP 1
<
i
§
|
i
i
u
-
Is g
III I
tn V
S 1
111 i
si
fc * 0
lii!
1
ii §
Sil i
C3460 - HWI-tWHIIOR-SVSTEM-DSSC Rtcord
C3461 - HUT-KOH-IND
C3462 - Kin-STREAM-MONITORED
CJ463 - HWT-MOHIIOR-DESC
C3464 - UVT-SAXPL-HETH
COHKMEHT-DCSC Ucord
OHP-IHD
TtH-CODE
JUIUFACTURER
i iiil
£ .1,U,
S S fn 11 r
• sss;
1
1>
K-cmKBU-oesc MCO
ISD
1ER-TVPE
HAS!!
1 P§!
1
U
O r-4
sfl §
2
ife f
slil
SlliyI
• jn >fl r«. « (r. o
;ssssss
is §tSB
u u u u C o u
67
-------
2
i- <
ijllsssigAgji
iliiiiiiilli
i i i i t i i i i i i t
QQouQQSdluCu
? is
I i!
i §§
ill
3 0
;
i-SJ
SSSSS
111 ..|g| 1 ill
!
i
FEED Reco
-UST-STRH
C3
C3
C3
o
a
Si
68
-------
vvv
Sco oo co a
c-j *v| M r
" 5
O ti
I 1
I 1
! T
69
-------
Data Collection
The second part of the implementation
phase is data collection. The MITRE Cor-
poration has completed a task to obtain
data from manufacturers of incineration
systems, documented in "Profile of the
Hazardous Waste Incinerator Manufacturing
Industry" (Frankel, et. al., 1981). Under
a separate task, MITRE has also begun
collecting information from operating
incineration facilities. Data is collected
through telephone and personal interviews
with facility operators who have submitted
RCRA permit applications. Data collected
during this and subsequent tasks will be
recorded on data entry forms designed to
correspond to the CRT screens used during
data entry to the system.
Incorporation of this data into the
automated data base will not occur until
the acceptance tests are underway. After
the initial loading of the data base, data
will be input to the system using full
screen CRT terminals connected to EPA's
PDF-11/70 minicomputers. Figure 2 is an
example of one of the screens which will
be used to input data for the Major Compon-
ent Description Record (C3430). The first
two lines identify the facility and design
records to which the data applies. The
screen then contains a list of the elements
in the record, along with blanks sized for
the maximum length of each element. The
cursor will automatically move from one
field to the next, allowing inputs only in
the appropriate positions. As each data
element is entered, the edit program will
check to ensure that it is a valid entry.
The edit procedure is a series of
computer programs which check that each
entry fits the description for the appro-
priate data element (i.e., correct type -
decimal, integer, or alphanumeric; and
correct number of characters) and that the
entry has a valid value (some of the data
elements, such as the Item Code, are re-
stricted to certain sets of common terms
or codes; other numeric data have valid
ranges, outside of which they will not be
accepted by the data base, e.g., no
combustion temperatures will be allowed
outside the range 500-4000°F). If the
screen entry does not pass the edit test,
the program will print an error message
and give the operator the opportunity to
correct it immediately. After data for
each element has been entered on the
screen, the program will, give the operator
another opportunity to correct any errors
not detected by the edit program.
After completing the entries for one
major incineration system component (e.g.,
.for the description of the waste feed
system), the operator may then request
another repetition of the screen for entry
of an additional record (e.g., for the
description of a venturi scrubber), or may
request the screen for a different record
group (e.g., for the data elements com-
prising the Monitor System Description
Record (C3460) ?t When the data meets the
approval of both the operator and the edit
procedure, the program will construct an
update file for later entry on the full
data base, which will physically reside in
EPA's IBM 370 in Research Triangle Park.
Acceptance Testing
The final portion of the implementa-
tion phase is the acceptance testing and
system initialization, which will begin
as CSC completes the system programming
and "turns over" the software to EPA. The
testing procedure will exercise each sys-
tem function (data entry, edits, update,
retrieval, etc.) with both valid and in-
valid "dummy" data. System performance
will also be evaluated relative to the
design requirements. Any deficiencies
will be noted for correction. All data
introduced during the acceptance test will
be deleted after successful completion of
the test procedure.
System initialization and performance
testing will begin with the initial load-
ing of "live" data on the data base, in-
cluding both the conversion of data from
the HWDMS and the entry of data collected
up to that time. System performance and
integrity will be monitored closely for
several months to detect previously un-
discovered problems. This effort will be
coordinated with user-training activities
and the establishment of procedures for
use of the data base by ORD and Regional
EPA personnel.
SYSTEM OPERATION
The system operation phase of this
project includes completion of documenta-
tion and user training efforts and contin-
uation of data collection and entry as
described above, long term data base
70
-------
I2J
o
M
H
M
Pi
O
co o
-i H
M g pa. >-t H M
fx< PM pd EH
MO MS H
c=> co ij w a
0- M EH pa
pa O CO g
H PL.
Pi fe CO M
0 !=)
>-o o"
^ W
** 12
pa
^
| >H O1 HO
EH
-------
management and maintenance, data base
Utilization, and continued system evalua-
tion and enhancement (as required).
Long term management and maintenance
will involve designating a data base admin-
istrator (DBA) who will be responsible for
access to the data base, data integrity,
file maintenance, backup and recovery
procedures, periodic reconciliation with
1IWDMS, and monitoring the continued devel-
opment and use of the data base. The DBA
will also control the data entry process,
taking the update files created by the
edit programs and incorporating them in
the data base.
,A user support group will be estab-"
lishcd to provide a retrieval service for
any users requiring assistance. This
group will maintain a library of special
retrieval programs of general interest to
the user community, provide assistance in
formulating retrieval requests and inter-
preting outputs, and will operate an on-
line news service to keep users informed
on new retrievals or other items of inter-
est pertaining to the data base. The user
support group will also coordinate the
creation and distribution of additional
documentation and the training .of new
users, as necessary.
The system is being designed so that
many of the common retrievals may be per-
formed quickly in interactive terminal
sessions using System 2000's Natural
Language. To perform such retrievals, the
user requires knowledge of the data base
logical structure (Figure 1), and usually
needs to attend a short user-training
course to learn the concepts of Natural
Language. No other computer-related ex-
perience will be required for simple
retrievals.
The basic retrieval procedures will
involve using Natural Language to identify
facilities with certain combinations of
characteristics (e.g., commercial liquid
injection units operating under starved
air conditions within a certain tempera-
ture range). After identifying the facil-
ity identification number (element C101)
and any other information needed to unique-
ly identify the records meeting the selec-
tion criteria (e.g., waste sequence number
or system index), the user may choose one
or a combination of several output report
formats. A list of the standard reports
and a summary of their contents is presen-
ted in Table 1. A user may augment the
reports by requesting additional data
elements, depending on his needs.
The final aspects of the system opera-
tional phase are continued evaluation and
enhancement, which include on-going evalua-
tion of system performance with respect to
both the original objectives and to the
continued evolution of EPA's programs. An
annual assessment .will be prepared to,
document the current system performance,
identify new program needs which could be
met by the data base, and suggest enhance-
ments which would increase the utility of
the data base or improve procedures iden-
tified by users as cumbersome. Recommended
enhancements could include the development
of additional retrieval capabilities and
new reports, modifications to the data
content or structure of the data base, or
simple expansion of the supporting documen-
tation.
REFERENCES
1. Barrett, K., D. Dratch, and E. Keitz.
1981. User Needs Interviews for a
Hazardous Waste Incineration Facility
Data Management System. WP-81W00140,
MITRE Corporation, McLean, Virginia.
2. Dratch, D., and E., Keitz. 1981. A
Review of Existing Information Systems
for Potential Use with Hazardous Waste
Incineration Facility Data.
WP-81W00268, MITRE Corporation, McLean,
Virginia.'
3. Frankel, I., N. Sanders, and G. Vogel.
1981. Profile of the Hazardous Waste
Incineration Manufacturing Industry.
WP-81W00443, MITRE Corporation, McLean,
Virginia.
4. Holberger, R. 1981. Recommended
Structure, Content, and Data Formats
for a Hazardous Waste Control Technol-
ogy Data Base (HWCTDB). WP-81W00340,
MITRE Corporation, McLean, Virginia.
5. U.S. EPA. 1980. Systems Information
Directory. Management Information and
Data Systems Division, National Compu-
ter Center, Research Triangle Park,
North Carolina.
72
-------
TABLE 1
STANDARD REPORTS FROM HWCTDB
REPORT
CONTENTS
Basic Identification Package
Expanded Identification
Package
Waste Stream
incinerator Design
Operating Conditions
Performance
Monitor Results
Facility name, location, identi-
fication number and permit status.
Contact name and phone, owner
name, mailing address.
Waste sequence number, description
and selected elements from the
Waste Characterization record and
its descendants
System index, System type, and
description, and other selected
elements from the Design-Instal-
lation record and Component Data
record and its descendants.
Operating conditions index, data
type and other selected elements
from .the Operating Conditions
record, Combustion Chamber re-
cord, Waste Feed record and Air
Pollution Control Device record.
Operating conditions index, data
type, total scrubber efficiency,
particulate emissions, and all
Emissions records containing a
value for DRE.
List all monitor records for a
specified monitor (given a facil-
ity ID, system index, and opera-
ting conditions index).
73
-------
TABLE 1
(Continued)
REPORT
CONTENTS
Burn Summary
Data Base Summary
Extracts selected data elements
from the Facility record, Waste
Characterization record, Design/
Installation record, Component
Data record and Operating Con-
ditions record, providing a
summary of the type of waste
burned, type of system used,
operating conditions used, and
resulting emissions.
Provides total of certain values,
(e.g., total liquid injection
capacity); ranges of values (e.g.,
range of temperatures used in
rotary kilns); or counts of the
number of occurrences of certain
values (e.g., number of units
combining rotary kiln and liquid
injection).
74
-------
Vogel, G., K. Brooks, J. Cross,
I. Frankel, S . Haus, and W. J'acobsen.
1981. Guidance Manual for Evaluating
Permit Applications for the Operation
of Hazardous Waste Incinerator Units.
WP-80W00628, Rev. 3. MITRE Corpora-:
tion, McLean, Virginia.
75
-------
OVERVIEW OF THE CONCEPT OF DISPOSING
OF HAZARDOUS WASTE IN INDUSTRIAL BOILERS
George L. Huffman
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45268
C. Dean Wolbach and Larry R. Waterland
Acurex Corporation
Mountain View, California 94042
ABSTRACT
This Paper discusses the results of the first phase of an overall study in which
the technical feasibility of the concept of disposing of hazardous waste by co-combusting
it with conventional fuels in industrial boilers is examined. The second phase of the
study involves the generation of appropriate sampling and analysis protocols and the
conduct of an investigation to locate candidate industrial boiler testing sites. Phase
three, just now getting into full swing, involves an actual Testing Program to determine
the effectiveness7of destroying hazardous wastes by co-firing them in industrial boilers.
In this Paper, the estimated quantities of hazardous wastes generated each year are
discussed, as are the projections of the number of industrial bpilers currently in opera-
tion. A kinetic model for the thermal destruction process that takes place in a boiler
is described. Waste decomposition rate curves are compared to boiler temperature/
residence time curves to allow prediction of the waste's destructibility in a given
industrial boiler. The kinetic model is applied to a real-life PCB test burn situation
and the prediction-versus-reality is compared.
INTRODUCTION
It has been estimated that approxi-
mately 41 million metric tons of hazardous
wastes are generated each year in the
United States (2,9). The environmentally
acceptable disposal of these hazardous
wastes is a mandated goal of the Resource
Conservation and Recovery Act (RCRA).
Regulations proposed and promulgated by
the U.S. Environmental Protection Agency
(EPA) in response to RCRA place restric-
tions on the generators of these wastes,
including requirements for the manifesting
of any wastes to be transported and chain-
of-custody liability. These requirements
provide, then, some economic incentives
for the onsite disposal of the hazardous
wastes that are generated on the Nation's
270,000 industrial plant sites (7).
Thermal destruction is a method of
disposing of those hazardous wastes which
are highly organic. One thermal destruc-
tion process is high-temperature inciner-
ation, a practice which is regulated by
RCRA. Another one is the co-combustion or
the co-firing of organic wastes in indus-
trial boilers along with conventional
fuels (coal, natural gas, fuel oil). Use
of this technique not only destroys the
wastes but also allows the recovery of the
waste's fuel value which, in turn, lessens
the Nation's overall consumption of scarce
fuels.
The primary purpose of industrial
boilers is the production of energy for
onsite process needs. Current RCRA
76
-------
regulations specifically exempt facilities
that burn wastes in energy-producing opera-
tions from complying with RCRA rules re-
garding the high-temperature incineration
of wastes. These rules require that 99^99
percent destruction and removal efficiency
be achieved for the principal organic
hazardous constituents (POHC's) contained
in the waste to be incinerated.
Industrial boilers offer great poten-
tial for the onsite thermal destruction of
hazardous waste. If the time and tempera-
ture profile of a given boiler is similar
to the residence time and exposure tempera-
ture required to destroy a given waste,
then the disposal of the hazardous waste
in the boiler becomes a distinct possibili-
ty, one worthy of further evaluation (2).
As a first step in this evaluation process,
the U.S. EPA awarded a contract to the
Acurex Corporation to study the technical
worth of the overall concept. This Paper
summarizes the findings of Phase I of that
study (2).
WASTE/BOILER AVAILABILITY
National figures regarding the amount
of hazardous wastes generated in this
country are extremely difficult to project.
Nonetheless, some attempts have been made.
Table 1 presents the results of one such
estimate (2,9). The values in this tabu-
lation were based 'on the assumption that
the ratio of the amount of hazardous waste
generated by an industry to the number of
employees in that industry is approximately
constant for each plant in each industry.
This is, to say the least, a highly specu-
lative assumption.
Table 1 shows that over 41 million
metric tons of hazardous waste are gener-
ated each year within 17 of the 19 stand-
ard industrial classification (SIC) codes
comprising the manufacturing industry plus
one non-manufacturing code. Not all of
this waste is amenable to thermal destruc-
tion however, since some portion of'the
wastes listed are inorganic. • The combusti-
ble or organic content of the wastes varies
from industry to industry, from perhaps a
low of 30 percent in SIC 33 to a high of
80 percent in SIC 24 (2). Consequently, a
significant fraction of the 41 million
metric tons of hazardous waste generated
each year is organic, is combustible, and
is amenable to destruction in industrial
boilers.
But, are there enough industrial boil-
ers in this country to do the hazardous
waste destruction job? There are currently
about 367,000 boilers of all types in the
United States (8). Of these, about 238,000
can be classified as industrial installa-
tions (7,8). Most of these boilers, how-
ever, are very small natural gas- or oil-
fired firetube units used primarily for
space heating. Such "packaged" units do
not easily lend themselves to waste co-
firing (7). Nonetheless, there remains a
substantial number of industrial boilers
having capacities larger than 2.9MW (10
million Btu's per hour) which are more
suitable for waste co-firing operations.
Table 2 reveals that there are about 43,000
industrial boilers in the U.S. over that
size (2,3). Table 2 also shows that the
boiler population matches fairly well the
distribution of waste generators across the
U.S., while providing an indication of the
regional breakdown of the total amount of
organic and inorganic hazardous wastes
generated (2).
KINETIC MODEL FOR THE THERMAL DESTRUCTION
PROCESS
The EPA/Acurex study set out to develop
a mathematical expression that would
calculate the amount of residence time
needed in an industrial boiler operating
over a relatively fixed temperature regime
to insure a hazardous waste destruction
efficiency of, say, 99.99 percent (or any
other required level of detoxification).
The physical parameters that influence
destruction efficiency are the three T's -
— time, temperature and turbulence. The
rate of thermal destruction is dependent
upon whether the boiler is operating in a
true oxidation mode or in a mode approach-
ing pyrplysis in some of the boiler's
starved-for-oxygen passages. It is also
dependent upon the values of the kinetic
constants for the hazardous waste compound
(or group of compounds) to be destroyed.
If the time, temperature and turbulence
requirements to meet the required destruc-
tion efficiency of a given compound can be
compared to the time, temperature and
turbulence conditions that exist in a
given boiler, then the ability of the
boiler to achieve the desired destruction
efficiency can be predicted (2).
77
-------
r
TABLE 1. SUMMARY OF ANNUAL HAZARDOUS WASTE GENERATION
(1,000 METRIC TONS PER YEAR)(2,9)
SIC
22
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Code
Textile mill products
Lumber and wood products
Furniture and fixtures
Paper and allied products
Printing and publishing
Chemicals and allied products
Petroleum and coal products
Rubber and miscellaneous plastics products
Leather and leather products
Stone, clay, and glass products
Primary metal industries
Fabricated metal products
Machinery, except electrical
Electric and electronic equipment
Transportation equipment
Instruments and related products
Miscellaneous manufacturing industries
Non-manufacturing industries
Waste
Total
203
87
36
1,296
153
25,510
2,118
248
474
18
4,061
1,997
323
1,092
1,241
90
319
1,971
Percent
of Total
0.5
0.2
0.1
3 . 1
0.4
62.0
5. 1
0.6
1.1
0. 1
9 8
4 8
0.8
2.6
3.0
0.2
0.8
4.8
TOTAL
41,237
100.0%
Acurex decided to build into their
model a measure of conservatism so that
any predictions made would be on the safe
side. Consequently, they chose to use a
first-order kinetic expression that de-
scribes a thermal cracking or pyrolytic
decomposition mode because the kinetics of
pyrolysis are slower than those of oxida-
tion (2). Slower kinetics translates into
conservative estimates of the residence
times needed for efficient waste destruc-
tion. The rate of disappearance of a
hazardous waste constituent by pyrolysis
can be expressed by:
dC
dt
- kC
(1)
where C is the concentration of the consti-
tuent at time t and k is the temperature
dependent rate constant. Solution of this
equation leads to:
In ±
C0
-Ae
-E/RT
At
(2)
where C = Concentration of material
at residence time At
C0 = Concentration of material at
time = 0
A = Arrhenius pre-exponential
factor, sec~l
E = JEnergy of activation for bond
rupture, cal/g-mole
R = Gas constant, 1.987 cal/g-
mole-°K
T = -Temperature, °K
At = Residence time in the firebox,
seconds
Using kinetic data such as that being
generated by the University of Dayton
Research Institute for the U.S. EPA's
Industrial Environmental Research Labora-
tory in Cincinnati, the A and E values can
be obtained for the hazardous waste in
78
-------
CO X-N
i«
O CM
[2*
Q W
Sw
°s
13 13
0
M CO
|"-i ^^
PQ Q
H N
CO
£»
^>
M
1-1
I-l
M
I-l
1— 1
I-l
r-- in
CM *d"
O O
oo . m
CO
VD O
CO VO
vO i — 1
CO CO
o- oo
VD r-l
CM
0
vO r— 1
I^ r-l
OS CM
vo in
CO CM
CO "•
O*\ r— 1
in r--
r^ vo
f\
CM
CO f~-
00 CM
CM m
00 r-l
m co
o m
CT> O
0\ r-l
co r--
CM 00
co 10
»^-
O CM
00 O
e~^ ^-j
»»
CO
CO CO
O CT>
CO rH
CM
C0_
CM «
to
M CO O 'CO
CO MO ' M X
r-l 0) r-l 0)
•r-l r-l 1 r-l O
O -r-t O -r-l O '-^
M O r-l O r-l M
r-< ^""M A -~^
W 'I-l 13 X! H-t v— ' 3
•H 0 g -- 0 4J
4J M OS 4-1 MS
to vo , . B CM r-i
C • 3-0 3 »
M S CM rH E3 ^
9
CM
f«^
0
CO
VO
C7N
vO
00
r~^
CM
CO
CO
00
r-l
CM
CO
o
CO
•\
CO
o
1— 1
oo
n
ON
00
m
#\
0
rH
o
I— 1
OS
n
^J-
CM
00
CO
CO
vO
OS
•*
CM
CO
M
rH
• H
o
rH
tO
4J
O
H
in
in
CO
oo
"*
o
0
in
i— i
m
rH
CO
-•*
o
p^^
m
0
r-
CM
CM
O
r-
CM
n
oo
o
o
oo
r.
CT>
o
CM
CTv
VO
O
oo
o
fl
in
o
o
CM
m
o
CO
CO
CM
CO .
M a)
O 4J
4-1 to
to rt
M IS
CD to
C >+-l M
CO O O
O 4-1
M n)
0) (!) M
4J XI CU
w EC
B) 3d)
13 SO
a
r-
CO
CM
i-H
"*
in
ON
OS
oo
CO
oo
CM
00
rH
CO
rH
O
CM
rH
, CO
CO
in
n
0
rH
OO
CM
n
VD
CO
in
CO
0
i— i
in
CO
,^
CO
I— 1
I— 1
• n
CO
^
o
rH
rH
0)
4-1
CO
CTj 3S
13 •*
M
CO T3 M
30) >-,
0 4J -~-
•O tr) EH
MM S
CD 01
N C CO
td 0) O
ffl O rH
•
79
-------
question (4). Then, assuming various
destruction efficiencies (for example, for
99.99 percent, C/C0 = 0.0001), one can
make a plot of Residence Time (At) versus
Destruction Temperature (T). Figure 1 is
an example of such a plot; it gives three
destruction curves for 2,2',4,5,5' penta-
chlorobiphenyl, one of the PCB's (2).
Figure 1 shows the comparison between
thermally decomposing PCB to the 99.99
percent level by either the pyrolysis mode
or the air oxidation mode, as well as the
pyrolysis mode of destruction to the 99
percent level. Note that pyrolysis is a
much slower destruction mode (at the same
residence time, a higher temperature is
required to get the same destruction effi-
ciency or, conversely, at the same tem-
perature, much more residence time is
needed in the pyrolysis mode than in the
oxidation mode).
Table 3 presents some A and E values
for certain selected hazardous waste con-
stituents (1,2,4,6). With data such as
this, use of Equation 2 will yield des-
truction rate curves similar to Figure 1
for hazardous waste streams of interest.
BOILER TEMPERATURE PROFILE VERSUS RESIDENCE
TIME CURVES
To determine what a prospective indus-
trial boiler can do in terms of providing
the time and temperature required for
efficient waste destruction, one must
3.000
ifoa
(2,240)
1.000
2,2', 4,5,5'
Pentachloroblphenyl
Pyrolysis (99.99%)
Oxidation {air -99.99%)
300° C
(540"F)
0.001 0,01 0.1
Time (Al), seconds
Rgura 1, Destruction curves (or pantach.lorobfprtonyl
ascertain what temperatures are reached in
its firebox,and offgas passages and relate
them to bulk gas residence times. A con-
ventional boiler temperature profile, such
as the simplified one shown in Figure 2,
provides that information (2). Figure 3,
on the other hand, is a more complex set
of boiler temperature profiles that
attempts to illustrate the range of boiler
conditions which are often encountered in
actual industrial plant operations. 'Point
A' on-Figure 3 depicts the exit gas tem-
perature at full-load and mean residence
time conditions. Point B" represents the
exit gas temperature at full-load and
"fast-path" (i.e., one-half of the mean
residence time) conditions. Points C' and
D" are the respective exit gas tempera-
tures but at half-load conditions. The
box connecting these four points has been
termed, by Acurex Corporation (2), the
"boiler kinetic operational zone", the
BKOZ, which of course is the range of
outlet conditions typically experienced
in normal boiler operations (2). Tempera-
tures above the exit temperature A', for
example, are readily found inside the
boiler firebox; they are found along the
A-A1 temperature path, but at lesser resi-
dence times.
WASTE DESTRUCTION VERSUS BOILER PROFILE
CURVES
To determine if a given hazardous
waste can be destroyed in a prospective
industrial boiler, one should construct a
chart in which the waste destruction rate
curves of Figure 1 are superimposed on the
boiler temperature profile curves of
Figure 3. One such chart is given in
Figure 4; on it, the decomposition curves
for waste compounds A through E are
overlaid onto the temperature profile
curves for a given boiler operating under
various conditions (2). In this example,
Compound E will readily be destroyed
because temperatures higher than that
required for its destruction are found at
every point within the boiler. Conversely,
Compound^ A will not be destroyed because
all temperatures found within the boiler
are lower than that needed for its des-
truction. If the boiler were to be opera-
ted in only the half-load condition, it
is questionable whether Compound B could
be destroyed' to the efficiency required.
80
-------
TABLE 3. KINETIC DATA FOR COMPOUNDS OCCURRING IN SELECTED
HAZARDOUS WASTE STREAMS (1,2,4,6)
Oxidation
(in air) Pyrolysis
Benzene
Hexachlorobenzene
Toluene
Pentachlorobiphenyl
Vinyl chloride
Acrylonitrile
A
(sec"1)
7.42 x
1.9 x
2.28 x
1.1 x
3.57 x
2.18 x
1021
10 16
1013
1016
10 I*
1012
E A E
(kcal/g-mole) (sec"1) (kcal/g-mole)
95.9
72.6
56.5 2.1 x l.O12 77.5
70.0 7.44 x 109 53.6
63.3
52.1
I
Log Time
Maximum Bulk
Gas Temperature
Exit Temperature
Tm full-load
m ,
Tm half-load
T full-load
Log At
Figure 2. Simplified boiler temperature profile2
Figure 3. Boiler temperature profiles under full- and half-load conditions
(at mean and one-half of mean residence times)!
THE KINETIC/BOILER MODEL APPLIED TO TEST
BURN DATA: IT WORKS II
The EPA/Acurex kinetic model des-
cribed above has been utilized to predict
the destruction efficiency to be expected
when PCB is co-fired along with conven-
tional fuels in an industrial boiler
and this result was compared with
actual test burn data taken .by the GCA
Corporation for the U.S. EPA during a May,
1980 PCB-burn at the General Motors plant
in Bay City, Michigan (2,5).
The GM boiler that was tested was a
17.4 Mw (60 million Btu/hour) watertube
unit burning No. 6 distillate oil. The
PCB-contaminated waste stream was a waste
oil containing approximately 500 ppm
Arochlor 1254. The waste stream and dis-
tillate oil were mixed in a ratio of 1:10
prior to injection. The boiler was
operated at half-load (i.e., the fuel
injection rate was half the design rate)
(2,5).
81
-------
The time and temperature profile
within the firebox was generated by
Acurex using their Zonal Heat Balance
Program. Kinetic data were derived from
the thermal destruction unit data on
2,2",4, 5,5' pentachlorobiphenyl generated
for the U.S. EPA by Duvall at the Univer-
sity of Dayton Research Institute (2,4).
Time-abovo-tamperaUire
cum}, full load,
-lastpam "
Time-above-temperature curve,
lull load, mean path
Extended BKOZ
_ Time-above-temperature curve,
halt loud, fast path
Original BKOZ
0.1
1n time
1.0
The input data for the kinetic model
is summarized in Table 4, and the pre-
dictive model diagram is shown in Figure
5. The model predicted that, for half-
load and 2.5 percent oxygen content in the
firebox, the boiler should achieve greater
than 99.99 percent destruction. The actual
test results are given in Table 5 and do,
in fact, confirm this prediction (2,4,5).
1,600
5. 1,500
o
o 1,400
?
£ 1,300
a.
£ 1,200
1,100
Full load
mean path
Pyrolysis (N2)
Oxidation
• (2.5% O2)
Oxidation (air)
0.1
1.0
Time (sec)
Figure 4, Bolter-compound overlay2
Figure 5. Predictive model diagram for the GM PCB test burn2
TABLE 4. INPUT PARAMETERS FOR EXAMPLE CASE'MODEL APPLICATION2
Compound (.2,2', 4, 5, 5' - Pentachlorobiphenyl)
Pyrolysis Oxidation (2.5% 02)
E (Kcal/g-mole)
A (sec"1)
Boiler
tnsean
(02% in flue gas
1700°K + 50° (Flame)
1144°K + 50° (Back wall)
0.4 sec
1.8 sec
}
Acurex boiler
zonal heat balance
Oxidation (21% 02)
53.6
7.44 x lO9
66.7 + 0.5
2.6 x 1012
70.0
1.10 x 1016
82
-------
TABLE 5. REPORTED DESTRUCTION EFFICIENCY OF PCB'S IN GM's
INDUSTRIAL BOILER2>5
Run
number
PCB-2
PCB-3
PCB-4
PCB in fuel3
concentration
range
•(mg/kg)
34-72
34-76
34-76
PCBb
in
(mg/min)
480-1,000
480-1,100
480-1,100
PCB
out
(mg/min)
<5.8 x ID"2
<5.6 x 10~2
<5.6 x ID"2
PCBC
destruction
efficiency
>99.99d
>99.99d
»99.99d
Note: Density of 1:10 Dilution of Waste Oil: No.. 6 fuel oil is 3.4 kg/gal.
aBased on GM and GCA reported results.
"Fuel combustion rate: 3.43 kg/gal x 4 gal/min fuel flow = 14 kg/min.
PCB in - PCB out
cPercent destruction = 100,
dThis assumes 100 percent sample collection efficiency. A validated test sample
collection efficiency was not conducted as part of this verification burn.
SUMMARY
This Paper can be summarized as
follows:
• One estimate shows that about 41
million metric tons of hazardous
wastes are generated each year
in the U.S.
• Of the 238,000 industrial boilers
in this country, about 43,000 of
them are larger in capacity than
2.9 Mw (10 million Btu's per
hour).
• A first-order kinetic expression
has been developed to model the
waste destruction efficiency to
be expected when hazardous wastes
are co-fired along with conven-
tional fuels in industrial boil-
ers. This expression relates
destruction efficiency to boiler
operating temperature and gas-
phase residence time.
• Superimposition of the model-
predicted waste destruction
curves onto a plot of boiler
temperature/residence times
allows the prediction of whether
a given hazardous waste can be
effectively destroyed in a
prospective industrial boiler.
The kinetic model has been
utilized to predict PCB destruc-
tion efficiency when co-fired
with distillate oil in an indus-
trial boiler and the result was
compared with actual field data.
The model predicted approxi-
mately 99.99 percent waste des-
truction efficiency; this effi-
ciency was corroborated by
analysis of the test burn data.
WHERE DO WE GO FROM HERE?
The verification of any kinetic model
cannot be achieved when there exists only
limited sets of field test data to analyze.
To fill the void, the U.S. EPA has embarked
upon a Testing Program for Hazardous Waste
Co-firing in Industrial Boilers. When the
Testing Program starts producing, in mid-
1982, additional test data from industrial
co-firing operations that are already
underway, the current model can be checked
for its ability to make accurate, repro-
ducible and verifiable predictions as to
what hazardous waste destruction efficiency
is possible in a particular kind of boiler.
83
-------
REFERENCES
1. Benson, S.W. 1960. The Foundations
of Chemical Kinetics, Chapters 10 and
11. McGraw-Hill.
2. Castaldini, C., H.K. Willard, C.D.
Wolbach and L.R. Waterland. 1981. A
Technical Overview of the Concept of
Disposing of Hazardous Wastes in
Industrial Boilers. Draft Report
prepared by Acurex Corporation,
Mountain View, California for the
U.S. EPA, Cincinnati, Ohio. 180 pp.
3. Devitt, T., et al. 1979. Population
and Characteristics of Industrial/
Commercial Boilers in the U.S. EPA
Publication No. EPA-600/7-79-178a.
4. Duvall, D.S., W.A. Rubey and J.A.
Mescher. 1980. High temperature
decomposition of organic hazardous
waste. In: Proceedings of the 6th
Annual Research Symposium on the
Treatment of Hazardous Waste, U.S.
EPA Publication No. EPA-600/9-80-011.
181 pp.
5. GCA Corporation. 1980. Evaluation of
PCB Destruction Efficiency in an
Industrial Boiler. Final Report
prepared for the U.S. EPA, Research
Triangle Park, North Carolina. EPA
Publication No. EPA-600/2-81-055a.
6. Lee, K.C., et al. 1979. Predictive
model of the time-temperature require-
ments for thermal destruction of
dilute organic vapors. 72nd Annual
Meeting of the Air Pollution Control
Association.
7. Olexsey, R.A. 1981. Alternative
thermal destruction processes for
hazardous wastes. To be presented at
the May, 1982 ASME National Conference
on Solid Waste Processing, New York
City.
8. Putnam, A.A., E.L. Kropp and R.E.
Barrett. 1975. Evaluation of National
Boiler Inventory. Final Report pre-
pared by Battelle-Columbus Labora-
tories, Columbus, Ohio, for the U.S.
EPA, Research Triangle Park, North
Carolina, NTIS # PB-248-100. 70 pp.
9. United States Environmental Protec-
tion Agency. May 2, 1980. Background
Document: Resource Conservation and
Recovery Act, Subtitle C Identi-
fication and Listing of Hazardous
Waste, Section 261.31 and 261.32.
Draft Document.
84
-------
BOILER SITE IDENTIFICATION, SAMPLING AND ANALYSIS
PROTOCOLS, AND CHARACTERIZATION OF EMISSIONS FROM BOILER TESTS
Richard S. Merrill
C. Dean Wolbach
Robert McCormick
Larry Water!and
Acurex Corporation
Mountain View, California 94042
The development of the Acurex mathematical model predicting the degree of destruction
of organic waste in industrial boilers must be followed by actual field testing and
further efforts to more quantitatively define the constituents in the exit gas stream.
Sponsored by the Incineration Research Branch, IERL, Cincinnati, Ohio, this subsequent
field testing program includes several support elements. Using public information,
facilities were identified that are cofiring wastes and fuels of interest to EPA; waste
and boiler combinations were prioritized; and each operator was approached to request
his participation. A sampling and analysis protocol was developed to reflect the data
needs of the Office of Environmental Engineering and Technology, the Office of Solid
Waste, and the Office of Air Quality Planning and Standards; it includes a pretest
engineering assessment, test plan preparation, sampling and analysis procedures, and
reporting requirements. Finally, air emissions are being evaluated for principal
organic species predicted to be present as products of incomplete destruction of input
materials, combustion products, and results of incremental changes to the priority
pollutant. This program will establish a data base to support the disposal of
wastes in boilers.
SITE IDENTIFICATION
After contacting EPA officials at
five regional offices, at all 50 state
EPA offices, and in Puerto Rico, we
concluded that state air discharge
permitting officials and, to a lesser
extent, solid waste disposal permitting
officials could provide the most
information on waste cofiring in
boilers. We requested that these
officials identify all sites within
their jurisdiction where potentially
hazardous wastes were being burned in
boilers or process heaters, giving as
much information as possible on the
boilers and wastes burned. . Since budget
and time limitations precluded a search
of permit files, we relied on the
personal knowledge of these officials
and their colleagues to initially
identify such sites, after which limited
information on boiler/waste
characteristics were obtained from the
permit files.
We also conducted a comprehensive
search of 20 data bases abstracting
government reports and technical
literature through the DIALOG
Information Retrieval Service offered by
Lockheed Information Services. This
literature search was extended to the
NEDS, SOTDAT, and EADS environmental
data systems offered by EPA. We made
limited contact with EPA/ORD, EPA/OSW,
and EPA/OAQPS personnel, private
consultants, and industry
representatives as well. Information
compiled from RCRA Part B permit
85
-------
applications, required for all hazardous
waste disposal, was unavailable at the
time of our data search.
Discussion of Results
We identified 129 sites in the
United States (not including the pulp
and paper industry) where hazardous or
potentially hazardous wastes are
currently being burned. Since wastes
were often identified only as "process
residues" or "waste solvents" with
occasional specification of the process
raw material or product, we could not
always determine whether the waste was
hazardous under RCRA 3001 guidelines;
hence, the qualification "potentially
hazardous" waste.
Even less data were available on
boiler characteristics. The apparent
trend was for primarily liquid wastes to
be cofired with fuel oil as the primary
fuel, although natural gas and
pulverized coal were also listed. We
suspect most of the boilers are
watertube designs, ranging from small
industrial heaters to utility sizes. A
significant fraction were apparently
designed for waste cofiring; to
distinquish between boilers and
incinerators with a heat recovery
capability, we identified as boilers
those units having steam generation as
their primary function.
The available data indicate that
cofiring of organic chemical process
byproducts, spent solvents, and waste
oils in boilers is widespread. However,
the sites we identified are not a
comprehensive listing. Many plants
burning wastes, particularly in modified
boilers, do not report this to the
permitting offices and are even less
likely to appear in the literature on
waste cofiring. The lack of time and
funds to search the EPA permit files and
the general limitation of the EPA data
bases to boilers having a total
uncontrolled criteria pollutant emission
potential over 100 tons/yr also preclude
a comprehensive listing. But the sites
identified were sufficient for
representative selection for sampling
and analysis.
Site Prioritization and Selection
Our primary objective was to select
four to six sites for field testing,
covering a range from minimum waste
destruction efficiency to maximum
performance, which would provide the
basis for an across-the-board
environmental assessment of hazardous
waste cofiring in boilers. Our
secondary objective was to select
boiler/waste combinations having
sufficient operational and kinetic data
available to test the Acurex computer
model for waste destruction.
Of the organic constituents listed
as hazardous on the basis of toxicity
under RCRA, light hydrocarbon organic
process wastes are the least thermally
stable and so represent the best
destruction potential. On the other
hand, halogenated or complex cyclic or
aromatic organics are the most difficult
wastes to destroy.
Taking into account the boiler's
time-temperatuare profile and how
representative it is of the industry, we
gave higher priority to standard
firetube, suspension-fired watertube,
and stoker-fired designs burning coal,
fuel oil, or natural gas. Of these
boilers, large watertube designs offer
superior time-temperature profiles for
organic compound destruction while small
firetube designs or small stokers
represent the worst case for such
destruction. We recommend that greater
emphasis be placed on the worst case
testing.
Table 1 lists the boiler/waste
combinations recommended for a six-test
matrix. Final site selection requires
further investigation since data on
boiler design and waste composition for
the identified sites were limited.
Therefore, we separated the sites into
classes based on the available waste
characterization data. Class A sites
are those burning wastes that can be
identified as hazardous. Class B sites
are those burning unspecified process
residues or waste solvents. Class C
sites are those burning waste oils, and
class D sites are those burning
86
-------
TABLE 1. BOILER/WASTE COMBINATIONS FOR A SIX-TEST MATRIX
Boiler type
Waste thermal stability
Large pulverized coal- or residual-oil-fired
watertube design
Large residual-oil-fired watertube design
Small gas/distillate-oil-fired watertube
design
Small gas/distillate-oil-fired watertube
design
Stoker-fired design, intermediate size range
Small gas- or oil-fired firetube design
Low
High
Low
High
High
High
PCB-contaminated oils, as well as all
utility-scale boilers.
We recommend that the class A and
class B sites be contacted for further
information on boiler design/operation
and waste composition before final
selection is made.
SAMPLING AND ANALYSIS PROTOCOL
The sampling and analysis protocol
was developed for use in the test
program to generate a data base from
which regulations and permit application
procedures can be established;
characterize multimedia emissions;
evaluate the impact of waste cofiring on
air pollution control devices; and try and
validate the Acurex predictive model for
waste destruction. Developed for
solid-fuel-fired boilers burning wastes
with a high thermal stability, the
protocol includes instructions for
conducting the field test program,
analyzing the collected samples, and
preparing the sampling/analysis reports.
Pretest Site Survey
This 1-day survey involves
collecting waste and fuel samples,
obtaining boiler design and operating
data, acquiring meteorological data, and
performing a site sampling evaluation.
From the waste and fuel samples,
the major organic components present are
identified and quantified. To minimize
the quantification effort, the
investigator should obtain as much
information as possible about the
process generating the waste and its
likely constituents. These samples will
also determine the ultimate and
proximate analysis of each, gross flue
gas compositions, boiler heat release
rates, and trace element content of each
for verification of the material balance
estimate.
Samples should be labeled and
logged on a data sheet. The
investigator should determine from plant
personnel that these samples are
representative of wastes to be fired
during the field test program.
From the boiler design and
operating characteristics, the
destruction efficiencies for organic
constituents and the formation of
products of incomplete combustion are
predicted. From meteorological data
available at the nearest weather
87
-------
Preliminary
Engineering
Assessment
Pretest
Site
Survey
Boiler
Engineering
Evaluati
Data
on
TDAS Analysis for
PICS Identification
and Decomposition
Rate Data
Boiler
Time-Temperature
Profile
Development
Identification
D9I-
PICS
Decomposition
Rates
DC torsi ration of
POHCS, PICS 1n
Fuel and PICS
In Haste
Waste and Fuel
Elemental
Analysis
1
Site Sampling
Engineering
Dhta Evaluation
Evaluation of
Available Emission
and Meteorological
nata
Estination of
Boiler Destruction
of POHCS and PICS
Establish Elemental
Mass Balance
, .,, 4*
Analytical Criteria
Establishment
of Test, Sampling,
and Analytical
Matrix
End
Figure 1. Preliminary engineering assessment information flow ehart.
88
-------
station, the maximum acceptable stack
emissions are calculated as a function •
of exposure and risk.
The site sampling evaluation covers
the plant configuration, components to
be sampled, each sampling point, its
accessibility, necessary modifications,
and other sampling requirements, as well
as information on provisions for the
test crew's safety, lodging, etc.
All data sheets and samples should
be returned to the contractor's office
for an analysis that may require up to
4 weeks. Elements involved in this
preliminary assessment are outlined in
Figure 1.
Waste and Fuel Evaluation
The waste and fuel samples are
analyzed to determine and quantify the
primary organic constituents present and
the waste/fuel combustion
characteristics; predict destruction
efficiencies and products of incomplete
combustion; and supplement the material
balance estimates. The analytical
protocols and the organic compounds of
interest are discussed in the Acurex
report prepared under this contract.
Determining the ultimate and
proximate analysis of the waste and fuel
allows the approximate stack gas
composition to be calculated at a given
air stoichiometric ratio. The results
of thermal gravimetric analysis for
volatilization rates and differential
scanning calorimetry for heats of
vaporization determine the boiler
time-temperature profile when cofiring
wastes, for input to the predictive
model. Finally, a Thermal Decomposition
Analytical System (TDAS) determines
potential, reaction products and their
decomposition rates; these data, with
the decomposition rates of the organic
constituents identified in the samples,
are also used in the predictive model.
The analytical procedures 'for obtaining
all these data are discussed in the same
report.
Boiler Evaluation
By establishing a
time-above-temperature profile for the
combustion zone, conditions can be
determined for destroying a given
waste. Time-temperature profiles are
computed using an Acurex zonal heat
balance computer code. Inputs include
the fuel and air properties and boiler
operating and design features obtained
at the site, and other variables such as
turbulence or emissivity based on our
experience with combustion processes.
Time-above-temperature profiles,
indicating the length of time combustion
gases are above a maximum; mean
temperature, are then generated from the
computer outputs. Overlaying a given
compound's time and temperature
destruction requirements on a boiler's
time-above-temperature profile allows
various compounds to be screened for
destruction efficiencies by noting where
the compound line /alls under the boiler
curve.
Component Destruction Efficiency and
Reaction Product Formation
Comparing the possible boiler
temperatures from the time-temperature
profile with the TDAS output at those
temperatures, allows those organic
species still present to be identified
and quantified and so determine the
primary species to look for in the stack
gas.
The concentration of these
components, which determines the length
of each stack test, are calculated using
the equation
In
where
f x
-A J exp C-E/RT(t)] dt
tm
C = Concentration of material
at time At
C0 = Concentration of material
at time o
A = Arrenhuis preexponential
factor
t0 = Time at Tmax
tx = Mean residence time
89
-------
T(t)
Reaction energy of
activation (cal/g mole)
Gas constant
(1.987 cal/g mole °K)
Temperature (°K) as
function of time in boiler
For compounds of known A and E, the
concentration (C) can be calculated.
When A and E are unknown, the IDAS data
must be further manipulated, using the
retention time at a given temperature
and the ratio of outlet to inlet
concentrations for a variety of test
cases to compute A and E.
Given a known input concentration,
flue gas volume, and temperature, stack
gas concentrations can be estimated.
Establishing the Test Matrix
Using the data assembled in the
preliminary assessment, a test plan is
formulated by balancing the size of the
program against available resources.
This test program assumes a large
solid-fuel-fired boiler is to be
sampled. The formulation process is
outlined in Figure 2.
For each operating condition to be
tested, including waste-to-fuel ratio,
boiler load, or excess air ratio, a
baseline test should identify discharges
typical of the primary fuel, followed by
triplicate tests at the desired
waste-to-fuel ratio.
Sampling points should be
established to include all fuel and
waste input streams, bottom ash
discharge, and inlet and outlet streams
from the air pollution control
device(s). Operating data requirements
should be established to include boiler
load, waste and fuel temperature, and
feed rate, process temperatures, and
data to determine fuel combustion
efficiency and operational stability.
Sampling methods are dictated by
the preliminary assessment. For solid
streams (i.e., the solid fuel or bottom
ash), grab samples should be taken at
hourly or shorter intervals. Liquid
samples (i.e., waste) should be
collected by grab sampling or automatic
compositing samples. Gas phase samples
should be collected using an EPA
Method 5 or SASS train.
Analytical sensitivities should be
reviewed against detection limits.
Sample sizes should then be set
according to the predicted constituent
concentrations and detection limits of
the sampling equipment.
Taking a set number of duplicate
samples, split samples, spike and
recovery sequences, and surrogate
samples for each set of test samples
ensures meaningful data. Each sample
should be labeled and identified on the
chain of custody and analysis request
forms.
The requirements of the test plan
should be compared to the available
resources. Funding constraints may
require that test parameters be
rewritten or low priority options
eliminated.
Preparing the Test Plan Document
The test plan document should
include the elements listed in Table 2.
Acurex computer simulations and
predictions will be included as well,
along with all pretest data. The EPA
technical project monitor and the plant
operator must approve the test plan.
Field Testing
Figure 3 outlines the field testing
process. A typical test program
includes one baseline test, firing the
boiler on fuel without waste, and three
runs cofiring the boiler on fuel and
waste standard settings. One day is
required for setup and the Methods 1
through 4 stack gas analyses which
involve measuring gas velocity and
temperature, determining the presence of
90
-------
Conduct
Sampling and
Analysis
Design Test
Program
From Pretest Site
Survey Determine
Maximum Info Requ,
to Satisfy
Program Objective
Can
Operating
Matrix be
Reduced
Establish Matrix
of Possible Oper.
Conditions
to be Tested
Submit to
TPH for
Approval
Is
Matrix
Maximized
Establish Matrix
of Samples and
Data Required for
Each Test Sequence
Are
There Sufficient
Resources
Can
Sample Matrix
Be Reduced
Is
Matrix
Maximized
Estimate Resource
Requirements for
Test Sequence
Identify Sampling
Methodology for
Individual Samples
and Data ~
Are
QA/QC Requirement
Satisfied
Identify
Analytical
Methodolony
Prepare
Project
QA/QC Plan
Are
Methodologies
Compatible
Have
All Samples
Been Reviewed
Review Analytical
Sensitivities Against
Desired Detection
Limits
Are
Sample Sizes
Compatible with
Methods
Figure 2. Matrix for preparation of testing program.
91
-------
TABLE 2. TEST PLAN DOCUMENT ORGANIZATION
Section
Description
1. Program Objective
2. Pretest Site Survey
3. Field Test Program
4. Analytical Methods
5. Cost and Schedule
Agreements with plant contractor commitments
Test matrix, identifying samples, equipment,
test times, personnel, and schedule
Process operating data
Sample storage/transport
Sample identification
Description of analyses in generic terms
QA/QC procedures
Site-specific cost impacts, manpower
requirements, and schedule
C02» 02, H20, and Ng, and
calculating molecular weight. Baseline
tests are run on the second day and
triplicate tests on the following
3 days, including cleanup and teardown.
A sixth day is used to finish
uncompleted tasks.
The team leader records the data,
one person staffs the continuous monitor
van, one person takes grab samples, and
a team runs the particulate trains. For
tests on solid-fuel-fired boilers, as
many as seven people may be needed.
Equipment Preparation
Sampling equipment is calibrated
and checked before field testing;
calibration sheets for gas meters, pilot
tubes, and magnahelic*gages are
completed and filed. Specialized
equipment or site modifications should
be arranged before arrival at the site.
Field Sampling
Preliminary gas characteristics are
determined using Methods 1 through 4 at
the beginning of each sampling task and
as necessary to account for changes in
process conditions.
Source sampling is conducted using
continuous monitors to record gaseous
emissions and operational variability,
and isokinetic-sampling to collect
organic and particulate materials. The
Method 5 particulate sampling train is
traversed to sample at several points
across the duct. The Source Assessment
Sampling System (SASS) which samples at
a single point in the duct has three
particulate sizing cyclones, an organic
module with XAD-2 sorbent, and impinger
bottles for trace metal capture. This
is used to collect larger sample sizes.
Process data measured to compute
boiler efficiency and evaluate cofiring
impacts on unit operation will be
collected. These operating data will be
used to estimate unit efficiency and
reduce the emissions sampling data.
92
-------
Sampling
and
Analysis
Field
Test
Program
_L
Conduct
Sampling
Activities
±
Gather
Operational
Data During
Testing
_L
Determine
Concentrations
Of POHCS & PICS
in Samples
Reduce Sampling
Field Data
(volume flow
Rates, etc. )
Determine
Concentration
of Elements
in Samples
Establish
Elemental
Material
Balance
±
Deternrine POHC
PIC Destruction
Efficiencies
Compare
Results to
Model
I
Prepare
Report
Figure 3. Sampling and analysis activities.
93
-------
Analysis
Data Reduction
Posttest analyses of fuel and waste
samples, ash grab samples, and the SASS
sample catch may include a flue gas
analysis of the SASS catch, inorganic
analyses for selected trace metals and
other species depending on the fuel and
waste composition, and organic analyses
when it is uncertain which organic
species are present. Should these data
be already available, these analyses may
be eliminated.
Semi- and low-volatility organic
compounds are determined with EPA Method
625 involving base/neutral extraction of
the sample and extract concentration;
the extract is analyzed by combined gas
chromatography/mass spectrometry
(GC/MS). Each sample is spiked with a
surrogate standard prior to extraction,
and quantitation is based on comparing
specific ion areas in each analysis to
that of a standard. Other organic
compounds are determined with Method 625
using acid extracts. For both
base/neutral and acid extracts, multiple
injections of diluted and concentrated
extracts are required to quantitate
trace and high level species.
Volatile organics are determined
using the purge and trap/GC/MS EPA
Method 624. Acurex uses a Tekman LSC2
automated purge and trap device
interfaced to a Finnegan 4023 GC/MS.
Once the grab sample has been
screened and the pollutants for that
particular matrix identified, routine
methods will be developed to confirm
that known methods work for the samples
in question. These routine methods are
based on the 600 series such as Method
610 for high-pressure liquid
chromatographic (HPLC) determinations.
All methods are evaluated for precision
and accuracy using replicate analyses
and spiked samples.
All analyses should be completed
within 6 weeks of sample receipt. All
samples are saved until the final report
has been accepted by the EPA technical
project monitor.
The analytical results will
establish the boiler operating
parameters and combustion efficiency,
destruction efficiency (per component
and total), PIC emissions, air pollution
control device efficiency, material
balance closure, chloride balance, and
particulate emissions and size
distribution.
The reduced data will be compared
to the Acurex model predictions.
EVALUATION OF AIR EMISSIONS
POHC emissions can be estimated by
conducting a zonal heat balance,
estimating volatilization rates,
determining the degree of POHC
destruction, and then calculating the
emission rate. Products of incomplete
combustion pose greater difficulties and
are discussed in the following
paragraphs.
Products of Incomplete Combustion
The complex mechanisms involved and
the lack of information on such
mechanisms hamper the prediction of the
formation/destruction of products of
incomplete combustion (PIC's). Using
chemical rate theory, however, upper
bounds can be established by reducing a
set of-pseudo first order kinetic
equations to simpler cases that can then
be used to develop concentration
relationships for PIC formation and
destruction.
A set of reaction schemes was
developed mathematically with each case
representing the disappearance of one
component (A) and the formation and
disappearance of a product component (?)
as follows:
Case I —
Case II
All of A converts
to P and then P
disappears
A converts to many
different products
94
-------
Case III —
Case IV —
P is generated from
many different
components
P is generated by a
sequence of reactions
in series
following relationships hold (n is the
number of steps in the sequence):
Case V -- A converts to P by two
different mechanisms
To find Cp/C/\j0 or the
concentration of P'at time t with
respect to the original concentration of
A, we: (1) reduced one or more cases to
a simpler, upper bound case; (2) reduced
one or more cases to the simple case of
the disappearance of P, and; (3)
directly compared the concentration of P
at time t with that of A at time t, both
with respect to A's original
concentration.
In case I, Cp/CAj0 can be
calculated using the empirical pseudo
first order disappearance rates for POHC
(A) and PIC (P). Calculating the time
at which Cp/C/\ 0 is a maximum
enables us in many cases to treat the
PIC (P) as a POHC. This occurs if the
rate of disappearance of P is greater
than 100 seconds-"*. Most of the other :
cases can be reduced to case I as an
upper bound.
For case II, the concentration of
any one P in this system can be
intuitively and analytically verified as
less than the concentration of P in case
I. Therefore, case II reduces to case I
as an upper bound.
For case III, the system is
mathematically equivalent to a system
where a POHC proceeds to a given PIC by
several mechanisms. This also reduces
to case I as an upper bound by the
simple expedient of assuming that all
the various waste components proceed to
the PIC of interest. This case is not
true, but it is a worst case. Case IV
represents a much more complicated
problem, and a "good" upper bound has
not yet been established for the general
case. With some stringent limitations,
solutions are available. For example,
if all the rates are equal, and if all
the concentration of the intermediate
product are zero at time zero, the
(k!t)n
exp (-kt),
= • and
exp (-n)
For the situation where all the
intermediate rates are faster than the
rate of disappearance of the PIC, it can
be shown that
max < F~
or, in general,
n
mm
However, the corresponding
relationships for concentration are not
true. At this time, it can only be said
that for t > n/km-jn
CT
A,o
« exp (-kt).
Case V has not been thoroughly
explored. It has been established that
kmin.
and for the situation where all
intermediate rates are greater than
and t > n/kp
< 2 kpt exp(-kpt).
In summary, steps have been taken
to set upper bounds on PIC emissions.
Some bounds have been set under
95
-------
stringent conditions that do not
represent the real world. Our approach
to more realistic bounds is currently
being pursued and looks promising.
Trace Element Emissions
These elements should partition as
in coal or residual-oil-fired boilers,
the quantity of any given metal
depending on its concentration in the
fuel, boiler combustion conditions, the
type and efficiency of the air pollution
control device (APCD) as a function of
particle size, and the properties of the
element itself.
Trace metals tend to concentrate in
certain waste particle streams of a
boiler, leading to the following
classifications by stream for
partitioning metals:
Class I — Bottom ash or slag and
in the APCD inlet and
outlet flyash in
approximately the same
mass concentrations
Class II — Flyash, mass entering
being less than mass
exiting
Class III — Vapors
Class I elements and oxides are not
volatilized in the combustion zone but
remain condensed throughout the boiler
cycle. Class II elements, however, do
volatilize and so concentrate in flyash
rather than slag; these elements either
form condensation nuclei or condense on
existing particulate. Waste burning is
not expected to affect trace metal
emissions.
Criteria Pollutant Effects
Waste firing will impact the
criteria pollutant emissions from a
source. The criteria pollutants of
concern are particulates, SOX, NOX,
and CO. In general, it can be expected
that emissions will change on a
unit-mass-for-unit-mass basis, although
this will vary from pollutant to
pollutant. If pollutant is generated
from element in the fuel, and the
element is in concentration Cf in the
fuel, then the original pollutant
emission rate will be kRfCf where k
is the fraction of the element converted
to the pollutant and Rf is the fuel
feed rate in mass per unit time. If the
conversion fraction is the same for the
fuel and the waste, the new emission
rate will be k(R'fCf + RWCW)
where R'f is the new fuel feed rate, R
is the waste feed rate, and Cw is the
concentration of element in the waste.
Particulates
Particulate emissions will probably
follow this pattern based on ash
content. The distribution of ash
between bottom ash and flyash may change
significantly depending on the nature of
the waste ash.
Units not designed to handle
ash-containing fuels (e.g., gas- and
distillate-oil-fired units) will
probably not fire wastes containing ash
or mineral-forming elements. Thus,
noncarbon particulate emissions are not
expected from these units. Carbonaceous
materials (soot) may increase or
decrease depending on the burnout rate.
This is usually controlled under normal
boiler operating procedures and is not
expected to change under waste firing
conditions.
Units designed to handle high-ash
content fuels such as the various
coal-fired boilers may experience wide
fluctuations in particulate emissions.
It is expected that, for the most part,
the fluctuations will be on the
downside. This prediction is based on
the fact that most high organic content
wastes with high fuel potential .will
have a lower ash content than the
replaced fuel.
Sulfur Oxides Emissions
The generation of sulfur oxides
will be directly related to the
incremental change in sulfur content of
the fuel-waste mixture. Reactivity of
flyash to sulfur oxide chemisorption
will effect the emissions. The process
96
-------
is straightforward but not simple.
Normally, the emissions would be
expected to be
ESOX = CS,FRF + C
- AS,WRW,A
where
ESQV = Emission rate, of SOX
A " • • "
CQ p = Molecular weight factor times
' concentration of sulfur fuel
C$ w = Molecular weight factor times
concentration of sulfur waste
Rp = Fuel feed rate
RW = Waste feed rate
AS j = Unit chemisorptivity of
SOX on fuel ash
AS,W = Unit chemisorptivity of
SOX on waste ash
Rp^A = Rate of ash generation
from fuel
RW,A = Rate °f asn generation
from waste
However, the synergistic effects of
the fuel and waste ash chemistries will
probably change the unit chemisorbtivity
factors in other than a linear manner.
Thus, more or less sulfur oxides may be
chemisorbed in the ash (on a unit mass
basis) than would normally be expected.
Waste and fuel ash chemistries also may
cause alterations in the sulfur trioxide
(503) to sulfur dioxide (S02)
ratio. Oxidation catalysts such as
vanadium and titanium are known to shift
the distribution to higher $03
content. These types of elements can be
expected to be present in many of the
waste streams under consideration for
boiler destruction.
NOx Emissions
Nitrogen oxides (NOX) are formed
in combustion processes through two
basic mechanisms: The fixation of the
nitrogen in the combustion air by
oxygen, termed thermal NOX, and the
oxidation of fuel-bound nitrogen termed
fuel NOX. The rate of thermal NOX
formation is highly dependent on local
temperatures and, less so, on 02
concentrations. The detailed mechanisms
of fuel NOX formation are poorly
understood, but fuel NOX formation
rates are most affected by local 02
concentrations, and are relatively
independent of temperature.
Since thermal NOX formation is
dependent on temperature and 02
concentrations, changes in a boiler's
combustion conditions attendant with
waste cofiririg may affect thermal NOX
levels. , For example, if excess air
levels are increased to facilitate
increasing destruction efficiencies,
then emitted NOX levels will
increase. Perhaps of greater importance
from a thermal NOX standpoint, though,
is if the heat content of the waste
cofired is sufficiently different than
that of the fuel displaced, then the
maximum flame temperature change along
with corresponding effects on thermal
NOX prediction. For example, if the
heat content per unit mass of the waste
cofired is substantially higher than the
fuel displaced then flame temperature
will increase and thermal NOX
prediction will increase. Corresponding
effects leading to lower thermal NOX
formation occur when waste content is
substantially less than that of the fuel
displaced.
Since fuel NOX formation derives
from oxidation of fuel- (or waste-)
bound nitrogen, emitted NOX levels
through this mechanism will be affected
depending on the relative nitrogen
contents of the waste cofired and the
fuel displaced. For example, if a high
nitrogen waste is cofired in a
distillate-oil-fired boiler (very low
nitrogen), emitted NOX levels due to
fuel NOX will increase. The increase
will not reflect total conversion of the
added waste nitrogen to NOX, though.
Total conversion of fuel nitrogen to
NOX does not occur for any combustion
process. Typical percent fuel nitrogen
conversions range from about 20 percent
for high nitrogen (1 to 2 percent N)
fuels such as some bituminous coals to
about 80 percent for very low (less than
0.1 percent N) fuels such as distillate
97
-------
oils (H. B. Mason, et al., Preliminary
Environmental Assessment of Combustion
Modification Techniques — Volume II,
Technical Results, EPA-600/7-77-119b,
October 1977).
In addition, however, other changes
in boiler operation attendant with waste
cofiring may have significant effects on
fuel NOX production from fuel- and/or
waste-bound nitrogen. As was the case
for thermal NOX formation, if excess
air levels are raised to help promote
waste destruction, then thermal NOX
formation will increase as well. If the
burner is retrofit to increase flame
distribution and mixing (thereby
removing fuel-rich eddies) then fuel
NOX formation will increase.
In summary, NOX formation by both
mechanisms will be increased by positive
incremental changes:
• Btu con tent/unit mass
• Heat release rate
• Excess air
• Flame turbulence
Carbon Monoxide
Carbon monoxide levels should not
be affected by combustion of wastes.
This is because carbon monoxide
concentrations are mechanically
controlled to maintain boiler
operational efficiency.
98
-------
EVALUATION OF FEASIBILITY OF INCINERATING HAZARDOUS WASTES
IN HIGH-TEMPERATURE INDUSTRIAL PROCESSES
F, D. Hall and W. F. Kemner, PEDCo Environmental, Inc.
Cincinnati, Ohio 45246-0100
L. J, Staley, Project Officer, U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
In the search for disposal alternatives, the U.S. Environmental Protection Agency is
evaluating the potential use of high-temperature processes for the incineration of haz-
ardous wastes. Many kinds of waste have already been disposed of in boilers and cement
kilns; this report considers the many other potential processes, such as metallurgical
furnaces, brick and lime kilns, glass furnaces, and sewage sludge corabustors. Each
process is examined against such criteria as time/temperature profile, geographical
location, product quality considerations, institutional factors, and environmental
impacts. Promising alternatives are identified.
INTRODUCTION
The Incineration Research Branch of
the U.S. Environmental Protection Agency
(EPA) has a variety of ongoing programs to
investigate, develop, and promote inciner-
ation as a means for ultimate destruction
of hazardous wastes. One of these pro-
grams concerns the use of high-temperature
industrial processes. The first processes
to be investigated under this program were
industrial boilers and cement kilns, which
were chosen partially because of their
ubiquity and partially because of the
interest many boiler and kiln operators
expressed in recovering the heating value
of wastes. At this point in time several
demonstration test burns are either com-
pleted or underway.
Interest in other processes prompted
EPA to initiate a feasibility screening
study for all high-temperature processes.
Such a study would provide consistent and
uniform criteria for selecting those
processes that offer feasible combustion
conditions and that should be examined
further for overall suitability.
PROCESS SELECTION CRITERIA
The first task was to select common
industries using high-temperature proces-
ses. Table 1 summarizes the screening
criteria against which each process was
evaluated.
TABLE 1. SCREENING CRITERIA FOR POTENTIAL
HIGH-TEMPERATURE PROCESSES FOR THE
DESTRUCTION OF HAZARDOUS WASTE
0 Compatibility of process
- Thermal destruction conditions:
temperature and residence time
(turbulence not considered)
- Product quality
- Potential for fugitive emissions
0 Number of facilities
0 Geographic matching of industry loca-
tion and waste generators
The initial list of processes was
based on operating temperature; any major
process with an operating temperature of
greater than 1200°F was included. The
99
-------
initial listing of approximately 100
processes included the metallurgical,
chemical, and mineral industries and
sludge and waste incinerators. When the
other criteria in Table 1 were applied,
several of these processes were eliminated
from consideration.
Gas residence time within the process
at the high-temperature condition was an
important consideration; the lower the
process temperature, the longer the resi-
dence time required for 99.99 percent
thermal destruction. Turbulence, the
other condition necessary for combustion,
was not considered because little informa-
tion is generally available on this param-
eter. Residence time and product quality
considerations eliminated organic and
miscellaneous chemical manufacturing
processes from consideration. These
processes often attain high temperatures
in reactors, but residence times are quite
short.
Product quality, both actual and
perceived, was an important consideration.
Destructing hazardous waste in a steel
furnace, for example, could increase trace
elements in the steel and affect its
physical properties. In all cases hazard-
ous waste disposal was considered second-
ary to product quality. This criterion
eliminated processes such as basic oxygen
and electric arc furnaces in the steel
industry because these processes produce a
high-quality steel that cannot tolerate
residues from hazardous waste destruction.
This criterion also affected chemical
manufacturing, food processing (human and
animal consumption), and petroleum refin-
ing processes.
The more potential a process had for
fugitive emissions, the less likely it is
to be suitable for hazardous waste dis-
posal. Raw materials handling procedures
affect the potential of a process for
fugitive emissions and for hazardous waste
disposal. In a sintering process at an
iron and steel mill, for example, open
conveyors are used to transport raw mate-
rials to the pug mill (mixing device) and
to the sinter strand, which is also open.
Because of the major modifications to the
materials-handling system that would be
required to correct this high potential
for fugitive emissions and because of the
short residence time at the maximum
temperature of about 2200°F,* the iron and
steel sintering process was eliminated
from consideration.
The number of facilities available in
the hazardous waste generation areas was
another major consideration. For example,
specialty processes that existed at only a
few widely scattered locations or develop-
ing processes that were currently opera-
tional at only a few sites were not con-
sidered. The processes selected for
detailed analysis were those that'had the
potential for disposing of significant
quantities of hazardous wastes.
Table 2 lists those industries and
processes that remained after several
processes had been eliminated from further
consideration by the application of the
screening criteria. These processes were
evaluated in more detail by use of a
hazardous waste destruction model derived
from an earlier model developed by Acurex
Corporation (1).
HAZARDOUS WASTE DESTRUCTION MODEL
The efficient destruction of hazard-
ous wastes in any industrial process is
primarily dependent on the temperature and
residence time of the combustion products
in the furnace employed by the process.
Hence, a matching of the residence time-
temperature relationship for the indus-
trial process with that required for
complete destruction (99.99%) of the
hazardous waste should enable one to
identify potentially destructible hazard-
ous waste categories. Based on kinetic
data for the constituents in the wastes,
characteristic curves showing various
time-temperature combinations for destruc-
tion of the wastes were obtained. For the
industrial process, knowing the furnace
volume, volumetric flow rate of gases, and
temperature range of the furnace made it
possible to plot "process curves" that
showed the residence time at which the
furnace gas is above a certain temperature
within the furnace. By laying the "proc-
ess curves" over the hazardous waste
characteristic curves, it was possible to
Refer to conversions table at the end of
the paper for metric equivalents to
English units used throughout.
100
-------
TABLE 2. INDUSTRIES AND PROCESSES SELECTED FOR FURTHER HAZARDOUS
WASTE DESTRUCTION EVALUATION
Industry
Brick
Carbon black
Primary copper
Primary lead
Iron and steel
Lime
Glass
Sewage sludge
Hazardous waste
incineration
Exit and maximum
Process or furnace temperatures (°F)
Tunnel kiln
Oil furnace process
Reverberatory furnace
Blast furnace
Blast furnace
Open hearth furnace
Long rotary kiln
Short rotary kiln with
stone preheater
Melting furnace
Fluid-bed furnace
Multiple-hearth furnace
Rotary kiln
Liquid injection
500-2150
1600-2550
2370-2570
1300-2200
2000^-3400
2200-3250
1265-3440
2115-3340
1140-2700
1400-1600
900-1800
1500-2900
1300-3000
Residence time
(seconds, unless
otherwise noted)*
4.3
1.1
2.2
5.9
1.1? •
2.0
8.3
7.6
4.1
1.4
0.5
2.0-2 hours8
0.5-2
Residence time above the greater of the exit temperature or 1400°F.
Represents estimated temperature at the top of the combustion zone in the
blast furnace.
T Residence time has been calculated using 33% of the total furnace volume
#
Included for comparison only.
o
8 The residence time for solids and sludges can be adjusted.
101
-------
predict the waste categories destructible
at various temperatures. The required
99.99 percent destruction is predicted for
those wastes whose lines are below and to
the left of the process curves.
During a preliminary review the
following waste categories were eliminated
from consideration because of their low
destructibility by high-temperature pro-
cesses or their high explosion potential:
inorganic pigments, inorganic chemicals,
explosives, iron and steel, and secondary
lead. Other wastes listed in the RCRA
Background Document (2) were included in
the model and grouped according to non-
specific sources (F series) and specific
sources (K series). These categories were
plotted according to the required tempera-
ture and time for 99.99 percent destruction
by use of the relationship:
£n At = Sin
where
At - time required to reach 99.99 per-
cent destruction, s
A = Arrhenius pre-exponent frequency
factor (s-1)
E = Energy of activation (Btu/1fa-
mo! e)
R = Universal gas constant (1.987
Btu/lb-mole - °R)
T = Absolute temperature (°R)
Figure 1 shows the destruction lines for
the selected hazardous waste categories.
Process curves were overlayed on the
destruction lines to determine the hazard-
ous waste categories that should be con-
sidered for destruction. These process
curves were calculated for the gas resi-
dence time above specific temperatures by
use of the following calculation:
t - 31800 V T
u — ~A' ' y-*. • '•'«.—I- jin =—
CTe - V
m
where
t = residence time above a given temper-
ature (T), s
Te = process exit temperature, °R
T = process maximum temperature, °R
V = volume of process temperature zone,
ft3
Q = volumetric flow rate, scfm
This relationship assumes that gas temper-
ature varies linearly with axial distance
between the point of maximum temperature
(T ) and the exit temperature. The rela-
tionship was plotted for each process for
both the mean residence time as calculated
by the equation and the "fast path" resi-
dence time (0.5 t). Figure 2 shows the
process curves for a short rotary lime
kiln.
EXAMPLE PROCESS ANALYSIS
The process analysis included model-
derived technical time-temperature consid-
erations, geographic matching of industry
location and waste generators, environ-
mental factors, and institutional factors.
Although all the processes listed in Table
2 will destruct at least some categories
of waste, application of all the analysis
criteria indicates that lime kilns provide
one of the best candidates.
Figure 2 also shows the time-tempera-
ture relationship for rotary lime kilns.
Figure 3 shows the 1978-79 geographical
distribution of lime kilns according to
the Department of Interior (2). Theoret-
ically, all the waste categories consid-
ered could be destructed in a typical lime
kiln, and as the figure shows, the geo-
graphic distribution of these kilns is
wide.
Most lime kilns are well controlled
by high-energy scrubbers, fabric filters,
or electrostatic precipitators. Regula-
tions dictate the use of a scrubber for
air pollution control of some wastes
(e.g., those containing halogenated hydro-
carbons). Any increase in particulate
loadings from the burning of hazardous
wastes should not create a problem, par-
ticularly if the plant is now operating
well within allowable particulate emission
limits.
Environmental problems could result,
however, from the potential formation of
hazardous products of combustion. The
model considers only the destruction of
1Q2
-------
o
o
CM
CO
I I
O
O
un
CM
I I
J I
01
o
S-
O)
Q.
CTl CO
<4- (O
(J
t/1
4J O)
C 4-3
O. O
E <1)
I =3
O)
-a
S-
13
Dl
o
o
CO
'1
103
-------
0)
104
-------
04
0)
to
"a.
O)
to
0)
O
-o
s~
O)
0)
CD
S-
1Q5
-------
the waste components and does not predict
the compounds generated during combustion.
Institutional factors could be the
greatest obstacle to disposing of hazard-
ous wastes in lime kilns. Preliminary
contacts with industry representatives
have been favorable, however, because of
the potential fuel savings from burning
combustible wastes.
CONCLUSIONS
The process evaluations and discus-
sions with many persons in industry and
government lead to the conclusion that the
major impediments to the use of high-
temperature processes are both technical
factors and institutional factors. The
economics of each case have not been
evaluated in detail, but will clearly be
proportional to the energy cost pressures
and the degree of hazard associated with
the waste. The term hazardous waste is an
unfortunate one because it raises a uniform
sceptre of alarm for a universe of mate-
rials which vary from deadly toxins to
everyday household substances. This
raises public concern over utilizing
existing high temperature sources, most of
Which are located in urbanized areas.
Similarly, it magnifies worker concern
over exposure and potential accidents.
The driving force to utilize high-
temperature processes must come from the
companies involved. The availability of
low- or zero-cost energy combined with
possible disposal fees from generators
must be sufficiently attractive to justify
the effort required to conduct test burns,
make the necessary modifications, and con-
vince disparagers of the overall environ-
mental benefit of this approach.
3. U.S. Environmental Protection Agency.
1980. Resource Conservation and
Recovery Act, Subtitle C; Background
Document. Office of Solid Waste,
Washington, D.C.
To convert
from
CONVERSION TABLE
To
Multiply by
°R
ft3
scfm
Btu/lb-mole
Btu/lb-mole °R
°c
°K
m3
Nms/s
J/kg-mole
J/kg-mole °K
5.56 (°F-32)
E-01
5.56 E-01
2.832 E-02
4.383 E-04
2.324 E+03
8.314 E+03
REFERENCES
1. Acurex Corporation. 1981. A Tech-
nical Overview of the Concept of
Disposing of Hazardous Wastes in an
Industrial Boiler. Prepared for U.S.
Environmental Protection Agency,
Cincinnati, Ohio.
2. U.S. Department of the Interior.
1980. Minerals Yearbook, Volume I,
Metals and Minerals, 1978-79. Bureau
of Mines, Washington, D.C.
106
-------
A SUGGESTED LABORATORY APPROACH TO SIMPLIFICATION OF THE
POHC - PIC DILEMMA*
Frank C. Whitmore
Principal Scientist
Versar, Inc.
Springfield, Virginia 22151
; Richard A. Carnes
Environmental Scientist
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Wayne A. Rubey
Environmental Engineer
University of Dayton
Dayton, Ohio 45469
INTRODUCTION
Recently promulgated regulations gov-
erning the operation of a hazardous waste
incinerator (1) are a significant departure
from those that govern the thermal disposal
of PCBs and PCB-contaminated wastes (2).
In the latter case, the PCB disposal regu-
lations carefully specify the range of al-
lowable operating conditions that are re-
quired to obtain a permit for the incinera-
tion regulations. Regulations issued under
the mandate of the Resource Conservation
and Recovery Act (RCRA) leave far more to
the engineering judgement of the individual
permit writer. Specifically, once those
hazardous organic compounds in the waste
that will most seriously challenge the
particular incineration technology, the
principal organic hazardous constituents
(POHCs) are chosen, it is then required
that the necessary operating conditions for
the incinerator be identified that will in-
sure the attainment of the destruction and
removal efficiency (ORE) that is required
for each POHC, i.e., a ORE of at least
99.99 percent.
Obviously, this approach requires a
considerable amount of information, both
*POHC - Principal Organic Hazardous Con-
stituent(s)
PIC - Product(s) of Incomplete Combustion
for the would-be permittee, as well as for
the permit writer. The question of avail-
ability of thermal stability information
will be treated below. Here it is import-
ant to discuss the problems that are in-
herent -in the requirement that the POHCs
be first identified from the large number
of compounds that make up a typical indus-
trial waste stream. We are specially con-
cerned with those waste that have been
classified and/or listed by the Agency as
hazardous. Therefore, one might suspect
that the compounds that cause the waste to
be classified as hazardous would be the
most likely POHCs. An example that illus-
trates this point is benzene. A waste con-
taining benzene would likely be classified
as hazardous due to the fact that it is a
suspected carcinogen. However, benzene
also contains sufficient heat content, more
than 15,000 Btu/lb, thus it has a consider-
able fuel value. Such a compound, although
properly classified as a POHC, would not
seriously challenge a properly operating
incinerator (i.e., one operating in a fuel
lean oxidizing environment).
A waste containing hexachlorobenzene
(HCB) may not be classified as hazardous
simply based on its known toxicity. How-
ever, HCB is known to be one of the most
thermally stable organic compounds under
most incineration conditions and must be
1Q7
-------
considered as a POHC for any waste known to
contain this particular compound. What is
needed is a distinction between those com-
pounds that are toxic by themselves and
those compounds that present a challenge to
the incinerator so a better POHC classifica-
tion system can be developed. One way to
do this would be through the detailed chem-
ical analysis route. On the basis of the
identification of the chemical constituents
of a waste, a comparison of the specific
thermal characteristics can.be made. The
identification of the POHCs can then be
made. Unfortunately, this body of inciner-
ability data does not presently exist and
the process of attaining it could be prohi-
bitively expensive.
In recognition of the fact that there
simply was not enough data on the* incinera-
tion characteristics of even the most com-
mon industrial compounds, and the fact that
full-scale test burns would.be. far too ex-
pensive and time consuming, 'the U.S. En-
vironmental Protection Agency (USEPA) spon-.
sored a research program at the University
of Dayton Research Institute (UDRI) (3) for
the design, fabrication, and operation of
a laboratory-scale process to simulate
nonflame thermal decomposition environ-
ments. What has evolved has become known
as the thermal decomposition analytical
system (IDAS) (4).
CONCEPT AND BASIC DESIGN OF TDAS
The rationale behind the design con- '
cept of the TDAS has not changed basically
from that of the earlier discontinuous
system (5, 6). The sample is still insert-
ed into the system and then gradually va-
porized in a flowing carrier gas. The va- .,
porized compounds are subsequently subject-
ed to a controlled, high-temperature expo-
sure. The components of the vaporized mix-
ture that emerge from the high-temperature
environment are then collected and subject-
ed to instrumental chemical analysis. This
same thermal analysis format has been em-
ployed with respect to the TDAS, but each
operation within the system is much more
sophisticated, thereby producing greatly
increased experimental versatility.
The major design changes over the ear-
lier system are centered around the design
of the reactor, the closed continuous sys-
tem concept, and also the vastly increased
analytical capability that is now provided
by an in-line gas chromatograph/mass spec-
trometer/dedicated computer (GCMS-COMP).
Many design objectives were associated
with the development of the TDAS. This sys-
tem should be capable of conducting precise
thermal decomposition tests. More precise-
ly, it should be capsble of experimentally
determining the effects of the five pro-
minent thermal decomposition variables—ex-
posure temperature, gaseous atmosphere,
pressure, mean residence time, and resid-
ence time distribution. In addition, the
TDAS should be able to accommodate almost
any type of organic material. Also, it
should be .capable of analyzing all of the
thermal decomposition effluent products.
This closed continuous system should be
capable of dealing with toxic materials.
Also, the TDAS should be capable of gener-
ating data on a quick response basis.
„. , Figure 1 shoys a bloqk,.. diagram of the
TDAS and Figure 2 is an artist's conception
of the assembled TDAS components. Studies
can be conducted with the TDAS using almost
any compressed gas as the carrier and ther-
mal decomposition'atmosphere. Indeed, py-
rolytic studies can be performed using in-
ert gases, and oxidative studies can be con-
ducted using air or any other oxygen-con-
taining carrier. Accurate measurements of
pressure and flow can be readily obtained
with the TDAS. The internal pressure in
the reactor and mass flow rate (thus mean
residence time) can be continuously moni-
tored by in-stre.am instrumentation.
In the earlier work with the discon-
tinuous system, only low-volatility organic
samples could be tested. The TDAS has been
designed for measuring the thermal decom-
position properties of a wide range of or-
ganic samples—gases, liquids, solids, and
even polymers. Complicated organic mix-
tures can also be tested with the TDAS,by
using the process of slow vaporization of
the sample. After vaporization, the gas
phase molecules are subjected to precise
high-temperature conditions (these range
between 200° and 1150°C and are held with
j^2°C) in a fused quartz tube reactor. The
emerging products are then rapidly swept
into a cryogenic, absorbent in-line trap
where the condensable products are captured
at temperatures down to minus 110°. The
collected effluent products are subsequent-
ly thermally desorbed from the trap and
then subjected to GC analysis using high-
resolution glass, open tubular columns.
108
-------
HIGH TEMPERATUftf TRANSFER
'iV«" ' -'.'..IP ' V ' iw
!! I CAPTURE
OF
EFFLUENT
PRODUCTS
". • ' 1
1
W/(///J
mrnjjt
-\i. .••
-
CONTROLLED
HIGH
TEMPERATURE
EXPOSURE
• '••''•'.',
\
'(//'///i
' '
_ SAMPLE
2 INSERTION
-r AND
PRESSURE AND
FLOW REGULATION
. .,.. ., ..... . . .... .^
' IN-LINE .
'GAS CHROMATOGRAPH
(HIGH RESOLUTION)
'
^
COMPUTER
SYSTEM
COUPLED
MASS
SPECTROMETER
(MAGNETIC!
r :
NIH-EPA
' CHEMICAL
INFORMATION
SYSTEM
COMPRESSED GAS
AND PURIFICATION
1
_
ANALYSIS Of EFFLUENT PROQUCTS
FIGURE 1
109
-------
110
-------
The separated compounds are then subjected
to detailed analysis by mass spectrometry
(MS). The thermal decomposition test para-
meters and the chemical analysis data are
then retained, and a complete analysis of
the thermal decomposition process is carri-
ed out.
Although the TDAS is a complicated
system with many interesting components,
the operation of this system and the con-
ducting of thermal decomposition tests can
be accomplished with relative ease (i.e.,
after suitable familiarization and experi-
ence). Most of the crucial instrumentation
for conducting experiments with the TDAS
have been mounted in an instrumentation
console, and most of the test functions can
be continuously monitored on this console.
The basic procedure for testing a par-
ticular sample* begins with selection of
the thermal decomposition atmosphere or
carrier gas. Next the mode of sample in-
sertion must be selected. Then the actual
exposure temperature is chosen and each of
the furnace zones are set to this tempera-
ture. The internal pressure under which
the test is to be conducted can be esta-
blished by adjusting the gas flow restric-
tors situated downstream of the reactor.
The next parameter to, be established is the
mean residence time. . Once the other vari-
ables have been stabilized and measured, it
is a simple matter of calculating the de-
sired carrier flow and dialing off that
value with the adjustable flow control.
After establishing the thermal decom-
position test parameters, the next step in
the procedure is to cool down the effluent
collection trap so that the condensable
products can be captured in the absorptive
cryogenic trap. The remaining steps are
relatively routine. The sample must then
be admitted, the products collected, and
after switching to helium carrier gas, the
captured products must be subjected to in-
line analysis by GCMS-COMP.
The above test procedures would yield
one set of data points in the thermal de-
composition procedure. Any one of the
above variables could now be changed for
*A.ll samples subjected to analysis by
the TDAS would be screened beforehand using
other chromatpgraphic instruments to verify
that the sample would indeed be conducive
to a TDAS examination.
the next thermal decomposition test, and
so on, until the thermal decomposition pro-
perties of the sample are adequately char-
acterized.
UTILIZATION OF TDAS DERIVED DATA
In support of the vast cleanup at the
Morris Foreman Wastewater Treatment Plant
in Louisville, Kentucky, UDRI was requested
to carry out a TDAS study on "Hex" waste.
"Hex" waste being derived from the produc-
tion of chlorinated pesticides and the
waste being established as the cause of
contamination of the Louisville sewer sys-
tem and the cause for shutdown of the treat-
ment plant. The source of the "Hex" was
the illegal dumping into the system by
waste haulers. Suffice it to say an ex-
tremely serious environmental situation ex-
isted that required sound remedies to bring
the plant back into operation.
Results of the analysis of the basic
sample as received are shown in Figure 3.
It must be noted here that the analysis in-
dicated that there was at a minimum, 185
compounds present in this waste stream.
The principal components of the waste (prin-
cipal with regard to concentration) are
listed in Table 1, with the exception of
compound "e" which has been tentatively
identified and classed as a major product
of incomplete combustion.
A series of controlled TDAS experi-
ments was conducted to determine the ther-
mal stability of the waste and its behavior
upon thermal stressing. The results are
presented in Figure 4 and show how several
major constituents of the waste thermally
decompose. Special-note should be made of
Compound C which was observed to increase
in concentration as the waste was exposed to
progressively higher temperatures. The com-
pound was identified as hexchlorobenzene
(HCB) and it was revealed through the lit-
erature that HCB has a high temperature for-
mation process known, to the industry. HCB
was of concern from several standpoints in
this study in that it was a POHC, it was a
PIC, and .it is one of the most thermally
stable compounds investigated on the TDAS.
Figure 3 goes on to show that compound
"e" was formed during decomposition of the
starting material ("Hex" waste) and has the
tentative identification of hexchloroinden-
one assigned to it. Figure 5 is a log-skel-
etal chromatogram of the "Hex" waste identi-
111
-------
o
Di
oo
-^ o_ rr
oo
>*-»» N^» NfcJ -_ NC
o
ir\
O
ir
O
O
NO
o
NO
o
0_
^; o
o
S
cn
112
-------
©
GAS CHHQMITOG3APRZC COHDITIOHS
It
AT
Di
latruatnt Varian 1800
.iv.t ». Rubey
M October 2, 1979
Colusni
C
Bt
D.
-I..-T -• P*"x
_.e±an«rv oluma 5P-210O-51J,1COHB
film thieluMH 0.1
irri-r tlSL!" °"" 30
«lr 300
coltmn outl«e BUDDloaiat 20
t>
m
u»
aba «Bw
o3 ^-1
«3 .in-l
«3 ain-1
=.3 nla-1
output tiqnal e«eardingi
full seal* raid oat 1'° »»
S
T
elwrc advance r*t« 0.5
°*m=«,ti.» ••»«' "»«••
»olT«ac Tolvena
connneratian in lolvuit SO+%
ln-tanion samla *i» Q*4
•nllt c»tio i-l 'to 25
"**J*^C"' a«
initial 3 0
initial bold °
£iaal 230
pxcKrtm rat« ^ • "
d.c«ur 300
can ain'1
ul
°C
°c
•in
«c
°C Bin"1
°c
©
10
20
1
30
TIME, minutes
FIGURE 3-
40
50
113
-------
THERMAL DECOMPOSITION PROFILES OF 'HEX' WASTE
0 200 400 600 800 1000
EXPOSURE TEMPERATURE PC)
FIGURE 4
114
-------
fying the six (6) major POHCs. It can be
observed from Figure 5, as we proceed from
an exposure temperature of 300°C through
1000°C, the change in chromatographic char-
acteristics. There are several PICs that
become noticeable during the studies, par-
ticularly in the 500° to 700° range.
From the data generated on the IDAS,
it was possible to say that the temperature
and residence times that were available at
the Morris Foreman Plant incinerator would
be sufficient for the final disposal of the
contaminated sludge. During the actual
field disposal of the waste, the incinera-
tor effluent was monitored (7) with the
findings that: (a) the contaminated waste
was successfully destroyed and (b) the
principal chemical species found in the ef-
fluent was HCB.
OBSERVATIONS/CONCLUSION
The availability of the TDAS data
greatly simplified the work at Louisville
by showing that disposal of the contaminat-
ed material was feasible and by flagging
specific compounds for which the incinera-
tor exhaust gases should be monitored.
The overwhelming complexity of the "Hex"
waste indicated that chemical analysis of
the waste for identification of individual
components would be a serious analytical
challenge, be very time consuming, and pose
severe economic problems. Through the use
of the TDAS and exposing a sample to a ser-
ies of temperatures in a controlled labora-
tory environment, the entire complexity of
the analysis and understanding of the ther-
mal stability of the sample were reduced
and clarified.
From the published regulations con-
cerning hazardous waste incineration (1)>
the definition of a POHC lies in its abil-
ity to challenge an incinerator or put an-
other way, its degree of incinerability.
At this time, a specific ranking of POHCs
via an incinerability index is inconclusive
at .best. From the results of the "Hex" re-
search on the TDAS and other compounds in-
vestigated, it appears that the TDAS may
serve as the primary analytical tool in the
characterization of a hazardous waste for
its POHCs. Again, this was demonstrated
in the early log-skeletal chromatograms of
the "Hex" waste when on exposure to 300°C,
the original waste was reduced'to approxi-
mately 5.5 percent of its starting com-
pounds. These then become the primary
POHCs for this waste and we then proceed
with developing the decomposition profiles
while being acutely aware of PIC forma-
tion during this process. By the discrete
interpretation of these results, it becomes
possible to identify.POHCs in the starting
waste for monitoring the incinerators per-
formance and to identify PICs for setting
up the analytical requirements for stack
gases sampled during the test period.
In conclusion, it is acknowledged that
while the TDAS-type of thermal analysis will
not take the place of field situation test
burns in support of RCRA incineration per-
mits, it can and does provide an invaluable
insight into planning and evaluating the re-
sults of such a test burn. Additionally,
as the data base of TDAS information grows,
there is every expectation that EPA and in-
dustry can readily predict the conditions
for adequate destruction and/or problems
associated with the incineration of indus-
trial wastes.
REFERENCES
1. Environmental Protection Agency. In-
cinerator Standards for Owners and Op-
ertors of Hazardous Waste Management
Facilities, Interim Final Rule and Pro-
posal Rule. Federal Register, Vol. 46,
No. 15, Friday, January 23, 1981, pp.
7666-7690.
2. Environmental Protection Agency. Poly-
chlorinated Biphenyls (PCBs): Disposal '•
and Marking. Federal Register, Vol. 43,
No. 15, Friday, February 17, 1978, pp.
7150-7690.
3. Duvall, D. S., et al., "High-Tempera-
ture Degradation Characteristics of
Hazardous Organic Wastes - A Laboratory
Approach," Draft Final Report for USEPA
Research Grant R805117, January 1980.
4. Rubey, W. A., "Design Consideration for
a Thermal Decomposition Analytical Sys-
tem (TDAS)." EPA/600-2-80-098, August
1980.
5. Duvall, D. S., et al., "High Tempera-
ture Destruction of Kepone and Related
Pesticides." Presented at 173rd Ameri-
can Chemical Society National Meeting,
New Orleans, LA, March 1977.
6. Carnes, R. A., et al., "A Laboratory
Approach to Thermal Degradation of Or-
ganic Compounds. In: Proceedings of
70th Annual Meeting of the Air Pollu-
115
-------
tion Control Association, Toronto, On-
tario, June 1977.
116
-------
FIGURE 5
u 10,000
?
I 1,000
§ e 100
1s
2
£ £ i.o
i-
S 0.1
O.OI
-
-
•
-
.
-
LOG-SKELETAL
GAS CHROMATOGRAM
0 10,000 -
xr\ /r\ jc
1 I.OOO
©
-— U
. © if too
g
Zui 10
gg , Q
300°C | o.l
£
O.OI
^
f
©
©
i)
G) ©
700°C
RETENTION TIME -
TIME-
og
i§
IO.OOO
1,000
100
10
•i.o
0.1
O.OI
400°C
TIME-
10,000
1,000
if loo
>£
0.,
0.01
800°C
TIME-
u IO.OOO
| 1,000
§*E IOO
5S I0
O Q
gS ''°
01 Q-
ca **•
H 0.1
2
O Ol
-
-
-
-------
SITING AND DESIGN CONSIDERATION FOR THE ENVIRONMENTAL
PROTECTION AGENCY COMBUSTION RESEARCH FACILITY
Richard A. Carnes
Environmental Scientist
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45268
Frank C. Whitmore
Principal Scientist
Versar, Inc.
6621 Electronic Drive
Springfield, Virginia 22151
BACKGROUND
In 1975, the EPA published the results
of a major research program entitled "De-
termination of Incinerator Operating Con-
ditions Necessary for Safe Disposal of
Pesticides" (1). This report presented the
results of a number of incineration tests
covering several pesticide formulations and
molecular structures. At that time, it
represented the Agency's first attempt to
define safe operating conditions for the
thermal destruction of hazardous sub-
stances. Further, under the authority
granted the Agency by the Resource Conser-
vation and Recovery Act of 1976 (RCRA), the
Agency used this report along with other
documents (2) to support the establishment
of incineration conditions that are requir-
ed for the incineration of hazardous wastes
as listed in appropriate Federal Register
publications (3, 4).
During the regulatory development ex-
ercise, it became apparent that a consid-
erably greater amount of incineration data
would be required for the enormous number
of compounds and mixtures of compounds to
be regulated under RCRA. A major research
and development effort was determined to
be necessary to provide that data. This
program was planned to consist of two ma-
jor efforts: (a) a laboratory research
program to provide basic data on the ther-
mal stability of specific hazardous com-
pounds, which would also support the
second effort; (b) an extensive series of
full-scale test burns using existing equip-
ment and facilities.
It soon became apparent in other relat-
ed work that there were significant differ-
ences in data observed in the large-scale
experiments from those derived from the
laboratory studies. Such differences are
thought to arise from the simplifications
that have been made in the incineration
conditions maintained in the laboratory-
scale experiments.
A need was recognized for an interme-
diate-scale study that would more nearly
approximate the thermal and chemical con-
ditions that exist in full-scale technolo-
gy, but at the same time be close enough
to the laboratory studies so as to provide
a bridge between the two. In July of 1978,
a research contract was awarded to conduct
parametric investigations of a pilot-scale
hazardous waste incinerator. The program
was originally scheduled to rent the tech-
nology at the manufacturer's facility and
have contract personnel conduct the nec-
essary experiments on an intermittent
basis so as not to disrupt activities of
the manufacturer. Unfortunately, this pro-
cedure was found to be unsatisfactory for
a number of reasons including the inability
of the manufacturer to make the equipment
available: Eventually, the EPA authorized
118
-------
the contractor to purchase the incinerator
and move it to a suitable site at which the
experiments would be carried out.
The contractor found a suitable iso-
lated site that was zoned for industrial
research and development and that was with-
in commuting distance from his office and
laboratories. Permits were received from
the state. All agreements were drawn up
and modifications to the site initiated,
when local citizens became concerned and
effectively blocked the continuation of the
program. A very vital lesson in communi-
cation was learned and subsequently report-
ed on in the literature (6).
NATIONAL SITE SEARCH
In September 1979, after the two sit-
ing failures, the authors were charged by
the Agency to find an acceptable site, pre-
ferably on Government-owned land, for the
incinerator and to conduct the research at
that site. An analysis of the earlier sit-
ing failures strongly suggested that the
public would never accept the concept of
this type of research being carried out in
anything suggesting a makeshift facility.
What was required would be a fully dedicat-
ed facility specifically designed for in-
cineration research and for the handling of
hazardous materials. As an essential first
step in the search for suitable facilities,
it was required that the criteria for such
a laboratory facility be carefully outlined
and a general layout for the facility be
designed. The details of the resulting fa-
cility will be treated below, suffice it'to
say here that the essential criteria were
defined as follows: (a) the laboratory
must be staffed and equipped so as to be
capable of on-site analyses of both the in-
coming candidate waste streams and of all
effluent streams from the facility; (b) the
operations should be entirely professional
and the resulting data made available to
the interested public; (c) safety of the
operating personnel, the facility, and the
surrounding area must be the first consid-
eration in all operations; and (d) since
recently published data (7) indicate that
the products of incomplete combustion
(PICs) may, in many cases be significantly
more hazardous than the components of the
waste that require the latter to be classi-
fied as hazardous, the laboratory must be
of a quality to allow the safe handling of
toxic materials including carcinogens.
With these criteria in hand, an exhaustive
site search was undertaken. The approach
was, in every case, begun by a search of
available Government space (this was se-
lected as a criterion since thereby it
would be possible to avoid zoning problems)
followed by direct contact with the respon-
sible authorities at the available sites.
In most cases, very little enthusiasm was
exhibited so that no further action was
necessary. Eventually, the National Center
for Toxicological Research (NCTR) in Jeffer-
son, Arkansas, suggested that they could
and would make space available. The
authors visited the decision officials,
were well received, and shown that the site
was ideal for the proposed facility. Spe-
cifically, NCTR is located on the northern
boundary of the Pine Bluff Arsenal on a
good all-weather road in a location'that is
sparesely populated'and served by a commu-
nity that is well conditioned to understand
the dangers inherent in the handling of ex-
plosive and toxic materials (the Arsenal is
the national depot of chemical weapons for
the Army).
PREPARATION FOR PERMIT OF OPERATIONS
With the approval in principle by the
management of NCTR, preparations were begun
to obtain the approval of the state and
local officials that would be concerned and
to properly inform the public of the plann-
ed facility (8). Previous experience had
shown that the latter element of the infor-
mational program was at least as critical as
the former. The steps taken were the fol-
lowing:
(1) Confer with and inform the cogni-
zant USEPA Regional Administrator and staff
of the proposed facility and its mission.
(2) Confer with and inform the highest
state officials of the proposed facility
and its planned mission.
When these individuals were satisfied
with the general notion and indicated that
they would actively support the facility,
the next series of steps were taken. Spe-
cifically, these consisted of the follow-
ing:
(3) A series of meetings with the Ar-
kansas Department of Pollution Control and
Ecology (DPCE) were held covering the gener-
al concept of the facility, its mode of
operation, and the nature of the staff
were outlined. The meetings finally led to
119
-------
an application for a construction permit
and the assurance that, if there are no
public objections, the operating permits
would be forthcoming.
(4) A series of meetings with state
officials and agencies that would be im-
pacted by the facility. This included the
health, police, and transportation depart-
ments among others.
(5) A series of meetings with local
officials and with the members of the state
legislation who represented the area that
would be most immediately impacted by the
facility.
(6) Civic groups were offered the
opportunity for informal (or for that mat-
ter, formal) presentations of the planned
facility and its mandate.
(7) The state industrial organization
were considered important in this informa-
tional program since they were concerned
with hazardous waste disposal problems,
maintained important informational services
for the state officials, and could be ex-
pected to serve as sources of test mate-
rials in the active phase of the research
program.
The format of the individual informa-
tional meetings was, of course, tailored
for the specific group to be addressed.
There were, however, many features of each
of these meetings that were common. In
order to be assured that all parties were
aware of the magnitude of the hazardous
waste disposal problem in the United
States, a slide show using selected exam-
ples of the poor practices used in the past
and the consequences that all too often
have accompanied such practices was pre-
sented by Agency officials. After this in-
troduction, the detailed program proposed
for the CRF was discussed. In each meet-
ing, sufficient time was allotted for ques-
tions from the concerned audience. In some
cases, for example with the DPCE of Ar-
kansas, there were a number of meetings,
most of which were working sessions where-
in the specific requirements for permitting
the facility and its operation were dis-
cussed and finally incorporated into the
construction permit and into the protocol
for the operation of the facility.
During this extensive series of infor-
tional meetings and briefings, the media
were informed, in detail, by both press con-
ferences and detailed handouts that care-
fully and fully described the proposed pro"
gram. Further, the national congressional
delegations was kept informed by frequent
letters that described the promotional ac-
tivities that were underway.
The culmination of this activity was a
public meeting in Pine Bluff, Arkansas, at
which the public was afforded the opport-
unity to present their questions and pos-
sible concerns. The meeting was chaired by
the Director of DPCE supported by respon-
sible staff from the Agency and from the
EPA Contractor. The attendance at the
meeting was approximately equally divided
between media persons and the technical and
administrative officials there to defend
the program. The informational program
that had been conducted had apparently an-
swered questions and there was no public
concern expressed. Shortly after this
meeting, the State of Arkansas issued a
contruction permit for the facility.
THE COMBUSTION RESEARCH FACILITY (CRF)
In the course of the discussion between
the authors and the officials of the State
explicit, as well as, several implicit con-
ditions existed in the permit-conditions
that would have a significant impact on the
design, staffing, and operation of the CRF.
• Safety assumes such an important as-
pect in the overall program that a
trained specialist (Safety Officer)
will be an integral part of the per-
manent staff.
• The concern about carcinogenic by-
products of combustion required
strict adherence to NIOSH and OSHA
requirements for the laboratories at
CRF.
• Continuous monitoring of all efflu-
ents must occur during any incinera-
tor tests.
. The facility must be equipped to con-
duct the majority of analyses in-
house.
. To the extent possible, all candi-
date waste streams should originate
from within Arkansas.
120
-------
• To the extent possible, local per-
sonnel should be hired to staff
the operations.
. The senior staff of the facility
must be full-time on site.
When all requirements are brought to-
gether, the resulting facility takes the
form shown in Figures T through 3. The
latter figure presents an artist's concep-
tion of the interior of the facility. Some
of the outstanding features of the faci-
lity are:
• The incinerator(s) will be housed
in an isolated room accessible
only through the change room/show-
er. All controls and monitoring
readout equipment will be monitor-
ed from the operations room.
• Each laboratory will be provided
with a hood and separate outside
air source and will be designed to
. safely handle carcinogens.
• Internal connections between la-
boratories have been designed to
reduce the potential of cross-con-
tamination to an absolute minimum.
. The characterization laboratory
will allow the complete determina-
tion of the required combustion
conditions for a given waste and
to determine the nature of the air
pollution control equipment re-
quired for each waste.
. The analytical laboratory will be
fully equipped to allow analysis
of all necessary organic com-
pounds.
FUNCTIONS OF THE CRF
The original purpose of the research
program, which eventually lead to the con-
cept of the CRF, was to carry out a number
of pilot-scale test burns on materials that
had previously been studied at the labora-
tory Scale (9). The purpose of these ex-
periments was to better allow the extrapo-
lation of the laboratory data to large-
scale systems and to indicate differences
between the idealized laboratory system and
the realtor Id equipment. In the coursfe of
the development of that program and spurred
by the needs of the EPA, and the require-
ments stipulated by the State of Arkansas,
a number of additional elements of the pro-
gram to be carried out at the CRF have sur-
faced. Several Of the more important of
these will be briefly described below.
WASTE CHARACTERIZATION
It has been stipulated that, prior to
the undertaking of a test series on a parti-
cular waste, it will be necessary for the
staff of the CRF to produce a detailed pro-
tocol for the proposed tests. Only after
review and acceptance by both the DPCE and
the EPA Project Manager, will the actual
tests be undertaken. Clearly such a re-
quirement means that the necessary inciner-
ation parameters must be established prior
to the preparation of such a protocol. It
is the function of the Characterization
Laboratory (Figure 3) to produce these
tests. A literal reading of the Incinera-
tion Regulations (4) (1/23/81 - FR) would
suggest that this will require the detailed
analysis of the wastes for the identifica-
tion of the principal organic hazardous
compounds (POHCs), a determination of which
of, these POHCs will most seriously chal-
lenge the incinerator [and thereby be the
compound(s) for which the minimum operating
conditions must be determined], and a de-
termination of the air pollution control
equipment that is most appropriate.
An alternative procedure to this dif-
ficult requirement has been proposed and
will be a major element of the program at
the CRF. This alternative procedure will,
among other results, serve to identify both
the POHC(s) for which the incinerator ef-
fluents must be sampled and will also serve
to suggest the nature of the products of
incomplete combustion (PICs) that might be
involved.
DEVELOPMENT OF STANDARD PROTOCOLS FOR HIGH
TEMPERATURE SAMPLING
The nature, of the incinerator technolo-
gy that will initially be located at the CRF
is such that the primary chamber can be
operated at sufficiently low temperatures
so as to allow the determination of de-
struction efficiency (DE) as a function of
temperature while at the same time operat- .
ing the secondary chamber at sufficiently
high temperatures that assures there are no
hazardous effluents. This ability carries
the necessity to develop and test various
121
-------
FIGURE 1: RELATIVE POSITION OF PLANNED' CRF TO EXISTING
POLLUTION CONTROL FACILITIES AT NCTR
FIGURE 2: CLOSE-UP VIEW OF CRF
FIGURE 3: ARTIST RENDERING OF CRF CUTAWAY SHOWING
LABORATORIES AND.INCINERATOR ROOM
122
-------
hot-zone sampling equipment and procedures.
There are a number of practical situations
in which a certified and widely accepted
sampling method would be of significance.
ANALTYICAL LABORATORY AND PROCEDURES
As has been mentioned, one serious
stipulation on the operation of the CRF is
the requirement for on-site laboratories to
be equipped and staffed so that analyses of
any and all effluent streams from the sys-
tem can be conducted. For many of the an-
ticipated trace compounds there does not
exist well established and widely accepted
analytical methods. It will be part of the
function of this laboratory to assist in
the development of such methods where nec-
essary. Furthermore, many of the accepted
sampling and collection methods and media
have not been adequately verified. This,
too, will be part of the mission of this
laboratory.
A somewhat esoteric aspect of the re-
search program will be the determination of
the effects of sample storage on the valid-
ity of analytical data. In most large
technology te'st burns, there is a consider-
able time lapse between the taking of sam-
ples and the subsequent analysis of those
samples. There is some concern that during
this extended interval relatively short
lived species might well be altered either
in nature or concentration, or both. A
program will be undertaken to determine the
possibility of this effect and acceptable
methods of control.
INCINERATION INSTRUMENTATION
Presently available original equipment
manufacturer (OEM) supplied control instru-
mentation is not of sufficient range or
reliability to adequately meet the control
requirements of, for example, the PCB In-
cineration Regulations (3). It will be
necessary that additional instrumentation
be developed and tested under standard con-
ditions for the measurement of (on-line and
in real time) such parameters as hot zone,
mass flow, temperature (both average and
temperature fluctuation magnitude), resi-
dence time and its fluctuation, and, if
possible, the actual concentration of or-
ganic compounds in the gaseous effluent
from the incinerator. Each of these
measurement sensors must be such as to pre-
sent its output in a form that can be used
to control the waste blend feed of an in-
cinerator.
AIR POLLUTION CONTROL DEVICES (APCD)
At present, the major source of 'opera-
tional information on the many different
devices that are used for air pollution
control is the vendors of that equipment.
It would be of great utility if there ex-
isted a standard series of tests and test
conditions whereby each such device could
be treated for its performance and chal-
lenged as to its design and capacity. In
this manner there would exist data that
would then allow the intercomparison of
apparently comparable devices under a set
of standard conditions. Clearly this would
better allow the designer of a facility to
make proper judgements. The incinerator
at the CRF will be sufficiently well char-
acterized as to allow the design of and the
carrying out of such'tests.
CONCLUSIONS
One of the disadvantages of the use of
trial burns in existing equipment lies in.
the difficulty in the intercomparison of
such data. The present difficulties with
the available data base derive principally
from this program. It was for the specific
purpose of circumventing this inherent un-
certainty that the CRF was designed. With
the availability of a dedicated facility
and a full-time dedicated staff, the rate
of return of comparable and reliable data
on the thermal characteristics of a variety
of industrial waste streams will soon be
forthcoming.
BIBLIOGRAPHY
1. Ferguson, Thomas L., et a!., "Determina-
of Incinerator Operating Conditions Nec-
essary for Safe Disposal of Pesticides,"
EPA-600/2-75-041, December 1975.
2. "Destroying Chemical Wastes in Commer-
cial-Scale Incinerators," Phase II.
Final Report (SW-155c) under Contract
68-01-2966, 1975.
3. Polychlorinated Biphenyls: Disposal and
Marking. Federal Register 43, No. 34,
February 17, 1978, pp. 7150^7164.
4. Environmental Protection Agency Incin-
erator Standard for Owners and Opera-
tors of Hazardous Waste Management
123
-------
Facilities; Interim Final Rule and
Proposed Rule. Federal Register, Vol.
46, No. 15, Friday, January 23, 1981,
pp. 7666-7690.
5. Bell, B. A. and F. C. Whitmore, "The
Kepone Incineration Test Program,"
EPA-600/2-78-108, May 1978.
6. Whitmore, F. C. and R. A. Games,
"Windmills, Incinerator, and Siting,"
In Press, J. of Hazardous Materials,
1981.
7. Lustenhouwer, S. W. A., et al.,
"Chlorinated Dibenzo-p-dioxins and
Related Compounds in Incinerator Ef-
fluents." Chemosphere 9, 501-522
(1980).
8. Private communication. Maryevelyn W.
Soller, USEPA, Cincinnati, Ohio 45268.
9. Rubey, Wayne A., "Use of Advanced
Thermal Systems for Studying Decom-
position of Hazardous Organic Wastes,"
EPA Cooperative Agreement CR 807815.
124
-------
ENVIRONMENTAL AND PERFORMANCE ASSESSMENT
AT HAMILTON COUNTY HAZARDOUS WASTE INCINERATOR
Boyd T. Riley Jr.
RYCON, Incorporated
Cincinnati, Ohio 45229
John H. Trapp
Metropolitan Sewer District of Greater Cincinnati
Cincinnati, Ohio 45204
ABSTRACT
This paper offers a brief description of the facility and an overview of the task
areas being developed in order to enhance the value of the MSD complex for USEPA and
MSD purposes. Another paper to be presented later in this conference describes the re-
sults of emissions and combustion efficiency tests conducted in the early fall of 1980.
INTRODUCTION
The MSD serving Cincinnati, Ham-
ilton County and associated municipali-
ties determined a need for a solid and
liquid waste incinerator complex to
serve the area needs in the late 1960's.
Operations were initiated at the facility
in 1979.
Recently, Hamilton County and the
U.S. Environmental Protection Agency,
Incineration Branch, formalized a co-
operative agreement whereby the
USEPA could begin to participate in the
management of the incinerator complex
for the purpose of generating data to
support permit writing activities for
other hazardous waste incinerator com-
plexes. Management of the day-to-day
operations of the MSD complex remains
with MSD personnel; however, review of
past experiences with the facility, possi-
ble modifications and improvements to
the MSD complex for data generation
purposes are presently being planned and
will be implemented on a cooperative
basis in the near future.
The MSD complex was constructed
on the premises of the Millcreek Sewage
Treatment Plant, located in the Lower
Price Hill area of downtown Cincinnati.
The sewage treatment plant includes pri-
mary settling, and conventional activated
sludge treatment facilities. Sludge is
treated in sludge digesters to produce
methane which fuels a stationary electri-
cal generating plant of 4 mw capacity.
Digested sludge is disposed of by
multiple hearth incinerators.
DESCRIPTION OF THE SYSTEM
Wastes are most commonly deliver -
ed by tanker truck from a variety of
sources to the receiving station of the
incinerator complex. The waste is
pumped from the truck through pre-
strainers whose round hole openings may
be varied from 1/8 in. to 3/8 in. open-
ings according to information available
on the waste. Selection of the screen is
125
-------
dependent on the viscosity of the material
and the known solids content. (The
screens are being replaced by continuous
shredders to minimize unloading delays.)
The wastes are pumped by one or more
of four pumps to a holding tank for place -
ment in selected storage tanks in the
tank field.
A small drum unloading and dis-
charge area is available for the manage -
ment of drummed wastes, although the
delivery of drummed wastes is not en-
couraged by MSD. An elevated unload-
ing dock is provided for the removal of
drums from trucks and a drum tilt and
discharge mechanism into a sump from
which wastes are pumped to a holding
tank is provided.
From the unloading point, the
wastes may be delivered either to the
6, 000 gal holding tanks or directly to
selected storage tanks of 25, 000 gal
capacity. The tank farm includes four
6, 000 gal preliminary holding tanks and
ten 25, 000 gal storage tanks, four
12, 000 gal storage tanks and an addition-
al 25, 000 gal uninsulated tank for fuel
oil storage to service the facility.
The entire tank farm is housed on a
concrete apron with a concrete spillwall
and trench drain which is valved to pre-
vent spills but will allow the removal of
rainwater. Drainage from the tank
farm is directed to the inlet of the sew-
age treatment plant or returned to one
or more storage tanks. Each waste
Storage tank is insulated and was origin-
ally designed with internal steam heat-
ing coils of the suction type. Due to
corrosion problems associated with the
heating coils and caking of the waste on
the coils, which negated their heating
efficiency, the heat exchangers in the
waste storage tanks were removed.
When necessary, wastes are heated by
operating a 40 HP mixer on each tank
until suitable temperatures and viscosi-
ties are achieved. A differential
pressure transducer is mounted on the
base of each tank to provide continuous
digital weight read-outs at the control
panel. Tanks are also monitored with a
liquid level gauge located in the control
room. In general, the tanks in the tank
farm are clustered in groups of two or
three and each group is served by two or
more pumps which may be operated
separately or concurrently to remove or
to add wastes to each tank. Under .nor -
mal operating conditions, two of the ten
storage tanks are planned to be used for
storage of skimmings from clarifiers and
primary settling tanks at the municipal
sewage treatment plant.
The total diked containment capacity
is approximately 650, 000 gal. Total
tank farm capacity including holding tanks,
storage tanks, fuel oil storage, etc. is
approximately 460, 000 gal, leaving
approximately a one and one-half ft free-
board in the event of a catastrophic spill.
A complete foamite fire protection sys-
tem is installed throughout the tank farm.
In addition, each tank is overlaid w.ith
nitrogen to reduce the possibility of oxi-
dation, polymerization or spontaneous
combustion events while waste is stored,
loaded or unloaded.
During incinerator operation, the
wastes are removed from the storage
tanks and combined in one or more of
four batching tanks. Wastes are blended
according to chemical compatibility and
heat content and, thus, provide a relative-
ly uniform fuel value to the combustion
systems. A target of 7, 500 Btu's per
pound is used for blending schedules.
The tanks are flat bottom in design and
are not washed between different waste
charges. To date, no difficulties have
been encountered with reactions between
waste residues in the tanks and new
wastes.
The blended wastes are removed
from the batching tanks and delivered to
either of two combustion systems. One
is a rotary kiln combustion system with
the capability of burning liquid wastes in
addition to certain types of granular
solid materials. The second is a liquid
126
-------
TABLE 1
Volumetric
Equipment
Kiln
Cyclone
Combustion
chamber
capacity
gal/hr
1200
1200
N/A
Maximum heat °F
release
MMBtu/ hr
65
77
N/A
Maximum
temperature
2400
3100
2500
Minimum
residence
time /sec
0.6
0. 1
2.0
injection combustion chamber. Both of
these primary combustion chambers dis-
charge into a secondary combustion
chamber.
Table 1 describes the design capaci-
ties of the rotary kiln, the liquid injection
c.ombustor and the secondary combustion
chamber.
Liquid wastes are delivered by heat
traced, insulated pipelines from the
batching tanks to externally atomized
type nozzles in both the rotary kiln and
the liquid injection combustion chamber.
Atomization was originally provided at
these nozzles with steam; however, high
steam .demand in other operations of the
waste water treatment plant created a
need to temporarily use compressed air
as the atomizing agent. Air is being.
supplied by portable rotary air compres -
sors pending installation of new dedicated
boilers. In addition, compressed air is
used to operate all controls in both the
tank farm and combustion facilities.
A solids handling system available
at the plant has two distinct capabilities.
The first is an in-ground hopper for re-
ceiving grit by dump truck delivered by
the municipal sewage treatment plant.
It consists basically of three receiving
chambers of approximately 8 cu yd
capacity from which solids are removed
by screw conveyor and fed to a bucket
elevator. The bucket elevator delivers
the solids to a lateral conveyor which
delivers the material to the gravity feed
chute of the rotary kiln. This system
could be used for contaminated combusti-
ble, solids that are finely divided, e.g.,
PCB-contaminated soil.
Bulky solids may also be handled by
the system. A single truck discharge
hopper is provided to receive oversize
material which is conveyed by a steel
plate conveyor to a 150 HP Eidal shred-
der. The output from the shredder is
carried by belt conveyor to the bucket
elevator which lifts the material, to the
discharge point of the rotary kiln.
Neither solids handling system is capable
of accepting semi-solid material at the
present time. In addition, due to the
open handling system, solids contamin-
ated with extremely volatile or powdery
materials cannot be accepted because of
the danger of excessive fugitive emissions
from the system. The design could be
modified to accept these materials should
a waste generator require this type of
service.
The rotary kiln combustion cham.ber
is approximately 20 ft in length and 12 ft
outside diameter. Inside diameter and
length of the kiln are 10 ft and 20 ft. The
injection combustion chamber is approxi-
mately 6 ft in diameter and 11 ft in
length. The secondary combustion
chamber is vertically oriented with hot
gases entering the base and leaving the
top of the chamber. It is cylindrical,
approximately 13 ft 6 in. in diameter by
40 ft in height, and surmounted by an
emergency stack bypass for use when
127
-------
power or scrubber water is lost. The
three combustion chambers are con-
structed of carbon steel, ASTM A-283,
and lined with refractory and insulating
brick. Maximum operating tempera-
tures are 3100° F in the liquid injection
chamber, 2400° F in the kiln and 2500°F
in the combustion chamber exit. Ash and
slag are removed in a quench tank and
transported via pipeline to a holding
lagoon located near the incinerator.
The hot gas flows from the secondary
combustion chamber into a horizontal
duct where quench water sprays reduce
the temperature of the gas stream to
200° F. Thence, they pass into a verti-
cally oriented variable throat venturi
scrubber where recirculated water is
added to remove particulate matter. The
duct work and venturi scrubber are lined
with silicon carbide block, while the
shell is constructed of Corten (ASTM A-
242). The gas stream turns 90° and
enters the base of an acid neutralization
scrubber. This absorber is vertically
oriented, approximately 15 ft in diameter
and 30 ft tall, and is of the sieve tray
type. Its function is to neutralize acid
gases. The scrubber shell is lined with
glass flake polyester to minimize corro-
sion. The trays are 316 stainless steel.
Absorber water is supplied from the
principal blowdown tank after pH adjust-
ment with 50% caustic (sodium hydroxide)
from two standby 10, 000 gal storage
tanks. The cooled, neutralized gas
flow exits the top of the scrubber, passes
through a demister and is conducted ver-
tically down to an induced draft fan
powered by a 1500 HP motor. The fan
housing is lined with rubber and the fan
blades are constructed of Inconel 625.
Gases exiting the induced draft blower
pass into a 7 ft diameter stack approxi-
mately 90 ft in height, incorporating
sound suppression baffles and construc-
ted of cylindrical sections of 3/16 in.
stainless steel. Three pumps are used
to supply water to the quench system and
to the venturi scrubber. Two more are
used to recirculate flow from the blow-
down tank to the neutralization scrubber.
An emergency water supply mounted on a
vertical tower approximately 120 ft in
height is designed to supply quench water
to the system should pumping capabilities
or power be lost to the system. Quench
water is supplied for 5 min of operation
which is sufficient to allow shutdown of
the system and opening of the emergency
cap.
After the complete scrubbing system,
gas exits at approximately 180° F; emis-
sions are designed to be a maximum of
. 03 grains per standard cubic foot at
70° F and 12% CO2. Emission tests
have indicated that acid emissions have
been completely neutralized, <2 ppm HC1,
• but that particulate emissions, while
within state limits, are about .08 grains
per DSCF. Better quenching and im,-
proved demisting should allow achieve-
ment of the original design goal.
Process control and operation is
carried out in a centrally located building
housing the main control panel. The
control panel offers complete monitoring
of both the tank farm and the combustion
facilities. The tanks are monitored for
both valve position, i.e., in-flow, out-
flow, non-operation, liquid level, and
weight of tanks'. Pump motors are indi-
cated to be in an on-off position and
valve openings are displayed. Digital
read-outs are provided for waste temper-
atures in each tank. On the combustion
system, temperatures are monitored both
by drum type direct reading gauge and
continuous recorder as well as digital
temperature read-out. Temperature
monitoring points are at the exits of the
rotary kiln and liquid combustion cham-
ber and at the exit of the secondary com-
bustion chamber. Overall system con-
trol is based on the Btu content of the
waste to be burned. Waste flow is auto-
matically varied to maintain set point
temperatures and negative draft in the
combustion chamber (about 1-1/2 in.
water), venturi pressure drop (28 - 30
in. water), and pH of recirculated water
128
-------
exiting the absorber (6.5 - 7.0) are conr
trolled.
A broad variety of liquid wastes can
be burned in the incinerator system.
Reasonable quantities of bulky solids may
be burned after shredding and granular
wastes can be incinerated. A general
category of wastes which are currently
excluded from the incinerator is semi-
solid material that is too viscous to be
handled as a pumpable fluid but too
sticky to be handled in a dry solids hand-
ling system, such as the bucket elevators
and conveyor represent. Should a
demand for this type of service be
sufficiently great, equipment could be
added to the system which would allow
its combustion.
FUTURE TASK AREAS
The purpose of the cooperative
venture between USEPA and MSD is to
allow EPA to develop in a cooperative
fashion a full scale test and evaluation
facility for evaluating the destructive
efficacy of the facility on various types
of hazardous wastes. The evaluation
will include: technical modifications re -
quired to increase the facility's produc-
tion or decrease costs; monitoring re-
quirements to more completely measure
emissions; and steps necessary to in-
crease the variety of hazardous wastes
which can be and are burned at the
facility. Throughout all operations,
data will be collected and analyzed to
support the permitting of other full scale
incineration facilities throughout the
United States. Thus, where questions of
compatibility of/an incinerator with a
specific hazardous material cannot be
resolved by other data or by the litera-
ture, it'will be possible for EPA to con-
duct a preliminary full scale, controlled
trial burn on the material in question.
A number of tasks are envisioned to
contribute to the modification of facility
operations in order to accommodate the
EPA requirements for a full scale
hazardous waste incinerator test facility.
Among those tasks presently being devel-
oped are the following:
1. Substantial additional effort will be
dedicated to the detailed monitoring
of on-going incineration operations
at the facility. Included will be a
definition and description of all
problem areas that are encountered
by the facility in order that this in-
formation may be passed along to
incinerator design personnel.- Many
areas will be examined as unit oper -
ations. This will include personnel
assignments, overall management,
laboratory operations for monitoring
and control of wastes,' receiving
area operations, storage operations
and compatibility control require-
ments, staging of wastes for com-
bustion, incinerator control oper-
ations, and the emission control re-
quirements . In short, a detailed
description of operations at the
facility will-be created to be used as
a reference manual for future
hazardous waste incinerator oper-
ations .
2. A detailed review and documentation
of the original plant design vs. the
actual plant as it exists will be made.
Included in the document will be
definitions and reasons for the
changes that have been brought about
and a description of the types of
problems which were overcome by
the changes.
3. A quarterly operating cost summary
for the facility will be generated in
order to indicate, the level of cost of
operation. Sufficient detail will be
developed in the costs in order, to
show those costs that are accrued
while the facility is not in operation,
as well as those that are accrued
during operation. Modification of
the facility costs will also be de-
fined.
4. One area of substantial interest to
129
-------
both EPA and MSD is the creation of
a system which will allow the collec-
tion of extensive data on various
hazardous waste incineration tests
at the lowest possible cost. As a
result, a number of alternative
approaches to long term sampling
and analysis problems of hazardous
waste incinerator facilities will be
examined. Principal among these
will be the preliminary consider -
ations for creating sampling slip-
streams from various points in the
facility coupled with automatic mon-
itoring instrumentation in an instru-
ment bank and computerized collec-
tion and reduction and presentation
of the data as it is generated. Such
an advanced system would, of course,
allow for the accumulation of large
quantities of data at very low cost
after the initial installation costs of
the system.
5. Substantial effort will be dedicated
to the classification of the principal
organic hazardous constituent
groups (POHC's) that have been
burned at the facility since its doors
opened several years ago. This
data collection effort will create a
backlog of information concerning
those hazardous materials that have
been successfully burned at this
type of facility.
6. Pursuant to the information devel-
oped in (5), a definition of POHC's
that may be burned at the facility
will also be created. Along with
these POHC's, appropriate sources
of typical wastes will be compiled.
Constraints on the incinerator
facility for accepting and burning
these materials will be defined.
This would include permitting re -
quirements, time required for per-
mits, and the possible quantities
and locations of waste that could be
accumulated for combustion at the
facility.
7. In order to carry out experiments
during the period of transition from
present operations at the facility to
those that might be exemplified by
improved data collection systems in
the future, an experimental data
collection and operating protocol for
the facility will be prepared in order
that data may be collected on
POHC's as soon as possible that may
be of immediate interest to EPA
permit writers. This protocol
would include a detailed definition of
operations by the various personnel
at the facility, as well as data which
should be collected and how it should
be collected.
The preceding tasks summarize .a
number of areas of significant interest to
EPA and MSD. A number of other tasks
are also being considered for inclusion
in the overall scope of the project, either
in this initial phase or in future phases.
Areas that may be of interest to others
in the hazardous waste incineration field
would be welcomed as possible suggested
tasks for inclusion in the cooperative
agreement. These should be submitted
to Mr. John.. Trapp of MSD of Greater
Cincinnati or Mr. Donald Oberacker of
the USEPA as soon as possible.
At the conclusion of the project, EPA
expects to have a full scale test facility
representing the most cost effective
approach to the gathering of data under
carefully controlled and monitored con-
ditions for a variety of POHC's of inter-
est to the Agency. This will be done
without altering or interfering with the
normal operation of the facility which
will continue to earn its own capital re -
payment and operating expenses through
the services rendered.
130
-------
TRIAL BURN VERIFICATION PROGRAM
FOR HAZARDOUS WASTE INCINERATION
K. P. Ananth, P. Gorman, E. Hansen
Midwest Research Institute
Kansas City, Missouri 64110
D. A. Oberacker
Incineration Research Branch, U.S. EPA
Cincinnati, Ohio 45268
ABSTRACT
The trial burn protocol described ift the EPA Guidance Manual for Evaluating Permit Ap-
plications for the Operation of Incinerator Units has been followed in a case study of
the Cincinnati Metropolitan Sewer District's (MSD) incineration facility. This paper
summarizes trial burn protocol requirements and presents the results of the protocol
verification tests carried out at the MSD incineration research facility.
INTRODUCTION
Hazardous waste incineration regula-
tions, as published in the Federal Regis-
ter, January 23, 1981, include stipula-
tions that will require "trial burns" in
incinerators. Therefore, EPA has pre-
pared a Draft Guidance Manual that pre-
sents a draft protocol for conducting the
trial burns (1). Midwest Research Insti-
tute (MRI) has a contract with EPA
through Rockwell International to verify
the protocol described in the guidance
manual, on a case study basis, at the
Cincinnati MSD incineration facility.
This paper describes the protocol re-
quirements, the incineration facility,
and the sampling and analysis that have
been carried out to verify the protocol
and to determine the performance of the
incinerator.
SUMMARY OF TRIAL BURN PROTOCOL REQUIRE-
MENTS
This section summarizes the salient
aspects of the trial burn requirements as
we understand and interpret them. A more
detailed description of the trial burn,
including waste analysis and evaluation
of incinerator performance, can be found
in Reference (1).
The purpose of a trial burn is to
(a) demonstrate incinerator operating
conditions such that a "new" waste can be
burned within the regulatory performance
standards and (b) demonstrate that a new
incinerator can operate similarly. "New"
wastes are defined as those hazardous
wastes with compositions different from
hazardous wastes specified in the facil-
ity permit. Either a trial burn plan
must be submitted with each request to
incinerate a new waste, or data showing
that the waste has been previously incin-
erated in a similar incinerator in an en-
vironmentally safe manner must be sub-
mitted. If a trial burn is required by
the permitting agency, the following ele-
ments must be contained in the trial burn
plan.
Waste Analysis
The analysis parameters in this cat-
egory are heating value, viscosity, con-
centrations of all hazardous constituents
present in the waste, PCB concentrations
above 50 ppm, chlorine content, hazardous
metals content, kinematic viscosity (for
131
-------
liquids), percent solids (insolubles),
ash, content, flash, point, elemental anal-
ysis, water content, and thermogravimet-
ric analysis. The rationale for selec-
tion of these parameters and the waste
sampling and analysis methods are desig-
nated in Reference (1). One of the major
objectives of the waste analysis is to
select principal organic hazardous con-
stituents (POHCs) for establishing the
effectiveness of the incinerator based on
the destruction and removal efficiency
(DRE) of the selected POHCs. The selec-
tion of POHCs is dependent upon the con-
centration of the POHC in the waste and
its heat of combustion, as presented by
EPA.
If the hazardous wastes are blended
prior to incineration, all the above
analyses should be performed on the
blend. Also, if an operator wishes to
vary the chlorine content, the waste
analysis should be reported for each of
the various chlorine-containing wastes.
If the auxiliary fuel is not a conven-
tional fuel, then the feed to the incin-
erator should be considered a blended
waste.
Descriptions of Incinerator Components
This category includes type of in-
cinerator, manufacturer's name and model
number of major components, dimensions of
major incinerator components including
combustion chamber, description of auxil-
iary fuel system, nozzle and burner de-
sign, capacities of prime movers and man-
ufacturer's curves, stack gas monitoring
and pollution control monitoring systems,
locations and descriptions for tempera-
ture, pressure and flow sensors, and con-
struction materials.
Provisions for Sampling/Analysis and Mon-
itoring of Incineration Process
Procedures and locations for moni-
toring combustion temperature, waste and
fuel feed rates, air flow rates, CO in
stack gas, and excess air in stack gas
should be described. In addition, equip-
ment and procedures must be specified for
determining POHCs in the waste, including
sampling and analysis of ash, scrubber
effluent, particulate emissions, stack
gas flow rate and temperature, POHC
concentrations in the waste and stack
gas, and HC1 concentration in the stack
gas.
Trial Burn Operating Conditions and
Schedule
Factors included here are combustion
zone temperature, waste feed rate, air
feed rate, auxiliary fuel feed rate, op-
erating conditions for pollution control
devices, and procedures for stopping feed
and safely shutting down the incinerator
and controlling emissions in case of an
emergency. Aspects connected with the
schedule are dates of trial burn, dura-
tion of trial burn, and quantity of waste
to be burned.
CINCINNATI MSD INCINERATION FACILITY
Only a brief summary of the inciner-
ation facility is provided in this paper,
since there will be another paper com-
pletely dedicated to discussing the in-
cineration facility, its various compo-
nents, and their design features.
The incineration facility at
Cincinnati is owned and operated by the
City of Cincinnati's MSD. It is a full-
scale, modern installation with a liquid
injection cyclone furnace and a rotary
kiln capable of handling both liquid and
solid waste. The system can handle about
151,413 £/day .(40,000 gal/day) of waste;
the cyclone furnace handles waste at the
rate of 75.7 £pm (20 gpm) and the kiln
about 37.8 Apm (10 gpm). The Btu ratings
on the cyclone furnace and the kiln are
65 x 106 kJ/h (62 x 106 Btu/h) and 55 x
106 kJ/h (52 x 106 Btu/h), respectively.
The cyclone furnace and the kiln are
connected to a common secondary combus-
tion chamber (combustor) which has a vol-
ume of 162 m3 (5,726 ft3). Combustion
gases leaving the combustor are quenched
before they enter a venturi scrubber
which has an operating pressure drop of
76.2 cm (30 in.) water. From the scrub-
ber, the gases go to a caustic absorber
and then exit through the stack. The ash
from the combustor and the rotary kiln
drop into a four-compartment ash tank
which is sluiced to a lagoon once every
8-h shift.
132
-------
SAMPLING AND ANALYSIS
In keeping with the requirements of
a trial burn, the overall objective of
the sampling and analysis program was to
characterize the waste burned, select the
POHCs, determine DSEs for the selected
POHCs for various incinerator operating
conditions, establish particulate grain
loadings, and determine chloride removal
efficiencies. Present RCRA regulations
require 99.99% efficiency for POHC de-
struction and removal, 99% hydrogen chlo-
ride removal efficiency, and a particu-
late grain loading of 0.183 g/dscm (0.08
gr/dscf) corrected to 12% C02.
A schematic illustration of the
Cincinnati MSB incineration facility
showing the various sampling points is
presented in Figure 1. As shown in the
figure, liquid waste which was fired in
the cyclone furnace as well as the rotary
kiln was sampled. In addition, one sam-
ple of the fuel oil was taken to deter-
mine whether any of the POHCs selected in
the waste were also present in the fuel
oil. Using the same rationale, quench
water samples were also taken for POHC
analysis. These two streams (fuel oil
and quench water) are not directly re-
quired by the trial burn protocol; but to
establish input of POHCs, one needs to
sample and analyze these streams for the
selected POHCs. Table 1 shows the test
matrix used in the trial burn at the MSB
facility and is complementary to Figure 1
with respect to sampling points, number
of samples taken per test day and the re-
sulting number of samples, and the analy-
ses carried out on the samples.
A few additional points should be
made with regard to Table 1 which bring
out the ambiguities in the protocol. The
first one deals with the duration of the
trial burn. It is not clear from Refer-
ence (1) what the duration should be.
For purposes of this program, we chose a
6-day duration for the trial burn based
on three operating temperatures and two
residence times. The operating tempera-
tures and residence times were selected
to provide a good range to evaluate DRE
as a function of both parameters. Also,
an 8-h test day was chosen. Secondly,
the EPA guidance document does not pre-
scribe stack sampling times. Since
sampling time is related to the concen-
tration of POHC in the waste, the ex-
pected concentration of the same POHC in
the stack, and the detectability of the
analysis procedure, we thought it prudent
to conduct stack sampling for both a 2-h
period and a 6-h period. The 2-h sam-
pling time corresponds to the EPA
Method 5 requirement, and the 6-h period
was chosen to enable collection of a suf-
ficient quantity of POHC if it was pres-
ent at the 100-ppm level in the waste and
was destroyed at the 99.99%-level. These
two sampling times would provide a better
perspective for selection of sampling
durations in future trial burns.
Waste sample collection times or in-
tervals are also not well prescribed.
Since industrial waste is often nonhomo-
geneous, one must ensure that the waste
characterized is indeed the waste that is
burned throughout the trial burn. To ad-
dress this aspect we took, as shown in
Table 1, grab samples of waste as well as
integrated samples over the 2-h period
and the 6-h period corresponding to the
2-h and 6-h stack sampling times.
The sampling train used during the
trial burn is shown in Figure 2. This
train is similar to that used for captur-
ing organic emissions from various emis-
sion sources. The XAD resin and the two
impingers downstream of the resin enable
capture of POHCs, and the impinger con-
taining silver-catalyzed solution is for
capture of volatile metals, viz As, Se,
and Hg. Due to the high moisture content
in the stack, we used a 4-liter bottle
downstream of the XAD resin instead of
the regular impinger, shown in Figure 2,
for a second set of tests.
Many of the analysis procedures for
POHCs and chlorides had to be developed
as part of this program, since methods in
Reference (1) were primarily based on
analysis of organics in wastewater and
could not be used here. In many of the
samples we encountered two separate or-
ganic phases and a semisolid phase.
In all cases, the four components of
each train (probe rinse, filter, XAD, and
condensate) were extracted and then com-
bined for POHC analysis by GC/MS so that
DRE could be calculated for each POHC.
133
-------
Quench
Water
Auxiliary
Fuel Oil
Liquid Waste
Feed
To Stack
Slowdown
Ash
Tank
Sluice
Gate
©
To Lagoon
V2CUU
Sampling Points
SIA, S1B - Liquid Waste Feed
SBA. S2B = Auxiliary Fuel Oil
S3 = Ash Sluicing
S4 — Scrubber Slowdown
S5 == Quench Water
S6 = Stack
Figure 1. Schematic illustration of incineration facility with sampling points.
134
-------
K
0
<,
W
fj*
i_ j
CJ
f— t
EH
gi
M
§
*-J
EH
^J
fe
f$
S
PQ
53
S
EH
1
X
1
3
13
r-t
a
S
CO
•H
CO
7—4
CO
M
0> CO
O CO
cO 4->
• & CO
O ' — .4)
/— , !ZI CO P
J-H 41
0 r-l rH X
O CO ft 'rt
O -P a co
>
M
CJ
O «M
CO CO
en 01
0 -P
to
*. at
5 POHCs to be selected)
mperatures— 899°C (1650°F)
s
days
No. of sampli
v-^ CU 41
•P S -P
41 -H CO
4-1 t)C -P 4)
co d P
CO -H 41
S -P O X -P
co d -H d
T3 H 01 CO -H
•H 4) T3 O
3 ft'H
I-H
•>!> 4-1
co Eb i— (
8-S
PMI O
rH H
in cj ft
rH
+ ^o
CM
1
co P*t
0) 01 CO
rH i-H T>
I1* J3
CO CO U
CO CO CO
41
13 T3
4) 0) Ol
4-1 4J 0
CO CO d
H M O
60 60
01 01 pD
4-> 4J «
d d ^
•H -H 60
f f S
CM vO X!
O 41 O
» d
EH O rH
«
M
CO 01
~-s 4-1
< CO
I-H CO
co 3:
1 1 1
1— 1 1— 1 1— t 1— 1
CJ CJ 0 CJ
CO CO CO U
1111
m m m m
\o *• O O
!-l CO fnl •• "PM
OH v— ' in o o cj m
-04-141
4J rd d 1— I
>> UCO >> COXt C060ft
iw CD co 'w d i d a
•H CO -H -H VO CM-H CO
tn CO M CO O !H to
0) X!ft 01 S-ICOCOO
> 0& > 4J3dM4-lCO
CO 3 O -H O -H CO
o oits omoi comdM
iv 4J d !-( O
WP t3ctt4->coS'0
>> od >> OP d11
r-l 4-ICO 1-1 Xll-H4l3COp
d r-l d +J3-OC3C8
•H 01 ft 'H 01 S -H O W
CO d ca S-HM6H36p
8 Od S cOO O
!H HJi! M M ^v-^ CMCOCJrH
60 60 d 60 60
rH -
-------
(0
c O J2
O ,_ p
4J
CM
0)
•H
136
-------
To ensure that volatile POHCs were ac-
counted for, we used integrated gas bag
samples to collect stack gases for analy-
sis, in addition to the sampling train
itself. In a similar vein, additional
analyses were carried out for the samples
collected by the chloride sampling trains
used at the stack. This was done because
the sampling/analysis method suggested in
Reference (1) measures total chlorides,
whereas the regulation specifies 99% HC1
removal.
In spite of the above difficulties,
we successfully completed trial burns
on two liquid wastes. Some of the avail-
able results of the first trial burn are
presented below; other results will be
discussed during the presentation of this
paper,
Waste Analysis and Selection of POHC
Analysis results of the six 6-h in-
tegrated waste samples for all potential
POHCs (both volatile and semivolatile)
are shown in Table 2, in the order of
their heat of combustion. This list is
limited to those compounds present at 100
ppm or greater, since, as noted in Refer-
ence (1), detection of compounds present
at < 100 ppm would present detection dif-
ficulties in the stack samples at a DRE
of 99.99%.
From the list of POHCs in Table 2,
the first four were selected for analysis
in all samples. In addition, we recom-
mended that hexachlorobenzene be included
as one of the selected POHCs because of
EPA's interest in this compound as a sur-
rogate for PCS. We also recommended that
hexachlorocyclopentadiene (HCCP) be in-
cluded as another of the selected POHCs
because it is present at relatively high
concentrations in the waste (14,000 to
39,000 ppm). It is a pesticide of par-
ticular interest to the Cincinnati MSB
and EPA. In summary, the six POHCs se-
lected for analysis in all samples are
carbon tetrachloride, hexachloroethane,
chloroform, tetrachloroethene, hexachlo-
robenzene, and hexachlorocyclopentadiene.
Results of the metals analysis on
the six 6-h integrated waste samples are
shown in Table 3. The analysis data are
presented for the liquid phases as well
as. the solid phase of each sample since
the waste samples were biphasic and non-
homogeneous. The metals listed in
Table 3 are those required by the proto-
col.
Destruction/Removal Efficiency (DRE) of
Selected POHCs
Determination of DRE is of utmost
importance in a trial burn. For the six
POHCs selected, analytical results are
available for both 6-h trains used in
Runs 1, 2, and 3. Using these stack re-
sults and the results of the waste analy-
sis, DREs have been computed for the six
POHCs of interest. These DRE results are
shown in Table 4.
These results show that the required
99.99% DRE was achieved for five of the
six POHCs. The DRE for tetrachloroethyl-
ene was slightly below 99.99% in most of
the tests. If one "rounds off" the re-
sults for this compound, it can be seen
that in each run one of the simultaneous
6-h trains would show 99.99% while the
other would show somewhat less (99.97 to
99.98%). Thus, further analyses of data
from, other runs will be needed to assess
the precision, of the sampling/analysis
results,
Review of the data given in Table 4
indicates, somewhat surprisingly, that
the DRE for each of the POHCs was unaf-
fected by the combustion temperatures
chosen for the tests (Run 1 - 899°C, Run
2 - 1093°C, and Run 3 - ^L3i60C). This,
in effect, means that 899^Cis) adequate to
obtain high DREs for the POHCs selected.
More data on DRE will be reported at
the meeting, after review of all analysis
data on the 2-h and 6-h trains.
Chloride Removal Efficiency
Chloride removal efficiency was
based on analysis for chlorine in the
waste . feed and two HC1 sampling trains
operated on each test day. Results are
given in Table 5.
The results show that there was only
one test (2A) in which the required re-
moval efficiency of 99% was achieved.
However, as we have discussed, the
137
-------
TABLE 2. ANALYSIS RESULTS AND RANKINGS FOR POTENTIAL POHCs
(> 100 (Jg/g) IN 6-h INTEGRATED WASTE FEED SAMPLES
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
He*
0.24
0.46
0.75
1.19
1.70
1.74
1.79
1.99
2.05
2.10
2.12
3.00
3.38
3.40
3.75
4.51
4.57
6.60
6.92
7.34
7.78
8.07
8.07
8.29
8.42
8.51
9.62
10.03
100/Hc*
416.7
117.4
133.3
84.0
58,8
57.5
55.9
50.3
48.8
47.6
47.2
33.3
29.6
29.4
26.7
22.2
21.9
15.2
14.5
13.6
12.9
12.4
12.4
12.1
11.9
11.8
10.4
10.0
Compound
Carbon tetrachloride
Hexachloroethane
Chloroform
Tetrachloroetheae
Methylene chloride
Trichloroethylene
Hexachlorobenzene
Trichloroethane
Penta chlo r ob enz ene
Hexachlorocyclopeatadiene
Hexachlorobutadiene
Dichloroethane
Isodrin
Trichlorobenzene
Aldrin
DDT
Dichlorobenzene
Chlorobenzene
Chlo ro toluene
Di-n-butylphthalate
Phenol
Toluene
Methyl ethyl keton
Butylbenzylphthalate
Bis (2-ethylhexyl)phthalate
Dimethy Ipheno 1
Naphthalene
Benzene
Concentration range
(Hg/g)
1,600-9,800
310-650
2,000-12,000
1,300-1,800
1,000-7,300
500-680
120-260
1,000-13,000
< 100-110
14,000-39,000
450-1,200
1,000-5,100
2,000-2,700
< 100-110
180-300
130-280
1,100-5,100
1,600-4,600
1,200-1,700
100-180
100-3,800
3,100-5,100
1,600-3,700
< 100-150
100-2,300
150-560
680-1,600
200-1,000
Class*
V or SV
V
SV
V
V
V
V
SV
V
SV
SV
SV
V
SV
SV
SV
SV
SV
V
SV
SV
SV
V
V
SV
SV
SV
SV
V
* HC = Heat of combustion in kcal/g.
/ Concentration range shown is the lowest and highest concentration found in the
six 6-h integrated waste feed samples.
£ Class designates whether the compound is a volatile (V) or semivolatile (SV).
This designation reflects the analysis techniques for the waste feed samples and
the fact that, subsequently, volatiles will most likely be detected in the inte-
grated gas bag samples while semivolatiles will most likely be detected in the
modified Method 5 train samples.
138
-------
*
""so
00
<— '
CO
1
CO
Q
Ctl
w
PH
w
CO
«5j
3»
J25
p
CO
M
CO
I—I
'^2
^ar
^M
CO
<
e-n
53
fa
0 _
H
. T
Q
CO
S
CO
1
r--OrHOOCMOOOOin
• CM C3 ^3 in G3 rH v£j ^D f^^ CM CM
rH CM • CT\ rH I— • rH
*v <^ *1 •*
t-H rH rH
o^^o^rS^^OrH0^1^
• VOOOVO rHLOVO COCN
0 0
V O
V
CMinOrHOOCOOOOOCO
• • rH <^ \^ CO * rH C5 f^* CO •
rHinCO- -* o* ^3* ^o c^
^^ • • ^3 rH O3 • • O^ i~H • •
• V^OO rHt^CO COCNI
0 0
v o
V
moocMooinoooiom
• CO O • ^O O rH *J; £5 f»*. ^ fsj
CM CO O ^f rH O CM
r\ *\ 4*
i-H V CM ocovooou">o
•COO •mOrHOOOCOCOCO
rH SO O rH OS i-H
•t M *»
rH CM r~
CMOr-CMrHVl-CMi-HOCO-*!— I
O • -OrHOO • •LOrH • •
•vfiOO rHOOin -* O •-* V CO
•* o •* •*
CO CM rH
V i-H
• • • O CM \O • rH CO CO • •
OCOOO O VO u">rH
o
o . •
V
comooooaFQ
d
_ft
5^
rH
d
03
,£j
4_)
01
tn
3
•o o*
•rl -H
i-H rH
o
W 0)
^d
0
o o
•rl
« 1?
M rH
*
139:
-------
ON ON O ON CO IO
ON ON ON ON ON ON
ON ON ON ON ON ON
ON ON ON ON ON ON
os ON ON ON ON ON
ON ON ON ON o> ON
ON ON ON ON ON ON
ON ON ON ON ON ON
A A A
a
A A
oo ON ON ON
ON ON ON ON
Jz; /z; ON ON ON ON
ON ON ON ON
sj-
I
a
^f ON o\ ON vo o)
ON ON oo ON ON ON
ON ON ON ON ON ON
ON ON ON ON ON ON
ON ON ON ON ON ON
•^ ON
-------
TABLE 5. SUMMARY OF CHLORIDE REMOVAL EFFICIENCY DATA
Run No.
1
1
2
2
3
3
4
4
5
5
6
6
HC1 test No.
1A
IB
2A
2B
3A
3B
4A
4B
5A
5B
6A
6B
Cl in waste
feed (%)
2.96
5.76
9.23
10.09
2.97
6.92
6.24
5.76
9.60
8.86
11.16
11,16
Chloride removal
efficiency (%)
Total
95.4
97.8
99.0
98.9
96.2
98.4
97.8
86.1
95.9
94.8
97.9
32.0
Excluding plug
97.7
99.7
99.8
99.7
99.4
99.8
99.5
98.2
98.6
99.0
99.8
80.0
TABLE 6. SUMMARY OF PARTICULATE DETERMINATIONS
Particulate concentration
Run No.
2
5
Train No.
M5(2)-2B
M5(2)-5B
g/dscm (gr/dscf)
0.2124 (0.0928)
0.0819 (0.0358)
g/dscm (gr/dscf)
corrected to
12% C02
0.3272 (0.143)
0.1467 (0.0641)
141
-------
regulations require 99% HC1 removal,
whereas the sampling method measures to-
tal Cl emissions. Thus, if there is any
mist or droplets of the scrubbing solu-
tion in the stack, they will probably be
captured by the glass wool "plug" used in
the front of the HC1 sampling probe, ac-
cording to the method specified in Refer-
ence (l). For this reason, Table 5 also
shows HC1 removal when excluding the Cl
determined in the plug component of the
sampling train. In all tests contribu-
tion of the plug to the total was quite
large.
Exclusion of the plug shows that the
99% removal efficiency requirement was
met in 8 of the 12 runs. In order to in-
vestigate the possibility that carryover
of scrubber solution was causing the low
removal efficiency, components of the
sampling train are presently being ana-
lyzed for sodium.
It was noted that there seemed to be
considerable variability in the chlorine
level in the waste feed samples. This is
presently being investigated.
Particulate Emissions
Two 2-h modified Method 5 trains
were used during each test, which re-
sulted in a total of 12 trains. Since
these were primarily intended for POHC
analysis, only two of these trains were
utilized for determination of particulate
catch. This involved evaporation and
weighing of the probe rinse and desicca-
tion and weighing of the filters. Re-
sults of these particulate determinations
are summarized in Table 6.
It can be seen in Table 6 that the
corrected particulate concentration for
Run 2 exceeded the RCRA limit of 0.183
g/dscm (0.08 gr/dscf), while the concen-
tration in Run 5 was below the limit.
However, neither of the above particulate
data is considered very reliable since it
was obvious that the fiberglass-packed
sound dampeners in the stack were deteri-
orating rapidly and pieces of the fiber-
glass were being emitted from the stack
and were interfering with the sampling
probes.
REFERENCES
Guidance Manual for Evaluating Per-
mit Applications for the Operation
of Hazardous Waste Incinerator
Units, prepared by the Mitre Corpo-
ration as a draft document under EPA
Contract No. 68-01-6092, April 17,
1981.
ACKNOWLEDGMENTS
We would like to acknowledge the as-
sistance of Dr. Larry Johnson of EPA's
IERL/RTP and Mr. John Trapp of the
Cincinnati MSD. Principal MRI partici-
pants on this project were Mr. George
Cobb, Mr. Steve Swanson, and Dr. Greg
Jungclaus.
142
-------
EVALUATION OF POTENTIAL VOC SCREENING INSTRUMENTS
Kenneth T. Menzies and Rose E. Fasano
Arthur D. Little, Inc.
Cambridge, Massachusetts
- Merrill Jackson
Technical Support Staff
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
ABSTRACT
This paper describes the evaluation of potential fugitive source emission screening
instruments for analysis of volatile organic compounds (VOC). An initial review of
available portable VOC detection instruments indicated that detectors operating on
several principles (i.e., flame ionization, catalytic combustion, photoionization, infra-
red absorption, and thermal conductivity) might be useful for VOC analysis. However,
flame ionization and catalytic combustion devices evaluated previously have shown poor
sensitivity for highly substituted aliphatic and aromatic organic compounds. Instruments
operating on the photoionization and infrared principles may be able to meet necessary
criteria for practical and accurate VOC analysis of highly substituted organics. There-
fore, three commercially available instruments were selected, modified, and evaluated for
32 such compounds in the concentration range of 100 to 10,000 ppmv. The results indicate
that the photoionization principle may be suitable for general VOC screening but a
reliable instrument/dilution system does not exist at present. The infrared absorption
principle will apparently not provide a suitable general VOC screening device but may be
useful for analysis of some classes of organic compounds.
INTRODUCTION
The U.S. Environmental Protection
Agency has issued guidelines (6) for vol-
atile organic compounds (VOC) from several
stationary source categories such as sur-
face coating operations. These guidelines
are for industries which emit significant
quantities of air pollutants. It has
become apparent that sources other than
classical point sources may also emit
VOC's into the workplace and surround-
ing atmosphere. The EPA1 s Office of
Air Quality Planning and Standards
(OAQPS) is, therefore, evaluating the
need for the control of fugitive emis-
sions of VOC's, from such sources as
valves, pumps, and drains. As described
in EPA Method 21, Determination of Volatile
Organic Compound Leaks (7), technically and
economically feasible devices suitable for
monitoring such leaks include only a few
portable detectors. These devices can be
placed near possible points of emissions
and will respond to releases of the organic
compounds. Specific instruments suitable
for this purpose include, but are not
limited to, catalytic oxidation, flame
ionization, infrared absorption, and
photoionization detectors.
Unfortunately, due to the chemical
complexity of many fugitive sources and
the lack of universal sensitivity of these
detectors, the detectors previously evalu-
ated cannot adequately measure all the
143
-------
volatile chemicals likely to be released.
Tills fact has been documented (2) for two
commercially available detectors using
flame ionization (FID) and catalytic com-
bustion principles. Among 168 compounds
tested, 23 showed sufficiently poor res-
ponse that the actual and measured
concentrations differed by a factor of
greater than five (Table 1). The classes
of compounds which show the poorest agree-
ment with the actual concentration
generally incorporate functional groups
such as halides, hydroxyls (alcohols),
carbonyls (aldehydes, ketones), and car-
boxylates (acids) and include both
substituted aromatic hydrocarbons and low
molecular weight, highly substituted
aliphatic compounds.
Additional portable devices which res-
pond accurately to these compounds are
needed for VOC screening. Instruments
other than flame ionization or catalytic
oxidation detectors which might meet this
goal operate on the principles of infrared
absorption, photoionization, and thermal
conductivity (1).
The first step in this task was to
select and procure one or more units ,of
those detectors which meet the mechanical
specifications of Method 21 (7). The VOC
instrument had to be rugged, reliable,
relatively inexpensive, portable, and easy
to operate. The instrument had to be
intrinsically safe for operation in explo-
sive atmospheres as defined by the appli-
cable National Electric Code. Of course,
it had to respond to the organic compounds
of interest and be able to measure the
leak definition concentration specified
in the regulations. At this time, few
TABLE 1. COMPOUNDS WITH RESPONSE FACTORS EQUAL TO OR GREATER THAN FIVE
OCPDB*
ID No.
120
—
490
790
810
830
—
—
—
—
2060
1221
2073
2105
—
2500
—
2690
1660
2770
2910
—
3291
Compound Name
Acetophenone
Acetyl-l-propanol,3-
Benzoyl Chloride
Carbon Disulfide
Carbon Tetrachloride
Chloro-Acetaldehyde
Dichloro-l-propanol,2 , 3-
Dichloro-2-propanol,l,3-
Diisopropyl Benzene,!, 3-
Dimethyl Styrene,2,4-
Formic Acid
Freon 12
Furfural
Glycidol
Hydroxyacetone
Methanol
Methyl-2, 4-pentanediol,2r:
Methylstyrene, a-
Monoethanolamine
Nitrobenzene
Phenol
Phenyl-2-propanol, 2-
Tetrachloroethane, 1, 1,2,2-
FID
Response Factor
10.98
10.87
6.40
571.92
21.28
13.40
61.51
29.34
9.43
37.09
34.87
9.65
7.96
8.42
8.70
5.69
96.34
10.24
28.04
29.77
11.75
89.56
6.06
*0rganic Chemical Producers Data Base
Response Factor
Source: (2)
Actual Concentration
Measured Concentration
144
-------
detectors are "approved."
The" second step in this task was to set
up a laboratory system capable of mixing
known volumes of vapors with air and
delivering the mixtures of known concen-
tration to the detectors. Tedlar bags and
a volumetric mixing system were selected
for sample preparation since they provide
adequate accuracy/precision and require.
little cost or time to set up.
The third step in' this task was evalu-
ation of the detectors for response to
the compounds of interest. The response
factors were determined at several con-
centrations over the range of 100 to
10,000 ppmv. Measurements were limited
to concentrations approaching about 90%
of the saturation concentration or 75% of
the lower explosive limit (LEL). In order
to permit statistically valid interpreta-
tion of the measured response factors,
five replicate measurements at three con-
centrations were conducted. Data analysis
included calculations of mean response
factors and confidence intervals.
Subsequent to this evaluation, a photo-
ionization detector attached to a gas
chromatograph was utilized to independently
confirm the r^sonse factors observed with
the portable detectors. This step provided
appropriate quality assurance and expanded
the available data base.
INSTRUMENT SELECTION
General Rationale
A recent summary of available portable
VOC detection devices (1) lists a number
of instruments operating on the following
principles:
Flame lonization (FID)
Photoionization (PID)
Infrared Absorption (IR)
Thermal Conductivity (TC)
Hot Wire/Catalyst Combustion
(Combustion)
The majority of available instruments
operate on one of three principles; i.e.,
FID, IR, or Combustion (1). As noted
previously: (2), two specific FID and Com-
bustion devices were shown to have poor
sensitivity to several substituted organic
compounds. Due to this observation and
with the assumption that other FID or
Combustion detectors available from dif-
ferent manufacturers probably do not differ
significantly in their basic response
factors, alternative VOC screening devices
were evaluated. These were selected from
instruments operating on other detection
principles, including PID, IR, and TC (1).
The selection of potential VOC detectors
from this list depends on several criteria
which are outlined in EPA Method 21. That
is, an instrument suitable for screening
should have the following .characteristics:
1. Fast response (<30 seconds);
2. Measurement range 100 to
10,000 ppmv;
3. Similar responsiveness to a variety
of organic vapors;
4. Portable;
5. Rugged;
6. Reliable;
7. Inexpensive;
8. Easy to operate; and
9. Intrinsically safe (as per National
Electric Code).
Each of the first three characteristics
is of primary importance in providing a
practical instrument for VOC screening.
Fast response time is necessary for rapid
screening of a large number of fugitive
sources. The specified measurement range
is required by the need to limit signifi-
cant leaks of VOC's.
Equal molar sensitivity to compounds of
widely differing functional character is
not achievable with currently evaluated
instruments but is a desirable goal. The
other characteristics, such as portability
and intrinsic safety, are also important
but none should be considered individually
critical to the acceptance of a potential
detector.
Assuming that characteristics of fast
response and appropriate measurement range
are available in potential VOC detectors,
the ability of the devices to meet the
criterion of similar responsiveness needs
to be reviewed prior to final instrument
selection. Thus, the efficacy of various
operating principles to meet this criterion
is discussed below.
Photoionization detectors utilize
ultraviolet radiation to ionize a small
fraction of molecules introduced into an
145
-------
ionization chamber. The ionization process
is initiated by absorption of a photon of
sufficient energy (i.e., greater than the
ionization potential) to remove an electron
from its ground state to infinity. A free
electron and positively charged ion are
thus formed:
R + hv •*- R+ + e~
If the energy of the UV lamp is less
than the ionization potential of the com-
pound, no ionization takes place. Ions
formed in the detector/ionization chamber
may reach the electrodes under the in-
fluence of an electric field and produce
a small current. The number of ions which
reach the electrode is proportional to the
concentration, although only a very small
fraction ("vO.01%) of the molecules in the
ionization chamber are ionized by incident
radiation. Depending on the character of
the electrons (e.g., sigma vs pi electrons)
the yield of ions (photoionization effi-
ciency) may vary as a function of the
energy of incident photons. At present,
UV sources are available for commercial
instruments which emit photons of approxi-
mately 9, 10, or 12 eV. Based on the
ionization potentials of organic com-
pounds (5), it is apparent that certain
classes of compounds (e.g., aromatics and
aliphatics greater than carbon number C7)
can be ionized by a 10 eV lamp while many
substituted aliphatics require photons of
at least 11 eV. This observation leads to
the conclusion that with sufficient energy
roost organic compounds can be ionized and
detected. A practical upper energy limit
for VOC analysis is about 12 eV since the
major components of air (such as nitrogen,
carbon monoxide, carbon dioxide, and water)
have ionization potentials above this
level. As well as the ionization
potential, the photoionization efficiency
is important since this parameter deter-
mines sensitivity of the technique to
different compounds. A recent paper (4)
indicates that the molar sensitivity of
aliphatic and oxygenated aliphatic com-
pounds is several times less than that of
aromatic compounds if incident radiation
is about 10.2 eV. In fact, for aliphatic
hydrocarbons of carbon number less than
C8, the relative sensitivity is less than
one-tenth that for benzene. If incident
radiation is about 11.7 eV, the relative
sensitivity of aliphatic and aromatic
compounds is similar (3) and perhaps
within a factor of two. Based on this
assumption, a commercially available
photoionization instrument with a lamp of
about 12 eV may provide a generally appli-
cable VOC detection technique.
Typical nondispersive infrared devices
operate by passing infrared radiation
through two separate absorption cells: a
reference cell and a sample cell. The
sealed reference cell is filled with non-
absorbing gas, such as nitrogen or argon.
The sample cell is physically identical to
the reference cell and receives a continu-
ous stream of gas being analyzed. Sub-
sequently, the net radiation in the two
beams are passed into and absorbed in
matched selective detectors (e.g., Luft
detector) containing the vapor to be
detected. When organic vapors are present
in the sample cell, energy is absorbed,
and the temperature and pressure in the
corresponding detector is reduced' relative
to that in' the detector on the reference
side.of the analyzer. A diaphragm between
the two detectors is displaced and the
amount of displacement is detected, elec-
tronically amplified, and an output signal
proportional to concentration produced.
In other NDlR systems, narrow bandwidth
filters which pass energy which corresponds
to that absorbed by the compound of
interest are used along with simple solid
state IR detectors. In both cases, inter-
ference from compounds with overlapping
absorption bands is possible. More
importantly, the maximum absorbing wave-
length for different organic species in
the sample gas may not correspond to .the
maximum absorbing wavelength of the cali-
bration compound used in the detector.
Within reason, several different calibra-
tion compounds could be used in the
detector to improve responsiveness for
several compounds. Alternatively, by
selection of a single narrow bandwidth
filter with a' wavelength corresponding to
a general aliphatic C-H stretch, many
aliphatic hydrocarbons might be detected
quite uniformly. Based on the maximum
absorption wavelength of aromatic hydro-
carbons, a separate filter or cell would
be needed for this class of compounds. In
practice, the specificity of the detection
principle has precluded the manufacture of
an NDIR device suitable as a general
(i.e., both aliphatic and aromatic)
organic vapor detector.
An alternative IR detection scheme
involves"dispersive IR analysis in which
146
-------
the specific wavelength absorbed by the
organic vapor of interest is passed through
a single sample cell. In this case, selec-
tivity is provided by a monochromatic light
source rather than a selective detector.
Such a device is inherently more selective
than an NDIR and thus may be less appro-
priate as a VOC screening device. However,
by successive, rapid monitoring of IR.
absorption at several selected wavelengths
corresponding to the maximum absorption
wavelengths for several organic functional
groups (e.g., aliphatic CH, aromatic CH,
C-C1, C=0), it may be possible to identify
and quantify a wide variety of organic
vapors in a fugitive emisson source.
Unit Selection
On the basis of the factors discussed
above, both the IR and PI principles might
be suitable for general VOC screening.
However, a comparison of the specifications
of commercially available instruments
operating by these principles and the
criteria of Method 21 leads to a rather un-
fortunate conclusion: No instruments of
these types are adequate for screening of
VOC emissions. In terms of a desire to
expand the list of potential detectors,
such a finding is unsatisfactory.
What criteria led to this finding? The
most obvious answer is the requirement for
an intrinsically safe device. No IR or PI
devices are certified for use in Class I,
Division 1 environments. One PID is cer-
tified for use in Class I, Division 2. It
should be noted, however, that in the future
other devices may be modified so as to
meet Class I, Division 1 certification.
Alternatively, the use of an instrument
only in less hazardous environments may not
be considered as particularly restrictive.
For these reasons, the criterion of intrin-
sic safety was given lesser significance
and not used to rule out potential devices
for screening in this program.
The other criteria listed previously
were ranked in approximately descending
order of importance for field application.
Thus, a response time of less than 30 sec-
onds was given the highest- importance. In
fact,, fast response time is very important
to practical measurement of VOC leaks, and
several instruments with faster than 30-
second response time are availabile. Thus,
it was decided that this criterion must be
met by any instrument to be evaluated. The
criteria of portability, ruggedness, and
ease of operation are also important but
were not chosen as absolute selection
criteria. Portability can be evaluated
subjectively, and devices operated with
automobile batteries placed on a small
cart may be considered to have adequate
portability. Devices operated with AC
power are less practical in many indus-
trial environments.
Current literature and additional
manufacturers' information were reviewed
using the modified criteria described
above. A list of potential VOC detectors
was developed (Table 2). This list in-
cludes instruments which operate on PI
and IR principles and which meet mo.st of
the Method 21 criteria. It is obvious
that not all of the IR instruments that
meet most of the criteria are included.
Since the goal of this study is to evalu-
ate the usefulness of the; IR principle
rather than all available IR devices, only
selected IR devices suitable on the basis
of the criteria for VOC screening were
included.
The list, therefore, includes a dis-'
persive IR device and NDIR devices with
or without solid state detectors. The'
NDIR devices may be useful for a specific
group of organics; e.g., aliphatic hydro-
carbons. Also included are two devices
operating on other detection principles;
i.e., ion capture and UV spark. These
instruments are only useful for a specific
class of organics (i.e., halogenated hydro-
carbons) but were included due to the
widespread industrial use of these
solvents.
The final instruments selected for
evaluation were:
Ins trument
AID Model 580
HNU PI-101
Foxboro Miran 80
Principle of Operation
Photoionization
Photoaonization
Dispersive Infrared
The .rationale for the selection of
these instruments is based on both suit-
ability and availability. As noted
previously, PI may be a particularly suit-
able VOC detector for aliphatic, aromatic,
and substituted organic vapors. Either of
the commercially available, portable
instruments meets most Method 21 criteria
and may be suitable for evaluation.
147
-------
3
1
1
P
8
d
1
£™4
S
P*
04
fO
1
cd
•H 0)
CO *4-4
(3 cd
•H cn
M
s
J_l
01
g
O
*
Cu
s
(2
*§
Q)
Q
0
eu
(0
OJ
X*Ni
a
Pu
t>0
5
CD
rH
U
£*
U
e
0)
Instrun
o o
u
0)
4J /~v
cd .a
pq H
H i — •
O cd co
< 0 ^
CO CO
0) (1)
!« >H
-a- CM
CM CO
m o
rH
O
o
o
o
V *
g a
H P,
rH rH
• E U
0 E rt
(3 m H
M V
•> ^3
« O CO
gg M a
K » >,
ij M
CU w
t ^ »
^
cu
4J
4-1
cd
M
CO
CU
t«
0,
m
o
o
CM
O
O
« O
0 0
CM CM
O O
a
o
•H
4J
cd
•H
a
o
•H
o
4-1
O
g
|
r\
CO
g
CU
4J ^-N
CO rH
cn H
23 P-(
§
o
S3
CO
CU
>>
co
CM
o
CM
1-1
O
0
0
O
rH
|
6
t
rH
O
y*^
CU CU
M i '
3 g.
e. cu
o o
1-1
(3 cd
°«
-w
(U
,— {
pr*]
General
(TVM 1)
S
u
<
CO
cu
!*
CO
m
o
o
o
n
0
H
1
O
n
0
0
H
1
O
s-^
CU
4J
Ctt
AS a
H 0)
CO 00
a. o
en H
Ej ^
U
)— ]
A
,fl /-N
ej 01
0) n
H v
t— i
co cd
4-1
•H
.§
^
0)
.5
CO
o
H
1
t— 1
111
[ "j
1
148
-------
The Miran 80 instrument is the one
available infrared device which permits a
selection of wavelength as opposed to
selection of a test compound in a reference
cell. This option permits the rapid (a
few seconds) assessment of the suitability
of several wavelengths for the measurement
of the substituted organics of interest.
Specific examples are an aliphatic C-H
stretch, aromatic C-H stretch, or a C=0
stretch. Other portable IR devices utilize
a filter at one specific wavelength band
corresponding, for example, to an aliphatic
C-H stretch. Thus, they have an inherent
selectivity against aromatic or substituted
species. Since all three classes of com-
pounds are of interest, the latter devices
are not preferred as VOC screening devices.
It is possible that one wavelength may be
suitable for analysis of a wide variety
of organic vapors. If that is the case,
other IR instruments could potentially
be used for VOC leak, detection. In
summary, the Miran 80 (with associated
microprocessor) permits the most rapid and
cost-effective assessment of the IR
principle as a general VOC detector. Other
IR devices were, therefore, not evaluated.
The halocarbon specific detectors
(i.e., General Electric TVM. 1 and Gas Tech
Halide Detector), were not selected for
evaluation despite the potential usefulness
of such a device in environments subject to
halocarbon solvent contamination. Neither
the GE TVM 1 nor an equivalent model is
now sold. The Gas Tech device could not
be modified to meet the intrinsic safety
requirements of Method 21 due to the
presence of a spark in the detector
section.
No other instruments appeared to have
a reasonable expectation of meeting the
Method 21 criteria and of providing a
significantly different performance than
those devices evaluated previously or
selected for evaluation in this study.
Unit Modifications
Both photoionization devices operate
with a maximum quoted linear range of
0-2000 ppmv. In fact, the linear range
is frequently reported to be only about
0-1500 ppmv. Since the maximum concen-
tration of concern in VOC screening is
10,000 ppmv, dilution of sample air is
necessary for both instruments to operate
in the linear range. Both HNU'Systems, Inc.
and AID, Inc. provided their instruments
with dilution systems designed in their
respective laboratories. The HNU Systems,
Inc. design consisted of (1) a fine bore
restrictor which limited the flow of sample
air and (2) a charcoal tube which passed an
excess (lOx) of hydrocarbon-free air
(methane is not removed but does not res-
pond in the detector). The sample stream
is thus diluted about 1. to 10. The AID,
Inc. design consisted of a pump and needle
valve which diverted 90% of the incoming
sample air through a charcoal tube and 10%
to the normal exhaust point. The hydro-
carbon-free (except for methane) sample
air is combined with the incoming sample
stream and thus a continuous tenfold
dilution is provided.
Problems were observed with .these
dilution systems and the UV lamps provided
with both instruments. The absolute
accuracy of the dilution ratios is in some
doubt since independent flow rates were
difficult to measure. The UV lamps
provided with both instruments were subject
to degradation during the life of the study.
In fact, the 11.8 eV lamp supplied with the
AID, Inc. device failed during the study
and, unfortunately, a replacement could
not be obtained in time to collect useful
data with this instrument. The 11.7 eV
lamp supplied with the HNU Systems, Inc.
device failed during the study and a
replacement was provided. The difference
in energy output from the two HNU Systems,
Inc. lamps was large (i.e., a factor of
three to ten depending on the age of the
lamp). This variation affected the linear
range of the•instrument and created prob-
lems in obtaining consistent results. In
some cases with the new lamp, saturation
of the detector occurred even with the
dilution probe attached to the instrument.
The test results reported must, therefore.
be carefully interpreted and conclusions
narrowly drawn.
The Miran 80 operates over the concen-
tration range from ppm to percent. The
wide dynamic range is provided by a cell
in which the pathlength of IR radiation
can be changed by optical folding of the
incident beam. At the concentration range
of interest (i.e., 100 to 10,000 ppmv), the
incident beam traversed a distance of about
0.75 m. At this pathlength, the full-scale
absorbance for vapors of interest at a
concentration of 10,000 ppmv was about
1 absorbance unit. Once the cell
149
-------
pathlength was set, no other modifications
of operating conditions were required.
COMPOUND SELECTION
As noted in the Introduction, 168
compounds had previously been tested for
response factor on two commercially avail-
able VOC detectors (2). Twenty-three
showed sufficiently poor response that
the actual and measured concentrations
differed by a factor of greater than
five (Table 1). The classes of compounds
showing poor agreement were generally
highly substituted aliphatic and aromatic
compounds and those compounds incorpo-
rating functional groups such as carbonyl
and hydroxyl groups. These 23 compounds
were selected for testing on the alter-
native VOC screening devices to be evalu-
ated in this study. Several other
compounds, which were not evaluated in
the previous work, were added to the list.
These compounds include only a portion
of those commonly used in chemical produc-
tion. At the request of OAQPS, other
industrial compounds which have a vapor
pressure greater than 0.3 kPa but which
were not considered previously were
reviewed. This extensive list of 76
compounds includes many species for which
an FID or catalytic combustion detector
would respond well. However, others are
highly substituted compounds which will
probably not give adequate response on
these two detectors. Selected substituted
compounds from the list were included in
the detector evaluation.
The selection criteria required a
response to several questions:
1. Are subsequent groups present
or absent? If absent, don't
test,
2. Are the compounds similar
(functionally and/or isomerically)
to others previously evaluated?
If they are, don't test,
3. Are response factors on an FID
instrument likely to exceed five?
If not, don't test, and
4. Do the compounds pose a serious
health hazard to laboratory
personnel? If they do, cautiously
consider evaluation.
As a result of the responses, the
compounds were separated into two groups:
compounds that should and those that need
not be analyzed. Within the first group,
the compounds were prioritized on the
basis of (1) their similarity to other
vapors to be analyzed (for example,
positional isomers of compounds selected
for testing were given lower priority),
and (2) their health hazard (extremely
toxic compounds with little commercial
application or likelihood of release and
which require complex/expensive handling
were given lower priority).
The compounds selected for evaluation
in this program are listed in Table 3.
They are listed in the approximate order
of testing.
EXPERIMENTAL PROCEDURES
Determination of response factors
required initial calibration of the VOC
detectors with a gas or gases of known
concentration. Methane had been used
previously for calibration of FID and com-
bustion analyzers (2). However, the PID's
do not respond to this compound, and the
multiple wavelength analysis by dispersive
IR spectrometry cannot be carried out by
the use of methane as a single calibration
gas.
Therefore, 1,2-dichloroethane was
selected as the calibration gas for the
PID's. This compound can be detected by
the instruments and it has a response
factor of about one (compared to methane)
when analyzed on an FID. As a result,
data collected in this study may be
comparable to data collected in a previous
EPA study (2).
Several calibration compounds were
used for the IR evaluation since several
wavelengths were scanned in the dispersive
IR instrument to determine if any wave-
length gave similar response factors for
all the compounds of interest. The wave-
lengths were selected to correspond to
key functional groups of the test com-
pounds to be analyzed. The wavelengths,
functional groups, and calibration compounds
are listed in Table 4.-
Once calibrated, the instruments were
150
-------
TABLE 3. COMPOUNDS FOR EVALUATION
Carbon Bisulfide
Carbon Tetrachloride
Chloro-Acetaldehyde
Bichloro-l-^propanol,.2,-3-
Dichloro-2-propanol,1,3-
Biisopropyl Benzene,1,3-
Dimethyl Styrene,2,4-
Formic Acid
Freon 12
Methanol
Methylstyrene, a-
Tetrachloroethane,1,1,2,2-
Ethanol
Formaldehyde
Ethylene Bichloride (Bichloroethylene)
Chlorinated Ethanes (C2H5C1, etc.)
Chlorinated Methanes (CHsCl, etc.)
Acetophenone
Benzoyl Chloride
Furfural
Monoethanolamine
Nitrobenzene
Phenol
Acetyl-1-propanol,3-
Glycidol
Hydroxyacetone
Methyl-2,4-pentanediol,2-
Phenyl-2-propanol,2-
Aniline Hydrochloride
Bifluoroethane
Diketene
Bimethylsulfide
Glyceroldichlorohydrin
Paraldehyde
Perchloromethylmercaptan
Propylene Chlorohydrin
Toluenesulfonic Acid
Toluene Sulfonylchloride
Ethylene Glycol Bimethyl Ether
Ethylene Glycol Monoethyl Ether Acetate
1-Pentanethiol
Acetal
Chlorobenzoylchloride
Chlorodifluoromethane
Chlorotrifluoromethane
Tr ichloro fluoromethane
Trichlorotrifluoroethane
Cyanoacetic Acid
Neopentaneoic Acid
AmyImercaptans
2-methyl-2-butanethiol
2-methyl-l-butanethiol
3-methyl-l-butanethiol
Glycols
Ethylene Glycol Monoethyl Ether
Ethylene Glycol Monomethyl Ether
Ethylene Glycol Monomethyl
Ether Acetate
Ethylene Glycol Monopropyl Ether
Glycol Methyl Ether (Bioxolane)
used to analyze the test compounds at three
concentrations over the range, of 100-10,000
ppmv. Details of the operation of each.
instrument are given in the manufacturer1s
instruction manual. Buring'the tests, 'the
lamp in Model 580 failed and a replacement
could not be obtained. Therefore, only the
PI-lOl and Miran 80 were evaluated. The
<^£Jsonse5factor was determined by calculating
the ratio of the actual concentration to the
concentration indicated by the instrument.
The following paragraphs describe the
procedures involved in calibration and
operation of the instruments, preparation
of test gas samples, and calculation of
response factors.
Gas mixtures tested in this study were
prepared in Tedlar gas sampling bags of a
nominal 25 liter volume. These bags
provide a relatively inert surface to pre-
clude absorption, reaction, or permeation.
They also permit visual inspection of the
bag interior to provide an indication of
sample condensation or reaction. The bags
are equipped with two valves to facilitate
flushing of sample gas and a septum to
permit injection of sample liquid with a
syringe.
Gas samples were prepared by the
following procedure:
1. Flush and evacuate bag three
times with hydrocarbon-free' air
•(i.e., until no hydrocarbons are
detected on each instrument).
2. Fill bag with 20.0 L of hydro-
carbon—free air.
3. Inject a known volume of test
compound into the bag.
4. Permit at least 1 hour equilibra-
tion to ensure adequate evaporation
and mixing.
5. Braw gas sample from bag with
each instrument.
The target concentrations prepared for
151
-------
TABLE 4. CALIBRATION SCHEME FOR MIRAN 80
Wavelength (pm)
Functional Group(s)
Calibration Compound
3.3
3.4
3.6
4.0
5.7
6.35
8.8
9.5
13.5
Aromatic & Unsaturated C-H
Saturated C-H
Aldehyde C-H
Reference Wavelength
Carbonyl C=0
Aromatic C-C,
conj C=C (also N-H, C-S)
Ether C-O-C
Alcohol C-O-H
C-C1
Toluene • .
Pentane
Butyraldehyde
Air
Acetone
Toluene
Diisopropyl Ether
Isopropanol
1, 2-dichloroethane
each compound were 500, 1000, 5000, and
10,000 ppmv. In several cases, it was not
possible to prepare the higher concentra-
tions due to the low vapor pressure of the
compound or due to safety reasons; that is,
such a concentration would exceed the lower
explosive limit. In these cases, a con-
centration of 100 ppm was often prepared.
For each target concentration, the required
volume of liquid was calculated and
measured in a microliter syringe.
Each instrument was initially
calibrated (spanned) with a gas sample
prepared in triplicate at a concentration
of 10,000 ppmv. Calibration curves were
then prepared by introducing samples of
the calibration gas, prepared in triplicate
at five concentrations over the range of
100 to 10,000 ppmv, into the instruments
and recording the response. The PI-101
was calibrated with 1,2-dichloroethane.
while the Miran 80 was calibrated at
individual analytical wavelengths with
the compounds listed in Table 1.
During subsequent analysis of each
test compound, the HNU PI-101 instrument
was spanned with an 8040 ppmv 1,2-dichloro-
ethane certified gas standard provided by
Scott Specialty Gas, Inc. of Plumsteadville,
PA. This span was carried out just'prior
to analysis of each set of sample bags for
each test compound.
The Foxboro Miran 80 instrument was
electronically zeroed and spanned according
to the manufacturer's instructions. This
zero and span check was carried out prior
to analysis of each set of sample bags
for each test compound.
The response factor reported in the
following test results section is the
number that, when multiplied by the
apparent concentration based on instrument
response, yields the actual concentration
as calculated to exist in the gas bag
sample. That is:
Response Factor (RF)
Actual Bag
Concentration (C)
Concentration
Calculated from
Instrument Response
Response factors were determined at
three actual concentrations; i.e., generally
100, 500, 1000, 5000, or 10,000 ppmv. No
attempt wa's made tp fit the three response
factors for each compound to a particular
function. For some compounds, the response
factor is nearly identical for each con-
centration; whereas, for others it differs
dramatically and in a complex manner. The
response factor for individual compounds
is, therefore, not reported for an observed
instrument response of 10,000 ppmv. In-
stead, the mean response factors calculated
from up to five replicate data points at
each of the three actual bag concentrations
are reported along with the standard devi-
ation. Also reported is the 95% confidence
intervals for the response factors as
calculated from Student's t-test.
To provide additional quality control,
expand the data base, and confirm response
factors, HNU Systems, Inc. photoionization
detector Model PI-51 was installed on a
Varian 2800 gas chromatograph. Using an
11.7 eV lamp, solutions 'of specific organic
compounds and toluene (reference compound)
152
-------
in several solvents were analyzed on a
3% SP 2250 on 100/120 ,S,upelcoport column.
Peak areas for the compound of interest
and toluene were measured and relative
detector:.sensitivities were determined on
a molar basis and normalized against the
detector response for toluene.
The molar sensitivity relative to
toluene was calculated using the following
equation:
RF =
A(tol)
B
where RF is the response,factor,
A is the peak area of the compound
of interest,
A(tol) is the peak area of the
toluene peak,
B is the molar concentration of
the compound of interest, and
B(tol) is the molar concentration
of the toluene in the solution.
RESULTS AND DISCUSSION
Photoionization Detection
The .photoipnization technique was
evaluated for a limited number of compounds
due to both chemical and, more signifi-
cantly, equipment problems. The PI-101
was calibrated with dichloromethane so as
to permit direct comparison with response
factors reported by Brown et al. (2). The
response factors observed for the 16 com-
pounds tested on the photoionization
detector, PI-101, range from 0.50 to 48.
Seventy-five percent (12) of the compounds
have response factors of less than five
and greater than 0.2. There appears to
be no obvious trend of response factor
with molecular weight (carbon number) or
functionality within this group. On the
other hand, it is interesting to note
that, both alcohols tested (i.e., methanol
and ethanol) have response factors greater
than five, and the values for methanol (Cl)
are higher than those for ethanol (C2).
Thus, it appears that non-bonding electrons
on the oxygen atom of the alcohols do not
provide a much greater photoionization
yield than other sigma-bonded electrons in
compounds with similar carbon numbers.
The high response factor for trichlorotri-
fluoroethane is consistent with its high
ionization potential (11.78 eV). In fact,
this, ionization potential is slightly
higher than the quoted energy of the UV
lamp used in the study. This may indicate
that thermal energy provides sufficient
additional.energy to permit some ioniza-
tion when coupled to the energy provided
by the UV light.
Although the specific response factors
for the limited number of compounds tested
do not unequivocally confirm the suit-
ability of photoionization as a general
VOC screening technique, an important but
cautious observation can be made. That is,
based on this, small sample of compounds
tested, which includes an aromatic compound
(i.e., toluene), an ether (i.e., acetal),
an alcohol (i.e., ethanol), and chlorinated
alkanes (i.e., trichloroethane and chloro-
form) , the response factor over a concen-
tration range of 500 to 10,000 ppmv may be
within a factor of five. This result is
consistent with an expectation of more
similar photoionization yields from sigma
and pi electrons when the compound is
influenced by UV radiation of approximately
12 eV rather than 10 eV. The expectation
that photoionization yields for aliphatic
and aromatic compounds may be similar
indicates the potential usefulness of
photoionization as a VOC screening tool.
In terms of current availability as a
potential VOC detector, the most signif-
icant result with respect to the photo-
ionization detector (HNU Systems, Inc.
PI-101 and AID, Inc. 580) is probably the
difficulty observed in operating the
prototype dilution system. Both dilution
probes were designed and fabricated by the
respective manufacturers under severe time
limitations. Neither probe was designed
in a manner which permitted reliable
independent measurement of dilution ratio
or reproducible adjustment. Thus, the
absolute dilution ratio is in some doubt.
The ability to adjust the dilution ratios
was practically nonexistent. As noted
previously, the fixed dilution ratios were
inappropriate for analysis of vapor
concentrations which yielded instrument
responses much above 10,000 ppmv or much
below 1000 ppmv. Detector saturation
was observed somewhat above an instrument
response of 10,000 ppmv. At the
span settings required for adequate
operation, the background instrument
response to zero air was quite
high.
153
-------
Whenever the intensity of the UV lamps
began to decrease (note that the AID, Inc.
lamp failed early in the program), the
instrument span had to be increased
regularly. Some alteration to the span
potentiometer setting could be made to
correct for this decrease in response.
However, for some tests the correction was
not sufficient to yield an identical cal-
ibration. Under these conditions, response
factors were calculated at a different
absolute instrument response. However,
since the calibration curve is linear
over the range of 0-10,000 ppmv (with
dilution, that is about 0-1000 ppmv), no
systematic error should occur due to the
change in absolute response.
Due to declining instrument response
and low vapor pressure of many compounds,
half of the compounds tested did not
yield reliable response factors. The
problems noted above and limited data
obtained indicate that, at the present
time, a reliable photoionization system
does not exist to operate over a VOC
concentration range of 100 to 10,000 ppmv.
More accurately, a reliable dilution/photo-
ionization system is not available.
As noted previously, a comparison of
response factors determined on the portable
PI-101 analyzer and the PI-51 attached to
a GO was conducted to independently confirm
the accuracy of the data. The response
factors for three chlorinated hydrocarbons
listed below (Table 5) indicate good
agreement (i.e., within ± 13%) on both
instruments.
As a result of this agreement, the GC
technique was utilized to effectively
gather response factor data for homologues
of several compound classes. Limited data
have been attained (Table 6).
It is apparent that the response
factor depends upon carbon number and
functional character. For the three com-
pound classes analyzed—aromatic hydro-
carbons, halogenated aromatics, and
chlorinated aliphatics—the response factor
(relative to toluene) increases as the
carbon number increases. Also, for the
aromatic halides, the response factor
increases as the electronegativity de-
creases.
Based on these observations it appears
that a photoionization detector with an
11.7 eV lamp acts as a rough carbon counter
and is able to respond to organic compounds
with as few as one carbon atom and with
significant substitution.
Infrared Detection
The results of the evaluation of the
Miran 80 are more complete. A total of 32
compounds were analyzed. As noted pre-
viously, other compounds were not tested
for several reasons, including (1) low
vapor pressure; (2) reactivity; (3) lack
of availability; and (4) close chemical
similarity to compounds previously tested.
Prior to testing, the instrument was
calibrated with individual span gases at
eight analytical wavelengths which cor-
respond to individual functional groups;
e.g., C-H, C-C1, C-OH. The calibration
curve data indicate that the absorbance
values observed over the concentration
range of 100 to 10,000 ppmv are linear.
Test compounds were then run and the .
instrument response calculated on the basis
of the response indicated by the specific
span gas used at individual analytical
wavelengths.
An analysis of the data indicates
that the response factors for most compounds
TABLE 5. COMPARISON OF RESPONSE FACTORS FOR PHOTOIONIZATION DETECTORS
Compound
1, 2-Dichloroethane
Methylene Chloride
Dichloroethylene
RF (PI-101)
N=5
0.56
0.37
0.41
RF (PI-51)
N=8
0.62
0.40
0.36
% Agreement
90.3
93.. 3
87.8
154
-------
TABLE 6. RESPONSE FACTORS DETERMINED ON THE HNU SYSTEMS, INC.
PI-51 DETECTOR
Compound Response Factor
Aromatic Hydrocarbons
Benzene :
Ethyl benzene
n-Propyl benzene
Cumene
n-Di±sopropylbenzene
Halogenated Aromatlcs
Fluorob enzene
Chlorobenzene
Bromobenzene
lodobenzene
Chlorinated Aliphatics
Dichloromethane
1,2-Dichloroethane
1,3-Dichloropropane
1,4-Dichlorobutane
0.49
1.34
1.58
1.30
2.30
0.60
1.45
2.28
3.31
0.40
0.62
1.12
1.54
with a particular functional group, deter-
mined at an analytical wavelength which
corresponds to that functional group
(Table 4), are generally less than a
value of twenty. This is consistent with
the general observation that the functional
group is more important than the remainder
of the molecule in determining the IR
extinction coefficient of the compound
at the wavelength of interest.
For example, three of the four aromatic
compounds tested have reasonable response
factors (less than five) at 6.35 ym as
shown below. This wavelength is within a
broad aromatic ring stretch area.
Compound
Diisopropyl Benzene
Dimethyl Styrene,2,4-
Methyl Styrene
Response Factor
Range
2.42 - 3.75
0.185 - 0.394
0.229 - 0.718
Within this group, the addition of the
large aliphatic group (isopropyl) on the
benzene ring appears to reduce the sensi-
tivity (larger response factor) at the
aromatic C -^ C stretch wavelength as
compared to less alkylated arbmatics.
In the case of aliphatic and substi-
tuted aliphatic compounds, the C-H stretch
wavelength of 3.3 ym yields suitable
response factors (less than five) for
about 52% of those tested. The classical
aliphatic C-H stretch is observed at 3.4ym,
but some overlap of 3.3 and 3.4 ym IR bands
may occur in the Miran 80 due to incomplete
resolution. Also, some shift of the CH
stretch wavelength probably occurs due to
nearby oxygen or halogens. A list of
aliphatic compounds and corresponding res-
ponse factor ranges at this wavelength are
shown in Table 7. If one-includes alkylated
aromatic compounds in the list of com-
pounds with response factors less than five
at 3.3-3.. 4 ym, the percentage of compounds
tested with suitable response factors
increases to 62%.
Ten chlorinated hydrocarbons tested in
this program yielded measurable response
factors at 13.5 ym. Seventy percent were
observed to yield response factors less
than five at this wavelength. The compounds
155
-------
and respective response factors are given
in Table 8.
Since the ultimate goal of this
instrument evaluation is to assess the "
suitability of IR as a general VOC screen-
ing technique, an assessment of the use-
fulness of a single wavelength for
measurement of organic compounds of varied
molecular weight and functionality is in
order. A review of the data indicates
that the number of test compounds (total
of 32) which yield response factors of
less than 20 or greater than 0.05 at each
analytical wavelength is as follows:
Wavelength (urn)
3.3
3.4
3.6
5.7
6.35
8.8
9.5
13.5
Number of Compounds
23
12
13
17
17
11
25
14
In some cases, the response factors at
a particular wavelength (e.g., 5.7 ym) are
Strongly a function of concentration. It
appears that this may be due to a concen-
tration broadening phenomenon which is
frequently observed ..in gas-phase IR spec-
trometry. ' If only those compounds which
show a response factor between 5 and 0.2,
and those which show no strong-variation
in response factor with concentration
(i.e., less than a factor of two from 1000
to 10,000 ppmv)"are summarized as above,
fewer compounds yield suitable response
factors:
Wavelength (um) , , Number of Compounds
3.3
3.4
3.6
5.7
6.35
8.8
9.5
13.5
12
4
3
1
3
4
15
7
The results indicate that only 3.3, 9.5,
and 13.5 ym analytical wavelengths respond
acceptably for a large number of compounds
(i.e., greater than 10% of the total number
of compounds). However, in any case,
fewer than 50% of the compounds are
reliably detected. ' The aliphatic and
aromatic compounds do not overlap at
6.35 ym but do overlap at 3.3 ym. However,
TABLE 7. SUBSTITUTED ALIPHATIC COMPOUNDS WITH RESPONSE FACTORS
LESS THAN TWENTY AT 3.3 ym
Compound
Response Factor Range
Acetyl-1-propano 1,3-
Chloro-acetaldehyde
Dichloro-1-propanol ,2,3-
Dichloro-2-propanol, 1, 3-
Diketene
Dimethylsulfide
Ethanol
Ethylene Glycol Dimethyl Ether
Ethylene Glycol Monoethyl Ether Acetate
Formaldehyde
Formic Acid
Glycidol
Methanol
Methylene Chloride
Pentanethiol, 1-
Propylene Chlorohydrin
Tetrachloroethane, 1, 1, 2,2-
Trichloroethane, 1,1,1-
1.23 - 2.02
2.73 - 8.62
18.5
5.29 ,
8.06 - 14.1
0.488 - 0.495
0.261 - 0.292
0.196 - 0.296
0.280 - 0.488
1.09 - 1.88
0.529 - 0.722
0.382
0.294 - 0.410
2.67 - 2.87
0.314 - 0.633
0.334 - 0.403
8.59, - 9.90
1.69 - 3.76
156
-------
TABLE 8, CHLORINATED COMPOUNDS AND RESPONSE FACTORS AT 13.5 ym
Compound
Response Factor Range
Carbon Tetrachloride •"
Chlorp-acetaldehyde
Chloroform
Dichloro-l-propanol,2,3-
Diehloro-1-propanol, 1,3-
Methylene Chloride
Propylene Chldrohydrin
Tetrachloroathane, 1, 1, 2, 2-
Trichloroethane, 1, 1,1-
Trichlorotrif luoroethane, 1, 1, 2-
0.233
0.435
0.150
0.431
0.467
0.0333
0.118
0.0500
0.243
4.41
- 0.302
- 0.705
- 0.292
- 0.538
- 0.522
- 0.0840
- 0.132
- 0.0990
- 0.344
- 4.55
note that only alkylated aromatics have
good response at 3.3 ym. Thus, there is
apparently no useful agreement in response
factors between, for example, a large
number of aromatic compounds and aliphatic
compounds (e. g., 50%, of those tested) at
analytical wavelengths specific to each
compound class* It is apparent that the
overlap of IR absorbance bands of different
functional groups is not sufficient to
yield one analytical wavelength which might
be used to quantify both compound classes
with the expectation of agreement within
a factor of five. This observation
Indicates that infrared-spectrophotbmetry
is hot particularly suitable for general
VOC screening.
On the other hand, the fact that the
response factors do not vary by large
values (i.e., greater than five) fbr some
classes of compounds (e.g., halbgenated
aliphatics at 13.5 ym and aliphatic and
alklyated aromatics at 3.3-3.4 ym) cor-
roborates the suitability of IR spectro-
photoinetry for VOC screening of compounds
belonging to one functional group. Even
in this case, only 30 to 80% of the com-
pounds in a given class may yield response
factors of less than five at a single
specific IR wavelength.
4.
5.
6.
VOC emissions of a single organic
functional group character.
IR screening of organic compounds
of a single functional class
(e.g., C-C1.) may be suitable for
as many as 80% of compounds in the
class.
IR screening at a wavelength cor-
responding to both aliphatic and
aromatic CH stretches may be suit-
able for as many as 30-50% of
organic compounds.
A portable PID is not currently
available for VOC screening in the
concentration range of 100 to
10,000 ppmv.
The development of" a reliable
dilution probe for use on a PID
is close'at hand.
With such a dilution probe, it
appears that a PID with an 11.7
or 11.8 eV UV lamp may be used
for reliable analysis of VOC
fugitive emissions.
LITERATURE CITED
CONCLUSIONS
In summary, based on the results of
this evaluation, it appears that:
1. IR spectrophotometry is not suit-
able for general VOC screening,
with the exception of -analysis of
1. Anastas, M.Y. and H.J. Belknap.
March 1980. Summary of available
portable VOC detection instruments.
EPA-340/1-80-010, U.S. Environmental
Protection Agency.
2. Brown, G.E., D.A. DuBose,
W.R. Phillips and G.E. Harris.
157
-------
January 1981. Response factors
of VOC analyzers calibrated with
methane for selected organic chemicals.
EPA-600/2-81-002 (OTIS PB81-136194),
U.S. Environmental Protection Agency.
3. Driscoll, J. June 1981.
HNU Systems, Inc. Personal
Communication *
4. Langhorst, M.L. February 1981.
Photoionization detector
sensitivity of organic compounds.
J. Chromat. Sci., 19:98-103.
5. Spain, D., J.J. Decorpo, and
J.R. Holtzclaw. August 1980.
Use of a photoionization detector
as a hydrocarbon trace gas analyzer.
Naval Research Laboratory, Memo
Report No. 4239.
6. U.S. Environmental Protection
Agency. September 1979. Measure-
ment of volatile organic compounds.
EPA-450/2-78-041 (NTIS PB80-221674).
7. U.S. Environmental Protection
Agency. January 5, 1981.
Method 21. Determination of
volatile organic compound leaks.
Proposed regulations. Federal
Register, 46.
158
-------
SURVEY METHODS FOR THE DETERMINATION
OF PRINCIPAL ORGANIC HAZARDOUS CONSTITUENTS (POHCs)
I. Methods for Laboratory Analysis
Ruby H. James, H. Kenneth Dillon, and Herbert C. Miller
Southern Research Institute
Birmingham, Alabama 35255
ABSTRACT
The U. S. Environmental Protection Agency (EPA) has established survey methods for
the analysis of a specific list of priority pollutants. There are potentially other
compounds whose presence or absence should be determined in analyzing incineration
streams. To assess the applicability of EPA methods or similar survey methods to the
determination of a broader range of principal organic hazardous constituents (POHCs),
we have evaluated gas chromatography/mass spectrometry (GC/MS) and high-performance
liquid chromatography (HPLC) methods for the determination of about 70 compounds (other
than priority pollutants) from the list of POHCs.
The methods evaluated are presented, analytical results are given, and the
potential applications of these survey methods are discussed.
INTRODUCTION
As part of the Resource Conservation
and Recovery Act (RCRA), the U. S.
Environmental Protection Agency (EPA) has
promulgated interim final and proposed
regulations for the owners and operators
of facilities that treat hazardous wastes
by incineration (4). The regulations
cover a range of activities including
operational performance standards, waste
analysis, trial burns, monitoring and
inspections, record keeping and reporting,
and1the establishment of emission control
criteria. The specific details for each
incinerator facility are authorized by
facility permits.
As part of the supporting
documentation for the permit writer and
for the incinerator facility owners and
operators, a manual of sampling and
analytical methods (5) has been compiled
by the EPA for use in measuring the levels
of principal organic hazardous constituents
(POHCs) in the various streams of an
incinerator facility, including inlet
wastes, stack gas, process waters, fly
ash, and bottom ash. This manual, referred
to as the "Methods Manual," expands upon
and augments the information in Reference
7. The Methods Manual is intended to be a
resource document for the preparation and
execution of a sampling and analysis plan
for hazardous waste incinerators. Existing
collections of sampling and analysis
methods documentation such'as References 1
and 6 have not been directly incorporated
into the Methods Manual, but are
incorporated by reference.
Although the Methods Manual includes
test procedures for proximate, survey,
159
-------
and directed (or specific-compound)
analysis, the emphasis of our work was
limited to directed analysis. The
directed-analysis portion of the waste
characterization scheme provides
qualitative confirmation of compound
identity and quantitative data with
appropriate quality control for the
potentially hazardous constituents that
might reasonably be expected to be present
in the waste, based on engineering judg-
ment and on the results of proximate and
survey analyses.
The Methods Manual recommends a
variety of directed-analysis techniques
for the determination of the list of POHCs
given in Appendix VIII, Part 261, 40 CFR.
Whenever documentation was available to
support the practice, the methods were
written to incorporate a high-resolution
analytical technique—fused-silica
capillary gas chromatography—and a highly
specific detection technique—mass
spectrometry. Also, whenever possible,
high-performance liquid chromatography
(HPLC) was recommended for the
determination of compounds that could not
be determined by gas chromatography with
mass spectrometric detection (GC/MS).
Nevertheless, there were many compounds
for which limited documentation was
available to support the recommendation of
the use of either GC/MS or HPLC.
Consequently, on the basis of the
demonstrated need to reduce the number and
complexity of analytical methods in the
Methods Manual, we developed generalized
GC/MS and HPLC techniques for the determi-
nation of as many as possible of the
chemicals on the list of hazardous wastes.
In the effort described in the subsequent
sections of this manuscript, we
concentrated on about 70 POHCs.
•ionization detection (FID)—in addition to
MS—was employed to aid in establishing
operating,conditions.
We gave no consideration to the .
modification of sampling procedures
or sample-preparation .procedures
reported in the Methods Manual. Changes
in these procedures were beyond the •
scope of the present assignment. The
investigation described here involved
the development,of instrumental
analytical methods using solutions of
the selected POHCs in solvents that
were .compatible with GC/MS or HPLC/UV
determinations. •
The laboratory work was structured
to lead systematically from the
determination .of the feasibility of
developing generalized test methods to the
calibration of the ensuing methods for
selected POHCs. First, we selected
(largely on the basis of commercial
availability) a variety of POHCs (from
Appendix VIII, Part -261, 40 CFR) for
preliminary investigation. Included in
this selection were a variety of compound
types including alcohols, esters,
chlorinated aliphatics and aromatics,
carboxylic acids and acid.anhydrides,
aliphatic and aromatic amines, nitrated
aromatics, nitrosamines, hydrazines,
nitriles, organosulfur compounds, and
polynuclear aromatics and heterocyclics.
We then analyzed standard solutions of
mixtures of these POHCs (in appropriate
solvents) to optimize instrumental
operating conditions. Once suitable
operating conditions had been established,
we analyzed a series of standard
solutions of each of the selected POHCs
to estimate detection limits and to
establish calibration curves.
TECHNICAL APPROACH
General Considerations
The focus of our experimental work
was to develop two generalized analytical
methods for the determination of POHCs in
appropriate organic solvents. One
analytical technique involved GC/MS with
the use of a capillary column, and the
other involved HPLC with reverse-phase
GIB columns and with ultraviolet spectro-
photometric detection (HPLC/UV). During
the optimization of the GC methods, flame
GC Analysis Procedures
Description of instruments and
general,operating conditions. We
developed the GC/MS generalized test
method on the Hewlett-Packard Model 5985
Gas Chromatograph-Mass Spectrometer-Data
System. The supplemental GC/FID work was
performed with a Hewlett-Packard Model
5840 Gas Chromatograph that was equipped
for use with papillary columns.
The work with both GC/MS and GC/FID
involved capillary chromatography with
matched, fused-silica SE-54 wall-coated
160
-------
capillary columns. The initial operating
conditions were chosen as a compromise of
the conditions given for several capillary
GC methods in the Methods Manual. The
initial starting column temperature was
40 °C; the temperature was- then programmed
at 10 °C/min to 280 °C and maintained at
280 °C for 15 min. The carrier gas,
helium, was maintained at' a volume flow
rate through the column of about
2 mL/min. In the GC/FID work, the
carrier gas flow was split downstream
from the-'injection port in the con-
ventional manner at a ratio of about 1
part to the column for every 40 parts
vented. Thus, only a few percent of an
injected sample was actually passed onto
the column. In the GC/MS work, the
"splitless" injection technique was
employed. Consequently, we assumed that
essentially all of the injected sample
reached the column in GC/MS determi-
nations.
Optimization and calibration of
GC/FID procedure. GC operating conditions
were optimized by analyzing methylene
chloride solutions containing a variety
of the candidate POHCs by the GC/FID
technique. The column head pressure was
adjusted to maximize the FID response
to test mixtures. We used these
adjustments to pinpoint the optimum
carrier gas velocity, which in turn
defined the splitting ratio.
Once we had established the optimum
operating conditions, we calibrated the
GC/FID procedure with external standards
and prepared a five-point calibration
curve for each of the POHCs investigated
by GC. Each curve was a plot of the FID
response (in peak area counts) as a
function of the quantity of the particular
POHC on the GC column. The detection
limit for each candidate POHC was
estimated as the quantity of the POHC on
the GC column that gave rise to a signal
approximately twice the background noise
level.
Optimization and calibration of GC/MS
procedure. Having established GC
operating conditions by the GC/FID
procedure, we then applied the method to
the determination of the candidate POHCs
by GC/MS. We also established reference
mass spectra for the identification of
the individual POHCs. The mass
spectrometer was operated in a full-mass-
scanning range (41 to 350 or 450 amu) in
the electron impact (El) mode. The scan
time was maintained at _
-------
we found that a few of the candidate POHCs
either were not retained by the Perkin-
Elmer column or did not chromatograph well
on the Perkin-Elmer column.
Rather than attempt to establish one
rigid set of HPLC operating conditions for
the method, our strategy in the develop-
ment of a generalized HPLC method was
to identify various procedural options
that would allow determinations of a
broad range of compound types. Conse-
quently, we investigated the use of
numerous isocratic and gradient elution
programs with the acetonitrile/water
mobile phase. In the determination of
several POHCs (including the phenoxy-
acetic acid), the eluent was acidified.
The wavelength of UV detection was also
varied as required to optimize
sensitivity. A Gary Model 17 Spectro-
photometer was used to establish an
absorption maximum in the range of 190
to 600 nm that would be suitable for
the quantification of each POHC by the
UV detector of the HPLC instrument.
As in the GC investigations, we
determined the precision of the method
with replicate injections of standard
solutions of the analytes; however, an
internal standard was not used in these
determinations. The precision of analysis
was determined for only a representative
group of the compounds investigated by
HPLC.
RESULTS AND DISCUSSION
GC Analysis
Table 1 presents the GC/FID results.
The retention time and on-column
detection limit are given for each
compound. The compounds are listed in the
order of their elution from the GC column.
Retention times are relative to that
observed for the internal standard,
dio-anthracene. The on-column detection
limit is the quantity of each analyte that
was estimated to yield an FID response of
twice the background noise; three-fourths
of the detection limits were in the
subnanogram range.
A chromatogram of a mixture of all
of the candidate POHCs investigated by
GC/FID is presented in Figure 1. The
chromatogram demonstrates the observed
absolute retention times, the peak shapes,
and any shifts in the base line that
occurred. Although two of the candidate
POHCs may not coexist in a field situation
and therefore may not require simultaneous
determination, the chromatogram
demonstrates the resolving power of the
capillary column.
Each reference calibration curve was
a point-to-point plot of FID response as
a function of the quantity of the analyte
on the GC column, which was calculated
from the quantity injected and the split
ratio. Each plot can be used to estimate
the sensitivity expected in the
determination of a particular analyte. In
generating these curves, we made no
attempt to determine the upper limit of
the linear range of determinations;
however, each plot demonstrated the
linearity of response within the range
of quantities investigated. Linear
regression analysis of the curves
typically yielded correlation coefficients
of 21.0.999. A typical curve (the curve for
N-nitrosopyrrolidine) is presented as
Figure 2.
Table 2 summarizes the GC/MS
determinations. As in Table 1, the
retention times are given relative to the
internal standard, dio-anthracene.
Table 2 also lists the five most abundant
mass fragments of each compound and,
moreover, specifies the mass of the ion
of each compound that was used for the
establishment of detection limits and for
the generation of calibration curves. The
detection limits are the quantities of the
analytes that were estimated to yield ion
currents '(of the selected ions) corre-
sponding to about twice the background ion
currents. Typical values of the detection
limits were 1 to 5 ng.
Reference ion chromatograms,
calibration curves, and mass spectra were
generated for each of the compounds
listed in Table 2. A chromatogram for the
determination of a mixture of all of the
compounds investigated by GC/MS is
presented in Figure 3.
Several compounds were determined by
GC/FID but were not determined by GC/MS.
These compounds—methylhydrazine,
dimethylhydrazine, and ethylenediamine—
were volatile enough to be partially swept
from the splitless injector of the GC/MS
162
-------
TABLE 1. SUMMARY OF GC/FID DETERMINATIONS OF CANDIDATE POHCs
Compound
Methylhydraz ine
1 , 2-Dimethylhydrazine
Ethylenediamine
IJ-Nitroso-N-methylethylamine
Malononitrile
Dimethyl sulfate
l,3-Dichloro-2-propanol
IJ-Nitrosodiethylamine
tl-Nitroso-N-methylurethane
Ethyl methanesulfonate"
Pentachloroethane
Benzyl chloride
N-Nitrosopyrrolidine
2 , 6-Dichlorophenol
4-Chloroaniline
Mh-Nitrosodibutylamine
Safrole
-3 , 4-Diaminotoluene
1,2,4, 5-Tetrachlorobenz ene
Nicotine
2,4, 5-Trichlorophenol
2 , 6 -Diamino toluene
2, 4-Diamino toluene
1 , 4-Naphthoquinone
£-Dinitrobenzene
m-Dinitrobenzene
Pentachlorobenzene
a-Naphthylamine
£-Nitroaniline
2-Acetamidofluorene
9 , 10-Dimethyl-l , 2-benzanthracene
Dibenz [a, j ] acridine
Relative
retention
time , min
0.06
0.07
0.08
0.18
0.20-
0.21
0.23
0.24
0.27
0.28
0.30
0.34
0.41
0.53
0.53
0.59
0.61
0.63
0.64
0.66
0.67
0.70
0.70
0.72
0.73
0.75
0.81
0.82
0.88
1.38
1.60
1.93 '•'"-
On-column
detection
limit, b ng
0.45
3.8
0.43
0.58 •
0.62
4.9
0.60
0.18
2.2.
0.70
1.1
0.20
0.22
1.5
0.20
0.21
0.26
0.39
0.49
0.35
1.7
0.11
1.7
0.44
0.53
0.12
0.67
0.23
"• 0.48
0.37
1.4
0.82
Relative to the retention time of d!0-anthracene, 16 min.
Quantity required to yield a response twice the magnitude
of the background signal.
163
-------
13
+14
27
29
10
4
+5
1
+2
JLi
6
+7
12
1618
21
+22
15
L9
2(
23
+24
25
26
28
31
JJLLu
30
10
15
20
25
30
35
RETENTION TIME, min
Figure 1. Chromatogrsm of all candidate POHCs by GC/FID. Peak 1: methylhydrazine; 2: 12-dtmethylhydrazine; 3: N-nitroso-N.-
nathylethylimine; 4: malononitrile; 5: dimethyl sulfate; 6: 1 £-dichloro-2-propanol; 7: fJ-fiitrosodiethy/amine; 8: fH-nitroso-N.-methy/-
unthina; 9: ethyl methaneailfonata; 10: pentachloroethane; } 1: benzyl chloride; 12: N'nitrosopyrrolidine; 13: 2,6-dichlorophenol;
14: 4-cbionaniHne; IS: N-nitrosodibutylamine; 16: safrote; 17: 3,4Jiaminoto/uene; 13: 1 £,4,B-tetrachlorobamane; 19: nicotine;
20: 2,4.5-trichtorophenol; 21: 2,&
-------
2400
2200
2990
1800
1600
1408
1200
1900
390
690
490
209
'» 29 30 «9 ii ta
OMMnTVONCOLUMH.ni
2 Calibration curve for N-nitrosopyrrolidine by GC/MS.
i
32
4 S •*./• * 9/101112 13 14 1516 1713 19292122232^262728 29 33 31 32> 3334
803-130
Figun 3. Chromatogram of all candidate POHCs by GC/MS.
165
-------
i—l en cr\ CM oo in iH CM p r^» vo
-»
6 I-t 1-1 r-t CM
O B
•r) CO
•a
•rt 3 VO »-% CO r~ O Op vp CM vp OO CO Op i-l CM CM CM tH CO vo ON vo O CM CM CT\ -
-------
system along with the solvent and thus
were not determined. It is likely that
these compounds could .be determined, with
some loss in sensitivity, by GC/MS with
split injection techniques.
Some other compounds investigated
could not be determined reliably by GC/FID
or GC/MS. These included maleic
anhydride, thiophenol, cyclophosphamide,
and ^-toluidine. Apparently these
compounds either were irreversibly sorbed
by the GC column or were decomposed in
solution, on the column, or in the
injection port.
Table 3 presents the results of
triplicate GC/FID determinations of
approximately 0.05-yg quantities of all
of the candidate POHCs listed in Table 1
except the hydrazines. The calculated
values of the standard deviation (SD) and
relative standard deviation (RSD) in
Table 3 indicate that most GC/FID
determinations were very precise. RSDs
greater than about 5% were, however,
obtained for several of the compounds—
11-nitrosodiethylamine, II-nitroso-N^
methylurethane, ethyl methanesulfonate,
pentachloroethane, benzyl chloride,
2,6-dichlorophenol, and 1,2,4,5-tetra-
chlorobenzene. The less precise
determinations were the result of
anomalously low responses obtained with
the first in the series of three
determinations for each of these
compounds. Perhaps at least one injection
was required to condition the GC column
and thereby to prevent loss of the
compounds in subsequent injections.
Table 4 presents the results of
triplicate GC/MS determinations of
about O.l-yg quantities of all of the
POHCs listed in Table 2 except malono-
nitrile. In general, the GC/MS
determinations were less precise than the
GC/FID determinations. Nevertheless,
about two-thirds of the compounds were
determined with an RSD of <5%. Several of
the compounds for which we obtained
relatively imprecise determinations by
GC/FID were also found to yield RSDs of
>5% by GC/MS; these included K[-nitroso-
diethylamine, pentachloroethane, and
benzyl chloride.
HPLC Analysis
Six procedural options were included
in the generalized HPLC/UV method. Three
were formulated for the Perkin-Elmer
reverse-phase C\e column; three for the
Waters column. The options for both
columns were either variations of the
isocratic composition of the acetonitrile/
water mobile phase or variations of the
solvent program. The various procedures
for the Perkin-Elmer column were as
follows:
• Option 1A
Solvent A: Distilled, deionized
water
Solvent B: Acetonitrile
Solvent program: 10% B, 5 min;
10 to 100% B in
35 min; 100% B,
10 min
Solvent flow rate:, 1 mL/min
• Option IB
Solvent A: 1% (v/v) acetic acid
in distilled,
deionized water
Solvent B: Acetonitrile
Solvent program: 20% B, 10 min;
20 to 50% B in
10 min; 50% B,
5 min.
Solvent flow rate: 2 mL/min
• Option 1C
Solvent A: 1% (v/v) acetic acid
in distilled,
deionized water
Solvent B: Acetonitrile
Solvent program: 10% B, 2 min;
10 to 100% B in
18 min
Solvent flow rate: 2 mL/min
The various procedures for the Waters
column were as follows:
• Option 2A
Solvent A: Distilled, deionized
water
Solvent B: Acetonitrile
Solvent program: 2% B, isocratic
Solvent flow rate: 1 mL/min
167
-------
r
TABLE 3. PRECISION OF GC/FID DETERMINATIONS
Relative GC/FID
response2
Compound
E thylened iamine
fl[-Nitroso-N_-inethylethylamine
Malononitrile
Dimethyl sulfate
l,3-D±chloro— 2-propanol
|l-Nitrosodiethy Iamine
jtf-Nitroso-IT-methylurethane
Ethyl methanesulfonate
Fentachloroethane
Benzyl chloride
tl-Nitrosopyrrolidine
2 , 6-Dichlorophenol
' 4-Chloroaniline
N_-Nitrosodibutylamine
Safrole
3 , 4 -D iamino to luene
1,2,4,5-Tetrachlorobenzene
Nicotine
2,4, 5-Trichlorophenol
2 , 6 -D iamino to luene
2 , 4-Diamino to luene
1 , 4-Naphthoquinone
£-Dinitrobenzene
m-Dinitrob enz ene
Pentachlorobenzene
a-Naphthylamine
jg_-Nitroaniline
2-Acetamido f luor ene
9 , 10-Dimethyl-l , 2-benzanthracene
Dibenz [a, j ] acridine
Detnb
No. 1
19.2
1.70
8.34
16.4
3.44
1.24
5.66
3.12
3.94
1.00
1.06
2.11
1.23
1.09
1.32
1.06
2.11
1.35
2.24
1.27
1.07
2.47
2.29
1.86
2.41
0.86
1.31
1.43
0.927
1.37
Detn
No. 2
19.6
1.75
9.12
16.3
3.76
1.45
6.16
3.81
4.47
1.12
1.15
2.30
1.32
1.17
1.35
1.07
2.39
1.38
2.29
1.27
1.04
2.61
2.29
1.84
2.66
0.88
1.27
1.44
0.927
1.35
Detn
No. 3
18.5
1.61
8.72
15.3
3.79
1.27
6.41
3.62
4.41
1.12
1.07
2.40
1.23
1.11
1.38
1.04
2.39
1.36
2.23
1.28
1.06
2.33
2.28
1.83
2.67
0.89
1.28
1.48
0.920
1.36
Mean
19.1
1.69
8.73
16.0
3.66
1.32
6.08
3.52
4.27
1.08
1.09
2.27
1.26
1.12
1.35
1.06
2.30
1.36
2.25
1.27
1.06
2.47
2.29
1.84
2.58
0.88
1.29
1.45
0.924
1.36
SD
0.6
0.07
0.39
0.6
0.19
0.11
0.38
0.36
0.29
0.07
0.05
0.14
0.05
0.04
0.03
0.02
0.16
0.02
0.03
0.01
0.01
0.14
0.002
0.01
0.15
0.02
0.02
0.02
0.004
0.01
RSD, %
2.9
4.2
4.5
3.9
5.3
8.5
6.3
10.2
6.9
6.5
4.4
6,4
4.2
4.0
2.0
1.7
7.0
1.1
1.3
0.6
1.2
5.6
0.1
0.7
5.7
1.9
1.6
1.7
0.4
0.8
(Peak area of internal standard)/(Peak area of POHC)
Detn ^ determination.
168
-------
TABLE 4. PRECISION OF GC/MS DETERMINATIONS
Relative GC/MS
a
response
Compound
N^Nitroso-iN-methylethylamine
Dimethyl sulfate
1 , 3-Dichloro-2-propanol
EJ-Nitrosodiethylamine
N-Nitroso-N-methylurethane
Ethyl methanesulfonate
Pentachloroethane
Benzyl chloride
N-Nitrosopyrrolidine
2 , 6-Dichlorophenol
4-Chloroaniline
tl-Nitrosodibutylamine
Safrole
3 , 4-Diaminotoluene
1,2,4, 5-Tetrachlor obenzene
Nicotine
2,4, 5-Trichlorophenol
2 , 6-Diaminotoluene
2 , 4-Diaminotoluene
1 , 4-Naphthoquinone
_p_-Dinitrobenzene
m-Dinitrobenzene
Pentachlor obenzene
ct-Naphthylamine
£-Nitroaniline
2-Acetamidof luorene
9 , 10-Dimethyl-l , 2-benzanthracene
Dibenz [a, j ] acridine
Detnb
No. 1
7.39
4.31
4.76
3.66
7.45
3.18
6.19
1.43
3.04
3.51
1.98
2.58
4.91
5.12
3.81
. 2.47
2.62
4.70
2.50
2.59
11.8
5.36
3.81
1.14
6.51
4.81
23.4
3.12
Detn
No. 2
7.71
4.60
4.79
3.45
7.56
3.14
5.33
1.77
3.35
3.64
1.87
2.92
5.08
5.32
3.68
2.87
2.54
4.80
2.50
2.77
12.0
4.82
4.03
1.12
5.76
5.59
24.6
3.35
Detn
No. 3
7.47
4.08
4.41
4.01
7.00
2.97
5.27
1.77
3.60
3.42
1.95
3.11
5.12
5.49
3.44
2.40
2.44
4.81
2.58
2.68
12.2
5.08
3.96
1.16
6.02
5.39
32.2
3.75
Mean
7.52
4.33
4.65
3.71
7.34
3.10
5.60
1.66
3.33
3.52
1.93
2.87
5.04
5.31
3.64
2.58
2.54
4.77
2.52
2.68
12.0
5.09
3.93
1.14
6.10
5.26
26.7
3.41
SD
0.17
0.26
0.21
0.28
0.29
0.11
0.51
0.20
0.28
0.11
0.06
0.27
0.11
0.19
0.19
0.25
0.10
0.06
0.05
0.09
0.2
0.27
0.11
0.02
0.38
0.40
4.8
0.32
RSD, %
2.2
6.0
4.5
7.6
4.0
3.6
9.2
11.8
8.4
3.2
3.1
9.2
2.2
3.5
5.1
9.8
3.6
1.3
1.8
3.3
1.4
5.2
2.9
1.6
6.2
7.7
17.9
9.2
(Peak area of internal standard)/(Peak area of POHC)
Detn = determination.
169
-------
• Option 2B
Solvent A: Distilled, deionized
water
Solvent B: Acetonitrile
Solvent program: 10% B, isocratic
Solvent flow rate: 1 mL/min
• Option 2C
Solvent A: Distilled, deionized
water
Solvent B: Acetonitrile
Solvent program: 20% to 100% B in
20 min; 100% B,
10 min
Solvent flow rate: 1 mL/min
In Table 5, the option used in the
determination of each compound
investigated by HPLC/UV is specified.
Two-thirds of the listed compounds were
determined by Option 1A. Option IB,
which involved the acidification of the
eluent, was instituted primarily for the
determination of the phenoxyacetic acids.
We observed that 4,6-dinitro-p_-cresol and
methyl yellow chromatographed very poorly
without the inclusion of acid in the
eluent; thus, Option 1C was established
for the determination of these two
compounds. Option 2A was suitable for
the determination of 11-nitroso-ll-
methylurea; Option 2B for saccharin.,
(Option 2C was used for the determination
of other compounds discussed later. See
Table 6.) For some of the compounds, it
is likely that another option—either one
of the other options listed here or a new
set of operating conditions—would have
given results comparable to those
presented in Table 5.
The application of the specified
procedures yielded the retention times and
detection limits given in Table 5. For
most of the compounds listed, the initial
determinations were made at a detector
wavelength of 254 nm, the wavelength of
maximum absorbance for the phenyl
functional group. After we had selected
optimum wavelengths for analysis, we
redetermined some of the compounds to
establish lower detection limits. The
limits established for the optimum wave-
lengths were typically <10 ng and usually
at least a factor of 10 lower than the
limits at 254 nm.
Reference chromatograms, UV spectra,
and calibration curves were generated for
the compounds listed in Table 5. The
calibration curves typically yielded
linear-least-squares correlation
coefficients of X).999. For some
exceptions, such as phenol, we obtained
calibration curves that were nonlinear.
Because these curves did appear to be
useful in determining the substances with
reasonable accuracy, however, we did not
consider it necessary to determine these
substances on other HPLC columns or with
alternative solvent programs.
In addition to the compounds cited
in Table 5, we found others that may be
determined by the generalized HPLC/UV
test procedure. Because of time
constraints, the investigation of these
substances was not completed; neverthe-
less , several compounds are presented in
Table 6 as potential candidates for
analysis by the HPLC procedure. Retention
times and approximate detection limits
were determined for the compounds on the
basis of a limited number of injections
of standard solutions. The retention
times of several of the compounds—
thiourea, thioacetamide, and ethylene
thiourea—were uncertain because several
major peaks were observed in their
chromatograms. The presence of more than
one major peak in a chromatogram was
interpreted as an indication of gross
contamination of the sample or as an .
indication of the decomposition of the
analyte on the HPLC column. Thus, the
feasibility of the application of the
method to the determination of these four
compounds is uncertain.
The precision of HPLC determinations
for a representative group of the
candidate POHCs is demonstrated by the
data in Table 7. All but two of the
triplicate sets of results were very
precise. The precision of the
determinations of acetophenetidin was
biased by one value that was 15% higher
than the other two. Perhaps additional
replicate determinations would have
proven the high result to be a statistical
outlier. The determinations of 5-nitro-o_-
toluidine yielded increasing responses
with each subsequent determination; such a
trend is usually indicative of column
conditioning. Perhaps the response would
have reached a stable value after repeated
injections.
170
-------
TABLE 5. SUMMARY OF HPLC/UV DETERMINATIONS OF CANDIDATE POHCs
ProceduraL
Compound option3
Streptozotocin
Phenol
4-Nitrophenol
o-Chlorophenol
Acetophenetidin
5-Nitro-o-toluidine
Tetramethylthiouram disulf ide
4-Chloro-m-cresol
2 , 4-Dichlorophenol
3-(a-Acetonylben2yl) -
4-hydroxycoumarin
2,4, 6-Trichlorophenol
2,3,4, 6-Tetrachlorophenol
Reserpine
Chlorambucil
2,4-Dichlorophenoxyacetic acid
2,4, 5-Trichlorophenoxyacetic
acid
2-( 2 , 4 , 5-Tr ichlorophenoxy) -
propionic acid
4 , 6-Dinitro-o-cresol
Methyl yellow
N-Nitroso-N_-methylurea
Saccharin
LA
1A
LA
1A
1A
LA
LA
1A
1A
LA
LA
LA
1A
1A
IB
IB
IB
1C
1C
2A
2B
Retention
time , min
1.4
5.4
9.5
12.4
12.6
14.3
16.3
16.8
17.6
19.8
,
20.0
21.5
22.7
23.9
7.6
14.2
16.5
7.6
13.1
1 8.4
3.2
On-column
detection
limit, ng
2
— —
78
—
54
6
.72
6
1
—
1
—
1
—
77
4
100
2
2
—
53
7
19
17
28
—
1
- — .
69
—
55
—
38
—
20
3
10
—
2
Wavelength of
detection, nm
254
c
230
254
f*
280
254
c
280
254
280C
254
C
248
254
253C
254
280C
254 .
280C
254
C*
. 280C
254
280
254
/•»
280
254
280
254
267°
254
r>
258
254
284°
254
287°
254
287C
378°
400C
254
234C
254c
224
See text for description of options.
Quantity injected that is required to yield a response twice the magnitude of
background signal.
CThis wavelength was selected from the referenced UV spectrum as the optimum
wavelength for analysis.
171
-------
TABLE 6. POTENTIAL CANDIDATES FOR ANALYSIS BY HPLC/UV
Compound
Trypan blue
Epinephrine
Thiosemicarbazide
Thiourea
Thioacetamide
Ethylene thiourea
Crotonaldehyde
Diethylstilbestrol
Mitomycin C
Melphalan
3 , 3 '-Dimethoxybenzidine
dihydrochloride
Daunomycin
Azaserine
Procedural
option3
2C
2C
2C
2C
2C
2C
2C
2C
1A
1A
1A
IB
2A
Retention
time, min
3
3
3
~3C
~4C
~4C
5
14
5
14
-19
8
4
Approximate
on-column
detection
limit, ng
20
60
5
6
2
8
1
4
17
10
, 9
75
2
Wavelength of
detection, nm
315
279
254
, 254
254,
254
230
240 ,
254
254
254
254
254
See text for description of options.
Quantity injected that should yield a response of 1000 counts on the integrator.
C '
The presence of several major peaks in the chromatogram made the assignment of
a retention time difficult.
TABLE 7. PRECISION OF HPLC DETERMINATIONS
Compound
Quantity
injected,
UK
HPLC/UV response^ '
area counts x 10 ?
Detna Detn Detn
No. 1 No. 2 No. 3 Mean
SD RSD, %
3-(ct-Acetonylbenzyl)- 9.39
4-hydroxycoumarin
Acetophenetidin 2.52
Chlorambucil 1.76
N^-Nitroso-N-methylurea 4.96
5~Nitro-p_-toluidine 2.48
Reserpine 7.50
Saccharin 0.98
Tetramethylthiuram 9.45
disulfide
895
1060
452
109
690
58.
43.
264
886
915
449
108
750
56.4
43.2
266
902
909
457
107
801
61.4
42.0
264
894
962
452
108,
747
58.8
42.7
264
8
85
4
1
56
2.5
0.7
0.9
8.9
0.9
0.8
7.4
4.2
•1.5
0.4
iDetn «• determination.
172
-------
SUMMARY AND CONCLUSIONS
A generalized GC/MS (and GC/FID) test
procedure and a generalized HPLC/UV
procedure have been developed for the
determination of approximately 70
candidate POHCs. Additional details of
the methods evaluation and reference
chromatograms, calibration curves, mass
spectra, and UV spectra are given in
Reference 2.
The candidate POHCs are of a variety
of compound types including alcohols,
esters, chlorinated aliphatics and
aromatics, carboxylic acids and acid
anhydrides, aliphatic and aromatic amines,
nitrated aromatics, nitrosamines,
hydrazines, nitriles, organosulfur
compounds, and polynuclear aromatics and
heterocyclics. The generalized procedures
are likely to be applicable to the
determination of compounds similar to
those tested.
The methods have been calibrated
for candidate POHCs over concentration
ranges of interest and have demonstrated
acceptable precision in the determination
of most of the POHCs. The ultimate
accuracy of the methods in the
determination of POHCs in incinerator
wastes and effluent will, of course,
depend on the prudent choice of
appropriate sampling procedures and
^ample-preparation procedures.
ACKNOWLEDGMENTS
This work was sponsored by the
U. S. Environmental Protection Agency
under Contract No. 68-02^-2685, Work
Assignment 111, Larry D. Johnson,
IERL-RTP, Project Officer.
REFERENCES
1. de Vera, E. R., B. P. Simmons,
R. D, Stephens, and D. L. Storm. 1980.
Samplers and Sampling Procedures for
Hazardous Waste Streams. EPA-600/
2-80-018 (NTIS PB 80-135353), U. S.
Environmental Protection Agency,
Cincinnati, Ohio.
2. Dillon, H. K., and R. H. James. 1981.
SATS Sampling and Analysis Methods:
Draft Report. EPA Contract 68-02-2685,
Work Assignment 111, U. S.
Environmental Protection Agency,
Research Triangle Park, North
Carolina.
3. Federal Register 44;69540-59; 1979.
4. Federal Register 46_:7666-90; 1981.
5. Rechsteiner, C., J. C. Harris,
K. E. Thrun, D. J. Sorlin, and
V. Grady. 1981. Sampling and Analysis
Methods for Hazardous Waste
Incineration: Draft Report. EPA
Contract 68-02-3111, Work Assignment
124, U. S. Environmental Protection
Agency, Research Triangle Park, North
Carolina.
6. U. S. Environmental Protection Agency.
1980. Test Methods for Evaluating
Solid Waste: Physical/Chemical
Methods. EPA Report No. SW-846, U. S.
Environmental Protection Agency,
Cincinnati, Ohio.
7. Vogel, G., K. Brooks, J. Cross,
I. Frankel, S. Huas, and W. Jacobsen.
1981. Guidance Manual for Evaluating
Permit Applications for the Operation
of Hazardous Waste Incinerator Units:
Draft Report. EPA Contract 68-01-6092,
U. S. Environmental Protection Agency,
Washington, D.C.
173
-------
NEW IDEAS IN HAZARDOUS WASTE MANAGEMENT TECHNOLOGY
Harry M. Freeman
California Office of Appropriate Technology
Sacramento, CA 9581*
ABSTRACT
This paper discusses the initial results of an EPA/State of California co-sponsored project to
evaluate emerging innovative technologies for managing hazardous wastes. Over 40 proposals
submitted by various waste management technology developers were evaluated by an Office of
Appropriate Technology (OAT) technical panel to select 12 for possible demonstration within
California. The selected proposals, which include five thermal, two biological and five physi-
cal/chemical processes, are briefly discussed.
INTRODUCTION
THERMAL PROCESSES
Under a cooperative agreement initiated
in March 1980 with the U.S. EPA, the California
Office of Appropriate Technology (OAT) has
begun a project to evaluate and possibly demon-
strate innovative processes for treating and
detoxifying hazardous wastes. The project is
part of a program supported by various State
agencies to encourage the detoxification, reduc-
tion, recycling, or destruction of hazardous
wastes as an alternative to landfills.
Processes for this project were obtained
primarily from a request for proposals which
was run in several prominent technical and
environmental magazines. A review panel con-
sisting of OAT staff members selected 12 pro-
cesses from the approximately 45 submitted.
The technologies chosen for further evaluation
are shown in Table I. Discussion of the tech-
nologies follows. The inclusion of these pro-
cesses in this paper is not intended to be an
endorsement of the processes by either OAT or
EPA. Neither OAT nor EPA, at this time, have
evaluated the claims made for the processes or
in any way determined their accuracy. This is
the purpose of the study which will follow.
Persons interested in learning more about these
technologies are urged to talk with the contact
person indicated.
Pyrolytic Incineration
Midland Ross Corporation
The system as described by the vendor
includes three components in sequence: a rotary
hearth furnace, a rich fume reactor, and a heat
recovery device. Waste is fed onto the rotary
hearth and held at 1000° to 1400°F for 15 to
30 minutes depending upon the type of waste.
As the waste is volatilized, the gaseous fraction
is passed to the reactor and is combusted to
complete the destruction of hazardous com-
pounds. The reactor is reported to operate at
temperatures in the range of 1800° to 3000°F
depending upon the toxicity of gaseous com-
ponents. Following combustion in the reactor,
the gaseous fraction is released to a boiler for
the extraction of heat energy.
The pyrolytic incineration process oper-
ates under quiescent atmosphere, so particulate
carry over is claimed to be lower than for the
conventional incineration process. Inorganic
chemical constituents containing sulfur, phos-
phorus, and halides may be retained in the char
because of lower pyrolysis temperature. Scrub-
bing equipment can be added, if need arises,
to remove chemical pollutants such as SO2,
174
-------
HC1, or ^2^5 to meet emission standards. The
process is commercially available.
Cement Kiln Co-Combustion
Systech Corporation
The process is to burn ignitable organic
wastes as a supplemental fuel in a cement kiln
at a location in Southern California. The proj-
ect is patterned on a similar operation being
run by the applicant at a plant in Ohio.
It is reported that at the Ohio plant,
selected waste materials, waste solvents and
industrial oils are procured directly from gener-
ators, transported to the facility using tank
trucks, and off-loaded at the receiving station
directly into the storage tanks located at the
cement manufacturing facility. Chemical anal-
yses are performed at the receiving station to
assure that the materials received comply with
the permit requirements, and to be sure that
the materials delivered to the cement kiln com-
ply with specifications designed to guarantee
that the cement manufacturing process is not
negatively impacted. The material is then
pumped to the cement kiln and fired in con-
junction with the standard fuel utilized in the
kiln.
No environmental emissions are expected
by the vendor from this process beyond the
normal products from combustion processes.
Particulate emissions, it is claimed, will remain
consistent with those from the fossil fuels nor-
mally used in the cement kiln, and will be
controlled by the existing air pollution control
equipment on the kilns. All the organics are
claimed to be completely destroyed in the kiln
due to the high temperature and long retention
times of the process. Metals or chlorine con-
tained in the waste products are said to be
incorporated in an essentially unleachable form
into the cement product or the dust. The
process may actually reduce the sulfur emission
from the cement kilns due to the fact that less
sulfur may be in the waste fuels than is present
in the coal that is replaced (coal is used in
95% of the kilns in the United States).
High Temperature Fluid Wall
Thagard Research Company
The Thagard HTFW process is a high tem-
perature process for quickly reducing organic
wastes to their elemental state. The reduction
is carried out in a patented reactor which
consists of a tubular core of porous refractory
material capable of emitting sufficient radiant
energy to activate reactants fed into the tubular
space. The core material is designed to be of
uniform porosity to allow the permeation of a
radiation-transparent gas through the core wall
into the interior. The core is completely jack-
eted and insulated in a fluid-tight pressure ves-
sel. Electrodes located in the annular space
between jacket and core provide the energy
required to heat the core to radiant tempera-
tures.
The principle, it is claimed, which dif-
ferentiates the HTFW Reactor from other high
temperature devices is the method of energy
transfer to the reactants. While the latter
employs convective or conductive techniques to
transfer energy to the Reactor feed, the HTFW
Reactor, according to claims, uses radiative
coupling to heat the reactants. The core is
heated by the external electrodes and its inner
surface re-radiates the energy into the tubular
space where the reactants are introduced.
Radiative power densities of over 1200 watts/
in^ are claimed, and the finely divided reactants
are heated by the direct impingement of elec-
tromagnetic radiation. The advantages of this
method of energy transfer are that the re-
actants are instantaneously heated, and the
chemical reactions are greatly accelerated
without the necessity for heating the entire
process stream to reaction temperature. The
rate of energy transfer is independent of either
the contact of the reactants with the Reactor
surfaces or of the flow regime of the reactants.
Another distinctive feature reported by
the builder for the HTFW Reactor is the pat-
ented fluid-wall. A gas, which is transparent
to the radiation (and therefore largely not
energy consuming), is introduced radially
through the porous walls of the Reactor to
produce an annular envelope of gas which blan-
kets the walls. This envelope reduces the
contact of the reactants with the Reactor sur^
faces to minimize deposition of reactants or
products on the radiator surface. This feature
offers an improvement over other gasifiers,
since the operational life of the Reactor com-
ponents is not. subject to slagging with its at-
tendant core wear.
Because energy transfer is accomplished
by radiation, heating of the reactants is claimed
to be very rapid — on the order of millions of
degrees per second for reactant surfaces. Nor-
mal operating temperatures are in the region
of 4000°F. Externally published literature does
not provide kinetic data and models for reac-
tions at these temperatures, and most data
175
-------
derived by Thagard is at the state-of-the art.
The vendor also claims evidence that in these
temperature regimes, the reactions differ in
kind, not simply in degree. However, empirical
data from several thousand hours of Thagard
testing shows that, in these ranges, the rate of
surface heating tends to preclude the formation
of intermediate compounds from partial re-
actions of feed material. Since these com-
pounds may contribute to downstream pollution
abatement requirements, their elimination
would assure a higher level of product purity
with reduced cost for downstream clean-up
equipment. Reaction times are almost instan-
taneous. Residence time for reactants can be
greatly reduced and reactor size can be mini-
mized. Consequently, capital costs lower than
systems based on lower-temperature processes
are claimed.
Catalytic, Low Temperature Fluidized Bed
Energy Incorporated
The incineration process is described by
the manufacturer as a catalytic, low tempera-
ture, fluidized bed process for the destruction
of toxic and hazardous wastes. The system
incorporates a dry exhaust gas clean-up method
for removal of the process by-products.
As presented, liquid waste is metered
(along with fuel oil when required) into an
injection tube which pneumatically injects the
waste near the bottom of a fluidized bed oper-
ating at a low temperature. The bed itself is
composed of a mixture of granular combustion
catalyst and limestone. Fluidizing air is forced
up through this bed at a high enough velocity
that the mixture behaves very similar to a
boiling liquid. During operation, limestone is
continuously added to the bed to replenish that
expended by reaction with combustion products.
Also during operation, bed material is periodi-
cally drained from the vessel to maintain the
appropriate bed height.
In the pilot plant, a single cyclone separa-
tor is used to remove paniculate matter from
the off-gas. In a deliverable full sized system,
it is likely that a multiclone or other more
efficient particulate removal system would be
necessary. During the past year, the El fluid-
ized bed incineration system has been demon-
strate^ for the destruction of chlorinated hydro-
carbons and organophosphates. The chlorinated
waste demonstration has used carbon tetra-
chloride as a representative aliphatic hydrocar-
bon and dichlorobenzene as a representative
aromatic hydrocarbon. Tributylphosphate has
been used to represent organo phosphate ma-
terials. The program for further waste types
to be evaluated includes phenols, chlorophenols,
hexacnlorobenzene, PCBs, and mixtures of these
wastes. Optimization of fuel (where necessary),
catalyst use, absorbent use, and system operat-
ing and design-parameters are not complete so
no complete process economics are now avail-
able. According to El, preliminary data suggest
that the technology may offer a considerable
economic advantage over other systems because
of low temperatures, reduced amount of fuel
oil where required, and less exotic materials of
construction.
Wet Air Oxidation
Zimpro, Inc.
Wet Air Oxidation is a process to oxidize
dissolved or suspended organic substances at
elevated temperatures and pressures. Water,
which makes ,up the bulk of the aqueous phase,
serves to catalyze the oxidation reactions so
that they proceed at relatively low tempera-
tures (350 F to 650°F) and at the same time
serves to moderate the oxidation rates removing
excess heat by evaporation. Water also provides
a heat transfer medium which, it is claimed,
enables the Wet Air Oxidation process to be
thermally self-sustaining with relatively low or-
ganic feed concentrations. The process is be-
lieved by the vendor to be most useful for
wastes too dilute to incinerate economically,
yet too toxic to treat biologically.
The oxygen required by the Wet Air Oxida-
tion reactions is provided by an oxygen-con-
taining gas, usually air, bubbled through the
liquid phase in a reactor used to contain the
process; thus the commonly used term "wet air
oxidation." The process pressure is maintained
at a level high enough to prevent excessive
evaporation of the liquid phase, generally be-
tween 300 and 3,000 psi.
As described by the manufacturer, a
wastewater stream containing oxidizable con-
taminants is pumped to the system pressure by
means of a positive displacement type pump.
The wastewater passes through a heat exchanger
which pre.heats the waste by indirect heat ex-
change with the hot oxidized effluent. The
temperature of the incoming feed is increased
to a level necessary to support the oxidation
reaction in the reactor vessel. Air and the
incoming liquid are injected into the reactor
where the oxidation begins to take place. As
oxidation progresses up through the reactor, the
heat of combustion is liberated, increasing the
176
-------
temperature of the reaction mixture. This heat
of oxidation is recovered by interchanging with
the incoming feed thus giving, in many in-
stances, thermally self-sustaining operation.
After energy removal, the oxidized effluent,
comprised mainly of water, carbon dioxide, and
nitrogen is reduced in pressure through a spe-
cially designed automatic control valve, accord-
ing to the manufacturer's designs.
BIOLOGICAL PROCESSES
Soil Enrichment
Veale Tract Farms
The applicant is an operating farm which
applies various inorganic industrial by-products
and wastes directly into the land to bring about
soil enrichment. Much of the material handled
by the process could be classified as hazardous
wastes if it were to be disposed of in a com-
mercial chemical landfill. Veale Tract Farms
is not classified as a waste disposal site. All
materials have been declassified as waste for
use as agricultural soil amendment, after ex-
tensive analysis and review. All materials bene-
ficial to agriculture are applied at recommended
rates, and are used in an environmentally safe
manner.
The applicant has conducted operations
since 1975 and is currently the only such opera-
tion in the United States. Some of the materials
utilized are limerock, paper pulp effluent, lime
sulfur slurry, ammonium sulfate, digested sew-
age sludges, lime sludge, and gypsum.
There are no direct emissions from the
operations. Indirectly, sodium and alkaline
water is leached from the cropland and drained
to a central location. The farm operates under
discharge requirements set by the California
Regional Water Quality Board. The operation
appears to offer a very attractive alternative
for at least some of the materials now going
into landfills.
Specialized Microbes For
In Situ Hazardous Materials Control
Polybac Corporation
The process involves the addition of mi-
crobes specially developed to break down highly
refractory toxic materials to conventional and
specialized biological treatment processes.
These microbes have been developed through a
process of "selective adaptation" through in-
creased exposure to model substrates, and sub-
sequent mutation to genetically fix the bio-
degradation enhancement. The wastes for which
this process can be used comprise a very broad
list, ranging from dilute aqueous solutions
through highly concentrated aqueous solutions,
and including semi-solid and solid materials.
Microbes have been developed for breaking down
extremely refractory compounds such as 2,4-D
and certain PCBs.
The biological process has been shown to
be less energy intensive and expensive than
other ultimate destruction technologies such as
oxidation or incineration. The process also
produces no secondary disposal problems for
toxic residuals.
The end product of biodecontamination is
carbon dioxide, water and cell mass. Where
small quantities of waste are being degraded
at a single site or where the decontamination
is occurring in the soil, the final residue (the
cell mass) is totally compatible with the natural
bioprocess and will be readily reassimilated into
the natural eco-system. Where very large
amounts of wastes are being degraded at a
single site, a substantial microbial residue will
develop which must be handled. This residue
can be handled through biological anaerobic
digestion with the concomitant production of
energy (methane gas). The final stable residue
from anaerobic digestion can be ultimately used
to re-establish the viability of lands which have
been made sterile by man's activities. This
would include stripped mine land reclamation,
reforestation, etc.
PHYSICAL/CHEMICAL PROCESSES
Ultra Violet Disinfection
Pure Water Systems, Inc.
The process as reported uses ultra-violet
light to accelerate the biological degradation
of organic wastes in dilute concentrations. Al-
though the mechanisms by which the UV process
detoxifies chemicals is still under investigation,
the reaction appears to be a chemical oxida-
tion/reduction reaction involving the generation
of ozone. Waste streams currently being studied
are halogenated aromatic compounds, chloro-
dioxins, copper wastes, and waste streams from
synthetic fuels production. The USDA pesticide
laboratory has carried out research on the pro-
cess and found that biological activity increased
177
-------
dramatically following pretreatment with UV.
The USDA has entered into an agreement with
the appplicant to construct some portable, 10
GPM, units for detoxifying waste lagoons.
These portable units could be made available
for a California demonstration.
Encapsulation
Environmental Protection Polymers
The process, according to the manufac-
turer, "fixes" hazardous waste for deposit in
the earth. It mixes participated contaminants
with polybutadiene, a material employed for
manufacturing rubber tires, and then thermosets
the mixture to form a block of waste. The
block is then encased in a seamless, 54-inch
thick jacket of polyethylene, a material used
In jacketing electrical cable. The process, it
is claimed, does not significantly increase the
volume of material requiring disposal, a factor
that could contribute greatly to the attractive-
ness of the process. The resulting product is
reported by the vendor to exhibit outstanding
resistance to water leaching, and chemical,
physical, and mechanical stresses. Additionally,
the process claims to offer the advantages of
"disposing" of the waste in a manner in which
it can be easily and safely retrieved.
Adsorption
Dr. Samia Fadl
The process is designed to remove haz-
ardous materials, mainly heavy metal ions, dis-
solved organics, phenol and cyanides from in-
dustrial sewage. The method uses carbon to
adsorb the contaminants, but rather than using
activated carbon, uses low rank lignite coals,
an abundant natural resource in North America.
Lignite coals cost approximately one tenth
the amount for activated carbon and have been
found by the applicant to be superior to acti-
vated carbon for selected contaminants. In
addition to clarifying effluent, the lignite might
also be useful as a source of energy when its
adsorbing capacity is exhausted. The process
has been tested at the bench scale using simula-
ted wastes and is at a stage for expansion to
a pilot scale.
Waste Oil Conversion
Soil Recovery, Inc.
The Soil Recovery Process is claimed to
be a method of converting oily wastes such as
those found in waste lagoons into an inert mass
which has several useful properties. The process
consists of mixing a specially adapted reagent
with the waste at the site of the pond. After
proper mixing, the reaction, which is exother-
mic, occurs and the waste is converted into a
dry, hydrophobic pulverized mass. Once the
waste has been converted into the solid, water-
repellant material, it can either remain on the
treatment site or be returned to the lagoon
from which it came.
Advantages claimed for the process are
1) waste can be treated on site; 2) the treated
material is environmentally safe; and 3) treated
material need not be disposed of in chemical
landfills. Process is claimed to be very price
competitive.
Encapsulation/Coagulation
Colloid Piepho
The process is described by the vendor to
utilize a one-step chemical encapsulation/coag-
ulation system followed by filtration dewatering
to clean oily and petrochemical wastewaters.
Typical wastestreams for which such a process
could be applicable include oily wastewaters
from the metal working industries, spray booths,
PCB waste from electrical transformers and
waste effluents from the computer chip in-
dustry.
According to the developer, a montmoril-
lonite based formulation is added directly to
oily wastewater and agitated vigorously using
a turbine mixer. The polymers disperse and as
the acids and bases dissolve, polymer-oil com-
plexes are formed and immediately covered by
sodium montmorillonite. Experience indicates
that most emulsions treat best if the pH is
first lowered. To do this, an organic acid is
chosen with a fast solubility rate and left un-
coated. Other acids and bases are coated to
prevent them from interfering with the lowering
of the pH at this point. After sufficient time
has been given for oil collection and sodium
montmorillonite covering, the pH is affected by
the slower dissolving acids/bases. Once again,
flocculation and montmorillonite covering oc-
curs. Another acid/base dissolves and the floc-
culation sequence occurs again until almost all
of the oil is fixated. The mixer is turned off
and conventional sludge settling and dewatering
can be done. The applicant has found that
most oily wastes can be effectively cleaned
with one of eleven formulations commercially
available.
178
-------
Operational results have demonstrated
that greater than 99% of its emulsified oil and
other dispersed contaminants are removed.
Clarified liquid is dischargeable to the sewer
and the sludge generated by the system is high
solids content and nonleachable.
SUMMARY
As is illustrated by the processes discussed
in this paper, there are either in commercial
existence or under development, many processes
that appear to offer promise as answers to
hazardous waste problems. It is simply no
longer necessary and certainly not desireable
to put highly toxic and persistent wastes into
landfills without treatment. There are many
options far more environmentally preferable.
There is presently underway within Cali-
fornia a program to phase out the disposal of
certain highly toxic wastes in landfills. This
program was initiated by a Governor's Executive
Order and seems to have broad support through-
out the State. We believe that it is only a
matter of time before similar approaches to
waste management, i.e. treatment rather than
burying, will be adopted in other industrial
states. We hope that the processes discussed
in this paper will accelerate these adoptions.
179
-------
TABLE 1. PROCESSES SELECTED FOR DEMONSTRATION
Applicant
Type of Process Waste Streams
Description
3.K. Shah
Midland Ross Corp.
900 N. Westwood
Toledo, OH 43696
(419) 537-6242
Arthur 3. Helmstetter
Systech Corporation
245 North Valley Road
Xenia, OH 45385
(513) 372-8077
Ed Matovich
Thagard Research Co.
2712 Kelvin Avenue
Irvine, CA 92714
(714) 523-2034
Roger D. Kuhl
Energy Incorporated
P.O. Box 736
Idaho Falls, ID 84301
(208) 529-1000
Phillip Schaefer
Zimpro, Inc.
Rothschild, WI 54474
(715) 359-7211
Brad Thomson
Veale Tract Farms
Knightsen, CA 94548
(415) 684-2193
Thomas G. Zitrides
Polybac Corporation
1251 South Cedar
Crest Blvd.
Allentown, PA 18103
(215) 433-1711
Incineration/
Pyrolysis
Mixed Wastes
Cement Kiln
Co-Combustion
Liquid
Combustible
Wastes
High Temperature
Pyrolysis
Organic Waste
Fluidized Bed
Incineration
Specific
Organic Waste
Wet Air
Oxidation
Organics in
Water
Soil
Beneficiation
Biological
Various Non-
toxic Indus-
trial Wastes
Toxic Organic
Wastes
180
A thermal oxidation process for li-
quids and sludges which combines
starved air pyrolysis with fume in-
cineration. The process requires no
participate clean-up and achieves
EPA required destruction efficien-
cies.
The process involves the co-com-
bustion of liquid wastes in a cement
kiln during the production of ce-
ment. Energy value of the waste
is recovered directly as heat to
make cement. Destruction effi-
ciencies for the wastes involved is
very high.
The process uses radiative heat
transfer in a proprietary reactor to
bring about temperatures of 4000°F
on the surface of waste particles.
Destruction is instantaneous and
complete with only combustion pro-
ducts remaining.
The incineration process is a cata-
lytic, low temperature, fluidized
bed system. The system incorpor-
ates a dry exhaust gas clean-up
method for the removal of process
emissions.
The process is a method of destruc-
tion and detoxification of dilute ha-
zardous wastewaters. The system
can be used as a means to recover
useable inorganics from streams
containing organics and inorganics
in wastewaters.
Various nontoxic solids and sludges
are disked into an operating farm
to improve the quality of the soil.
Wastes are approved by the State
and water run-off from the farm
closely monitored.
Microbes have been developed to
selectively break down toxic or-
ganic wastes such as 2,4-D; 2,4,5T;
and some PCBs. The microbes can
be added to conventional systems
and are added through specialized
treatment systems (CTX Treatment
Systems).
-------
Tom Creeden
Pure Water Systems,
Inc.
4 Edison Place
Fairfield, NJ 07006
(201) 575-8750
H.R. Luowitz
Environmental Protection
Polymers, Inc.
13414 Prairie Avenue
Hawthorne, CA '90250
(213) 675-3555
Dr. Samia Fadl
Simon Fraser
University
Burnaby, B.C. Canada
219-3573
Joel Shofel
Soil Recovery, Inc.
P.O. Box 2147
95 Madison Avenue
Mqrristown, NJ
(201) 540-0566
Rick Brinkman
Colloid Piepho
5100 Suf field Ct.
Skokie, IL 60077
(313) 966-5720
Ultfa-Violet
Encapsulation
Adsorption
Waste Oil •
Conversion
Encapsulation/
Coagulation
Organics in Water containing toxic organic
Water wastes is exposed to ultra-violet
light which accelerates the degra-
dation of halogenated aromatic
compounds by naturally occurring
bacteria. Process is most appli-
cable to detoxifying on-site la-
goons.
Toxic Non- The process is a technology using
Combustibles polymers to encapsulate extremely
toxic wastes. The wastes can be
either dry and unconfined, sludges,
or containerized. The end result
of the encapsulation process is the
isolation of the waste in question
for either retrievable storage or
burial.
Wastewater An adsorption technique for remov-
ing organics and metals from
wastewaters using low value lignite
coals. Spent coal from the process
could be used as an energy source.
•Organic The process involves adding a rea-
Sludges gent to sludgey wastes to produce
a dry, hydrophobic pulverized mass.
The resulting material which is sol-
id and water repellent can be used
as clean fill.
Contaminated A one-step chemical additive sys-
Aqueous tern for small volumes of emulsified
Streams and water insoluble wastes.
181
-------
RECOVERING METALS FROM METAL FINISHING WASTES
Alfred B. Craig, Jr.
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
This paper gives an overview of recovery methods used in metal finishing operations. It
provides some details on a pilot plant designed to recover metals from waste treatment
sludges and other manufacturing byproducts.
INTRODUCTION
Recycling of metals by industry is
practiced at significant levels. But this
practice is largely restricted to the pure,
solid, scrap metal. As an example, recycled
copper constitutes about 20 percent of the
total copper metal available in the market.
This is the scrap that needs simple melting
and refining. The scrap comes from two
sources: removal from capital goods such as
buildings, industrial machinery, etc., or
consumer goods such as automobiles, appli-
ances, etc. as well as scrap generated in
their manufacture.*
On the other hand, permanent metal
losses are also tremendous. Of all the
recorded production of copper in this coun-
try, 72.5 million short tons, about 60% are
traceable as being in use presently. Forty
percent or 29 million short tons (58 billion
pounds) are lost permanently due to wear,
abrasion, chemicals, fabrication, etc. At
today's prices, this loss amounts to 58
billion dollars figured at the price of
finished copper product. The loss is stag-
gering when figured in terms of the finished
capital and consumer goods.2
An example of this loss from chemical
and metal processing is provided, by the
electroless copper plating industry. The
printed circuit board industry comprises ten
percent of all the electroplaters in busi-
ness. They generated about a million gallons
of liquid waste (sludges) annually. Typ-
ically., one gallon of this sludge weighs about
ten pounds and contains ten percent copper,
so about a million pounds of copper are
contained in this sludge. Disposal of such
waste is not always carried out with future
recovery in mind, so a very significant
amount is lost each year.
Metal-containing wastes are not treated
for recovery for many reasons. Generally,
recovery is not an economical proposition in
terms of capital and operating costs, it is
an incompatible operation, or individually
the quantities generated are not large e-
nough to justify a recovery system when
evaluated simply on the economics of the
metals- themselves. The broader picture
which includes disposal and treatment costs
as well as liabilities merit additional
cons iderat ion.
Regulations
Federal regulation of pollution control
has been closely followed by state and local
governments issuing their own wastewater and
hazardous waste programs. These programs
can be, and often are, more stringent than
federal programs. Their basic goal has been
to protect the health and safety of the
general public by controlling industrial
discharges of all types to the environment.
This has been accomplished by increasing the
degree of complication and precaution re-
quired at each waste treatment or disposal
operation. This has usually come in the form
of more advanced technology, often at a
greater cost. Further, elimination of some
182
-------
Table 1. Economic Penalty for Losses of Plating Chemicals
Cost ($/lb)
Replacement Treatment3 Disposal*1
Nickel:
As NiS04
As NiCI,
Zinc cyanide as Zn(CN)2:
Using CI2 for cyanide oxida-
tion
Using NaOCI for cyanide oxida-
tion
Chromic acid as H2CrO4:
Using SO2 for chromium reduc-
tion
Using NaHSO3 for chromium
reduction
0.76
1.04
1.41
1.41
0.78
0.78
0.28
0.29
0.72
1.53
0.48
O.69
0.17
0.24
0.25
0.25
0.32
O.37
Total
1.21
1.57
2.38
3.19
1.58
1 79
Copper cyanide as Cu(CN)2:
Using CI2 for cyanide oxidation ...
Using NaOCI for cyanide oxida-
tion
Copper sulfate as CuSO.
1.95
1.95
0.56
0.72
1.53
0.28
0.25
0.25
0.17
2.92
3.73
1.01
a Based on chemical system described in Tables 2 and 3.
b. Based on Figure 1 at $0.10/gal sludge disposed.
Table 2, Chemical.and Sludge Disposal Cost:Example System'
a 4
Treatment chemicals Sludge disposal
Tntol
Treatment step
Chromium reduction ....
Neutralization:
Chrome effluent .
Cyanide effluent .
Acid/alkali waste
Precipitation
Flocculation
Total
Chemicals
Sludge disposal
Waste streams
30 gal/min = 1 800 gal/h
0.75 Ib/h Cr+6
0.15 Ib/h Cr+3
20 gal/min — 1 200 gal/h
0.80 Ib/h CN~
0.60 Ib/h Zn+2
1 800 gal/h
1 200 gal/h
60 gal/min = 3 600 gal/h
0.90 Ib/h Cr+3
0.60 Ib/h Zn+2
3.01 Ib/h Fe+2
2.41 Ib/h Ni+2
1 .51 Ib/h Cu+2
110 gal/min = 6 600 gal/h
Rates
(Ib/h)
1 .86 S02
0.36 H2S04
5.6 CI2
6.4 NaOH
2.7 NaOH
(b)
3.6 NaOH
2.1 NaOH
1.2 NaOH
6.0 NaOH
4.8 NaOH
3.0 NaOH
17.1 NaOH
0.66 polyelectrolyte
1 .8 SO2
0.36 H,SO.
5.6 CI2
29.8 NaOH
0.66 polyelectrolyte
Cost Dry solids
rates generated
($/h) (Ib/h)
0.18
0.93
0.22
0.29
0.17 1.80
0.10 0.91
0.48 4.83
0.38 3.80
0.24 2.30
1.37 13.64
0.66
3.65 1 3.64°
Disposal annual costs
cost ($)
($/h)
0.53
0.28
1.44
1.10
0.68
4.03
4.03
1 7,500
1 9,300
aSystem described in Table 3
bp"H adjustment not required
^Sludge volume at 4% solids=40 gal/h.
183
-------
Table 3. Investment Cost: Example System*
Component
Cpst
Chromium reduction unit (continuous system rated at 30 gal/min, from Figure 2)
Cyanide oxidation unit (continuous system rated at 20 gal/min, from Figure 3)....
Noutralizer(single-stage continuous system rated at 110 gal/min, from Figure 4).....
Flocculation/clarification unit (continuous system rated at 110 gal/min, from
Figure 5) ,.....'..
Polymer feed tank, mixer, and feed pump.
Sludge storage tank (5,000-gal tank to provide sludge storage volume, from Figure
6) - ....,.:.;.
Total equipment and installation cost
Contingency (10% of total equipment and installation cost).
Total installed cost
23,000
40,000
28,000
26,000
3,000
13.000
133,000
13,000
146,000
*References listed in table are irrelevant to
this report.
Table 4. Annual Cost Summary: Example System*
Component
B Hii, ' M ™
Operating labor (based on 2 h per shift)
*References listed in table are irrelevant -to
this report.
Tables 3 and 4 describe the capital and operating costs for an example
waste treatment system.
184
-------
100 i-
o
o
£
03
a
3
<
75
50
25
Legend:
disposal cost at $0.30/gal
disposal cost at $0.10/gal
0 25 50 75
SLUDGE VOLUME (gal/h)
Note.—Based on operating time of 3,000 h/yr. Costs in 1979 dollars.
100
Figure 1
Annual Cost for Sludge Disposal
Figure 1 shows what sludge disposal did cost, what it costs
now, or will cost in the future. The dotted line reflects
the loaded disposal cost for electroplating sludges disposed
in a secure landfill at $1.00/gal,
185
-------
treatment options has occurred because of
public opposition to landfills and inciner-
ators, etc. Today, landfilling of metal
bearing wastes is at best a difficult and
costly option. It begs the question 'of
liability if something should go wrong in the
future at the landfill site. The future may
show that while land disposal is today's most
widely used option, resource recovery and
associated detoxification is the optimum
approach.
Economics^
Do you know what it costs-?
The "whys" of pollution control have a
strong economic side. Pollution is really
process chemicals going down the drain.
Efforts to control discharges will provide
real dollar savings. Consider this, these
chemicals have been bought and paid for once
and now, as pollution, require additional
expenditures of capital for treatment equip-
ment and, of course, more chemicals. Fur-
ther, the cost of disposal of these process
and treatment chemicals carry a hefty dump
fee, not to mention any contingent liability
which may exist if the disposal facility
fails to maintain proper disposal standards.
How long will these facilities for land
storage of wastes exist? As they become more
scarce their users' costs will increase.
Table 1 describes what it costs to waste
chemicals in 1979 dollars.^ Capital goods
have increased about 20 percent since then-
and the cost of disposal has increased to
$0.50-$1.00/gal, far beyond the $0.10/gal
used as the basis for this chart, a summary
of Tables 2, 3, and 4 and Figure 1.
Table 2 describes what chemicals are
used for treatment and what they cost.^
Tables 3 and 4 describe the capital and
operating costs for an example waste treat-
ment system.41
Figure 1 shows what sludge disposal
used to cost, what it costs now or will cost
in the future. The dotted line reflects the
disposal cost for electroplating sludges
disposed in a secure landfill.
These charts show that pollution costs
all of us money. If the waste treatment
costs are tagged onto some profit centers,
some of them may turn out to be costing money
to operate. Their profitability may be
restored by cutting waste; that which re-
sults from inefficient operations and ex-
travagant wastes of process chemicals. This
is the subject of this paper—how to ac-
complish waste reduction and recovery from
metal finishing operations.
METALS RECOVERY
Do An Inventory
The word recovery indicates that some-
thing has been lost and in many metal fin-
ishing plants, indeed something has. The
quickest and surest way of conserving and
recovering metals is to prevent them from
becoming part of a mixed metal waste stream.
Recover them before they are lost as pol-
lution.
This recovery process must be started
with a good inventory of your plant and its
wastes generation characteristics. Review
each plating tank, rectifier, hoist, steam
coil or other piece of equipment; and do it
in a new suit and shoes. The place you are
most reluctant to go in your new clothes is
probably the area that needs the most atten-
tion. Determine the actual flow of each line
using the bucket and stop watch method.
Review your water bills to see if your total
water use ifnventory is consistent with what
you've been buying. Chemical purchases
representing a six-month period should be
studied carefully. This inventory of water
and chemical usage will be the beginning of
identifying where the metals are exiting
your process and entering the sewer. Deter-
mine how frequently dumping of process chem-
icals occurs and why' it occurs. Contam-
ination that could be eliminated may reduce
the associated pollutant load and maintain
more of the process chemicals in the process.
Equipment design and layout can be a
major source of pollution if one considers
how much chemical waste occurs if parts must
be carried across a plant between opera-
tions. Proper work patterns can restore
efficiency to an operation while further
minimizing 'floor spills and contamination.
If reorganizing the shop is not feasible,
consider letting the batch of parts remain
above the process for an extra 15 seconds to
allow additional draining. This keeps the
process solutions in the tanks; off the floor
and equally important, out of other pro-
cesses. Orienting pieces to facilitate
drainage will also aid this process.^
186
-------
Salvaging improperly prepared parts is
expensive from several perspectives. First,
the pollution has been generated during its
first preparation. Stripping or reversing
completed preparation processes will gener-
ate additional pollution. Then the piece has
to be reworked. Look at the cause of the
rework required. Is there a contaminant
causing this? If so, eliminate it at the
source and reduce the associated pollution
caused by rework and salvage. It is of major
consequence.
Conserve that which you have bought.
Conservation of these materials is a
major factor in controlling pollution. Keep
them out of the sewer. Many processes can be
operated at a lower concentration than con-
ventional wisdom would have you believe.
Experiment with this possibility and pursue
it with any and all knowledgeable people;
including your suppliers. Chemical sub-
stitutions play a role here. Will a chloride
bath do just as good a job as a cyanide bath?
If so, you've just conserved $10-40,000
worth of a cyanide waste treatment system.5
Water conservation will help conserve
other things as well. In plating, counter-
flow rinsing is a monumental step towards
conserving chemicals and water. The concept
can be applied to other uses of chemicals.
Basically, use a chemical more than once if
you can and then recycle it; e.g., use
sulfuric acid for steel pickling and then
reuse it to neutralize a caustic stream.
Flow restrictors can further reduce water
usage where counterflow rinses have been
installed or where room is insufficient' for
additional tanks. A three-stage counterflow
rinse can recover about 90 percent of the
chemicals which would be lost in a single
stage rinse. Spray and timed rinses can also
be of major consequence for water saving.
Only after rigid water and chemical
conservation at the point of generation is a
common practice, should one begin to con-
sider recovery equipment as part of the
process. The cost of any piece of treatment
equipment will be some function of its hy-
draulic capacity. The bigger the flow, the
greater the capital cost and floor space
requirement.
"It is extravagant and often futile to
consider any recovery process equip-
ment until the maximum flow reduction
has been achieved in each and every
operation that has recycling poten-
tial. The best and most reliable way
to specify any recovery method is by
measuring the flow and assiduously
monitoring and analyzing the waste-
water in concern. Without specific
data, equipment may be recommended
purely on speculation, thus increasing
the possibility of under- or over-
design. The latter is not a func-
tional disaster—just a waste of mon-
ey. Conversely, under-design will
spur financial and emotional problems
in that the recovery system will be
expected to perform in ways never
intended. Again, complete your water
conservation program and then consider
and select the appropriate treatment
technology for the specific applica-
tions in your plant.""
IN PROCESS RECOVERY SYSTEMS
Recovery equipment should be chosen
with the understanding that it can be a
money saver but that it is a piece of
process equipment just like any other chem-
ical system. It will not operate by itself
but must have some form of maintenance and
operational attention. Diminished waste
treatment and sludge disposal costs should
be factored into the equation as, you re-
call, we are going to look at the plating
line as a profit center and take into
account the total cost of its operation.
The least expensive and the most cost
effective recovery system is a simple drag-
out tank. The first rinse is a still rinse
which recovers much of the material from the
plating bath which is dragged out. If the
plating bath is above ambient temperature,
the drag out or dead rinse tank contents can
be used to restore the plating bath level.
Impurity control is important in
both tanks so extra makeup water purifica-
tion is essential.
Ion exchange uses polymeric resins
which remove metal ions from wastewater by
exchanging them for ions already in the
resin. These metal ions eventually sat-
urate the resins and they must be regen-
erated usually by replacing the metal ions
with sodium from brine solutions or hy-
drogen from acid solutions. Anions can also
be captured in resin columns of the appro-
priate material. They are restored with
caustic. These recovered solutions may be
added back to the plating tanks, however the
187
-------
r
volume tends to grow which limits its use to
hot baths or where evaporation can be used
to further decrease the amount of solution.
In electrodialysis—basically electro-
lytic dialysis—ions move through membranes
which only allow the passage of only either
anions or cations. By stacking these membrane
types alternately and applying a DC current,
the ions in the waste streams are forced to
migrate from the dilute-wastewater so as to
form a series of channels into collecting-
concentrating-channel , where it is available
for reuse. The greater the concentration, the
greater the efficiency of the system. Its
best applications are where dragout is high.
Reverse osmosis is used for the removal
of dissolved organic and inorganic materials
and control of such wastewater parameters as
soluble metals, IDS, and TOG. It separates
dissolved materials in solution by filtration
through a semipermeable membrane at a pres-
sure greater than the osmotic pressure caused
by the dissolved materials in the wastwater.
With existing membranes and equipment, oper-
ating pressures vary from atmospheric to
1,500 psi. Products from the process are (1)
the permeate or product stream with dissolved
material removed, and (2) concentrate stream
containing all removed material. Removal
levels obtainable are dependent on membrane
type, operating pressure, and the specific
pollutant of concern. Removal of multi-
charged cations and anions is normally very
high, while most low molecular weight dis-
solved organics are not removed or are only
partially removed.
Distillation is one of the simplest
methods of separating plating chemicals from
water. The distilled water is condensed in a
chiller. Distillation can be accomplished at
ambient temperatures but is less likely to
destroy heat sensitive chemicals if accom-
plished under a vacuum, i.e., at a lower
boiling point. Fabrication of equipment
should be made with considerations of the
corrosive environment caused by the materials
being recovered.7
Metal Recovery from Hydroxide Sludges**
Treatment of hydroxide sludges for metal
recoveries falls into two categories: wet and
dry. In wet processes, separation reactions
are carried out usually at atmospheric tem-
peratures and pressures. In dry processes,
separation reactions take place at elevated
temperatures. The former processes are gen-
erally classified as hydrometallurgy and the
latter as pyrometallurgy.
Pyrometallurgical treatments for seg-
regated single metal sludges are:
• Direct smelting after dewatering,
• Direct alloying with other metals.
For economical feasibility, the opera-
tion of pyrometallurgical processes would
have to be on a very large scale, and requires
concentrated feed material. The mixed metal
hydroxide sludges fail to meet these two
requirements. But, pyrometallurgical treat-
ment using the two above bulleted processes
are presently available.
Hydrometallurgical treatments for mixed
and segregated sludges use aqueous dissolu-
tion of metals by leaching followed by re-
covery of individual metals or metal com-
pounds from this solution.
There are certain advantages for using
hydrometallurgical processes. Metals may be
removed directly from the solution by con-
centration, electrolysis, or hydrogen re-
duction. Fuel requirements are low as the
processes are carried out at low tempera-
tures. Disadvantages are that solution puri-
fication and concentration upgrading may be
absolutely necessary before metal recoveries
are possible.
The hydrometallurgical treatment of the
hydroxide sludges requires that the process
be specific, reproducible, and controllable.9
The process should also be possible with a
minimum number of unit operations and/or unit
processes. Those needed depend on the chem-
ical state and also on the physical environ-
ment of the metals. Electroplating wastes,
after the hydroxide neutralization, will con-
sist mainly of metal hydroxides such as Cu-
(OH)2) Ni(OH)2, Cd(OH)2, Zn(OH)2, etc. With
lime neutralization, the sludge will also
contain calcium salts.
Solvents for these metallic constituents
are sulfuric acid, ferric sulfate, ammonia
and ammonium carbonate, sulfur dioxide, fer-
ric chloride, hydrochloric acid, and nitric
acid and are used to solubilize metal ions.
The order given is about that of their im-
portance. Metal recovery from these solu-
tions by such operations as cementation (co-
precipitation with or solution substitution
by another metal) or electrolysis is most
188
-------
commonly practiced from the metal sulfate
solution.
Sulfuric acid is the most important
leaching reagent. The main advantages for its
use are the cost, minor corrosion problems,
and dissolution of many metal forms. Its main
disadvantage is that it sulfatizes every-
thing, which does not provide any selectiv-
ity. Ferric sulfate may be obtained cheaply
from spent pickle liquors. Its primary func-
tion is to provide a sulfate solution. Am-
monia and ammonium carbonate may be suitable
reagents for some hydroxide sludges because
they possess better selectivity for solu-
bilizing metal constituents in a hydroxide
sludge. But the reagent cost is very high and
their recovery in the process would be man-
datory from an economical point of view.
The primary aim for a solvent is to bring
the metal or metals into solution from which
individual separations can provide pure solu-
tions, with major impurities removed.
Leaching of hydroxides by either sul-
furic acid or ammonia will provide dilute
solutions. Copper may be recovered from lean
sulfate solutions by cementation with iron,
displacing copper quantitatively from a solu-
tion. The product, called cement copper,
although, rich in copper, may be highly contam-
inated with iron so that only a smelter may be
able to handle it.
A technique which is becoming widely
accepted for concentrating solutions is one
of ion exchange. The ion exchange medium may
be solid or liquid. Generally speaking,
concentrating by solid organic resins is
known as ion exchange (IX) and by liquid
organic resins is known as solvent extraction
(SX). Ion exchange has tremendous applica-
tions in uranium metallurgy as the solvent
extraction technique is becoming accepted in-
copper metallurgy. Its acceptance is on such
a large scale that the future may see its
application in areas where the ore grade has
decreased or residual metallic wastes are to
be treated as in this case.
For sulfuric acid leach solutions, cop-
per may be recovered from the polymetal solu-
tion by cementation under oxidizing solution.
The cement copper may be sold directly to a
smelter or releached for electrolysis. Other
metals are selectively precipitated after
iron removal and may be recovered as pure
metals or metal compounds. Electrolysis
regenerates the sulfuric acid for recycle
which is a very important factor in economic
evaluation because regeneration is part of
the process requiring no additional or extra
regeneration equipment installation.
Copper may also be removed from the
polymetal sulfate solution by solvent extrac-
tion and strip copper then is removed by
electrolysis. Other metal recoveries may be
achieved by further solvent extraction or by
cementation with zinc.
If an ammonia leach is employed, metals
such as copper, nickel, and zinc are selec-
tively brought into solution as amines from
the hydroxide sludges. Other impurities such
as iron and calcium are left behind as solids.
Copper may be removed from this polymetal
solution by solvent extraction, stripping
with sulfuric acid, and electrowinning. Oth-
er metals are further treated for separation.
Additional reagent recovery equipment is nec-
essary.
Generally, there are three types of
liquid wastes in a plating shop: (1) rinse
waters, (2) spent baths, and (3) spills.
Spent baths are collected separately.
They require only cyanide destruction and
thus become concentrated metal solutions.
Rinse waters and spills are treated con-
ventionally by heavy metal precipitation.
This presorting provides two distinct feed
materials: (1) a concentrated solution, and
(2) a hydroxide sludge. The leach solution
from hydroxide sludge leaching with sulfuric
acid is mixed with the spent bath solution for
metal extraction.
Metals conservation and recovery has
economicj environmental, and socio- political
benefits. The recommendations made in this
paper will help to alleviate some of the
economic consequences of waste treatment,
reduce environmental impacts from this in~
dustry, and reduce the rate at which permanent
metal loss occurs.
189
-------
REFERENCES
1. Mehta, Anil, 1981 Proposal to EPA,
"Recovery of Metals from Metal Finish-
ing Waste," Unpublished.
2. Ibid.
3. Roy, Clarence, "Establishing a Program
for Pollution Control," excerpted
from Plating and Surface Finishing,
Vol. 68, No. 10, October and November
1981.
4. Control and Treatment Technology for
the Metal Finishing Industry: In
Plant Changes, U.S. EPA, January 1982
and Alternatives; Economics of Waste-
water Treatment for the Metal Finish-
ing Industry, U.S. EPA, June 1979.
5. Roy, Clarence, "Establishing a Pro-
gram for Pollution Control," excerpted
from Plating and Surface Finishing,
Vol. 68, No. 10, October and November
1981.
6. Ibid.
7. Ibid.
8. Excerpted from "Routes to the Recovery
of Metals from Hydroxide Sludges,"
Third Conference on Advanced Pollution
Control from the Metal Finishing
Industry, U.S. EPA, 1981.
9. Ibid.
10. Ibid.
190
-------
EPA MINING WASTE RESEARCH
S. JACKSON HUBBARD
NONFERROUS METALS AND MINERALS BRANCH
ENERGY POLLUTION CONTROL DIVISION
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY - CINCINNATI
USEPA
INTRODUCTION
The mining industry generates an es-
timated 2.3 billion tons of solid waste
each year. Historically, the total ac-
cumulated solid waste from both active
and inactive mine sites approaches 30
billion tons. Five segments of the mining
industry generate 85 percent of mine
solid waste: copper, phosphate, iron,
uranium, and coal.
In recent years, Congress has passed
three acts indicating increasing interest
in the proper disposal of solid waste
generated by the mining industry. The
Reource Conservation and Recovery Act
(RCRA) is the most comprehensive
regulatory authorization. It was passed
in 1976. RCRA is implemented and enforced
by the Environmental Protection Agency
(EPA) and is intended to: control municipal
and industrial solid wastes. The passage
of the Surface Mining Control and Re-
clamation Act (SMCRA) of 1977, imple-
mented and enforced by the Department of
Interior, controls coal surface mining
and the surface effects of underground
coal mining. The Uranium Mill Tailings
Radiation Control Act (UMTRCA) of 1978 is
implemented by EPA and enforced by the
Nuclear Regulatory Commission (NRC) to
control uranium mill tailings disposal.
RCRA
The Resource Conservation and Recov-
ery Act addresses the disposal of waste
from municipal or industrial sources.
Specific to mining, RCRA recognizes the
special problems associated with this
industry, and requires that a special
study be made. This study will be dis-
cussed later. Two types of mining waste
which are exempted from control under
RCRA are uranium mill tailings (covered
by UMTRCA) and overburden intended for
return to the mine site (except when
specifically included by EPA).
The basic regulatory philosophy
underlying RCRA is that wastes are one of
two types: either hazardous or nonhazardous,
and the regulatory scheme applicable to
each type is different. Hazardous wastes
are those currently listed by EPA in its
Section 3001 regulations or those which
fail specified tests for toxicity, corro-
sivity, reactivity, and/or ignitability.
Wastes which are not listed by EPA and
which do not fail the tests for these
four characteristics are presumed to be
nonhazardous.
Hazardous wastes are regulated under
a "cradle-to-grave" approach. The gener-
ator, transporter, storer, treater and
disposer are all subject to regulation.
The regulations provide a manifest system
to track the waste from the point of
generation to the point of ultimate dis-
posal. Specific standards are set for
transport, treatment, storage, and dis-
posal. In addition to the technical
standards, administrative standards cov-
ering recordkeeping, reporting, and fi-
nancial responsibility are also imposed.
Presently, EPA's RCRA Subtitle C. regu-
lations do not apply to mining wastes.
Congress amended RCRA on October 21, 1980
so that solid waste generated from the
"...extraction, beneficiation, and proc-
essing of ores and minerals..." are ex-
cluded from regulation under Subtitle C.
Subtitle C does however, have important
191
-------
implications for the raining industry.
While no mining wastes are included in
the RCRA control system now, EPA may
eventually bring these wastes under regu-
lation. It is important to note that the
RCRA mining wastes exclusion is not a
permanent exclusion. At the same time
Congress excluded mining waste from RCRA
subtitle C regulation, Congress directed
EPA to perform a special study of the
raining industry and waste generated from
mining operations. EPA can, if it deter-
mined it necessary, issue RCRA Subtitle C
regulations for the mining industry six
months after it completes this study.
The mining exclusion would end if and
when EPA issues Subtitle C regulations.
EPA views mining wastes as "special wastes.
That is, unlike many low volume, high
toxicity waste streams regulated under
RCRA (and for which Subtitle C of RCRA
was designed), most mining operations
produce high volume, low toxicity waste
streams. Obviously, EPA could never
apply its existing RCRA Subtitle C regula-
tions to these waste streams. If EPA
decided there was a need to control the
disposal of mining waste, the RCRA regula-
tions it would issue would be custom fit
to Che raining industry and the unique
nature of wastes generated from mining
operations.
SMCRA
The Surface Mining Control and Re-
clamation Act applies to coal surface
raining and the surface effects of under-
ground coal mining. Interim and permanent
regulatory programs have been promulgated
by the Department of the Interior to
implement the authority of that law. The
regulations include specific requirements
for the construction and operation of
mine waste disposal sites, including both
mining and processing wastes. The re-
quirements are intended to prevent the
formation of acid mine drainage and other
adverse environmental conditions, and
assure that the mine site is returned to
a condition as good or better than before
raining. Monitoring of surface water and
groundwater is provided in the regu-
lations. The Department of the Interior
requires the posting of a bond to provide
the funds to reclaim lands damaged by
raining operations, and has the power to
shut down mine sites which do not comply
with its regulations.
UMTRCA
The Uranium Mill Tailings Reclamation
and Control Act requires the promulgation
of standards for the disposal of both
radioactive and nonradioactive mill tail-
ings at both active and inactive sites.
The standards are to be promulgated by
EPA, but the Nuclear Regulatory Com-
mission (NRG) is the enforcement au-
thority, through its permitting procedure
for uranium mining operations.
SPECIAL STUDY - MINING WASTE
Authority
Section 8002(f) of RCRA requires
that EPA conduct a special study of the
wastes generated by the mining industry.
The requirements of this study are deline-
ated in the following excerpt from RCRA:
"The Administrator, in consultation
with the Secretary of the Interior,
shall conduct a detailed and compre-
hensive study on the adverse effects
of solid wastes from active and
abandoned surface and underground
mines on the environment, including,
but not limited to, the effects of
such wastes on humans, water, air,
health, welfare, and natural re-
sources, and on the adequacy of
means and measures currently employed
« by the mining industry, Government
' agencies, and others to dispose of
and utilize such adverse effects.
Such study shall include an analysis
'pf'.-,
(1) The sources and volume of
, discarded material
generated per year from
mining;
(2) present disposal practices;
(3) potential dangers to human
, , .health and the environment
from surface runoff of
leachate and air pollution
by dust;
(4) alternatives to current
disposal methods;
,(5) the cost of those alter-
natives in terms of the
impact on mine product
costs; and
(6) potential for use of dis-
carded material as
a secondary source of the
mine product."
192
-------
Scope of Study
EPA's Office of Solid Waste (OSW)
requested that the Office of Research and
Development (ORD) help it conduct the
mining waste study required by Congress.
Due to the time constraints imposed by
Congress and the lack of available data
concerning these wastes, ORD was forced
to approach the problem at two different
levels. Phase I consisted of an explor^-
atory investigation of 45 mining.sites, to
evaluate the solid waste disposal prac-
tices currently being utilized by the
mining industry. Phase II is comprised
of a presurvey which included collecting
mine waste samples from 65 active and
abandoned mines and detailed environmental
evaluations at eight of these sites.
Phase I
Phase I consisted of a survey of cur-
rent solid waste management practices.
The selection of the management practices
to be studied in Phase I was based on the
volume of solid wastes produced and their
potential for causing harm to the envi-
ronment .
A Mine Solid Waste Coordinating
Committee (MSWCC) was established (con-
sisting of representatives from the EPA
Regions, the Office of Solid Waste, Efflu-
ent Guidelines Division, the Office of
Research and Development, and other
Federal Agencies including the Bureau of
Mines, Geological Survey, Fish and Wild-
life Service and Soil Conservation Ser-
vice) to assist the Agency in selecting
the management practices to be evaluated
during the entire study. The American
Mining Congress and the Interstate Mining
Compact Commission also participated on
the commi'ttee in an advisory capacity.
This committee nominated candidate
mines that utilized solid waste management
practices that were representative of the
industry. EPA compiled the list and
selected forty five mines for final review
and selection.
Based on information obtained from
published literature, interviews with
industry representatives, reports and
research of Federal and state agencies,
onsite visits to the forty five mines,
and information supplied by EPA, a listing
of types of management practices for the
disposal of solid waste from mining and
beneficiation processes was derived.
Management practices from the following
industries were considered:
1. Metallic ore mining and ben-
eficiation (e.g., copper, lead,
zinc, molybdenum, gold, and
silver).
2. Phosphate ore mining and ben-
eficiation
3. Uranium mining
The evaluation of each practice was
based upon the following criteria: (1)
protection of public health and welfare,
(2) protection of the quality of ground
and surface waters from leachates, (3)
protection of the surface waters from
runoff, (4) protection of air quality,
e.g., from fugitive dust, (5) and aesthetr
ics.
Evaluation of each of the practices
was based on visual observation; dis<-
cussions with the mine operator and regular
tory agencies; and data available from
mine operators, literature, and regulatory
agencies. No sampling or analysis of
water, air, or solid waste was conducted.
The results from this phase of the
study provided input to Phase II and will
also be used in the preparation of the
final report for the entire study.
Phase-II - Presurey
Sixty five mines were visited during
the presurvey. These sites were selected
from a list of candidate sites suggested
by the Mine Solid Waste Coordinating
Committee and from the results of the
Phase I study.
During the presurvey, samples were
collected from active, inactive, and
abandoned waste management practices.
Depending on the specific site surveyed,
the following residuals were sampled and
analyzed: mine waste, which included
overburden, waste rock, and development
rock; tailings, including fresh tailings,
backfilled tailings, settled solids,
water, and dike material from tailings
ponds; low grade ore; heap leach material;
mine pumpout water and settled solids
from mine pumpout water; and stockpiled
193
-------
topsoil. In addition, when waste mate-
rials of a certain type v?ere segregated
for disposal by separate waste management
practices (e.g., backfilled and surface
impounded tailings), each portion was
sampled separately. The sampling pro-
tocol and the sampling equipment speci-
fied in the "Sampling and Analysis Manual",
prepared for this study and approved by
EPA, were used. Over 400 samples were
collected from the 65 mine sites during
this sampling effort.
A composite sample consisting of
five to ten subsamples and totalling
approximately ten pounds was collected
from each management practice sampled.
The subsamples were collected in a manner
which assured that the most representative
sample was obtained. Typically, samples
were taken along a transect or from ran-
domly selected areas of the disposal
site. There were exceptions made to the
sampling protocol due to physically lim-
iting factors such as extremely steep
slopes and flooded or unstable areas
which prevented the safe acquisition of
the samples. In these instances the
selection of an appropriate sampling
method was left to the discretion of the
sampler. Subsamples were taken using an
aluminum hand scoop from a depth of one
foot; however, when physical limitations,
such a frozen ground, prohibited sampling
at this depth, the sample was taken from
a more shallow excavation. Material that
was too large to sample in this manner
was reduced to a more manageable size
with a crack hammer. Samples were placed
in appropriately labeled one-gallon poly-
ethylene containers for subsequent ship-
ment.
Samples of fresh tailings were taken
using a one-liter polyethylene bottle
attached to an extension pole or, if
possible, by filling the one-gallon sample
bottle directly from the mill discharge
or the tailings pond influent. A minimum
of one gallon of sample was collected.
An additional gallon was collected if the
solids content of the tailings appeared
to be excessively low.
Liquid samples from tailings ponds
or mine pumpout ponds were obtained by
compositing five to ten subsamples taken
along a transect across the pond when
possible or from the perimeter when a
transect was not possible. Subsamples
were collected in two one-gallon poly-
ethylene containers and subsequently
homogenized and transferred to appro-
priately labeled smaller sample bottles
and preserved for subsequent analyses.
Samples of settled solids were taken
at the same locations as the liquid sam-
ples from tailings ponds and mine water
pumpout ponds. Samples were taken with a
core sampler, cork and bottle sampler, or
a scoop, depending on the consistency of
the solids.
Duplicate samples were provided to
the mine site personnel on request.
Analytical Procedures
Analyses were performed on the samples
both in the field and the laboratory. Field
analyses included alkalinity, acidity,
conductivity, temperature, and pH, and
were conducted according to procedures
prescribed in "Methods for Chemical
Analysis of Water and Wastes," U.S.
EPA-600/4-79-020, March 1979. Samples
for laboratory analysis were shipped to
the contractor's laboratory. Subsequent-
ly, selected samples and aliquots of
samples were forwarded to subcontractor
laboratories for specific analyses.
Standard chain of custody procedures were
employed in all shipments.
Solid samples were subjected to
total digestion, RCRA's Extraction Pro-
cedure (EP) toxicity test with acetic
acid, and RCRA's EP using deionized water.
Analyses for radionuclides were performed
on all solid and liquid samples from the
phosphate and uranium industries, on a
selected small population of solid samples
from the metal mining industries, and on
RCRA Extraction Procedure extracts when
the solid sample contained a significant
level of radionuclides. Quality assurance
procedures approved by EPA were adhered
to for both the field and laboratory
analyses to .ensure both the precision and
accuracy of the results.
Summary of Results
The results of the Phase II pre-
survey waste characterization study show
several important findings:
194
-------
Extraction Procedure Analysis
• As expected, a number of the
waste streams analyzed exhibited
elevated levels of 'certain
parameters such as copper,
lead, cadmium, selenium, and
mercury. However, the con-
centrations of these parameters
for the majority of the samples
tested were considerably lower
(in some cases less than detec-
tion limits) in both the acid
and water extracts. In fact,
no sample tested exhibited
toxicity based on the results
for deionized water extracts,
and only 12 samples out of 300
tested (4 percent) were judged
toxic as a result of lead con-
centrations based on the re-
sults for the acetic acid ex-
tract tests. These results
suggest that in a neutral or a
slightly acidic environment,
most of the mining wastes tested
have a relatively low leacha-
bility potential.
• Of the 12 samples tested in
this study which had toxic
extracts from the EP test, 9
were tailings, 2 were mine
waste, and one was settled
solids from a mine water
settling pond.
Radionuclide Analysis
• All liquid and solid samples
collected from metal mining
operations exhibit levels of
radium 226 less than 5 pCi/g
of solid sample and 50 pCi/1 of
liquid sample. Concentrations
of radium 226 in the acid and
water extracts of these samples
were less than 1 pCi/1 of the
extract.
• As expected, levels of radium
226 in wastes collected from
phosphate mining operations
were higher than the values
recorded for samples from metal
mining operations. The levels,
however, were still relatively
low with values ranging from
0.1 to 25 pCi/g of the solid
samples,
• Levels of radium 226 in some of
the samples collected from the
uranium industry were elevated.
Highest values were detected
for settled solids from mine
pumpout water ponds (most values
ranged from 90 to 250 pCi/g)
and low grade ore (values ranged
from 37 to 150 pCi/g of sample).
Samples of overburden and seg-
regated mine waste exhibited
much lower concentrations of
radium 226. Most values were
within a range of 1 to 10 pCi/g
of sample and no sample exceeded
50 pCi/g.
Corrosivity (pH)
. Copper leach liquor was the
only sample type that ex-
hibited the characteristic of
corrosivity. Three samples of
this material were collected
and the pH values of these
samples ranged between 2.1 and
1.9.
Phase II - Comprehensive Monitoring
Shortly after the completion of the
sixty five site visits the comprehensive
monitoring program was initiated. A
total of eight management practices were
selected for this phase of the study:
•Florida phosphate combined mine
waste and clay slimes disposal—
Agrico's Fort Green Mine.
.Southwest copper tailings pond—Cyprus
Pima's Pima Mine.
•Southwest copper dump leach—Kennecott
Copper's Chino Mine.
•South Dakota gold tailings pond—
Homestake's Blackhills Mine.
•Nevada gold tailings pond—Newmont's
Carlin Mine.
..Missouri lead tailings pond—St.
Joe's Viburnum Mine.
•New Mexico uranium mine water pond—
Kerr McGee's Churchrock #1 Mine.
195
-------
•Idaho phosphate waste rock dump—
Stauffer's Wooley Valley Mine.
Specific research plans have been
developed for evaluation of each of the
waste management practices at these sites.
The plans include the design of the mon-
itoring system (air, surface water,
groundwater and climatological); sampling
methods, equipment, scheduling, and han-
dling; quality assurance procedures; mine
solid waste evaluation criteria; and data
handling and analysis procedures.
Monitoring and sampling during this
phase of the study will continue for up
to a 12-month period, and samples will be
collected during four to six separate
sampling periods. After the initial
sampling period at each site, the sam-
pling program may be modified due to the
absence of specific parameters, inappro-
priate sampling frequency, or other in-
adequacies of the initial site-specific
research plan. Monitoring at each site
will include several or all of the fol-
lowing activities:
Surface Runnoff and Seepage: When
flow is present, samples will be
taken prior to exposure to solid
waste material and after exposure to
solid waste. Inflow records will be
maintained.
Storm Event Runoff; During storm
events, runoff samples will be col-
lected on an hourly basis. Contin-
uous flow measurements will be taken
during the storm.
Climatological Data; A continuous
record of precipitation, relative
humidity, temperature, and wind
direction and magnitude will be
collected where appropriate.
Groundwater Samples: Representative
samples will be collected from mon-
itoring wells at a predetermined
frequency. A continuous record of
water level will be obtained when
appropriate and other tests will be
performed when it is necessary to
define groundwater movement.
Solid Waste Samples: At several
sites, representative samples of
solid waste will be subjected to the
EPA EP extraction procedures and the
extract will be analyzed. Samples
will be composited from core borings
at each site and will be analyzed.
Fish and Wildlife: At selected
sites a fish and wildlife-related
assessment will be made. This will
include literature searches, compi-
lation of properties of toxicity,
mobility, and/or bioaccumulation of
selected materials, or the identifi-
cation and elaboration of a specific
situation of interest such as bene-
ficial effect on fish and wildlife
resulting from a specific waste
management practice.
. CURRENT STATUS
It is anticipated that all of the
monitoring stations will be in place by
the end of January 1982. Sampling and
analysis has been initiated at seven of
the sites and will be completed during
the summer of 1982. At the conclusion of
the comprehensive monitoring phase of the
study, a detailed final report will be
prepared describing all the results of
the mining solid waste study. EPA will
use this data, as well as the results
from the studies as a part of a final
report to Congress fulfilling the Agency's
mandate under Section 8002(f) of RCRA.
196
-------
CHEMICAL TREATMENT OF PCBs IN THE ENVIRONMENT
.,. . bv •.' '
Charles J. Rogers
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
.Cincinnati, Ohio 45268
ABSTRACT
During the 50 years that PCBs were manufactured and used in this country, an esti-
mated 400 million pounds of 'these chemicals entered the environment. The primary sources
of PCBs in the environment are the disposal of waste chemicals by consumer industries
and of the waste products of municipal treatment plants. In recent years, PCBs and
other hazardous chemicals have been discovered as contaminants at sites throughout the
country. Conventional waste management methods cannot be used, in many cases, for in-
situ treatment of contaminated soils..
The EPA has initiated studies to determine the efficacy of sodium polyethylene gly.co-
lates (NaPEG) as reagents for decomposing PCBs and other halogenated materials in the
environment. The overall objective of these studies is to demonstrate that NaPEG, in
less-than-stoichiometric proportions, can be applied as spray-on reagents for treatment
of chemically contaminated areas. This report summarizes research aimed at development
of NaPEG for the detoxification/destruction of halogenated chemicals in the environment.
INTRODUCTION
During the past ten years, newspapers
and scientific journals have reported the
discovery of discarded hazardous and toxic
chemicals at sites throughout the country.
In addition to these, the Environmental
Protection Agency (EPA) has identified a
group of materials that have been manu-
factured on a large scale during the past
several years and that are now known to be
harmful to living organisms, to be resistant
to biodegradation and to accumulate in the
food chain. Because of their toxicity,
some of these materials have been removed
from the marketplace. These materials
must now be detoxified or disposed of in
an environmentally acceptable manner. A
partial list of priority chemicals
scheduled for detoxification includes
aldrin, kepone, benzidine, DDT, dieldrin,
dioxin-contaminated materials (2,4-D,
2,4,5-T), endrin, and polychlorinated bi-
phenyls (PCBs). Many of the current dis-
posal methods are not adequate for environ-
mental control of these chemicals. Efforts
are therefore underway to develop effective
treatment methods.
Disposal of PCBs
PCBs are .among the most widespread
synthetic-chemical contaminants in the
environment. During the 50 years that PCBs
have been manufactured and used in this
country, more than 400 million pounds of
these chemicals have;entered'the environment.
The primary sources of PCBs in the environ-
ment are the disposal of wastes by consumer
industries and disposal of the products in
municipal treatment plants. Several dis-
posal practices have contributed to environ-
mental contamination (Figure 1):
• Open burning,or incomplete incinera-
tion of municipal and industrial
waste.
197
-------
§
S
o
Ji
g
§
S
198
-------
« Discharge of PCB-containing fluids
into waterways with municipal and
industrial water effluents.
• Dumping of sewage sludge, solid waste,
and sludge spoils at sea and into
sanitary landfills or dumps. (2)
There is general agreement that atmos-
pheric transport is the major mode of
global dispersal of chlorinated aromatic
compounds such as pesticides and PCBs.
The compounds are volatilized into the
atmosphere and can be carried thousands
of miles from the original source, either
vaporized or adsorbed onto dust particles.
Pesticides and PCBs have been found in
ice samples collected in Arctic and Ant-
arctic regions. (3)
Mason and Hanger (4) report that vent
gas from an ordinary landfill site near
Chicago contained 3240 ng/m3 of PCBs.
Also, air samples taken by the New York
Department of Environmental Conservation
from an uncontrolled dumpsite north of
Albany showed PCB concentrations of 300,000
ng/m3. Studies by Eisenreich, (5) Murphy,
(6) and Bidleman (7) have confirmed that
precipitation events (rain and snow) are
the means of net transfer of PCBs from
the atmosphere to large bodies of water.
Volatilization of PCBs is the principal
reason for concern over the use of land-
fills or lagoons for ultimate disposal of
waste PCBs or PCB-contaminated materials.
Current PCB Waste Management Practices
The chemical stability of PCBs pre-
cludes their destruction by conventional
refuse incineration methods. (8) Most
municipal incinerators cannot achieve the
high temperatures necessary to destroy
PCBs in refuse. The surprisingly high
volatilization rates of PCBs and other
chlorinated aromatic compounds raises
questions over the use of land disposal
for these materials. Further, many land-
fill disposed PCBs may be recovered and
destroyed by chemical treatment or high
temperature incineration in the future.
Currently, some commercial chemical
and physical methods are used to chemically
alter or remove PCBs in PCB-contaminated
oils. These methods include an adsorption
system developed by RTE Corporation. The
chemical methods, developed by Acurex,
Goodyear, and SunOhio, involve dispersion
of metallic sodium in oil or the use of
sodium-biphenyl or naphthalene mixtures.
Because of the reactivity of sodium with
water, these reagents cannot be used effi-
ciently to decompose PCBs in soils, sludges,
and sediments, and dredgings.
Biological treatment of PCBs and other
hazardous pollutants is also receiving
attention. Cytox, Flow Laboratories, and
Sybron Biochemical are some of the companies
marketing strains of bacteria that report-
edly degrade PCBs and other hazardous pollu-
tants.
Although it has been demonstrated that
biological systems can decompose phenols,
fats, waxes and other organics, many skeptics
remain unconvinced that microorganisms can
reliably decompose highly chlorinated iso-
mers of PCBs or other highly halogenated
pollutants.
New Dehalogenation Reactions
Scientists at the Franklin Research
Institute began a'study 3 years ago to
devise a reaction system that would lead to
the cleavage of carbonhalogen bonds. Such
a reaction must be rapid, complete, exo-
thermic, and self-sustaining. After exten-
sive study, the investigators have identi-
fied a promising chemical reagent for this
purpose, prepared from sodium, polyethylene
glycols, and oxygen. The EPA is supporting
further research, which has been broadened
to explore application of dehalogenation
reagents for treatment of soils and lake
sediments as well as oils contaminated with
PCBs or other halogenated compounds.
Typically, the dehalogenation reagent
is prepared by reaction of 60 grams of
sodium with 1.0 liter of polyethylene glycol
with an average molecular weight of 400.
The molar ratio of sodium to solvent should
be 1.1:1.0. About one-half hour after the
sodium has melted, the NaPEG reagent prepa-
ration is complete.
An important feature of the NaPEG
reagent is that dechlorination can be readily
controlled by several parameters:
• Temperature: high temperature favors
rapid and complete dechlorination.
» Reactivity: addition of water increases
reaction times and reduces the degree
of dechlorination.
199
-------
• Concentration of 02: dehalogenation
can be stopped at any time by elimi-
nating 02 from the system.
• Quantity of NaPEG reagent: excess
sodium or other cation ions must be
present to react with liberated halo-
gens.
Kinetic studies have been conducted on
the model compound, 1,3,5-trichlorobenzene,
using 6% by weight of the reagent. In all
cases, the dechlorination rate was first
order with respect to the chlorinated
compound. (9)
Also, a kinetic study of the dechlori-
nation of dichlorobiphenyl (DCS) was carried
out at a single temperature, 59°C. When
DCS was treated with the NaPEG reagent, the
ficst-order kinetic rate constant was
0.1396 rain"1, with a correlation coefficient
of 0.9874. After 1 hour, the reaction
ceased completely. At that point, approxi-
mately 30% of the DCS added to the reaction
system had been dechlorinated.
Pytlewski (9) has proposed mechanisms
for dechlorination using NaPEG.
MECHANISMS
The mechanisms proposed by Pytlewski
involve nucleophilic substitution and
oxidative dehalogenation of halo-organic
compounds. Alkoxides and hydroxides are
potent nucleophiles that can dehalogenate
even the least activated halogenated aro-
DMtic compound. (10)
Hydroxide and alkoxide ions displace
halides of halogenated aromatics to yield
phenols and aromatic ethers, respectively.
The two reactions that take place are as
follows:
AR-X + OH"
Arylhalide Hydroxide
AR-X + RO-
Alkoxide
AR-OH +
X
Arylhydroxide
+ halide
AR-OR + X~
Arylether
The displacement of a chloride ion f
a chlorinated aromatic structure (i.e.,
PCBs) is rapid only if the reagent such as
peroxide, hydroxide, alkoxide, or glycoxide
is in the same phase as the aromatic sub-
strates. The NaPEG reagents are miscible
with halogenated aromatics in general and
possess activated nucleophilic groups
(RO~, NaO2> capable of displacement of un-
activated as well as activated halides under
realtively mild conditions.
The important component of the NaPEG
reagent is the sodium salt of long chain
polyethylene glycols. As proposed, the
sodium cation is complexed by the long chain
polyethylene glycol as shown in (Figure 2).
Activation of the terminal alkoxide group
is due to the lack of ion pairing and anion
solvation. The alkoxide group is associated
with electron abstraction from atmospheric
oxygen, and in the generation of nucleo-
philes. The free radical signals of the
superoxide ion in NaPEG have been measured
by Electron Spin Resonance in concentrations
of 10~3 M.
Roberts and Sawyer (11) report the
degradation by superoxide ion of carbon
tetrachloride, chloroform, methylene chlor-
ide, and p,p,-DDT in aprotic media. They
describe the overall reactions as multi-
step processes and discuss the reactant
product stoichiometry.
By-products of the reaction are salts
and hydroxylated biphenyl derivatives.
Ongoing NaPEG studies will attempt to demon-
strate that less-than-stoichiometric
proportions of NaPEG can be applied to clean
chemically contaminated facilities, soils,
and other materials. Specifically, the
studies will attempt to demonstrate that
the removal of one Cl~ is all that is re-
quired for subsequent degradation of the
hydroxylated structure by naturally present
soil microorganisms.
Applications of NaPEG Reagents
In laboratory studies, NaPEG reagents
reduced PCS concentrations of dielectric
fluid from 1000 ppm to less than 1 ppm and
in soils from 1000 to 481 ppm. (9) Con-
ceptually, since NaPEG dehalogenation
mechanisms are not based on a dispersed
metallic sodium reaction, the reagent can
be used to treat contaminated soils, dredg-
ings, sediments, and low-moisture sludges to
remove organohalogens./
200
-------
A proposed procedure is being evaluated
for treatment of low-moisture 15-20% PCB-
contaminated waste. In two. established
experimental studies at Appleton, Wisconsin
and Coventry, Rhode Island, soil contaminated
with PCBs (42 to 200 ppm) is spread over a
heat-adsorbent liner (PVC) to a depth of 12
inches. The NaPEG reagents, which are very
viscous, are diluted with isopropyl alco-
hol, and sprayed in less-than-stoichiometric
proportions (0.5 M NaPEG/1.0 M PCBs) onto
the contaminated soils.
Soil samples including controls and a
standard (approximately 200 ppm) are taken
before treatment and at intervals of 15,
30, 60, 90, and 180 days after treatment.
The treated areas are covered with a liner
throughout the study to prevent the intro-
duction of excess water from rain or snow
precipitation. It is hoped that this in-
vestigation, scheduled for completion in
June 1982, will prove the method to be more
cost-effective for PCB control than conven-
tional decontamination procedures such as
incineration or containment.
CONCLUSION
Conventional chemical treatment of halo-
genated compounds necessitates the exclu-
sion of moisture and air from the reaction.
Detoxification of PCB-contaminated soils
and sediments by conventional methods would
be extremely costly and cumbersome, parti-
cularly on a commercial scale. A new
chemical reagent has been developed in the
laboratory that requires oxygen and toler-
ates low amounts of moisture which could
facilitate treatment. A reactive reagent
is prepared from sodium, polyethylene gly-
cols, and a continuous supply of oxygen.
In laboratory studies, the NaPEG reagents
have successfully reduced the PCB level of
contamianted oils and soils from 1000 ppm
to less than 1 ppm. The continued success
of ongoing field and laboratory studies
could lead to a -viable method for destruc-
tion of PCBs, chlorinated pesticides,
dioxins, and other toxic organohalogens
dispersed in the environment.
REFERENCES
1. Nisbet, I.D.T. and A. F. Serafim, "Rates
and Routes of Transport of PCBs in the
Environment," Environmental Health
Perspectives, Exp. 1, 21-28, 1972.
2.
7.
8.
9.
10.
11.
Fuller, B., J. Gordon, and M. Korn-
reiches, "Environmental Assessment of
PCBs in the Atmosphere". U. S.
Environmental Protection Agency,
Contract No. 68-02-1495, Mitre Corpor-
ation, April 1976.
Mackay, Donald and Aaron W. Wolkoff,
"Rate of Evaporation of Low-Solubility
Contaminants from Water Bodies to
Atmosphere", Environmental Science and
Technology, 7, p. 611 (July 1973).
Mason and Hanger-Silas Mason Co., Inc.,
Lexington, KY. "Volatilization of PCBs
During Palnned Waukegan Harbor Cleanup
Operations", submitted to U. S. EPA,
Region V, Chicago, IL, May 1981.
Eisenreich, Steven J., Brian B. Looney,
and David Thornton, "Airborne Organic
Contaminants in the Great Lakes Eco-
system" , Environmental Science and
Technology, 15, p. 30 (January 1981).
Murphy, Thomas J. and Charles P.
Rzeszutko, "Eolychlorinated Biphenyls
in Precipitation in the Lake Michigan
Basin". U.S. EPA, Duluth, Minnesota,
EPA 600/3-78-071 (1978).
Bidleman, T. F., C. P. Sice and C. E.
Olney, "High Molecular Weight Hydro-
carbon in the Air and Sea: Rates and
Mechansism of Air/Sea Transfer".
Marine Pollutant Transfer, published
by Lexington Books, D. C. Hearth and
<3o. , Lexington, MA.
Gustafson, C. G., "PCBs-Prevalent and
Persistent", Environmental Science and
Technology, £, 10, 814-819, 1970.
Pytlewski, Louis, "A Study of the
Reaction of Molten Sodium and Solvent
with PCBs", EPA Grant R806649, March
1979.
Bunnett, J., "Aromatic-Nucleophilic
Substitution," Chemical Review, 49,
273-412 (1951).
Robert, J. L. Jr., Sawyer, D. T.,
(Dept. of Chemistry, University of
California, Riverside, CA), "Facile
Degradation by Superoxide Ion of Carbon
Tetrachloride, Chloroform, Methyl-
chloride, and ,p-DDT in Aprotic Medic,"
Journal of American Chemical Society,
103: (3):712-714.
201
-------
DESTRUCTION OF HAZARDOUS WASTE USING SUPERCRITICAL WATER
Michael Modell, Gary G. Gaudet, Morris Simson, Glenn T. Hong and Klaus Biemann
MODAR, Inc.
Natick, Massachusetts 01760
ABSTRACT
MODAR has developed a new process that is capable of destroying highly toxic wastes
with efficiencies of greater than 99.99%. The process uses supercritical water as the
processing medium for oxidation. The organic components of the waste are converted to
carbon dioxide and water with potential recovery of the heating value; inorganic
components are separated out as solid salts, which can be reused or discarded safely;
water in aqueous waste is recovered in a highly purified form, which can be reused
without further treatment. The wastes are treated in an enclosed system so that escape
of pollutants to the environment is eliminated. Furthermore, there is a minimal number
of processing steps so that the capital investment compares very favorably with
alternative processes. The most significant feature of this new approach is that it can
be cost-effective. When the organic content of the waste,is 5% or more, the process is
capable of generating sufficient net power to cover the entire cost of waste treatment.
INTRODUCTION
The MODAR process is based, in part,
on a discovery made in 1975 by Modell and
co-workers at M.I.T. In the course of
investigating the effects of treating
aqueous solutions of organics at high
temperatures and pressures, it was found
that complex organic substances can be
dissolved and broken down into low
molecular weight products when treated
under conditions where water is
supercritical (i.e., above 374°C and
218 atra). The discovery is the basis of
U.S. Patent No. 4,113,446 [Modell, ^t
al. (7)J, originally assigned to M.I.T.
and now exclusively licensed to MODAR.
In 1979, Modell carried this work
one step further. He proposed a new
method of destroying wastes by dissolving
both organic material and oxygen in
supercritical water (SCW), and then
carrying out the oxidation in the
supercritical water medium. MODAR was
formed in 1980 to develop and
commercialize this new SCW oxidation
process. The SCW oxidation process has
been reduced to practice with outstanding
technical results. In addition,
preliminary economic analyses indicate
that the process is potentially much less
expensive in both capital and operating
costs when compared to high temperature
incineration.
TECHNICAL CONCEPTS
In the supercritical region, water
exhibits properties that are far
different from normal liquid water [see,
e.g., Franck (3)]. The density of,
supercritical water (0.05-0.5 g/cm ) is
low enough and the temperature high
202
-------
enough to essentially eliminate hydrogen
bonding [Franck (4)]. As a result, the
dielectric constant is diminished to
about 3 to 10 and water becomes an
excellent solvent for organic substances
[Connolly (2)]. At somewhat higher
temperatures (above 500°C), the
density of water is reduced to 0.05 - 0.1
g/cm and the dielectric constant is
less than 2. Under these conditions,
inorganic salts become only sparingly
soluble [Martynova (6)]. Salts such as
NaCl act as weak electrolytes with
little, if any, dissociation into ions
[Marshall (5)]. Thus, the solubility
characteristics of supercritical water
are the inverse of those of normal liquid
water.
Above 350°C, water reacts with
organic materials in a way that leads to
the formation of low molecular weight
products [Amin (1)]. Whereas many
organic compounds tend to form a high
molecular weight char at temperatures
below 350°C, at supercritical
conditions the same organics are reformed
to gases (e.g., CO, H CH,, CO-)
and volatile organic liquids (alcohols,
aldehydes, furans) without producing any
char.
The products of SCW reforming can be
subjected to oxidation while still under
supercritical conditions. It is well-
known that aqueous solutions of organics
will undergo oxidation at temperatures of
200 to 300°C. This phenomenon is the
basis of the conventional wet oxidation
or Zimpro process, which is operated
under subcritical conditions [see, e.g.,
Wilhelmi and Knopp (8)]. The wet
oxidation process requires residence
times of 20 minutes to 1 hour to achieve
destruction efficiencies which are, at
best, mediocre (70 to 95% reduction in
total organic carbon). On the other
hand, under supercritical conditions, the
residence time required for oxidation is—
less than 1 min, which greatly reduce"!
the volume of the oxidizer vessel. In
addition, oxygen is completely miscible
with supercritical water,—-aad"tfie —
oxidation can be,conducted under
homogeneous (i.e., single phase)
conditions. Thus, oxidation under SCW
conditions is an efficient means of
ultimate disposal of organics.
When toxic or hazardous organic ,
chemicals are subjected to SCW oxidation,
carbon is converted to CO- and hydrogen
to H.O. The chlorine atoms from
chlorinated organics are liberated as
chloride ions. Similarly, nitro-
compounds can be converted to nitrates,
sulfur to sulfates, phosphorus to
phosphates, etc. In other words, hetero-
atoms form oxy-acid anions. Upon
addition of appropriate cations (e.g.,
Na , Mg , Ca ), inorganic salts
can be formed.
Finally, when the concentration of
organics is above 5 wt%, the heat of
oxidation is sufficient to bring the
supercritical stream to temperatures in
excess of 550°C. At these conditions,
inorganic salts have extremely low
solubilities in water. Inorganic salts
will be precipitated out and readily
separated from the aqueous phase. After
removal of inorganics, the resulting
aqueous phase is a highly purified stream
of water at high temperature (>500°C)
and high pressure (3700 psia). It can be
used as a source of high-temperature
process heat or fed to conventional
supercritical steam turbines for
generating power.
PROCESS DESCRIPTION
The MODAR system for hazardous waste
destruction makes use of water in its
supercritical state (SCW) as the process
medium for carrying out the destruction
of organic materials by oxidation. Key
to the success of the process is the fact
that organic substances and gases,
including oxygen, are completely soluble
in SCW, whereas inorganic salts exhibit
greatly reduced solubilities under
process conditions. Thus., it becomes
conceptually possible to carry out
"-combustion" reactions by dissolving
organic substances and oxygen in SCW,
bringing them into intimate contact in a
_jsiogle-phase "medium. The temperatures
and molecular densities allow the
conventional oxidation reactions to
proceed rapidly and essentially to
completion. In fac.t, one might expect
these conditions to be more favorable for
carrying combustion reactions to
203
-------
completion than those of conventional
incineration processes, where
volatilization and mass transport of
reaction species are limiting factors.
Furthermore, the reduced solubility of
salts makes possible the direct removal
of undesirable reaction products through
precipitation.
A schematic flowsheet for the MODAR
process described above is given in Fig.
1. The process consists of the following
steps:
(1) The toxic or hazardous waste is
slurried with make-up water to
provide a mixture of about 5 to 10
wtZ organics. The slurry is
pressurized and heated to
supercritical conditions to avoid
char formation. Heating is attained
by mixing the feed with superheated
SOW, which is generated in a
subsequent step. During a short
residence time in the tube leading
to the oxidizer, organics in the
feed are converted to combustible
gases, low to intermediate molecular
weight compounds (furans, furfurals,
alcohols, aldehydes) and inorganic
salts.
(2) Air or oxygen is pressurized and
mixed with the feed. Since the
water is still supercritical, the
oxidant is completely miscible with
the solution (i.e., the mixture is a
single, homogeneous phase).
Organics are oxidized in a
controlled but rapid reaction.
Since the oxidizer operates
adiabaticallyf the heat released by
combustion of readily oxidized
components is sufficient to raise
the fluid phase to temperatures
where all organics are oxidized
rapidly. For a feed of 5 wt%
organics, the heat of combustion is
sufficient to raise the oxidizer
effluent to at least 550°C.
(3) The effluent from the oxidizer is
fed to a salt separator, where
inorganics originally present in the
feed are removed as a solid slurry.
At 500°C and above, the
solubility of inorganics in SCW is
extremely low.
(4) A portion of the superheated SCW is
recycled to an eductor upstream of
the SCW oxidizer. This operation
provides for sufficient heating of
the feed to bring the oxidizer
influent to supercritical
conditions.
(5) The remainder of the superheated SCW
(with some CO- and N2) is
available for power generation or
use as high-pressure steam. A
portion of the available energy is
used to generate the power required
to pressurize feed and oxidant.
Note that the energy required to
pressurize the oxidant is recovered
in the expansion of the products of
combustion in the superheated SCW
turbine. Thus, the method of
oxidation is analogous to that of a
gas turbine.
As a waste destruction process, the
MODAR concept has several advantages over
conventional processes. The chemical
reactions which occur are carried out in
a closed system, making it possible to
maintain total physical control of waste
materials from storage, through the
oxidation process, to the eventual
discharge of the products of combustion,
and any associated wastes. This feature
provides positive assurance of
environmental protection. In addition,
bench-scale results indicate essentially
complete destruction of chemically stable
materials (such as PCB's) at projected
costs which are well within those
presently associated with hazardous waste
operations. Finally, the process can be
adapted to a wide range of feed mixtures
and scales of operation. Skid-mounted,
transportable systems are being designed
as well as larger-scale stationary units.
EXPERIMENTAL
Under the joint sponsorship of the
U.S. Army Medical Bioengineering Research
and Development Laboratory (USABMERL) and
EPA's Office of Research and Development,
MODAR conducted a series of experiments
to demonstrate the technical feasibility
of the SCW oxidation process. A
204
-------
to
to
o
Q_
z
o
<
X
o
o
s:
LU
o
CO
205
-------
continuous flow bench-scale system with
an organic throughput of 1 gallon per day
was used for these tests. The reactor
was constructed from Hastelloy C-276,
with an i.d. of 0.88 in. and a length of
2 ft. The reactor effluent was cooled to
room temperature and depressurized to 1
a tin. Liquid effluents were analyzed for
total organic carbon (TOG), pH, and by
ion-selective electrodes and gas
chromatography/mass spectrometry (GC/MS).
Gaseous effluents were analyzed by gas
chromatography (GC) for low molecular
weight hydrocarbons and permanent gases.
The terminology used to describe the
results is given in Table 1.
Since the GC/MS analysis of liquid
effluents is of particular importance,
some details will be described. Liquid
effluent samples were first extracted
with dichlororaethane and then
concentrated in a Kuderna-Danish
apparatus. One tnicroliter of this
solution was injected into a stand-alone
gas chromatograph and an appropriate
aliquot (1-2 ul, dependent on the
concentration of the solution as judged
from the GC experiment) injected into the
GC/MS system. A Varian 3700 gas
chromatograph, equipped with flame
ionization detector and automatic
integrator, was used for stand-alone GC
analysis. Columns were 15 m fused quartz
coated with either SE-30 or SE-52.
Another Varian 3700 gas chromatograph was
coupled to a Varian-mat 212 double
focusing mass spectrometer via an
open-split interface. The conditions on
the gas chromatograph were the same as on
Che corresponding stand-alone GC run.
The mass spectrometer was scanned from
M/Z of 30 through 500 at cycle times
ranging from 2.0-2.8 seconds. The
resolution of the mass spectrometer was
set to 1:1000. The GC/MS was interfaced
to a Varian-mat SS-200 data system. This
computer contains all the mass
spectrometry data processing and
evaluation programs developed at M.I.T.
over the years. The data were searched
for compounds of interest either by
inspection of the mass spectra at
GC-peaks in the proper retention areas or
by plotting mass chromatograms of the
characteristic masses of the compounds
(e.g., M/Z of 154, 188, 222, 256, 290,
324, 358, 392 and 426 for biphenyl
through octachlorobiphenyl), followed by
inspection of the mass spectra at these
maxima.
Alternatively, selected mass spectra
were compared with the NIH/EPA/MSDC mass
spectral library or automatically
microfilmed for easier visual inspection.
As a search for tetrachlorodioxins
and tetrachlorodibenzofurans, the mass
chromatograms of M/Z 320 and 304,
respectively, were plotted and the mass
spectra at any maximum in the plot were
inspected for the characteristic M/Z 320,
322, 324, 326, 328 pattern (304, 306,
308, 310, 312 for TCDF). If found, the
entire spectrum would be compared to the
authentic spectra.
After identification of the
components of interest, their
concentrations were estimated from the
stand-alone GC trace. Peak heights were
determined and converted to nanograms
using known amounts of naphthalene,
methyl naphthalene and phenanthrene for
the estimation of aromatics and the
phthalates, and 4,4'-dichlorobiphenyl for
the PCB's and other aromatic chlorinated
compounds.
RESULTS AND DISCUSSION
MODAR has conducted an extensive
series of bench-scale tests to establish
the technical feasibility of SCW
oxidation. A typical series of tests
with organic chloride feeds is described
herein.
The feed materials and compositions
for each run in this series are given in
Table 2. Chronologically, we began this
series with a model compound, DDT, and
progressed to more complex mixtures.
A summary of the results for this
series is given in Table 3. Note that
the compounds identified by GC/MS
analysis are identified by letters at the
bottom of Table 3; the structures are
given in Table 4. Using the methods
described in the prior section, the mass
spectra were searched for each of the
compounds shown in Table 4; only those
found in one experiment or another are
206
-------
TABLE 1. DEFINITION OF TERMS
Residence Time
Volume of reactor divided by volumetric flow rate of
process fluid.
Organic Carbon In (ppm)
Carbon content of organic/water feed mixture as it
enters the reactor.
Organic Carbon Out (ppm)
Total carbon in liquid effluent after sparging or
total carbon minus inorganic carbon.
Destruction Efficiency
(Organic Carbon In - Organic Carbon Out) x 100
Organic Carbon In
Combustion Efficiency
[C02/(C02 + CO + CH4)] x 100
concentrations in the effluent vapor, as measured by
GC analysis.
Organic Chloride In (ppm)
Chloride content of organic/water feed mixture, as
it enters the reactor.
Organic Chloride Out (ppm)
Residual organic chloride as determined by GC/MS.
Organic Chloride Conversion
(Organic Chloride In - Organic Chloride Out) x 100
Organic Chloride In
=====3=:=:==:=:=:
207
-------
TABLE 2. COMPOSITION OF FEED MIXTURES FOR RUNS 11-15
Run 11
DDT
MEK
Run 12
1,1, 1-trichloroethane
1,2-ethylene dichloride
1,1,2, 2-tetrachlorethylene
o-chlorotoluene
1 ,2,4-trichlorobenzene
biphenyl
o-xylene
MEK
Run 13
hexachlorocyc lohexane
DDT
4,4'-dichlorobiphenyl
hexachlorocyclbpentadiene
MEK
Run 14
PCB 1242
PCB 1254
transformer oil
MEK
Run 15
4,4' -dichlorobiphenyl
MEK
C14H9C15
C H 0
4
C2H3C13
C2H2C12
c2ci4
C7H7C1
C6H3C13
C12H10
C8H10
C H 0
4 8
C6H6C16
G14H9C15
C12HgCl
c5ci6
C,HR0
4- 0
C12HXC14-6
C12HXC15-8
C10-C14
C,HaO
4 8
C12HaCl
C,HQ0
wt %
4.32
95.68
100.0
1.01
1.01
1.01
1.01
1.01
1.01
5.44
88.48
100.0
0.69
1.00
1.57
0.65
96.09
100.0
0.34
2.41
29.26
67.99
100.0
3.02
96.98
wt % Cl
2.133
-
2.133 ,
0.806
0.739
0.866
0.282
0.591
-
-
-
3.284
0.497
0.493
0.495
0.505
-
1.99
0.14
1.30
-
-
1.44
.96
.
208
-------
TABLE 3. SUMMARY OF RESULTS: OXIDATION OF ORGANIC CHLORIDES
Run No.
Residence Time (min)
11
1.1
12.
1.1
13
1.1
14
1.1
15
1.3
Carbon Analysis
Organic Carbon In (ppm)
Organic Carbon Out (ppm)
Destruction Efficiency (%)
Combustion Efficiency (%)
Gas Composition
26,700.
2.0
99.993
100.
25,700.
1.0
99.996
100.
24,500.
6.4
99.975
100.
38,500.
3.5
99.991
100.
33,400
9.4
99.97
100.
25.58
59.02
32.84
51.03
37.10
46.86
10.55
70.89
19.00
70.20
Chloride Analysis
Organic Chloride In (ppm) 876. 1266. 748. 775. 481.
Organic Chloride Out (ppm) .023 .037 <.028 .032 .036
Organic Chloride Conversion (%) 99.997 99.997 99.996 99.996 99.993
GC/MS Effluent Analysis
Compound B (ppb Cl) - - • - - -
C _____
E - 9 - 14
F 18 12 18. -
H - - <4. - -
K 5 16 <5. 6
M - - 0.2 -
N '- 0.3 - 36
0 12
209
-------
A:
B:
C:
TABLE 4. COMPOUNDS SEARCHED BY GC/MS ANALYSIS
CHC1
I:
Cl
Cl - \
CHO
CO - CH3
J:
D:
Cl
COOH
L:
Isomer of I.
E:
Cl M: Cl
: Cl
G:
H:
N: Cl
O:
P:
Notes: p-Isomers are assumed, based on the position
of the chlorine atoms in the starting materials.
Compound F - No authentic MS
Compound G - No MS available in literature
210
-------
given in Table 3. In other words,
compounds A,D,G,I,J,L and P were not
present in the test results shown in
Table 3. It should be especially noted
that chlorinated dibenzo-p-dioxins were
never found in any of our effluents.
even though the mass spectra were
specifically searched for them.
The results of run 11, as shown in
Table 3, are representative of one of our
most important findings: organic
chlorides can be destroyed by SCW
oxidation with efficiencies above 99.99%.
In run 11, both carbon destruction
efficiency and organic chloride
conversion exceed 99.99%.
The feed for run 12 consisted of a
mixture of five organic chlorides (see
Table 2), two of which were aromatic
chlorides (o-chlorotoluene and
1,2,4-trichlorobenzene). Once again, the
carbon destruction efficiency exceeded
99.99%. The organic chloride conversion
is 99.997%.
It should be noted that chlorinated
compound K, which is DDE (see Table 4),
was found in the liquid effluent of run
12, although the concentration was minute
(16 ppb). It is extremely difficult to
accept the hypothesis that the DDE was
formed by reaction or rearrangement of
any of the chlorinated organics in the
run 12 feed. We are more apt to believe
that we had cross-contamination of
samples from the low pressure, room
temperature portion of our bench scale
system. To prevent cross-contamination,
an acetone wash system was later
installed.
The feed for run 13 was a mixture of
four non-volatile chlorinated organics,
including hexachlorocyclohexane (Lindane)
and 4,4'-dichlorobiphenyl (a model PCB
compound), both of which should be
relatively stable. As shown in Table 3,
the carbon destruction and chlorine
conversion efficiencies were very good,
the latter being somewhat higher than the
former.
The feed for run 14 was a mixture of
spent transfomer oil, containing two
PCB's, which was diluted with MEK. Once
again, both carbon destruction and
chlorine conversion exceeded 99.99%.
Note that the residual organic chlorides
identified by GC/MS (compounds E,K and 0)
are all DDT-related species. Thus, those
trace quantities are probably cross-
contaminants from run 13. It should also
be noted that for run 14, mass spectra
were searched for all chlorinated
biphenyls. That is, we specifically
looked for trace quantities of tri-,
tetra-, penta- and hexa-chlorobiphenyl,
in addition to the mono- and di-chloro
species (compounds M and N in Table 4).
At the detection limit of < 10 ppb, none
of these chlorinated biphenyls were
found. If we discount compounds E,K and
0 as cross-contaminants, then the
chlorinated biphenyl conversion
efficiency is >99.9994%.
Run 15 was an experiment with a
model PCB, 4,4'-dichlorobiphenyl. The
results obtained in run 15 were once
again excellent.
CONCLUSION
The present results indicate that
SCW oxidation gives essentially complete
destruction of refractory organic
compounds in residence times on the order
of 1 min. As shown in Table 5, a broad
spectrum of organic materials have been
tested; the results are similar to those
described above. In general, the SCW
oxidation is non-specific.
Although not demonstrated here, the
process has other advantages which make
it an excellent means of hazardous as
well as non-hazardous waste disposal.
Auxiliary pollution control equipment for
effluent streams is unnecessary. The
process is applicable to wastes in the
form of liquids, solids, or sludges. It
is particularly applicable to aqueous
waste streams containing mixtures of
organic and inorganic materials, with
inorganics being recovered as a solid.
Upon removal of inorganic compounds, the
high temperature, high pressure mixture
of steam and gases is available to be
used as process heat or for generation of
power. When the organic content of the
waste is greater than 5%, sufficient net
power may be generated to cover the
entire cost of treatment. The process
211
-------
cost compares very favorably with
alternative schemes, and is expected to
see wide application in the near
future.
TABLE 5. MATERIALS TESTED UNDER SOW
OXIDATION
Dextrose
Cyclohexane
Biphenyl
Benzene
1,1,1-Trichloroethane
1,2-Ethylene Dichloride
o-Chloro toluene
1,2,4-Trichlorobenzene
o-Xylene
Hexachlorocyclohexane
4,4'-Dichlorobiphenyl
Hexachlorocyclopentad iene
DDT
PGB 1234
PCB 1254
2,4-Dinitrotoluene
Urea
Methyl ethyl ketone
REFERENCES
1. Amin3 S., R.C. Reid and M. Modell.
1975. Reforming and Decomposition of
Glucose in an Aqueous Phase, Am. Soc.
Mech. Eng., paper no. 75-ENAs-21.
2. Connolly, J.F. 1966. Solubility of
Hydrocarbons in Water Near the Critical
Solution Temperature, J. Chem. Eng.
Data, 11, 13.
3. Franck, E.U. 1970. Water and
Aqueous Solutions at High Pressures and
Temperatures, Pure Applied Chem., 24,
13.
4. Franck, E.U. 1976. Properties of
Water, in High Temperature, High
Pressure Electrochemistry in Aqueous
Solutions (NACE-4), p. 109.
5. Marshall, W.L. 1976. Predicting
Conductance and Equilibrium Behavior of
Aqueous Electrolytes at High Temperatures
and Pressures, in High Temperature, High
Pressure Electrochemistry in Aqueous
Solutions (NACE-4), p. 117.
6. Martynova, O.I. 1976. Solubility of
Inorganic Compounds in Subcritical and
Supercritical Water, in High Temperature,
High Pressure Electrochemistry in Aqueous
Solutions (NACE-4), p. 131.
7. Modell, M., R.C. Reid and S. Amin.
1978. Gasification Process, U.S. Patent
4,113,446, Sept. 12.
8. Wilhelmi, A.R. and P.V. Knopp. 1979.
Wet Air Oxidation - An Alternative to
Incineration, Chem. Eng. Progress, 75
(8), 46.
212
-------
THE DESTRUCTION OF VARIOUS ORGANIC SUBSTANCES
BY A CATALYZED WET OXIDATION PROCESS
Richard A. Miller
Mark, D. Swientoniewski
IT Enviroscience
Knoxville, Tennessee 37923
ABSTRACT
Research was initiated to determine the ability of the catalyzed wet oxidation process to
destroy organic substances that are representative of materials found in hazardous wastes.
Catalyzed wet oxidation uses homogenous catalysts in conventional chemical reaction
equipment to^ultimately oxidize an organic substance to carbon dioxide. Fifteen organic
compounds were studied in a laboratory-scale batch reaction system. These compounds were
destroyed with reaction conditions ranging from 165°C to 275°C, 150 psig to 1000 psig and
reaction time from 5 minutes to 420 minutes. Preliminary designs using a portable 1000
gallon treatment process have been made based on demonstrated destruction rates. The
cost of treating some organics range from $0.12/lb of organic to $1.04/lb of organic.
INTRODUCTION
The United States is facing a major en-
vironmental problem as it becomes more
aware of the production and handling of
toxic and hazardous wastes. These wastes
have been produced for many years, but
their disposal has not been controlled.
The recent increase in public awareness
and concern has prompted the United States
Environmental Protection Agency (USEPA) to
evaluate technologies aimed at solving the
problem. Of particular interest is the
development and application of new tech-
nologies to the problems associated with
abandoned chemical waste disposal sites
such as Love Canal and the Valley of the
Drums. The estimate of the number of such
sites in the United States and the cost to
clean them up is staggering. Treatment of
the leachate and ground waters and ulti-
mate disposal of the sludges, soils,
liquids, solids, and containers in the
site are typical disposal problems.
The most desirable technology for dealing
with the problem of chemical disposal
sites is. a process that 1). achieves com-
plete destruction of toxic and hazardous
chemicals and minimizes by-products and
chemical remnants requiring disposal;
2) minimizes the energy required for
disposal, especially supplemental fuels;
3) has few unit operations; 4) has low
chemical usage; 5) has low volume
effluents, so that polishing treatment
to achieve total containment can be
readily implemented; 6) is easily pilot-.
planted; and 7) is transportable to the
landfill sites.
The present study evaluates a new cata-
lytic wet oxidation technology for
the treatment of toxic and hazardous
materials. It is applicable to both
aqueous wastes and organic residues.
Earlier work resulted in U.S. Patent
3,984,311 (1) and two additional patent
applications that describe catalyst
systems for destruction of various
chemicals. The data show fast de-
struction rates for compounds such as
nitrated and chlorinated phenols, 2,4-D,
aklylphosphorothioates, and short chain
fatty acids. On the basis of this
earlier range-finding testing and eco-
nomic evaluation, the technology appears
to have wide-ranging applicability and
potential for alleviating significant
waste disposal problems facing the
213
-------
United States. The data on which this
paper was developed was supported by funds
from the USEPA.
CATALYZED WET OXIDATION PROCESS CONCEPT
The catalyzed wet oxidation process is
based on U.S. Patent 3,984,311, originally
assigned to The Dow Chemical Company with
right, title, and interest now owned by
IT Enviroscience for development and com-
mercialization. The patent teaches the
use of a co-catalyst system consisting of
bromide and nitrate anions in an acidic,
aqueous solution. Continued research
with the co-catalyst system has led to the
development of a new catalyst mixture
which is more effective for oxidizing
insoluble organics. The new catalyst
system consists of bromide, nitrate, and
manganese ions in acidic solution. A
patent for the new catalyst has been
allowed (U.S. 4,276,198) and assigned to
IT Enviroscience (2). The destruction of
most organics by these catalyst systems
is rapid and essentially complete.
The keys to the operation of this catalyst
system are the mechanism of oxygen fix-
ation and the fact that it is a water-
soluble, single-phase catalyst system—
a homogeneous catalyst as contrasted to a
heterogeneous catalyst. In conventional
wet oxidation, heat and pressure are used
to drive the dissolution of oxygen from
air and the reaction with dissolved or-
ganics in aqueous solution. In the
bromide-nitrate based catalyst systems
the transfer of oxygen to the dissolved
state is speeded up by using very rapid
gas and liquid reactions associated with
the catalyst components. 'The importance
of the enhanced oxygen transfer is the
ability to oxidize organics at much lower
temperatures than uncatalyzed wet oxi-
dation, 165-200°C versus 250-325°C. The
lower operating temperatures also mean
lower operating pressures which not only
reduces capital cost but operational
problems.
The second important aspect of the catalyst
system is its homogeneous nature which
permits application to the destruction of
toxic or hazardous organic residues, such
as still bottoms or other organic wastes.
The advantages of a homogeneous catalyst
are best utilized by using a reactor design
which is different from conventional wet
oxidation processes. In simplest form the
reactor, a continuously stirred tank
reactor (CSTR) contains the catalyst
solution. Air and the waste are continu-
ously pumped into the reactor and the
organics are oxidized with the heat of
reaction driving off water. The only
materials to leave the reactor are CO ,
N , water vapor, any volatile organics
and inorganic solids formed. Water is
condensed and returned to the reactor, if
necessary, as are condensable organics.
Any inorganic salts or acids which may be
formed have to be removed by treatment of
a closed loop stream of catalyst solu-
tion. Such treatment is- individually
designed utilizing conventional tech-
nologies, such as filtration or distil-
lation. The vent gases from the reactor
are low in volume and may, if necessary,
be .treated by conventional techniques,
such as absorption, adsorption, or
scrubbing. The most important features
of this process concept are that non-
volatile organics remain in the reactor
until destroyed, and that there is no
aqueous bottoms product.
COMPOUND SELECTION
The study of a process associated with
chemical wastes needs to evaluate a wide
range of chemical compounds to evaluate
the applicability of the process. In
order to meet this requirement fifteen
compounds were selected from various
chemical groupings of the EPA priority
pollutant list. The specific compounds
chosen from each grouping were made based
on the likelihood of finding the chemipal
in industrial waste and chemical disposal
sites. In certain cases a lower toxicity
compound was substituted for a highly
toxic compound, e.g., diphenyl hydrazine
substituted for benzidine, when it pre-
sented an unacceptably high risk to lab-
oratory personnel. The chemical groups
and the compounds studied were:
- Halogenated hydrocarbons
Ethylene dibromide
Hexachlorobutadiene
Trichloropropane
- Pesticides
Atrazine
DDT
Malathion
Mirex
- Phenols
Pentachlorophenol
- Phthalates
214
-------
Di-n-butyl phthalate
Polynuclear aromatics
Chloroanthracene
Nitrogen containing compounds
Acetonitrile
Chloroaniline
Diphenyl hydrazine
Nitrobenzene
Aromatic
o-Xylene
organics), by ion chromatography. All
of these methods followed quality assur-
ance/quality control (QA/QC) procedures
as outlined in the IT Enviroscience
Analytical Chemistry Control Manual.
These measurements were used to deter-
mine the completeness of the organic
destruction.
EXPERIMENTAL RESULTS
EXPERIMENTAL METHOD
Summary
The experimental evaluation of catalyzed
wet oxidation was conducted in a 1-liter
agitated titanium autoclave. The de-,
struction rate of the compounds was
measured at various operating conditions
with batch reactions. Typical operating
conditions were 165-250°C, 0.5% Br , 5.0%
NO ~ 0.25% Mn+ , 30-120 minutes re-
action time, and 20% excess oxygen over
stoichiometric. The procedure for con-
ducting the oxidation reactions con-
sisted of loading the reactor with, the
desired quantity of deionized water, HBr,
and organic to be oxidized. The reactor
was sealed, purged with oxygen, and
pressurized with sufficient oxygen to
totally oxidize the organic. The re-
actor was then heated to the desired
operating temperature and the remaining
catalysts, HNO , and MnSO , were added to
the reactor with oxygen pressure. The
reaction was run for the desired time
period. At the termination of the re-
action, the reactor was cooled to room
temperature with cooling water. The
pressure was then vented and any free
bromine was reduced to bromide by the
addition of 10 ml of sodium bisulfite
solution to prevent operator exposure to
toxic bromine vapors. The reactor was
opened and the aqueous contents were
aspirated from the reactor. All internal
surfaces of the reactor were rinsed twice
with solvent, methylene chloride. The
solvent rinsings were aspirated from the
reactor and combined with the aqueous
effluent. The aqueous phase was extracted
with the solvent rinse and two successive
50-ml portions of solvent to recover
unreacted organics and reaction by-pro-
ducts. The combined solvent extracts
were analyzed by gas chromatography and
gas chromatography/mass spectrometry
(GC/MS). The destruction rate of the
organic was also measured by the formation
of CO , by gas chromatography, and the
formation of Cl (from chlorinated
For the purpose of evaluating the appli-
cability of the process to treat hazard-
ous wastes in chemical landfills, the
results of the individual compound
studies are grouped into two major cate-
gories, fast and slow destruction rate
compounds. The definition of fast and
slow destruction rates can be set at
any level, but the results of the com-
pounds studied can be split as follows.
Compounds that were destroyed at low
temperatures, less than 200°C, and short
reaction times, less than 60 minutes,
were classified as fast destruction rate
compounds. This group includes atrazine,
butyl phthalate, chloroaniline, diphenyl'
hydrazine, ethylene dibromide, malathion,
and pentachlorophenol. Compounds in the
other group (slow destruction rate) were
more difficult to destroy. These com-
pounds required higher reaction tem-
peratures, up to 250°C, and longer
reaction times. This group includes the
other compounds studied, acetonitrile,
Chloroanthracene, DDT, hexachlorobuta-
diene, mirex, nitrobenzene, trichloro-
propane, and o-xylene.
Atrazine
The destruction rate for atrazine was
very fast, apparently due to both the
instability of the heterocyclic nitrogen
ring at high .temperatures and the oxi-
dation by the catalysts. The instability
of the atrazine molecule was observed in
the first experiment without., catalysts.
At a temperature of 163°C, 51.5% of the
atrazine was oxidized with release of a
comparable amount of chloride (58.3%) but
very little CO2 formation. This cor-
responding release of Cl with atrazine
decrease suggests an initial C-C1 bond
rupture. Whether this reaction occurs in
the presence of the catalyst system is
unknown and beyond the scope of this
investigation. However, the addition of
215
-------
the Br-NO -Mn catalysts increased the oxi-
dation of atrazine (99.9%) and the forma-
tion of CO, (45.3%) and chloride (83.1%).
A later experiment using only bromide and
nitrate catalysts had a similar oxidation
rate for atrazine (99.9%) and a formation
rate of chloride (100%) but very little
formation of CO, (0.6%), indicating the
manganese plays a critical role in the
complete oxidation of atrazine to carbon
dioxide. Because of the demonstrated
importance of manganese in the catalyst
system, the remaining experiments were all
performed with the three-catalyst system,
Br-NO,-Mn. The experimental variables of
reaction time, temperature, and atrazine
concentration were further evaluated to
determine the maximum destruction rate for
the catalytic oxidation of atrazine.
The rapid oxidation rate of atrazine and
the limited accuracy of the CO, and
chloride analysis made it difficult to
evaluate optimum oxidation conditions.
However, the experimental data can be ex-
amined for trends toward optimum reaction
parameters. Increasing the reaction tem-
peratures from 165 to 200°C increased the
complete destruction rate to CO2 by ap-
proximately 40%, during a 30 minute re-
action: The effect of time during these
experiments was minimal, with less than
an 18% increase in formation of C02 with
a four-fold increase in reaction time,
from 30 to 120 minutes. The effect of
atrazine concentration was determined by
increasing its concentration ten-fold from
0.1% to 1.0%. The destruction rate
measured by the oxidation of atrazine and
the formation of CO, remained constant,
indicating a first-order kinetic relation-
sip for atrazine concentrations. The
first-order relationship of concentration,
coupled with the fast oxidation rate,
permits the design of a treatment process
that can destroy large quantities of atra-
zine in relatively small equipment.
Therefore, a waste treatment process with
small equipment (low capital cost) and
high throughput (via high destruction
rates) would be a very economical process
to install and operate. A catalyzed wet
oxidation process offers a technically
and economically viable alternative for
destroying atrazine or similarly
structured compounds.
Butyl Phthalate
The oxidation of butyl phthalate was dif-
ferent from the behavior exhibited by
atrazine. Butyl phthalate showed con-
siderable resistance to wet oxidation
without the presence of catalysts. At
164°C only 6% of the initial butyl
phthalate was oxidized in 30 minutes with
1.2% of the initial amount being com-
pletely oxidized to CO.. The addition of
the Br-NO -Mn catalyst increased the
oxidation rate of butyl phthalate to
86.6% in 30 minutes at 163°C with 50.6%
of the initial material being totally
oxidized to CO.. The effect of the in-
creased reaction time increases both the
oxidation of the butyl phthalate and the
formation of CO,. A four-fold increase
in time, 30 to 120 minutes, increased the
oxidation rate by 13%. This small effect
on reaction time indicates that the oxi-
dation of butyl phthalate probably
follows a multistep process and that
although the initial steps are fast, the
final oxidation reactions are slow.
Based on earlier catalyzed oxidation
experiments with carboxylic acids, e.g.,
acetic acid, it is postulated that the
early oxidation reactions yield rela-
tively stable by-products which oxidize
at a much slower rate.
The disappearance of butyl phthalate is
first-order, with similar oxidation rates,
96.0% vs. 97.6% for 0.1% and 1.0% butyl
phthalate experiments. This result is
consistent with early studies of water-
soluble organics and the data from the
study of atrazine. The effect of various
catalyst mixtures on the oxidation rate
of butyl phthalate and the formation rate
of CO was similar to the results re-
ported for atrazine. Although the actual
effects of the catalysts are small, the
trends are Similar. The removal of
manganese from the mixture decreases the
oxidation rate of butyl phthalate and the
formation of CO,. Increasing either the
bromide or nitrate concentration slightly
increases the oxidation rate of butyl
phthalate but has little effect on the
formation of CO,.
Analysis of selected experiments by GC/MS
revealed that the extractable by-products
from the oxidation of butyl phthalate
fall into two categories, brominated
organics and phthalic acid. Brominated
by-products, bromoform and dibromopropane,
were present in the solvent extracts of
216
-------
all the samples analyzed by GC/MS.
Phthalic acid was identified in the '
extracts of two later experiments, both
high butyl phthalate concentration experi-
ments. •_ • ,
Chloroaniline
The oxidation of chloroaniline was similar
to the oxidation of atrazine in that
chloroaniline.was partially oxidized with
no catalyst present. Two experiments
were conducted with no catalyst at tem-
peratures of 165 and 199°C. It is not
known whether the. uncatalyzed reaction was
due to a wet oxidation reaction or to a
rearrangement reaction. However, the ease
of oxidizing chloroaniline without cata-
lysts usually results, as in the case with
atrazine, in rapid and essentially com-
plete oxidation with the Br-NO -Mn
catalyst system. The solvent-extracted
effluent from the third experiment was
analyzed by GC/MS for reaction by-products.
The compounds identified were small
amounts of dibromochloromethane, bromoform,
tetrabromoethane, and various brominated
chlorobenzenes. These by-products are
consistent with other analyses of reactor
effluents. The by-products or partially
oxidized compounds are ultimately oxidized
to CO , HC1, and HO.
^ ^ . ' . '
Chloroanthracene
The oxidation of Chloroanthracene was very
different from any of the earlier com-
pounds studied. The initial experiment
with no catalysts showed very little oxi-
dation. However, the rest of the ex-
periments showed very high destruction
rates by disappearance of Chloroanthracene
and by the limited formation of CO arid
Cl . Analysis of the reactor effluent by
gas chromatography showed the presence of
a major reaction by-product, which was
identified by GC/MS as chloroanthra-
cenedione with smaller amount of bromo-
form, tetrabromoethane, and'dibromochloro-
benzene. The significance of the chloro-
anthracenedione by-product is that it is
the first major by-product identified
which was not brominated. Later'experi-
ments show that chloroanthracenedione is
oxidized further at longer reaction times
or under more stringent conditions. The
similar destruction rates as measured by
CO and Cl formation indicate that the
chloroanthracenedione can be completely
oxidized.
Ethylene Dibromide
The oxidation of ethylene dibromide (EDB)
follows a multistep sequence that
initially results'in the'disappearance
of EDB with very little CO formation.
This occurred in the first and third
experiments when no catalysts were used
at a reactor temperature of 164°C. The
mechanism which causes this reaction is
not known, but the addition of the
Br-NO -Mn catalyst to the reactor re-
sults in the oxidation of the organics
to CO . The excess formation of CO» is
attributed to an analytical error as-
sociated with the CO analysis rather
than to the complete oxidation of all
the EDB placed in the reactor. The only
identified by-product of the oxidation
reaction was bromoform in the second
experiment.
Hexachlorobutadiene
The oxidation of hexachlorobutadiene was
more difficult than for any of the com-
'pounds tested earlier. Wet oxidation at
165°C without catalysts was ineffective
in destroying hexachlorobutadiene. Even
the addition of the Br-NO -Mn catalyst
did not improve the oxidation rate at
165°C. The resistance of hexachloro-
butadiene to oxidation required a higher
reaction temperature and a longer re-
action time to achieve destruction.
After two hours at 250°C, 25% of the
hexachlorobutadiene was oxidized to CO'
and Cl . Although this destruction rate
is relatively slow, the catalyst system
successfully destroyed hexachlorobuta-
diene. Further experiments with longer
reaction times increased the destruction
rate but the overall rate, 45% de-
struction in 7 hours, was still too slow
to effectively destroy large quantities
of hexachlorobutadiene. However, ex-
periments with higher concentrations of
hexachlorobutadiene yielded much higher
destruction rates. A later experiment
had the greatest destruction of hexa-
chlorobutadiene, 91% in 2 hours'at
242°C.• 'It'is unknown whether the high
destruction rate was due to the presence
of greater quantities of chloride or to
an insoluble'organic phase.. The reason
for the improved destruction rate could
be the subject of future laboratory
studies. Analysis of the reactor ef-
fluent from the hexachlorobutadiene
experiments by GC/MS revealed the
217
-------
presence of hexachloroethane, an impurity
of hexachlorobutadiene, and bromoforra, a
by-product of the oxidation.
Malathion
Malathion was so unstable at temperatures
higher than 165°C that none could be de-
tected in the reactor effluent. The
affect of temperatures was isolated by
running an uncatalyzed oxidation at room
temperature. Greater than 91% of the
malathion was recovered from the reactor.
This experiment demonstrated that mala-
thion is thermally unstable. The only
significant experimental difference ob-
served was that uncatalyzed oxidation gave
limited CO , indicating only partial oxi-
dation of malathion.
In the presence of the catalyst, malathion
was oxidized completely to CO . The high
CO_ formation results, 105% to 113%, indi-
cate complete oxidation. A few trace
organics were identified in the reactor
effluent by GC/MS. These were methylene
bromide and bromoform, which could be
final remnants from the oxidation of
malathion.
Hirex
Of the fifteen organics studies, mirex was
the most resistant to oxidation. The
first experiment, no catalysts, showed no
oxidation of mirex and greater than 99%
recovery at 199°C. The addition of the
standard Br-NO_-Mn mixture only increased
the oxidation rate to 8.5% in 65 minutes.
With the apparent high resistance to oxi-
dation later experiments utilized much
higher reaction temperatures to attempt
mirex oxidation. At temperatures as high
as 274°C, only 15.9% of the mirex could
be oxidized. No extra compounds were
indicated by analysis of the reactor ef-
fluent, so no analyses by GC/MS were
performed.
Nitrobenzene
Nitrobenzene was the first compound evalu-
ated during this study and the experi-
mental results are difficult to interpret.
The uncatalyzed oxidation of nitrobenzene
was evaluated at 200°C with reaction times
from 4 to 60 minutes. The loss of nitro-
benzene in these runs ranged from 18 to
31% with no apparent relationship between
loss of nitrobenzene and reaction time.
Inconsistent data were attributed to
start-up problems associated with a new
operator. The formation of CO in the
uncatalyzed oxidations was low, 0.2 to
8.2%, indicating that very little nitro-
benzene was being totally oxidized. The
addition of catalysts to the experiments
at 200°C had little effect on the partial
oxidation, or loss, of nitrobenzene.
When the reaction temperature was raised
to 250°C, the oxidation rate increased
to 46% in 60 minutes with 12.2% of the
nitrobenzene being completely oxidized
to CO. Although the results of the
nitrobenzene experiments are too varied
for detailed analysis, the overall re-
sult of the evaluation is that nitro-
benzene can be oxidized by the Br-NO -Mn
system at slow rates.
Pentachlorophenol
The oxidation of pentachlorophenol was
evaluated in a series of six experiments.
The oxidation rate with no catalysts was
measured at reaction temperatures of 165
and 194°C. The results indicate that
pentachlorophenol is oxidized by wet
oxidation at moderate rates and that the
oxidation is complete, as measured by
CO . However, the addition of the Br-
NO_-Mn catalyst drastically increased the
oxidation rate by all three methods of
measurement, loss of pentachlorophenol
and formation of CO and Cl . The
initial catalyst reactions with penta-
chlorophenol were so fast that after 30
minutes at 165°C more than 99% of the
pentachlorophenol was partially oxidized.
Total oxidation can be achieved with
either increased reaction time or in-
creased reaction temperature. The re-
action by-products from later experiments
were analyzed by GC/MS to identify the
partial oxidation products of penta-
chlorophenol. The only compounds present
were dibromochloromethane and bromoform.
Such short chained organics are usually
the oxidation products prior to the final
oxidation to CO .
Trichloropropane
The oxidation of trichloropropane was
evaluated at a few experimental con-
ditions. The first experiment was an
uncatalyzed oxidation at 163°C for 30
minutes. At these conditions, 14.5% of
the trichloropropane was oxidized, 5.3%
completely to chloride. The next experi-
218
-------
merits evaluated the effect of adding the
catalysts and increasing the reaction
temperature. The addition of the Br-
NO -Mn catalyst doubled the destruction
rate by loss of trichloropropane and by
formation of CO and Cl~. These de-
struction rates were doubled again by in-
creasing the reaction temperature from
165 to 200°C. The ability to oxidize tri-
chloropropane was demonstrated during
these 30-minute reactions. The by-
products of the reaction were analyzed by
GC/MS and identified as debromochloro-
methane, bromoform, and dichlorobromo-
propane.
o-Xylene
The oxidation experiments with xylene
generated some interesting data. The
uncatalyzed oxidation at 165°C destroyed
more than half the xylene, 54.3%, but
generated only 0.3% of CO . The wide
difference in destruction rates may be
attributable to partial recovery of un-
reacted xylene with the solvent ex-
traction procedures used, but this was
not verified during the xylene oxidation
experiments. The destruction rates were
increased by the addition of the Br-NO -
Mn catalysts but not by increasing the
reaction temperature.
The most surprising result was the in-
crease in destruction rates caused by
increasing the initial concentration of
xylene. During these experiments the
destruction rate by loss of xylene was
greater than 98% with almost half the
xylene being totally oxidized to CO .
This result was also observed with Hexa-
chlorobutadiene but the reason for the
increase was beyond the scope of this '
project and was not studied. However,
this observation would indicate a
slightly higher overall order of reaction
for xylene than the first-order rate of
reaction observed for the other compounds.
Keeping in mind the complexity and number
of reactions occurring in the liquid and
vapor phases, this slight increase in
order may indicate a different reaction
mechanism during the destruction process
of this compound. The by-products of
xylene oxidation were identified by GC/MS
as bromoxylene and dibromoxylene. The
presence of the brominated by-products
reinforces the hypothesis that bromine
species are the major oxidant in this
catalyst system.
Diphenyl Hydrazine
The oxidation of diphenyl hydrazine is
fast and comparable to the oxidation of
butyl phthalate. Diphenyl hydrazine was
oxidized slowly without catalysts. At
164°c, 19.3% of the initial diphenyl
hydrazine was oxidized in 30 minutes and
only 3.6% of the initial amount was
completely oxidized to CO . . Addition of
the Br-NO3-Mn catalysts increased the
destruction of diphenyl hydrazine to
73.3% in 30 minutes at 164°C, and 44.5%
of the initial amount was totally oxi-
dized to CO2. Increasing the reaction
temperature from 164°C to 200°C caused
destruction of 97.9% of the initial
diphenyl hydrazine in 30 minutes and
69.8% of this initial amount was oxi-
dized completely to CO .
Surprisingly, when the initial diphenyl
hydrazine concentration was increased
from 0.1% to 1.0% the destruction de-
creased to 57.9% vs. 97.9% by disap-
pearance of organic and 39.5% vs. 69.8%
by conformation. The reason for this •
sharp decrease was not studied nor was
the experiment repeated to confirm this
result. The trace by-products of oxi-
dation of diphenyl hydrazine were
identified by GC/MS as tribromomethane,
various bromobenzenes, and bromodiphenyl
hydrazine.
DDT . . :
The oxidation of DDT without catalyst is
slow. At 166°C only 9.6% of the initial
DDT oxidized in 60 minutes and 2.8% of
the initial amount oxidized to CO and
5.6% by Cl formation. The addition of
the Br-NO3-Mn catalyst raised the de-
struction of DDT to 59.8% in 60 minutes
at 165°C and 5.4% of the initial material
oxidized to C02, 19.9% to Cl~ ions. In-
creasing the temperature- to 249°C in-
creased the destruction to 93.5% but
did not significantly change the CO or
Cl destruction numbers. Also, an in-
crease in time from 30 to 60 minutes
increased the destruction by only 10.2%.
As in the case of butyl phthalate oxi-
dation, the small effect of reaction
time on destruction indicates that the
oxidation of DDT follows a multistep
process.
219
-------
The destruction of DDT is first-order
with respect to initial DDT concentration.
Destruction was 100% vs. 98.4% for 0.1%
and 1.0% DDT concentration experiments.
Experimental conditions were 150°C for
120 minutes for these experiments.
As in the case of butyl phthalate, the
actual effects of various catalyst mix-
tures on the destruction of DDT is minor.
The removal of manganese from the catalyst
mixture decreases the destruction of DDT
and the formation of CO and Cl ions.
Increasing either bromide or nitrate
slightly increases the destruction of
DDT as measured by the formation of CO2
and Cl ions.
Analysis of the extractable by-products
by GC/MS revealed a variety of materials.
By-products formed in an early experiment
include bromoform, dibromochloromethane ,
dichlorobenzene , bromochlorobenzene ,
dibromochlorobenzene , dichlorophenyl
chlorophenyl ketone, chlorophenyl ketone,
bromochlorophenyl chlorophenyl ketone,
and l,l-dichloro-2,2-bis (p-chlorophenyl)
ethylene. Later experiments include
toluene, benzaldehyde, tetrachlorobenzene ,
dichloroacetophenone , pentachTorobehzene ,
trichlorobenzaldehyde, and hexachloro-
benzene. This variety of by-products
from the oxidation of DDT is supportive
of the fact that initial experiments
supported a multi-path decomposition as
indicated by a large disappearance_of DDT
but a low appearance of CO2 and Cl ions.
Although these by-products are somewhat
more stable toward oxidation than the
original DDT molecule these materials
are ultimately oxidized to CO2/ HC1, and
Acetonitrile
Acetonitrile is resistant to wet oxi-
dation. Without catalyst only 5.4% of
the initial material oxidized in 30
minutes at 165°C. When the catalyst
mixture was added, 89.9% of the acetoni-
trile disappeared at the same reaction
conditions. An increase in temperature
to 200°C and initial acetonitrile con-
centration from 0.1% to 1.0% resulted in
destruction increasing to 96.3% and 99.3%
respectively. An increase in reaction
time from 30 minutes to 60 minutes had
no effect on the destruction of acetoni-
trile.
PROCESS DESIGN AND ECONOMICS
With data generated from the batch oxi-
dation reactions of the 15 compounds
preliminary capital and operating costs
can be estimated for a full-scale con-
tinuous process to destroy these com-
pounds. Two processes were designed to
destroy aqueous wastes or organic resi-
dues.
Process Description; Organic Residues
A continuous large-scale system for the
destruction of organic residues by
catalyzed wet oxidation is based on the
use of a conventional continuous stirred
tank reactor (CSTR). The organic resi-
due and compressed air enter the pres-
sure vessel continuously. A reflux
condenser returns condensate to the
reactor. Noncondensables are scrubbed
with caustic and passed through a carbon
adsorber. For chlorinated residues it
is also necessary to remove HC1 from the
reactor, which is accomplished through a
distillation of the reactor contents and
a return of the catalysts to the reactor.
Makeup catalysts are also added to the
reactor to compensate for losses through
the reactor vent.
A CSTR in this system has the advantage
of requiring the relatively insoluble
and nonvolatile organic to remain in the
reactor until it is oxidized to CO2 or a
volatile organic. However these vola-
tile organics are condensed and returned
to the reactor through the condensate
accumulator. This advantage is more
important to a "real world" waste stream
composed of easily oxidized or more dif-
ficult to oxidize organic compounds.
Therefore a CSTR-based wet oxidation
system allows for the destruction of the
complete range of organics if the design
is based on the most difficult to oxi-
dize organic compound.
Process Description: Aqueous Wastes
A large-scale system for the destruction
of an aqueous organic waste by catalyzed
wet oxidation has the same major compo-
nents and functions as the organic resi-
due oxidation system. The major dif-
ference in these two systems is that any
feed water entering the oxidation re-
actor can only exit the reactor as water
220
-------
vapor. The heat required to boil up this
incoming water could be supplied by the
heat of combustion of the organic if the
organic is at a high enough concentration.
The water vapor leaving the reactor is
then condensed. The condensate may re-
quire subsequent treatment, depending on
the particular waste profile being oxi--
dized. All other equipment units function
the same as the organic residue process
previously described.
Economics
Capital and operating costs were estimated
for several catalyzed oxidation processes
using the experimental destruction rates
obtained in this study. The processes
were designed for compounds with dif-
ferent destruction rates to show the
impact of destruction rates on the process
design and economics. The design basis is
a 1000-gal reactor system for destroying
a pure compound in water. 'By fixing the
system design at 1000-gal, the major
factor affecting the cost of the treat-
ment system is the destruction rate of
the organic substance. For example the
feed rate of organic for a pentachloro-
phenol treatment system is 2.3 times the
feed rate for a butyl phthalate treatment
system. Therefore the corresponding
treatment cost per pound of organic would
be greater for the butyl phthalate treat-
ment system. However, due to the added
constraint of removing hydrochloric acid
from the pentachlorophenol system, .the
operating cost of this system is almost
twice the cost of the butyl phthalate
system. Therefore, the cost per pound :
of organic of ..the pentachlorophenol treat-
ment system is correspondingly higher and
approaches the cost of the butyl
phthalate treatment system* Other minor
factors affecting the treatment costs
include air compressor duty, condenser
size, cooling water requirements,
neutralization or scrubbing requirements,
and catalyst losses.
The cost for treating waste organic resi-
dues similar to the compounds evaluated
in this program ranged from $0.12/lb of
organic for compounds with fast de1-
struction rates, such as pentachlorophenol
to $1.04/lb for compounds with slow de-
struction rates such as hexachlorobuta-
diene. 'A second type of process was
designed to permit the treatment of •
aqueous waste contamined with high levels
of a soluble organic compound. This
process assumes the destruction rate of
acetonitrile in a 7.25% aqueous solution
and results in costs of $0.45/lb of
organic or $0.27/gal of waste.
Funds for this research came from the
program budget of the former Solid and
Hazardous Waste Division of the USEPA
Minicipal Environmental Research
Laboratory, with that program responsi-
bility now part of the Industrial Envi-
ronmental Research Laboratory in
Cincinnati, Ohio, Contract No. 68-03-2568,
Work Directive T-7016.
REFERENCES .
1. Diesen, R. W. and J. R. Moyer, "Wet
Combustion of Organics," U.S. Patent
3,984,311 (October 5, 1972).
2. Miller, R. A., "Destruction Method
for the Wet Combustion of Organics,"
U.S. Patent 4,276,198 (June 30,
1981). ,
221
-------
SHIPBOARD INCINERATION OF HAZARDOUS CHEMICAL WASTE
Routine Disposal of Liquids
and
Experimental Destruction of Solids
Gerald 0. Chapman
U.S. Environmental Protection Agency
Washington, D.C. 20460
Robert J. Johnson
TRW Environmental Division
Redondo Beach, CA 90278
Daniel W. Leubecker
Maritime Administration
Washington, D.C. 20590
Donald A. Oberacker
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
In February 1980, an interagency work group undertook a study of at-sea incineration
and the alternatives available to the Federal Government for encouraging the design,
construction, and operation of U.S. flag incinerator ships. The group examined previous
incineration operations, various federal assistance programs, safety and control measures,
Incinerator ship conceptual designs, environmental impacts, and waterfront facilities. In
September 1980, the work group concluded that chemical waste incineration at sea aboard
specially designed and equipped vessels is a cost-effective, technically efficient, and
environmentally acceptable technology for destroying many combustible hazardous wastes.
The work group was expanded in October 1980, designated the Interagency Review Board
for the Chemical Waste Incinerator Ship Program, and now coordinates and expedites all _
Federal Government activities related to developing an incineration at sea capability in
the United States.
This paper presents the findings of the work group and the work program which the
Interagency Review Board has initiated. Important ship design factors, such as the regu-
latory requirements, incinerator technologies, and incinerator system research recommenda-
tions, are explained. A conceptual design for a dual-mission ship, that can incinerate
both liquid and solid wastes is presented. Anticipated operations permits, environmental
monitoring, vessel design requirements, and waterfront facilities are discussed.
INTRODUCTION
The United States currently confronts
a serious and massive hazardous materials
disposal problem. The public health and
the nation's environment are being
threatened by ever increasing volumes of
hazardous wastes. The Environmental
Protection Agency (EPA) has estimated that
tens of millions of tons of hazardous waste
are generated annually across the nation.
Furthermore, there are thousands of disposal
222
-------
sites throughout the country being improp-
erly maintained, and a large number may
pose health problems. (1-6)
These environmental and public health
problems have prompted federal agencies to
examine various technologies for treatment
and destruction of waste materials. Con-
trolled high temperature incineration,
whether on- land or at sea, is the
most effective method available for
the destruction of combustible hazardous
wastes. Incineration at sea removes the
destruction site from populated areas and
the vulnerable freshwater environment,
which is of special importance when in-
cinerating the most toxic wastes.
The U. S. faces a deficiency in
incineration capacity for chlorinated
wastes as well as a difficulty in siting
new or expanded incinerators for certain
high toxicity or problem wastes. Incin-
eration at sea could provide a significant
portion of the needed disposal capacity
for short-term as well as a continued
long-term disposal option over the next
15 to 30 years. '
Several foreign flag incinerator
ships have been operating in Western Europe
for about a decade. In 1974, 1975 and 1977
three officially sanctioned U.S. at-sea
incineration operations were successfully
conducted on the foreign incinerator ship,
M/T Vulcanus. (The Vulcanus was purchased
by a U.S. firm, Chemical Waste Management,
Inc., in 1980). Eight shiploads of indus-
trial organochlorine wastes and three
shiploads of Air Force.Herbicide Orange
were destroyed under research, special,
and interim permits issued by EPA. These -
operations and those in Europe have demon-
strated incineration at sea of liquid com-
bustible chemical wastes to be an available
industrial operation which can be implemen-
ted in the U.S. without much of the pre-
liminary testing that other ultimate dis-
posal alternatives may require. (7-16)
In February 1980, an interagency work
group (including EPA, Maritime Administra-
tion, Coast Guard, and National Bureau of
Standards) undertook a study of at-sea in-
cineration technology and the alternatives
available to the Federal Government for •
encouraging the design, construction, and
operation of U.S. flag incinerator ships.
The work group examined previous incinera-
tion operations, various federal assistance
programs, safety and control measures, in-
cinerator ship conceptual designs, environ-
mental impacts, and waterfront facilities.
In September 1980, the work group issued its
report (17) concluding that chemical waste
incineration at sea aboard specially
designed and equipped ships is a cost-
effective, technically efficient, and en-
vironmentally acceptable technology for the
destruction of many types of combustible
hazardous waste.
INCINERATOR SYSTEMS AND RELATED RESEARCH
The EnvironmentaLProtection Agency
and its predecessor organization have con-
ducted federally supported incineration
studies for some 15 years, including 7
years of engineering and scientific studies
of hazardous waste incineration technolo-
gies, both on land and at sea. These
studies have supported increasingly strin-
gent incineration performance and emissions
standards of federal, state, and local
agencies. Such standards have helped
eliminate unsafe incineration practices and
have encouraged the development and use of
improved technologies to protect public
health and the environment.
EPA's work on incineration at sea be-
gan in 1974 and continued in 1977, when the
Agency permitted and monitored the burning
of U.S. industrial and military wastes
aboard the incinerator ship M/T Vulcanus.
These organochlorine wastes were incinera-
ted at sea with no detectable impact on the
environment. The liquid injection incinera-
tion equipment on board the ship is simi-
lar in basic design to the best available
land-based incineration equipment. The
major difference between the two technolo-
gies is that land-based incinerators for
chlorinated materials must have scrubbers
to remove effluent acid gases, whereas the
marine environment neutralizes the acidic; .
stack emissions from the at-sea incinera-
tion process.
Applicable Hazardous Waste
Orie advantage of incineration is that
the energy content of the chemical wastes
can be used to maintain combustion, so that
wastes assist the destruction process. High.
energy wastes can even be destroyed with-
out using supplemental fuel.. The EPA has
studied hazardous waste types and incinera-
tors which can handle different wastes.
223
-------
Chemical waste streams suitable for thermal
destruction fall into four classes:
1. C-H and C-H-0 compounds, yielding
C02 and HgO
2. C-H-N and C-H-O-N compounds, yielding
COgj HpO, and nitrogen oxides
3. C-H-C1 and C-H-0-C1 compounds, yield-
ing COp. HgO, and HC1 (gas)
4. Other wastes including organic com-
pounds containing both nitrogen and
chlorine; compounds containing sul-
fur, bromine, fluorine, phosphorus,
or silicon; and varied wastes not
included in the first three major
classes
The candidate wastes for incineration
at sea may be either pumpable liquids or
slurries, sludges, tars, or discrete solids.
The liquids and slurries are highly vari-
able mixtures, with specific gravities as
low as .85 or as high as 1.5. The solid
wastes also have highly variable physical
properties, but they must be suitable for
a screw feed or a waste container. A ship
that has both liquid injection and rotary
kiln incinerators could destroy virtually
any organic waste stream, except those "con-
taining more than "trace quantities" of
heavy metals or other nonincinerable
materials which would not be permitted for
incineration at sea without appropriate
air pollution control devices installed.
Incineration Equipment
Available marine liquid injection in-
cinerators (Figure 1) can safely and effec-
tively dispose of combustible liquid or-
ganic chemical wastes. Similar units are
used in land-based operations and on the
M/T Vulcanus and the K/B Vesta. Although
fluidized bed, multiple chamber, and molten
salt incinerators are also available today,
these designs are not as well developed,
proven, or widely used as is liquid injec-
tion technology.
The EPA seeks to advance at-sea incin-
eration by investigating the thermal de-
struction of certain types of solid hazard-
ous wastes including dry flowable granules,
containerized wastes, and nonpumpable
slurries or semi-solid materials. For
destroying these materials, rotary kiln
incinerators (Figure 2) have been used for
decades in land-based applications. .How-
ever, at-sea applications would constitute
WASTE TYPES
PUMPABLE LIQUIDS
SLURRIES. SLUDGES
TARS
SOLIDS:
— GRANULAR
— IRREGULAR
— CONTAINERIZED
MAXIMUM OPERATING
TEMPERATURE. C
MAINTENANCE
COMMERCIAL APPLICATIONS
VES
NO
NO
NO
NO
NO
1600
LOW
WIDELY USED FOR LIQUID WASTES
CURRENTLY USED FOR AT-SEA
[INCINERATION
Figure 1. Liquid injection incinerator.
a unique mobile environment for these units,
and a research or demonstration period
would be necessary. Also, unlike the
relatively "ash-free" or "clean-burning"
incineration of liquid wastes, many solid
wastes contain significant amounts of
uncombustible ash materials which must be
retained on board and properly disposed of
after incineration in the rotary kiln. A
stack scrubber might also be needed to pre-
vent particulates from being released with
the flue gases.
WASTE TYPES
PUMPASLE LIQUIDS
SLURRIES SLUDGES
TARS
SOI IDS:
— GRANULAR
— IRREGULAR
— CONTAINERIZED
MAXIMUM OPERATING
TEMPERATURE. C
MAINTENANCE
COMMERCIAL APPLICATIONS
VES
VES
VES
VES
VES
VES
1600
MEDIUM
WIDELY USED ON LAND FOR ALL WASTES
Figure 2. Rotary kiln incinerator.
224
-------
Recommended Research
A U.S. flag incinerator ship would pro-
vide a safe facility to research hazardous
waste incineration. To advance shipboard
incineration at sea, as well as other
vital technology development, the following
needs have been identified:
1. Demonstrate full-scale rotary kiln
incineration of solid hazardous waste
aboard ship;
2. Develop seawater scrubber technology
for shipboard incinerators; and
3. Investigate shipboard use of emerging
technologies at pilot-scale, such as
fluidized bed, molten salt, and'
oxygen-blown incineration and various
air pollution control devices.
These studies would take advantage of
available space on an incinerator ship
along with supplies of wastes and other
facilities. A rotary kiln demonstration
program would cost approximately $1.5
million with its associated construction,
installation, and engineering evaluation
studies. A pilot-scale seawater scrubber
program would cost $500 thousand or more.
Both programs conducted concurrently would
be completed .in 12 to 18 months.
INCINERATION AT SEA PERMITS AND
ENVIRONMENTAL MONITORING REQUIREMENTS
Permits for Incineration at Sea
The Environmental Protection Agency
has permit granting authority for all ocean
dumping except dredged material; this
authority includes incineration at sea.
Ocean Dumping Regulations (18) describe
the criteria for permit application and
procedures for issuing or denying permits
for ocean dumping. EPA regulates incin-
eration at sea according to the Marine
Protection, Research and Sanctuaries Act
(MPRSA) (19) and the London Dumping Con-
vention (20), applying the Intergovern-
mental Maritime Consultative Organization
(IMCO) mandatory regulations (21) and tak-
ing full account of the IMCO technical
guidelines (22). A Research Permit is
granted for the initial burn of a particu-
lar waste and requires extensive monitor-
ing. Special permits, for more routine
incineration operations, are granted for
a maximum of three years with an option for
renewal.
A current revision of the Ocean Dump-
ing Regulations (18) is progressing which
will align these rules more closely with the
needs of the Chemical Waste Incinerator
Ship Program. In addition, further revi-
sions are under discussion which will lead
to the development and implementation of
policies and procedures for the mixing,
blending, and treating of multiple waste
streams. It is also acknowledged that the
present permit system is time-consuming,
costly, and inflexible. A solution to this
regulatory problem in support of incinera-
tion at sea is of prime concern to EPA.
Under the present system, a permit
application must describe specified chemi-
cal and physical characteristics of the
waste and the results of specified tests
done on the waste. The method of incinera-
tion, the proposed incineration location,
and any special considerations regarding
probable impacts of the disposal must also
be included. A permit application is re-
viewed in consultation with other federal,
state, and local agencies. After a tenta-
tive determination is reached, time is
allowed for public comments, which may
include a public hearing. When all infor-
mation has been collected and evaluated,
a final determination is made and announced.
Administrative appeal procedures and re-
source to legal authority are available if
anyone objects to the action of the per-
mitting authority.
The design and operation requirements
for land-based and shipboard incineration
systems are divided into three main cate-
gories: regulatory, technical, and safety.
(23) The regulatory requirements include
the ocean dumping permit requirements, the
Resource Conservation and Recovery Act (24)
requirements which regulate the land-based
portion of the integrated system, and the
pertinent Coast Guard and IMCO requirements.
IMCO mandatory regulations and technical
guidelines are considered as technical re-
quirements, and all may appear in the per-
mit.
Open ocean sites for at-sea incinera-
tion are selected to minimize the inter-
ference of waste disposal activities on
the marine environment. Disposal site
evaluations are based on EPA criteria. The
results of these studies are presented as
an environmental assessment of,the impact
of using the site for disposal. An environ-
mental impact statement is subsequently
prepared for each site designation. (8,25)
225
-------
A site has been designated in the Gulf of
Mexico and one has been proposed and evalu-
ated for the North Atlantic.
Environmental Monitoring
Environmental monitoring during at-sea
incineration is imposed by the MPRSA regu-
lations and the IMCO mandatory regulations
adopted under the London Dumping Conven-
tion. The IMCO mandatory regulations,
which went into effect in the United States
in March 1979, impose minimum operational
monitoring requirements for protection of
the marine environment. Additional re-
quirements may be in the permit for at-sea
•incineration. For example, specific per-
mit monitoring requirements were imposed
on the at-sea disposal actions for organo-
chlorine waste and Herbicide Orange. Two
general cases involving different degrees
of environmental monitoring are discussed
below.
Initial Incineration Monitoring—
The London Dumping Convention and the
MPRSA require a survey during the first
use of an at-sea incineration facility to
determine compliance with the regulations.
The survey requires stack gas sampling and
analysis; monitoring of the stack gas for
CO, C0£» 02, total hydrocarbons, and halo-
enated organics; and combustion and de-
struction efficiencies of at least 99.9
percent. The survey must be repeated every
two years.
Monitoring in excess of the minimum
regulatory requirements would be performed,
for example, under the following conditions:
1. An unusually hazardous waste (e.g.,
PCBs, which may contain TCDFs) or a
waste containing an unusually hazard-
ous substance (e.g., Herbicide Orange
and its 2,3,7,8-TCDD contaminant) is
to be incinerated.
2. A new type of waste is to be incin-
erated, unless it is deemed similar
to a previously tested waste.
3. A new incinerator or a new type of
incinerator is used, or extensive
system component changes are made.
When any of these three situations
occur, it is expected that detailed stack
sampling followed by shipboard or land-
based analysis of the samples will be
required. Additionally, monitoring of
solid residue from a rotary kiln incinera-
tor and influent and effluent from a
scrubber would be required.
When waste containing appreciable
amounts of sulfur and/or nitrogen (e.g.,
greater than 5 percent) are incinerated,
stack gases should be monitored for oxides
of sulfur and/or nitrogen. Additionally,
during any research phase, oxides of
nitrogen should be monitored to establish
baseline values. In some cases, seawater
sampling could be required to determine
actual impacts of incinerator effluents
on the sea and on marine organisms. For
example, marine water and organism samp-
ling was performed during several Research
Permit burns of organochlorine waste in the
Gulf of Mexico, and marine waste samples
were taken during the Research Permit burn
of Herbicide Orange in the Pacific Ocean.
Routine Incineration Monitoring—
The MPRSA regulations and London Dump-
ing Convention mandatory regulations do not
require environmental monitoring during
routine operations. These regulations do,
however, require operational monitoring.
The operational monitoring requirements are:
1. Flame temperature not less than
1250°C, unless studies on the incin-
erator have shown that a lower tem-
perature will achieve the required
combustion and destruction efficien-
cies.
2. Combustion efficiency is at least
99.95 + 0.05%, based on
Combustion
Efficiency
= 100 x
:co2 - cco
CC00
Mhere CGO? and Cco are, respectively,
concentrations of carbon dioxide and
carbon monoxide in the stack gas.
3. No black smoke or flame extension
above the exit plane of the stack.
The London Dumping Convention .manda-
tory regulations also direct that Contrac-
ting Parties "take full account of the
technical guidelines." The IMCO technical
guidelines establish additional operational
standards. Several relevant guidelines are:
226
-------
1. Minimum 3 percent oxygen in stack gas
near the exit plane of the stack.
2. Incinerator wall temperature should
be not less than 1200°C (unless tests
on the unit have shown that adequate
waste destruction can be achieved at
lower wall temperatures).
3. Residence time of all wastes in the
incinerator should be of the order
of one second or longer at a flame
temperature of 1250°C.
4. Waste type and rate of input to the
incinerator should be recorded.
WATERFRONT SUPPORT FACILITIES
Waterfront integrated hazardous waste
management facilities are necessary to
support incinerator ships. Such a facility
would include waterfront storage tanks;
waste receiving, processing, and handling
equipment; a laboratory for waste analysis,
an'd a transfer terminal 'for loading wastes
aboard ship. In addition, an inland trans-
portation system is required to safely
haul wastes to the waterfront facility.
(16,17)
The waterfront facility must:
1. Receive liquid and solid hazardous
wastes either by land or by water-
borne barge transport;
2. Analyze, blend, shred, and process
the materials as appropriate for in-'
cineration at sea;
3. Load the waste aboard ship in a safe
and efficient manner; and
4. Remove and receive residues from the
incinerator ship for analysis and
disposal either on land or at sea, in
the case of incineration of wastes
producing a collectable residue dur-
ing disposal.
The facility will accommodate waste in
almost any physical form and in several
types of containers, some of which may be
older, corroded, and possibly leaking.
Ideally, the facility would service three
transportation modes for delivery of
wastes: truck, rail, and barge. It should
consecutively accommodate at least two in-
cinerator ships, each on a 10 to 14 day
cycle. Figure 3 presents a conceptual
layout of the facility.
Liquid waste, solid waste, and ash
residue from incineration will be pro-
cessed and stored separately. Liquid waste
in drums and other containers will be sent
through a shredder Tn the dedrumming facil-
ity. Liquid from both the containers and
the decontamination of the containers will
be blended to optimize transfer and combus-
tion processes and pumped to storage tanks.
Liquid waste arriving in tank trucks or
tank cars, along with the tanker decontami-
nation rinse, will also be blended and
pumped to the storage tanks.
Solid waste arriving at the site will
be unloaded at the unloading rack, prepared
for incineration by shredding, and placed
in bulk material containers to be loaded
on the ship. Any ash residue from the at-
sea burn will be returned to the waterfront
facility and kept in the residue storage
area until removed for ultimate disposal,
probably in a landfill approved for hazard-
ous waste disposal.
DUAL-MISSION INCINERATOR SHIP
Design Alternatives and Operating Scenario
In addition to a conventional ship,
several alternative marine technologies
may be viable for an incineration at sea
system capable of destroying solid as well
as liquid waste, including:
1. Integrated tug/barge unit.(a ship-
shaped barge propelled by a specially
designed pushing tug that.is mechani-
cally linked to the barge stern).
2. A towed tug/barge combination (a
traditional tug and tethered incin-
erator barge).
3. A ship conversion (an existing tanker
may be the most easily converted
ship).
4. A tug/supply vessel (using either
existing vessels or new vessels).
A conventional new ship design was
selected, -based on the following operating
scenario:
1. Two week operating cycle, with ten
days "on-site;"
-2. Both liquid and solid wastes accepted;
3. Three liquid incinerators operated
continuously at full capacity for ten
227
-------
i TANKER UNLOADING— •
O O @C8"6AN©OOO
I Illiiliilliil minimi Illlllllllllllll l II il|ii|ii|!!i!!!!!!!!HMH
260 M
tBOOFT)
SOLID AND
CONTAINERIZED
LIQUID STORAGE
!R.R.
EXPANSION
FENCE
O
O
DIKE
RESIDUE
RECEIVING
AND
LOADING
RAMP
n
Figure 3. Waterfront facility layout.
days for routine destruction of
large volumes of liquid wastes;
One solid waste incinerator operated
for research purposes;
No "abandoned site" wastes accepted
without prior analyses;
Solid waste containerized before
loading onto ship;
Wastes loaded at terminal by auto-
matic equipment;
Drifting or slow steaming during
incineration operations; and
Heading into wind maintained by bow
thruster during incineration.
Adopting a dual-mission scenario has
resulted in a conservative conceptual
design that exceeds the needs of a vessel
built only for routine liquid incineration.
A first-generation U.S. flag commercial
incinerator vessel need not incorporate all
the features of this design.
Incinerator Ship Conceptual Design
A typical mild steel tankship hull,
as shown in Figure 4, has been equipped
with extensive auxiliary processing sys-
tems to handle and incinerate liquid and
solid hazardous chemical waste. The incin-
eration-plant, liquid cargo pumps, and
228
-------
If?
\=
\
8 §§
"
4&
§a
UJ-
:n
c
o
ra
c
o
-------
unattended electric propulsion motors arle
located aft, with the deckhouse, labora-'
tory, and attended propulsion diesel-gen-
erator machinery spaces located forward.
This arrangement locates personnel living
and working spaces as far as possible from
the incinerator plume, heat, and potential
chemical contamination. Design specifica-
tions of the dual-mission incinerator ship
are presented in Table 1.
Liquid waste tanks and solid waste
container storage are midships. The'four
center!ine tanks provide IMCO Type I cargo
protection. The eight port and starboard
tanks, which are protected by wing ballast
tanks, are ample for Type II cargo protec-
tion. Type I chemicals are products that
require maximum preventive measures to pre-
clude escape of such cargo; Type I pro-
ducts require significant preventive mea-
sures to preclude escape. (26,27) The
large amount of ballast tankage can be
sequentially filled to help the ship main-
tain a constant draft and uniform ship
motions as wastes are incinerated. Solid
waste containers are stored on deck,
directly above the liquid waste tanks, but
separated from those tanks by a three foot
high inert gas-filled void across the full
width of the cargo area.
The rotary/kiln liquid injection in-
cinerator combination is on the ship center-
line, with the two liquid injection incin-
erators located out-board of the kiln's
forward end. The rotary kiln is an experi-
mental facility and the problems of safely
loading, handling, and incinerating the
solid wastes have influenced the overall
design in several respects. The kiln axis
on the ship center!ine minimizes the effect
of ship's roll on the waste residence time.
The estimated maximum ship pitch angle of
about 2 1/2 degrees is unlikely to disrupt
the rotary kiln's operation. The ship's
vibrations and roll motions impose loads on
the kiln's external rotary drive, but two
major induced vibration sources, the pro-
pulsion motors and the propeller, would be
idle much of the time that the rotary kiln
is operating. The diesel generators,
another major source of induced vibration*
are forward, remote from the incinerators.
Therefore, ship source vibrations should
not disrupt the rotary kiln. Isolation
mountings and a modified rotary drive can
be provided to alleviate the effect of
vibration and roll.
The ship is powered by a diesel elec-
tric propulsion plant, providing a service
speed of 12 Kts. The propulsion motors
must be located aft, but the diesel gen-
erators are below the forward deckhouse,
so that the machinery space, where the
engineers will work most of the voyage, is
separated from both the incineration sys-
tem and the waste cargo. The same diese!
generators power the incineration plant
when the propulsion plant is idle or opera-
ting at reduced speeds. This "power pool"
arrangement makes possible a single instal-
lation, rather than separate ship's propul-
sion and ship's service generator installa-
tions.
TABLE 1. INCINERATOR SHIP DESIGN SPECIFICATIONS
Length Overall
Beam, Molded
Draft, Full Load
Range
Liquid Waste Capacity
Solid Waste Capacity
Incinerator Specifications
Number
Flame Temperature
Capacity
Residence Time
Up to 1600°C
10 mt/hr (each)
1 to 1.5 sec
129.5m (425'-0")
23.8m (78'-0")
7.5m (24'-7")
5000 Nautical miles
7200 mt (7088 long tons)
360 mt (355 long tons)
Solid
1
Up to 1600°C
1.5 mt/hr
Minutes to hours
230
-------
TABLE 2. ESTIMATED SHIP CONSTRUCTION AND OPERATING COSTS
Contract Drawings
Construction Contract
(1)
Plan Approval, Inspection, and Spares
$ 1 million
$50- million
$ 1 million
Projected Ship Cost ^ '
fo\
Operating Cost ^ '
Ship Related Disposal Cost
Total $52 million
$75 million
$ 1.1 million/14 day voyage
$145/metric ton of waste
(1) Includes $14 million for subcontracted incineration systems.
(2) Based on October 1982 contract award and March 1985 vessel delivery, assuming
1no/ annual inflation rate.
U.S.-flag incinerator ship de-
velopment;
2- Establishment of a coordinated
certification and permitting pro-
gram for an applicant who wishes
to establish a commercial incin-
eration at sea service;
3. Encouraging and assisting state and
local authorities to develop water-
front integrated hazardous waste
management facilities to support in-
cinerator ships;
4. Development of technical requirements
based on existing international
standards and national regulations
for the design and operation of in-
cinerator ship systems;
5. Planning and conducting research to
advance the state-of-the-art of in-
cineration at sea, e.g., polymeric
liner materials for containment of
hazardous wastes, kinetics of the in-
cineration process, 'and incineration
of solid wastes at sea;
6. Evaluation of loan guarantee appli-
cations from private industry for
U.S.-flag incinerator ship develop-
ment; and
7- Environmental assessment and desig-
nation of ocean incineration sites.
Early in 1981, the National Advisory
Committee on Oceans and Atmosphere (NACOA)
(3) Based on 1985 start of operations.
Estimated Costs and Schedule
Ship construction cost is conserva-
tively estimated at $75 million for vessel
delivery in 1985 (Table 2). Thirty months
total construction time is required.
Operating cost includes ship capital cost
depreciated over 15 years, but does not
include waterfront terminal operations or
waste land transportation costs.
CONCLUSION
Incineration at sea aboard specially
designed and equipped ships has been deter-
mined to be a cost-effective, technically
efficient, and environmentally acceptable
technology for the destruction of many
types of combustible hazardous waste, in-
cluding chlorinated organic chemical and
petrochemical wastes (17). An Interagency
Review Board for the Chemical Waste Incin-
erator Ship Program has been formed to co-
ordinate and expedite all Federal Govern-
ment activities related to incineration at
sea. (28) Principal participants at this
time are the Environmental Protection
Agency, Maritime Administration, Coast
Guard, and National-Bureau of Standards.
The work program of the Interagency Review
Board includes the following projects:
1. Investigation and pursuit of viable
legislative amendments to provide
additional federal assistance for
231
-------
released a Report to the President and
Congress entitled, "The Role of the Ocean
In a Waste Management Strategy." (29)
This special report is the culmination of
a two-year effort by NACOA to formulate
recommendations on use of the ocean as a
waste disposal medium. NACOA is firmly
convinced that the United States must
reexamine present national policies and
regulations concerning waste disposal in
the oceans, and notes a high probability
that land, deepwell, and atmospheric waste
disposal activities will be reduced during
the 1980's in favor of ocean waste dis-
posal . NACOA recommends that ocean dis-
posal of industrial wastes should continue
at sites where evidence indicates no un-
reasonable environmental degradation and
when human health, environmental, and
economic considerations indicate this is
the preferable option. It is the opinibn
of this paper's authors that chemical
waste incineration at sea, when properly
regulated and control!edj thoroughly
satisfies these NACOA criteria and will be
a preferred, readily available, disposal
option for many hazardous wastes.
ACKNOWLEDGEMENTS
The authors wish to express their
sincere gratitide to the members of the
Interagency Review Board for the Chemical
Waste Incinerator Ship Program ahd its
predecessor, the Interagency Ad Hoc Work
Group, for their contributions to this
program.
REFERENCES
1. , U.S. Environmental Protection Agency.
EPA Journal, Hazardous Waste Fact
Sheet, Volume 5, Number 2, February
1979, page 12.
2. U.S. EPA. Information prepared for
the U.S. Senate Subcommittee on Health
and Scientific Research, June 6, 1980.
3. U.S. EPA. EPA Activities Under the
Resource Conservation and Recovery
Act of 1976, Annual Report to the
President and the Congress for Fiscal
Year 1978, SW-755, March 1979.
4. U.S. EPA. Everybody's Problem -
Hazardous Waste, SW-286, 1980.
5. ILS. EPA. EPA Journal, Cleaning Up
in New Jersey, Volume 6, Number 6,
June 1980, pages 10-11.
6. U.S. EPA. Siting of Hazardous Waste
Management Facilities and Public
Opposition, SW-809, November 1979.
7. Wastler, T.A., C.K. Offutt, C.K.
Fitzsimmons, and P.E. Des Rosiers,
Disposal of Organochlorine Wastes by
Incineration at Sea. EPA-430/9-75-
014, July 1975.
8. U.S. EPA. Final Environmental Impact
Statement - Designation of a Site
in the Gulf of Mexico for Incinera-
tion of Chemical Wastes. EPA-EIS-WA
76X-054, July 8, 1976.
9i Clausen, J.F., H.J. Fisher, R.J.
Johnson, E.L. Moon, C.C. Shih, R.F.
Tobias, and C.A. Zee, At-Sea Incinera-
tion of Organochlorine Wastes On-
board the M/T Vulcanus. EPA-600/2-
77-196, September 1977.
10. Ackerman, D.G., H.J. Fisher', R.J.
Johnson, R.F. Maddalone, B.J.
Matthews, E.L. Moon, K.H. Scheyer,
C.C. Shih, and R.F. Tobias, At-Sea
Incineration of Herbicide Orange
Onboard the M/T Vulcanus. EPA-600/
2-76-086, April 1978.
11. Paige, S.F., L.B. Baboolal, H.J.
Fisher, K.H. Scheyer, A.M. Shaug, R.L.
Tan, and C.F. Thome, Environmental
Assessment: At-Sea and Land-Based
Incineration of Organochlorine Wastes.
EPA-600/2-78-087, April 1978.
12. Shih, C.C.j J.E. Cotter, D. Dean,
S.F. Paige, E.P. Pulaski, and C.F.
Thorne, Comparative Cost Analysis
and Environmental Assessment for
Disposal of Organochlorine Wastes.
EPA-600/2-78-190, August 1978.
13. U.S. Environmental Protection Agency,
U.S. Department of State, Final
Environmental Impact Statement for the
Incineration of Wastes at Sea Under
the 1972 Ocean Dumping Convention,
February 9, 1979.
14. U.S. Department of Commerce/Maritime
Administration Chemical Waste Incinera-
tor Ship Project, MA-EIS-7302-76-
041F, July 2, 1976.
232
-------
15. Halebsky, M., A Study .of the Economics
and Environmental Viability of a U.S.
Flag Toxic Chemical Incinerator :Ship.
MarAd Report No. 04068-002, NTIS Report
No. PB 291932to4, December 1978.
16. Martinez, L.A., Hazardous Chemical In-
cineration at Sea: A Disposal Alter-
native for the United States, Unpub-
lished Research Report Supported by
the Maritime Administration (M.S.
Thesis at the Massachusetts institute
of Technology), February 1980.
17. U.S. Environmental Protection Agency,,
U.S. Department of Commerce/Maritime
Administration, U.S. Department 'of
Transportation/Coast Guard, U-.S.
Department of Commerce/National Bureau
of Standards, Report of the Ad Hoc
Work Group for the Chemical.Waste
Incinerator Ship Program, NTIS Report
No. PB-81-112849, September 1980. '
18. U.S. EPA. Code of Federal Regulations,
Title 40, Subchapter H - Ocean Dumping.
19. Marine Protection, Research, and
Sanctuaries Act, 1972, (Public Law
92-532), as amended by Public Law
93-254. •.:
20. International Convention on the,
Prevention of Marine Pollution by
Dumping of Waste and Other Matter,
1972, as amended.
21. IMCO Mandatory Regulations for the
• Control of Incineration of Wastes
and other Matter at,Sea, 1978,
(Addendum to Annex I of the London
Dumping Convention).
22. IMCO technical Guidelines on the Con-
trol of Incineration of Wastes' and
Other Matter at Sea, 1979. '',
23. Johnson, R.J., D.G. Aekermanv,and
L.L. Scinto, Preliminary Criteria for
Evaluation of At-Sea Incineration
Activities, Volumes I and II. EPA
Contract No. 68-02-3174, Work
Assignment No. 61, November 1981.
24. Resource Conservation and Recovery
Act, 1976, (Public Law 94-580).'
25. U.S. Environmental Protection Agency,
Draft Environmental Impact Statement
for the Proposed North Atlantic
Incineration Site Designation,
October 1980.
26. IMCO' Code for the Construction and
Equipment of Ships Carrying Danger-
ous Chemicals in Bulk, as amended, '
- (IMCO Resolution A212 (VII)).
27. U.S... Coast Guard. Code of Federal ;
Regulations, Title 46, Part 153•- ;;.
Safety Rules for .Selt-Propelled
Vessels Carrying Hazardous Liquids.
28. Interagency Memorandum of Agreement
... . from the Administrator, U.S. Environ-
mental. Protection Agency., and the , .
Assistant Secretary for Maritime '
Affairs, U.S-. Department of Commerce,
Report .of the Interagency Ad Hoc Work
Group for the Chemical Waste Incinera-
tor Ship Program, October 8, 1980.
29.- National Advisory Committee on Oceans;
and Atmosphere, A Special Report to the
President and Congress - The Role of
the Ocean in a Waste Management
Strategy, January; 1981.
233
-------
ELIMINATION OF HAZARDOUS WASTES BY THE
MOLTEN SALT DESTRUCTION PROCESS
James 6. Johanson, Samuel J. Yosim, Larry G. Kellogg, Seymour Sudar
Rockwell International, Energy Systems Group
Canoga Park, California 91304
ABSTRACT
The Resource Conservation and Recovery Act (RCRA) imposes strict regulations on the
management of hazardous wastes. Disposal of the wastes by any means that does not
result in conversion of the waste matter to ecologically harmless materials sets up a
continuing responsibility for prevention of damage to public health and to property.
Under RCRA, the waste generator must bear this continuing responsibility with its con-
comitant liability potential. The generator's responsibility, however, can be terminated
by destruction of the waste material in conformance with RCRA regulations, which require
at least 99.99% destruction and removal efficiency (ORE) of all principal organic
hazardous constituents (POHC). An exception to this rule is polychlorinated biphenyls
(PCBs) for which a minimum of 99.9999% ORE is required. To meet these stringent require-
ments, much attention is being given to high-temperature combustion systems for hazardous
waste treatment. This paper describes the Rockwell Molten Salt Destruction (MSD) process
and recent experimental results which demonstrate the high degree of destruction achieva-
ble with the MSD process. These experiments were conducted with the support of the EPA,
Incineration Research Branch, Contract 68-03-3014, TMS II, Task 21, Subtask 1. These
tests emphasized the treatment of chlorinated hydrocarbons with MSD. Results of prior
MSD tests performed with a variety of hazardous waste materials are also described herein.
MOLTEN SALT DESTRUCTION PROCESS
This process is based on the use of a
molten salt, such as sodium carbonate, as
a heat transfer and reaction medium. In
the process, combustible waste and air are
continuously introduced beneath the surface
of a sodium carbonate-based salt at 1450 to
1850°F. Figure 1 shows the molten salt
combustor schematically. The combustible
material is added in such a manner that any
gas formed during combustion is forced to
pass through the melt before it is emitted
into the atmosphere. The off-gas contain-
ing carbon dioxide, steam, nitrogen, and
unreacted oxygen is cleaned of particulates
in commercial applications by passing it
through a baghouse.
Acidic gases, such as HC1 and S02,
produced from halogenated organic compounds
and sulfur-containing organic compounds,
STACK
f C02.H2O.H2.02
REMOVED PARTICULATES
Figure 1. Schematic of molten salt
waste combustion system.
234
-------
respectively, are neutralized and absorbed
by the alkaline Na2C03- The ash intro-
duced with the combustible waste is also
retained in the melt. The char from the
fixed carbon is consumed in the salt. The
temperatures of oxidation are too low to
permit a significant amount of NOX to be
formed by fixation of the nitrogen in the
air.
The scrubbing function of the sodium
carbonate leads to the formation of other
sodium salts. Ash, introduced by the waste,
must be removed when it becomes excessive
to preserve the fluidity of the melt. An
ash concentration of about 20 wt % provides
an ample margin of safety for the melt
fluidity. In some applications, the melt
can be removed batchwise. In other applica-
tions, a side stream of the melt is with-
drawn either batchwise or continuously and
is processed. If desired, the melt may be
quenched in water, the solution filtered to
remove the ash, and processed to recover
the carbonate which is then recycled to the
combustor.
Sodium carbonate is used because it is
compatible with combustion products, C02
and H20, and because it reacts with acidic
gases such as HC1 (produced from organic
chloride compounds). It is stable, non-
volatile, inexpensive, and nontoxic.
Data are reported from tests conducted
in two different-sized systems. One system
is a 6-in.-diameter bench-scale unit for
feasibility and optimization tests (up to
approximately 3 Ib/h). The other system
is a pilot-scale unit (up to approximately
250 Ib/h) for obtaining engineering data
for the design of an actual plant.
DESTRUCTION OF HEXACHLOROBENZENE AND
CHLORINE
A program to test the destruction of
solid hexachlorobenzene (HCB) and liquid
chlordane by the MSD process was completed
for the Environmental Protection Agency
under Contract 68-03-3014, Task 21, Sub-
task 1. The HCB was selected as a stand-in
for PCBs; chlordane was selected as an
example of a liquid organic chlorinated
waste. The experimental .portion of the
program consisted of bench-scale tests
and pilot-scale tests.
Bench-Scale Tests
Test Apparatus and Waste Streams
The overall objective of the bench-
scale tests was to provide a data base
for the pilot-plant tests. A schematic
of the bench-scale combustor is provided
in Figure 2.
The hexachlorobenzene, mixed with
coke (70 wt % HCB - 30% coke) to increase
the heating value, contained (on a weight
basis) 98% HCB and 2% inerts. The chlor-
dane, mixed with kerosene (60 wt %
chlordane, as received -40% kerosene),
contained 61.1% carbon, 27.8% chlorine,
and 8.7% hydrogen.
HCB-coke feed rates varied from 1.7
to 3.2 Ib/h, and chlordane-kerosene feed
rates varied from 0.9 to 1.4 Ib/h. The
INJECTOR TUBE
AIR IN
1
xs^
I
_J
o
o
\.l
v
— Z.-— _~L
: SEALED 1
. HOPPER
il- ^
"— L "-I
^VIBRATOR
/^D— » OFF GAS OUTLET
_)jC (ON-LINE Sft
-
O
o
\j
MPLER .
ATTACHED HERE)
^sL— STAINLESS-STEEL
RETAINER VESSEL
^- 3.7 cm ID ALUMINA
FEED TUBE
*. MARSHALL FURNACE
^^15 cm ID
ALUMINA
TUBE
—-15 cm DEPTH OF
MOLTEN SALT
Figure 2. Bench-scale molten salt
combustor schematic.
235
-------
air feed rate was 2.5 scfm. The super-
ficial air velocity through the melt was
I ft/s. The NaCl content of the melt
increased with the amount of waste proc-
essed. It reached 88 wt % in the case of
HCB and 55 wt % in the case of chlordane.
Sampling and Analysis Procedures
Portions of the off-gas were sampled
with a sampling line consisting of:
(1) a heated glass fiber filter which re-
moved particulates, (2) a heat exchanger
which cooled the gas to a proper absorption
temperature, and (3) polyurethane foam
plug(s) which capture the HCB or chlordane
in the off-gas followed by toluene-contain-
ing impingers to trap any material which
escaped the polyurethane foam plug. These
samples, along with samples of the melt,
were analyzed for undestroyed HCB and
chlordane. The sampling and chemical
analyses for HCB and chlordane were done
in accordance with EPA-appr6ved modifica-
tions to EPA Methods 612 and 608, respec-
tively. A gas chromatograph equipped with
an electron capture detector was used in
this analysis. A schematic of the sampling
:rain for capturing undestroyed HCB and
chlordane is shown in Figure 3.
The analytical procedures described
below follow the methods recommended in
EPA Methods 608 and 612 for organochloride
pesticides.
Sample Extraction and Connection. The
polyurethane plug and glass fiber filter
were placed in Soxhlet extractors and
extracted with methylene chloride for
about 10 h. Melt samples were dissolved
in water. The resulting solutions were
then extracted with methylene chloride
using a separatory funnel. The extracts
were quantitatively transferred to Kuderna-
Danish (K-D) evaporators. When the apparent
volumes of each extract reached 1 to 2 ml,
the K-D apparatus was moved and allowed to
drain while cooling. Fifty (50) ml of
hexane was added to each extract, and the
resulting solutions were then concentrated
to a final known volume of 1 to 5 ml. Known
aliquots of impinger solutions and probe
washings were directly injected into the
gas chromatograph for analysis.
Gas Chromotographic Analyses. Analyses
of hexachTorobenzene and chlordane were
conducted using a chromatograph with an
electron capture detector. The chromato-
graphic column was a 6-ft x 4-mm-ID glass
column packed with 100/120 Supelcoport
coated with 1.5% SP2250 and 1.95% SP2401.
This column is recommended in EPA Method
608 for organochlorine pesticides and PCS
analyses. Confirmation of hexachloroben-
zene or chlordane was done on selected
samples by a gas chromatograph-mass spec-
trometer (6C-MS) using characteristic ions
at m/e 284 for HCB and m/e 373 for chlor-
dane. Only the electron impact mode was
used in this GC-MS work. Degradation pro-
ducts were also analyzed, and their mass
spectra was searched and compared to EPA/
NIH or special priority pollutant libraries
to identify the most probable chemical
compounds. The GC-MS analysis of the
samples has shown mainly the presence of
unburned hydrocarbons anthracene and
other polynuclears along with some chlor-
inated derivatives (during off-spec
conditions). Furans and dioxins were not
detected using conventional GC-MS analysis.
HCB Results
The results of the HCB destruction
tests are shown in Table 1 and include
the concentration of HCB in the off-gas
and in the melt as well as the destruc-
tion removal efficiency. The destruction
removal efficiency is defined as 100%
minus the percent of HCB in the feed
emitted to the atmosphere and thus does
not include the contribution from the
melt. No HCB was detected in any of the
melt samples, despite the very low detec-
tion limits (less than 1 ppb) for HCB.
Excellent destruction was obtained
in all cases. The best destruction
removal efficiency, greater than 7/9's
(>99.9999986%), was obtained at 1839°F
with 118% of stoichiometric air. Excel-
lent destruction removal efficiencies
were also obtained at 127% stoichiometric
air and 1656°F (>99.999974%) and at 70%
stoichiometric air and 1848°F (>99.999982%)
The last result indicates excellent
destruction even when excess air is not
used.
236
-------
cc
III
o
X UJ
JLL
£
CC
g
LU
237
-------
TABLE 1. SUMMARY.OF HCB RESULTS
Temp
(°F)
1839
1848
1656
% stoich
air
118
70
127
Cone of HCB
in off-gas
ppmv
<0. 00017
<0.0031
<0.0028
mg/m3
<0.0019
<0.037
<0.033
Destruction
Cone of HCB removal
in melt efficiency
(ppmw) («)
<0. 00100
<0. 00040
<0. 00110
>99. 9999986
>99. 999982
>99. 999974
No HCB was detected in any melt sample
even though melt samples were taken while
HCB was being fed. Thus, the concentration
of HCB in the melt was, at most, ,1 ppb in
spite of the short residence time of the
feed in the expanded melt (^3/4 s). This
is an excellent indication that the spent
melt is nontoxic.
The concentration of HCB. in the off-
gas was extremely low in all cases. In
the 1839°F test with excess air, the con-
centration of HCB was only 0.00017 ppmv or
0.0019 mg/m3. In the 1656°F test and in
the test with 70% of stoichiometric air,
the concentrations were higher by an order
of magnitude but still extremely low
(0.003 ppmv or 0.03 mg/m3).
Chlordane Results
The results of the chlordane destruc-
tion tests are shown in Table 2.
The sensitivity of the electron capture
detector is significantly less for chlordane
(60 ng/ml) than for HCB (1 ng/ml) due to the
relative chlorine contents of the molecules
and the number of different chlordane com-
pounds in the feed. Thus, the detection
limits were considerably higher than those
of HCB.
Excellent-chlordane destruction was
obtained in the excess air tests. In
these tests, two peaks with the same
elution times as two of the four charac-
teristic chlordane peaks were present.
These peaks were assumed to be chlordane
for the calculation of destruction removal
efficiencies. With this conservative
assumption, the destruction removal
efficiencies were greater than 99.9999%.
In the test with 81% of stoichiomet-
ric air, in which larger peaks were found,
three peaks with the same elution time as
three of the four characteristic chlordane
peaks were present in all samples except
the impinger and the melt sample. How-
ever, when the first plug was examined
by gas chromatograph-mass spectrometry,
the three peaks with elution times similar
to chlordane were identified as not due
to chlordane isomers. Thus, the~cFflordane
destruction efficiencies were probably
much greater than 99.99%, which was
calculated assuming the three "chlordane"
peaks were actually due to chlordane.
No chlordane was detected in any
melt sample even though melt samples were
taken while the chlordane was being fed.
Thus, the concentration of chlordane in
the melt was, at most, 12 ppb.
TABLE 2. SUMMARY OF CHLORDANE RESULTS
Temo
1930
1836
1648
% stoich
air
121
81
118
Cone of chlordane
in off-gas
ppmv mg/nv-'
<0. 00132 <0.0447
O.243 <4.08
<0. 00267 <0.0220
Cone of chlordane
in melt
(ppmw)
<0.0090
<0.0120
<0.0080
Destruction
removal
efficiency
(*)
>99. 9999951
>99.9934
>99. 999903
238
-------
The concentration of chlordane in the
off-gas was very low. In the 183QOF test
and in the 16480F tests (both with excess
air), the concentration of chlordane was
<0.001 and <0.003 ppm, respectively. The
concentration of chlordane in the test
with 81% of stoichiometric air was
<0.243 ppm. However, the fact that the
peaks (at least in the plug) could not be
confirmed as chlordane by the GC-MS indi-
cates that the chlordane concentration is
much less than this value.
Pollutant Concentrations from HCB
and Chlordane Tests
No chlorine or phosgene was detected
in the off-gas. The limits of detection
of chlorine and phosgene are 0.2 and
0.3 ppmv, respectively. The HC1 concentra-
tions were <100 ppm. No H2 or CH4 was
detected except with less than 100% of
stoichiometric air. As expected, the NOX
levels were very low (<50 ppm). The CO
levels were higher than desired (up to
2.4%) in the excess, air tests. Such high
values have been noted before in this lab-
oratory when chlorinated aromatic compounds
were destroyed in NagCOs melts containing
large amounts of NaCl. However, the CO
concentrations decreased markedly with
increasing melt temperature and with in-
creasing percent stoichiometric air.
While destruction efficiency and destruc-
tion removal efficiencies of 99.9999% and
better can be achieved with excess air at
1830QF and at 16480F, a temperature of
1830QF is preferred over 16480F since CO
emissions are much lower at the higher
temperature.
Pilot-Scale Tests
The Molten Salt Test Facility (MSTF)
was operated for a total of 96 h to provide
five data points each for demonstrating
destruction of HCB and chlordane. A
schematic of.the MSTF, including the two
sample locations used for these experi-
ments, is shown in Figure 3. Sample Loca-
tion 1 was immediately after the molten
salt destructor, i.e., just upstream of the
particulate drop box. Sample Location 2
was just downstream of the baghouse. The
feed rates of HCB and chlordane were as
high as 269 and 72 Ib/h, respectively, in
these tests. The tests were conducted at
melt temperatures from 1645 to 18160F and
from 78 to 151% of stoichiometric air.
Sampling for unreacted HCB and chlordane
was done in the exhaust of the molten
salt combustor, in the exhaust of the
particulate baghouse downstream of the
combustor, in- the residual baghouse salt
bottoms, and in the salt melt overflow
from the combustor.
The maximum total air feed rate
corresponded to a 2.7-ft/s superficial
velocity in the bed. Testing was done
in 0.5- to 2.0-h test periods with a
minimum 1 h of vessel purging and up to
3 h of process stabilization between
sampling periods. Destruction gas
sampling was done during each sampling
period. Auxiliary coke fuel was used
throughout the various waste burn periods
as required to maintain nominal operating
conditions. Between burn periods, coke
or diesel oil was the only combustible
material fed to the vessel. The salt bed
composition in all cases was maintained
at 50 wt % sodium chloride and 50 wt %
sodium carbonate to simulate nominal
steady-state bed compositions for organor-
chloride waste destruction.
HCB Test Results
The HCB feed rate varied from 46 to
269 Ib/h. Coke was used as auxiliary
fuel in all tests except 3B, where no
auxiliary fuel was needed. Bed tempera-
tures varied from 1685 to 18050F.
The test results are summarized in
Table 3. Runs 1 and 2A were conducted
essentially under the same test conditions.
An HCB concentration of 0.27 yg/m3 was
determined in the combustor exhaust
(Sample Location 1) on Run 1 using a plug
sampler. On Run 2A, no HCB was detected
in the combustor exhaust using an impinger
sampler (HCB 0.8 yg/m3). " No HCB was
detected in the exhaust from the baghouse
(Sample Location 2) for Run 1. Based on
the HCB detection limit and the size of
the baghouse gas sample, this resulted in
an HCB concentration in the final system
exhaust of less than 0.006 yg/m3. An HCB
sample was not taken in the baghouse
239
-------
GAS
VENT
SPENT
MELT
OVERFLOW
INTO
DRUMS
LIQUIDS
FEED
n ' [^.Iffi-iffi
jpjL^^w^^
MOLTEN SALT
DESTRUCTOR
Figure 4. Molten Salt Test Facility including sample locations.
TABLE 3....SUMMARY QF PILOT-SCALE TEST RESULTS
Combustor off -gas
ppmv
Baghouse
mg/m3
ppmv
Spent melt, ppmw
NOX, ppmv
HC, ppmv
3
Parti cul ate, mg/m
ORE, %
Note: The pH of the
line remained
HC1 emission.
2.7
2.3
<6
<5.2
liquid
basic
x 10
x 10
x 10
x 10
<6.2
HCB
-4 ?
-5-6!
-6-l.
-7-l.
0.001
x 10"3
11-9's
in a small
throughout
Chlordane
1 x
1 x
6 x
4 x
—
ID'2
ID"3
io-4
io-5
0.104
70 - 125
35 - 110
-0.107
—
9-9's
sampling
the test
5.3
3.2
<3.6
<2.1
4.1
x
x
x
x
x
10
10
10
10
10
-3 _
-4 _
-4 _
-5 _
0
-3 _
6.8 X
4.1 x
<4.4 X
<2.6 X
io-2
io-3
10"3
io-4
.0044 - 1.2
0.5 -630
0.4 - 60
1.75 x IO"2
8-9's. -7-9's
scrubber
indicating
in the off-gas
essentially no
240
-------
exhaust (Sample Location 2) for Runs 2A or
3A, and the DRE values reported for these
two runs are based on the Sample 1 HCB
values. The DREs for Runs 1 and 2A were
greater than 10/9's (99.99999999) and
8/9's, respectively.
Run 2B was made at 75.8 Ib/h HCB feed
(approximately 30% of the total thermal
load). Measurements for HCB showed
22.0 yg/m3 at Sample Location 1 and less
than 0.027 yg/m3 at Sample Location 2.
The molten salt bed temperature was held
at 1685°F during the run. The DRE was
>9/9's.
Run 3A was made at 112 Ib/h HCB feed.
An impinger measurement of less than
1.3 yg/m3 in the vessel exhaust corre-
sponded to a calculated DRE of greater
than 8/9's.
Run 3B was the last run of the test
series, and since results up to then were
so promising, it was decided to test the
molten salt unit with 100% HCB feed,
i.e., no auxiliary fuel coke was used. A
feed rate of 269 Ib/h HCB was used to main-
tain the vessel at a temperature of 1805°F
The HCB in the off-gas was 71 yg/m3 down-
stream of the combustor and 0.16 yg/m3
downstream of the baghouse. Note that
this was the only run in which HCB was
actually detected in the system's ex-
haust to the atmosphere. The 71
yg/m3 and 0.16 yg/m3 figures imply a
baghouse removal efficiency of 99.77%.
Post-test inspection of baghouse filters
verified the presence of tears in the
fabric.
Measured residual HCB in the molten
salt overflow is shown in Table 3 for
Tests 2B and 3B. The measured HCB con-
centration in the molten salt overflow
was only 0.001 yg/g of melt in Test 2B
but was 0.1 yg/g of melt in Test 3B. It
may be that the last value was a result
of contamination.
The salt dust from the baghouse
contained <0.001 yg/g of dust; it could
be recycled continuously back to the
combustor if the HCB concentrations in
the dust were of concern. Otherwise,
this residual material would be disposed
of as nonhazardous waste, as would the
melt overflow.
TEST CONCLUSIONS
Bench-scale testing of HCB and chlor-
dane was performed to optimize process
parameters and verify analytical techniques
to be utilized in pilot-scale tests which
followed. For HCB destruction at a nomi-
nal salt bed temperature of 1839°F, a
stoichiometry of 118%,and a residence time
of 2-3/4 s, a destruction removal efficiency
of _>99.9999986% was achieved. Concentra-
tion of HCB in the off-gas and the salt
bed, sampled during HCB feed, was
<0.000017 ppmv and <0.00100 ppmw, respec-
tively. For chlordane destruction, at a
lower salt bed temperature of 1648°F and
similar stoichiometry and residence time,
a destruction removal efficiency of
>99.999903% was achieved. Concentrations
of chlordane in the off-gas and salt bed
were <0.00267 ppmv and <0.0080 ppmv,
respectively. Additionally, no chlorine
or phosgene was detected within the ana-
lytical limits. HC1 and NO^ were deter-
mined to be <100 ppm and <50 ppm, respec-
tively. Monitoring, sampling, and analyti-
cal techniques to ensure conformance to
EPA standards were verified. Additionally,
a data base was established and subsequent-
ly verified for correlation of bench-scale
and pilot-scale tests.
Pilot-scale tests of the same materials
under similar process conditions produced
similar or better destruction efficiencies.
Typically, destruction efficiencies were
>99.99999%. Improved performance charac-
teristics in the pilot-scale units were
attributed to greater salt inventory and
greater residence times, the latter
approaching 2 s. System limits were
challenged by deliberate introduction of
plant upset conditions. Stoichiometry
was reduced to 0.78% from the nominal
120%; bed temperature was reduced'over
115°F from the nominal operating tempera-
ture of 1830°F; superficial air velocity
was increased from 2.0 fps to 2.7 fps;
feed throughput was more than doubled; and
auxiliary fuel was cut off in one case.
None of these conditions resulted in off-
gas emissions or POHCs in the spent melt
which.would be considered hazardous by
RCRA criteria. Under these upset condi-
tions, maximum concentrations of feed
material in the combustor off-gas and
spent melt was <0.0061 ppmw and <1.2 ppmw,
respectively. Stack emissions were
241
-------
<640 ppmv NOX and <110 ppmy hydrocarbons.
A small sampling scrubber in the off-gas
remained basic throughout the test series,
indicating essentially no HC1 carryover.
In summary, these tests provided quan-
titative data on the performance of the
Molten Salt Process in the destruction of
hazardous waste. The data indicates more
than adequate destruction efficiencies in
accordance with RCRA criteria. These
efficiencies were obtained with stack
emissions and process spent melt residue
characteristics v/ell below minimum
criteria for environmentally acceptable
discharges. The data supports considera-
tion of the Molten Salt Process as an
environmentally acceptable alternative
technology for the destruction of hazard-
ous waste.
ENGINEERING STATUS
Reference Commercial System Design
Energy Systems Group has completed
conceptual design of an MSD plant capable
of 2000-lb/h throughput for liquid PCB.
The plant is also capable of destroying
other liquid and solid wastes. Design
modifications may be needed on a case-by-
case basis after evaluation of performance
requirements. Figure 5'is a schematic of
this plant and Figure 6 is an artist's
concept.
COMBUSTORS
SALT DISPOSAL
Figure 5. Simplified plant schematic.
Figure 6. Molten salt destruction plant for hazardous wastes.
242
-------
EVALUATION OF HAZARDOUS WASTE
INCINERATION IN A .DRY PROCESS CEMENT KILN
Gregory M. Higgins and
Arthur J. Helmstetter, P.E.
SYSTECH Corporation
Xenia, OH 45385
ABSTRACT
This report presents the preliminary results of a test program conducted
by SYSTECH Corporation at the Marquette Cement Plant in Oglesby, Illinois.
The objective of this program was to compare the emissions resulting from co-
firing low chlorine, high Btu liquid waste and coal in a dry process cement
kiln with the emissions resulting from firing coal only.
The characteristics of the liquid waste burned during the test were exam-
ined by performance of standard analytical methods, with particular emphasis
on organic composition. Destruction and removal efficiencies (DREs) were
calculated for four principal organic hazardous constituents (POHCs) of the
fuel: methylene chloride; methyl ethyl ketone; 1,1,1-trichloroethane; and
toluene. Additional analyses were conducted on the stack gases to determine
particulate loading, S02, NOX, total gaseous nonmethane organics (TGNMO),
HC1, and metals emissions. The kiln dust was also sampled and analyzed for
metals and Extraction Procedure (EP) toxicity.
The results of these tests indicate that the cement kiln may be an ideal
method of disposal for low chlorine, high Btu liquid wastes. The burning of
liquid wastes in the kiln did not lead to any significant increase in particu-
late loading, S02, NOX, TGNMO, or HC1 over the levels observed during base-
line coal-only test periods. Among the metals examined, only lead was found
to significantly increase in emission rate during the liquid waste firing. No
significant differences were observed in the EP toxicity of the kiln dusts
sampled during the liquid waste and baseline tests, and only the concentration
of lead was found to significantly increase in the kiln dust. Within the
detection limits of the test method employed, the four POHCs measured were
completely destroyed in the kiln.
INTRODUCTION
Marquette Company operates a ce-
ment plant in Oglesby, Illinois, which
produces approximately 450,000 tons
of cement yearly. Pulverized coal
has been the primary fuel for this
facility. Marquette proposes to
243
-------
construct and operate a liquid waste fuels
resource recovery system at the Oglesby
plant to facilitate the use of select
combustible liquid waste materials such as
high Btu (10,000 to 14,000 Btu/lb), low
chlorine (2 to 5 percent) waste solvents
as supplemental fuel in the kiln. The
primary fuel will remain coal with an
estimated 25 to 40 percent of the heat
required by the manufacturing process
supplied by the energy content of liquid
waste materials* In order to assess the
environmental effects of burning liquid
wastes in the kiln, Marquette, in con-
junction with the U.S. Environmental Pro-
tection Agency (EPA), the Illinois Envi-
ronmental Protection Agency (IEPA), and
SY.STECH Corporation performed a series of
tests on the cement kiln at Oglesby on
October 26 through 31, 1981.
The liquid waste fuels burned at the
Oglesby facility were low chlorine, rela-
tively high Btu waste solvents from ink
and paint manufacturing and solvent re-
covery processes. Wastes of this type are
classified as hazardous under the Resource
Conservation and Recovery Act (RCRA) regu-
lations primarily because of their ignita-
bility, and have generally been considered
unsuitable for recycling or reclaiming.
Materials of this type have traditionally
been disposed in landfill or incinerated
without recovering their useful energy
content* Incineration of such wastes in a
ceaent kiln has the potential of minimiz-
ing any environmental damage from their
disposal by destroying their combustibles
while simultaneously recovering their
useful energy content. Because of the high
temperatures (2700° to 3000°F), long gas
retention times (approaching 10 seconds),
and high alkalinity in cement kilns, even
highly toxic wastes such as PCBs have been
demonstrated to be effectively destroyed.
Earlier tests have demonstrated destruc-
tion and removal efficiencies (DREs) of
such compounds exceeding 99.99 percent.1»2
The primary objective of the tests
conducted in this program was to charac-
terize the baseline emissions from the
process when only coal was fired and to
compare the results from this situation
with those obtained when a combination of
liquid waste and coal was being fired in
the kiln. Specific emphasis was placed on
characterizing the DREs of four principal
organic hazardous constituents (POHCs) of
the liquid wastes: methylene chloride;
methyl ethyl ketone; 1,1,1-trichloro-
ethane; and toluene. The experimental
design and test methods used were develop-
ed to permit a comparison to be made be-
tween these two conditions rather than to
accomplish a definitive environmental
assessment of the burning of liquid wastes
in a cement kiln. Therefore, it should be
emphasized that the main purpose of this
program was to determine whether co-firing
liquid wastes increased the emissions over
those normally observed during baseline
conditions. The only statistically valid
conclusions which can be derived from this
test program fall into two categories:
either the emissions were significantly
increased or else no significant increase
was observed. Although the mean values of
the data collected may at first glance
seem to indicate differences, only by
considering the numbers of samples taken
and the calculated sample standard devia-
tions can statistically valid conclusions
be reached. This summary report focuses
on system emissions while the final EPA
report will include information regarding
any effects that the liquid waste may have
had on the characteristics of the cement
product. These effects are expected to' be
negligible and should not impact the en-
vironmental considerations regarding this
project.
METHODS
The test program conducted at the
Oglesby facility was designed to collect
2 days of samples and data during both the
baseline and liquid waste burns. Because
of difficulties encountered in receiving
and delivering a sufficient quantity of
liquid waste and periodic unscheduled kiln
down times, however, only 1 day of testing
was accomplished with liquid waste and
only 1 1/2 days during baseline condi-
tions.
Figure 1 shows a schematic flow dia-
gram of the dry process cement kiln tested
at Oglesby. Samples and data were col-
lected from the sites labeled A through E.
Liquid waste was pumped -from a storage
244
-------
tanker into the flame of the kiln through
a specially designed delivery nozzle.
The liquid waste burned in the test
program was sampled at Site A and analyzed
for heating value, chlorine, sulfur, ni-
trogen, and water content by standard ASTM
reference methods. Method 8.01 (SW 846
Test Methods for Evaluating Solid Waste;
Physical/Chemical Methods) was employed to
identify the concentrations of methylene
chloride; methyl ethyl ketone; 1,1,1-tri-
chloroethane; and toluene. Other organics
were qualitatively identified by a general
scanning procedure employing gas chroma-
tography with flame ionization detection
(FID). Metals were identified by in-
ductively coupled plasma emission spec-
troscopy (ICP). Atomic absorption spec-
troscopy (AA) was employed for mercury
analysis.
.Stack emissions were sampled at
Site C and examined for total gaseous non-
methane organics (TGNMO) by EPA Method 25.
Integrated bag samples of stack gases were
analyzed by gas chromatography with FID
according to EPA Method 23 for methylene
chloride; methyl ethyl ketone; 1,1,1-tri-
chloroethane; and toluene (see Federal
Register, June 11, 1980, and EPA 600/4-80-
003). Particulate emissions were measured
by EPA Method 5, and S02 was measured by
a modified EPA Method 8 procedure employ-
ing 3 percent H202 in the impingers to
collect S02 and sulfuric acid mist. Metals
emissions were determined from an analysis
of the Method 5 filters and the acetone
probe rinses by ICP spectroscopy. Mercury
emissions were determined by AA spectros-
copy. Emission of nitrogen oxides was
determined by EPA Method 7. A midget
impinger train containing sodium hydroxide
was used to collect hydrochloric acid
emissions. Analysis of the impinger con-
tents was accomplished by a mercuric ni-
trate titration.
The kiln dust collected at Site D
from the last two banks of the electro-
static precipitator (ESP_) was analyzed for
metals by ICP spectroscopy. Mercury was
analyzed by AA spectroscopy. The poten-
tial for leaching of metals from this dust
was assessed by conducting EP toxicity
tests.
RESULTS AND DISCUSSION
Figure 2 illustrates the testing
and process operating log for the test
period. It is evident that the kiln ex-
perienced several periods of down time
during the tests. Three tests, desig-
nated Test Nos. 1, 2, and 3, were con-
ducted during the liquid waste burns on
October 28, 1981. Test Nos. 4, 5, and 6
were conducted during baseline burns on
October 30 and 31, 1981. Each test con-
sisted of approximately 2 hours during
which samples and data were collected.
The liquid waste used for this test
was typical of that which would be speci-
fied for a permanent facility operation.
The liquid waste was sampled every hour
during the test period, and composites
were made of the hourly samples to repre-
sent the material fired during each of
the three co-firing conditions. The re-
sults of these analyses are presented in
Table 1. The liquid waste had a mean
heating value of 12,350 Btu/lb and a mean
chloride content of 4.54 percent. The
primary metallic constituent of the waste
was lead, with a mean concentration of
1,800 ppm. No PCBs were detected in the
waste and the POHCs designated for these
tests comprised an average of 24.7 percent
of the liquid waste. The other organic
constituents listed in Table 1 were quali-
tatively identified by retention time in
the GC analysis.
It is anticipated that in normal
operation a substitution rate of 25 to
40 percent liquid waste would be used;
however, during the test period a limi-
tation on the quantity of liquid waste
available precluded firing the waste at
these rates. The substitution rates for
the liquid waste expressed as a percent of
the total heat input of the combined coal
and liquid waste stream were 14.0 percent
for Test No. 1, 11.2 percent for Test
No. 2, and 12.8 percent for Test No. 3.
Stack emissions measured during the
test program are shown in Table 2. None
of the POHCs measured were detected In the
stack gas samples during either co-firing
or coal-only combustion. Table 3 shows
245
-------
the DREs of the kiln for the POHCs meas-
ured during the liquid waste tests. Since
none of the POHCs were detected at the
minimum detection limit of the test method
(nominally 0.1 ppra), the DREs expressed
can be considered minimum values. The
efficiency of combustion of organic com-
pounds is also supported by the results of
the Method 25 tests which show no signifi-
cant increase at a 95 percent confidence
level in TGNMO emissions during co-firing
when compared with baseline results. These
results indicate that there is no measur-
able contribution to the hydrocarbon emis-
sions from the stack due to the addition
of liquid waste to the kiln.
In addition to hydrocarbons, Table 2
also presents the particulate, S02, NOX,
and HC1 emissions under the co-firing con-
dition for Test Nos. 1, 2, and 3 versus
the coal condition during Test Nos. 4, 5,
and 6. Although unstable operating con-
ditions contributed considerable varia-
bility within the data sets, there was no
significant increase at a 95 percent con-
fidence level over baseline conditions in
particulate, S02, NOx, or HC1 emissions
while cofiring liquid waste.
Table 4 shows the results of metals
analyses of .stack particulates sampled
during this test. A slight increase in
lead emissions over baseline conditions
(from .04 to .072 Ib/hr) was observed dur-
ing the co-firing tests. Other metals
examined did not increase significantly at
a 95 percent confidence level with the
burning of the liquid wastes. Table 4
also shows the results of metals analyses
of the kiln dust sampled during this test.
An increase in the lead concentration in
the kiln dust was observed during the
liquid waste firing. This increase was
from approximately .02 percent during
coal-only conditions to .05 percent dur-
ing the co-firing conditions. The other
metals present in the liquid waste were
not observed to significantly increase in
concentration at a 95 percent confidence
level in the kiln dust when the liquid
waste was co-fired in the kiln. Further,
the leachability of metals from the kiln
dust did not increase significantly and
remained orders of magnitude below the
allowable limits specified by the EPA.
The inclusion of metals in the liquid
waste burned in the cement kiln appears to
have had a minimal impact on the discharge
of metals from the stack or on the amount
occurring in the kiln dust. This may in-
dicate that many metals are deposited in
the cement clinker and become incorporated
in the crystalline structure of the cement
product.
CONCLUSIONS
The general conclusion that can be
drawn from this test program regarding the
the environmental impacts of firing liquid
wastes in a cement kiln is that the liquid
waste does not significantly increase the
emissions of particles, hydrocarbons, S02,
NOX, or HC1 from the stack. Within the
detection limits of the test methods em-
ployed, the four POHCs examined were com-
pletely destroyed in the kiln. A slight
increase in lead in the particulate and
kiln dust was observed, but other metals
examined did not show a significant in-
crease either in stack emission or in the
process collected kiln dust with the
burning of the liquid waste. The kiln
dust was determined by the EP toxicity
test to be nonhazardous and could be land-
filled even when liquid wastes are co-
fired. In general, the use of liquid
waste in this test produced only minor
changes in the emissions from the kiln and
revealed that the kiln may provide a prom-
ising waste disposal method for high Btu,
low chlorine wastes.
REFERENCES
1. Ahling, B., "Combustion Test with
Chlorinated Hydrocarbons in a Cement
Kiln at Stora Vika Test Center,"
Swedish Water and Air Pollution Re-
search Institute, Stockholm, March 16,
1978.
2'. Knut Trovaag, "Hazardous Waste Incin-
eration in a Cement Kiln," Inter-
national NATO/CCMS Hazardous Waste
Symposium, October 5, 1981.
246
-------
TABLE 1. WASTE FUEL ANALYSIS
Heat content (Btu/lb)
Ash (%)*
Chloride (%)
Sulfur (%)
Nitrogen (%)
Water (%)
Metals (ppm)
Cadmium
Chromium
Copper
Lead
Manganese
Mercury
Nickel
Zinc
Quantified organics (%)
Methylene chloride
1,1, 1— trichloroethane
Methyl ethyl ketone
Toluene
PCBst '
Other organics (%)§
Methanol
Ethanol
2-Propanol
Acetone
Methyl acetate
Ethylene chloride
2-ethoxy-ethyl acetate
Methyl isobutyl ketone
Butyl acetate
4-hydroxy-4-methyl-2-pentanone
Benzene
Ethyl benzene
Styrene
Cg Alkyl benzenes
Paraffins (Cg-C]^)
Xylene
Unidentified
(2 compounds)
Test 1
12,210
12.13
3.56
.09
.13
10.70
3
725
195
2520
47
.076
52
1770
23.88
2.72"
1.86
7.51
11.79
N.D.
53.29
Test 2
13,012
7.82
4.28
.09
- .10
10.3
2
565
. 167
1820
35
.104
42
1290
22.01
2.94
1.63
8.90
8.54
N.D.
59.87
Test 3
11,823
6.85
5.80
.06
.20
' -11.80
1
446
177
1050
42
<.032
107
905
28.26
6.27
1.97
8.18
11.84
N.D.
53.09
* All (%) are weight percent.
t None detected at a detection limit of .005 percent.
§ Weight percent of other organics = 100 percent - ash - water —'quanti-
fied organics. Compounds listed in this category were not quantified.'
247
-------
CO
t*
i
DQ
H
C/J
13
|
CO
CO
•g
S
H
S
1
.
CvJ
W
. *j
PQ
3
S
«
S
o
P
•o
S
0
o
4J
S>
u
•H
4J
m
3
CO
^i
0
01
0
P
01
P4
rH P
B^
H
dg,
a CM
n
§ ja
r-l
M f^
o o<
P
CM i0
CO rQ
H
CM a
s a
• P
SH
i^l
• o
4J CO
P ^*
BJ P
CM 60
So
p o.
«
CO
o
W
O
FW
^
•°
4J CO
3 01
(X 4-1
0 CO
•rt
• •
o id
• •
5= 2:
CM CO
r-l CM
r^ *••!
^
iH
a
•i
o
CM CO O
vo m o
en . •»
CO vo h~
r-l r-l
• • •
o o e
Jgj J2» igj
II 1
I
III
«* in vo
.
0)
0
cd
4J
8
o
i-H
O
Tt
4J
r-l.
r-l ,
r-4
'
0
cd
. a^
01
0
o
4J
OJ
-W
(H
4J
01
iH
4J
i
• *»
01
t-l
O
4J
Ol
T3
•H
P
5
O
Ol
0
01
^
_r*
4J
.,
Ol
01
is
P
0
•o
01
4J
CO
01
4J
CO
O
g
0
p
S
g,
CM
CO
cd
T)
Ol
CO
CO
01
p
CM
Ol
CO
0
1
cd
o
o
p
X
^£3
in
CM
•o
o
W
+.
.
CM
Q
iH
O
M-l
O
4J
•H
r-l
0
o
•H
4J
O
Ol
4J
01
•d
r-t
0
*3
§
I3
cd
4J
cd
01
4J
O
Ol
4J
rp
Ol
,0
un
2A8
-------
TABLE 3. DREs CALCULATED FOR THE CO-FIRING TESTS
Test
numbers
1
2
3
Average
Methylene
chloride
99.869
99.851
99.917
99.879
Minimum
Methyl ethyl • >
ketone l.lj
99.960 , /,.;
99.959
99.961
99.960
DREs*
, 1-trichloroethane
;99.718 '
99.604 :
99.7,10
99.677
, :(
Toluene
99,968
99.947 ;
99.968 ;
99.961 '
* The DREs indicated have been adjusted by a sensitivity analysis '
approach to include the impact of random measurement errors oil the ORE
calculations. The stated values therefore actually represent the
minimum DREs observable in ;the cement kiln by employing the test methods
described in this report. '•
249
-------
TABLE 4. METALS TEST RESULTS
Element
Cd
Cr
Cu
Pb
Mn
NI
Hg
Zn
As
Ba
Cd
Cr
Pb
Hg
Se
Ag
Cd
Cr
Cu
Pb
Mn
Hg
Ni
Zn
Kiln Dust Analysis (ppm)
Liquid waste Coal only
Run
No. 1 2 3 Avg. 456
18 16 15 16 19 22 30
23 24 31 26 23 28 24
26 24 25 25 24 30 30
524 511 488 507 173 186 167
632 656 795 694 681 759 655
25 31 31 29 22 27 29
.22 .30 .40 .31 .40 .22 .26
71 66 70 69 49 57 75
Kiln Dust EP Toxicity Test Results (mg/Jl)
Liquid waste Coal only
Run
No. 123 Avg. 4 5 6
<.003 <.003 <.003 <.003 .006 .004 .003
.593 1.020 .685 .766 .753 .649 1.020
.004 .481 <.003 .167 .073 .017 .897
<.020 .088 .032 .047 <.020 .020 .035
<.043 .919 <.043 .335 <.043 <.043 .713
<.0008 <.0008 <.0008 <.0003 <.0008 <.0008 <.0008
.215 .343 .179 .246 .350 .414 .502
•C.007 <.007 <.007 <.007 <.007 .013 <.007
Metals Emissions On Particulates (Ib/hr)
Liquid waste Coal only
Run
No. 1 23 Avg. 456
.003 .001 .001 .002 .002 .003 .002
.075 .030 .054 .053 .020
— — — — — — —
.105 .060 .052 .072 .042 .036 .043
.007 .004 .003 .005 .003 .002 .003
.0002 .0002 .0001 .002 .0003 .0003 .0003
.160 .061 .065 .095 .044 .018 .032
.049 .005 .001 .018 .002 .002
Avg.
24
25
28
175
698
26
.29
60
Avg.
<.005
.807
.329
.025
.266
<.0008
.422
<.009
Avg.
.002
.020
—
.040
.003
.003
.031
.002
250
-------
FEED
MATERIALS
STACK
GASES
COAL
LIQUID
B
KILN
CEMENT
CLINKER
ESP
DUST RETURN
DUST
DISPOSAL
Figure 1. Marquette-Oglesby cement kiln schematic.
251
-------
Hour
Date
10/27
10/28
10/29
10/30
10/31
0 2 4 6
10 12 14 16 18 20 22 Ik
Coal
Co-firing
Downtime
Test 1 Test 2 Test 3
Test 4 Test 5
Test 6
Y/////////////:
Figure 2. Process operations log.
252
*U.S. GOVERNMENT PRINTING OFFICE 1fl83-659-095/1930
-------