United States
Environmental Protection
Agency
Hazardous Waste Engineering
Research Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S2-86/098 Jan. 1987
\>EPA Project Summary
Preliminary Assessment of Air
Emissions from Aerated
Waste Treatment Systems at
TSDFs
C. C. Allen, D. A. Green, J. B. White, and J. B. Coburn
Aerated wastewater treatmet unit
operations are used for the removal of
organic compounds from hazardous
waste and other industrial wastewater
streams. Some operations use aeration
to supply oxygen for aerobic decompo-
sition of organics; in other operations,
incidental air/water contact occurs to
varying degrees. Methods for estimat-
ing emissions resulting from air strip-
ping of volatile organic compounds
that may accompany aerated treatment
are applied to full scale and pilot plant
wastewater treatment plants.
Treatment plant examples are inves-
tigated for aerated industrial waste
treatment systems employing trickling
filters, activated sludge and aerated la-
goons. The size and configuration of full
size treatment plants are highly vari-
able, corresponding to the highly vari-
able volume and strength of industrial
wastewater. Typical size and residence
time specifications for unit operations
within these plants are discussed.
The recommended mathematical
models are used to generate predic-
tions of the fate of volatile organic com-
pounds in wastewater treatment sys-
tems. Where full scale plant data are
available, predictions generally agree
with measurements (within the limits
of accuracy which result from varia-
tions in sampling and chemical analy-
sis). For this reason, the mathematical
models can be used to estimate emis-
sions for those systems where no field
data are available.
This Project Summary was devel-
oped for the EPA's Hazardous Waste
Engineering Research Laboratory,
Cincinnati, OH, to announce key find-
ings of the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
The EPA Office of Air Quality Planning
and Standards (OAQPS) is developing
regulations under the 1976 Resource
Conservation and Recovery Act (RCRA)
to control air emissions from hazardous
treatment, storage, and disposal facili-
ties (TSDFs). The purpose of the air
emissions regulations is to protect
human health and the .environment
from emissions of volatile organic com-
pounds (VOCs), particulates, and aero-
sols.
The sources of TSDF emissions in-
clude storage tanks, treatment proc-
esses, surface impoundments, lagoons,
landfills, land treatment, and drum stor-
age and handling facilities. Those proc-
esses involving the use of air (biological
treatment and cooling) or subject to the
introduction of air (stirred equalization
and neutralization) may emit VOCs as a
consequence of air stripping.
To further understand and better esti-
mate the sources and extent of VOC
emissions from TSDFs, typical aerated
treatment facilities are investigated.
Various mathematical models are
presented to predict the mechanism
and extent of VOC emissions during the
-------
different process conditions encoun-
tered. The models include a methodol-
ogy for estimating the relative impor-
tance of competing removal pathways
(i.e., adsorption and biological oxida-
tion). The study is limited to those bio-
logical processes or physical chemical
processes with potential to vent VOC to
the air.
The basic problem in evaluating air
emissions from aerated waste treat-
ment processes is determining the im-
portance of competing mechanisms. In
an activated sludge biological treatment
process, for example, the major mecha-
nisms for removal of dissolved contam-
inants from the aqueous waste are bio-
logical oxidation, adsorption on
biomass, and mass transfer into the air.
In a rough analysis of "removal effi-
ciency" concentrations of the contami-
nants in the aqueous influent and efflu-
ent are measured and the fraction of the
contaminant which disappears in the
process is reported as an efficiency.
This procedure gives no information
about the relative importance of com-
peting removal mechanisms.
Where one or more of the removal
mechanisms is destructive of the con-
taminants (chemical or biological oxida-
tion) and others non-destructive (air
stripping or aerosol dispersal), measur-
ing the relative importance of the com-
peting mechanisms becomes compli-
cated. In such a case, the contaminant
removed by the non-destructive mecha-
nism must be recovered and the bal-
ance assumed to have been destroyed.
Errors in chemical analyses, and partic-
ularly sampling errors, have a great im-
pact on calculations of the fraction of
the destroyed contaminant.
Alternatively, data may be obtained
under controlled laboratory conditions
and adapted to provide reasonable esti-
mates of emissions expected under
field conditions. One approach is to con-
duct experiments in covered vessels to
permit accurate sampling of vent gases
and accurate metering of vent gas flow
rates. Another approach involves in-
hibiting the destructive removal mecha-
nisms by taking steps to chemically
sterilize the system or to substitute an
inert gas such as nitrogen for air. In this
latter case, the mass transfer to the gas
phase continues to occur and all of the
destructive mechanisms are eliminated.
Influent and effluent aqueous phase
sampling are used to determine the ex-
tent of removal resulting from stripping.
Vapor phase sampling of such a labora-
tory system should close the mass bal-
ance for the system. The extent to which
under-recovery or over-recovery of con-
taminants is found can be used to evalu-
ate the adequacy of laboratory vapor
phase sampling techniques.
One approach to the air stripping
problem is to measure the total disap-
pearance in the field and then apply
mathematical models obtained from
laboratory data to simultaneously esti-
mate the rate of stripping and biological
oxidation. When these models account
for significantly more or significantly
less contaminant removal than the ob-
served total removal, several steps may
be taken. The simplest approach is to
assume that the mass transfer model is
more accurate than the biological oxi-
dation model and account for biological
oxidation (and other destructive re-
moval mechanisms) by difference. Be-
cause they are less affected by changes
in operating conditions and influent
wastewater composition, the mass
transfer models are in fact more accu-
rate predictors than are the biological
oxidation models. In some cases, how-
ever, air sampling must be conducted to
verify the prediction of stripping mod-
els. The interpretation of such data ob-
tained under field conditions is, how-
ever, subject to considerable error.
The objectives of this study are to
present and evaluate the various kinetic
and predictive fate models that are ap-
plicable to aerated wastewater treat-
ment systems. Four different waste
treatment systems will be considered:
(1) trickling filters, (2) activated sludge,
(3) aerated lagoons, and (4) spray
ponds. When possible, model predic-
tions will be compared with experimen-
tal data.
This study investigated mathematical
models and correlations available in the
literature for mass transfer in systems
similar to wastewater treatment units.
Classical engineering approaches to
strippers and absorbers, packed bed de-
sign, and stream and lake reaeration
were considered and adapted as neces-
sary to wastewater treatment systems.
Procedure
Pretreatment was modeled as a plug
flow system with the gas phase mass
transfer coefficient calculated from a
"wind speed" correlation and the liquid
phase mass transfer coefficient calcu-
lated using a stream reaeration correla-
tion. Primary and secondary clarifiers
were modeled as plug flow systems
using the same methods as described
above to calculate mass transfer coeffi-
cients. Surface aeration systems (e.g.
some activated sludge units) were mod-
eled using a correlation with applied
aerator horsepower. Biological oxida-
tion was incorporated using a pseudo-
first order rate constant to be calculated
from the best available zero-order data
and an assumption as to steady state
composition.
Diffused aeration systems (e.g. some
activated sludge systems with sub-
merged aerators) were modeled with
stripping based on the attainment of
equilibrium during the time of bubble
rise, which is combined with surface dif-
fusion calculated as for clarifiers. Bio-
logical decay is treated in the same
manner regardless of the means of
aeration. Suggested procedures for cal-
culating mass transfer coefficients ap-
propriate for equalization ponds, dis-
solved air flotation units, spray ponds
and cooling towers are also given.
As a step in verifying the accuracy of
the suggested mathematical models,
data from an extensive sampling pro-
gram at a well characterized (with re-
spect to residence time, dimensions
and influent and effluent concentra-
tions) wastewater treatment system
were used to test the models. In addi-
tion, pilot scale data from well con-
trolled experiments and some less ex-
tensive industrial wastewater treatment
system data from EPA plant surveys
were also used.
A list of input specifications which de-
scribe the system to be modeled is
given in Table 1; values given are from
a wastewater treatment system at a
Union Carbide chemical manufacturing
plant. For cases where units of a particu-
lar type are not used in a particular sys-
tem (e.g. dissolved air flotation in the
example given), a specification of zero
for area, number, or air flow eliminate
the inappropriate part of the model.
Results and Discussion
The results of the calculations on the
sample plant specification are given in
Table 2. The rate of loss of volatiles
from the plant as predicted by the math-
ematical models agreed well with
measured results, with the exception o1
the UNOX biological treatment unit
The rate of removal in the UNOX sys
tern was overpredicted by the model
particularly in the case of tetralin. Al
though the loss in the aeration basir
was overpredicted by the model foi
tetralin and naphthalene (89% was the
observed loss in both cases), the abso
lute errors were still only 5-10%.
-------
Table 1. Model Input Data for Union Carbide Plant (Alsop, et al., 1984)*
Water Temperature (Deg C) 25
Wind Velocity (cm/S) 200
Concentration of Benzene (Mole Fract) .001
Concentration of Naphthalene (Mole Fract) .005
'Concentration of Ethylbenzene (Mole Fract) .001
Concentration of Methylcellosolve (Mole Fract) .001
Concentration of Tetralin (Mole Fract) .0002
Biorate of Benzene (Hours) 400
Biorate of Naphthalene (Hours) 400
Biorate of Ethylbenzene (Hours) 400
Biorate of Methylcellosolve (Hours) 800
Biorate of Tetralin (Hours) 200
Waste Flow Rate (M3/Sec) .07
Area of Pretreatment Basin (M2) 50
Depth of Pretreatment Basin (M) 3
Diameter of Clarifier (M) 19.4
Depth of Clarifier (M) 2.4
Number of Clarifiers 1
Length of Aeration Basin (M) 170
Width of Aeration Basin (M) 170
Depth of Aeration Basin (M) 3.6
Area of Agitators (Each) (M2) 96
Number of Agitators 30
Power of Agitation (Each) (HP) 75
Impeller Diameter (cm) 30
Impeller Rotation (rpm) 2000
Number of Secondary Clarifiers 1
Diameter of Secondary Clarifiers (M) 37
Depth of Secondary Clarifiers (M) 3
Area of Equalization Basin (M2) 5185
Depth of Equalization Basin (M2) 3
Air Flow in DAF (Pretreat) (SM3/S) 0
Flow of Submerged Air (M3/Sec) .4
Power of Sub-Agitation, Each (HP) 50
Length of Sub. Aeration Basin (M) 33.5
Width of Sub. Aeration Basin (M) 8.4
Depth of Sub. Aeration Basin (M) 8.5
Area of Sub. Agitators (Each) (M2) 70.1
Number of Sub. Units in Series 3
Impeller Rotation (rpm) 70
Impeller Diameter (cm) 243
Weir Height (cm) 30
Thickness of Weir Flow (cm) 1
Enter {1} for Covered Subm. Agit. 1
*Fate of Specific Organics in an Industrial Biological Wastewater Treatment Plant, EPA Draft
Report (June 29, 1984).
In an EPA pilot investigation, a num-
ber of VOCs were processed in a clari-
fier before entering an aerated biologi-
cal oxidation unit. Enough data were
provided to specify the input parame-
ters of the model. Although the treat-
ment plant was not directly exposed to
the atmosphere, air was circulated and
the mass transfer is considered to be
liquid phase controlled. For both the
clarifier and the aerator, the model pre-
dicted VOC losses in close agreement
with the pilot data. The model predicted
that most of the VOCs were lost to the
circulating air.
The fractional loss of VOCs to the air
was estimated for the du Pont Belle, WV
plant. The model predictions for
methanol, acetone, and chloroform
were within a factor of 5 for the primary
process, but did not agree with the esti-
mated losses of butanol. The secondary
treatment estimations did not agree
with model predictions. Not enough in-
formation was provided to determine
the source of the difference. At the
Union Carbide Sisterville plant, the pre-
dicted fractional loss was within a factor
of 5 of the estimated values. Only pri-
mary treatment was assumed; except
for the flow rate, no information was
provided about the source characteris-
tics.
Conclusions and Recommenda-
tions
The available theoretical and semi-
empirical methods for predicting the
rates of mass transfer of volatile organic
compounds from dilute solution into air
can be used to accurately predict emis-
sions from aerated wastewater treat-
ment processes in the absence of com-
peting removal mechanisms. Where
liquid phase concentration data and
system operating conditions are well-
characterized, VOC emissions from aer-
ated lagoons, activated sludge proc-
esses, trikckling filters, and Clarifiers
associated with wastewater treatment
systems may be estimated within the
accuracy of sampling and chemical
analysis results.
The presence of competing removal
mechanisms, of which biooxidation is
the most important, reduces the accu-
racy of emissions predictions based on
plant influent concentration data. If bulk
wastewater concentrations are known
for individual treatment units, then rea-
sonably accurate predictions may be
made. However, when only the concen-
tration of the influent to the entire plant
is known, accurate rate data for biooxi-
dation become more important, and
their absence has a greater influence on
the accuracy of emissions estimation.
Biooxidation rate data is available in
the literature for various organic com-
pounds which may be present in indus-
trial wastewater. The data are highly de-
pendent on the conditions under which
they were obtained and are not easily
adapted to real systems which may
have different initial concentrations,
biomass loadings, nutrient and inhibitor
concentrations, etc. A table of bioxida-
tion rate data obtained from the litera-
ture is included in the final report to pro-
vide a starting point for calculations.
Industrial waste treatment systems
vary widely in design reflecting the wide
variation in the waste streams to be
treated. The recommended predictive
mathematical models are adapatable to
widely varying volumetric flow rates.
Residence time, which in actual system
designs accounts for the concentration
and increases with the difficulty of re-
moval of organic compounds and sus-
pended solids, must be specified from
actual system designs or estimated
(less accurately) from removal rate
data. Model plant designs are given in
the report which fit within the wide
range of suitable designs. The best
mathematical models, as determined
-------
Table 2. A Comparison of the Experimental Loss of Volatiles from the Union Carbide Plant by this Study, were compared with ex-
Field Test to the Predicted Loss from Mathematical Models perimental measurements of full-scale
Fractional Loss Fractional Loss to Air and P»ot-*cale wastewater treatment
UniWolatile (experimental) (predicted) systems. Sample calculations are in-
/Nlnrl
-------
C. C. Allen, et a/., are with Research Triangle Institute, Research Triangle Park,
NC 27709.
Ronald J. Turner is the EPA Project Officer, see below.
The complete report, entitled "Preliminary Assessment of Air Emissions from
Aerated Waste Treatment Systems at Hazardous Waste Treatment Storage
and Disposal Facilities," (Order No. PB87-113 783/AS; Cost: $18.95, subject
to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Hazardous Waste Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S2-86/098
0000329 PS
------- |