-------
trenches connected to a main sewer line. The drains and trenches are usually open to the
atmosphere. Junction boxes, sumps, trenches, lift stations, and weirs will be located at points
requiring waste water transport from one area or treatment process to another.
A typical POTW facility collection system will contain a lift station, trenches, junction
boxes and manholes. Waste water is received into the POTW collection system through open
sewer lines from all sources of influent waste water. As mentioned previously, these sources may
convey sanitary, pretreated or untreated industrial, and/or storm water runoff waste water.
The following paragraphs briefly describe some of the most common types of waste water
collection system components found in industrial and POTW facilities. Because the arrangement
of collection system components is facility-specific, the order in which the collection system
descriptions are presented is somewhat arbitrary.
Waste water streams normally are introduced into the collection system through individual
or area drains, which can be open to the atmosphere or sealed to prevent waste water contact
with the atmosphere. In industry, individual drains may be dedicated to a single source or piece
of equipment. Area drains will serve several sources and are located centrally among the sources
or pieces of equipment that they serve.
Manholes into sewer lines permit service, inspection and cleaning of a line. They may be
located where sewer lines intersect or where there is a significant change in direction, grade or
sewer line diameter.
Trenches can be used to transport industrial waste water from point of generation to
collection units such as junction boxes and lift stations; from one process area of an industrial
facility to another, or from one treatment unit to another. POTWs also use trenches, to transport
waste water from one treatment unit to another. Trenches are likely to be either open or
covered with a safety grating.
Junction boxes typically serve several process sewer lines, which meet at the junction box
to combine multiple waste water streams into one. Junction boxes normally are sized to suit the
total flow rate of the entering streams.
Sumps are used typically for collection and equalization of waste water flow from trenches
or sewer lines before treatment or storage. They are usually quiescent and open to the
atmosphere.
Lift stations are usually the last collection unit before the treatment system, accepting
waste water from one or several sewer lines. Their main function is to lift the collectedwaste
water to a treatment and/or storage system, usually by pumping or by use of a hydraulic lift, such
as a screw.
Weirs can act as open channel dams, or they can be used to discharge cleaner effluent
from a settling basin, such as a clarifier. When used as a dam, the weir's face is normally aligned
perpendicular to the bed and walls of the channel. Water from the channel usually flows over the
weir and falls to the receiving body of water. In some cases, the water may pass through a notch
or opening in the weir face. With this type of weir, flow rate through the channel can be
measured. Weir height, generally the distance the water falls, is usually no more than 2 meters.
A typical clarifier weir is designed to allow settled waste water to overflow to the next treatment
9/91 Evaporation Loss Sources 4.13-3
-------
process. The weir is generally placed around the perimeter of the settling basin, but it can also be
towards the middle. Clarifier weir height is usually only about 0.1 meters.
Treatment And/or Storage Systems - These systems are designed to hold liquid wastes or
waste water for treatment, storage or disposal. They are usually composed of various types of
earthen and/or concrete-lined basins, known as surface impoundments. Storage systems are used
typically for accumulating waste water before its ultimate disposal or for temporarily holding batch
(intermittent) streams before treatment.
Treatment systems are divided into three categories, primary, secondary or tertiary,
depending on their design, operation and application. In primary treatment systems, physical
operations remove floatable and settleable solids. In secondary treatment systems, biological and
chemical processes remove most of the organic matter in the waste water. In tertiary treatment
systems, additional processes remove constituents not taken out by secondary treatment.
Examples of primary treatment include oil/water separators, primary clarification,
equalization basins, and primary treatment tanks. The first process in an industrial waste water
treatment plant is often the removal of heavier solids and lighter oils by means of oil/water
separators. Oils are usually removed continuously with a skimming device, while solids can be
removed with a sludge removal system.
In primary treatment, clarifiers are located usually near the beginning of the treatment
process and are used to settle and remove settleable or suspended solids contained in the influent
waste water. Figure 4.13-2 presents an example design of a clarifier. Clarifiers are generally
cylindrical and are sized according to both the settling rate of the suspended solids and the
thickening characteristics of the sludge. Roating scum is generally skimmed continuously from
the top of the clarifier, while sludge is typically removed continuously from the bottom of the
clarifier.
Equalization basins are used to reduce fluctuations in the waste water flow rate and
organic content before the waste is sent to downstream treatment processes. Row rate
equalization results in a more uniform effluent quality in downstream settling units such as
clarifiers. Biological treatment performance can also benefit from the damping of concentration
and flow fluctuations, protecting biological processes from upset or failure from shock loadings of
toxic or treatment-inhibiting compounds.
In primary treatment, tanks are generally used to alter the chemical or physical properties
of the waste water by, for example, neutralization and the addition and dispersion of chemical
nutrients. Neutralization can*6ontrol the pH of the waste water by adding an acid or a base. It
usually precedes biotreatment, so that the system is not upset by high or low pH values. Similarly,
chemical nutrient addition/dispersion precedes biotreatment, to assure that the biological
organisms have sufficient nutrients.
An example of a secondary treatment process is biodegradation. Biological waste
treatment usually is accomplished by aeration in basins with mechanical surface aerators or with a
diffused air system. Mechanical surface aerators float on the water surface and rapidly mix the
water. Aeration of the water is accomplished through splashing. Diffused air systems, on the
other hand, aerate the water by bubbling oxygen through the water from the bottom of the tank
or device. Figure 4.13-3 presents an example design of a mechanically aerated biological
treatment basin. This type of basin is usually an earthen or concrete-lined pond and is used to
treat large flow rates of waste water. Waste waters with high pollutant concentrations, and in
4.13-4 EMISSION FACTORS 9/91
-------
Effluent Weir
Drive Unit
\
Scraper Blades
Sludge Drawoff Pipe
Figure 4.13-2. Example clarifier configuration.
Cable Ties
Mechanical
Aerators
A
•In
Figure 4.13-3. Example aerated biological treatment basin.
9/91
Evaporation Loss Sources
4.13-5
-------
particular high flow sanitary waste waters, are typically treated using an activated sludge system
where biotreatment is followed by secondary clarification. In this system, settled solids containing
biomass are recycled from clarifier sludge to the biotreatment system. This creates a high biomass
concentration and therefore allows biodegradation to occur over a shorter residence time. An
example of a tertiary treatment process is nutrient removal. Nitrogen and phosphorus are
removed after biodegradation as a final treatment step before waste water is discharged to a
receiving body of water.
Applications - As previously mentioned, waste water collection, treatment, and storage are
common in many industrial categories and in POTW. Most industrial facilities and POTW collect,
contain, and treat waste water. However, some industries do not treat their waste water, but use
storage systems for temporary waste water storage or for accumulation of waste water for ultimate
disposal. For example, the Agricultural Industry does little waste water treatment but needs waste
water storage systems, while the Oil and Gas Industry also has a need for waste water disposal
systems.
The following are waste water treatment and storage applications identified by type of
industry:
1. Mining And Milling Operations - Storage of various waste waters such as acid mine water,
solvent wastes from solution mining, and leachate from disposed mining wastes.
Treatment operations include settling, separation, washing, sorting of mineral products
from tailings, and recovery of valuable minerals by precipitation.
2. Oil And Gas Industry - One of the largest sources of waste water. Operations treat brine
produced during oil extraction and deep-well pressurizing operations, oil-water mixtures,
gaseous fluids to be separated or stored during emergency conditions, and drill cuttings
and drilling muds.
3. Textile And Leather Industry - Treatment and sludge disposal. Organic species treated or
disposed of include dye carriers such as halogenated hydrocarbons and phenols. Heavy
metals treated or disposed of include chromium, zinc and copper. Tanning and finishing
wastes may contain sulfides and nitrogenous compounds.
4. Chemical And Allied Products Industry - Process waste water treatment and storage, and
sludge disposal. Waste constituents are process-specific and include organics and organic
phosphates, fluoride, nitrogen compounds, and assorted trace metals.
5. Other Industries - Treatment and storage operations are found at petroleum refining,
primary metals production, wood treating, and metal finishing facilities. Various industries
store and/or treat air pollution scrubber sludge and dredging spoils sludge (i. e., settled
solids removed from the floor of a surface impoundment).
4.13.2 Emissions
VOCs are emitted from waste water collection, treatment, and storage systems through
volatilization of organic compounds at the liquid surface. Emissions can occur by diffusive or
convective mechanisms, or both. Diffusion occurs when organic concentrations at the water
surface are much higher than ambient concentrations. The organics volatilize, or diffuse into the
air, in an attempt to reach equilibrium between aqueous and vapor phases. Convection occurs
4.13-6 EMISSION FACTORS 9/91
-------
when air flows over the water surface, sweeping organic vapors from the water surface into the
air. The rate of volatilization relates directly to the speed of the air flow over the water surface.
Other factors that can affect the rate of volatilization include waste water surface area,
temperature, and turbulence; waste water retention time in the system(s); the depth of the waste
water in the system(s); the concentration of organic compounds in the waste water and their
physical properties, such as volatility and diffusivity in water; the presence of a mechanism that
inhibits volatilization, such as an oil film; or a competing mechanism, such as biodegradation.
The rate of volatilization can be determined by using mass transfer theory. Individual gas
phase and liquid phase mass transfer coefficients (kg and kj, respectively) are used to estimate
overall mass transfer coefficients (K, KQJJ, and Kj)) for each VOC1"2 Figure 4.13-4 presents a
flow diagram to assist in determining the appropriate emissions model for estimating VOC
emissions from various types of waste water treatment, storage and collection systems.
Tables 4.13-1 and 4.13-2, respectively present the emission model equations and definitions.
VOCs vary in their degree of volatility. The emission models presented in this section can
be used for high, medium and low volatility organic compounds. The Henry's Law Constant
(HLC) is often used as a measure of a compound's volatility, or the diffusion of organics into the
air relative to diffusion through liquids. High volatility VOCs are HLC > 10"-* atm-m3/gmol;
medium volatility VOCs are 10'3 < HLC < 10'5 atm m3/gmol; and low volatility VOCs are
HLC < 10'5 atm-mS/gmol.1
The design and arrangement of collection, treatment and storage systems are facility -
specific, therefore the most accurate waste water emissions estimate will come from actual tests of
a facility (i. e., tracer studies or direct measurement of emissions from openings). If actual data
are unavailable, the emission models provided in this section can be used.
Emission models should be given site-specific information whenever it is available. The
most extensive characterization of an actual system will produce the most accurate estimates from
an emissions model. In addition, when addressing systems involving biodegradation, the accuracy
of the predicted rate of biodegradation is improved when site-specific compound biorates are
input. Reference 3 contains information on a test method for measuring site-specific biorates,
and Table 4.13-4 presents estimated biorates for approximately 150 compounds.
To estimate an emissions rate (N), the first step is to calculate individual gas phase and
liquid phase mass transfer coefficients kg and kj. These individual coefficients are then used to
calculate the overall mass transfer coefficient, K. Exceptions to this procedure are the calculation
of overall mass transfer coefficients in the oil phase, KQJI, and the overall mass transfer coefficient
for a weir, KT> KQJI requires only kg and Krj does not require any individual mass transfer
coefficients. The overall mass transfer coefficient is then used to calculate the emissions rates.
The following discussion describes how to use Figure 4.13-4 to determine an emission rate. An
example calculation is presented in 4.13.2.1 below.
Figure 4.13-4 is divided into two sections: Waste water treatment and storage systems, and
2) waste water collection systems. Waste water treatment and storage systems are further
segmented into aerated/nonaerated systems, biologically active systems, oil film layer systems, and
surface impoundment flowthrough or disposal. In flowthrough systems, waste water is treated and
discharged to a POTW or a receiving body of water, such as a river or stream. All waste water
collection systems are by definition flowthrough. Disposal systems, on the other hand, do not
discharge any waste water.
9/91 Evaporation Loss Sources 4.13-7
-------
Yes
/ Diffusedl\
Wastewater
Treatment and \Aerated?
Storage
No
•TI
£
a Numbered equations are presented In Table 4.13-1
K| - Individual liquid phase mass transfer coefficient, m/is
individual gas phase mass transfer coefficient, m/s
Overall mass transfer coefficient In the oil phase, m/s
Volatilization • reaeratton theory mass transfer coefficient
K - Overall mass transfer coefficient, m/s
N » Emissions, fl/s
Wastewater Collection
Weir?
>
3
>
S
>
3
Yes
\
^
No
t
^
r
./
X
>
•to
\
^
res
— | Ftowthrough
— ] Disposal
— 1 Ftowthrough
1 Disposal
— | Ftowthrough
1 — | Disposal
— I Ftowthrough
1 1 Disposal
— | Ftowthrough
' — | Disposal
— | Ftowthrough
' — | Disposal
~ ' 1 Junction OOK
1 Ult Station
— 1 Sump
H Weir
J
Equations Used to Obtain:8
Ji_ Jg 1^ Jfc_ K N
12 7 14
12 7 13
1.3 2,4 7 16
1.3 2,4 7 15
12 7 16
12 7 15
12 7 12
12 7 11
29 18
29 17
29 22
29 23
32 7 12
32 7 12
1 1 2 7 12
10 21
OarHlerWelr
5 6
8 24
Figure 4.13-4. Flow diagram for estimating VOC emissions from
waste water collection, treatment and storage systems.
4.13-8
EMISSION FACTORS
9/91
-------
Table 4.13-1. MASS TRANSFER CORRELATIONS
AND EMISSIONS EQUATIONS".
Equation Equations
No.
Individual liquid (k^ and gas (kg) phase mass transfer coefficients
1 k, (m/s) = (2.78 x 10-6)(Dw/Dether)2/3
For: 0 < UJQ < 3.25 m/s and all F/D ratios
k, (m/s) = [(2.605 x 1Q-9)(F/D) + (1.277 x 10-7)](U10)2(Dw/Dether)2/3
For: U > 3.25 m/s and 14 < F/D < 51.2
k, (m/s) = (2.61 x 10-7)(U10)2(Dw/Dether)2/3
For: UIQ > 3.25 m/s and F/D > 51.2
k| (m/s) = 1.0 x 10-6 -I- 144 x 10'4 (U*)2-2 (Scr)'0-5; U* < 0.3
k| (m/s) = 1.0 x lO'6 + 34.1 x 10"4 U* (ScL)-°-3; U* > 0.3
For Ujo > 3.25 m/s and F/D < 14
where:
U* (m/s) = (0.01)(U10)(6.1 + 0.63(Uio))°-5
F/D = 2 (A/n)
kg (m/s) = (4.82 x 10-3)(U10)°-78 (ScG)-°-67 (de)'0'11
where:
de(m) =
k| (m/s) = [(8.22 x 10-9)(J)(POWR)(1.024)(T-20)(Ot)(lo6) *
(MWL)/(VavpL)](Dw/D02,w)6-5
where:
POWR (hp) = (total power to aerators)(V)
Vav (ft2) = (fraction of area agitated)(A)
kg (m/s) = (1.35 x lQ-7)(Re)l-42 (P)°-4 (ScG)°-5 (Fr)-°-21(Da MWa/d)
where:
Re = d2 w pa/na
P = [(0.85)(POWR)(550 ft-l
Fr = (d^w/^
k| (m/s) = (fair,«)(Q)/[3600 s/min (hc)(7idc)]
where:
fair | = 1 - 1/r
'r= exp [0.77(hc)0-623(Q/ndc)°-66(Dw/Do2,w)a66]
9/91 Evaporation Loss Sources 4.13-9
-------
Table 4.13-1. MASS TRANSFER CORRELATIONS
AND EMISSIONS EQUATIONS8.
Equation Equations
No.
kg (m/s) = 0.001 + (0.0462(U**)(ScG)-0.67)
where:
U** (m/s) = [6.1 + (0.63)(U1o)]°-5(Uio/100)
Overall mass transfer coefficients for water (K) and oil KQJI phases
and for Weirs (KjV)
7 K = (kt Keq kg)/(Keq kg + kj)
where:
Keq = H/(RT)
8 K (m/s) = [[MWL/(kjpL*(10o cm/m)] + [MWa/(kgPaH*
55,555(100 cm/m))]]-1 Mw£/[(100 cm/m)pL]
ii = kgKeq0ii
where:
il = P*PaMWoil/(poil MWa Po)
10 KD = 0.16h (Dw/Do2,w)°-75
Air emissions ( N)
11 N(g/s) = (1 - Ct/Co) V Co/t
where:
Ct/Co = exp[-K A t/V]
12 N(g/s) = K CL A
where*
CL(g/m3) = Q Co/(KA + Q)
13 N(g/s) = (1 - Ct/Co) V Co/t
where:
Ct/Co = exp[-(KA + KeqQa)t/V]
14 N(g/s) = (KA + QaKeq)CL
where:
= QCo/(KA + Q + QaKeq)
15 N(g/s) = (1 - Ct/Co) KA/(KA + Kmax bj V/Kg) V Co/t
where:
Ct/Co = exp[-Kmax bj t/Kj - K A t/V]
4.13-10 EMISSION FACTORS 9/91
-------
Table 4.13-1. MASS TRANSFER CORRELATIONS
AND EMISSIONS EQUATIONS'.
Equation Equations
No.
16 N(g/s) = K CL A
where:
CL(g/m3) = [-b + (b2 - 4ac)°-5]/(2a)
and
a = KA/Q + 1
b = Ks(KA/Q + 1) + Kmax bj V/Q - Co
17 N(g/s) = (1 -
where:
Ctoji/Co0ji = exp[-Koii t/D0ii]
and:
Co0ii = Kow Co/[l - FO + FO(Kow)]
V0il = (FO)(V)
il = (FO)(V)/A
18 N(g/s) =
where:
CL,oil(g/m3) = QoilCo0il/(KoilA + Qoii)
and:
Co0jl = Kow Co/[l - FO + FO(Kow)]
Qoil = (FO)(Q)
19 N(g/s) = (1 - Ct/Co)(KA + QaKeq)/(KA + QaKeq + Kmax bj V/Kg) V Co/t
where:
Ct/Co = exp[-(KA + KeqQa)tA^ - Kmax b{
20 N(g/s) = (KA + QaKeq)CL
\vhcrc"
CL(g/m3) = [-b +(b2 - 4ac)°-5]/(2a)
and:
a = (KA + QaKeq)/Q + 1
b = Ks[(KA + QaKeq)/Q + 1] + Kmax bj V/Q - Co
c = -
21 N (g/s) = (1 - exp[-KD])Q Co
22 N(g/s) =
where:
CL,oil(g/m3) = Qoil(Co0il*)/(KoilA + Qoil)
and:
Cooil* = Co/FO
Qoil = (FO)(Q)
9/91 Evaporation Loss Sources 4.13-11
-------
Table 4.13-1. MASS TRANSFER CORRELATIONS
AND EMISSIONS EQUATIONS8.
Equation Equations
No.
23 N(g/s) = (1 -
where:
Ctoil/Co0ii* = exp[-Koii t/Doii]
and:
Co0a* = Co/FO
Veil = (FO)(V)
il = (FO)(V)/A
24 N (g/s) = (1 - exp[-K TC dc VQ])Q Co
aAll parameters in numbered equations are defined in Table 4.13-2.
4.13-12 EMISSION FACTORS 9/91
-------
Table 4.13-2. PARAMETER DEFINITIONS FOR MASS TRANSFER
CORRELATIONS AND EMISSIONS EQUATIONS.
Parameter
A
bi
CL
CL,oil
Co
Cooji
COoil
Q
oca
d
D
d*
Da
dc
de
Dether
D02,w
Definition
Waste water surface area
Biomass concentration (total biological solids)
Concentration of constituent in the
liquid phase
Concentration of constituent in the
oil phase
Initial concentration of constituent
in the liquid phase
Initial concentration of constituent
in the oil phase considering mass
transfer resistance between water
and oil phases
Initial concentration of constituent in
the oil phase considering no mass transfer
resistance between water and oil phases
Concentration of constituent in the
liquid phase at time = t
Concentration of constituent in the
oil phase at time = t
Impeller diameter
Waste water depth
Impeller diameter
Diffusivity of constituent in air
Clarifier diameter
Effective diameter
Diffusivity of ether in water
Diffusivity of oxygen in water
Units
m2 or ft2
g/m3
g/m3
g/m3
g/m3
g/m3
g/m3
g/m3
g/m3
cm
m or ft
ft
cm2/s
m
m
cm2/s
cm2/s
Code"
A
B
D
D
A
D
D
D
D
B
A,B
B
C
B
D
(8.5xlO-6)b
(2.4xlO-5)b
9/91 Evaporation Loss Sources 4.13-13
-------
Table 4.13-2. PARAMETER DEFINITIONS FOR MASS TRANSFER
CORRELATIONS AND EMISSIONS EQUATIONS.
Parameter
Doil
Dw
fair,l
F/D
FO
Fr
gc
h
he
H
J
K
KD
Keq
Keq0il
kg
Definition
Oil film thickness
Diffusivity of constituent in water
Fraction of constituent emitted to the
air, considering zero gas resistance
Fetch to depth ratio, dg/D
Fraction of volume which is oil
Froude number
Gravitation constant
(a conversion factor)
Weir height (distance from the
waste water overflow to the receiving
body of water)
Clarifier weir height
Henry's Law Constant of constituent
Oxygen transfer rating of
surface aerator
Overall mass transfer coefficient for
transfer of constituent from liquid
phase to gas phase
Volatilization-reaeration theory mass
transfer coefficient
Equilibrium constant or partition
coefficient (concentration in gas
phase/concentration in liquid phase)
Equilibrium constant or partition
coefficient (concentration in gas
phase/concentration in oil phase)
Gas phase mass transfer coefficient
Units
m
cm^/s
dimensionless
dimensionless
dimensionless
dimensionless
Ibm-ft/s2-lbf
ft
m
atm-m^/gmol
Ib 02/(hr-hp)
m/s
dimensionless
dimensionless
dimensionless
m/s
Code"
B
C
D
D
B
D
32.17
B
B
C
B
D
D
D
D
D
4.13-14
EMISSION FACTORS
9/91
-------
Table 4.13-2. PARAMETER DEFINITIONS FOR MASS TRANSFER
CORRELATIONS AND EMISSIONS EQUATIONS.
Parameter
k«
Kmax
Koil
Kow
KS
MWa
MWoil
MWL
N
Nl
ot
P
P*
PO
POWR
Q
Qa
Qoil
r
Definition
Liquid phase mass transfer coefficient
Maximum biorate constant
Overall mass transfer coefficient for
transfer of constituent from oil
phase to gas phase
Octanol-water partition coefficient
Half saturation biorate constant
Molecular weight of air
Molecular weight of oil
Molecular weight of water
Emissions
Number of aerators
Oxygen transfer correction factor
Power number
Vapor pressure of the constituent
Total pressure
Total power to aerators
Volumetric flow rate
Diffused air flow rate
Volumetric flow rate of oil
Deficit ratio (ratio of the difference
between the constituent concentration
at solubility and actual constituent
concentration in the upstream and the
downstream)
Units
m/s
g/s-g biomass
m/s
dimensionless
g/m3
g/gmol
g/gmol
g/gmol
g/s
dimensionless
dimensionless
dimensionless
atm
atm
hp
m3/s
m3/s
m3/s
dimensionless
Code"
D
A,C
D
C
A.C
29
B
18
D
A,B
B
D
C
A
B
A
B
B
D
9/91
Evaporation Loss Sources
4.13-15
-------
Table 4.13-2. PARAMETER DEFINITIONS FOR MASS TRANSFER
CORRELATIONS AND EMISSIONS EQUATIONS.
Parameter Definition
R Universal gas constant
Re Reynolds number
SCQ Schmidt number on gas side
SCL Schmidt number on liquid side
T Temperature of water
t Residence time of disposal
U* Friction velocity
U** Friction velocity
Units
~v«
dimensionless
dimensionless
dimensionless
°C or Kelvin (K)
s
m/s
m/s
UIQ Wind speed at 10 m above the liquid surface m/s
V Waste water Volume
Vav Turbulent surface area
Voil Volume of oil
w Rotational speed of impeller
pa Density of air
PL Density of water
Poil density of oil
|ia Viscosity of air
UL Viscosity of water
aCode
m3 or ft3
ft2
m3
rad/s
g/cm3
g/cm3 or lb/ft3
g/m3
g/cm-s
g/cm-s
Code"
8.21xlO-5
D
D
D
A
A
D
D
B
A
B
B
B
(1.2xlO-3)b
lb or 62.4b
B
(1.81xl(H)b
(8.93xlO-3)b
A = Site-specific parameter.
B = Site-specific parameter. For default values, see Table 4.13-3.
C = Parameter can be obtained from literature. See Attachment 1 for a
list of —150 compound chemical properties
D = Calculated value.
bReported values at 25°C (298°K).
at T = 25°C (298°K).
4.13-16
EMISSION FACTORS
9/91
-------
Figure 4.13-4 includes information needed to estimate air emissions from junction boxes,
lift stations, sumps, weirs, and clarifier weirs. Sumps are considered quiescent, but junction boxes,
lift stations, and weirs are turbulent in nature. Junction boxes and lift stations are turbulent
because incoming flow is normally above the water level in the component, which creates some
splashing. Waste water falls or overflows from weirs and creates splashing in the receiving body
of water (both weir and clarifier weir models). Waste water from weirs can be aerated by
directing it to fall over steps, usually only the weir model.
Assessing VOC emissions from drains, manholes and trenches is also important in
determining the total waste water facility emissions. As these sources can be open to the
atmosphere and closest to the point of waste water generation (i. e., where water temperatures
and pollutant concentrations are greatest), emissions can be significant. Currently, there are no
well established emission models for these collection system types. However, work is being
performed to address this need.
Preliminary models of VOC emissions from waste collection system units have been
developed4. The emission equations presented in Reference 4 are used with standard collection
system parameters to estimate the fraction of the constituents released as the waste water flows
through each unit. The fractions released from several units are estimated for high, medium and
low volatility compounds. The units used in the estimated fractions included open drains,
manhole covers, open trench drains, and covered sumps.
The numbers in Figure 4.13-4 under the columns for kf, kg, KQJI, Kj), K, and N refer to
the appropriate equations in Table 4.13-1.* Definitions for all parameters in these equations are
given in Table 4.13-2. Table 4.13-2 also supplies the units that must be used for each parameter,
with codes to help locate input values. If the parameter is coded with the letter A, a site-specific
value is required. Code B also requires a site-specific parameter, but defaults are available.
These defaults are typical or average values and are presented by specific system in Table 4.13-3.
Code C means the parameter can be obtained from literature data. Table 4.13-4 contains
a list of approximately 150 chemicals and their physical properties needed to calculate emissions
from waste water, using the correlations presented in Table 4.13-1. All properties are at 25°C.
(A more extensive chemical properties data base is contained in Appendix C of Reference 1.)
Parameters coded D are calculated values.
Calculating air emissions from waste water collection, treatment and storage systems is a
complex procedure, especially if several systems are present. Performing the calculations by hand
may result in errors and will be time consuming. A personal computer program called the Surface
Impoundment Modeling System (SIMS) is now available for estimating air emissions. The
program is menu driven and can estimate air emissions from all surface impoundment models
presented in Figure 4.13-4, individually or in series. The program requires for each collection,
treatment or storage system component, at a minimum, the waste water flow rate and component
surface area. All other inputs are provided as default values. Any available site-specific
information should be entered in place of these defaults, as the most fully characterized system
will provide the most accurate emissions estimate.
*A11 emission model systems presented in Figure 4.13-4 imply a completely mixed or uniform
waste water concentration system. Emission models for a plug flow system, or system in which
there is no axial, or horizontal mixing, are too extensive to be covered in this document. (An
example of plug flow might be a high waste water flow in a narrow channel.) For information on
emission models of this type, see Reference 1.
9/91 Evaporation Loss Sources 4.13-17
-------
The SIMS program with user's manual and background technical document, can be
obtained through state air pollution control agencies and through the U.S. Environmental
Protection Agency's Control Technology Center in Research Triangle Park, NC, telephone (919)
541-0800 (FTS 629-0800). The user's manual and background technical document should be
followed to produce meaningful results.
The SIMS program and user's manual also can be downloaded from EPA's Clearinghouse
For Inventories and Emission Factors (CHIEF) electronic bulletin board (BB). The CHIEF BB
is open to all persons involved in air emission inventories. To access this BB, one needs a
computer, modem, and communication package capable of communicating at 1200, 2400, or
9600 baud, 8 data bits, 1 stop bit, and no parity (8-N-l).
This BB is part of EPA's Technical Support Division bulletin local system and its telephone
numbers are:
1) (919) 541-5742 (1200 or 2400 baud), and
2) (919) 541-5384 (9600 baud).
First-time users must register before access is allowed.
Emissions estimates from SIMS are based on mass transfer models developed by Emissions
Standards Division (ESD) during evaluations of TSDFs and VOC emissions from industrial waste
water. As a part of the TSDF project, a Lotus* spreadsheet program called CHEMDAT7 was
developed for estimating VOC emissions from waste water land treatment systems, open landfills,
closed landfills, and waste storage piles, as well as from various types of surface impoundments.
For more information about CHEMDAT7, contact the ESD's Chemicals And Petroleum Branch
(MD 13), US EPA, Research Triangle Park, NC 27711.
4.13-18 EMISSION FACTORS 9/91
-------
Table 4.13-3. SITE-SPECIFIC DEFAULT PARAMETERS8
Default
Parameter*
Definition
Default Value
General
T Temperature of water
UIQ Windspeed
Biotreatment Systems
bj Biomass concentration (for biologically
active systems)
Quiescent treatment systems
Aerated treatment systems
Activated sludge units
POWR Total power to aerators
(for aerated treatement systems)
(for activated sludge)
W Rotational speed of impeller
(for aerated treatment systems)
d(d ) Impeller diameter
(for aerated treatment systems)
Vav Turbulent surface area
(for aerated treatment systems)
(for activated sludge)
J Oxygen transfer rating to surface aerator
(for aerated treatment systems)
O{ Oxygen transfer correction factor
(for aerated treatment systems)
Nj Number of aerators
Diffused Air Systems
Qa Diffused air volumetric flow rate
298°K
4.47 m/s
50g/m3
300 g/m3
4000 g/m3
0.75 hp/1000 ft3 (V)
2 hp/1000 ft3 (V)
126 rad/s(1200 rpm)
61 cm(2 ft)
0.24 (A)
0.52 (A)
3 Ib O2/hp»hr
0.83
POWR/75
0.0004(V) m3/s
9/91
Evaporation Loss Sources
4.13-19
-------
Table 4.13-3. SITE-SPECIFIC DEFAULT PARAMETERS"
Default
Parameter Definition Default Value
Oil Film Layers
MWoji Molecular weight of oil 282 g/gmol
D0il Depth of oil layer 0.001 (V/A) m
VOH Volume of oil 0.001 (V) m3
°-oil Volumetric flow rate of oil 0.001 (Q) m3/s
Poil Density of oil 0.92 g/cm3
FO Fraction of volume which is oilc 0.001
Junction Boxes
D Depth of Junction Box 0.9 m
Nj Number of aerators 1
Lift Station
D Depth of Lift Station 1.5 m
NI Number of aerators 1
Sump
D Depth of sump 5.9 m
Weirs
dc Clarifier weir diameter11 28.5 m
h Weir height 1.8m
^ Clarifier weir height" 0.1 m
"Reference 1.
bAs defined in Table 4.13-2.
"Reference 4.
Reference 2.
"Reference 5.
4.13-20 EMISSION FACTORS 9/91
-------
4.13.2.1 Example Calculation
An example industrial facility operates a flowthrough, mechanically aerated biological
treatment impoundment that receives waste water contaminated with benzene at a concentration
of 10.29 g/m*
The following format is used for calculating benzene emissions from the treatment process:
I. Determine which emission model to use
II. User-supplied information
III. Defaults
IV. Pollutant physical property data and water, air, and other properties
V. Calculate individual mass transfer coefficient
VI. Calculate the overall mass transfer coefficients
. Calculate VOC emissions
I. Determine Which Emission Model To Use - Following the flow diagram in Figure 4.13-4,
the emission model for a treatment system that is aerated, but not by diffused air, is
biologically active, and is a flowthrough system, contains the following equations:
Equation Nos.
Parameter Definition from Table 4.13-1
k| Individual liquid phase mass
transfer coefficient, m/s 1,3
kg Individual gas phase mass
transfer coefficient, m/s 2,4
K Overall mass transfer coefficient, m/s 7
N VOC emissions, g/s 16
II. User-supplied Information - Once the correct emission model is determined, some site-
specific parameters are required. As a minimum for this model, site-specific flow rate,
waste water surface area, and depth, and pollutant concentration should be provided. For
this example, these parameters have the following values:
Q = Volumetric flow rate = 0.0623 m3/s
D = Waste water depth = 1.97 m
A = Waste water surface area = 17,652 m^
Co = Initial benzene concentration in the liquid phase = 10.29 g/m3
III. Defaults - Defaults for some emission model parameters are presented in Table 4.13-3.
Generally, site-specific values should be used when available. For this facility, all available
general and biotreatment system defaults from Table 4.13-3 were used:
jQ
T
= Wind speed at 10 m above the liquid surface = e = 4.47 m/s
= Temperature of water = 25°C (298°K)
j = Biomass concentration for aerated treatment systems = 300
g/m3
J = Oxygen transfer rating to surface aerator = 3 Ib O2/hp-hr
9/91 Evaporation Loss Sources 4.13*21
-------
POWR = Total power to aerators = 0.75 hp/1,000 ft3
Oj = Oxygen transfer correction factor = 0.83
Vav = Turbulent surface area = 0.24 (A)
d = Impeller diameter = 61 cm
d* = Impeller diameter = 2 ft
w = Rotational speed of impeller = 126 rad/s
NI = Number of aerators = POWR/75 hp
IV. Pollutant Physical Property Data, And Water, Air and Other Properties - For each
pollutant, the specific physical properties needed by this model are listed in Table 4.13-4.
Water, air and other property values are given in Table 4.13-2.
A. Benzene (from Table 4.13-4)
Dw,benzene = Diffusivity of benzene in water = 9.8 x 10"*> cm^/s
Da benzene = Diffusivity of benzene in air = 0.088 cm^/s
"benzene = Henry's Law Constant for benzene = 0.0055 atm-
m^/gmol
Kmaxbenzene = Maximum biorate constant for benzene = 5.28 x 10"^
g/g-s
KS benzene = Half saturation biorate constant for benzene = 13.6 g/m-*
B. Water, Air and Other Properties (from Table 4.13-3)
pa = Density of air = 1.2 x 10^ g/cm^
PL = Density of water = 1 g/cm^ (62.4 lbm/ft3)
fia = Viscosity of air = 1.81 x 10'^ g/cm-s
DQ2,w = Diffusivity of oxygen in water = 2.4 x 10~5 cm^/s
Aether = Diffusivity of ether in water = 8.5 x 10"^ cm^/s
M\VL = Molecular weight of water = 18 g/gmol
MWa = Molecular weight of air = 29 g/gmol
gc = Gravitation constant = 32.17 lbm-ftAbf-s2
R = Universal gas constant = 8.21 x 10'^ atm-m^/gmol
V. Calculate Individual Mass Transfer Coefficients - Because part of the impoundment is
turbulent and part is quiescent, individual mass transfer coefficients are determined for both
turbulent and quiescent areas of the surface impoundment.
Turbulent area of impoundment - Equations 3 and 4 from
Table 4.13-1.
A. Calculate the individual liquid mass transfer coefficient, kj:
k|(m/s) = [(8.22 x 10-9)(J)(POWR)(1.024)(T-20) *
(Ot)(lo6)MWL/(VavPL)](Dw/Do2)W)0-5
The total power to the aerators, POWR, and the turbulent surface area, Vav, are
calculated separately [Note: some conversions are necessary.]:
1) Calculate total power to aerators, POWR (Default presented in III):
POWR (hp) = 0.75 hp/1,000 ft3 (V)
4.13-22 EMISSION FACTORS 9/91
-------
V — waste water volume, m3
V (m3) = (A)(D) = (17,652 m2)(1.97 m)
V = 34,774 m3
POWR = (0.75 hp/1,000 ft3)(ft3/0.028317 m3)(34,774 m3)
POWR = 921 hp
2) Calculate turbulent surface area, Vav, (Default presented in III):
Vav (ft2) = 0.24 (A)
Vav = 0.24(17,652 m2)(10.758 ft2/m2)
Vav = 45,576 ft2
Now, calculate k|, using the above calculations and information from II, III, and IV:
k, (m/s) = [(8.22 x 10-9)(3 lbO2/hp-hr)(921 hp) *
(1.024)(25-20)(0.83)jfl06)(18 g/gmol)/
((45,576 ft2)(l e/cm3))] *
[(9.8 x 10-6 Cm2/s)/(2.4 x 10'5 cm2/s)]°-5
kt (m/s) = (0.00838)(p.639)
kj = 5.35 x 10-3 m/s
B. Calculate the individual gas phase mass transfer coefficient, k«:
kg (m/s) = (1.35 x 10-7)(Re)l-42(P)0-4(ScG)°-5(Fr)-°-21(DaMWa/d)
The Reynolds number, Re, power number, P, Schmidt number on the gas side, SCQ,
and Froude's number Fr, are calculated separately:
1) Calculate Reynolds Number, Re:
Re = d2wPa/ua
Re = (61 cm)2(126 rad/s)(1.2 x 10'3 g/cm3)/(1.81 x 10'4 g/cm-s)
Re = 3.1 x 106
2) Calculate power number, P:
P = [(0.85)(POWR)(550 ft-lbf/s-hp)^] gc/(pL(d*)5 w3)
NI = POWR/75 hp (default presented in III)
P = (0.85)(75 hp)(POWR/POWR)(550 ft-lbf/s-hp) *
(32.17 lbm-ft/lbf-s2)/[(62.4 Ibm/ft3)(2 ft)5(126 rad/s)3]
P = 2.8 x 10-4
3) Calculate Schmidt Number on the gas side, SCQ:
ScG = ua/(paDa)
ScG = (1.81 x 10-4 g/cm-s)/[(1.2 x 10'3 g/cm3)(0.088 cm2/s)]
ScG = 1-71
4) Calculate Froude Number, Fr:
Fr = (d*)w2/fc
Fr = (2 ft)(126 rad/s)2/(32.17 lbm-ft/Ibf-s2)
Fr = 990
Now calculate kg using the above calculations and information from II, III, and IV:
kg (m/s) = (1.35 x 10-7)(3.1 x lO6)1-42^ x 10-4)0-4(1.71)0-5 *
9/91 Evaporation Loss Sources 4.13-23
-------
(990)-0.21(o.088 cm2/s)(29 g/gmol)/(61 cm)
kg = 0.109 m/s
Quiescent surface area of impoundment - Equations 1 and 2 from
Table 4.13-1
A. Calculate the individual liquid phase mass transfer coefficient, kf :
F/D = 2(A/n)°-5/D
= 2(17,652 m2/it)0.5/(i.97 m)
= 76.1
= 4.47 m/s
For Ujo > 3.25 m/s and F/D > 51.2 use the following:
k, (m/s) = (2.61 x lO-^oftDy/Dethe 0
k| (m/s) = (2.61 x 10-r)(4.47 m/s)2[(9.8 x ID"6 cm2/s)/
(8.5 x 10-6 cm2/s)]2/3
k| = 5.74 x 10-6 m/s
B. Calculate the individual gas phase mass transfer coefficient, kg:
kg = (4.82 x 10-3)(U10r7«(ScG)-0-67(de)-0.11
The Schmidt number on the gas side, SCQ, and the effective diameter, dg, are
calculated separately:
1) Calculate the Schmidt Number on the gas side, SCQ:
= Ha/(Pa^a) = 1-71 (same as for turbulent impoundments)
2) Calculate the effective diameter, de:
de (m) = 2(A/n)0-5
de (m) = 2(17,652 m2/*)0-5
de = 149.9 m
kg (m/s) = (4.82 x 10-3)(4.47 m/s)0-78 (l.71)-0-67 (149.9 m)-0-*1
kg = 6.24 x 10-3 m/s
VI. Calculate The Overall Mass Transfer Coefficient. Because part of the impoundment is
turbulent and part is quiescent, the overall mass transfer coefficient is determined as an
area-weighted average of the turbulent and quiescent overall mass transfer coefficients.
(Equation number 7 from Table 4.13-1)
Overall mass transfer coefficient for the turbulent surface area of impoundment.
KT
KT (m/s) = (kjKeqkgXKeqkg + k<)
Keq = H/RT
Keq = (0.0055 atm-m3/gmol)/[(8.21 x 10'5 atm-m3/ gmol-°K)(298°K)]
Keq = 0.225
KT (m/s) = (5.35 x lO"3 tn/s)(0.225)(0.109)/[(0.109 m/s)(0.225) +
(5.35 x lO-6 m/s)]
KT = 4.39 x lO'3 m/s
4.13-24 EMISSION FACTORS 9/91
-------
Overall mass transfer coefficient for the quiescent surface area of
impoundment.
KQ (m/s) = (k|Keqkg)/rKeqkg + k|)
KQ (m/s) = (5.74 x fir* m/s)(0.225)(6.24 x 10-3
[(6.24 X 10-3 m/s)(0.225) + (5.74 x 10-° m/s)]
KQ = 5.72 x 10-6 m/s
Overall mass transfer coefficient. K weighted by turbulent and quiescent surface
areas. AT and AQ
K (m/s) = (KTAT + KQAQ)/A
AT = 0.24(A) (Default value presented in HI: AT = Vay)
AQ = (1 - 0.24)A
K (m/s) = [(4.39 x 10~3 m/s)(0.24 A) + (5.72 x 10-6 m/s)(l - 0.24)A]/A
K = 1.06 x lO-3 m/s
VII. Calculate VOC Emissions for an Aerated Biological Flowthrough Impoundment - Equation
number 16 from Table 4.13-1
N (g/s) = K CL A
where:
CL (g/m3) = [-b + (b2 - 4ac)°-5]/(2a)
and:
a = KA/Q + 1
b = Ks(KA/Q + 1) + Kmax bj V/Q - Co
Calculate a, b, c, and the concentration of benzene in the liquid phase, CL, separately:
1) Calculate a:
a = (KA/Q + 1) = [(1.06 x 10'3 m/s)(17,652 m2)/(0.0623 m3/s)] + 1
a = 301.3
2) Calculate b (V = 34,774 m3 from IV):
b = Kg (KA/Q + 1) + Kmax bj V/Q - Co
b = (13.6 g/m3)[(1.06 x 10'3 m/s)(17,652 m2)/(0.0623 m3/s)] +
[(5.28 x 10-o g/g.s)(300 g/m3)(34,774 m%0.0623 m3/s)] - 10.29 g/m3
b = 4,084.6 + 884.1 - 10.29
b = 4,958.46 g/m3
3) Calculate c:
c = -(13.6 g/m3)(10.29 g/m3)
c = -139.94
4) Calculate the concentration of benzene in the liquid phase, CL, from a, b, and c above:
CL (g/m3) = I-b +• (b2 - 4ac)0-5]/(2a)
CL (g/m3) = [(4,958.46 g/m3) + [(4,958.46 g/m3)2 -
[4(301.3)(-139.94)]]°-5]/(2(301.3))
9/91 Evaporation Loss Sources 4.13-25
-------
CL = 0.0282 g/m3
Now calculate N with the above calculations and information from II and V:
N (g/s) = K A CL
N (g/s) = (1.06 x lO'3 m/s)(17,652 m2)(0.0282 g/m3)
N = 0.52 g/s
4.13.3 Controls
The types of control technology generally used in reducing VOC emissions from waste water
include: steam stripping or air stripping, carbon adsorption (liquid phase), chemical oxidation,
membrane separation, liquid-liquid extraction, and biotreatment (aerobic or anaerobic). For
efficient control, all control elements should be placed as close as possible to the point of waste
water generation, with all collection, treatment and storage systems ahead of the control
technology being covered to suppress emissions. Tightly covered, well maintained collection
systems can suppress emissions by 95 to 99 percent. However, if there is explosion potential, the
components should be vented to a control device such as an incinerator or carbon adsorber.
The following are brief descriptions of the control technology listed above and of any
secondary controls that may need to be considered for fugitive air emissions.
Steam stripping is the fractional distillation of waste water to remove volatile organic
constituents, with the basic operating principle being the direct contact of steam with waste water.
The steam provides the heat of vaporization for the more volatile organic constituents. Removal
efficiencies vary with volatility and solubility of the organic impurities. For highly volatile
compounds (HLC greater than 10"3 atm-nv'/gmol), average VOC removal ranges from 95 to
99 percent. For medium volatility compounds (HLC between 10~* and 10~3 atm-m3/gmol),
average removal ranges from 90 to 95 percent. For low volatility compounds (HLC < 10"^ atm-
m3/gmol), average removal ranges from less than 50 to 90 percent.
Air stripping involves the contact of waste water and air to strip out volatile organic
constituents. By forcing large volumes of air through contaminated water, the surface area of
water in contact with air is greatly increased, resulting in an increase in the transfer rate of the
organic compounds into the vapor phase. Removal efficiencies vary with volatility and solubility
of organic impurities. For highly volatile compounds, average removal ranges from 90 to
99 percent, for medium to low volatility compounds, removal ranges from less than 50 to
90 percent.
Steam stripping and air stripping controls most often are vented to a secondary control, such
as a combustion device or gas phase carbon adsorber. Combustion devices may include
incinerators, boilers and flares. Vent gases of high fuel value can be used as an alternate fuel.
Typically, vent gas is combined with other fuels such as natural gas and fuel oil. If the fuel value
is very low, vent gases can be heated and combined with combustion air. It is important to note
that organics such as chlorinated hydrocarbons can emit toxic pollutants when combusted.
Secondary control by gas phase carbon adsorption processes takes advantage of compound
affinities for activated carbon. The types of gas phase carbon adsorption systems most commonly
used to control VOC are fixed bed carbon adsorbers and carbon canisters. Fixed bed carbon
adsorbers are used to control continuous organic gas streams with flow rates ranging from 30 to
4.13-26 EMISSION FACTORS 9/91
-------
over 3000 m3/mm Canisters are much simpler and smaller than fixed bed systems and are usually
installed to control gas flows of less than 3 m^/min.4 Removal efficiencies depend highly on the
type of compound being removed. Pollutant-specific activated carbon is usually required.
Average removal efficiency ranges from 90 to 99 percent.
Like gas phase carbon adsorption, liquid phase carbon adsorption takes advantage of
compound affinities for activated carbon. Activated carbon is an excellent adsorbent, because of
its large surface area and because it is usually in granular or powdered form for easy handling.
Two types of liquid phase carbon adsorption are the fixed bed and moving bed systems. The fixed
bed system is used primarily for low flow waste water streams with contact times around 15
minutes, and it is a batch operation (i. e., once the carbon is spent, the system is taken off line).
Moving bed carbon adsorption systems operate continuously with waste water typically being
introduced from the bottom of the column and regenerated carbon from the top (countercurrent
flow). Spent carbon is continuously removed from the bottom of the bed. Liquid phase carbon
adsorption is usually used for low concentrations of nonvolatile components and for high
concentrations of nondegradable compounds.5 Removal efficiencies depend on whether the
compound is adsorbed on activated carbon. Average removal efficiency ranges from 90 to
99 percent.
Chemical oxidation involves a chemical reaction between the organic compound and an
oxidant such as ozone, hydrogen peroxide, permanganate, or chlorine dioxide. Ozone is usually
added to the waste water through an ultraviolet - ozone reactor. Permanganate and chlorine
dioxide are added directly into the waste water. It is important to note that adding chlorine
dioxide can form chlorinated hydrocarbons in a side reaction. The applicability of this technique
depends on the reactivity of the individual organic compound.
Two types of membrane separation processes are ultrafiltration and reverse osmosis.
Ultrafiltration is primarily a physical sieving process driven by a pressure gradient across the
membrane. This process separates organic compounds with molecular weights greater than 2000,
depending on the size of the membrane pore. Reverse osmosis is the process by which a solvent
is forced across a semipermeable membrane because of an osmotic pressure gradient. Selectivity
is, therefore, based on osmotic diffusion properties of the compound and on the molecular
diameter of the compound and membrane pores.4
Liquid-liquid extraction as a separation technique involves differences in solubility of
compounds in various solvents. Contacting a solution containing the desired compound with a
solvent in which the compound has a greater solubility may remove the compound from the
solution. This technology is often used for product and process solvent recovery. Through
distillation, the target compound is usually recovered, and the solvent reused.
Biotreatment is the aerobic or anaerobic chemical breakdown of organic chemicals by
microorganisms. Removal of organics by biodegradation is highly dependent on the compound's
biodegradability, its volatility, and its ability to be adsorbed onto solids. Removal efficiencies
range from almost zero to 100 percent. In general, highly volatile compounds such as chlorinated
hydrocarbons and aromatics will biodegrade very little because of their high volatility, while
alcohols and other compounds soluble in water, as well as low volatility compounds, can be almost
totally biodegraded in an acclimated system. In the acclimated biotreatment system, the
microorganisms easily convert available organics into biological cells, or biomass. This often
requires a mixed culture of organisms, where each organism utilizes the food source most suitable
9/91 Evaporation Loss Sources 4.13-27
-------
to its metabolism. The organisms will starve and the organics will not be biodegraded if a system
is not acclimated, i. e., the organisms cannot metabolize the available food source.
4.13.4 Glossary Of Terms
Basin - an earthen or concrete-lined depression used to hold liquid.
Completely mixed - having the same characteristics and quality throughout or at all times.
Disposal - the act of permanent storage. Flow of liquid into, but not out of a device.
Drain - a device used for the collection of liquid. It may be open to the atmosphere or be
equipped with a seal to prevent emissions of vapors.
Flowthrough - having a continuous flow into and out of a device.
Plug flow - having characteristics and quality not uniform throughout. These will change in the
direction the fluid flows, but not perpendicular to the direction of flow (i. e., no axial
movement)
Storage - any device to accept and retain a fluid for the purpose of future discharge.
Discontinuity of flow of liquid into and out of a device.
Treatment - the act of improving fluid properties by physical means. The removal of undesirable
impurities from a fluid.
VOC - Volatile Organic Compounds, refering to all organic compounds except the following,
which have been shown not to be photochemically reactive: methane, ethane,
trichlorotrifluoroethane, methylene chloride, 1,1,1,-trichloroethane, trichlorofluoromethane,
dichlorodifluoromethane, chlorodifluoromethane, trifluoromethane, dichlorotetrafluoroethane,
and chloropentafluoroethane.
4.13-28 EMISSION FACTORS 9/91
-------
Table 4.13-4.
SIMS CHEMICAL PROPERTY DATA FILE
9/91 Evaporation Loss Sources 4.1V9
-------
5 *
ii
Hi
UJ M
3 5 «
32!
S S S
a
UJ
li-
i &
li
H
2 . a
at =
^ o u
I a
? 3
*
a 8 5
X*~
S 5 3 S S
s^ fV f** *> «X
s 5 s s a c
^- M « £
— — — o>
**• O- »n
» IN Wt
•
» o
(> O
-: ^ rw > ^
S*- -o -o -*
»^ f* •- fM
N. N. s» irt
U"i -^ k/"i /%
« »rt ix o « -*
• »rt M * O fy
a „
S^ -^ ^ •- 4 g *n
K Z tn o 3 o »
5 g £ S 3 3
i!
5 =
III
u< fe p
si 5 1 £
S w •• o ^ ^ *J
3 X C a «- £ 5 5
Ss§,i,CC8SS
4.13-31
-------
« &
UJ
i *
li
5?
ill
Is
i S
•£ S
4.13-32
s -
SS
A
u ^
jl 5
i i
£ S
rf> ^ fl -O M *t »- • >A > *
• O IM O ^ *
« •* • AJ « PI -* 5
N. .* ^- in ry IN* ry 5 *
s s x » » a 3 a s 3 §
*. S S S 5 •: * s
-
as 2^25 s S3 li cs§ § 3sass|5 | „ „
l-lcsaSss^^^^g^sssfi^ssSS^r^s^
g 2
K0 ^ *••
fw h> xi
3 » 3 2
SO. > « •-
'O •" 2 S
S g
K « c a
**53 —oRJo
8p> K> »•
P- »A P* •- r«
tv <) 00
O > >
00^0-0
** "* fe ^
2«2^^ 55 S
s3aS£ S S 29S
32955'ON.SS '"
^S^^WN^**OrMfMO»-
ssssssslisls
50% .l»
-------
« a
uj
3 5
ii
III
s ; -
3 2 S
UJ ~ •+ rg 03
3 I s 5 5 2 I s « « "• S 5-83 a~""
j* ie <« 1*1 O « * .J "* "" * * _ _ ^. -. r « .^N.j^*>-rgiA>N*hn
|X"^ ft * 2 "£"-"^ 52^'*"^'^ H'*^ ^
^ OO r~~t**~-^iS*tf%-*t*~ ^l*» f^-O
O> ^ -atn O rg 4 O O &> rg ui o rg oo r« t/\ *- i*y
• m rg .. . . , . ^ . ... ..
?j ^ "^-"^rg' n?p?*f.°; T°!T°;ri/! **o*? * * .
O»
•Ow-rg rg rgrg^^->*
*^ ^ -O J^ *^
NT ^ v- eg rg o 9 rg «j rg ^"S "* ^ *°
K i, £
o £ £ fi S3 ^ fi S? ? ® S>£ IAIO^O**
s2S~!3sS!g55§3ss!?Js3issis!S5a2
s s
- !K
• ••oe&»a*MB
SSSS5S6SCSS
'
X « 1 i » X
4.13-33
-------
3
§
IN
3 ; i
825
IU <* 1
SB!
Ul f»
3 < »
s « —
38"
3 *
>.«->.
r 5 -3
i i s
4.13-34
I!
II
K 8 « R 8
«*nw
»- — tn
.* fV
*r N-
2 S
•O
>
*-h/limt*«* 'O -C f\ •OK>—i/«^
xr tt N — C *••» O>-r-«<<> «*-» > — -O « ^ i«
O»*io O'O'C^'*^'*''*^**'^"-!^^' ^ -»O P1**" ^-^^
-O K K XJ5-S — of-owi^OK.© t> r^M t.*- --sro
rgfyrgorvjr^r*J(M«>
tnm^ o3«*
g> « • ** h- •o-
Aj«tfl N.tr««O
KI f? £ vr*--^rvt
— £
flO
O
rg
•
K^
>
o
M ^ !o ^S 4SCS«£^$)C«><85C^a»ZSS£ ^ ^ o Sfe
-o "* 5
N, N. lf\ fl p> i/S ^ 0i« O> U^ -^ O
^" hr» IM «N» "^ ^ !S 2 M ^ SJ O »^- *- r^ O
_ > rfj -O
§0 O ^ >0 ^5|o^^2« o
g^^^^ss^o.sso^s^i-l^S^gdS^s^^oS
^^ *^«M»n^^*^ • ** ~ . . . . ^ . . ^. ^ ,«. • ^ (^ *t/\Aj •
go~goooa28g~S22aSSo~S2~8gg-c,o =
ss5K^s~§§"stS53gs|ss25:ss»:^2
* * **
15!
-i^
iis
W *^
w ^ in &
1 ss s a *
« ^ 5 2 £ ~ 8
2 sg 8 IK
a s s « a
-------
5 *
ii
u. 3 5
a"
s ; i
8 2 i
UJ V) •
|5 '
X S ?
i s s
li
5 « ~
!< i
r i $
i i •
i S ;
v ~- 9
f»» N v
OO
**^*^
S» *M *M ^ •-
«n rsj IM *•» --
-
O ^ O» *rt rw
fM >O f%* ^ -O
tO*~^-s»
*— *o o O
S*~
^
*- «3 -O
K> u> rg A
O ^ rg h* £
O nj^»O
•
• > fy o 3 S *- •n«
^ «o -o •• F^ K» >m *<^*-
O» Q
-O W» -O —
-^^!»
4.13-35
-------
References for Section 4.13
1. Hazardous Waste Treatment. Storage. And Disposal Facilities (TSDF) - Air Emission Models.
EPA-450/3-87-026, U. S. Environmental Protection Agency, Research Triangle Park, NC, April
1989.
2. Waste Water Treatment Compound Property Processor Air Emissions Estimator (WATER 7).
Office of Air Quality Planning And Standards, U. S. Environmental Protection Agency,
Research Triangle Park, NC, available early 1992.
3. Evaluation of Test Method For Measuring Biodegradation Rates Of Volatile Organics. Draft,
EPA Contract No. 68-D90055, Entropy Environmental, Research Triangle Park, NC,
September 1989.
4. Industrial Waste Water Volatile Organic Compound Emissions - Background Information For
BACT/LAER Determinations. EPA-450/3-90-004, U. S. Environmental Protection Agency,
Research Triangle Park, NC, January 1990.
5. Evan K. Nyer, Ground Water Treatment Technology. Van Nostrand Reinhold Company, New
York, 1985.
4.13-36 EMISSION FACTORS 9/91
-------
5.13 PLASTICS
5.13.1 POLYVINYL CHLORIDE AND POLYPROPYLENE
5.13.1.1 Process Description1
The manufacture of most resins or plastics begins with the polymerization or linking of the
basic compound (monomer), usually a gas or liquid, into high molecular weight noncrystalline
solids. The manufacture of the basic monomer is not considered part of the plastics industry and
is usually accomplished at a chemical or petroleum plant
The manufacture of most plastics involves an enclosed reaction or polymerization step, a
drying step, and a final treating and forming step. These plastics are polymerized or otherwise
combined in completely enclosed stainless steel or glass-lined vessels. Treatment of the resin after
polymerization varies with the proposed use. Resins for moldings are dried and crushed or
ground into molding powder. Resins such as the alkyd to be used for protective coatings are
usually transferred to an agitated thinning tank, where they are thinned with some type of solvent
and then stored in large steel tanks equipped with water-cooled condensers to prevent loss of
solvent to the atmosphere. Still other resins are stored in latex form as they come from the
kettle.
5.13.1.2 Emissions And Controls1
The major sources of air contamination in plastics manufacturing are the raw materials or
monomers, solvents, or other volatile liquids emitted during the reaction; sublimed solids such as
phthalic anhydride emitted in alkyd production, and solvents lost during storage and handling of
thinned resins. Emission factors for the manufacture of polyvinyl chloride and polypropylene are
shown in Table 5.13-1.
Table 5.13.1-1. UNCONTROLLED EMISSION FACTORS FOR
PLASTICS MANUFACTURING8
EMISSION FACTOR RATING: E
Type of Plastic
Polyvinyl chloride
Polypropylene
Participate
kg/Mg
17.5b
1.5
Ib/ton
35b
3
Gases
kg/Mg
8.5°
0.35d
Ib/ton
17c
0.7d
"References 2-3.
bUsually controlled with fabric filter, efficiency of 98-99%.
cAs vinyl chloride.
dAs propylene.
9/91 Chemical Process Industry 5.13.1-1
-------
Much of the control equipment used in this industry is a basic part of the system, serving
to recover a reactant or product. These controls include floating roof tanks or vapor recovery
systems on volatile material, storage units, vapor recovery systems (adsorption or condensers),
purge lines venting to a flare system, and vacuum exhaust line recovery systems.
References for Section 5.13
1. Air Pollutant Emission Factors. Final Report. Resources Research, Inc. Reston, VA.
Prepared for National Air Pollution Control Administration, Durham, NC, under Contract
Number CPA-22-69-119. April 1970.
2. Unpublished data. U. S. Department of Health and Human Services, National Air Pollution
Control Administration, Durham, NC, 1969.
3. Communication between Resources Research, Inc., Reston, VA, and State Department of
Health, Baltimore^ MD, November 1969.
5.13.1-2 EMISSIONS FACTORS 9/91
-------
5.13.2 POLY(ETHYLENE TEREPHTHALATE)1'2
5.13.2.1 General
Poly(ethylene terephthalate), or PET, is a thermoplastic polyester resin. Such resins may
be classified as low viscosity or high viscosity resins. Low viscosity PET typically has an intrinsic
viscosity of less than 0.75, while high viscosity PET typically has an intrinsic viscosity of 0.9 or
higher. Low viscosity resins, which are sometimes referred to as "staple" PET (when used in
textile applications), are used in a wide variety of products, such as apparel fiber, bottles, and
photographic film. High viscosity resins, sometimes referred to as "industrial" or "heavy denier"
PET, are used in tire cord, seat belts, and the like.
PET is used extensively in the manufacture of synthetic fibers (i. e., polyester fibers),
which compose the largest segment of the synthetic fiber industry. Since it is a pure and
regulated material meeting FDA food contact requirements, PET is also widely used in food
packaging, such as beverage bottles and frozen food trays that can be heated in a microwave or
conventional oven. PET bottles are used for a variety of foods and beverages, including alcohol,
salad dressing, mouthwash, syrups, peanut butter, and pickled food. Containers made of PET are
being used for toiletries, cosmetics, and household and pharmaceutical products (e. g., toothpaste
pumps). Other applications of PET include molding resins, X-ray and other photographic films,
magnetic tape, electrical insulation, printing sheets, and food packaging film.
5.13.2.2 Process Description3'15
PET resins are produced commercially from ethylene glycol (EG) and either dimethyl
terephthalate (DMT) or terephthalic acid (TPA). DMT and TPA are solids. DMT has a melting
point of 140°C (284°F), while TPA sublimes (goes directly from the solid phase to the gaseous
phase). Both processes first produce the intermediate bis-(2-hydroxyethyl)-terephthalate (BHET)
monomer and either methanoi (DMT process) or water (TPA process). The BHET monomer is
then polymerized under reduced pressure with heat and catalyst to produce PET resins. The
primary reaction for the DMT process is:
CH3OOC <=> COOCH3 + HOCH2CH2OH-^HO - (OC O- COOCH2CH2O)nH + 2nCH3OH
DMT EG PET
The primary reaction for the TPA process is:
HOOC -O COOH + HOCH2CH2OH-^HO - (OC COOCH2CH2O)nH + 2nH2O
TPA EG PET
Both processes can produce low and high viscosity PET. Intrinsic viscosity is determined
by the high polymerizer operating conditions of (1) vacuum level, (2) temperature, (3) residence
time, and (4) agitation (mechanical design).
9/91 Chemical Process Industry 5.13.2-1
-------
The DMT process is the older of the two processes. Polymerization grade TPA has been
available only since 1963. The production of methanol in the DMT process creates the need for
methanol recovery and purification operations. In addition, this methanol can produce major
VOC emissions. To avoid the need to recover and purify the methanol and to eliminate the
potential VOC emissions, newer plants tend to use the TPA process.
DMT Process - Both batch and continuous operations are used to produce PET using
DMT. There are three basic differences between batch process and continuous process, (1) a
column-type reactor replaces the kettle reactor for esterification (ester exchange between DMT
and ethylene glycol), (2) "no-back-mix" (i. e., no stirred tank) reactor designs are required in the
continuous operation, and (3) different additives and catalysts are required to ensure proper
product characteristics (e. g., molecular weight, molecular weight distribution).
Figure 5.13.2-1 is a schematic representation of the PET/DMT continuous process, and
the numbers and letters following refer to this figure. Ethylene glycol is drawn from raw material
storage (1) and fed to a mix tank (2), where catalysts and additives are mixed in. From the mix
tank, the mixture is fed, along with DMT, to the esterifiers, also known as ester exchange reactors
(3). About 0.6 pounds (Ibs) of ethylene glycol and 1.0 Ibs of DMT are used for each pound of
PET product. In the esterifiers, the first reaction step occurs at an elevated temperature
(between 170 and 230°C [338 and 446°F]) and at or above atmospheric pressure. This reaction
produces the intermediate BHET monomer and the byproduct methanol. The methanol vapor
must be removed from the esterifiers to shift the conversion to produce more BHET.
The vent from the esterifiers is fed to the methanol recovery system (11), which separates
the methanol by distillation in a methanol column. The recovered methanol is then sent to
storage (12). Vapor from the top of the methanol column is sent to a cold water (or refrigerated)
condenser, where the condensate returns to the methanol column, and noncondensables are
purged with nitrogen before being emitted to the atmosphere. The bottom product of methanol
column, mostly ethylene glycol from the column's reboiler, is reused.
The BHET monomer, with other esterifier products, is fed to a prepolymerization reactor
(4) where the temperature is increased to 230 to 285°C (446 to 545°F), and the pressure is
reduced to between 1 and 760 millimeters (mm) of mercury (Hg) (typically, 100 to 200 mm Hg).
At these operating conditions, residual methanol and ethylene glycol are vaporized, and the
reaction that produces PET resin starts.
Product from the prepolymerizer is fed to one or more polymerization reactors (5), in
series. In the polymerization reactors, sometimes referred to as end finishers, the temperature is
further increased to 260 to 300°C (500 to 572°F). The pressure is further reduced (e. g., to an
absolute pressure of 4 to 5 mm Hg). The final temperature and pressure depend on whether low
or high viscosity PET is being produced. For high viscosity PET, the pressure in the final (or
second) end finisher is less than 2 mm Hg. With high viscosity PET, more process vessels are
used than low viscosity PET, to achieve the higher temperatures and lower pressures needed.
The vapor (ethylene glycol, methanol, and other trace hydrocarbons from the
prepolymerization and polymerization reactors) typically is evacuated through scrubbers (spray
condensers) using spent ethylene glycol. The recovered ethylene glycol is recirculated in the
5.13.2-2 EMISSION FACTORS 9/91
-------
§
8
§
PL,
UJ
O
ea
I
'•o
1
•8
IS
"5.
CO
9/91
Chemical Process Industry
5.13.2-3
-------
scrubber system, and part of the spent ethylene glycol from the scrubber system is sent to storage
in process tanks (13), after which it is sent to the ethylene glycol recovery system (14).
The ethylene glycol recovery system (14) usually is a distillation system composed of a low
boiler column, a refining column, and associated equipment. In such a system, the ethylene glycol
condensate is fed to the low boiler column. The top product from this column is sent to a
condenser, where methanol is condensed and sent to methanol storage. The noncondensable vent
(from the low boiler condenser) is purged with nitrogen and sent to the atmosphere (Stream G in
the flow diagram). The bottom product of the low boiler column goes to its reboiler, with the
vapor recycled back to the low boiler column and the underflow sent to the refining column. The
refining column is under vacuum and is evacuated to the atmosphere. Top product from the
refining column goes through a condenser, and the condensate is collected in a reflux tank. Part
of the ethylene glycol condensate returns to the refining column. The remaining liquid goes to
refined ethylene glycol storage (15). The reflux tank is purged with nitrogen. (The purge gas
vented to the atmosphere from the reflux tank consists of only nitrogen.) The bottom product of
the refining column goes to a reboiler, vapor returns to the column, and what remains is a sludge
byproduct (16).
The vacuum conditions in the prepolymerization and polymerization reactors are created
by means of multi-stage steam jet ejector (venturi) systems. The vacuum system typically is
composed of a series of steam jets, with condensers on the discharge side of the steam jet to cool
the jets and to condense the steam. The condensed steam from the vacuum jets and the
evacuated vapors are combined with the cooling water during the condensation process. This
stream exiting the vacuum system goes either to a cooling tower (17), where the water is cooled
and then recirculated through the vacuum system, or to a waste water treatment plant (once-
through system) (18).
Product from the polymerization reactor (referred to as the polymer melt) may be sent
directly to fiber spinning and drawing operations (6). Alternatively, the polymer melt may be
chipped or pelletized (7), put into product analysis bins (8), and then sent to product storage (9)
before being loaded into hoppers (10) for shipment to the customer.
TPA Process - Figure 5.13.2-2 is a schematic diagram of a continuous PET/TPA process,
and the numbers and letters following refer to this figure. Raw materials are brought on site and
stored (1). Terephthalic acid, in powder form, may be stored in silos. The ethylene glycol is
stored in tanks. The terephthalic acid and ethylene glycol, containing catalysts, are mixed in a
tank (2) to form a paste. In the mix tank, ethylene glycol flows into a manifold that sprays the
glycol through many small slots around the periphery of the vent line. The terephthalic acid and
ethylene glycol are mixed by kneading elements working in opposite directions. Combining these
materials into a paste is a simple means of introducing them to the process, allowing more
accurate control of the feed rates to the esterification vessels. A portion of the paste is recycled
to the mix tank. This paste recycle and feed rates of TPA and ethylene glycol are used to
maintain an optimum paste density or weight percent of terephthalic acid.
The paste from the mix tanks is fed, using gear pumps to meter the flow, to a series of
esterification vessels (referred to esterifiers, or ester exchange reactors). Two or more esterifiers
may be used. Residence time is controlled by valves in the transfer lines between each vessel.
These esterifiers a*re closed, pressurized reactors. Pressure and temperature operating conditions
in the primary esterifier (3) are between 30 and 50 pounds per square inch gauge (psig) and 230
to 260°C (446 to 500°F), respectively. Vapors, primarily water (steam) and glycol, are vented to a
5.13.2-4 EMISSION FACTORS 9/91
-------
o
2
3
O
3
s
I
Pk
-------
reflux column or distillation column. A heat exchanger cools the vapors. Recovered glycol is
returned to the primary esterifier. The water vapor is condensed using 29°C (85°F) cooling water
in a shell-and-tube condenser and then is discharged to the waste water treatment system. The
monomer formed in the primary esterifier and the remaining reactants are pumped to the
secondary esterifier.
The secondary esterifier (4) is operated at atmospheric pressure and at a temperature of
250 to 270°C (482 to 518°F). The vapors from the secondary esterifier, primarily water vapor, are
vented to a spray condenser, and this condensate is sent to a central ethylene glycol recovery unit
(12). The condensate water is cooled by cooling water in a shell-and-tube heat exchanger and
then recycled.
At one plant, the secondary esterifiers for the staple PET lines have a manhole (or rotary
valve on some lines) through which chips and reworked yarn pellets were recycled. These
manholes are not present on the secondary esterifiers for the industrial PET lines. Water vapor
and monomer are emitted from the manholes, and the monomer sublimates on piping near the
manhole.
Monomer (BHET) from the secondary esterifier is then pumped to the polymerization
reactors. The number of reactors and their operating conditions depend on the type of PET
being produced. Typically, there will be at least two polymerization reaction vessels in series, an
initial (low) polymerizer and a final (high) polymerizer. The former is sometimes referred to as a
prepolymerizer or a prepolycondensation reactor. The latter is sometimes called an end finisher.
In producing high viscosity PET, a second end finisher is sometimes used.
In the initial (low) polymerizer (5), esterification is completed and polymerization occurs
(i. e., the joining of short molecular chains). Polymerization is "encouraged" by the removal of
ethylene glycol. This reactor is operated under pressures of 20 to 40 mm Hg and at 270 to 290°C
(518 to 554°F) for staple (low viscosity) PET, and 10 to 20 mm Hg and 280 to 300°C (536 to
572°F) for industrial filament PET. The latter conditions produce a longer molecule, with the
greater intrinsic viscosity and tenacity required in industrial fiber. Glycol released in the
polymerization process and any excess or unreacted glycol are drawn into a contact spray
condenser (scrubber) countercurrent to a spent ethylene glycol spray. (At one facility, both the
low and high polymerizer spray condensers have four spray nozzles, with rods to clear blockage by
solidified polymer. Care is taken to ensure that the spray pattern and flow are maintained.)
Recovered glycol is pumped to a central glycol recovery unit, a distillation column. Vacuum on
the reactors is maintained by a series of steam jets with barometric intercondensers. At one plant,
a two-stage steam ejector system with a barometric intercondenser is used to evacuate the low
polymerizer. The condensate from the intercondensers and the last steam jets is discharged to an
open recirculating water system, which includes an open trough (referred to as a "hot well") and
cooling tower. The recirculation system supplies cooling water to the intercondensers.
In the production of high viscosity PET, the polymer from the low polymerizer is pumped
to a high polymerizer vessel (6). In the high polymerizer, the short polymer chains formed in the
low polymerizer are lengthened. Rotating wheels within these vessels are used to create large
surface exposure for the polymer to facilitate removal of ethylene glycol produced by the
interchange reaction between the glycol ester ends. The high polymerizer is operated at a low
absolute pressure (high vacuum), 0.1 to 1.0 mm Hg, and at about 280 to 300°C. Vapors evolved
in the high polymerizer, including glycol, are drawn through a glycol spray condenser. If very
"hard" vacuums are drawn (e. g., 0.25 mm Hg), such spray condensers are very difficult, if not
5.13.2-6 EMISSION FACTORS 9/91
-------
impossible, to use. At least one facility does not use any spray condensers off the polymerizers
(low and high). Recovered glycol is collected in a receiver and is pumped to a central ethylene
glycol recovery unit. At one plant, chilled water between -3.9 and 1.7°C (25 and 35°F) is used on
the heat exchanger associated with the high polymerizer spray condenser.
At least one facility uses two high polymerizers (end finishers) to produce high viscosity
PET. At this plant, the first end finisher is usually operated with an intermediate vacuum level of
about 2 mm Hg. The polymer leaving this reactor then enters a second end finisher, which may
have a vacuum level as low as 0.25 mm Hg.
Vapors from the spray condenser off the high polymerizers are also drawn through a
steam jet ejector system. One facility uses a five-jet system. After the first three ejectors, there is
a barometric intercondenser. Another barometric intercondenser is located between the fourth
and fifth ejectors. The ejectors discharge to the cooling water hot well. The stream exiting the
vacuum system is sent either to a cooling tower (16) where the water is recirculated through the
vacuum system, or to a waste water treatment plant (once-through system) (15).
Vacuum pumps were installed at one plant as an alternative to the last two ejectors.
These pumps were installed as part of an energy conservation program and are used at the
operator's discretion. The vacuum pumps are operated about 50 percent of the time. The
vacuum system was designed for a maximum vapor load of about 10 kilograms per hour (kg/hr).
If vacuum is lost, or is insufficient in the low or high polymerizers, off-specification product
results. Each process line has a dual vacuum system. One five-stage ejector/vacuum pump system
is maintained as a standby for each industrial filament (high viscosity) process line. The staple
(low viscosity) lines have a standby ejector system, but with only one vacuum pump per process
line. Steam ejectors reportedly recover faster from a slug of liquid carryover than do vacuum
pumps, but the spare system is used in the production of either high or low viscosity PET.
At many facilities, molten PET from the high polymerizer is pumped at high pressure
directly through an extruder spinerette, forming polyester filaments (7). The filaments are air
cooled and then either cut into staple or wound onto spools. Molten PET can also be pumped
out to form blocks as it cools and solidifies (8), which are then cut into chips or are pelletized (9).
The chips or pellets are stored (10) before being shipped to the customer, where they are
remelted for end-product fabrication.
Ethylene glycol recovery (12) generally involves a system similar to that of the DMT
process. The major difference is the lack of a methanol recovery step. At least one TPA facility
has a very different process for ethylene glycol recovery. At this plant, ethylene glycol emissions
from the low and high polymerizers are allowed to pass directly to the vacuum system and into
the cooling tower. The ethylene glycol is then recovered from the water in the cooling tower.
This arrangement allows for a higher ethylene glycol concentration in the cooling tower.
5.13.2.3 Emissions And Controls3'5'11'13-16'21
Table 5.13.2-1 shows the VOC and particulate emissions for the PET/DMT continuous
process, with similar levels expected for batch processes. The extensive use of spray condensers
and other ethylene glycol and methanol recovery systems is economically essential to PET
production, and these are not generally considered "controls".
9/91 Chemical Process Industry 5.13.2-7
-------
TABLE 5.13.2-1. EMISSION FACTORS FOR PET/DMT PROCESS8
Stream
Identification
A
B
C
D
E
El
E2
E3
F
G
H
I
J
Total Plant
Emission Stream
Raw material storage
Mix tanks
Methanol recovery system
Recovered methanol storage
Polymerization reaction
Prepolymerizer vacuum system
Polymerization reactor vacuum system
Cooling tower8
Ethylene glycol process tanks
Ethylene glycol recovery condenser
Ethylene glycol recovery vacuum system
Product storage
Sludge storage and loading
Nonmethane Paniculate
vocb
0.1 0.165°
negligible"1
0.3e
0.09f
0.009
0.005
0.2
3.4
0.0009
0.01
0.0005
0.0003h
0.02
0.73! 0.17
3.9)
Emission
Factor
Rating
C
C
C
C
C
C
C
C
C
C
C
C
References
17
13
3,17
3,17
17
17
18-19
17
17
17
17
17
"Stream identification reters to Figure 5.13.2-1. Units are grams per kilogram of product.
Dash = no data.
bRates reflect extensive use of condensers and other recovery equipment as part of normal
industry economical practice.
Trom storage of DMT.
dAssumed same as for TPA process.
Reference 3. For batch PET production process, estimated to be 0.15 grams VOC per kilogram
of product.
fReflects control by refrigerated condensers.
SBased on ethylene glycol concentrations at two PET/TPA plants. The lower estimate reflects
emissions where spray condensers are used off the prepolymerizers and the polymerization
reactors. The higher estimate reflects emissions where spray condensers are not used off the
prepolymerizers and the polymerization reactors. A site-specific calculation is highly
recommended for all cooling towers, because of the many variables. The following equation may
be used to estimate windage emissions from cooling towers:
E =
where
xCT,
E
WR
4.2
3.78
cr x 60 x WR] x [(4.2 x EG^) + (3.78 x H2
= Mass of VOC emitted (kilograms per hour)
= Concentration of ethylene glycol, weight percent (fraction)
= Cooling tower circulation rate, gallons per minute
= Windage rate, fraction
= Density of ethylene glycol (kilograms per gallon)
= Density of water (kilograms per gallon)
5.13.2-8
EMISSION FACTORS
9/91
-------
= Concentration of water, weight percent (fraction)
60 = Minutes per hour
Example: The VOC emissions from a cooling tower with a ethylene glycol concentration of
8.95% by weight, a water concentration of 91.05% by weight, a cooling tower
circulation rate of 1270 gallons per minute, and a windage rate of 0.03% are
estimated to be:
E = [0.0895 x 1270 x 60 x 0.0003] x [(4.2 x 0.0895) + (3.78 x 0.9105)]
= 7.8 kilograms per hour
hEmission rate is for "controlled" emissions. Without controls, the estimated emission rate is 0.4
grams per kilogram of product.
'With spray condensers off all prepolymerizers and the polymerization reactors.
JWith no spray condensers off all prepolymerizers and the polymerization reactors.
Total VOC emissions will depend greatly on the type of system used to recover the
ethylene glycol from the prepolymerizers and polymerization reactors, which give rise to emission
streams El, E2, E3, F, G, H, and J. The emission streams from the prepolymerizers and
polymerization reactors are primarily ethylene glycol, with small amounts of methanol vapors and
volatile impurities in the raw materials. Of these emission streams, the greatest emission potential
is from the cooling tower (Stream E3). The amount of emissions from the cooling tower depends
on a number of factors, including ethylene glycol concentration and windage rate. The ethylene
glycol concentration depends on a number of factors, including use of spray condensers off of the
polymerization vessels, circulation rate of the cooling water in the cooling tower, blowdown rate
(the rate are which water is drawn out of the cooling tower), and sources of water to cooling
tower (e. g., dedicated cooling tower versus plant-side cooling tower).
Most plants recover the ethylene glycol by using a spent ethylene glycol spray scrubber
condenser directly off these process vessels and before the stream passes through the vacuum
system. The condensed ethylene glycol may then be recovered through distillation. This type of
recovery system results in relatively low concentrations of ethylene glycol in the cooling water at
the tower, which in turn lowers emission rates for the cooling tower and the process as a whole.
At one PET/TPA plant, a typical average concentration of about 0.32 weight percent ethylene
glycol was reported, from which an emission rate of 0.2 grams VOC per kilogram (gVOC/kg) of
product was calculated.
Alternatively, a plant may send the emission stream directly through the vacuum system
(typically steam ejectors) without using spent ethylene glycol spray condensers. The steam
ejectors used to produce a vacuum will produce in contaminated water, which is then cooled for
reuse. In this system, ethylene glycol is recovered from the water in the cooling tower by drawing
off water from the tower (blowdown) and sending the blowdown to distillation columns. This
method of recovering ethylene glycol can result in much higher concentrations of ethylene glycol
in the cooling tower than when the ethylene glycol is recovered with spray condensers directly off
of the process vessels. (The actual concentration of ethylene glycol in the cooling water depends,
in part, on the blowdown rate.) Higher concentrations in the cooling tower result in greater
ethylene glycol emissions from the cooling tower and, in turn, from the process as a whole. At
9/91 Chemical Process Industry 5.13.2-9
-------
one PET/TPA plant recovering the ethylene glycol from the cooling tower, emissions from the
cooling tower were approximately 3.4 gVOC/kg of product.
Next to the cooling tower, the next largest potential emission source in the PET/DMT
process is the methanol recovery system. Methanol recovery system emissions (Stream C) from a
plant using a continuous process are estimated to be approximately 0.3 gVOC/kg of product and
about 0.09 gVOC/kg of product from the recovered methanol storage tanks. The emissions from
the methanol recovery system (Stream C) for a batch process were reported to be 0.15 gVOC/kg
of product, and typically are methanol and nitrogen.
The other emission streams related to the prepolymerizer and polymerization reactors are
collectively relatively small, being about 0.04 gVOC/kg of product. VOC emissions from raw
material storage (mostly ethylene glycol) are estimated to be about 0.1 gVOC/kg of product.
Fixed roof storage tanks (ethylene glycol) and bins (DMT) are used throughout the industry.
Emissions are vapors of ethylene glycol and DMT result from vapor displacement and tank
breathing. Emissions from the mix tank are believed to be negligible.
Particulate emissions occur from storage of both raw material (DMT) and end product.
Those from product storage may be controlled before release to the atmosphere. Uncontrolled
particulate emissions from raw material storage are estimated to be approximately 0.17 g/kg of
product. Particulate emissions from product storage are estimated to be approximately 0.0003
g/kg of product after control and approximately 0.4 g/kg of product before control.
In summary, total VOC emissions from a PET/DMT continuous process are approximately
0.74 gVOC/kg of product, if spray condensers are used off all of the prepolymerizers and
polymerization reaction vessels. For a batch process, this total decreases to approximately 0.59
gVOC/kg of product. If spray condensers are not used, the ethylene glycol concentration in the
cooling tower is expected to be higher, and total VOC emissions will be greater. Calculation of
cooling tower emissions for site-specific plants is recommended. Total particulate emissions are
approximately 0.17 g/kg of product, if product storage emissions are controlled.
Table 5.13.2-2 summarizes VOC and particulate emissions for the PET/TPA continuous
process, and similar emission levels are expected for PET/TPA batch processes. VOC emissions
are generally "uncontrolled", in that the extensive use of spray condensers and other ethylene
glycol recovery systems are essential to the economy of PET production.
Emissions from raw material storage include losses from the raw materials storage and
transfer (e. g., ethylene glycol). Fixed roof storage tanks and bins with conservation vents are
used throughout the process. The emissions, vapors of ethylene glycol, TPA, and TPA dust, are
from working and breathing losses. The VOC emission estimate for raw materials storage is
assumed to be the same as that for the PET/DMT process. No emission estimate was available
for the storage and transfer of TPA
VOC emissions from the mix tank are believed to be negligible. They are emitted at
ambient temperatures through a vent line from the mixer.
VOC emissions from the esterifiers occur from the condensers/distillation columns on the
esterifiers. Emissions, which consist primarily of steam and ethylene glycol vapors, with small
amounts of feed impurities and volatile side reaction products, are estimated to be 0.04 gVOC/kg
of product. Exit temperature is reported to be approximately 104°C (220°F). At least one plant
5.13.2-10 EMISSION FACTORS 9/91
-------
Table 5.13.2-2. EMISSION FACTORS FOR PET/TPA PROCESS3
Stream
Identification
A
B
C
D
Dl
D2
D3
E
F
G
Total Plant
Emission Stream
Raw material storage
Mix tanks
Esterification
Polymerization reaction
Prepolymerizer vacuum
system
Polymerization reactor
vacuum system
Cooling tower6
Ethylene glycol process
tanks
Ethylene glycol recovery
vacuum system
Product storage
Nonmethane Paniculate
vocb
O.lc
negligible
0.04d
0.009C
0.005C
0.2
3.4
0.0009°
0.0005C
0.0003c'f
0.36«
3.6h
Emission
Factor
Rating
C
C
A
C
C
C
C
C
C
References
17
13
20-21
17
17
18-19
17
17
17
Stream identification refers to Figure 5.13.2-2. Units are grams per kilogram of product.
Dash = no data.
bRates reflect extensive use of condensers and other recovery equipment as part of normal
industry economical practice.
cAssumed same as for DMT process.
dAt least one plant controls the primary esterifier condenser vent with a second condenser.
Emissions were 0.0008 grams VOC per kilograms of product with the second condenser
operating, and 0.037 grams VOC per kilogram of product without the second condenser
operating.
eBased on ethylene glycol concentrations at two PET/TPA plants. The lower estimate reflects
emissions where spray condensers are used off the prepolymerizers and the polymerization
reactors. The higher estimate reflects emissions where spray condensers are not used off the
prepolymerizers and the polymerization reactors. It is highly recommended that a site-specific
calculation be done for all cooling towers as many variables affect actual emissions. The
equation found in footnote g for Table 5.13.2-1 may be used to estimate windage emissions from
cooling towers.
fReflects control of product storage emissionss. Without controls, the estimated emission rate is
0.4 grams per kilogram of product.
8With spray condensers off all prepolymerizers and the polymerization reactors.
hWith no use of spray condensers off all prepolymerizers and the polymerization reactors.
9/91
Chemical Process Industry
5.13.2-11
-------
controls the primary esterifier condenser vent with a second condenser. At this plant, emissions
were 0.0008 gVOC/kg of product with the second condenser operating, and 0.037 gVOC/kg of
product without the second condenser operating. The temperature for the emission stream from
the second condenser was reported to be 27 to 38°C (80 to 100°F). The emissions from the
second condenser were composed di-iso-propyl amine (DIPA) and acetaldehyde, with small
amounts of ethylene.
Emissions from the prepolymerizers and polymerization reaction vessels in both PET/TPA
and PET/DMT processes should be very similar. The emissions were discussed earlier under the
DMT process.
The estimates of VOC emissions from the ethylene glycol process tanks and the ethylene
glycol recovery system, and of particulate emissions from product storage, are assumed to be the
same as for the DMT process.
In summary, total VOC emissions from the PET/TPA process are approximately 0.36
gVOC/kg of product, if spray condensers are used with all of the prepolymerizers and
polymerization reaction vessels. If spray condensers are not used with all of these process vessels,
the concentration in the cooling tower can be expected to be higher, and total VOC emissions
will be greater. For example, at one plant, emissions from the cooling tower were calculated to
be approximately 3.4 gVOC/kg of product, resulting in a plant-wide estimate of 3.6 gVOC/kg of
product. Calculation of cooling tower emissions for site-specific plants is recommended.
Excluding TPA particulate emissions (no estimate available), total particulate emissions are
expected to be small.
References for Section 5.13.2
1. Modern Plastics Encyclopedia, 1988. McGraw Hill, New York, 1988.
2. Standards Of Performance For New Stationary Sources: Polypropylene. Polyethylene,
Polystyrene. And Polvfethvlene terephthalatel. 55 FR 51039, December 11, 1990.
3. Polymer Industry Ranking By VOC Emissions Reduction That Would Occur From New
Source Performance Standards. Pullman-Kellogg, Houston, TX, August 30, 1979.
4. Karel Verschueren, Handbook of Environmental Data on Organic Compounds. Van
Nostrand Reinhold Co., New York, NY, 1983.
5. Final Trip Report To Tennessee Eastman Company's Polyester Plant. Kingsport. TN. Energy
and Environmental Analysis, Inc., Durham, NC, October 2, 1980.
6. Written communication from R. E. Lee, Tennessee Eastman Co., Kingsport, TN, to A.
Limpiti, Energy and Environmental Analysis, Inc., Durham, NC, November 7, 1980.
7. Written communication from P. Meitner, E. I. duPont de Nemours and Company, Inc.,
Wilmington, DE, to Central Docket Section, U. S. Environmental Protection Agency,
Washington, DC, February 8, 1988.
5.13.2-12 EMISSION FACTORS 9/91
-------
8. Written communication from P. Meitner, E. I. duPont de Nemours and Company, Inc.,
Wilmington, DE, to J. R. Farmer, U. S. Environmental Protection Agency, Research Triangle
Park, NC, August 29, 1988.
9. Final Trip To DuPont's Polvfethvlene terephthalate) Plant. Kinston. NC. Pacific
Environmental Services, Inc., Durham, NC, February 21, 1989.
10. Telephone communication between R. Purcell, Pacific Environmental Services, Inc., Durham,
NC, and J. Henderson and L. Williams, E. I. duPont de Nemours and Company, Inc., Kinston,
NC, December 1988.
11. Final Trip Report To Fiber Industries Polyester Plant. Salisbury, NC. Pacific Environmental
Services, Inc., Durham, NC, September 29, 1982.
12. Written communication from D. V. Perry, Fiber Industries, Salisbury, NC, to K. Meardon,
Pacific Environmental Services, Inc., Durham, NC, November 22, 1982.
13. Written communication from R. K. Smith, Allied Chemical, Moncure, NC, to D. R. Goodwin,
U. S. Environmental Protection Agency, Research Triangle Park, NC, October 27, 1980.
14. Final Trip Report To Monsanto's Polyester Plant. Decatur. Alabama. Energy and
Environmental Analysis, Durham, NC, August 27, 1980.
15. Written communication from R. K Smith, Allied Fibers and Plastics, Moncure, NC, to J. R.
Farmer, U. S. Environmental Protection Agency, Research Triangle Park, NC, April 15, 1982.
16. Written communication from D. Perry, Fiber Industries, Salisbury, NC, to K. Meardon, Pacific
Environmental Services, Inc., Durham, NC, February 11, 1983.
17. Written communication from D. O. Quisenberry, Tennessee Eastman Company, Kingsport,
TN, to S. Roy, U. S. Environmental Protection Agency, Research Triangle Park, NC, August
25, 1988.
18. K. Meardon, "Revised Costs For PET Regulatory Alternatives," Docket No. A-82-19, Item II-
B-90. U. S. EPA Air Docket Section, Waterside Mall, 401 M Street, SW, Washington, DC,
August 20, 1984.
19. Written communication from J. W. Torrance, Allied Fibers and Plastics, Petersburg, VA to J.
R. Farmer, U. S. Environmental Protection Agency, Research Triangle Park, NC, September
4, 1984.
20. Written communication from A T. Roy, Allied-Signal, Petersburg, VA to K. Meardon, Pacific
Environmental Services, Inc., Durham, NC, August 18, 1989.
21. Telephone communication between K. Meardon, Pacific Environmental Services, Inc.,
Durham, NC, and A Roy, Allied, Petersburg, VA August 18, 1989.
9/91 Chemical Process Industry 5.13.2-13
-------
5.13.3 POLYSTYRENE1'2
5.13.3.1 General
Styrene readily polymerizes to polystyrene by a relatively conventional tree radical chain
mechanism. Either heat or initiators will begin the polymerization. Initiators thermally
decompose, thereby forming active free radicals that are effective in starting the polymerization
process. Typically initiators used in the suspension process include benzoyl peroxide and di-tert-
butyl per-benzoate. Potassium persulfate is a typical initiator used in emulsion polymerizations.
In the presence of inert materials, styrene monomer will react with itself to form a homopolymer.
Styrene monomer will react with a variety of other monomers to form a number of copolymers.
Polystyrene is an odorless, tasteless, rigid thermoplastic. Pure polystyrene has the
following structure.
The homopolymers of styrene are also referred to as general purpose, or crystal,
polystyrene. Because of the brittleness of crystal polystyrene, styrene is frequently polymerized in
the presence of dissolved polybutadiene rubber to improve the strength of the polymer. Such
modified polystyrene is called high impact, or rubber-modified, polystyrene. The styrene content
of high impact polystyrene varies from about 88 to 97 percent. Where a blowing (or expanding)
agent is added to the polystyrene, the product is referred to as an expandable polystyrene. The
blowing agent may be added during the polymerization process (as in the production of
expandable beads), or afterwards as part of the fabrication process (as in foamed polystyrene
applications).
Polystyrene is the fourth largest thermoplastic by production volume. It is used in
applications in the following major markets (listed in order of consumption): packaging,
consumer/institutional goods, electrical/electronic goods, building/construction, furniture,
industrial/machinery, and transportation.
Packaging applications using crystal polystyrene biaxial film include meat and vegetable
trays, blister packs, and other packaging where transparency is required. Extruded polystyrene
foam sheet is formed into egg carton containers, meat and poultry trays, and fast food containers
requiring hot or cold insulation. Solid polystyrene sheet is formed into drinking cups and lids, and
disposable packaging of edibles. Injection molded grades of polystyrene are used extensively in
the manufacture of cosmetic and personal care containers, jewelry and photo equipment boxes,
and photo film packages. Other formed polystyrene items include refrigerator door liners, audio
and video cassettes, toys, flower pots, picture frames, kitchen utensils, television and radio
9/91 Chemical Process Industry 5.133-1
-------
cabinets, home smoke detectors, computer housings, and profile moldings in the
construction/home-building industry.
5.13.3.2 General Purpose And High Impact Polystyrene1"2
Homopolymers and copolymers can be produced by bulk (or mass), solution (a modified
bulk), suspension, or emulsion polymerization techniques. In solution (or modified bulk)
polymerization, the reaction takes place as the monomer is dissolved in a small amount of solvent,
such as ethylbenzene. Suspension polymerization takes place with the monomer suspended in a
water phase. The bulk and solution polymerization processes are homogenous (taking place in
one phase), whereas the suspension and emulsion polymerization processes are heterogeneous
(taking place in more than one phase). The bulk (mass) process is the most widely used process
for polystyrene today. The suspension process is also common, especially in the production of
expandable beads. Use of the emulsion process for producing styrene homopolymer has
decreased significantly since the mid-1940s.
5.13.3.1.1 Process Descriptions1"3
Batch Process - Various grades of polystyrene can be produced by a variety of batch
processes. Batch processes generally have a high conversion efficiency, leaving only small
amounts of unreacted styrene to be emitted should the reactor be purged or opened between
batches. A typical plant will have multiple process trains, each usually capable of producing a
variety of grades of polystyrene.
Figure 5.13.3-1 is a schematic representation of the polystyrene batch bulk polymerization
process, and the following numbered steps refer to that figure. Pure styrene monomer (and
comonomer, if a copolymer product is desired) is pumped from storage (1) to the feed dissolver
(2). For the production of impact grade polystyrene, chopped polybutadiene rubber is added to
the feed dissolver, where it is dissolved in the hot styrene. The mixture is agitated for 4 to 8
hours to complete rubber dissolution. From the feed dissolver, the mixture usually is fed to an
agitated tank (3), often a prepolymerization reactor, for mixing the reactants. Small amounts of
mineral oil (as a lubricant and plasticizer), the dimer of alpha-methylstyrene (as a polymerization
regulator), and an antioxidant are added. The blended or partially polymerized feed is then
pumped into a batch reactor (4). During the reactor filling process, some styrene vaporizes and is
vented through an overflow vent drum (5). When the reactor is charged, the vent and reactor are
closed. The mixture in the reactor is heated to the reaction temperature to initiate (or continue)
the polymerization. The reaction may also be begun by introducing a free radical initiator into
the feed dissolver (2) along with other reactants. After polymerization is complete, the polymer
melt (molten product), containing some unreacted styrene monomer, ethylbenzene (an impurity
from the styrene feed) and low molecular weight polymers (dimers, trimers, and other oligomers),
is pumped to a vacuum devolatilizer (6). Here, the residual styrene monomer, ethylbenzene, and
the low molecular weight polymers are removed, condensed (7), passed through a devolatilizer
condensate tank (9), and then sent to the byproduct recovery unit. Overhead vapors from the
condenser are usually exhausted through a vacuum system (8). Molten polystyrene from the
bottom of the devolatilizer, which may be heated to 250 to 280°C (482 to 536°F), is extruded (10)
through a stranding die plate (a plate with numerous holes to form strands), and then immersed
in a cold water bath. The cooled strands are pelletized (10) and sent to product storage (11).
5.13.3-2 EMISSION FACTORS 9/91
-------
en
in
1
tt,
9/91
Chemical Process Industry
5.13J-3
-------
Continuous Process - As with the batch process, various continuous steps are used to
make a variety of grades of polystyrene or copolymers of styrene. In continuous processes, the
chemical reaction does not approach completion as efficiently as in batch processes. As a result, a
lower percentage of styrene is converted to polystyrene, and larger amounts of unreacted styrene
may be emitted from continuous process sources. A typical plant may contain more than one
process line, producing either the same or different grades of polymer or copolymer.
A typical bulk (mass) continuous process is represented in Figure 5.13.3-2. Styrene,
polybutadiene (if an impact grade product is desired), mineral oil (lubricant and plasticizer), and
small amounts of recycled polystyrene, antioxidants, and other additives, are charged from storage
(1) into the feed dissolver mixer (2) in proportions that vary according to the grade of resin to be
produced. Blended feed is pumped continuously to the reactor system (3) where it is thermally
polymerized to polystyrene. A process line usually employs more than one reactor in series.
Some polymerization occurs in the initial reactor, often referred to as the prepolymerizer.
Polymerization to successively higher levels occurs in subsequent reactors in the series, either
stirred autoclaves or tower reactors. The polymer melt, which contains unreacted styrene
monomer, ethylbenzene (an impurity from the styrene feed) and low molecular weight polymers,
is pumped to a vacuum devolatilizer (4). Here, most of the monomer, ethylbenzene, and low
molecular weight polymers are removed, condensed (5), and sent to the styrene recovery unit (8
and 9). Noncondensables (overhead vapors) from the condenser typically are exhausted through a
vacuum pump (10). Molten polystyrene from the bottom of the devolatilizer is pumped by an
extruder (6) through a stranding die plate into a cold water bath. The solidified strands are then
palletized (6) and sent to storage (7).
In the styrene recovery unit, the crude styrene monomer recovered from the condenser
(5) is purified in a distillation column (8). The styrene overhead from the tower is condensed (9)
and returned to the feed dissolver mixer. Noncondensables are vented through a vacuum system
(11). Column bottoms containing low molecular weight polymers are used sometimes as a fuel
supplement.
5.13.3.2.2 Emissions And Controls3"9
As seen in Figure 5.13.3-1, six emission streams have been identified for batch processes,
(1) the monomer storage and feed dissolver vent (Stream A); (2) the devolatilizer condensate
tank (Stream B); (3) the reactant vent drum vent (Stream C); (4) the devolatilizer condenser vent
(Stream D): (5) the extruder quench vent (Stream E); and (6) product storage emissions (Stream
F). Table 5.13.3-1 summarizes the emission factors for these streams.
The major vent is the devolatilizer condenser vent (Stream D). This continuous offgas
vent emits 0.25 to 0.75 grams of VOC per kilogram (gVOC/kg) of product, depending on the
molecular weight of the polystyrene product being produced. The higher emission factor is more
likely during the manufacture of lower molecular weight products. The emissions are unreacted
styrene, which is flashed from the product polymer in the vacuum devolatilizer, and it is extremely
diluted in air through leakage. The stream is exhausted through a vacuum system and then
through an oil demister to the atmosphere. The oil demister is used primarily to separate out
organic mist.
5.13.3-4 EMISSION FACTORS 9/91
-------
Table 5.13.3-1. EMISSION FACTORS FOR BATCH PROCESS POLYSTYRENE"
EMISSION FACTOR RATING: C
Stream
Identification
A
B
C
D
E
F
Total Plant
Emission
Stream
Monomer storage and feed
dissolver tanks
Devolatilizer condensate tank
Reactor vent drum vent
Devolatilizer condenser vent
Extruder quench vent
Product storage
Nonmethane
VOC
0.09b
0.002b
0.12 - 1.35C
0.25 - 0.75C
0.15 - 0.3C
negligible
0.6 - 2.5
References
3
3
3-4
3-4
3-4
3
Stream identification refers to Figure 5.13.3-1. Units are grams VOC perKilogram of product.
bBased on fixed roof design.
Reference 4. The higher factors are more likely during the manufacture of lower molecular
weight products. Factor for any given process train will change with product grade.
The second largest vent stream is likely to be the reactor vent drum vent, with an emission
rate ranging from 0.12 to 1.35 gVOC/kg of product, this range also being associated with the
molecular weight of the polystyrene product being produced. The higher emission factor is more
likely during the manufacture of lower molecular weight products. These emissions, which are the
only intermittent emissions from the process, occur only during reactor filling periods and they are
vented to the atmosphere. The rate of 0.12 gVOC/kg of product is based on a facility having two
batch reactors that are operated alternately on 24 hour cycles.
Stream E, the extruder quench vent, is the third largest emission stream, with an emission
rate of 0.15 to 0.3 gVOC/kg of product. This stream, composed of styrene in water vapor, is
formed when the hot, extruded polystyrene strands from the stranding die plate contact the cold
water in the quenching bath. The resulting stream of steam with styrene is usually vented through
a forced draft hood located over the water bath and then passed through a mist separator or
electrostatic precipitator before venting to the atmosphere.
The other emission streams are relatively small continuous emissions. Streams A and B
represent emissions from various types of tanks and dissolver tanks. Emissions from these streams
are estimated, based on fixed roof tanks. Emissions from product storage, Stream F, have been
reported to be negligible.
There are no VOC control devices typically used at polystyrene plants employing batch
processes. The condenser (7) off the vacuum devolatilizer (6) typically is used for process reasons
(recovery of unreacted styrene and other reactants). This condenser reduces VOC emissions, and
its operating characteristics will affect the quantity of emissions associated with batch processes
(Stream D in particular).
9/91 Chemical Process Industry 5.13.3-5
-------
u
I
8.
1
A
•s
i
=1
5.13J-6
EMISSION FACTORS
9/91
-------
Total process uncontrolled emissions are estimated to range from 0.6 to 2.5 gVOC/kg of
product. The higher emission rates are associated with the manufacture of lower molecular
weight polystyrene. The emission factor for any given process line will change with changes in the
grade of the polystyrene being produced.
Emission factors for the continuous polystyrene process are presented in Table 5.13.3-2,
and the following numbered steps refer to that figure. Emissions from the continuous process are
similar to those for the batch process, although the continuous process lacks a reactor vent drum.
The emission streams, all of which are continuous, are (1) various types of storage (Streams A
and G); (2) the feed dissolver vent (Stream B); (3) the devolatilizer condenser vent (Stream C);
(4) the styrene recovery unit condenser vent (Stream D); (5) the extruder quench vent (Stream
E); and (6) product storage emissions (Stream G).
Industry's experience with continuous polystyrene plants indicates a wide range of emission
rates from plant to plant, depending in part on the type of vacuum system used. Two types are
now used in the industry, one relying on steam ejectors and the other on vacuum pumps. Where
steam ejectors are used, the overheads from the devolatilizer condenser vent and the styrene
recovery unit condenser vent are composed mainly of steam. Some companies have recently
replaced these steam ejectors with mechanical vacuum pumps. Emissions from vacuum pumps
usually are lower than from steam ejectors.
It is estimated that the typical total VOC emission rate for plants vising steam ejectors is
about 3.34 gVOC/kg of product. The largest emission stream being the devolatilizer condenser
vent (2.96 gVOC/kg of product). Emissions from the styrene recovery condenser vent and the
extruder quench vent are estimated to be 0.13 and 0.15 gVOC/kg of product, respectively,
although the latter may vary significantly depending on overall plant design. One plant designed
to minimize emissions reported an emission factor of 0.0012 gVOC/kg product for the extruder
quench vent.
For plants using vacuum pumps, it is estimated that the total VOC emission rate is about
0.21 gVOC/kg of product. In these plants, emissions from the devolatilizer condenser vent and
the styrene recovery condenser vent are estimated to be 0.05 gVOC/kg of product. Styrene
monomer and other storage emissions can be the largest emission sources at such plants,
approximately 0.1 gVOC/kg of product. Some plants combine emissions from the dissolvers with
those from the devolatilizer condenser vent. Other plants may combine the dissolver, devolatilizer
condenser vent, and styrene recovery condenser vent emissions. One plant uses an organic
scrubber to reduce these emissions to 0.004 gVOC/kg of product.
Condensers are a critical, integral part of all continuous polystyrene processes. The
amount of unreacted styrene recovered for reuse in the process can vary greatly, as condenser
operating parameters vary from one plant to another. Lowering the coolant operating
temperature will lower VOC emissions, all other things being equal.
Other than the VOC reduction achieved by the process condensers, most plants do not
use VOC control devices. A plant having controls, however, can have significantly reduce the
level of VOC emissions. One company, for example, uses an organic scrubber to reduce VOC air
emissions. Another uses a condenser downstream from the primary process condensers to control
VOCs.
9/91 Chemical Process Industry 5.13.3-7
-------
Table 5.13.3-2. EMISSION FACTORS FOR CONTINUOUS PROCESS POLYSTYRENE8
EMISSION FACTOR RATING: C
Stream
Identification
Al
A2
A3
B
C
D
C+D
E
F
Gl
G2
Total Plant
Emission
Stream
Styrene monomer
storage
Additives
General purpose
High impact
Ethylbenzene storage
Dissolvers
Devolatilizer
condenser ventb
Styrene recovery unit
condenser vent
Extruder quench vent
Pellet storage
Other storage
General purpose
High impact
Nonmethane VOC
Uncontrolled Controlled
0.08
0.002
0.001
0.001
0.008
0.05C 0.04d
2.96e
0.05C
0.13e
0.024 - 0.3f 0.004s
0.01C
0.15e-8'h
negligible
0.008
0.007
0.21C
3.34e
References
3,5
5
5-6
5
3,5
4-5,7
3
4,7
3
5-6,8
4
3
3
3,5
3,5
"Reference 9. Larger plants may route this stream to the styrene recovery section. Smaller
plants may find this too expensive.
Tor plants using vacuum pumps.
dCondenser is used downstream of primary process condensers; includes emissions from dissolvers.
Plant uses vacuum pumps.
Tor plants using steam jets.
fLower value based on facility using refrigerated condensers as well as conventional cooling water
exchangers; vacuum pumps in use. Higher value for facility using vacuum pumps.
gPlant uses an organic scrubber to reduce emissions. Nonsoluble organics are burned as fuel.
hThis factor may vary significantly depending on overall process. Reference 6 indicates an
emission factor of 0.0012 gVOC/kg product at a plant whose process design is "intended to
minimize emissions".
5.13.3-8
EMISSION FACTORS
9/9\
-------
5.13.3.3 Expandable Polystyrene1-2-10-11
The suspension process is a batch polymerization process that may be used to produce
crystal, impact, or expandable polystyrene beads. An expandable polystyrene (EPS) bead typically
consists of high molecular weight crystal grade polystyrene (to produce the proper structure when
the beads are expanded) with 5 to 8 percent being a low boiling aliphatic hydrocarbon blowing
agent dissolved in the polymer bead. The blowing agent typically is pentane or isopentane
although others, such as esters, alcohols, and aldehydes, can be used. When used to produce an
EPS bead, the suspension process can be adapted in one of two ways for the impregnation of the
bead with the blowing agent One method is to add the blowing agent to a reactor after
polymerization, and the other is to add the blowing agent to the monomer before polymerization.
The former method, called the "post-impregnation" suspension process, is more common than the
latter, referred to as the "in-situ" suspension process. Both processes are described below.
EPS beads generally are processed in one of three ways, (1) gravity or air fed into closed
molds, then heated to expand up to 50 times their original volume; (2) pre-expanded by heating
and then molding in a separate processing operation; and (3) extended into sheets. EPS beads
are used to produce a number of foamed polystyrene materials. Extruded foam sheet is formed
into egg cartons, meat and poultry trays, and fast food containers. In the building/construction
industry, EPS board is used extensively as a low temperature insulator.
5.13.3.3.1 Process Description1-10-12
Post-impregnation Suspension Process - This process is essentially a two part process using
two process lines in series. In the first process line, raw styrene monomer is polymerized and a
finished polystyrene bead is produced. The second process line takes the finished bead from the
first line, impregnates the bead with a blowing agent, and produces a finished EPS bead. Figure
5.13.3-3 is a schematic representation of this process.
In the first line, styrene monomer, water, initiator, and suspending agents form the basic
charge to the suspension reactor (1). The styrene-to-water ratio varies with the type of
polystyrene required. A typical ratio is about one-quarter to one-half monomer to water volume.
Initiators are commonly used because the reaction temperature is usually too low for adequate
thermal initiation of polymerization. Suspending agents are usually protective colloids and
insoluble inorganic salts. Protective colloids are added to increase the viscosity of the continuous
water phase, and insoluble inorganic salts such as magnesium carbonate (MgCO3) are added to
prevent coalescence of the drops upon collision.
In the reactor, the styrene is suspended, through use of mechanical agitation and
suspending agents, in the form of droplets throughout the water phase. Droplet size may range
from about 0.1 to 1.0 mm. The reactor is heated to start the polymerization, which takes place
within the droplets. An inert gas, such as nitrogen, is frequently used as a blanketing agent in
order to maintain a positive pressure at all times during the cycle, to prevent air leaks. Once
polymerization starts, temperature control is typically maintained through a water-cooled jacket
around the reactor and is facilitated by the added heat capacity of the water in the reactor. The
size of the product bead depends on both the strength of agitation and the nature of the
monomer and suspending system. Between 20 and 70 percent conversion, agitation becomes
extremely critical. If agitation weakens or stops between these limit*, excessive agglomeration of
the polymer particles may occur, followed by a runaway reaction. Polymerization typically occurs
within several hours, the actual time varying largely with the temperature and with the amount
9/91 Chemical Process Industry 5.13.3-9
-------
H
4)
s
t
U
3
•a
CO
•s
CO
c*5
T—I
•o
5.13J-10
EMISSION FACTORS
9/91
-------
and type of initiator(s) used. Residual styrene concentrations at the end of a run are frequently
as low as 0.1 percent.
Once the reaction has been completed (essentially 100 percent conversion), the
polystyrene-water slurry is normally pumped from the reactor to a hold tank (2), which has an
agitator to maintain dispersion of the polymer particles. Hold tanks have at least three functions,
(1) the polymer-water slurry is cooled to below the heat distortion temperature of the polymer
(generally 50 to 60°C [122 to 140°F]); (2) chemicals are added to promote solubilization of the
suspension agents; and (3) the tank serves as a storage tank until the slurry can be centrifuged.
From the hold tanks, the polymer-water slurry is fed to a centrifuge (3) where the water and
solids are separated. The solids are then washed with water, and the wash water is separated
from the solids and is discarded. The polymer product beads, which may retain between 1 and 5
percent water, are sent to dryers (4). From the dryers, they may be sent to a classifier (5) to
separate the beads according to size, and then to storage bins or tanks (6). Product beads do not
always meet criteria for further processing into expandable beads, and "off-spec" beads may be
processed and sold as crystal (or possibly impact) polystyrene.
In the second line, the product bead (from the storage bins of the first line), water,
blowing agent (7), and any desired additives are added to an impregnation reactor (8). The beads
are impregnated with the blowing agent through utilization of temperature and pressure. Upon
completion of the impregnation process, the bead-water slurry is transferred to a hold tank (9)
where acid may be added and part of the water is drained as wastewater. From the hold tanks,
the slurry is washed and dewatered in centrifuges (10) and then dried in low temperature dryers
(11). In some instances, additives (12) may be applied to the EPS bead to improve process
characteristics. From the dryers, the EPS bead may undergo sizing, if not already done, before
being transferred to storage silos (13) or directly to packaging (14) for shipment to the customer.
In-situ Suspension Process - The in-situ suspension process is shown schematically in
Figure 5.13.3-4. The major difference between this process and the post-impregnation suspension
process is that polymerization and impregnation takes place at the same time in a single reactor.
The reaction mixture from the mix tank (1), composed of styrene monomer, water, polymerization
catalysts, and additives, are charged to a reactor (2) to which a blowing agent is added. The
styrene monomer is polymerized at elevated temperatures and pressure in the presence of the
blowing agent, so that 5 to 7 percent of the blowing agent is entrapped in the polymerized bead.
After polymerization and impregnation have taken place, the EPS bead-water slurry follows
essentially the same steps as in the post-impregnation suspension process. These steps are
repeated in Figure 5.13.3-4.
5.13.3.3.2 Emissions And Controls10'12"16
Emission rates have been determined from information on three plants using the post-
impregnation suspension process. VOC emissions from this type of facility are generally
uncontrolled. Two of these plants gave fairly extensive information, and of these, one reported
an overall uncontrolled VOC emission rate of 9.8 g/kg of product. For the other, an overall
uncontrolled VOC emission rate of 7.7 g/kg is indicated, by back-calculating two emission streams
controlled by condensers.
The information on emission rates for individual streams varied greatly from plant to
plant. For example, one plant reported a VOC emission rate for the suspension reactor of 0.027
g/kg of product, while another reported a rate of 1.9 g/kg of product. This inconsistency in
9/91 Chemical Process Industry 5.13.3-11
-------
&•
TTt
5.133-12
EMISSION FACTORS
9/91
-------
emission rates may be because of differences in process reactors, operating temperatures, and/or
reaction times, but sufficient data to determine this are not available. Therefore, individual
stream emission rates for the post-impregnation process are not given here.
Particulate emissions (emissions of fines from dryers, storage and pneumatic transfer of
the polymer) usually are controlled by either cyclones alone or cyclones followed by baghouses.
Overall, controlled particulate emissions are relatively small, approximately 0.18 g particulate/kg of
product or less. Control efficiencies of 99 percent were indicated and thus, uncontrolled
particulate emissions might be around 18 g particulate/kg of product.
Table 5.13.3-3 summarizes uncontrolled VOC emissions factors for the in-situ process,
based on a study of a single plant. An uncontrolled emission rate of about 5.4 gVOC/kg of
product is estimated for this suspension EPS process. Most emission streams are uncontrolled at
this plant. However, reactor emissions are vented to the boiler as primary fuel, and some of the
dryer emissions are vented to the boiler as supplementary fuel, thereby resulting in some VOC
control.
The blowing agent, which continually diffuses out of the bead both in manufacturing and
during storage, constitutes almost all of VOCs emitted from both processes. A small amount of
styrene is emitted from the suspension reactors in the post-impregnation process and from the mix
tanks and reactors in the in-situ process.
Because of the diffusing of the blowing agent, the EPS bead is unstable for long periods
of time. Figure 5.13.3-5 shows the loss of blowing agent over time when beads are stored under
standard conditions. This diffusion means that the stock of beads must be rotated. An up-to-date
analysis of the blowing agent content of the bead (measured as percent volatiles at 100°C [212°F])
also needs to be maintained, because the blowing agent content determines processing
characteristics, ultimate density, and economics. Expandable beads should be stored below 32°C
(90°F) and in full containers (to reduce gas volume space).
Since pentane, a typical blowing agent, forms explosive mixtures, precautions must be
taken whenever it is used. For example, after storage containers are opened, a time lag of 10
minutes is suggested to allow fumes or pentane vapors to dissipate out of the containers. Care
must be taken to prevent static electricity and sparks from igniting the blowing agent vapors.
9/91 Chemical Process Industry 5.13.3-13
-------
Table 5.13.3-3. EMISSION FACTORS FOR IN-SITU PROCESS
EXPANDABLE POLYSTYRENE"
EMISSION FACTOR RATING: C
Stream
Identification
A
B
C
D
E
F
G
H
Total Plant
Emission
Stream
Mix tank vents
Regranulator hoppers
Reactor vents
Holding tank vents
Wash tank vents
Dryer vents
Product improvement
vents
Storage vents and
conveying loses
Nonmethane
VOC
0.13
negligible
1.09b
0.053
0.023
2.77"
0.008
1.3
537*
References
16
16
17
16
16
16
16
16
0 Stream identification refers to Figure 5.13.3-4. Units are grams VOC per kilogram
of product
b Reference 16. All reactor vents and some dryer vents are controlled in a boiler.
Rates are before control.
c At plant where all reactor vents and some dryer vents are controlled in a boiler
(and assuming 99% reduction), an overall emission rate of 3.75 is estimated.
800
7.75
7.50
7.25
^700
^,675
^6.50
I 625
=• 6.00
5.75
550
525
500
Reg. crystal grade
polystyrene
.s~ Self extinguishing _
4 6 8 10 12 14 16
Weeks
5.133-14
Figure 5.133-5. EPS beads stored in fiber drum at 21 - 24°C (70 - 75°F).
EMISSION FACTORS
9/91
-------
References for Section 5.13.3
1. L. F. Albright, Processes For Major Addition-type Plastics And Their Monomers. McGraw-
Hill, New York, 1974.
2. Modern Plastics Encyclopedia. 1981-1982. McGraw Hill, New York, 1982.
3. Written communication from E. L. Bechstein, Pullman Kellogg, Houston, TX, to M. R.
Glowers, U. S. Environmental Protection Agency, Research Triangle Park, NC, November 6,
1978.
4. Written communication from J. S. Matey, Chemical Manufacturers Association, Washington,
DC, to E. J. Vincent, U. S. Environmental Protection Agency, Research Triangle Park, NC,
October 19,1981.
5. Written communication from P. R. Chancy, Mobil Chemical Company, Princeton, NJ, to J. R.
Fanner, U. S. Environmental Protection Agency, Research Triangle Park, NC, October 13,
1988.
6. Report Of Plant Visit To Monsanto Plastics and Resins Company. Port Plastics. OH. Pacific
Environmental Services, Inc., Durham, NC, September 15,1982.
7. Written communication from R. Symuleski, Standard Oil Company (Indiana), Chicago, IL, to
A. Limpiti, Energy And Environmental Analysis, Inc., Durham, NC, July 2, 1981.
8. Written communication from J. R. Strausser, Gulf Oil Chemicals Company, Houston, TX, to
J. R. Farmer, U. S. Environmental Protection Agency, Research Triangle Park, NC,
November 11,1982.
9. Written communication from J. S. Matey, Chemical Manufacturers Association, Washington,
DC, to C. R. Newman, Energy and Environmental Analysis, Inc., Durham, NC, May 5,1981.
10. Calvin J. Benning, Plastic Foams: The Physics And Chemistry Of Product Performance And
Process Technology. Volume I: Chemistry And Physics Of Foam Formation. John Wiley And
Sons, New York, 1969.
11. S. L. Rosen, Fundamental Principles Of Polymeric Materials. John Wiley And Sons, New
York, 1982.
12. Written communication from K. Fitzpatrick, ARCO Chemical Company, Monaca, PA, to D.
R. Goodwin, U. S. Environmental Protection Agency, Research Triangle Park, NC, February
18, 1983.
13. Written communication from B. F. Rivers, American Hoechst Corporation, Leominster, MA,
to J. R. Farmer, U. S. Environmental Protection Agency, Research Triangle Park, NC, May 4,
1983.
14. Written communication from B. F. Rivers, American Hoechst Corporation, Leominster, MA,
to K. Meardon, Pacific Environmental Services, Inc., Durham, NC, July 20, 1983.
9/91 Chemical Process Industry 5.13.3-15
-------
15. Written communication from T. M. Nairn, Cosden Oil And Chemical Company, Big Spring,
TX, to J. R. Fanner, U. S. Environmental Protection Agency, Research Triangle Park, NC,
March 30,1983.
16. Written communication from A. D. Gillen, BASF Wyandotte Corporation, Parsippany, NJ, to
J. R. Farmer, U. S. Environmental Protection Agency, Research Triangle Park, NC, February
18, 1983.
17. Telephone communication between K. Meardon, Pacific Environmental Services, Inc.,
Durham, NC, and A. Gillen, BASF Wyandotte Corporation, Parsippany, NJ, June 21, 1983.
5.13.3-16 EMISSION FACTORS 9/91
-------
emissions of gaseous ammonia, gaseous fluorides (HF and SiFij) and partic-
ulate ammonium phosphates. These two exhaust streams generally are
combined and passed through primary and secondary scrubbers.
Exhaust gases from the dryer and cooler also contain ammonia,
fluorides and particulates, and these streams commonly are combined and
passed through cyclones and primary and secondary scrubbers. Partic-
ulate emissions and low levels of ammonia and fluorides from product
sizing and material transfer operations are controlled the same way.
Emission factors for ammonium phosphate production are summarized
in Table 6.10.3-1. These emission factors are averages based on recent
source test data from controlled phosphate fertilizer plants in Florida.
Exhaust streams from the reactor and ammoniator-granulator pass
through a primary scrubber, in which phosphoric acid recovers ammonia
and particulate. Exhaust gases from the dryer, cooler and screen go
first to cyclones for particulate recovery, and from there to primary
scrubbers. Materials collected in the cyclone and primary scrubbers are
returned to the process. The exhaust is sent to secondary scrubbers,
where recycled gypsum pond water is used as a scrubbing liquid to control
fluoride emissions. The scrubber effluent is returned to the gypsum
pond.
Primary scrubbing equipment commonly includes venturi and cyclonic
spray towers, while cyclonic spray towers, impingement scrubbers, and
spray-crossflow packed bed scrubbers are used as secondary controls.
Primary scrubbers generally use phosphoric acid of 20 to 30 percent as
scrubbing liquor, principally to recover ammonia. Secondary scrubbers
generally use gypsum and pond water, for fluoride control.
Throughout the industry, however, there are many combinations and
variations. Some plants use reactor-feed concentration phosphoric acid
(40 percent ^2^5) in both primary and secondary scrubbers, and some use
phosphoric acid near the dilute end of the 20 to 30 percent P20s range
in only a single scrubber. Existing plants are.equipped with ammonia
recovery scrubbers on the reactor, ammoniator-granulator and dryer, and
particulate controls on the dryer and cooler. Additional scrubbers for
fluoride removal are common but not typical. Only 15 to 20 percent of
installations contacted in an EPA survey were equipped with spray-
crossflow packed bed scrubbers or their equivalent for fluoride removal.
Emission control efficiencies for ammonium phosphate plant control
equipment have been reported as 94 - 99 percent for ammonium, 75 - 99.8
percent for particulates, and 74 - 94 percent for fluorides.
10/80 Food and Agricultural Industry 6.10.3-3
-------
TABLE 6.10.3-1. AVERAGE CONTROLLED EMISSION FACTORS FOR THE
PRODUCTION OF AMMONIUM PHOSPHATES3
EMISSION FACTOR RATING: A
Emission Point
Reactor /ammoniator-granulator
Fluoride (as F)
Particulates
Ammonia
Dryer /cooler
Fluoride (as F)
Particulates
Ammonia
Product sizing and material transfer
Fluoride (as F)c
Particulates
Ammonia
Total plant emissions
Fluoride (as F)d
Ammonia
Controlled
Ib/ton P20,
0.05
1.52
b
0.03
1.50
b
0.01
0.06
b
0.08
0
0.14
Emission Factors
. kg/MT P205
0.02
0.76
b
0.02
0.75
b
0.01
b
0.04
0.07
^Reference 1, pp. 80-83, 173.
No information available. Although ammonia is emitted from these unit
operations, it is reported as a total plant emission.
^Represents only one sample.
EPA has promulgated a fluoride emission guideline of 0.03 g/kg P?0,.
input.
Based on limited data from only 2 plants.
Reference for Section 6.10.3
1. J. M. Nyers, et al., Source Assessment; Phosphate Fertilizer
Industry, EPA-600/2-79-019c, U.S. Environmental Protection Agency,
Research Triangle Park, NC, May 1979.
6.10.3-4 EMISSION FACTORS 10/80
-------
7. METALLURGICAL INDUSTRY
The metallurgical industry can be broadly divided into primary and secondary metal production
operations. Primary refers to the production of metal from ore. Secondary includes the production of
alloys from ingots and the recovery of metal from scrap and salvage.
The primary metals industry discussed in this chapter includes both ferrous and nonferrous
operations. These processes are characterized by the large quantities of sulfur oxides and paniculate
emitted. Secondary metallurgical process are also discussed, and the major air contaminant from such
activity is paniculate in the forms of metallic fumes, smoke and dust.
9/91 Metallurgical Industry 7.0-1
-------
8.6 PORTLAND CEMENT MANUFACTURING
8.6.1 Process Description
Most of the hydraulic cement produced in the United States is portland cement, a
cementitious, crystalline compound composed of metallic oxides. It is produced by a pyroprocess
in a rotary kiln from raw materials, such as limestone containing calcium carbonate and aluminum,
iron, and silicon oxides, shale, clay and sand. A diagram of this process is shown in Figure 8.6-1.
This manufacturing process may be conveniently divided into five stages, correlated with location
and temperature of the materials in the rotary kiln.
1. Uncombined water evaporates from raw materials as material temperature increases to
100°C (212°F).
2. As the material temperature increases from 100°C to approximately 430°C (800°F),
dehydration and precalcination occur.
3. Between 430°C and 900°C (1650°F), calcination occurs in which CO2 is liberated from
the carbonates.
4. Following calcination, sintering of the oxides occurs in the burning zone of the rotary
kiln at temperatures up to 1510°C (2750°F).
5. Following sintering, cement clinker is produced as the temperature of the material
decreases from 1510°C to 1370°C (2500°F).
The raw material mix enters the kiln at the elevated end, and the burner is at the opposite
end as shown in Figure 8.6-2. The raw materials are then changed into cementitious oxides of
metals by a countercurrent heat exchange process. The materials are continuously and slowly
moved to the lower end by rotation of the kiln. The fuel burned in the kiln may be natural gas,
oil or coal. Many cement plants burn coal, but supplemental fuels such as waste solvents, chipped
rubber, shredded municipal garbage, and coke have been used in recent years.
There are three variations in cement manufacturing, wet, dry, and dry preheater/
precalciner processes. These processes are essentially identical relative to the manufacture of
cement from raw materials. However, the type of process does affect the equipment design,
method of operation, and fuel consumption. Fuel combustion differs between the wet and dry
processes and the preheater/precalciner process. In the former two, all fuel combustion occurs in
the kiln. In the latter, some fuel combustion occurs in a precalcining or calcining vessel before
the materials enter the kiln. See Figure 8.6-2. Generally speaking, preheater/precalciner
equipment uses less fuel and requires a shorter kiln, and the wet process uses the most fuel and
takes the longest kiln, but the relationship is not linear.
9/91 Mineral Products Industry 8.6-1
-------
^DRILLING RIO
.OVERBURDEN
-v. TO CRUSHER
LIMESTONE
n«w m»Uf i«u centlcf oi
corftMnattom O4 Um*«lana.
iron *••. Mod and clar or male.
CRUSHER
To *•!«. tin of (•••
Quarrying and blending of raw ma erials
RECLAIMING PROPORTIONING
SCRAPER BINS
STACKER-RECLAIMING SrSTEM.
STORAGE AND BLENDING
\\ HOT GASES TO
RAW MATERIALS — \2|
ARE PROPORTIONED
ROLLER MILL
ORT MIXING AND
BLINDING SILOS
OROUNO RAW
MATERIAL STORAGE
Proportioning and fine grinding of raw materials.
HOT GASES TO I
ROLLER MILL
RAW MATERIAL FEED
*
ROLLER WILL
n
Kiln system. Preheating, burning, cooling and clinker storage.
BULK STORAGE BULK BULK BOX PACKAdMO TRUCK
TRUCK CAR CAR MACHINE
Finish grinding and shipping.
Figure 8.6-1. Steps in the manufacture of portland cement by dry process
with preheater.
8.6-2
EMISSION FACTORS
9/91
-------
8.6.2 Emissions And Controls
Participate NOr SO2, CO and CO2 are the primary emissions in the manufacture of
Portland cement, and emissions may also include minute particles from the fuel and raw materials.
Sources of particulate at cement plants include (1) quarrying and crushing, (2) raw
material storage, (3) grinding and blending (in the dry process only), (4) clinker production,
(5) finish grinding, and (6) packaging. The largest emission source within cement plants is the
three units of kiln operation: the feed system, the fuel firing system, and the clinker cooling and
handling system. The most desirable method of disposing of the collected dust is injection into
the kiln burning zone and production of clinkers from the dust. If the alkali content of the raw
materials is too high, however, some of the dust is discarded or leached before returning to the
kiln. In many instances, the maximum allowable cement alkali content of 0.6 percent (calculated
as sodium oxide) restricts the amount of dust that can be recycled. Additional sources of
particulate are raw material storage piles, conveyors, storage silos, and loading/unloading facilities.
The complications of kiln burning and the large volumes of material handled have led to
the adoption of many control systems. The industry may use mechanical collectors, electrostatic
precipitators, fabric filters (baghouses), or combinations of these devices to control emissions,
depending on the material emitted, the temperature of plant effluents, and applicable particulate
emission standards and community practices.
Oxides of nitrogen (NOX) are generated during fuel combustion by oxidation of chemically
bound nitrogen in the fuel and by thermal fixation of nitrogen in the combustion air. As flame
temperature increases, the amount of thermally generated NOX increases, and the amount of NOX
generated from fuel increases with the quantity of nitrogen in the fuel. In the cement
manufacturing process, there are two areas which may generate NO,, the burning zone of the kiln
and the burning zone of a precalcining vessel. Fuel use will affect the quantity and type of NOX
generated. Natural gas combustion with a high flame temperature and low fuel nitrogen may
generate a different quantity of NOX than would oil or coal, which have higher fuel nitrogen but
lower flame temperatures.
Fuel use varies in the cement manufacturing process. Generally, natural gas is used only
in the kiln, while coal and oil are used in the kiln and precalcining vessel. Therefore, the
generation and emission of NOX relate to the type of fuel burned and to the extent to which fuel
affects flame temperature and contains chemically bound nitrogen.
Currently, there are data to support only two types of reduction of NOX in the cement
industry. First, for conventional wet and dry process kilns, NOX emissions are reduced by fuel
conversion, with coal producing the least NOr For new construction, the data are not yet clear.
Some preheater/precalciner systems have low emissions and others have high.
There are at least ten different preheater/precalciner systems used in the cement industry,
and each appears to have unique emission properties. However, it is evident that for a single
system, burning oil in the calciner produces less NOX than coal. The NOX emissions from the
preheater/precalciner appear to relate to design. Some have very low emissions and others have
emissions in a mid range of some conventional or wet processes.
9/91 Mineral Products Industry 8.6-3
-------
Kiln Burner
Figure 8.6-2. Conventional portland cement kiln.
Raw
Material
Precalclnlng
Burner
Kiln
Burner
Figure 8.6-3. Typical portland cement preheater/precalciner.
Sulfur dioxide may be generated both from the sulfur compounds in the raw materials, and
from sulfur in the fuel. The sulfur content of both raw materials and fuels will vary from plant to
plant and with geographic location. The alkaline nature of the cement, however, provides for
direct absorption of SO2 into the product. Using a baghouse that allows the SO2 to come in
contact with the cement dust provides inherent reduction of 75 percent or more of the raw
material and fuel sulfur content. The percent reduction, of course, will vary with the alkali and
sulfur content of the raw materials and fuel.
CO emissions are associated with the efficiency of the combustion process, and the CO2 is
generally a release of 33 percent of the weight of the limestone in the calcining process.
Currently, there are no methods available for reducing CO or CO2 except process control for CO
and reduced production for CO2.
8.6-4
EMISSION FACTORS
9/91
-------
Tables 8.6-1 through 8.6-4 give emission factors for cement manufacturing, including
factors based on particle size. Size distributions for particulate emissions from controlled and
uncontrolled kilns and clinker coolers are also shown in Figures 8.6-4 and 8.6-5.
NOTICE
The revised information in this Section
involves only SO2 and NOr The Emission Inventory
Branch intends to update material on particulate and
to add CO information in the future. Toward this end,
we would welcome any emissions data, comments or
suggestions from the reader.
9/91 Mineral Products Industry 8.6-5
-------
U
1
•o "*
M
o c
73 O
I 1
1 1
••i ""
*„ I
^^ ^
3 ~
« -IP
j*
1
(S| g O
o' o o
*O r$ ir>
O O O
C) C5 O*
•o ' 2
(A W
fj •»
O O
^O \£ o
«O ON ^ .
CS (S
oo oo o
r5 "* cs
1-H 1-H
a, ^
a- I
O i i i
o
O i i i
o
O O
«0 »
«r> •*
& ^
cs H
u u
& fe
O 1-1
c5 c3
cs
SO
Hj
n "T- « —
| 1 »
O 1H U
« ti"
§^.
g,^
i ^
g
o Jj g
^ 3 8
SI'S
2
5 o
5 s
4^ g
8.6-6
EMISSION FACTORS
9/91
-------
Table 8.6-2. CONTROLLED PARTICULATE EMISSION FACTORS FOR
CEMENT MANUFACTURING*
Participate
Type of
source
Wet process kiln
Dry process kiln
Clinker cooler
Primary limestone
crusher0
Primary limestone
screen"
Secondary limestone
screen and crusher0
Conveyor transfer0
Raw mill system0-*1
Finish mill system6
Control
Baghouse
ESP
Multiclone
Multicyclone
+ ESP
Baghouse
Gravel bed
filter
ESP
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
Baghouse
kg/Mg
clinker
0.57
0.39
130b
0.34
0.16
0.16
0.048
0.010
0.00051
0.00011
0.00016
0.000020
0.034
0.017
Ib/ton
clinker
1.1
0.78
260b
0.68
0.32
0.32
0.096
0.020
0.0010
0.00022
0.00032
0.000040
0.068
0.034
Emission
Factor
Rating
C
C
D
C
B
C
D
C
D
D
D
D
D
C
"Factors are for kg particulate/Mg (Ib particulate/ton) of clinker produced, except as noted.
ESP = electrostatic precipitator.
bBased on a single test of a dry process kiln fired with a combination of coke and natural gas.
Not generally applicable to a broad cross section of the cement industry.
cExpressed as mass of pollutant/mass of raw material processed.
Includes mill, air separator and weigh feeder.
eExpressed as units of cement produced. Includes mill, air separators) and one or more
material transfer operations.
9/91
Mineral Products Industry
8.6-7
-------
8 I
M 35
£
1
C/5
C/l
d d o o d o
22222-
8.6-8
EMISSION FACTORS
9/91
-------
1000.0
i
8
100.0
si
3°
Jl
V. 10.0
:l
i
3
i.o
o.i
CD Uncontrolled Wet Proem Kiln
Uncontrolled Dry Procea Kiln
Dry Procett Kiln with MulHclono
Wet Proceu Kiln with ESP
Dry Procet* Kiln with BaghouM
I I I I I i t I
100.0
10*0 j
*
2
i.o
i
0.1
1.0 10
Aerodynamic Particle Diameter (/imA)
100
0.01
Figure 8.6-4. Size specific emission factors for cement kiln operations.
9/91
Mineral Products Industry
8.6-9
-------
1
10.Op
10
100
8
17
s .»
1
I o.i
u
0.01
i i i
JJ Uncontrolled CooUn
m CooUn with Grov.l B«d Filf.r
I I I I M I
I
I I I I t I
1.0 10.0
Aerodynamic Parrici* Oianwrvr
10.0
15
o
i o
II
0.1 II
I
0.01
100.0
Figure 8.6-5. Size specific emission factors for clinker coolers in
a portland cement process.
8.6-10
EMISSION FACTORS
9/91
-------
Table 8.6-4. SIZE SPECIFIC EMISSION FACTORS FOR
CLINKER COOLERS"
EMISSION FACTOR RATING: E
Particle
sizeb
Cumulative mass %
* stated sizec
(urn)
Uncontrolled
2.5
5.0
10.0
15.0
20.0
Total mass emission
0.54
1.5
8.6
21
34
factor
Gravel bed filter
40
64
76
84
89
Cumulative emission factor
* stated sized
Uncontrolled
kg/Mg
0.025
0.067
0.40
0.99
1.6
4.6e
Ib/ton
0.050
0.13
0.80
2.0
3.2
9.2e
Gravel bed filter
kg/Mg
0.064
0.10
0.12
0.13
0.14
0.16f
Ib/ton
0.13
0.20
0.24
0.26
0.28
0.32f
bAerodynamic diameter.
founded to two significant figures.
dUnit weight of pollutant/unit weight of clinker produced. Rounded to two significant figures.
"From Table 8.6-1.
'From Table 8.6-2.
References for Section 8.6
1. T. E. Kreichelt, et al.. Atmospheric Emissions From The Manufacture Of Portland Cement. 999-AP-
17, U. S. Environmental Protection Agency, Cincinnati, OH, 1967.
2. Background Information for Proposed New Source Performance Standards: Portland Cement Plants.
APTD-0711, U. S. Environmental Protection Agency, Research Triangle Park, NC, August 1971.
3. A Study Of The Cement Industry In The State Of Missouri. Resources Research Inc., Reston, VA,
December 1967.
4. Portland Cement Plants - Background Information For Proposed Revisions To Standards. EPA-
450/3-85-003a, U. S. Environmental Protection Agency, Research Triangle Park, NC, May 1985.
5. Standards Of Performance For New Stationary Sources. 36 FR 28476, December 1971.
6. Particulate Pollutant System Study. EPA Contract No. CPA-22-69-104, Midwest Research Institute,
Kansas City, MO, May 1971.
7. Restriction Of Emissions From Portland Cement Works. VDI Richtlinien, Duesseldorf, Germany,
February 1967.
9/91
Mineral Products Industry
8.6-11
-------
8. J. S. Kinsey, Lime And Cement Industry - Source Category Report. Vol. II. EPA Contract No. 68-
02-3891, Midwest Research Institute, Kansas City, MO, August 14, 1986.
9. M. S. May, "NOX Generation, Emission And Control From Cement Kilns In The United States",
Proceedings: 1982 Joint Symposium On Stationary Source Combustion NC^ Control. EPA-600/9-88-
026a and 026b, U. S. Environmental Protection Agency, Cincinnati, OH, December 1988.
10. M. S. May, et al.. "Nitrogen Oxide Emissions From Cement Kiln Exhaust Gases By Process
Modification", Proceedings: 1987 Joint Symposium On Stationary Source Combustion NO., Control.
EPA-600/9-88-026a and 026b, U. S. Environmental Protection Agency, Cincinnati, OH, December
1988.
11. J. Croom, et al.. "NOX Formulation In A Cement Kiln: Regression Analysis", Proceedings: 1987 Joint
Symposium On Stationary Source Combustion NO^ Control EPA-600/9-88-026a and 026b, U. S.
Environmental Protection Agency, Cincinnati, OH, December 1988.
12. Methodology For Development Of SO^/NO^ Emission Factor. PSM International, Inc., Dallas, TX,
July 26,1990.
13. F. Bergman, Review Of Proposed Revision To AP-42 Section 8.6. Portland Cement Manufacturing.
EPA Contract No. 68-02-4395, Midwest Research Institute, Kansas City, MO, September 30, 1990.
8.6-12 EMISSION FACTORS 9/91
-------
TABLE 8.19.1-1. UNCONTROLLED PARTICIPATE EMISSION FACTORS
FOR SAND AND GRAVEL PROCESSING PLANTS3
Uncontrolled Operation
Process Sources0
Primary or secondary
crushing (wet)
Open Dust Sources0
Screening"*
Flat screens
(dry product)
Contlnous dropc
Transfer station
Pile formation - stacker
Batch dropc
Bulk loading
Active storage piles?
Active day
Inactive day (wind
erosion only)
^^Unpaved haul roads
^B Wet materials
Emissions by Particle Size Range (aerodynamic diameter)11
Total
Particulate
NA
NA
0.014 (0.029)
MA
0.12 (0.24)
NA
NA
1
TSP
(< 30 ym)
0.009 (0.018)
0.08 (0.16)
NA
0.065 (0.13)
0.028 (0.056)f
14.8 (13.2)
3.9 (3.5)
1
PM10
(< 10 ym)
NA
0.06 (0.12)
NA
0.03 (0.06)e
0.0012 (0.0024)f
7.1 (6.3)e
1.9 (1.7)«
1
Units
kg/Mg (Ib/ton)
kg/Mg (Ib/ton)
kg/Mg (Ib/ton)
kg/Mg (Ib/ton)
kg/Mg (Ib/ton)
kg/hectare/day11
(Ib/acre/day)
kg/hectare/day11
(Ib/acre/day)
Emission
Factor
Rating
D
C
E
E
E
D
D
D
aNA - not available. TSP - total suspended partlculate. Predictive emission factor equations, which generally!
provide more accurate estimates of emissions under specific conditions, are presented in Chapter 11. Factors
for open dust sources are not necessarily representative of the entire Industry or of a "typical" situtation.
bTotal partlculate is airborne particles of all sizes in the source plume. TSP is what is measured by a standard
high volume sampler (aee Section 11.2).
cReferences 5-9.
^References 4-5. For completely wet operations, emissions are likely to be negligible.
Extrapolation of data, using k factors for appropriate operation from Chapter 11.
fFor physical, not aerodynamic, diameter.
KReference 6. Includes the following distinct source operations in the storage cycle: (1) loading of aggregate
onto storage piles (batch or continuous drop operations), (2) equipment traffic in storage areas, (3) wind
erosion of pile (batch or continuous drop operations). Assumes 8 to 12 hours of activity/24 hours.
"Kg/hectare (lb/acre) of storage/day (Includes areas among piles).
*See Section 11.2 for empirical equations.
References for Section 8.19.1
!• Air Pollution Control Techniques For Nonmetallic Minerals Industry,
EPA-450/3-82-014, U. S. Environmental Protection Agency, Research
Triangle Park, NC, August 1982.
2. S. Walker, "Production of Sand and Gravel", Circular Number 57, National
Sand and Gravel Association, Washington, DC, 1954.
3. Development Document For Effluent Limitations Guidelines And Standards -
Mineral Mining And Processing Industry. EPA-440/l-76-059b, U. S. Environ-
mental Protection Agency, Washington, DC, July 1979.
9/91
EMISSION FACTORS
8.19.1-3
-------
4. Review Emissions Data Base And Develop Emission Factors For The Construc-
tion Aggregate Industry, Engineering-Science, Inc., Arcadia, CA, September
1984.
5. "Crushed Rock Screening Source Test Reports on Tests Performed at Conrock
Corp., Irwindale and Sun Valley, CA Plants", Engineering-Science, Inc.,
Arcadia, CA, August 1984.
6. C. Cowherd, Jr., et al., Development Of Emission Factors For Fugitive Dust
Sources, EPA-450/3-74-037, U. S. Environmental Protection Agency, Research
Triangle Park, NC, June 1974.
7. R. Bohn, et al., Fugitive Emissions From Integrated Iron And Steel Plants,
EPA-600/2-78-050, U. S. Environmental Protection Agency, Washington, DC,
March 1978.
8. G. A. Jutze and K. Axetell, Investigation Of Fugitive Dust, Volume I;
Sources, Emissions and Control, EPA-450/3-74-036a, U. S. Environmental
Protection Agency, Research Triangle Park, NC, June 1974.
9. Fugitive Dust Assessment At Rock And Sand Facilities In The South Coast
Air Basin, Southern California Rock Products Association and Southern
California Ready Mix Concrete Association, P.E.S., Santa Monica, CA,
November 1979.
8.19.1-4 Mineral Products Industry 9/85
-------
CO
0)
a)
o
o
u
cd
co
c
VI
u
4-1
CO
U
•H
a
co
g
(0
l-i
ai
a
o
es
I
-*
es
•
ao
cu
t-1
9/88
Mineral Products Industry
8.24-3
-------
%
__
to
w
0
prf
t3
0
C/D
HtO
CO x™v
^-i f r\
|_r l/J
Q H
M
t^ 5^
KU ^-i
U !=>
PM
0 U
M
Q Prf
W H
d §
>— 1 S
O ^
Prf
gCO
w
o a
U M
*Zt "^
nJ
prf <:
0 0
tin O
CO N
a o
0 <
£S
< &
p w
fY
^y
W 55
Prf U
O H
H CO
CJ W
<: s
PH
S5
O
M
W
W
)— f
S
rH
1
sr
CM
co
H
1-J
a
c
0
0!
Cf
°E
U
U
f
/^
h
01
4J
«
•O
U
•fH
8
g
1
V
«
4)
!
h
>-
4
•
g
•H
i
^H
«
Materi
( Operation
u bO
0 C
4J —1
O ij
R) (Q
fa. a?
=
I
a.
(O
H
*••»
§
W>
0-4
\/
•o
O
V|
s
»n
Ol
0
Oi
%
H
u co ca ca co < oo « •< iu
0
^^
u
c
u -^
n H H H H O
^ 00 u u r", S S S S b
J3 5: f £ B >> > > M «
*^. "^i *^» ^^ "**. *•»•» "**^ ""^. *"S» £& U
i?i?2I?£ i?£## 1
^
ONtMinr*. \o <— » O r*»
— t (N O ^ CN< fn •* -M
< 0 0 — 0 O O O O
z • • • • . . • • 4;
o o o o o o o o z
cMinmmm o o O o
inr*F*.r-*r»* <
* • » • • •••• B
o o o o o o o o o
m
N
*£
*~*
•+
rH
OD o m
1^. S^ a •
in m • csi n
• • O ^o ^-\ ^^
-i -• *-s 1 M >
*-% >-•» "O O s^ vx
sO^B-tftt-*"-"^ -H If*
O\ • -*«*• • S^* a • ^) 1 « <4>
tno -^ — • o\ o MIA in]"* ~H
< O'^<4>/BNin«^vc4'^ o o ^N o <
as *z:^x«xoz NO • r o «
O ^> • — ' •
e
"a 3
•^ -w
^%
* m «
^H S^ • •
m CM • w m
a • N ^ SO ^^ ^^
— • ^ • -^ i co k
CM y-N -H -O O »-/ »S
CM tw 0 rt ^^ cn ^^ M ^^ o
§O*<~"*cn* • ^ ,.o»
oo — ^ w -» « o K «n nk» ~* 3
O int ^v *o /-s ^s ^ y-v O *<>] ^s O
• r • x ^0 r o r *> <=> • x o »
O Oj^-'ir,*^ •<*^o*«-» • • ^1 ^-^ • •
C")l CM • O*O O -^
O
e
01
•o c c
t-i « V
3 -O -0
U J3 L. U.
°S 5 J
^H > t-H «-H h U ,_(
* O * « V 01 «
-i u -~. u u « «
tOO-H 01 tl 4J 3 • U
c i-j N G u > «c «i .e C -o c
-i o -H u a e 1-1 w w u B -^
w j,: -a r-i 5. u -. o 4? >«•
« 0 rH M « U -0 ,J ^ rf 45 S 1
•1 9 fH M UN^VidS^ 944V
fHU9 hUt4tl 4U
•0 H CB 0 M 0 > M4
•
1
S
tl
H
M
•>
"8 9
1
D. CM.
5^^> vl
% .
•c «
r-t x^ M W
4 CM V
U . B
0 — S •*
W rt ^S^N B
J*X v< • •
1 B X 0. 13 0 /-.
o ^^ j< • • e >e
B. -In O^i o -H 1
M W W II fH • «
HO e -o « «N
u M V £ *> -H •
• « r!**'<'lto
• « « IX ^4 • «
*3 S *••?"§. "8
«•• vv 41 ,5 a •
SiHV^ tHrHbCI«4H
•H ouvoSlm
«l « U *4« *3 X • ^ i «•*
•IJBJL A»l599VAfr4
£"•§. 5 p g S • .
£ 0 8 B B CtJ 8"°*
* e S S g g^5 o .
• i « « o P " •
0 1 Jj i 1 I M C _, .H
.grH
0 > CM V •
•H • . one
u -a f v u n •
• « H w <^
9 f-< -H * B
cr 41 x " o • >
at > o. «7 *H M
5-o « . u 3
41 n m -ox: uw«
•HUD u • a
'SL-'o B°g^ iJS1"
M S »• 2 M M
««4J AJ^^ «BM
MUD • • -HO
04} V tls-N f«4*
Uifl *H(MBM «J«4M
o e f> o v -~- m
U fi >, W K 3 •
• vl J3 £ 41 O 4J •« ? 5
J< W U O B * 8.
•-< T) ri « O 41 B >>
« 41 2 > lu u^> OIIH
Orrfki 41 D 0 «Sk
•Deference 1, except for c
partlculata. VKT • vehlc
blSP denotea what Is Meaau:
cSyMbola for equatlona:
A - horizontal area,
Hot for vertical
M • Material Molatur
a • Material allt coi
u - vlnd apeed (a/ae
d - drop height (M)
^Multiply the <15 UM equat:
eNultlply the TSP predlctl
f Dating applicable to Mine
8.24-4
EMISSION FACTORS
9/91
-------
9. PETROLEUM INDUSTRY
The petroleum industry involves the refining of crude petroleum and the processing of natural gas
into a multitude of products.
9/91 Metallurgical Industry 9.0-1
-------
In discussing prescribed burning, the combustion process is divided into
preheating, flaming, glowing and smoldering phases. The different phases of
combustion greatly affect the amount of emissions produced. 5~7 The preheating
phase seldom releases significant quantities of material to the atmosphere.
Glowing combustion is usually associated with burning of large concentrations
of woody fuels such as logging residue piles. The smoldering combustion phase
is a very inefficient and incomplete combustion process that emits pollutants
at a much higher ratio to the quantity of fuel consumed than does the flaming
combustion of similar materials.
The amount of fuel consumed depends on the moisture content of the
For most fuel types, consumption during the smoldering phase is much greatest
when the fuel ia driest. When. lower layers of the fuel are moist, the fire
usually is extinguished rapidly. ^
The major pollutants from wildland burning are particulate, carbon monoxide
and volatile organics. Nitrogen oxides are emitted at rates of from 1 to 4
grams per kilogram burned, depending on combustion temperatures. Emissions of
sulfur oxides are negligible. ^
Particulate emissions depend on the mix of combustion phase, the rate of
energy release, and the type of fuel consumed. All of these elements must be
considered in selecting the appropriate emission factor for a given fire and
fuel situation. In some cases, models developed by the U. S. Forest Service
have been used to predict particulate emission factors and source strength. ^
These models address fire behavior, fuel chemistry, and ignition technique, and
they predict the mix of combustion products. There is insufficient knowledge
at this time to describe the effect of fuel chemistry on emissions.
Table 11.1-3 presents emission factors from various pollutants, by fire
and fuel configuration. Table 11.1-4 gives emission factors for prescribed
burning, by geographical area within the United States. Estimates of the
percent of total fuel consumed by region were compiled by polling experts
from the Forest Service. The emission factors are averages and can vary by
as much as 50 percent with fuel and fire conditions. To use these factors,
multiply the mass of fuel consumed per hectare by the emission factor for the
appropriate fuel type. The mass of fuel consumed by a fire is defined as the
available fuel. Local forestry officials often compile information on fuel
consumption for prescribed fires and have techniques for estimating fuel
consumption under local conditions. The Southern Forestry Smoke Management
Guidebook ^ and the Prescribed Fire Smoke Management Guide^-* should be consulted
when using these emission factors.
The regional emission factors in Table 11.1-4 should be used only for
general planning purposes. Regional averages are based on estimates of the
acreage and vegetation type burned and may not reflect prescribed burning
activities in a given state. Also, the regions identified are broadly defined,
and the mix of vegetation and acres burned within a given state may vary
considerably from the regional averages provided. Table 11.1-4 should not be
used to develop emission inventories and control strategies.
To develop state emission inventories, the user is strongly urged to con-
tact that state's federal land management agencies and state forestry agencies
that conduct prescribed burning to obtain. the best information on such activities.
9/88 Miscellaneous Sources 11.1-7
-------
^
O
PQ
a
u
a
B
s
e
s
j- eo
2 <=
o '5
H
E
I
8
J
s
I
j
OH
2
£
a
/fuel
figura
<<< <<< fflfflffl 03 ffl CO Q Q CQ PQ « PQ M PQ <<<
PI r- fir-~
<-/•>
' odviH
8 2223 8 gj 8 S
i/i
g
E
£
&o p.,
«
Mil
1
I
l
11.1-8
EMISSION FACTORS
9/91
-------
1
i
CQ
S
9
g
P^^
2
EMISSION FA
3
« .fc
£ £
§
1
c
o
43
I
Fire/fuel
configura
QQ Q QQ U
II 1 1
(S)
O '
O
o o i- v» • in o
\n ?M »-i «-i >-i <-i
S 0 0, 2
00
i • 00
%% s
T3 'S3 "O
co "3 a> n
w ca *• w
ffi ffl E ffi
-
II §
a
"351
O
u w
1-1 *-
O TO
CS T3
Al
§
ii
..
Sit:"!
2"cr2''w
'S i T i o
j=
o
Sfsllf 1
•O > o MH fe « ^
2 o Q £ SP
|
1
i
M J3 .
9/91
Miscellaneous Sources
11.1-9
-------
TABLE 11.1-4.
EMISSION FACTORS FOR PRESCRIBED BURNING
BY U. S. REGION
Regional
configuration and
fuel type*
Pacific Northwest
Logging slash
Piled slash
Douglas fir/
Western hemlock
Mixed conifer
Ponderosa pine
Hardwood
Underburning pine
Average for region
Pacific Southwest
Sagebrush
Chaparral
Pi nyon/ Juniper
Underburing pine
Grassland
Average for region
Southeast
Palmetto/ gallberry
Underburning pine
Logging slash
Grassland
Other
Average for region
Rocky Mountain
Logging slash
Underburning pine
Grassland
Other
Average for region
North Central and Eastern
Logging slash
Grassland
Underburning pine
Other
Average for region
Percent
of fuelb
42
24
19
6
4
5
100
35
20
20
15
10
100
35
30
20
10
5
100
50
20
20
10
100
50
30
10
10
100
Pollutant0
Particulate
(«/kg)
1*2.5
4
12
12
13
11
30
9.4
8
PMio
5
13
13
13
12
30
10.3
9
9
13
30
10
13.0
15
30
13
10
17
18.8
4
30
10
17
11.9
13
10
30
17
14
PM
6
17
17
20
18
35
13.3
15
15
17
35
10
17.8
16
35
20
10
17
21.9
6
35
10
17
13.7
17
10
35
17
16.5
CO
37
-
175
175
126
112
163
111.1
62
62
175
163
75
101.0
125
163
126
75
175
134
37
163
75
175
83.4
175
75
163
175
143.8
aRegional areas are generalized, e. g., the Pacific Northwest includes
Oregon, Washington and parts of Idaho and California. Fuel types
generally reflect the ecosystems of a region, but users should seek
advice on fuel type mix for a given season of the year. An average
factor for Northern California could be more accurately described as
chaparral, 25Z; Underburning pine, 15Z; sagebrush, 15Z; grassland,
5Z; mixed conifer, 25%; and Douglas fir/Western hemlock, 15Z.
Dash » no data.
bBased on the Judgment of forestry experts.
cAdapted from Table 11.1-3 for the dominant fuel types burned.
11.1-10
EMISSION FACTORS
9/90
-------
References for Section 11.1
1. Development Of Emission Factors For Estimating Atmospheric Emissions From
Forest Fires. EPA-450/3-73-009, U. S. Environmental Protection Agency,
Research Triangle Park, NC, October 1973.
2. D. E. Ward and C. C. Hardy, Advances In The Characterization And Control
Of Emissions From Prescribed Broadcast Fires Of Coniferous Species Logging
Slash On Clearcut Units, EPA DW12930110-01-3/DOE DE-A179-83BP12869, U. S.
Forest Service, Seattle, WA, January 1986.
3. L. F. Radke, et al., Airborne Monitoring And Smoke Characterization Of
Prescribed Fires On Forest Lands In Western Washington and Oregon,
EPA-600/X-83-047, U. S. Environmental Protection Agency, Cincinnati, OH,
July 1983.
4. H. E. Mobley, et al., A Guide For Prescribed Fire In Southern Forests,
U. S. Forest Service, Atlanta, GA, 1973.
5. Southern Forestry Smoke Management Guidebook, SE-10, U. S. Forest Service,
Asheville, NC, 1976.
6. D. E. Ward and C. C. Hardy, "Advances In The Characterization And Control
Of Emissions From Prescribed Fires", Presented at the 77th Annual Meeting
of the Air Pollution Control Association, San Francisco, CA, June 1984.
7. C. C. Hardy and D. E. Ward, "Emission Factors For Particulate Matter By
Phase Of Combustion From Prescribed Burning", Presented at the Annual
Meeting of the Air Pollution Control Association Pacific Northwest
International Section, Eugene, OR, November 19-21, 1986.
8. D. V. Sandberg and R. D. Ottmar, "Slash Burning And Fuel Consumption In
The Douglas Fir Subregion", Presented at the 7th Conference On Fire And
Forest Meteorology, Fort Collins, CO, April 1983.
9. D. V. Sandberg, "Progress In Reducing Emissions From Prescribed Forest
Burning In Western Washington And Western Oregon", Presented at the Annual
Meeting of the Air Pollution Control Association Pacific Northwest
International Section, Eugene, OR, November 19-21, 1986.
10. R. D. Ottmar and D. V. Sandberg, "Estimating 1000-hour Fuel Moistures In
The Douglas Fir Subregion", Presented at the 7th Conference On Fire And
Forest Meteorology, Fort Collins, CO, April 25-28, 1983.
11. D. V. Sandberg, et al.. Effects Of Fire On Air - A State Of Knowledge
Review, WO-9, U. S. Forest Service, Washington, DC, 1978.
12. C. K. McMahon, "Characteristics Of Forest Fuels, Fires, And Emissions",
Presented at the 76th Annual Meeting of the Air Pollution Control
Association, Atlanta, GA, June 1983.
13. D. E. Ward, "Source Strength Modeling Of Particulate Matter Emissions From
Forest Fires", Presented at the 76th Annual Meeting of the Air Pollution
Control Association, Atlanta, GA, June 1983.
9/90
Miscellaneous Sources 11.1-11
-------
14. D. E. Ward, et al., "Particulate Source Strength Determination For Low-
intensity Prescribed Fires," Presented at the Agricultural Air
Pollutants Specialty Conference, Air Pollution Control Association,
Memphis, TN, March 18-19, 1974.
15. Prescribed Fire Smoke Management Guide, 420-1, BIFC-BLM Warehouse,
Boise, ID, February 1985.
16. Colin C. Hardy, Emission Factors For Air Pollutants From Range
Improvement Prescribed Burning Of Western Juniper And Basin Big
Sagebrush, PNW 88-575, Office Of Air Quality Planning And Standards,
U. S. Environmental Protection Agency, Research Triangle Park, NC,
March 1990.
17. Colin C. Hardy and D. R. Teesdale, Source Characterization And Control
Of Smoke Emissions From Prescribed Burning Of California Chaparral,
CDF Contract No. 89CA96071, California Department of Forestry And
Fire Protection, Sacramento, CA, 1991.
18. Darold E. Ward and C. C. Hardy, "Emissions From Prescribed Burning
of Chaparral," Proceedings Of The 1989 Annual Meeting Of The Air And Waste
Management Association, Anaheim, CA, June 1989.
11.1-12 EMISSION FACTORS 9/91
-------
11.4 WET COOLING TOWERS
11.4.1 General1
Cooling towers are heat exchangers which are used to dissipate large heat loads to the
atmosphere. They are used as an important component in many industrial and commercial
processes needing to dissipate heat. Cooling towers may range in size from less than 5.3(10)6
kilojoules (5(10) British Thermal Units per hour) for small air conditioning cooling towers to
over 5275(10)6 kilojoules per hour (5000(10)6 Btu/h) for large power plant cooling towers.
Although cooling towers can be classified several ways, the primary classification is into
dry towers or wet towers. However, some hybrid wet-dry combinations exist. Subclassifications
can include the type of draft and/or the location of the draft relative to the heat transfer medium,
the type of heat transfer medium, the relative direction of air movement, and the type of
distribution system.
When water is used as the heat transfer medium, wet or evaporative cooling towers may
be used. Wet cooling towers rely on the latent heat of water evaporation to exchange heat
between the process and the air passing through the cooling tower. The cooling water may be an
integral part of the process or provide cooling via heat exchangers.
In wet cooling towers, the heat transfer is measured by the decrease in the process
temperature and a corresponding increase in the moisture content and wet bulb temperature of
the air passing through the cooling tower. (There may also be a change in the sensible, or dry
bulb, temperature; however, its contribution to the heat transfer process is very small and is
typically ignored when designing wet cooling towers.) Wet cooling towers typically have a wetted
media called "fill" to promote evaporation by providing a large surface area and/or by creating
many water drops with a large cumulative surface area.
Cooling towers can be categorized by: the type of heat transfer; the type of draft and
location of the draft relative to the heat transfer medium; the type of heat transfer medium; the
relative direction of air and water contact; and the type of water distribution system. Since
evaporative cooling towers are the dominant type, and they also generate air pollutants, this
Section will address only that type of tower. Diagrams of the various tower configurations are
shown in Figures 11.4-1 and 11.4-2.
11.4.2 Emissions And Controls1
Because wet cooling towers have direct contact between the cooling water and the air
passing through the tower, some of the liquid water may be entrained in the air stream and be
carried out of the tower as "drift" droplets. Therefore, the constituents of the drift droplets, i. e.,
paniculate matter, may be classified an emission.
The magnitude of drift loss is influenced by the number and size of droplets produced
within the cooling tower, which in turn are determined by the fill design, the air and water
patterns, and other interrelated factors. Tower maintenance and operation can also influence the
formation of drift droplets. For example, excessive water flow, excessive air flow, and water
bypassing the tower drift eliminators can promote and/or increase drift emissions.
9/91 Miscellaneous Sources 11.4-1
-------
Since the drift droplets generally have the same water chemistry as the water circulating
through the tower, they may compose airborne emissions. Large drift droplets settle out of the
tower exhaust air stream and deposit near the tower. This can lead to wetting, icing, salt
deposition, and related problems such as damage to equipment or vegetation. Since other drift
droplets may evaporate before being deposited in the area surrounding the tower, they can also
result in PM-10 emissions. PM-10 is generated when the drift droplets evaporate leaving fine
paniculate matter formed by crystallization of dissolved solids. Dissolved solids found in cooling
tower drift can consist of mineral matter, chemicals for corrosion inhibition, etc.
Water OulM
MrOulM
AirOutM
OiKM
Coumarflow Natural On* Toi
Fan
MrOulM
Mr
Mr
FanAuMCountorttnr
MueadOnft
FanAMbtCrowflow
Muead On*
Figure 11.4-1. Atmospheric and natural draft cooling towers.
11.4-2
EMISSION FACTORS
9/91
-------
AirOulW
Fan
AirOutM
: Air
• Inl*
Fornd DraH Countarfbw Toimr
Muc*d Draft CounMrflow Toww
WtfwInM
n I i II i
WlterOulM
Fornd Draft Crow Flow Toww
Induced Draft CmMflow To
Figure 11.4-2. Mechanical draft cooling towers.
In order to reduce the drift from cooling towers, drift eliminators are usually incorporated
into the cooling tower design to remove as many droplets as practical from the air stream before
exiting the tower. The drift eliminators used in cooling towers rely on inertial separation caused
by direction changes while passing through the eliminators. Drift eliminator configurations
include herringbone (blade-type), wave form, and cellular (or honeycomb) designs, with the
cellular units generally being most efficient.
Like cooling tower fill materials, drift eliminators may include various materials such as
ceramic, fiber reinforced cement, fiberglass, metal, plastic, and wood installed or formed into
closely spaced slats, sheets, honeycomb assemblies, or tiles. The materials may have other
features such as corrugations and water removal channels to enhance the drift removal further.
Table 11.4-1 provides available paniculate emission factors for wet cooling towers.
Separate emission factors are given for induced draft and natural draft cooling towers. Also note
that the factors shown in Table 11.4-1 most closely represent older towers with less efficient mist
elimination.
9/91
Miscellaneous Sources
11.4-3
-------
1,
I
TO
It,
+^
a
i
i
«
u.
ts E
1 S
x 5s
6|
1
K
P
u
g
H
m
0\ .,
TH ^
0 ^
C5 Z
ro ^^
c5 IZ
Q U
- i
TH Q
^^
O »
H c>
.
o d>
£2 •!-•
1 1
3 9
i «
^ z
~ M »-> S CO
klSslgl
te . *c ^L BoC^ -
I «o
& s
1 •=! 73
Sc* g>^a^
1 ~ r. .S ^ "" 73
e*rr "n 3 o a
§r.
•s-.a ^ * J I o A
-ll^llll
2||S.|:|§
53 S 2 » s § § §
fls!
^s-§
' .a j £ o 2
-------
References for Section 11.4
1. J. S. Kinsey, et al.. Development Of Participate Emission Factors For Wet Cooling Towers.
EPA Contract No. 68-DO-0137, Midwest Research Institute, Kansas City, MO, September
1991.
2. N. M. Stich, Cooling Tower Test Report. Drift And PM-10 Tests T89-50. T89-51. and T89-52.
Midwest Research Institute, Kansas City, MO, February 1990.
3. Cooling Tower Test Report. Typical Drift Test. Midwest Research Institute, Kansas City, MO,
January 1990.
4. Mass Emission Measurements Performed On Kerr-McGee Chemical Corporation's Westend
Facility. Kerr-McGee Chemical Corporation. Trona. California. Environmental Systems
Corporation, Knoxville, TN, December 1989.
5. Cooling Tower Drift Test Report For Unnamed Client Of The Cooling Tower Institute.
Houston. Texas. Midwest Research Institute, Kansas City, MO, January 1989.
6. Cooling Tower Drift Test Report For Unnamed Client Of The Cooling Tower Institute.
Houston. Texas. Midwest Research Institute, Kansas City, MO, October 1988.
7. Cooling Tower Drift Test Report For Unnamed Client Of The Cooling Tower Institute.
Houston. Texas. Midwest Research Institute, Kansas City, MO, August 1988.
8. Report Of Cooling Tower Drift Emission Sampling At Argus And Sulfate #2 Cooling Towers.
Kerr-McGee Chemical Corporation. Trona. California. Environmental Systems Corporation,
Knoxville, TN, February 1987.
9. Cooling Tower Drift Test Report For Unnamed Client Of The Cooling Tower Institute.
Houston. Texas. Midwest Research Institute, Kansas City, MO, February 1987.
10. Cooling Tower Drift Test Report For Unnamed Client Of The Cooling Tower Institute.
Houston. Texas. Midwest Research Institute, Kansas City, MO, January 1987.
11. Isokinetic Droplet Emission Measurements Of Selected Induced Draft Cooling Towers. Kerr-
McGee Chemical Corporation. Trona. California. Environmental Systems Corporation,
Knoxville, TN, November 1986.
12. Cooling Tower Drift Test Report For Unnamed Client Of The Cooling Tower Institute.
Houston. Texas. Midwest Research Institute, Kansas City, MO, December 1984.
13. Cooling Tower Drift Test Report For Unnamed Client Of The Cooling Tower Institute.
Houston. Texas. Midwest Research Institute, Kansas City, MO, August 1984.
14. Cooling Tower Drift Test Report For Unnamed Client. Midwest Research Institute, Kansas
City, MO, November 1983.
15. J. H. Meyer and William Stanbro, Chalk Point Cooling Tower Project. Volumes 1 and 2. JHU
PPSP-CPCTP-16, John Hopkins University, Laurel, MD, August 1977.
9/91 Miscellaneous Sources 11.4-5
-------
16. J. K. Chan and M. W. Golay, Comparative Evaluation Of Cooling Tower Drift Eliminator
Performance. MIT-EL 77-004, Massachusetts Institute Of Technology, Energy Laboratory
And Department of Nuclear Engineering, Cambridge, MA, June 1977.
17. G. O. Schrecker, et al. Drift Data Acquired On Mechanical Salt Water Cooling Devices.
EPA-650/2-75-060, U. S. Environmental Protection Agency, Cincinnati, OH, July 1975.
11.4-6 EMISSION FACTORS 9/91
-------
11.5 INDUSTRIAL FLARES
11.5.1 General
Flaring is a high temperature oxidation process used to burn combustible components,
mostly hydrocarbons, of waste gases from industrial operations. Natural gas, propane, ethylene,
propylene, butadiene and butane constitute over 95 percent of the waste gases flared. In
combustion, gaseous hydrocarbons react with atmospheric oxygen to form carbon dioxide (CO^)
and water. In some waste gases, carbon monoxide (CO) is the major combustible component.
Presented below, as an example, is the combustion reaction of propane.
CjHg + 5 O2—> 3 CO2 + 4 H2O
During a combustion reaction, several intermediate products are formed, and eventually,
most are converted to CO2 and water. Some quantities of stable intermediate products such as
carbon monoxide, hydrogen and hydrocarbons will escape as emissions.
Flares are used extensively to dispose of 1) purged and wasted products from refineries, 2)
unrecoverable gases emerging with oil from oil wells, 3) vented gases from blast furnaces, 4)
unused gases from coke ovens, and 5) gaseous wastes from chemical industries. Gases flared from
refineries, petroleum production, chemical industries, and to some extent, from coke ovens, are
composed largely of low molecular weight hydrocarbons with high heating value. Blast furnace
flare gases are largely of inert species and CO, with low heating value. Flares are also used for
burning waste gases generated by sewage digesters, coal gasification, rocket engine testing, nuclear
power plants with sodium/water heat exchangers, heavy water plants and ammonia fertilizer plants.
There are two types of flares, elevated and ground flares. Elevated flares, the more
common type, have larger capacities than ground flares. In elevated flares, a waste gas stream is
fed through a stack anywhere from 10 to over 100 meters tall and is combusted at the tip of the
stack. The flame is exposed to atmospheric disturbances such as wind and precipitation. In
ground flares, combustion takes place at ground level. Ground flares vary in complexity, and they
may consist either of conventional flare burners discharging horizontally with no enclosures or of
multiple burners in refractory-lined steel enclosures.
The typical flare system consists of 1) a gas collection header and piping for collecting
gases from processing units, 2) a knockout drum (disentrainment drum) to remove and store
condensables and entrained liquids, 3) a proprietary seal, water seal, or purge gas supply to
prevent flash-back, 4) a single or multiple burner unit and a flare stack, 5) gas pilots and an
ignitor to ignite the mixture of waste gas and air, and if required, 6) a provision for external
momentum force (steam injection or forced air) for smokeless flaring. Natural gas, fuel gas, inert
gas or nitrogen can be used as purge gas. Figure 11.5-1 is a diagram of a typical steam-assisted
elevated smokeless flare system.
Complete combustion requires sufficient combustion air and proper mixing of air and
waste gas. Smoking may result from combustion, depending upon waste gas components and the
quantity and distribution of combustion air. Waste gases containing methane, hydrogen, CO and
ammonia usually burn without smoke. Waste gases containing heavy hydrocarbons, such as
paraffins above methane, olefins and aromatics, cause smoke. An external momentum force, such
9/91 Miscellaneous Sources 11.5-1
-------
as steam injection or blowing air, is used for efficient air/waste gas mixing and turbulence, which
promotes smokeless flaring of heavy hydrocarbon waste gas. Other external forces may be used
for this purpose, including water spray, high velocity .vortex action or natural gas. External
momentum force is rarely required in ground flares.
Steam injection is accomplished either by nozzles on an external ring around the top of
the flare tip or by a single nozzle located concentrically within the tip. At installations where
waste gas flow varies, both are used. The internal nozzle provides steam at low waste gas flow
rates, and the external jets are used with large waste gas flow rates. Several other special purpose
flare tips are commercially available, one of which is for injecting both steam and air. Typical
steam usage ratio varies from 7:1 to 2:1, by weight
Waste gases to be flared must have a fuel value of at least 7500 to 9300 kilojouks per
cubic meter (200 to 250 British Thermal Units per cubic foot) for complete combustion, otherwise
fuel must be added. Flares providing supplemental fuel to waste gas are known as fired, or
endothermic, flares. In some cases, flaring waste gases even having the necessary heat content
will also require supplemental heat If fuel bound nitrogen is present, flaring ammonia with a
heating value of 13,600 kJ/m3 (365 Btu/ft3) will require higher heat to minimize nitrogen oxide
(NOJ formation.
fUU STACK
mm
STUM
ICMIM TIM
•— STACK SIM.
IOMIM
-CD-
IGNITION
'(AS
• moT CAS
ouniw
WttUI CIIHI KAMI
TIAtmillW
Figure 11.5-1. Diagram of a typical steam-assisted smokeless elevated flare.
11.5-2
EMISSION FACTORS
9/91
-------
At many locations, flares normally used to dispose of low volume continuous emissions are
designed to handle large quantities of waste gases which may be intermittently generated during
plant emergencies. Flare gas volumes can vary from a few cubic meters per hour during regular
operations up to several thousand cubic meters per hour during major upsets. Flow rates at a
refinery could be from 45 to 90 kilograms per hour (100 - 200 pounds per hour) for relief valve
leakage but could reach a full plant emergency rate of 700 megagrams per hour (750 tons per
hour). Normal process blowdowns may release 450 to 900 kg/hr (1000 - 2000 Ib/hr), and unit
maintenance or minor failures may release 25 to 35 Mg/hr (27 - 39 tons/hr). A 40 molecular
weight gas typically of 0.012 cubic nanometers per second (25 standard cubic feet per minute) may
rise to as high as 115 cubic nanometers per second (241,000 scfm). The required flare turndown
ratio for this typical case is over 15,000 to 1.
Many flare systems have two flares, in parallel or in series. In the former, one flare can
be shut down for maintenance while the other serves the system. In systems of flares in series,
one flare, usually a low-level ground flare, is intended to handle regular gas volumes, and the
other, an elevated flare, to handle excess gas flows from emergencies.
11.5.2 Emissions
Noise and heat are the most apparent undesirable effects of flare operation. Flares are
usually located away from populated areas or are sufficiently isolated, thus minimizing their effects
on populations.
Emissions from flaring include carbon particles (soot), unburned hydrocarbons, CO, and
other partially burned and altered hydrocarbons. Also emitted are nitrogen oxides (NOJ and, if
sulfur-containing material such as hydrogen sulfide or mercaptans is flared, sulfur dioxide (SO^.
The quantities of hydrocarbon emissions generated relate to the degree of combustion. The
degree of combustion depends largely on the rate and extent of fuel-air mixing and on the flame
temperatures achieved and maintained. Properly operated flares achieve at least 98 percent
combustion efficiency in the flare plume, meaning that hydrocarbon and CO emmissions amount
to less than 2 percent of hydrocarbons in the gas stream.
The tendency of a fuel to smoke or make soot is influenced by fuel characteristics and by
the amount and distribution of oxygen in the combustion zone. For complete combustion, at least
the stoichiometric amount of oxygen must be provided in the combustion zone. The theoretical
amount of oxygen required increases with the molecular weight of the gas burned. The oxygen
supplied as air ranges from 9.6 units of air per unit of methane to 38.3 units of air per unit of
pentane, by volume. Air is supplied to the flame as primary air and secondary air. Primary air is
mixed with the gas before combustion, whereas secondary air is drawn into the flame. For
smokeless combustion, sufficient primary air must be supplied, this varying from about 20 percent
of stoichiometric air for a paraffin to about 30 percent for an olefin. If the amount of primary air
is insufficient, the gases entering the base of the flame are preheated by the combustion zone,
and larger hydrocarbon molecules crack to form hydrogen, unsaturated hydrocarbons and carbon.
The carbon particles may escape further combustion and cool down to form soot or smoke.
Olefins and other unsaturated hydrocarbons may polymerize to form larger molecules which crack,
in turn forming more carbon.
The fuel characteristics influencing soot formation include the carbon-to-hydrogen ratio
and the molecular structure of the gases to be burned. All hydrocarbons above methane, i. e.,
those with a C-to-H ratio of greater than 0.33, tend to soot. Branched chain paraffins smoke
9/91 Miscellaneous Sources 11.5-3
-------
more readily than corresponding normal isomers. The more highly branched the paraffin, the
greater the tendency to smoke. Unsaturated hydrocarbons tend more toward soot formation than
do saturated ones. Soot is eliminated by adding steam or air, hence most industrial flares are
steam assisted and some are air assisted. Rare gas composition is a critical factor in determining
the amount of steam necessary.
Since flares do not lend themselves to conventional emission testing techniques, only a few
attempts have been made to characterize flare emissions. Recent EPA tests using propylene as
flare gas indicated that efficiencies of 98 percent can be achieved when burning an offgas with at
least 11,200 kJ/m3 (300 Btu/ft3). The tests conducted on steam-assisted flares at velocities as low
as 39.6 meters per minute (130 feet per minute) to 1140 m/min (3750 ft/min), and on air-assisted
flares at velocities of 180 m/min (617 ft/min) to 3960 m/min (13,087 ft/min) indicated that
variations in incoming gas flow rates have no effect on the combustion efficiency. Flare gases
with less than 16,770 kJ/m3 (450 Btu/ft3) do not smoke.
Table 11.5-1 presents flare emission factors, and Table 11.5-2 presents emission
composition data obtained from the EPA tests.1 Crude propylene was used as flare gas during
the tests. Methane was a major fraction of hydrocarbons in the flare emissions, and acetylene was
the dominant intermediate hydrocarbon species. Many other reports on flares indicate that
acetylene is always formed as a stable intermediate product. The acetylene formed in the
combustion reactions may react further with hydrocarbon radicals to form polyacetylenes followed
by polycyclic hydrocarbons.2
In flaring waste gases containing no nitrogen compounds, NO is formed either by the
fixation of atmospheric nitrogen with oxygen or by the reaction between the hydrocarbon radicals
present in the combustion products and atmospheric nitrogen, by way of the intermediate stages,
HCN, CN, and OCN.2 Sulfur compounds contained in a flare gas stream are converted to SO2
when burned. The amount of SO2 emitted depends directly on the quantity of sulfur in the flared
gases.
Table 11.5-1. EMISSION FACTORS FOR FLARE OPERATIONS8
EMISSION FACTOR RATING: B
Component
Emission Factor
(lb/106 Btu)
Total hydrocarbonsb
Carbon monoxide
Nitrogen oxides
Sootc
0.14
0.37
0.068
0 to 274
Reference 1. Based on tests using crude propylene
containing 80 % propylene and 20 % propane.
bMeasured as methane equivalent.
°Soot in concentration values: nonsmoking flares, 0 ug/liter;
lightly smoking flares, 40 ug/1; average smoking flares,
177 jig/1; and heavily smoking flares, 274 jig/1.
11.5-4
EMISSION FACTORS
9/91
-------
Table 11.5-2. HYDROCARBON COMPOSITION OF FLARE EMISSION*
Composition
Methane
Ethane/Ethylene
Acetylene
Propane
Propylene
Average (range),
Volume %
55
8
5
7
25
(14 - 83)
(1 - 14)
(0.3 - 23)
(0 - 16)
(1 - 65)
Tieference 1. Ranges in parentheses. The composition
presented is an average of a number of test results
obtained under the following sets of test conditions:
steam-assisted flare using high Btu content feed;
steam-assisted using low Btu content feed; air-assisted
flare using high Btu content feed; and air-assisted
flare using low Btu content feed. In all tests,
"waste" gas was a synthetic gas consisting of a
mixture of propylene and propane.
References for Section 11.5
1. Flare Efficiency Study. EPA-600/2-83-052, U. S. Environmental Protection Agency, Cincinnati,
OH, July 1983.
2. K. D. Siegel, Degree Of Conversion Of Flare Gas In Refinery High Flares. Dissertation,
University of Karlsruhe, Karlsruhe, Germany, February 1980.
3. Manual On Disposal Of Refinery Wastes. Volume On Atmospheric Emissions. API Publication
931, American Petroleum Institute, Washington, DC, June 1977.
9/91
Miscellaneous Sources
11.5-5
-------