-------
of the distance from the surface leads to an imbalance in the molecular
collisions on a particle close to the surface of the droplet. This
imbalance causes the particle to migrate toward the site of the condensa-
tion. The opposite effect occurs when the droplet is undergoing evaporation.
2.2.1.4 Thermophoresis. When there is a distinct temperature gradient,
there is an imbalance between the forces exerted on the sides of a particle
by the molecular collisons. The molecules on the higher temperature side
have a greate.r ye,lpcity^a,nd thereby transfer more mqmen.tum to the particle
. "~g ' " • " '- * *e '—i'* *(> -n jH'MtV, *f"7MM'•* t~} \ . f t^j ,j> * , j i AU ,si i I •
when colliding than those molecules on the colder side. The result is a
net movement toward the colder temperatures.
jon z~ norjsijsnsq no norauttfo nernwoiS -TO *aa't^9. fel'O'f^snad sriT ,9sl~a
fe Rjaytt-iclas'&igei airfl rfr J'nsisqqB
^^^V.f^W^^^^^.
etration curve as a function of the particle size generally appears as
shown in Figure 2-6. The peak penetration corresponds to the 0.2 to 0.5
?'1?/^":'-,:F0 fll .'W
1.0
DIFFUSION
MORE
EFFECTIVE
I I I 11
I i i i 11
O
o.io
UJ
•z.
UJ
a.
O.OI
IMF-ACTION MORE
EFFECTIVE
I i i i i i 11
O.I 0.2 0.5 1.0 234 56789IO
PARTICLE DIAMETER,
Figure 2-6.
Hypothetical curve illustrating general relationship
between penetration and particle size.
2-12
-------
micrometer size range, in which inertia! impaction is relatively ineffective
and the particles are too large for significant diffusional capture.
The emissions from a given scrubber system depend primarily on the
quantity of parti cul ate matter in the size range from 0.1 to 5 micrometers
and the penetration level achievable for the particle sizes. The latter
is generally considered to be a function of the total power input into
the scrubber and the type of scrubber itself.
'"'• tlie:|iartiCle: s!i'z^':d1ependehce frf venturi scr:ubbers: is il 1 ustrated' in
Figure 2-7. The highest penetration is for particles of 0.5 micrometers in
,. .."nu.^'tawriaJ 'teDiOD srtj tnswp;? tnjamevqm tan
size. The beneficial effect of Brbwnia'n diffusion on penetration is not
apparent in this ft^e^fDMosjK^^ s>i%tfi#
determine mass concentrations as the total mass with diameters less than a
^ -.-,.. -, e, -. »pt • j-f. T-. • _ _, ......... :.„.-.,.,- . . '.< ..... '*.
''"han 0.2
If the Affect of Brownian
diffusion had been included, the curve shown in Figure 2-7 would begin
1.0
§
cc
Lu
Q
5
or
5
UJ
0.
UJ
o
p
0.01
1 — r~n MIII
T 1 I I I I II-
' ' i ' * ' ' i
1.0
PARTICLE DIAMETER,
10
Figure 2-7. Fractional efficiency curve for venturi scrubbers.7
2-13
-------
to resemble the curve presented in Figure 2-6. A number of similar
particle size-penetration curves are presented in references 3, 4, and 5.
On a given scrubber system a shift in the particle size distribution
in the gas stream has a major impact on the performance. The can be
illustrated by use of the venturi scrubber model described in references
6 through 9. For a hypothetical set of operating conditions, the model was
used to calculate the penetration. The mass mean particle size was
varied from 5 to 20 micrometers and a standard-deviatri on of,2v5-.was,-used
^ „.-.-,
in all three cases. The results are shown in Figure 2-8. The predicted
ibyr a-.fjacfcorr o,fi 3s,whenf«thaj sjzBijdlfetnSbiifclm shifts
5 10 15 20
GEOMETRIC MASS MEAN DIAMETER, microns
Figure 2-8. Predicted relationship between overall penetration
and particulate mass mean diameter.
2-14
-------
2.2.3 Factors That Modify Particle Size
Numerous process-related operating factors and upset conditions
affect the particle size distribution at the scrubber inlet. Thus a wet
scrubber system cannot be considered as a separate entity operating
independently of the process it serves.
2.2.3.1 Process Operational Change. Any change that modifies the
quantity of particulate and vaporous materials evolved from the process
can cause a decrease; in the particle size distribution. For example, an
increase in the sodium content of the lime sludge feed to a lime kiln at
a, kraf t Kpulp: mi 3 1 woul du subs tan ti a! Ty; i"n:cheases the Jqu DtSM *ov.£i
micrometer size particulate matter
the
! ' I '••
plants can lead to a shift in the quantity of organic material volatilized
and thereby increase the quantity of of very fine hydrocarbon aerosol
particulate which must be removed.
The indications of a change in the process operation may include
some of the following: a major increase or decrease in the gas tempera-
ture entering the scrubber, a change in the types of raw materials (to
the extent this can be determined), an increase in the residual opacity
of the plume downstream from the scrubber (bluish white material), and a
change in the fuel characteristics. As many symptoms as possible should
be considered together to reliably identify process changes as responsible
for a particular shift in the particle size disbribution.
2.2.3.2 Evaporative Release. Many commercial wet scrubbers are
preceded by a quench tower which serves to cool the gas stream to the
125° to 200°F temperature range prior to its entry into the scrubber.
This is necessary to protect scrubber internal parts and may also improve
particle capture.
The liquor quality used in the quench tower (or presaturator) varies
from source to source depending on the types of spray nozzles used in the
quench tower and the need to recycle liquor to minimize treatment costs.
The total solids content can sometimes be as high as 15% by weight.
Since a major portion of the liquor injected into the quench tower and/or
presaturator evaporates, some of the solids entering as suspended and
dissolved solids in the liquor are released as particulate matter.
2-15
-------
Kalika has presented some calculated effects of evaporative release
based on conditions observed at an incinerator scrubber.10 The results
are provided in Table 2-1. In preparing these data, he assumed an inciner-
ator exit temperature of 1600°F with a humidity of 0.08 pound water per
pound of dry air. He also assumed that 85% of the quench water droplets
evaporate to dryness.
TABLE 2-1. EFFECT OF RECIRCULATION LIQUOR SOLIDS10
Solids
in Recirculated
Water, %
»•
2
5
10
Water
Recirculated
to Quench, Gpm
0
22
25
26
Outlet
Dust Loading,
•Lb./Min.
0.032
0.114
0.259
0.512
Apparent
Scrubber
Efficiency, %
98.0
92.9
83.9
68.2
As the solids content of the liquor is increased to 10% by weight, the
emissions from the scrubber are predicted to increase by more than an
order-of magnitude. These data provide only a rough guide to the potential
effect; the actual effect will depend on the droplet size distribution
in the quench tower, the gas temperature at the quench tower inlet, the
liquid-to-gas ratio at the quench tower, the residence time within the
quench tower, and the manner in which the droplets shatter during the later
stages of evaporation. A droplet must be completely evaporated before
the suspended and dissolved particulate matter are released.
At the present time, the extent to which evaporative release con-
tributes to additional particulate matter emissions is not well documented.
Considerable study will be necessary to determine the evaporative con-
ditions within a quench tower and the effect of the released material on
the particle size distribution.
Indirect indications .of potential or actual evaporative release in-
clude: high turbidity liquor going to the quench tower, frequent pluggage
problems with the quench tower nozzles, high gas inlet temperatures, and
high residual opacity.
2-16
-------
2.2.3.3 Vapor Condensation. Some industrial sources generate organic
and/or metallic vapor in addition to partial!ate matter. If these vaporous
materials begin to condense in a homogeneous manner while passing through
the scrubber, very little of this material will be'captured. To the
extent possible, the condensation process should be initiated as far
upstream of the scrubber as possible to enhance heterogenous condensation
on the surfaces of existing aerosols.
Symptoms of increased vapor condensation aerosol include a change in
the process operating temperature, a change in the use of raw materials
and/or corrosion inhibitors, and increased residual opacity plume downwind
of the stack exit.
2.3 PRESSURE DROP AS A PERFORMANCE PARAMETER
The static pressure drop of the gas stream passing through a particu-
late wet scrubber is used extensively by both operators and inspectors to
evaluate scrubber performance. In some cases, conclusions regarding
adequacy of performance have been based solely on these data. The uses and
limits of pressure drop as a performance parameter are examined in this
section.
2.3.1 Definition of Pressure Drop
Pressure drop is the reduction in potential energy (through conversion
to kinetic energy) of the gas stream as it passes from one point to
another. It results from frictional losses on the ductwork and scrubber
internal components, from acceleration of the gas and liquid streams,
from the atomization of the liquor, and from any changes in elevation of
the gas and liquid streams.
The scrubber pressure drop is simply the arithmetic difference
batween the static pressures of the gas stream at the inlet and at the
outlet of the collector. If there are two collectors in series, the
total pressure drop across the system is the sum of• the two individual
pressure drops. The additive nature of static pressure drop (P^) is
illustrated in Equation 2-2.
Pt =
P2
Equation 2-2
2-17
-------
where:
Pt = total pressure drop of system
Pj[ = pressure drop across unit 1
?2 = pressure drop across unit 2
Pn = pressure drop across unit n
The same concept applies to a single scrubber with multiple trays. The
pressure drop across the scrubber is the sum of the pressure drops across
each tray plus the pressure drop across the demister.
If there are two scrubbers in parallel on a given effluent stream, as
shown in Figure 2-9, the pressure drop across the system is equivalent to
the pressure drop across either unit. In other words, the pressure drop
along each path must be identical. The equivalent nature of static pressure
along parallel paths is illustrated in Equation 2-3.
Pt = P2 - Pi = P4 -
Equation 2-3
where:
P-t = total pressure drop of system
?l = static pressure at inlet of path A
?2 - static pressure at outlet of path A
?3 = static pressure at inlet of path B
P4 = static pressure at outlet of path B
While the pressure drops across each path must be identical, the
pressure drops across each scrubber may not be identical. If one of the
paths has longer ductwork, more elbows, or a flow restriction such as a
damper or solids deposit, then the pressure drop of the scrubber on this
path will be less than that in the other path. This difference is illus-
trated in Figure 2-10.
With certain types of venturi scrubbers, there is static pressure
recovery in the divergent section; in other words, the pressure drop
across the throat is greater than the pressure drop across the entire
venturi section (convergent section - throat - divergent section). This
occurs in units where the divergent section expansion angle is equal to
2-18
-------
Figure 2-9. Parallel scrubber trains.
Figure 2-10. Parallel scrubber trains with unequal duct resistance.
or less than 7.5 degrees as shown in Figure 2-11. With units of this
type, boundry layer separation probably does not occur.H Pressure
recovery can also occur in scrubbers with a slightly greater expansion
angle; however, the extent of the gain would be slight. Most commercial
designs have a divergent angle ($ ) of greater than 25° and many are in
the 35° to 45° range. A few types do not even include a divergent
section to the venturi; pressure recovery in these would be negligible.
In order to properly define the pressure drop, it is very important
to specify the locations where the "inlet" and "outlet" measurements are
made. A sketch of the system showing the scrubber, all major internal
parts, and flow restrictions of the inlet and outlet ductwork is helpful
in avoiding misinterpretation of the pressure drop value.
2-19
-------
Figure 2-11. Converging and diverging venturi angles.
2.3.2 Relationship Between Penetration and Pressure Drop
There is a useful relationship between the collection efficiency of
a particulate wet scrubber and the pressure drop of the gas stream passing
through it. For this reason, operators have been using pressure drop as
an indicator of scrubber performance for a long time.
Tn 1956, Kemack and Lapple proposed that penetration is a function
of the total energy consumption within the scrubber, not just the energy
loss represented by the gas phase pressure drop.12 This theory was
expanded and stated in its present form by Semrau in I960.13 (See also
references 14-16.) Referred to as the Contact Power Theory, it has
enjoyed broad acceptance by both operators and regulatory agencies.
Since much of the available data are presented in the Contact Power format,
the basis of this theory is briefly examined.
According to the Contact Power Theory, scrubber efficiency is directly
related to the total energy consumption of the system. The power can be
expended by the gas stream, the liquid stream, by a mechanical rotor, or
by a combination of all three. Other variables such as collector size,
scrubber design characteristics, liquid-to-gas ratio, liquor surface
tension, and gas velocity are assumed to have no independent effect on
the scrubber performance. Accordingly, it should be possible to quantify
the emissions from a specific wet scrubber system using only those para-
meters with an effect on the power input.
2-20
-------
The Contact Power Theory has also been used to compile industry-by-
industry mass emission-contact power curves. In this section, the adequacy
of these unit specific and the industry wide applications of the theory
are evaluated .
The Contact Power equations commonly presented in the literature are
reproduced below. These apply to all particulate wet scrubbers.*
PQ = 0.158 AP
PL = 0.583(QL/QG)
PT = PG + PL
H = 1 - e
P = 1 - ( i?/100)
Equation 2-6
Equation 2-7
Equation 2-8
Equation 2-9
Equation 2-10
Equation 2-11
where:
AP
P
77
a,
Pt
PL
PQ
gas phase pressure drop, inches w.c.
penetration
collection efficiency, percent
number of transfer units, dimension! ess
empirical constants, dimension! ess
total power input, hp/1000 ACF
power input from liquid stream, hp/1000 ACF
power input from gas stream, hp/1000 ACF
liquid-to-gas ratio, gal/1000 ACF
The constants a and 0 are functions of the particle size and particle
characteristics. Equations 2-10 and 2-11 simply indicate the relationship
between transfer units, efficiency, and penetration.
The derivation of these equations, which was only briefly described
by Semrau,14 was apparently based on a mechanical energy balance across
the scrubber. To fully evaluate the adequacy of these equations and the
means to apply the Contact Power approach, the derivation was reconstructed.
It is presented below. The starting point is the basic Bernoulli Equation
for incompressible flow.
*An additional term for shaft power would be required if a mechanical
rotor were included in the scrubber.
2-21
-------
P! - P21 g F 1 faiVi2 - a?Vo2]
_L £ G + — Zi - Z2 G + P-i ±J-
PG J ScL1 2J I 2 gc J
Wp -
L +
V 2
PVP
2V2
2 gc
L +
G = 0
Equation 2-12
*)
2gc
Equation 2-14
2-22
-------
p"p
2gc
Equation 2-15
Equation 2-16
PT =
Equation 2-17
Since the liquor pressure is zero psig, AP|_ is equivalent to P|_.
PT APG
(L/G)
Equation 2-18
The left hand term, Py/G is identical to the contact power form
of horsepower per 1000 ACF. To convert the gas phase energy term,
PG/PG» to tne fom of Equation 2-6, the gas density at 70°F, 14.7 psia
must be inserted. This reconstructed derivation is intended to demonstrate
a very important point, namely that total energy input is a function of
gas density. Too often Equation 2-6 has been used without questioning
the applicability of the "constant," 0.158. The actual gas density can
vary substantially as a function of the inlet gas temperature, the scrub-
bing liquor inlet temperature, the liquid-to-gas ratio, and the outlet
static pressure^
As stated earlier, the form of the Bernoulli Equation used in the
derivation of the Contact Power equation applies only to incompressible
flow. Since it is conceivable that the gas density varies more than 10%
while passing through some types of scrubbers, these equations would not
2-23
-------
be strictly applicable.17 Gas density data at various humidities, tem-
peratures, and static pressures are presented in the Appendix A to
illustrate the degree of variability possible.
The presssure drop term in Equation 2-18 applies only to the throat
in a venturi scrubber. If there is some degree of pressure recovery in the
divergent section of the venturi, the observed pressure drop will be lower
than that used in this equation.
Despite the theoretical limits of the Contact Power Theory, there
appears to be a number of cases in which it provides reasonable cor-
relations between performance and energy input. A number of correlations
originally compiled by Semrau are provided in Figures 2-12, 2-13, 2-14,
and 2-15.13 The gas phase pressure drop is probably the dominant energy
loss in these correlations.
6.0
4.0
.3.0
t 2.0
6
z
cr
K
CO
u.
o
DC
UJ
m
S
1.0
0.8
0.6
0.4
0.3
0.2
SPRAY NOZZLE
LOCATION
O CYCLONE
• CYCLONE
A INLET DUCT
I
SPRAY
DIRECTION
DOWNWARD OVER INLET
HORIZONTAL, ACROSS INLET
COUNTER CURRENT
I i i i i i i
_L
J_
111!
0.3 O.4 0.6 0.8 1.0 2.0 3.0 4.0
CONTACTING POWER, HP/(IOOOcuft/min)
6.0 8.0
Figure 2-12. Contact power curve for a cyclone scrubber controlling talc dust.
2-24
-------
All of these data are from very early studies and were obtained with
test methods that are not as accurate as current methods. However, the
data do support the contention that emissions corelations can be devel-
oped for a group of similar sources using only the power input and possibly
just the pressure drop.
Figure 2-15 is presented specifically to demonstrate that the liquid-
to-gas ratio does not have any independent effect.13 Semrau qualifies
the statement, however, by suggesting that some effect may occur at very
low liquid-to-gas ratios of 1 to 3 gallons per 1000 ACFM.18 No data
in this range was apparently available when the data in Figure 2-15 was
originally compiled. ,
8
7
tr
u.
o
OL
UJ
00
AND MELTING
SCRAP MELTING, HOT METAL WORKING
WORKING
2 3 45 6 7 8 9 10
CONTACTING POWER, HP/dOOO cu ft/sec)
Figure 2-13. Contact power curve for venturi scrubbers on open hearths.13
2-25
-------
3.0
H-
Z
ff
Z
u
2.0
0.8
0.6
I
£ 0.4
O 03
£C '
i 0.2
z —
01
I I I
-D
I I I I I I I I
IITIIIFT
cfia
• VENTURI SCRUBBER REF.
D CYCLONIC SPRAY SCRUBBER
I I
I I I I I I
l tit
0.06 0.1 02 0.3 04 0.6 0.8 1.0 2.0 3.0 4.0 6.0 8.0 100
CONTACTING POWER, HP/(IOOO cu ft/min)
Figure 2-14. Contact power curve scrubbers controlling ferrosilicon
furnace fume.13
IO.O
< 40
DC
H-
DC
2.0
1.0
I I I I M I
I I I I I L
I I I I i I I I I
LEGEND
(L/G,gal/IOOOft3)
olO
A 15
•20
I I 1 I I I I I
4 6 8 10 20
EFFECTIVE FRICTION LOSS, inches
40 60 80
Figure 2-15.
Contact power curve for venturi scrubbers illustrating
lack of independent effect of the liquid-to-gas ratio. 18
2-26
-------
Walker and Hall19 have compiled a contact power correlation for
flooded disc venturi scrubbers serving four commercial lime kilns. Data
from a pilot plant were also included in the data set. The resulting
emissions-power input correlation presented in Figure 2-16 suggests a
linear relationship between the pressure drop (the only significant energy
input is the gas phase) and the measured emissions. To further evaluate
its usefulness, the curve has been replotted (see Figure 2-17) into a
more conventional format and the confidence interval prepared (using the
procedures presented in Appendix B).
T 1 1 1—
• PLANT A
A PLANTS
A PLANTC
O PLANTD
• PILOT PLANT
20
Q.
O
O 10
LU
ir
Z>
en
UJ
£ 5
T 1
A A
j.
J L
0.02
0.05 0.10 0.20
EMISSIONS, grains/SCF
0.50
Figure 2-16.
Relationship between pressure drop and emissions for
flooded disc scrubbers on lime kilns.
2-27
-------
It is readily apparent in the modified curve of Figure 2-17 that
there is considerable scatter in this correlation. For example, at a
pressure drop of 9 inches, the emissions can vary from 0.075 grain per.
ACF to as high as 0.20 grain per ACF. Obviously, the degree of scatter
would preclude application of this type of curve for the purposes of
evaluating compliance with an emission standard. During the preparation
of this curve, it was not possible to determine whether individual site-
specific correlations would have less scatter, because the raw data was
not included in reference 19.
o.io
co" 0.08
g 0.06
tn
| 0.04
LU
0.01
Figure.2-17.
I 2 3 10 20 30
PRESSURE DROP, inches
Replotted Walker and Hall data; 90% confidence interval is
denoted by shaded area.
Kashdan and Ranade have summarized emissions data and contact power
levels (uncorrected for saturated gas density) for pulverized coal and
lignite-fired utility boilers 20. The modest degree of scatter in
Figure 2-18 is due in part to the inclusion of data from various types of
scrubber systems incuding the Krebs Preformed scrubber, the UOP Turbulent
2-28
-------
Contact Absorber, and several major types of venturi scrubbers. Considering
the diversity of scrubber designs and the major differences in the fuel
characteristics, it is remarkable that any correlation is apparent.
2 3 4 56789
THEORETICAL POWER CONSUMPTION, hp/1000 ACFM
Figure 2-18. Pressure drop versus emission for wet scrubbers serving coal-
fired utility boilers. 20
A subset of Kashdan and Ranade's data was used to prepare a contact
power curve for power plants using only venturi-type scrubbers (see
Figure 2-19). It is apparent that the scatter is substantially reduced.
The range of variables for this subset of data is presented along with
the curve to demonstrate that even with this small data group, there is
considerable variation. Additional testing and evaluation is necessary
to confirm that a curve representing all types of coal-fired boilers with
venturi scrubbers can be prepared. Data from Kansas City Power's LaCygne
Station are considerbly higher than the other points.
2-29
-------
_
o
v>
.050
.045
.040
.035
.030
I
co~ .025,
O
CO
CO
5 .020
tu
.015
5 10 15 20 25
PRESSURE DROP, inches
Figure 2-19. Replotted data of Kashdan and Rariade using only scrubbers;
90% confidence interval is denoted by shaded area.
Genoble, et al. have collected a number of emission-contact power
correlations, two of which are shown in Figures 2-20 and 2-21.I6 In both
of these cases, there is considerable scatter in the data. The scatter
in the data set for the coal preparation plant thermal dryers has been
partially attributed to differences introduced by the various organizations
conducting these tests.
2-30
-------
.20
.
cn"
z
g
en
en
LU
.02
10 20
PRESSURE DROP, inches
Figure 2-20.
Pressure drop versus emissions data for venturi scrubbers
serving asphalt plants.16
The data scatter in the asphalt concrete plant curve, the thermal
dryer curve, and many of the curves presented earlier, is evidence of the
weakness of industry wide correlations. These correlations inherently
cannot take into account the numerous important site-specific variables
which include, but are not limited to, particle size distribution, particle
size distribution, particle surface characteristics, liquid-to-gas ratio,
gas-liquor maldistribution, and liquor droplet size. Correlations prepared
on a site-specific basis should be less vulnerable to these variables and
hence, should be more useful for evaluating scrubber performance.
2-31
-------
.08
.07
.06
Q
.04
to" .03
z
o
en
CO
UJ
.02
.01
10
20 30 40 50 607080
PRESSURE DROP, inches
Figure 2-21. Pressure drop versus emissions data for venturi scrubbers
serving coal thermal dryers.16
Site-specific data are very limited since few operating wet scrubber
systems have been tested enough times for an emissions - contact power
correlation to be prepared. Available data are usually limited to pilot
scale equipment.
Pilot plant wet scrubber data for the Minnesota Power and Light
system have been reported by Johnson and also by Nixon and Johnson.21
The scrubber tested handled a 3200 ACFM slip stream from the Clay Boswell
Station, Unit No. 3. Emission tests conducted during periods when two
different coal supplies were being burned are summarized in Figure 2-22.
The data for the scrubber are very consistent, despite the slight differ-
ences apparently introduced by the coal supplies.
2-32
-------
o
0.04
0.03
o
en
in
c
s
o>
cT
5
<
3
? 0.02
<
cc
O.Oi
MCKAY COAL
ROSEBUD COAL
_L
_L
_L
-L
5 IO 15 2O
VENTURI PRESSURE DROP, inches water
25
Figure 2-22. Relationship between pressure drop and emissions for a pilot
scale venturi scrubber on a coal-fired boiler slip stream.
A contact power relationship for a full scale Q-Basic .Oxygen Process
(Q-BOP) furnace is presented in Figure 2-23. It was prepared from a
tabulated data set provided by Kemner and Mcllvaine.22 The data are for
three identical vessels controlled by adjustable throat Baumco venturies
in series with venturi prequenchers.
The values shown on Figure 2-23 represent the average pressure drops
during the period of 5 to 11 minutes into the oxygen blow. The pressure
2-33
-------
drop variation during cycle is due to the rate of gas evolution. The
pressure drop values have not been corrected for gas density since the
necessary temperature data were not included.
The correlation between the average pressure drop and the outlet
dust loading has very little scatter. This may be partially due to the
identical design and operating conditions of the three vessels. A site-
specific relationship of this type would be adequate for an emissions
correlation. The relationship, however, wuold probably not be applicable
to other basic oxygen furnaces because of the significant differences
made by the various types of hooding and process conditions on the particle
characteristics.
Since performance is affected by so many variables, from a number of
different scrubbers should not be grouped into a single correlation. It
would be preferable to use the contact power approach only on a site-
specific basis. Even in this case, it would be necessary to confirm that
there have been no major operating changes that would invalidate the mass
emission-power correlation. •,
o
u.
o
.04
1 .03
s
o>
CD"
§.02
CD
CD
J I L
60 62 64 66 68 70
PRESSURE DROP, inches (minutes 5-11)
Figure 2-23.
Pressure drop versus emissions data for venturi scrubbers
serving three Q-BOP furnaces ( plotted using data of
Kemner and Mcllvaine).22
2-34
-------
2.4 PLUME AND BREECHING OPACITIES AS PERFORMANCE PARAMETERS
Regulatory agencies and control equipment operators routinely use
plume opacity as an indicator of the particulate matter removal performance.
In the case of wet scrubbers, however, the use of opacity has been limited
by the practical problems created by the condensation of water droplets
i n the piume.
Theoretically, opacity could be a meaningful operating indicator
since particles that scatter visible light most effectively are in the
range of 0.2 to 2 micrometers and this coincides with the peak penetration
range for the particles treated in a wet scrubber system (as was shown in
Figure 2-6.) Thus, as the penetration through the scrubber system in-
creases in this important size range, the opacity should also increase
substantially.
Mass emission-opacity correlations have rarely been developed based
on manual observations. This is partially due to the lack of sufficient
emission test data and partially due to the difficulty in making these
observations. For example, in compiling these correlations it is impor-
tant that the observational path length through the plume remain the same.
This is possible only when the condensation of water droplets occurs at
a point downstream of the stack. Under these somewhat unusual conditions,
it is possible to make the observation of opacity at a point directly
above the stack, where tne path length is known. Observation of the
residual opacity of a plume downwind of the stack is not as useful since
the path length at this point can be highly variable due to meteorological
conditions and dilution of the plume becomes a factor.
As early as 1963, an extractive instrument was tested as a means to
evaluate the stack opacity without interference from water condensation.23
The device consisted of a photocell mounted in a flow-through gas cell.
The source of light was a simple flashlight bulb. The sample was extract-
ed through a heated sample line to prevent condensation, and the instrument
was calibrated using neutral density filters. While successful performance
was claimed, it is doubtful that this instrument could satisfy the specifi-
cation requirements of present day transmissometers.
2-35
-------
In some cases, a conventional transmissometer has been successfully
utilized as a continuous monitor. Very often condensation of water drop-
lets downsteam of the scrubber does not occur in the ductwork. Therefore,
a transmissometer installed in these locations should perform satisfactor-
ily. A study by Nixon and Johnson at the Minnesota Power and Light plant
using a pilot scale scrubber system included the evaluation of the plume
opacity as a function of the mass concentration in the effluent.21 The
opacity was determined using a Lear Siegler RM41 transmissometer mounted
along a 6'6" section of the outlet duct. As seen in Figure 2-24, the study
showed a linear relationship between the mass concentration and the opacity.
This supports the logical presumption that opacity should be a useful indi-
cator of mass emissions when the opacity can be accurately determined.
More recent test work has been done using a heated extractive
sample line similar in concept to this early instrument.24 The opacity
monitor in this case, however, is a Nephleometer which measures light
scatter. Test work done at a kraft pulp mill recovery boiler equipped
30
20
Q x
w fc
05 O
GQ UI
ll'l
^8_ 8
6
5
o to
0. O
I
I
I
I
I
.005 .010 .015 .020 .025 .030
OUTLET GRAIN LOADING, grains/SCFD
.035
.040
Figure 2-24. Opacity versus emissions data for a venturi scrubber on a
coal-fired boiler slip stream.2^
2-36
-------
with a venturi scrubber was encouraging. With certain refinements, this
approach may soon provide a convenient means to continuously monitor the
performance of wet scrubbers.
2.5 LIQUID-TO-GAS RATIOS AS PERFORMANCE PARAMETERS
The liquid-to-gas ratio has a significant impact on the penetration
through a wet scrubber. However, it is questionable whether the effect of
the liquid-to-gas ratio1needs to be considered independently of the
pressure drop, because, for most types of wet scrubbers, the pressure
drop is linearly related to the liquid-to-gas ratio.7»8
The possible effect of the liquid-to-gas ratio has been examined
using the venturi scrubber analytical model discussed earlier. The pene-
tration was calculated for a number of combinations of gas velocities and
liquid-to-gas ratios. Then the penetration and corresponding pressure
drop data was plotted using the liquid-to-gas ratio as a parameter. The
results are illustrated in Figure 2-25.
O.IO
O.O8
0.06
ui
Q_
0.04
0.02
LIQUID-TO-GAS RATIO,
gals/IOOOACF
20 4O 60 80 100
PRESSURE DROP, centimeters
120
Figure 2-25.
Predicted effect of liquid-to-gas ratio on penetration as
calculated using the venturi scrubber model.
2-37
-------
Based on the model predictions, the optimum liquid-to-gas ratio is
between 8 and 12 gallons per thousand ACFM. At this rate, the penetration
is lowest at a given pressure drop as illustrated by the dotted line at
40 centimeters of water. There is only a slight difference in performance
as the liquid-to-gas ratio varies over the range of 6 to 20 gallons per
thousand ACFM. The model suggests that there is essentially no independent
effect over this range, which is the most common operating range of com-
mercial wet scrubber systems.
At very low liquid-to-gas ratios, the venturi scrubber model predicts .
that there is an effect which is independent of the pressure drop. The
4 gallon per thousand ACFM curve is significantly higher than the other
liquid-to-gas ratio curves. This prediction is consistent with the
observations of Semrau and others during the testing of operating systems.^
At low liquid-to-gas ratios, it would be necessary to monitor the liquid
flow rate since a change of plus or minus 1 gallon per thousand ACFM
would have a major impact on scrubber performance.
As normally used, the liquid-to-gas ratio applies to the entire scrub-
ber. In other words, the liquid-to-gas ratio is the total recirculation
liquor flow divided by the total scrubber outlet gas flow rate. Unfortu-
nately, the distribution of the liquor across the gas stream is not entirely
uniform; therefore, there can be significant variation in the local liquid-
to-gas ratios. This variation can have a significant impact on the penetra-
tion without a proportional impact on the pressure drop. It is very diffi-
cult to identify the maldistribution problems.
Poor gas-liquid distribution is often the result of improper treat-
ment of the recirculation liquor. Excessive suspended solids can lead to
pluggage of the spray nozzles, the main liquor pipes, and the strainer
before the pump. In some spray tower scrubbers, as many as half of the
nozzles have been found to be plugged by suspended solids in the liquor.
In venturi scrubbers and traytype scrubbers, scaling of the scrubber
internals can also result in serious gas-liquor maldistribution. To
minimize scaling, it is important to maintain a proper pH (generally
below 10) and a proper purge/blowdown rate.
Buildup of material in the wet-dry interface can lead to maldistri-
bution, too. To prevent buildup of material in this location, the gas
approach must be wetted.
2-38
-------
2.6 EFFECT OF LIQUOR SURFACE TENSION
The surface tension of the scrubbing liquor can potentially affect
the penetration rate whenever impaction on droplets is an important unit
mechanism. This is true for gas atomized scrubbers such as Venturis and
orifice units, impingement plate and sieve plate scrubbers, mechanically
aided scrubbers, and spray tower scrubbers. Low surface tension reduces
the liquor droplet size and improves coalescence (the inclusion of the
particle into the droplet). In one study by Hesketh, it was found that
lowering the surface tension from 50-60 dynes per centimeter to 10 dynes
per centimeter reduced the penetration by 50%.25
The impact of reduced droplet size can be predicted using the ven-
turi scrubber analytical model developed by Calvert et alJ The data
presented in Table 2-2 was computed by Woffinden, Markowski, and Ensor^^
using the Calvert Model and the conditions of an 8000 cm/sec throat
velocity and a monodisperse aerosol of 1 /imA. The model suggests no
improvement if the droplet diameter without surfactant is below approxi-
mately 50 micrometers. This is not a commercially important case since
the large majority of nozzles used in wet scrubbers produce sprays in
the 200 to 800 micrometer range. 27,28
TABLE 2-2. PREDICTED EFFECT OF SURFACE TENSION (a) ON THE PENETRATION
THROUGH A VENTURI SCRUBBER 26
Droplet
a= 72 dynes/cm
205
144
51
Diameter
a= 30 dynes/cm
144
102
36
Penetration
a= 72 dynes/cm a= 30
0.020
6.020
0.020
dynes/cm
0.011
0.013
0.021
The effect of improved particle coalescence into a droplet is diffi-
cult to estimate. Presumably this would benefit sources with marginally
wettable particulate ;matter.
Liquor surface tension will vary according to the quantities of mater-
ials absorbed from the gas stream, the quantities of surfactant added, and
2-39
-------
the rates of the purge and makeup streams. Chemicals added to the liquor
to improve the rate of settling of solids in the scrubber clarifier and
recirculation tank can have an impact opposite to that for the surfactant.
The influence of foaming suppressants on the surface tension of the liquor
is not known. Due to the potential variability of the surface tension and
the moderate effect this may have on scrubber performance, the quantities
of all surfactants, flocculants, and anti-foam agents should be recorded
during each stack test and during subsequent evaluation periods.
If it is necessary to quantify the surface tension, ASTM Method
D 1590-60 can be used.29 This technique is based on the ring method of
measurement and has a precision of 0.3 dyne per centimeter.29 Most com-
mercial laboratories have the necessary equipment for this test procedure.
2.7 CONDENSATION AND EVAPORATION EFFECTS ON PERFORMANCE
Impaction can be enhanced by the condensation of water vapor on the
surfaces of particles and, to a lesser degree, by the condensation of water
vapor on existing water droplets. Evaporation of water vapor from existing
droplets has an adverse effect on scrubber performance.
The beneficial effect of condensation is primarily due to the in-
creased mass present on a particle. The increased inertia of the particle
yields higher impaction efficiency. A secondary benefit results from
the mass gradiant which exists whenever condensation is occurring. The
imbalance in the momentum imparted by the water vapor molecules on each
side of the particle results in a modest movement toward the surface of
the condensation site.30 This phenomenon is termed diffusiophoresis. The
negative effect of evaporation is the opposite of that described above.
The quantity of water available for the inadvertent condensation is
usually quite limited. This can be estimated by use of psychrometric
chart (actual physical conditions are dependent on the prevailing gas
pressures). It is usually in the range of 0.01 to 0.15 pound of water
per pound of dry air.
Calvert and Jhaveri have presented information indicating that the
quantities necessary to achieve substantial improvement in scrubber per-
formance are in the range of 0.2 to 1.0 pound of water per pound of dry
air.31 This is illustrated in Figure 2-26.
2-40
-------
0.4
0.2
O.I
_L
I
I
O.I 0.2 O.3 0.4 0.5
GRAMS H20 CONDENSED PER GRAM DRY AIR
Figure 2-26. Beneficial effect of condensation.31
There are few commercially operating scrubbers which have a quantity
of water vapor condensing that is this large. In order to induce this phe-
nomenon, it is possible to evaporate a large quantity of water in a high
temperature zone of the process duct work or to inject large quantities of
low pressure steam. Cooling of the inlet liquor stream would also favor
condensation. Lemon has presented operating data on a novel scrubber system
which used artificial means to induce condensation.32 This system was also
able to achieve compliance with all applicable requirements at a pressure
drop of only 20 inches w.c., while the earlier scrubber on this source re-
quired a pressure drop of 40 inches w.c. for a similar penetration rate.
At the moderate water vapor levels present in most effluent streams,
the presence of condensation would have little effect. The plume appear-
ance would provide an initial indication of a dramatic change in the water
vapor content of the efflue'nt stream. Other parameters which would be
useful indicators include the gas stream inlet temperature and the
liquor stream inlet temperature.
2-41
-------
-------
3.0 FACTORS AFFECTING SCRUBBER SYSTEM RELIABILITY
The performance of a scrubber system is dependent on the operating
characteristics discussed in Section 2.0 and on the integrity of the
scrubber internals. The latter is addressed in this section. Failure
of these components can lead to extended downtime of the system or
operation in temporary noncompliance. Both the operator and the regu-
latory agency inspector have an interest in minimizing these problems
by identifying the emergence of factors which threaten scrubber compo-
nents and ultimately degrade scrubber performance.
3.1 CORROSION AND EROSION OF SCRUBBER SHELL
The susceptibility of a scrubber to erosion is a direct function
of the gas velocities within the entry ductwork and the scrubber shell.
The venturi scrubber throat is most prone to these problems due to the
extreme gas velocities of 20,000 to 40,000 feet per minute at this
point. Other areas of common erosion failure are illustrated in
Figure 3-1.
Regulatory agency inspectors usually do not have the opportunity
or safety equipment to perform internal inspections of control equipment,
therefore, they must use the less direct, external symptoms of erosion.
This includes the obvious holes worn through the shell at critical
high velocity points or points of gas stream turning. In extreme cases,
these holes can lead to severe inleakage of ambient air and a reduction
of the quantity of gas pulled from the process.. The extent of the air
infiltration can be quantified using pi tot tubes before and after the
collector (and accounting for absorption and condensation/evaporation
within the scrubber).
A valuable sign of potential erosion problems is the quantity of
suspended solids in the liquor. The potential for erosion is directly
related to the percent suspended solids. The performance of the clari-
fier and/or the settling pond can also have an impact on the susceptibil-
ity to erosion in that the large, abrasive particles should settle out.
The greater the recirculation flow rate relative to the make-up and purge
flows, the greater the potential for build-up of high suspended solids.
3-1
-------
Figure 3-1. Areas of a venturi scrubber vulnerable to erosion.
Also, systems which purge only once per day or once per week can develop
high suspended solids levels late in the operating period. A rough
check on the quantity and character of the suspended solids can be made
by obtaining a sample of the liquor leaving the scrubber sump and
observing the initial turbidity and the rate of settling.
Due to the absorption of corrosive materials such as sulfur dioxide,
sulfur trioxide, hydrogen chloride, and hydrogen fluoride, corrosion of
the scrubber shell and ancillary equipment is a common problem. It is
3-2
-------
important to maintain the pH of the liquor well above the levels at which
carbon steel (the most common material used in scrubber construction) is
attacked. A pH of 6 or greater is usually satisfactory. An appropriate
pH is usually maintained using alkaline additives such as soda ash, lime
or limestone. The performance of the alkaline additive delivery system
and the process control instrumentation are very important in minimizing
short term low pH excursions. The operation of the pH monitor used to
control the rate of additive injection should be checked on at least a
daily basis unless long term operating experience justifies less frequent
inspection. A portable pH meter or pH paper can be used for this check.
The sample should be collected at the sump of the scrubber since this is
often the point of minimum pH.
The recirculation rate is an important factor in determining the
extent to which halogenated compounds are building up in the scrubbing
liquor. Relatively small levels of chlorides and fluorides can attack
many types of materials, especially the 300 series stainless steels.
One convenient means to monitor potential corrosion problems is to
prepare small coupons (small circular samples) of the various materials
used throughout the scrubber system. These are placed in racks which
can be mounted at various locations in the scrubber. During every
outage these are visually inspected for pitting and cracking and are
weighed for material loss. This information provides an early indication
of developing corrosion problems.
3.2 EROSION AND PLUGGAGE OF SPRAY NOZZLES
Spray nozzles are extremely susceptible to erosion and pluggage
problems due to the high velocities of the liquid stream and due to the
suspended solids within the stream. The most common types of nozzles
in use include the hollow cone and the full cone. The latter is particu-
larly prone to pluggage due to the presence of an internal spinner vane.
The vane is installed to achieve the full cone spray pattern which is
necessary for distribution of liquor on a packed bed.
Damage to the nozzles can sometimes be determined by observing the
spray angle while the nozzles are operating at normal line pressures.
3-3
-------
(This can usually be done safely while the system is down and only the
pumps are operating.) If the spray angle is considerably greater than pre-
viously, then it is very possible that the nozzle orifice has been enlarged
by erosion. If the spray angle has decreased or if a distinct spray pat-
tern is no longer achieved, it is likely that the nozzle is partially or com-
pletely plugged. Both conditions lead to severe gas-liquid maldistribution.
Erosion problems on the tangential inlet to the entrainment separa-
\
tor are common to many types of scrubbers because of the concentration
of large, abrasive particulates in the outer part of the inlet duct. The
venturi-rod type scrubber is also prone to erosion of the rods within the
throat, hence these must be checked on a routine basis.
Venturi scrubbers in which the throat is oriented vertically may have
erosion/abrasion failure at the bottom of the divergent section as the gas
stream turns 90° to enter the entrainment separator. One means of minimiz-
ing this problem is to include, below the venturi throat, a flooded elbow
containing 6" to 9" of trapped liquor.
Erosion can best be identified during a routine internal inspection
of the scrubber shell, entrainment separator, entry ductwork, and other
such components. The erosion of corrosion protective linings such as
rubber or synthetic coatings is particularly important since gaps and
crevices in the lining can lead to rapid corrosion underneath the lining.
The entire lining can be defeated by erosion of a small area in an ero-
sion prone zone of the scrubber.
The visual check of spray angle usually cannot be done by a regula-
tory agency inspector or an operator while the entire system is operating.
Some indirect indications of the nozzle condition can be used in lieu of
these observations. If a nozzle has plugged significantly, the pipe
leading to that nozzle in the scrubber begins to cool. Comparision of
the skin temperature of one pipe against that of the pipes leading to
other nozzles is a reasonably reliable indicator of pluggage. It is
common to find a skin temperature 5 to 10 degrees below that of adjacent
pipes. Pluggage of a nozzle will also be indicated by an increase in
the line pressure near the nozzle. Erosion of the nozzle orifice yields
the opposite conditions. Unfortunately, line pressure is difficult to
interpret since it is strongly dependent on the liquid flow rate. The
3-4
-------
relationship between pressure and flow rate (for a nozzle in original
condition) is presented in Equation 4-1.28
_Ll_
t-2
Equation 3-1
where:
LI
l_2
P!
?2
= liquid flow rate at pressure 1
= liquid flow rate at pressure 2
— rvv»Q e* o 11 v»a 1
pressure 1
pressure 2
During a routine inspection it is difficult to determine whether a
change in line pressure is due to a change in the internal condition of
the nozzle or simply due to a change in the liquid flow rate.
If possible, all nozzles for venturi scrubbers should have external
rod-out capability. This is usually not feasible for spray tower scrubbers
and the nozzles used to clean demisters. In these cases, an automatic
flush system may be of benefit. If a pond is used for settling, it is ad-
visible that it have several zones separated by weirs and the pick-up for
the pump be near the water's surface to prevent entrainment of silt.
3.3 FANS
Fans can either be installed before or after the scrubber. Only
radial blade designs are used ahead of the scrubber due to the large
quantities of suspended particulate in the gas stream and the fact that
other fan wheel designs are prone to fan wheel particulate buildup.
Fans installed following the scrubber should be the radial blade type.
The principal problems of fans on wet scrubber systems include fan
wheel erosion, fan wheel buildup, and bearing failure.33 Corrosion is
also a common problem. Both fan wheel erosion and fan wheel buildup
will eventually result in increased vibration which is usually audibly
and visually detectable. For systems prone to this problem, vibration
sensors are advisable, enabling a source to have the fan automatically
tripped if the vibration reaches undesirable levels. If during a routine
inspection of a scrubber system, excessive fan vibration is noticed,
extreme caution and immediate action is necessary. Disintegrating fans
can fling metal parts over a wide area.
3-5
-------
Fan bearings should be inspected on a frequent basis for oil level,
oil color, oil temperature, and vibration. Bearings found to be inade-
quately lubricated may require a greater frequency of lubrication or the
installation of a forced lubrication system. Excessive bearing wear may
be caused by a higher than originally specified fan operating temperature
or by misalignment of the bearing mountings.
Fan belts should be checked periodically for wear. At the time of
installation, belts should be checked for proper tension. Loose belts
can cause the entire dust collector system to malfunction. Out-of-line
sheaves will destroy belts; uneven wear on grooves of sheaves and the
surfaces of the belts indicates misalignment.
Fan vibration can sometimes be caused by air flow factors and, in
these cases, can be eliminated by adjusting the inlet or outlet dampers,
modifying the inlet or outlet dampers, modifying the inlet and/or outlet
ductwork, or changing the sheaves. The latter will have a direct affect
on the gas flow rate and will result in a change in the tip speed (which
should never exceed the manufacturer's recommended rate).
Operating problems experienced by fans downstream of scrubbers are
generally only symptoms of more fundamental problems with the particle
capture and/or entrainment separation.
3.4 PUMPS
The two types of pumps in service on wet scrubber systems are the
centrifugal and positive displacement pumps. The centrifugal pump is
the most common.
Accelerated wear of the pump impeller occurs at high suspended sol-
ids levels. As is the case with spray nozzles, the best way to minimize
scrubber downtime due to pump malfunctions, is to minimize the total
quantity of suspended solids and especially the larger suspended particles
(greater than 25 micrometers). A strainer ahead of the pump will improve
the quality of liquor passing through and thereby minimize wear.
Another common problem with pumps is leakage through the packing
glands; pumps must be repacked whenever necessary. This can be mini-
mized through the use of doub.le seals with a clean water, pressurizing
line. All bearings on the pumps and the pump couplings should be lubricat-
ed periodically in accordance with the manufacturer's instructions.
3-6
-------
3.5 PIPING AND VALVES
All piping is subject to erosion, especially at elbows. The rate
of erosion is a function of the flow velocity, the quantity of suspended
solids, the particle size distribution of the suspended solids, and the
presence of corrosive agents. The proper materials of construction
must be selected in order to minimize this problem.
Valves are also subject to erosion. Those which are operated in
either a fully open or closed position are affected much less than those
use.d in a partially open mode for flow throttling purposes.
In order to reduce pluggage of lines during outages, the piping
should be completely drained. This requires that each line have a modest
slope so that a pocket of liquid cannot remain after draining. The drain
should be at the lowest point of each piping system. Draining is also
necessary to prevent the freezing of lines during outages.
During routine inspections, one should be aware of the piping design
with regard to drainage capability. If it appears inadequate, and the
systems runs in a cyclic nature, occasional pluggage and/or freezing prob-
lems should be anticipated. Periods of noncompliance can occur initially
after start-up of the system and prior to the rectification of a piping
freezing/pluggage problem. In such cases, the pump discharge pressure
will be high, the nozzle pressure will be low, and the gas exit tempera-
ture from the scrubber will be high. The gas phase pressure drop across
the scrubber will be low due to the lack of typical liquor flow rates.
3.6 DUCTS
The most common problems with the ducts leading to a scrubber include
dust buildup, erosion/abrasion, flex failure, and failure of expansion
joints. The last three of these conditions result in air infiltration
and consequently, a reduction in the quantity of gas pulled from the
process equipment. The extent of air infiltration can be quantified by
preparing an oxygen profile at various points along the duct from the
source to the scrubber. . ,
Dust buildup occurs in low gas velocity zones of the duct work.
This can occur wherever the duct work is oversize for the effluent gas
flow rate and during periods of operation at reduced process rates. Clean
3-7
-------
out ports should be located at convenient locations along the ductwork '
to facilitate occassional removal of accumulated material. Without this
cleaning the duct work can collapse (unless the supports have been de-
signed to take the combined weight of the duct and the deposits).
3-8
-------
4.0 SCRUBBER INSPECTION AND PERFORMANCE EVALUATION
The purpose of a wet scrubber performance evaluation is to determine
whether the penetration has increased or decreased since the previous eval-
uation and to identify any conditions which threaten the proper operation
of the scrubber system. The performance evaluation must incorporate the
operating variables and factors presented earlier in Sections 2.0 and 3.0.
A brief summary of these are presented below in Table 4-1.
TABLE 4-1. PERFORMANCE EVALUATION PARAMETERS
Penetration Variables
Inlet and Outlet Static Pressure
Demister Pressure Drop
Bypass Duct Pressure Drop
Gas Inlet and Outlet Temperatures
Liquid Inlet and Outlet Temperatures
Gas Inlet 02 Levels
Gas Inlet COg Levels
Liquid Suspended Solids and Dissolved Solids
Liquid Surface Tension
Bypass Duct Temperature Profile
Rate of Liquor Evaporation and Condensation
Condensible Vapor Levels in Inlet Gas
Particle Size Distribution
Absolute Humidity of Inlet Gas
Operating Factors
Liquor pH
Liquor Suspended and Dissolved Solids
Liquor Supply Line Pressure
Nozzle Spray Pattern ,
Fan Vibration
Integrity of Ductwork
The list of variables is long and some of them cannot be measured or
observed directly. The performance evaluation approach must include some
indirect means to consider these difficult-to-measure factors.
The routine performance evaluation is structured to aquire only the
minimum data necessary to detect a change in wet scrubber performance.
When abnormal conditions are found, the remainder of the measurements and
other inspection procedures are performed to determine the nature of the
specific problem.
4-1
-------
4.1 BASELINE PERFORMANCE EVALUATION TECHNIQUE
The routine evaluation of wet scrubbers must be done in a logical
sequence in order to obtain all the necessary information and to minimize
the time requirements. The technique recommended is the Baseline Inspection
Technique.34 This approach is based on the assumption that the operating
characteristics and performance of each control system is unique. The
approach is designed to control for the myriad of process variables and
control device design factors, any one of which can singly or collectively
influence performance. ... , .
To ensure the accuracy of the performance evaluation, the Baseline
Inspection Technique utilizes a comparison of present operating conditions
with historical baseline levels for the unit. Each variable which has
shifted significantly is considered a "symptom" of possible operating prob-
lems. While compiling the operating data, the general condition of the
wet scrubber is observed to identify problems which could affect future
reliability of the system. The baseline data are compiled during periods
when the wet scrubber is operating well, preferably during a compliance
or acceptable stack test.
The evaluation is accomplished by proceeding in a counterflow manner
through the entire system; the stack and plume opacities are observed first,
and then the scrubber system is inspected component by component proceeding
backwards through the system in the direction opposite to the flow of the
gas stream. This is illustrated in Figure 4-1.
The time consuming inspection of process equipment and operating con-
ditions is done by the agency inspector only when, deemed-necessary based on
data collected during inspection of the scrubber and observation of the stack
opacity. The advantages of this approach include the following: (1) poten-
tially confidential process information is obtained only when absolutely
necessary, (2) the inconveniecnce to the operator is minimized, and (3) on-
site time is minimized. Since the inspection data are organized in a coherent
manner as the inspection proceeds, the effort can gradually be focused on
only the problems of direct interest. In other words, the intensity of the
evaluation can be minimized and many of the measurements can be avoided.
This is a particularly important point considering the extensive list of
operating variables and reliability factors presented earlier. The typical
order of inspection for a routine evaluation is presented in Table 4-2.
4-2
-------
CONTINUOUS
MONITOR
WASTES
CONCENTRATED
POLLUTANTS
Figure 4-1. Counterflow inspection sequence.
TABLE 4-2. ROUTINE INSPECTION"POINTS
1, Stack Discharge
2. Fan
3. Demi ster
4. Scrubber
5. Bypass Duct
6. Ductwork
7. Process
Plume Opacity at the Stack Discharge
Presence or Absence of Entrainment
Presence or Absence of Condensed Water
Presence or Absence of "Mud" Lip
Vibration (Qualitative Check)
Pressure Drop
Pressure Drop
pH
Liquor Line Skin Temperatures
Turbidity
Continuous or Interim'ttant Purge
Use of Anti-Foaming Agents and Surfactants
Damper Position
Holes and Leaks
Integrity of Expansion Joints
Hood Capture Effectiveness (Qualitative)
4-3
-------
Depending on the size of the scrubber system and the convenience of
the measurement ports, the routine inspection should take between 1 and 4
hours. This is a reasonable time commitment considering that this type
of evaluation can aid in the early identification of operating problems
that might ultimately result in extended periods of noncompliance and
expensive repairs.
4.1.1 File Review
Prior to inspecting a scrubber system, an agency inspector should
review the agency files to familiarize himself with all pertinent infor-
mation on the particular scrubber, types of emissions, and operating
history. The following items should be checked:
1. Pending compliance schedules and variances
2. Construction and/or operating permits pertaining
to source processes
3. Past conditions of noncompliance
4. Malfunctions reports
5. History of abnormal operations
The inspector should know what regulations are applicable to the
facility; and if he has any questions, he should consult the agency
administration or legal staff prior to conducting the inspection. It is
important to know what data are useful in confirming compliance and what
evidence is necessary to document possible violations of each of the
applicable regulations.
The inspector should also refer to plant layout drawings to fami-
liarize himself with emission point locations, to draw flow diagrams,
and to prepare the inspection report. If available in the files, the
inspector should note the facility's personal safety equipment requirements.
Based on his review of agency files, the inspector should schedule an
inspection time when plant processes will be operating at representative
conditions. The scheduling of a time to visit plants having batch opera-
tions or other irregular operating schedules (e.g., seasonal) is especially
important. The files should always be reviewed briefly before entry to
the plant so that important plant characteristics will be more easily
remembered.
4-4
-------
4.1.2 Evaluation of Plume Opacity
The initial step in any scrubber inspection is to observe the visible
emissions exiting the stack under representative operating conditions and
under weather conditions which permit opacity observations. The visible
emission observation, while more difficult than on other control devices
because of steam plume interferences, can provide a good indication of
the scrubber's performance.
In order to compare the opacity against baseline levels, the opacity
must be observed at the point of stack discharge where the pathlength
is fixed. Very often this is not possible due to the presence of light
scattering condensed water droplets. If this is the case, the opacity
of the residual plume should be observed after the point of water droplet
dissipation. The latter provides only a very rough check on the overall
performance of the scrubber.
Experienced observers are able to distinguish between a plume
obscured by water droplets and a plume free of water droplets. Parti-
culate matter in a plume usually has a bluish-white or brownish-white
color. Water vapor, on the other hand, is usually a brilliant white color.
A plume containing water droplets also appears to have more texture and
is more wispy in nature than a plume free of water droplets. To confirm
the difference between the plume being observed and a water droplet
plume, it is usually possible to simultaneously note the behavior of a
nearby steam vent plume.
After the plume observation is completed, ,the discharge from any
bypass stacks (if separate) should be observed. A general check should
also be made for fugitive emissions from process sources.
During these initial observations, a check should be made for any
entrained droplet fallout from the plume. The presence of dried spots on
adjacent equipment and in the surrounding area is an indication of demister
or entrainment separator problems. When the entrainment of solids is
severe, a lip of mud forms around the stack mouth and the stack is
discolored several feet down from the opening.
4-5
-------
4.1.3 Fan Evaluation
The location of the fan relative to the scrubber system should be
ascertained. If the fan is ahead of the scrubber, then the remainder of
the scrubber system and ductwork is under positive pressure. Leaks in
the gas handling system or an open measurement port will allow the gas to
escape into the areas immediately around the scrubber. Personnel conduct-
ing the performance evaluation should avoid any partially confined areas
where leaking gas could result in high concentrations of toxic agents.
If the fan is located downstream of the scrubber, then the shell
and the inlet ductwork is under negative pressure. In this case, any
leaks or open measurement ports will result in inleakage of ambient air.
This presents less of a personnel exposure problem, however, it should
not be assumed that pockets of toxic contaminants are not present in
some partially confined areas. Fans on the downstream side are vulnerable
to buildup of material on the fan wheel. This can lead to imbalance of
the wheel, particularly if the fan tip speed is high. If this remains
uncorrected, severe vibration and eventual disintegration of the fan is
possible. Personnel evaluating scrubber operation should leave the vi-
cinity of a "severly vibrating fan immediately; operators of such systems
should take immediate action to take the fan off line.
The inspector should ask about the recent operating history of the fan
with respect to particulate buildup, vibration, bearing failure, or cor-
rosion. These are indications of potential carryover from the demister.
4.1.4 Demister Pressure Drop "
If there is entrained solids fallout from .the plume or vibration of
the induced draft fan, the operation of the demister should be checked.
The upstream and downstream static pressures should be measured and com-
pared against baseline levels. A significant increase in the pressure
drop usually means partial pluggage of the demister, while pressure drop
values much lower than baseline values can signal that there are openings
through the demister beds.
It is generally advisable to have several static pressure ports
around the scrubber shell both before and after the demister. In this way
the upstream and downstream static pressures can be averaged.
4-6
-------
The measurements may first be made using a portable slack tube type
manometer or a Minihelic® differential pressure gauge. Due to the low
pressure drops involved, however, it may be necessary to repeat the
measurements using an inclined manometer or, at least, a low range Magna-
helic® gauge. The Magnahelic® and Minihelic® gauges should be calibrated
regularly.35
The presence or absence of a demister flush system should be noted.
This often consists of an array of spray nozzles above and/or below the
demister. Use of the sprays on a once per shift and a once per day basis
substantially reduces the potential for demister pluggage. The agency
inspector should inquire about the cleaning system operating schedule.
If an inspector is being conducted when the scrubber is down, it may
be possible to visually check for buildup on the demister. (Care is
necessary to avoid exposure to toxic gases which may be trapped within
the scrubber vessel.)
4.1.5 Scrubber Pressure Drop
4.1.5.1 Measurement of Pressure Drop. The penetration of parti-
culate matter through the scrubber is partially a function of the pressure
drop. Thus, measurement of the pressure drop is an essential step in the
routine performance evaluation. Unfortunately, the permanent on-site
monitors are not always reliable because of the potential for line pluggage.
For this reason, it is usually advisable to measure the pressure drop
using portable gauges.
Locating proper measurement ports is the first step in measuring the
pressure drop. The fundamental rule in locating and using measurement
ports is only those that are safely accessible should be used. Access to
ports in some scrubber systems may involve close contact with hot ducts,
unsafe climbing, and/or exposures to fugitive gas leaks. Heroic efforts
should not be taken to reach ports which have been incorrectly located.
A second rule is ports must be in a location relatively free of inter-
ference from sludge and entrained water droplets. Recommended locations
for ports in tray-type scrubber and venturi scrubber systems are shown in
Figure 4-2.
4-7
-------
It is important that the fan and the fan dampers jiot be included in
the system between the two measurement ports. Inclusion of the demister
(or cyclonic entrainment separator in venturi scrubbers) is also not
ideal; however, it must often be included because the region between the
venturi throat and the demister has a high flow rate with high concen-
trations of entrained water droplets and other materials, making measure-
ments of static pressure difficult. Inclusion of the demister usually
adds only 0.5 to 2 inche's of pressure drop (if there are no deposits),
only a small fraction of the venturi throat pressure drop.
GAS INLET
PREFERRED
OUTLET PORT
PREFERRED
INLET PORT
\\\\\\\\\\\
DEMISTER
*-«—*- PURGE
Figue 4-2. Preferred static pressure measurement locations.
4-8
-------
Once the appropriate measurement ports have been located, the
measurement work can begin. However, before opening the port, the
operating pressure within the duct or scrubber shell should be
considered; it,can either be roughly estimated or previous records and
measurements can be reviewed. If the system operates under positive
pressure, the opening of the port can result in the unintentional
fumigation of the area, in which case everyone present should be prepared
with the necessary personal protective equipment. If the system operates
under negative pressure, then it is necessary to be aware of the pos-
sibility that air infiltration through the port can result in measurement
errors by the aspiration effect. This is illustrated in Figure 4-3.
NEGATIVE
PRESSURE
DUCT
-20"
-25"
AIR/INFILTRATION
COPPER PROBE
RUBBER STOPPER
Figure 4-3. , Aspiration effect error in static pressure measurement.
Suction induced by air infiltration around the probe leads to
measured static pressures which are less (more negative) than actual.
The magnitude of the error can be as great as 10 to 15 inches w.c. for
some venturi scrubbers. Even tray-type units operating in the 6 to 12
inch pressure drop range, can have aspiration errors of 1 to 2 inches.
One way to minimize the aspiration effect in the measurement of static
pressure is to use a probe inserted well into the duct or scrubber
shell. This is illustrated in Figures 4-4a and b.
4-9
-------
DUCT WALL
COPPER TUBE
RUBBER STOPPER
TUBING
I"*"! pr-i *U
ELECTRICAL
BONDING WIRE
GROUNDING CLAMP
GROUNDING TAB
Figure 4-4a. Measurement of static pressure using a copper tube.
-DUCT
S TYPE PITOT TUBE
RUBBER STOPPER
TUBING
•J
-GROUNDING
CLAMP
ELECTRICAL
BONDING WIRE
-GROUNDING CLAMP
-GROUNDING TAB
Figure 4-4b. Measurement of static pressure using a pitot tube.
4-10
-------
Figure 4-4a illustrates the use of a 1/4" O.D. copper tube, while
Figure 4-4b illustrates use of the pitot tube. Both approaches are very
effective in avoiding problems with aspiration; however, both involve the
risk of water droplet and solids impaction on the probe. The measurement
port must usually be at least 3/4" in diameter to use the S-type pitot
tube.
The higher the negative static pressure is at a given location, the
smaller the measurement port should be in order to ensure that the port
can be effectively sealed during measurement. If it is necessary to use
an oversized port, care, must be taken to insure a good seal. The use of
gloves is not recommended since they can be sucked into the gas stream.
On the same account, sampling probes can also be lost.
One quite useful assembly for oversized ports is illustrated in
Figure 4-5. This consists of a 1/4" O.D. copper tube which is inserted
through a #4 rubber stopper and then a commercial sanding disk (with a
1/4" O.D. hole). The sanding disk serves as an effective flange to
prevent air infiltration and the rubber stopper holds the copper tube.
This assembly has been used with no difficulties on measurement ports
serving ducts at up to 100" negative pressure. The sanding disk should
be at least 1" greater in diameter than the port on which it is used.
SANDING DISK
COPPER TUBE
RUBBER STOPPER
)UCT WALL
Figure 4-5. Measurement port seal.
4-11
-------
In measuring the static pressure drop across a scrubber, it is
good practice to measure the upstream and downstream static pressures one
at a time. First of all, this guarantees that errors will not occur due
to pluggage of one of the ports. In the case of wet scrubbers, it is
possible (and even likely) that a port will plug minutes after it has
been rodded-out. Therefore, hooking up two separate lines to one gauge
as shown in Figure 4-6a is not recommended. It is preferable to make
the measurements one at-;a time as illustrated in Figure 4-6b. In this
case, a zero reading clearly indicates a port which was not originally
cleaned out or which has become blocked again. Furthermore, less tubing
is necessary when making the measurements one at a time. Since most of
the ports are some distance apart, the connecting tubing necessary to
measure pressure drop directly could represent a safety hazard; it could
also pass next to a hot gas duct which would cause it to collapse or burn.
Figure 4-6a.
Poor approach for measuring pressure drop across a venturi
scrubber. -
Figure 4-6b.
Preferred approach for measuring pressure drop across a
venturi scrubber.
4-12
-------
Before opening any static pressure port, it is important to confirm
that the port is not connected to any type of electrical transducer. In
some systems, transducers are an integral part of the overall process con-
trol system; the opening of the port can result in a low pressure signal
which will trigger the tripping off-line of the fan and process equipment.
If there is any doubt about the existence of a transducer somewhere along
the static pressure line, it should be assumed that one exists. Preferably
there should be separate, isolated static pressure ports installed strictly
for the purposes of routine evaluaton using portable gauges.
When recording the static pressure data, the location of each measure-
ment should be carefully noted. The data has no meaning whatsoever with-
out the accompanying data concerning measurement location. A sketch of
the system, showing the ports, the scrubber, the demister, dampers,
bypass ducts, and fan(s) should be included in the notes. As discussed in
the next section, the temperature of the gas stream should be measured at
the same points.
The pressure drop measurements should be conducted over a moderate
time period at each location to determine if there are any short term
temporal variations. These can be caused by cyclic process operating
conditions or by "seeking" type adjustable throat mechanisms (for venturi
scrubbers).
4.1.5.2 Measurement of Temperature. Immediately after measuring the
static pressure at the inlet and outlet ports, the gas stream temperature
should be recorded at the same points. These values can then be used to
convert the static pressures back to a standard gas density. (See Sec-
tion 2.3 for a discussion of the relationship between pressure drop, gas
density, and energy consumption.) In making these temperature measurements,
caution must be exercised to avoid (1) air infiltration at the port which
causes cooling of the measurement probe and (2) water droplet impaction
on the probe which also causes cooling.
The temperature at the scrubber inlet can also be used as an indica-
tion of potential operating problems. If it is greater than 300° F and
the gas stream contains a high concentration of condensible metallic or
organic vapors, then submicrometer particles can condense after the gas
has passed through the zone of optimal inertia! impaction. If it is over
4-13
-------
300° F in combination with a high solids content liquor, then "regeneration"
can result from the droplets evaporating to dryness.
4.1.5.3 Analysis of Pressure Drop. Low static pressure drop across
a wet scrubber, coupled with a high stack or residual opacity suggests
that the penetration rate has increased relative to the baseline level.
The logical starting point when confronted with these conditions is to
carefully inspect the entire liquid pumping and distribution system.
The reason for this is that pressure drop is a direct function of the
liquid-to-gas ratio at any specific gas velocity.
The most common reason for low liquid flowrate is nozzle or pipe plug-
gage. This may be indicated by an increase in the pump discharge pressure
(from the baseline levels). Other problem symptoms include low pipe skin
temperatures and high suspended solids in the recirculation line. If pos-
sible, the operator should rod out each nozzle and determine if there is a
change in the pressure drop and in the pump discharge pressure.
A second reason for the low static pressure drop may be a decrease
in the gas velocity through the system. In the case of adjustable throat
Venturis, the movement of the throat mechanism (damper, vanes, plump bob)
can change the throat velocity by a factor of 3. The relationship between
the gas velocity and throat velocity is linear, therefore, the pressure
drop can also vary by a factor of 3. In some designs, it is possible to
check for relative movement of the throat mechanism. If the wet scrubber
is a tray-type, spray tower, or simple venturi, then throat velocity adjust.
ments cannot be made independently of the gas flow rate. In these cases,
it may be advisable to measure the inlet gas flaw rate to determine if
the wet scrubber system is operating at less than design flow rates;
most scrubbers are less efficient under these conditions since particle
impaction is less effective.
The recommended technique for the measurement of gas flow rate up-
stream of a wet scrubber is the Pitot tube (EPA Reference Method 2). Due
to the large quantities of suspended material upstream of the scrubber, it
is generally necessary to use the S-type probe which is less prone to
pluggage. It should be noted that wet scrubbers are often used on sources
which interim"ttantly or continuously generate effluent gases in the
4-14
-------
explosive range. Accordingly, any probes inserted upstream of the scrubber
should be carefully grounded. A less accurate, but safer approach is to
measure the gas flow rate downstream of the scrubber. The decreased
accuracy results from evaporation of some of the liquor, and condensation
in the scrubber. Nevertheless, the downstream value is generally close
to the upstream value (after correction for gas density differences).
Another indication of the gas flow rate can be obtained indirectly
by using the fan motor current. A decrease in the corrected fan motor
current usually accompanies a decrease in the gas flow rate. Correction
of the motor current to standard conditions can be accomplished by multi-
plying the fan current at actual conditions by the appropriate factor
presented in Table 4-3. The data in Table 4-3 is accurate only when the
fan inlet static pressure is not less than -20 inches w.c.
Other problems which can lead to low pressure drop all involve the
pump and the liquid distribution system. Pump impeller wear and pump
seal leakage can both contribute to a decrease in the liquid flow rate.
At the earliest opportunity, the operator should inspect the pump packing,
the pump impeller, and strainers mounted before the pump. The position
of any flow control valves should also be checked.
TABLE 4-3. FAN DATA TEMPERATURE CORRECTION9
Temp
°F
20
40
60
80
100
120
140
160
180
200
220
240
260
280
300
^From
tion
Factor
0.91
0.94
0.98
1.92
1.06
1.09
1.13
1.17
1.21
1.25
1.28
1.32
1.36
1.40
1.43
"Basic Energy/Envi
Temp
°F
320
340
360
380
400'
420
440
460
480
500
520
540
560
580
600
ronment Analysis",
Factor
1.47
1.51
1.55
1.59
1.62
1.66
1.70
• 1.74
1.77
1.81
1.85
1.89
1.92
1.96
2.00
NAP A informa
series 67, by C. Heath, August 1978.
4-15
-------
4.1.6 Liquid-Gas Distribution
Liquid-gas distribution is, unfortunately, a difficult factor to con-
sider during a routine evaluation. Usually, the only conclusive evidence
of poor distribution (also called liquor maldistribution) is the observation
of severe scaling and deposits inside the unit. In extreme cases, there
may be a decrease in the pressure drop across the scrubber or across an
individual tray, at a constant gas flow rate.
Wet scrubber systems with moderate to high suspended solids levels
should be checked to the extent possible for maldistribution problems.
Partially or completely plugged nozzles are a major cause of poor gas-
liquid distribution which is particularly disruptive in venturi scrubbers.
One external check for plugged nozzles which can be made is to compare
the skin temperatures of the various liquor supply lines going to the
scrubber. If one of these lines is plugged, the flow of hot recirculated
liquor through it will be substantially reduced and its skin temperature
will begin to fall. A difference of 5 to 10 degrees F between two supply
lines often indicates a plugged line. (The temperature difference is not
as great as might be expected since conduction of heat along the pipe
continues despite the plug.) Supply line pressure can also provide
evidence of pluggage.
4.1.7 Liquor Quality
4.1.7.1 pH Measurement and Evaluation. The pH of the recirculated
liquor should be measured during each routine evaluation. This can be
done using color indicating pH paper or a portable pH meter. The pH paper
is adequate assuming it is not necessary to have highly accurate data
and the scrubber liquor will not attack the paper. Highly colored liquor,
liquor with high levels of foam, liquor with high levels of colloidal
matter, and oxidizing liquor all preclude the use of the pH paper.
The liquor sample should be drawn from a region of the system
where its pH is at the lowest value. This is commonly at the scrubber
sump effluent point, prior to the discharge into the recirculation tank.
Another sampling point of interest is the collected liquor entrainment
at the region of the demister. This sample can be acquired using the
probe developed and discussed by Schifftner.36
4-16
-------
A battery powered portable pH meter is ideal for pH evaluation
of wet scrubber systems. These normally have an accuracy of-_+ 0.2 pH
units and should be standardized using at least two buffers.
4.1.7.2 Solids Content. The liquor turbidity should be quali-
tatively analyzed. If the turbidity is high, nozzle erosion and/or line
piuggage could occur in the system. The rate by which the material
settles to the bottom of the sampling flask should also be considered,
since it is an indirect indication of the particle size. Larger parti-
culate material is more abrasive.
If the turbidity appears high, the inspector should evaluate the
purge rates. A reduction in this purge rate can lead to an increase in
the total solids content. The settling pond (if any) should also be
inspected to ensure there is minimum total solids at the pump intake.
The presence of a persistent, high opacity plume at apparently ade-
quate pressure drop levels can be due to evaporative regeneration of
particulate in the presaturator or the quench tower. Samples of the
liquor fed to the chamber should be obtained over a representative period
of time; these should be analyzed for both suspended solids and dissolved
solids. Similar samples should be obtained of liquor drained from the
chamber which did not evaporate. The difference in the total solids con-
tent of the feed and discharge streams will give a rough indication of the
loss of particulate to the gas stream. This is only approximate since
some of the solids are lost as entrained water droplets from the chamber,
and it is possible that some particulate from the inlet gas stream is
captured in the chamber discharge stream.
4.1.7.3 Surface Tension. If the opacity of the discharge has
increased without any other apparent changes, then one possible cause
is a change in the surface tension. This change could affect both the
droplet size distribution and the effectiveness of impaction. The surface
tension can be measured using ASTM Method D-450. The quantity of anti-
fearning agents, surfactants and flocculants should be recorded, if known.
4.1.8 Nozzle PIuggage
Nozzle piuggage should be suspected when there is increased opacity
and one or more liquor feed lines with lowered skin temperatures. If it
4-17
-------
is impossible to rod the nozzles out during operation, then it must be
done during the next outage. At this time, each nozzle should be visually
inspected for erosion and pluggage; if damaged, the nozzles should be
replaced. In some cases, it will be possible to open'an access hatch and
observe the operation of the nozzles (with no gas flowing through the
system). If a nozzle is partially plugged, the spray angle and pattern
will be severely distorted. If it is completely plugged, at best, only a
trickle of liquor will be observed. A high powered light (explosion
proof) is useful for inspection of the nozzles from the access hatch.
4.1.9 Bypass Duct Damper
The damper position of the bypass duct should be checked. If there is
any indication of substantial leakage across the duct, follow up testing
is necessary.
4.1.10 Presaturator and Ductwork-
Presaturator (or quench) water quality is particularly important; a
sample of this liquor should be obtained and checked. Turbidity of this
liquor should be very low, in fact, it should resemble drinking water
quality.
The general physical condition of the ductwork leading from the process
to the scrubber should be briefly checked. Fugitive leaks from positive
pressure systems are a direct bypass of the scrubber system. Leaks into
negative systems can reduce the quantity of gas pulled from the source and
thereby affect capture and result in fugitive emissions at the process.
The locations of any obvious holes or gaps/tears in expansion joints
should be noted on a sketch.
Air infiltration along the inlet duct can be evaluated by determin-
ing the 02 levels (combustion sources only) at various points along the
duct. An increase in the 02 levels between two measurement points provides
clear indication of the location of the leakage. A material balance can
be used to estimate the quantity.
Air infiltration along the ductwork can also be detected by compari-
son of the gas inlet temperature to the baseline value. A substantial
decrease is indicative of air infiltration.
4-18
-------
4.1.11 Process Conditions
As discussed earlier, the process operating conditions and raw material
characteristics can have a major impact on scrubber operating conditions
including, but not limited to: the particle size distribution, the gas
flow rate, the particle surface characteristics, and the concentration of
condensible vapors. Nonetheless, it is easier to identify the emergence of
problems by observing the stack and performing a routine evaluation of the
scrubber, than by first evaluating the process equipment. Thus, the evalua-
tion of process conditions should be used primarily as a follow-up to
problems identified by inspection of the air pollution control equipment.
On a routine basis, the evaluation should be limited to ascertaining
the overall process operating rate and determining the adequacy of hood cap-
ture (if applicable). The hood capture can usually be evaluated visually
from a safe vantage point; an alternative method is to use the hood static
pressure as a qualitative means to estimate the hood gas flow rate. Compari-
son of the present hood static pressure against the baseline period hood sta-
tic pressure provides an estimate of any shifts in hood capture efficiency.
An increase in the vapor concentration (material which may condense
in the scrubber to form particulate matter) of the gas stream can be indi-
cated by an increase in the stack opacity as compared to baseline values
and perhaps by a change in the process operating conditions. The presence
of vaporous material can be confirmed by an extractive stack test using a
front filter (unheated) and a set of impingers immersed in ice. Note that
this does not yield a quantitative measurement, only an indication of
the pretense of vaporous material.
4.2 APPLICATION OF BASELINE PERFORMANCE EVALUATION TECHNIQUE TO SPECIFIC
TYPES OF SCRUBBERS
There is substantial diversity in the design of particulate wet
scrubbers. The various types of scrubbers serve different applications,
are vulnerable to different operating problems, have different main-
tenance requirements, and have different particulate removal capabi-
lities. To a certain extent, the means of evaluating performance is
also different. Some of the unique factors which must be considered
while evaluating performance for common types of scrubbers are described
in this section. It is by no means an exhaustive summary of inspection
techniques, rather it is intended to illustrate use of the evaluation
approach presented earlier.
4-19
-------
Six major categories of participate wet scrubbers are presented in
Table 4-4. These are listed in the approximate order of both increasing
effectiveness and increasing operating cost. The preformed spray scrubbers
range from simple "homemade" units for small scale material transfer
applications to very large, sophisticated systems used in flue gas desul-
furization systems. Packed tower scrubbers are generally used for combina-
tion particulate and gaseous contaminant removal. They are rarely used
solely for particulate control since there are more economical and simple
approaches of equal or greater effectiveness. Moving bed scrubbers were
developed as an alternative to packed beds for applications where the gas
stream contains sticky particulate which could clog a packed bed. This is
a comparatively new approach, having been only introduced during the last
20 years. There are numerous tray-type scrubbers almost all of which offer
TABLE 4-4. MAJOR TYPES OF WET SCRUBBERS3>b
Category
Common
Application
Specific Type
Preformed Spray
Packed Bed
Moving Bed
Tray-type
Impingement Plate
Mechanically Aided
Gas-atomized
Material Handling
Asphalt
Phosphate Fertilizer
Primary Aluminum
Coal-fired Boilers
Spray Towers
Cyclonic Spray Towers
Vane-type Cyclonic Towers
Multiple-tube Cyclones
Standard Packed Bed
Fiber-bed
Cross-flow
Grid-packed
Turbulent Contact Absorbers
Municipal Incinerators Perforated Plate
Material Handling
Foundries
Coal-fired boilers
Asphalt
Steel Mills
Sewage Sludge
Incinerators
Wet Fans
Standard Venturis
Variable Throat Venturis
Orifice Scrubbers
List not intended to be all inclusive.
Adapted from a similar table presented in Reference 37.
4-20
-------
moderate to high removal efficiencies at low pressure drops. They are
currently used in a very wide set of applications. The principal category
of scrubber for high efficiency collection of submicron particulate is
the gas-atomized type, the most common of which is the venturi scrubber.
4.2.1 Preformed Spray Scrubbers
The preformed spray scrubber is the simplest type of wet scrubber
and it has the lowest overall particulate collection capability. It
usually consists of a vertical contact chamber with an array of spray
nozzles as shown in Figure 4-7. The particuTate laden gas stream enters
near the bottom and passes upward past the spray headers. The nozzles
may consist of sophisticated liquid atomizing nozzles or may simply be
small holes drilled into the spray header.
Clean gas
Dirty gas
Figure 4-7. Typical preformed spray scrubber.
(Courtesy of EPA's Air Pollution Training Institute.)
This type of scrubbing system is often difficult to evaluate due to
the lack of instrumentation and accessibility. Rarely do preformed spray
scrubbers include continuous recording pH meters and recirculation liquor
flow meters. There is usually a pump discharge pressure gauge but rarely
pressure gauges on the individual pipes leading to the nozzle.
The types of data which are useful for evaluating the performance are
listed in Table 4-5. Emphasis is given to analyses of the recirculation
liquor quality since nozzle erosion and pluggage can be major problems
4-21
-------
for preformed spray scrubbers. Both problems have a direct and significant
impact on the particulate collection efficiency by changing the spray
coverage and average droplet sizes. During a routine inspection, the
turbidity of the recirculation should be observed (classified as "light",
"moderate", and heavy"). If the turbidity is moderate or heavy, the
condition of the spray nozzles should be evaluated to the extent possible.
Pluggage of some of the nozzles will be indicated by an increase in the
supply pipe pressure. Erosion of the nozzle will yield the opposite
effect on the supply pipe pressure gauge.
Many of these systems operate on an intermittent basis. The condition
of the nozzles can be checked while the scrubber is off-line by operating
the recirculation pump for a short period of time and observing each
nozzle with a high powered flashlight (explosion proof). Pluggage or
erosion of the nozzle will result in a distorted spray angle. This check
should only be done if there is an appropriate access hatch and if the
scrubber vessel is purged of any toxic gaseous and particulate.
If there is any indication of liquor solids related operating prob-
lems, the condition of the pond (if any) should be checked. The pond
should have sufficient capacity to allow suspended materials to settle;
multiple zones separated by weirs will permit maximum settling. The pump
intake line should be located well above the elevation in the pond where
silt and collected solids can be entrained in the liquor.
TABLE 4-5. PREFORMED SCRUBBER EVALUATION DATA
Data Baseline
Recirculation Liquor Turbidity
(Light, Moderate Heavy)
Recirculation Liquor Total Solids
Recirculation Liquor Suspended Solids
Recirculation Liquor pH
Spray Nozzle Operating Pressure
Physical Condition of Shell
Scrubber Inlet 02 or C0£ Levels
(Combustion Sources Only)
Quantity of Surfactants Used
Quantity of Anti -Foaming Agents Used
Liquor Flow Rate
X
X
X
X
X
X
X
X
X
X
Routi ne
X
X
X
X
X
X
X
4-22
-------
It is important to measure the pH of spray tower scrubbers since many
of these are constructed of carbon steel. If the pH of the recirculation
liquor falls below 6, corrosion is probable. If the pH increases above 10,
precipitation of calcium and magnesium compounds can result in scaling of
the scrubber and pluggage of the pipes and nozzles.
The liquor flow rate to the scrubber is an important operating
variable. The on-site monitor should be checked. If one is not available,
the flow rate should be estimated at the point where the liquor returns
to the pond or the recirculation tank. One parameter which is not of
interest is the pressure drop across the scrubber. This is not directly
related to the effectiveness of the particle capture and the magnitudes
of the pressure drops are small.
4.2.2 Packed Bed Scrubbers
Packed bed scrubbers (also called packed tower scrubbers) have
primarily been used for gas absorption or for gas cooling. Both functions
are facilitated by the large liquid surface exposed to the gas stream as
the liquid flows downward over the packing material. A typical packed
bed scrubber is illustrated in Figure 4-8. Another common type of packed
bed system is a cross flow scrubber. It is particularly advantageous
when a liquid mist is being collected since this facilitates drainage of
the accumulated material from the bed.
Mist eliminator
•;•>. Dirty gas
Figure 4-8. Typical packed bed scrubber.
(Courtesy of EPA's Air Pollution Training Institute.)
4-23
-------
The packing materials used in packed beds can consist of crushed
granite, raschig rings, pall rings tellerettes, and intalox saddles.
These are illustrated in Figure 4-9 (with the exception of the gravel).
All of the packing materials are usually the size range of 0.5 to 3
inches (1.25-7.5 centimeters), and the packing is usually randomly orien-
ted in the bed. Some commercial units utilize several beds in series,
each with a separate liquor supply and distribution system.
t
The types of data most useful for the evaluation of packed bed
scrubbers (particulate removal only) are listed in Table 4-6. As in the
case with the preformed scrubbers, the primary emphasis is the liquor
quality. One of the major operating problems of packed bed scrubbers is
pluggage of the bed due to deposition and/or precipitation of solids.*38
The flow rate of liquor is often insufficient to remove the accumulated
material. These solids can cause increased static pressure drop across
the unit, reduced gas flow through the control system, and channeling of
the gas stream through only a part of the bed.
Berl saddle
Intalox saddle
Pall ring
Tellerette
Figure 4-9. Typical packing materials.
(Courtesy of EPA's Air Pollution Training Institute.)
4-24
-------
TABLE 4-6. PACKED BED SCRUBBER EVALUATION DATA
Data
Recirculation Turbidity
(Light, Moderate, Heavy)
Recirculation Liquor Total Solids
Recirculation Liquor pH
Recirculation Liquor Nozzle Pressures
Recirculation Liquor Flow Rate
Inlet and Outlet Gas Temperatures
Apparent Shell Corrosion and Erosion
Baseline
X
X
X
X
X
X
X
Routine
X
X
X
X
X
X
The turbidity of the recirculation liquor should be low. The condi-
tion of the packing should be checked if the turbidity is moderate to
heavy. This can usually be done by opening the access hatches above and
below each bed (the scrubber must be off-line).
Typical liquid-to-gas ratios for packed bed scrubbers are 1 to 6 gal-
lons per 1000 ACF (0.1 to 0.5 liters per AM3).39 A decrease in the liquor
flow rate relative to the baseline period may indicate a reduction in the
particulate removal rate and a major decrease in the gaseous removal
efficiency. The pH is measured primarily to ensure that corrosion is not
occurring.
4.2.3 Moving Bed Scrubbers
Moving bed scrubbers were developed primarily as an alternative to
packed bed scrubbers in applications having very sticky particulate
material and/or high levels of gaseous contaminants. The turbulent
motion of the lightweight packing material allows self cleaning of the
deposited material without sacrificing the good gas transfer capability
common to packed bed scrubbers. Principal applications of this type of
scrubber include primary aluminum Soderberg pptlines and coal-fired
boilers (for flue gas desulfurization). A simple schematic of a moving
bed scrubber is shown in Figure 4-10. The trade name commonly used is
Turbulent Contact Absorber (TCA).
As shown in Figure 4-10, the gas enters at the lower side of the unit
after passing through a presaturator chamber (optional). The pollutant
laden gas stream passes upward through a series of beds each of which is
10% to 25% full of the lightweight packing. The liquor is introduced at
4-25
-------
Clean gas
Mist eliminator
Figure 4-10. Typical moving bed scrubber.
(Courtesy of EPA's Air Pollution Training Institute.)
the top through a set of nozzles and passes through the beds counter-
current to the gas flow. The bed is fluidized by the moving gas stream
and this results in the formation of liquid droplets and sheets in the
turbulent zone of the bed. Often one or more beds of packing are operated
dry (above the point of liquid injection) in order to serve as an initial
demister. A chevron demister or equivalent serves as the main entrainment
separator.
Evaluation data for moving bed scrubbers are presented in Table 4-7.
Unlike the previous types of scrubbers discussed, the liquor turbidity is
not critical to the performance of the scrubber. Therefore, an increase
or decrease in the liquor turbidity would not be cause for major concern,
this would, however, eventually degrade the liquor distribution nozzles at
the top of the scrubber.
TABLE 4-7. TURBULENT CONTACT ABSORBER EVALUATION DATA
Data
Static Pressure Drops Across Each
Recirculation Liquor pH
Recirculation Liquor Flow Rate
Spray Nozzle Header Pressure
Physical Condition of Shell
Baseline
Stage x
X
X
X
X
Routine
X
X
X
X
4-26
-------
The pressure drop the entire scrubber would be measured by measuring
the static pressure at the scrubber inlet, and the pressure drop prior to
the demister. The procedures for measuring static pressure discussed
earlier should be used. Typical pressure drops are 4 to 20 inches of
water. The pressure drop is proportional to the gas velocity, the rate
of liquor addition, and to the number of beds in series.
The recirculation liquor flow rate needs to be confirmed only if
there has been a decrease in the pressure drops or an increase in the
plume residual opacity.
4.2.4 Tray-type Scrubbers
A tray-type tower consists of a vertical shell with one or more
plates or stages mounted horizontally. The liquor is introduced at the
top through a simple delivery pipe and flows onto the top stage. The
height of liquor on this stage is maintained at 0.5 to 3 inches (1.3 to
7.6 centimeters) by a downcomer on the side opposite the inlet pipe.
While the liquor is passing across the stage it is exposed to a relatively
high velocity gas stream which passes up through numerous holes in the
stage. A sketch of a tray-type scrubber is shown in Figure 4-11. Two
common stage designs are illustrated in Figure 4-12.
Penetration of particles greater than 1 jimA is low in most tray type
scrubbers. The degree of control for submicron particles is partially
dependent on the number of stages, the type of stage, the size of the
holes in the stage, and the design gas velocities through the holes. The
static pressure drop is several inches of water per stage.40,41
As with almost all types of particulate wet scrubbers, tray-type
units are vulnerable to corrosion and pluggage. Both corrosion and pluggage
have an impact on the particulate collection efficiency in addition to
their obvious effect on the scrubber's useful life.
The static pressure drop across the scrubber is directly related to
the particulate removal effectiveness, therefore, this should be measured
during each inspection. Measurement posts should be located before the
first tray, between each tray, and after the last tray before the gas
stream enters the demister section. The liquor quality is important,
4-27
-------
Clean gas
Figure 4-11. Tray-type scrubber.
(Courtesy of EPA's Air Pollution Training Institute.)
Sieve Plate
Impingement Plate
Figure 4-12. Types of stages.
(Courtesy of EPA's Air Pollution Training Institute.)
4-28
-------
especially for the impingement tray type scrubbers. Due to the extremely
small holes in the tray (approximately 3/16ths of an inch), suspended
solids concentrations of more than 1% by weight can lead to rapid pluggage.
Obviously, the turbidity of the recirculation liquor should be very low.
If the opacity of the unit has increased, it may be necessary to
evaluate the condition of the trays. This can be done by opening access
hatches above and below each of the trays. A build-up of materials in
part of the tray can lead to the channeling problem. Partial pluggage
of the holes leads to an increased pressure drop. Evaluation data for
tray-type scrubbers are presented in Table 4-8.
TABLE 4-8. TRAY-TYPE SCRUBBER EVALUATION DATA
Data
Pressure Drop Across Each Stage/Tray
Recirculation Liquor Turbidity
(Light, Moderate, Heavy)
Recirculation Liquor Suspended Solids
Recirculation Liquor pH
Physical Condition of Shell
Physical Condition of Trays
Baseline
X
X
X
X
X
X
Routine
X
X
X
X
4.2.5 Gas-atomized Scrubbers
Gas-atomized scrubbers are the most energy intensive and efficient
type of particulate wet scrubber now in large scale commercial utiliz-
ation. Regardless of the specific design, they all operate by accelera-
ting the gas stream to high velocities, usually in the range of 10,000
to 40,000 feet per minute (3,050 to 12,2000 meters per minute). The
design differences are found in the method of injecting the scrubber
liquid into the gas stream at the converging section of the scrubber.
Here, during the period before the atomized liquid reaches the same
velocity as the accelerated gas stream, the droplets serve as inertial
impaction targets for the entrained particles.
4.2.5.1 Orifice scrubbers. The common denominator in orifice
scrubber design is that the constricted area which serves as the "throat"
is partially flooded by a stationary pool of liquor. Atomization and
4-29
-------
acceleration of the liquor occurs as the gas stream passes through
the constricted area with typical liquid-to-gas ratios up to 20 gallons
per 1000 ACFM. Evaluation data requirements are provided in Table 4-9.
TABLE 4-9. ORIFICE SCRUBBER EVALUATION DATA
Data
Pressure Drop
Liquor pH
Liquor Level
Physical Condition of Shell and
Internal Components
Baseline
X
X
X
X
Routine
X
X
X
X
One commercial type of orifice scrubber, a Western Precipitation
Tubulaire Gas Scrubber, is illustrated in Figure 4-13. The particulate
laden gas stream passes down through the central delivery tube and then
makes a 180 degree turn after it passes around the central cone. The
baffle to the left of the inlet tube serves as a demister. It is very
difficult to evaluate the liquid-to-gas ratio with this kind of device,
since the liquor flows by gravity back to the main sump without the need
for a recirculation pump and piping (some systems have recycle pumps);
nevertheless, the liquor level control is very important for the proper -
operation of the scrubber.
GAS INLET
GAS
OUTLET
DRAIN
SCRUBBING LIQUOR
Figure 4-13. Typical orifice type gas-atomized scrubber.
4-30
.
-------
A second common type of orifice scrubber is illustrated in Figure 4-14,
The contaminated gas stream enters from the top of the collector and flows
down the rectangular inlet channel. The gas turns near the bottom and
enters a set of venturi nozzles inclined slightly upward. The liquid
level is controlled to a height that ensures that the desirable amount of
liquid is entrained and atomized. There are a number of venturi nozzles
arranged in parallel; particulate impact ion on the droplets occurs primar-
ily in the throat of these nozzles due to their high relative velocites
in this region. The entrained droplets are later removed in the series
of stationary baffles located downstream of the nozzles.
GAS OUTLET
BAFFLES
VENTURI
LIQUOR
DRAG
CHAIN
Figure 4-14. Typical orifice type gas-atomized scrubber.43
4-31
-------
As shown in Figure 4-14, a drag conveyor at the bottom of the sump re-
moves the accumulated solids. As with the previously mentioned type of
orifice scrubber, there is no recirculation line and pump since the liq-
uor is retained inside the collector. For this reason, it is difficult
to independently evaluate the liquid-to-gas ratio.
Since the fan is often mounted directly above the collector, care
must be taken in the measurement of the pressure drop across the unit.
The inlet static pressure may be taken on the inlet duct or anywhere
along the inlet chamber portion of the scrubber. The outlet pressure
must be taken before the gas enters the inlet to the fan. For this
reason, the static pressure port must be located along the collector
wall in the general vicinity of that illustrated in Figure 4-14. Loca-
tion of the port below the first baffle is not recommended due to the
high quantity of sludge and liquor which impacts on the wall and could
cause pluggage.
One of the principal problems of orifice scrubbers is maintaining
proper liquid levels. Sufficient make-up water must be added to allow
for evaporation, carry-over of droplets, and loss of liquid in the sludge.
Corros'ion and erosion of the constricted area or venturi nozzles are
also common problems and can lead to substantially reduced collection
efficiency. Accordingly, these nozzles often are simply bolted in to
facilitate replacement.^
4.2.5.2 Venturi scrubbers. This type of gas-atomized scrubber is
generally used in applications requiring large scale systems. There are
numerous varieties available with the principal differences being in:
1. Methods for varying the throat area to adjust for changes in gas
flow rate.
2. Methods for injecting the liquor ahead of the throat.
3. Presence or absence of a diverging section.
4. Design of the entrained liquor droplet separator following
the throat.
A standard fixed throat venturi scrubber is illustrated in Figure
4-15. The particulate laden gas stream enters the converging section of
the venturi which usually has an angle of approximately 25°. The liquor
4-32
-------
is injected in the converging section and is atomized near the inlet to
the throat. Immediately following the converging section, the gas
stream enters a diverging section (the throat having a negligible length
in this case). In the diverging section of the throat of the unit
shown in Figure 4-15 the angle is less than 15°; here the gas is gradually
decelerated prior to entry to the cyclonic chamber. As shown in the
plan view sketch, the gas stream enters the entrainment separator tangen-
tial ly. Some commercial units incorporate an additional demister tray
to further remove the entrained droplets.
THROAT
LIQUOR INLET
CONVERGENT SECTION—^
DIVERGENT SECTION
GAS OUTLET
CYCLONIC DEMISTER
SUMP
Figure 4-15. Venturi scrubber with long divergent section.
Data necessary for the evaluation of venturi scrubbers is presented
in Table 4-10. The pressure drop is particularly important since this
is related to the impaction effectiveness. As described earlier, the
pressures must be converted back to a standard gas density, if the inlet
and outlet gas densities have changed substantially since the baseline
period.
4-33
-------
TABLE 4-10. EVALUATION DATA FOR VENTURI SCRUBBERS
Data
Pressure Drop
Recirculation Liquor Flow Rate
Recirculation Liquor Total Solids
Recirculation Line Skin Temperatures
Presaturator Total Solids
Presaturator Turbidity
(Light, Moderate, Heavy)
Inlet and Outlet Gas Temperatures
Physical Condition of Shell
Quantity of Surfactants Used
Quantity of Flocculants Used
Quantity of Anti -Foaming Agents Used
Recirculation Liquor pH
Baseline
X
X
X
X
X
X
X
X
X
X
X
X
Routine
X
X
X
X
X
X
X
X
Due to the gradual slope of the divergent section in this unit, the
kinetic energy of the gas/liquor stream is partially converted back to po-
tential energy* As a result, the gas stream static pressure will increase
moderately as the gas passes through the divergent section. The pressure
drop of direct interest to the inertial impaction phenomenon is the pres-
sure drop between points 1 and 2 in Figure 4-15. The actual pressure
drop measured in many commercial systems is representative of the pressure
drop between points 1 and 3. The latter value will be lower according
to the extent of static pressure recovery.
A variation of the classical fixed throat design is the annular ring
throat. The flange separating the converging and diverging sections has
a small extension lip which supports a ring insert. The area of the
opening in this insert determines the maximum throat velocity at a given
gas flow rate, and thereby controls the collection efficiency. This
insert can be changed and will vary the collection efficiency.45 On
some commercial units it is difficult to determine if such a change
has been made or even that the scrubber has an insert. Erosion of an
insert over the time can lead to a reduction in gas velocities, and a
corresponding reduction in the particulate control efficiency. Erosion
of the insert or installation of a different insert can often be de-
tected by comparing the present pressure drop at a given gas flow rate
and liquor flow rate with baseline data. One generally accepted rela-
tionship for the pressure drop is shown in Equation 5-1.
4-34
-------
AP = 0.001 Vt2 (L/6)
Equation 5-1
where:
AP = pressure drop, centimeters of water
2
Vj. = throat velocity, cm/sec
L/G = liquid-to-gas ratio, (m3/g)/(m'3/g)
In a common type of adjustable venturi,. a hand cranked assembly
moves a center cone up or down to change the annular throat area. The
length of the throat also varies slightly as the cone moves. The liquor
is injected from a set of four tangential pipes (no nozzles) and from a
center deluge pipe oriented directly above the center cone. A flooded
elbow (a depression of 6" to 9" filled with liquid) is used at the bottom
of the venturi section to absorb the impact of the turning gas stream.
This reduces erosion of the elbow.
Another variation of the annular area variable throat venturi is the
flooded disc scrubber. As shown in Figure 4-16, the center assembly
includes a large flat disc mounted horizontally at a point just upstream
of the constricted area. The liquor passes up through the center column
and out over the disc. As the gas stream is accelerated around the
edges of the disc, the liquor is entrained and atomized. Particulate
impaction on the atomized liquor occurs directly downstream of the
disc. The gas velocity can be controlled by slight adjustments in
the elevation of the flooded disc with relation to the constricted area.46
There is no divergent section per se, therefore, no substantial pressure
recovery would be expected downstream of this type of collector.
One variable throat design common to many commercial units is depicted
in Figure 4-17. In this design, two damper blades are used to vary the
open area in the rectangular throat and thus control the gas velocity
through the throat. The position of the damper blades are controlled
either manually or by a differential pressure sensor and controller. The
liquor can be injected by a number of methods including the deluge nozzles
with deflectors shown in Figure 4-17. Other designs have a single variable
4-35
-------
damper on the throat with a side mounted controller. In these units, the
liquor spray is directed towards the section opposite the damper.
The dampers are in the high velocity zone, therefore, erosion can
be a problem. The scrubber may gradually lose the capability to achieve
the baseline pressure drops. To diagnose this problem an internal
inspection may be necessary when the scrubber is off-line and purged
out (agency inspectors should jTot^ enter equipment).
clean gas out
dirty gas in
water in
slurry but
Figure 4-16. Flooded disc 'type venturi scrubber.
(Courtesy of Research Cottrell.)
4-36
-------
GAS INLET-
LIQUOR INLET
THROAT DAMPERS
GAS OUTLET
Figure 4-17. Adjustable venturi throat mechanism.
A variable rod-type, adjustable venturi scrubber is illustrated in
Figure 4-18. In this design, the throat is comprised of the multiple
slots between the rows of horizontal rods. The gas velocity can be mod-
ified by changing the diameter of the rods, the spacing of the rods, and
the spacing between adjacent layers of the rods.47 This type of scrubber
is conceptually similar to the annular ring type discussed earlier, except
that changes in the throat area can be made while the unit is on-line.
However, it is difficult for regulatory agency personnel to independently
determine that the rod arrangement has been changed. Erosion and corro-
sion of the rods are potential problems. The pressure drop can provide
a good indication of these conditions.
4-37
-------
Dirty gas
« ^^ »
•^"i Water sprays
Figure 4-18. Variable rod venturi scrubber.
4.3 APPLICATION OF BASELINE PERFORMANCE EVALUATION TECHNIQUE TO SPECIFIC
INDUSTRIES
This section describes a number of wet scrubbers applications and
the problems unique to each application to a specific source category.
It is not an exhaustive summary of all applications of the various types
of wet scrubbers. However, each application has been included because it
illustrates conditions and potential problems that must be considered
when evaluating the performance of the wet scrubbers on many different
processes.
4-38
-------
4.3.1 Sludge Lime Kilns at Kraft Pulp Mills
Calcination of the sludge produced in the kraft pulping process is
necessary to reuse the lime. The sludge, which is normally 55% to 65%
solids, is fed to a rotary kiln which is usually fired by either oil or
natural gas.48,49,50 Typically rotary kilns are 8 to 13 feet in diameter
and 100 to 400 feet long,49 and have exit temperatures in the range of
300° to 500° F.48 A venturi scrubber having either a fixed or variable
throat is the most common type of air pollution control device used on
rotary kilns. Some plants have previously used impingement plate scrub-
bers, but the Venturis have gradually gained the major share of the market
because of their lower capital cost and higher efficiency.49
Typical static pressure drops for the venturi scrubbers range from
10 to 30 inches of water; however, the large majority of the systems op-
erate in the much narrower range of 15 to 20 inches of water.48 The
liquid-to-gas ratios vary somewhat depending on the design of the equipment,
but a representative ratio would fall in the range of 10 to 25 gallons
per 1000 ACF.49 Due to the alkaline nature of the entrained dust, the pH
of the recirculating liquor is high, often above 11. Consequently,
scaling has been a problem in some units
The mass loadings of the inlet gas stream from the rotary kiln are
relatively high at 3 to 20 grains per ACF.48,49 The large majority of
these particles are calcium oxide dust particles which have a fairly large
geometric mean size. However, the gas stream may also contain a small
small quantity of sodium oxide material which evolves as a very small dia-
meter fume.50*51 During the periods when substantial quantities of
submicron sodium oxide particulate are present, the mass emissions can
be higher than the baseline period even though the pressure drop has not
changed. This illustrates an important point: A wet scrubber cannot be
evaluated independent of basic process operational checks. In this case,
the amount of sodium material evolved is strongly dependent on the quantity
of sodium in the limestone mud feed, which varies from 0.1% to 2.5% by
weight. The inspector should evaluate the operation of the mud washers
if there is any question concerning the emissions from the scrubber
system.
4-39
-------
Compared with other applications of wet scrubbers, the quantity of
suspended solids in the recirculation liquor of lime kiln scrubbers can
be high. The solids usually range from 10% to 30% by weight,49 and plug-
gage of the inlet lines to the venturi scrubber is a frequent problem.
Venturi scrubbers on lime kilns are rarely equipped with presatura-
tion chambers, thus, the inlet gas temperatures to the throat are usually
greater than 300°F, causing some liquor evaporation. This can have an
adverse effect on particle collection and will increase the gas volume
handled by the induced draft fan.
The typical operating characteristics of particulate wet scrubbers
used on rotary lime kilns are summarized in Table 4-11.
TABLE 4-11.
TYPICAL OPERATING CHARACTERISTICS OF SCRUBBERS ON
ROTARY LIME KILNS AT KRAFT PULP MILLS50.5!
Parameter
Liquid-To-Gas
Liquor Solids
Pressure Drop,
Ratio, Gal
Content, %
Inches H2
s/1000
by wei
0
ACF
ght
Scrubber
Impingement
4-5
1-2
5-7
Type
Venturi
10-25
10-30
10-30
the principal problem of the impingement plate scrubber as applied
to rotary line kilns, is the pluggage of the very small holes in the
plates. For this reason, the solids content of the recirculation liquor
must be maintained at less than 2% by weight. The venturi scrubber on
the other hand, is less expensive to purchase and can be operated at a
higher solids level, therefore, use of the venturi facilitates the over-
all plant-wide water and lime balances. The venturi operates at a higher
pressure drop and achieves greater particulate removal efficiencies at
these pressure drops. The venturi scrubber is the type used on systems
installed in the last 10 years.
4.3.2 Asphalt Batch Plants
There are several major processes used to make asphalt concrete, the
two most common being the hot mix batch plants and the newer drum mix
plants. There are substantial differences in the characteristics of the
4-40
-------
particulate matter from these two processes. In both cases, subtle
process changes change the particle size distribution substantially and
may even alter the wettability of the materials.
In the conventional hot mix plant, the various grades of aggregate
necessary to produce the types of asphalt are stored in open piles or
bins near the dryer. These are fed from the bins at a controlled rate
using weigh belt feeders. The accumulated material passes into the top of
the dryer where it is exposed to high temperature flue gas. The aggregate
is dried as it passes through the dryer countercurrent to the gas stream.
A single burner, usually fired by either gas or oil, is used to provide
the necessary heat.
The dried aggregate is then transferred to the top of the tower where
it is mixed with heated asphalt purchased from local refineries. The ag-
gregate and asphalt are mixed in the pug mill and discharged to delivery
trucks. The flue gas from this process passes through a cyclone for re-
moval of the potentially useful aggregate dust, and then goes to the
scrubber. Most newer plants, being subject to the Standards of Perfor-
mance fpr New Sources, use a venturi type scrubber; however, numerous
older stationary plants use either a simple spray tower or a cenrifugal
washer. In some cases, the spray tower scrubbers in-use have either
been designed or extensively modified by the operators.
In the drum mix process, the aggregate drying and mixing with asphalt
are both accomplished in the rotary drum. The asphalt is injected at a
point midway down the dryer where the gas temperature should be suffi-
ciently low to prevent significant volatilization of the asphalt components,
However, the injection point should not be too far removed from the
firing end, or the gas temperature will be too low to permit proper
coating of the aggregate and the resulting product will not meet customer
specifications.
The particulate matter emitted from drum mix plants is significantly
different from that of conventional hot mix plants. Emissions from drum
mix plants, especially if the injection point is too close to the flame
zone, will include a portion of condensed hydrocarbon particles.
These can be significantly smaller in size than the entrained aggregate
itself and the material may be less wettable due to the organic nature
4-41
-------
of the emissions. A similar problem can result when a drum mix plant
is used for the recycling of asphalt. In this case, the old road bed
material is ground and added through a chute near the middle of the drum.
Volatilization of the asphalt materials can lead to a very difficult to
control, partially organic particulate matter.52
When there is partial volatilization of the asphalt binder, the wet
scrubber baseline data may be misleading. Due to the large number of
very small particles, the emissions may be substantially higher than
during the baseline period even though there has been no significant
change in the observed pressure drop or the liquid to gas ratio. If the
plume residual opacity is high and there has been no significant change
in the scrubber operating parameters, the inspector should inquire about
changes in the asphalt binder injection.
Typical operating conditions for the wet scrubbers used in asphalt
plants are listed in the following Table 4-12. The range pf pressure
drops is quite large. This is partially due to the variable quantities
of organic material volatilized (drum mix plants) and the variable particle
size distribution resulting from the drying of the aggregate itself (drum
mix and hot mix batch). The latter is due to variations in the friability
of the aggregate which can breakdown in certain plants to form a high con-
centration of fines and the presence of clay fines in certain aggregates.
Both act to substantially increase the amount of small diameter dusts
entering the scrubber. The process and raw material variables discussed
TABLE 4-12. TYPICAL OPERATING CHACTERISTICS OF ASPHALT PLANT
PARTICIPATE WET SCRUBBERS53,54,55
Parameter
Liquid-to-Gas Ratio,
Gals/1000 ACF
Pressure Drop, Inches H20
Inlet Gas Temperature, °F
Hot Mix
Spray Tower Scrubbers and
Centrifugal Scrubbers
3-10
2-6
325-425
Venturi
Scrubbers
5-10
15-25
. 325-425
Drum Mix
Venturi
Scrubbers
5-10
15-30*
225-350
*Asphalt Concrete NSPS Background Document (EPA) recommends a venturi
scrubber with a pressure drop of at least 20 inches of water.
4-42
-------
above underscore the importance of evaluating each system on a site-
specific basis using the baseline ;data. The data presented in Table 4-12
is not intended to be applied as performance criteria for a specific
asphalt concrete scrubber.
Regardless of the type of asphalt plant, the size of the facility is
usually quite small. The range of gas volumes is 15,000 to 50,000 SCFM,54>55
well below that for wet scrubbers in other applications. Due to the size
of the plants and the intermittent operating schedules, there is rarely
an engineering staff available to assist the operator in diagnosing
operating problems. Furthermore, many of the wet scrubber systems at
these plants have no instrumentation for determining pressures, liquid
flow rates, and gas temperatures. Another problem common to asphalt
concrete plants is reentrainment.
Many of the plants construct a make-shift pond for settling of the
solids collected. In some cases, inadequate ponds or inadequate pump
intake designs result in excessive solids pick-up. This can lead to
pluggage and erosion in the scrubber. Ponds sho.uld be separated by several
overflows in order to improve settling. General design criteria for
settling ponds are listed below.. _
Number of zones =3
Volume =1/2 day circulating pump capacity
Depth = 6 feet (minimum)
Width = twice the length
Additional design criteria are discussed in reference 56.
Scrubbers used on hot mix plants burning natural gas have few problems
with corrosion; however, those burning oil with a high sulfur content can
have low pH problems, especially if the recirculation rate is high. Corro-
sion of the scrubber can also be caused by mildly acidic ground waters in
certain locations and by acidic components in the aggregates being dried.
It is important to check the liquor pH frequently and neutralize with
hydrated lime or caustic soda, if necessary.
4-43
-------
4.3.3 Grey Iron Foundries
A grey iron foundry cupola is a firebrick, refractory lined cylin-
drical furnace used for the melting of scrap to produce grey iron. The
cupola is alternately charged with coke, limestone and soda ash, and
scrap. The effluent gases pass through the top of the cupola and exit
at temperatures as high as 1800 °F. The entrained particulate matter is
partially composed of incompletely combusted coke and entrained charge
materials. Because of the high temperatures involved, there is a sub-
stantial quantity of submicron fume and condensed organic particles. The
gas stream also contains very high concentrations of carbon monoxide (CO).
The concentration of CO can vary from as low as several percent to as
high as 25% depending on the iron to coke ratio charged and the degree of
maldistribution of the material across the cross section of the cupola.57
The extremely high negative static pressures present near the I.D. fan,
can cause severe air inleakage. If this inleakage results in inadequate
exhaust of the cupola, then some discharge may be evident from the cupola
top. Fugitive leaks of exhaust gases during periods of positive pressure
operation can pose a risk to inspection personnel because of the high
concentrations of CO and particulate matter present in the gas stream.
Due to the batch type operation of cupolas, their effluent conditions
are not constant. Thus, stack test data taken at one point in time may
not be representative of other stages of a heat. Significant differences
in the particle size distribution, the carbon dioxide/carbon monoxide
ratio, and the gas temperature may occur making it very difficult to
prepare a baseline data set. The cupola conditions must be well docu-
mented in order to make the baseline data reasonable.
Venturi scrubbers are the most frequently used for control of cupola
emissions. These are necessary due to the small particle sizes involved.
Often the pressure drop of the scrubber must exceed 40 inches w.c. in order
to achieve an adequate removal efficiency. In some cases the required
pressure drop exceeds 120 inches of water. At the typical pressure
drops of cupola venturi scrubbers, the outlet static pressures (between
the scrubber and the fan) are in the range of -25 inches to -125 inches.
4-44
-------
Care must be taken to avoid very large errors in the measurement of the
pressure drop due to the aspiration effect discussed earlier.
As shown in Figure 4-19 the effluent concentration from the scrubber
can be a function of the scrap quality. The increased emissions with
oily scrap is probably due to increased concentrations of organic vapor
which condense in the scrubber to form a very small, difficult to wet
particulate material. Therefore, pressure drop is not acceptable as a
single indicator of performance. As in other scrubber applications,
the process operation and material characteristics must be taken into
account.
cc.
UJ
o
z
o
o
en
o
t-
Ul
_J
fe
o»
0.4
0.3
0.2
O.I
O.O5
'CLEAN SCRAP
0 10 20 3O 40 5O 6O 70 8O 90 100
SCRUBBING PRESSURE DROP, inches-water gauge
Figure 4-19. Pressure drop collection efficiency relationship.
Typical operating conditions for the wet scrubber used on cupolas are
listed in Table 4-13. It appears that basic cupola design characteristics
do not have a significant effect on emissions or specific melting rates.
The use of briquettes increases emissions. The blast rate also has a
significant effect causing greater emissions and smaller particle sizes
as the blast rate increases.
4-45
-------
TABLE 4-13. TYPICAL OPERATING CHARACTERISTICS OF PARTICULATE WET
SCRUBBERS ON GREY IRON CUPOLAS58
Parameter
Values
Liquid-to-Gas Ratio, Gals/1000 ACF
Pressure Drop, Inches H20
Gas Inlet Temperature, °F
Liquor Solids Content. % by weight
10-25
25-45
160-200
1-10
4.3.4 Coal-fired Boilers
Particulate wet scrubbers used on coal-fired boilers are generally
found in conjunction with S02 absorption systems. The most common
types of particulate scrubbers in this application include variable
throat Venturis (e.g. plumb bob, venturi rod) and moving bed scrubbers.
This type of installation is large, with gas flows exceeding 106 ACFM;
typical operating conditions are presented in Table 4-14. In most cases,
there are several scrubbing trains in series so that changes in boiler
•load can be accommodated. Due to the high cost of such systems, there
is almost always reasonable instrumentation including pH monitors, static
pressure gauges, and liquor flow rate meters.
TABLE 4-14. TYPICAL OPERATING CHARACTERISTICS OF PARTICULATE WET SCRUBBERS
ON COAL-FIRED BOILERS59*60
Parameter
Liquid-to-Gas Ratio, Gals/1000 ACF
Pressure Drop, Inches H20
Gas Inlet Temperture, °F
Liquor pH
Venturi
(all types)
16-30
3-20
280-450
6-8
Moving
Bed
50
10-15
280-350
6-8
The particle size distribution and mass loadings from a given boiler
are a function of the fuel quality and the boiler firing conditions (such
as excess air rates). Compared with other types of sources, the gas
4-46
-------
quantities and the particle size characteristics are relatively stable.
Thus, scrubbers at pulverized coal-fired boilers are suitable candidates
for the development of relatively simple baseline performance curves.
Two of the principal problems reported from these types of scrubbers
are corrosion and scaling, both of which are partially dependent on the pH
of the recirculation liquor. Operation of the particulate scrubber at
pH levels below 6 can lead to severe corrosion of carbon steel components.
Scaling results from precipitation of calcium and magnesium compounds
which normally happens at pH levels of 9 and above. One factor of impor-
tance in this phenomenon seems to be the degree of oxidation of sulfite
to sulfate in the liquor.
Recent water quality requirements have resulted in greater recircu-
lation rates than previously would have been considered. This, in turn,
has increased (1) the risk of corrosion due to the concentration of chlo-
rides and other corrosive agents, (2) the risk of erosion due to the
higher solids levels, and (3) the risk of pluggage.
High opacity conditions have been observed at a number of wet scrubbers.
which"appear to be operating satisfactorily. The light scattering aerosol
formation is not well understood, however, it may be due to condensation
of vaporous material formed in the boiler. The most likely compound is
sulfuric acid vapor.61
4.3.5 Municipal Incinerators
During the last ten years, the number of municipal incinerators in
operation has been gradually decreasing due to the increased use of
landfills. Nevertheless/ a number of wet scrubber systems are used
for particulate control on the incinerators still in operation. The
types of scrubbers commonly found in this application include impingement
plate scrubbers and various types of Venturis. Typical operating charac-
teristics for these venturi systems are presented in Table 4-15.
One of the major problems of wet scrubber operation as applied to
municipal incinerators is control of the condensible, partially oxidized
organic matter which forms when the incinerator operating temperature is
too low. This material condenses while passing through the scrubber.
The final particle sizes are too small for effective impaction in the
4-47
-------
scrubber, and therefore, collection efficiency is low. Few scrubbers
have been able to achieve the NSPS limits because of this problem.
The residual plume from these units is generally a bluish white. Due to
the high chloride content of the material charged, corrosion of the
scrubbing systems applied to incinerators can be severe.
TABLE 4-15.
TYPICAL OPERATING CHARACTERISTICS OF MUNICIPAL INCINERATOR
VENTURI SCRUBBERS62,63,64
Parameter
Venturi Scrubber
Liquid-to-Gas Ratio, Gals/1000
Pressure Drop, Inches F^O
Inlet Gas Temperature, °F
Liquor Solids Content, % by weight
4-8
15 - 60
300 - 500
1 - 10
As with other applications of scrubbers, some process operational
data must be evaluated to determine if there has been a significant shift
in the baseline conditions. The incinerator operating temperature and
the charging rate should be checked at a minimum. It may also be necessary
to determine if there has been a change in the types of materials charged
to the incinerator.
4.3.6 Basic Oxygen Furnaces
Venturi scrubbers have commonly been used to control the fine parti-
culate emissions from basic oxygen furnaces (BOF's). This application is
of interest due to the very high static pressure drops involved and the
cyclic nature of the gas flow rate over the 30 to 45 minute operating
cycle; the latter raises a question concerning the definition of pressure
drop.
The basic oxygen furnace is a batch reactor in which up to 350 tons
of pig iron and scrap are converted to steel by the oxidation of the
carbon, phosphorus, silicon, and magnesium impurities. The materials
charged include a minimum of 70% molten iron, 10% to 30% scrap, and
limited quantities of flux materials such as lime and fluorspar.65 The
offgas from the BOF consists of 75% to 90% carbon monoxide and the gas
temperature is approximately 3000°F before dilution .
4-48
-------
The gas evolution rate varies substantially during the oxygen blow
period. An example gas flow rate versus time curve is presented in Figure
4-20. Since the pressure drop across a venturi scrubber is a function of
the gas flow rate, the pressure drop will also have major variations
during the production cycle. To be explicit, the time must be specified
along with pressure drop value. The typical operating conditions of
venturi scrubbers and quenchers serving BOF's are presented in Table 4-16.
10-MINUTE RECOVERY
67% VOLUME (215 Btu/cf) = 533,594 scf
OFF-GAS FLOW RATE
% CO
10 12 14 16 18
Figure 4-20. Gas flow rate and CO content during a blow.22
The particulate matter entrained in the gas stream is very fine with
85% to 95% of the mass smaller than 1 micrometer in diameter and as much
as 20% of the mass less than 0.5 micrometers.66 The venturi scrubbers
are operated at very high pressure drops to achieve reasonable control
efficiencies. The typical range is 40 to 70 inches of water.22»65 Due
to these high negative static pressures, pressure drop measurements must
4-49
-------
be made carefully. The particle size distribution generated by the
process and consequently the pressure drop necessary, is a function of
the raw material charge characteristics, the rate of oxygen blow, the
type and quantity of ladle additions, and the condition of the refractory.
A quencher is used to decrease the gas temperature prior to entry to
the venturi scrubber. As much as 80% of the particulate matter is removed
in the quencher, therefore, the pressure drop across this collector is also
of interest.
TABLE 4-16.
TYPICAL OPERATING CONDITIONS OF VENTURI SCRUBBERS
SERVING BASIC OXYGEN FURNACES22*65.66
Parameter
Liquid to Gas Ratio,
Gallons/1000 ACF
Pressure Drop, Inches 1^0
Gas Inlet Temperature, °F
Venturi Scrubber
25-50
50-70
120-150
Quencher
Variable
5-10
3000
4-50
-------
5.0 EVALUATION OF OPERATION AND MAINTENANCE PROCEDURES
This section addresses basic operation and maintenance procedures
for wet scrubber systems. Since there is considerable diversity in
the design of scrubber systems and in applications, there is no one
set of operation and maintenance procedures which is adequate in every
case. Instead, it is necessary to tailor the procedures to the
specific site involved; this is an evolutionary process in which the
initial procedures are modified in response to the specific operating
problems experienced at the site.
The procedures presented in this section represent a skeleton
of the initial operation and maintenance requirements for a wet
scrubber system. For some systems, these procedures include steps ,
which are unnecessary and for some systems, important maintenance
items may not be included. Therefore, regulatory agency personnel
should not conclude that the following procedures necessarily consti-
tute a complete description of good operating practice.
5.1 STARTUP AND SHUTDOWN PROCEDURES
Proper startup and shutdown procedures are important in preventing
(1) damage to the corrosion and abrasion resistant lining in the scrubber
(if used), (2) overheating of the fan motor, and (3) damage to the pump
impeller and motor. Each scrubber manufacturer provides instructions
for proper startup and shutdown of its system which should be strictly
followed. To illustrate the major elements, a condensed version of
these procedures is provided in the two lists that follow.
Routine Startup
o Inspect piping and valves and open or close valves to permit proper
flow of fluid from the system to recycle tank, pumps, and return lines.
o Inspect ductwork system to assure that all ductwork leading to the
scrubber is intact and that damper blades move with damper control
activation. Flow of exhaust gas to scrubber should not be initialed
until liquid circuit has been fully started. Inspect and test in-
struments to be sure they are operating properly.
o Make sure that the unit is level (especially important for tray-
type scrubbers).
o Turn on all electrical switches for controls and instruments.
5-1
-------
o Fill system with scrubbing liquid by opening plant water lines
to recycle tank, allow it to fill to proper level, and then
start recycle pump.
o Open liquid line to quench/presaturator sections and adjust to
proper flow.
o Open damper to direct exhaust gas flow to venturi scrubber and
start the fan.
Prior to the startup of a wet scrubber, a complete internal inspec-
tion should be conducted. A partial list of the items to inspect in-
clude: the condition of the spray nozzles; the condition of the supply
lines to the scrubber; presence of internal deposits; presence of worn
or eroded linings; presence of partially plugged or warped demisters;
and presence of corrosion on the interior trays, downcomers, and other
components. For some types of scrubbers, it may be advisable to visually
confirm that the inlet liquor flow is sufficient to completely wet the
venturi convergent section to avoid any wet-dry interface problems.
During startup, the bypass duct (if any) should be gradually closed
as the inlet damper is gradually opened. If the fan inlet damper is opened
too quickly, it is possible to overload the fan motor.
After operating conditions have stabilized, a complete set of readings
should be taken and compared against the baseline operating levels. It
is advisable to conduct a routine evaluation, as discussed earlier, to
confirm that the system is operating properly.
It may become necessary to adjust the liquid flow rates to maintain
the recycle tank level. This requires balancing the water make-up rate,
the purge rate, the sludge removal rate, and the recycle rate.
Routine shutdown
o Turn off the fan and close the inlet or outlet damper. Open the
bypass damper to ensure that the scrubber is isolated from any
process gas flow.
o Continue pump operation and all internal water sprays until the
scrubber system has cooled.
o If the system will be exposed to cold weather conditions, or if
an extended shutdown is anticipated, all pipes should be drained.
5-2
-------
o An internal inspection should be conducted to determine if there are
eroded or plugged nozzles, worn (corrosion or abrasion resistant)
liners, damaged demisters, or other problems which warrant
.maintenance attention prior to startup.
o All deposits which could accelerate corrosion of the scrubber
should be washed out.
During shutdown, it is particularly important not to expose the scrub-
ber system to any high temperature excursions. For this reason the pumps
and internal sprays should be operated until well after the scrubber has
been isolated from the process. It is also important to continue the
alkali additions to the system long enough to prevent low pH excursions
after shutdown.
5.2 ROUTINE MAINTENANCE
This section presents general information concerning the operation
and maintenance of wet scrubber components; specific maintenance require-
ments vary substantially depending on the design of the scrubber and its
application.
5.2.1 Fans
The types of operating problems experienced with fans depend on
whether the fan unit is before or after the scrubber. If the fan is in
the upstream position where the particulate mass concentration is high,
erosion of the fan wheel is of concern. Due to the high gas temper-
atures prior to a scrubber, it may also be necessary to cool the fan
bearings. If the fan is in the downstream position, it is vulnerable to
the buildup of solids and corrosion.
Regardless of location, all fan wheels should be visually inspected
on a regular basis. A typical frequency is monthly, although in some appli-
cations weekly inspection is warranted. If the fan vibration increases
noticeably, the fan wheel should be inspected as soon as possible.
Fan bearings require frequent inspection; they should be checked
for oil level, oil color, and operating temperature. Bearing wear can
cause vibration problems. The bearing housing and mountings should
also be checked on a weekly basis.
5-3
-------
On belt-driven fans, the belt tension should be checked on a weekly
basis. Loose belts are prone to a distinctive squealing which suggests
an unintentional reduction of several hundred R.P.M. and a decrease in
the gas flow delivered by the fan. The rotational speed of the fan can
be checked using a manual tachometer, a phototachometer, or a strobo-
tachometer. Use of these instruments is discussed in the paper by Richards
and Segall.35 Fans should never be operated at tip speeds greater than
those recommended by the fan manufacturer.
5.2.2 Pumps
The most common pump related problems include erosion of the impellers,
erosion and chipping of the liner, and leakage of the pump glands. All
pump bearings should be lubricated regularly and inspected on a frequent
basis. The actual schedule depends somewhat on the severity of the
application, with more frequent attention required if the total solids
content of the liquor is high and/or the pH is low. High wear pump
parts should be kept on hand since delivery times are sometimes long. If
possible a spare pump should be available.
5.2.3 Valves
The operational status of all control valves should be checked on
at least a weekly basis. They are vulnerable to sticking, pluggage, and
erosion. Air activated valves are also subject to freezing if the com-
pressed air supply is not properly dried.
5.2.4 Nozzles
The frequency of nozzle inspection varies with the type of scrubber
and the quantity of suspended solids in the recirculation liquor. If
the suspended solids level is greater than several percent, erosion
and/or pluggage are probable. As an initial check, the nozzle operating
pressures should be recorded on a daily basis. An increase or decrease
in the operating pressure without a corresponding change in the liquor
flow rate strongly suggests that the nozzle condition has deteoriated.
If pluggage is suspected due to an increase in the nozzle pressure,
the nozzles.should be rodded out (if possible) and then the presssure
rechecked.
5-4
-------
When the scrubber is down, nozzle conditions can be evaluated
by an internal inspection or by observing the spray angle while the liquor
recirculation system is operating. A high intensity (explosion proof)
flashlight is useful for checking the individual spray nozzles. It is
generally advisable to keep a set of replacement nozzles when they are
used in severe applications.
5.2.5 Recirculation Tank
The gradual accumulation of sludge at the bottom of the recirculation
tank is undesirable. Whenever the system is down for maintenance, the
quantity of the deposits should be checked either by draining the system
or simply by the use of a rod. Draining the tank is preferable since
it is possible to check for corrosion and erosion when the tank is empty.
If a mixer is used in the recirculation system, the condition of the
impeller should be checked on at least a monthly basis.
5.2.6 Dampers . - • '
It is advisable to exercise all dampers at least once a month to
confirm that they have not "frozen" in the open position. The damper
drive mechanism and bearings should be lubricated in accordance with the
manufacturers specifications.
If a purge air blower is necessary to keep the damper seating
surfaces clean, this should be checked on a daily basis to ensure that
the blower is operating.
5.2.7 Instruments
Instrumentation systems are particularly prone to operating problems
and therefore should be checked on a daily basis. The status of a pH
control system can be checked using either a portable pH meter or pH
paper. It is important to take these measurements at approximately the
same location as the pH probe of the instrumentation system, since pH
can vary substantially throughout a wet scrubber system. Deviations
from the baseline level and/or a difference in the two readings suggest
either a problem with the installed pH meter or failure of the alkaline
additive delivery system. The most common pH meter problems are scaling
scaling of the probe and breakage of the probe.
5-5
-------
5.2.8 Motors
The motor bearings and the motor starting switches should be lubri-
cated on a routine basis. The manufacturers' guidelines should be followed.
All contacts, cooling fans, and other components should be inspected at
least once per year.
5.2.9 Demisters
\
All chevron and meslr pad type demisters should be inspected on at
least a weekly basis to prevent excessive solids buildup. Gradual scaling
of the demister will lead to an increase in local gas velocities through
the open areas and this in turn will result in entrainment. Presence of
solids on the demister can be determined by using a high intensity
(explosion proof) flashlight. The condition of the demister flush nozzles
should be evaluated at the same time. Any deposits should be removed
as soon as possible using either the existing spray system or an external
hose.
5.3 ELEMENTS OF MODEL RECORD AND REPORT SYSTEM
A comprehensive recordkeeping system addressing the operation, main-
tenance, and performance of the wet scrubbers at a particular source can
be of valuable assistance to both the control agency inspector and the
plant manager. The information contained in such a system can provide
the inspector with a preliminary indication of the overall condition and
performance of the control equipment. It can also direct the inspector's
attention to specific problem areas that may require special attention
during the inspection.
The records will allow the operator to ascertain the effectiveness
of his operation and maintenance program, as well as determine the extent
of deterioration of control system components. The latter benefit can,
in a well-maintained recordkeeping system, minimize the time that the pro-
cess and control equipment is out of service due to repairs. In addition,
significant cost savings can be realized by diagnosing and repairing
major equipment failures before they occur.
For either the inspector or the plant manager to benefit from a
recordkeeping and report program, there are several important aspects
of the use of the program that should be realized. For the program to
5-6
-------
be effective, recordkeeping should routinely executed on a daily basis.
Some advance warning is usually given prior to equipment failure; how-
ever, the time between the initial indication of a problem and the re-
sultant failure can be very short. The earlier a problem is discovered
and rectified, the greater the reduction in repair time and cost.
A recordkeeping program should never be considered an entirely inde-
pendent system on which decisions are based. The data contained in the
records can, for many reasons, be inaccurate or misleading. The major
cause of such problems is that the data are obtained from faulty instru-
ments. Thus, agency inspectors and plant operators should always conduct
a physical inspection of the equipment to verify the accuracy of instru-
ment data. Quality assurance is another important part of such a program,
particularly in ensuring that the data are that useful.
The recommended recordkeeping practices that follow are intended to
provide an overview example of elements of a model record and report pro-
gram. The items that are discussed are not for a particular scrubber, but
rather encompass many different scrubber types and applications. Each
recordkeeping system should be tailored for the specific scrubber configu-
ration and the process to which ft is applied. In tailoring a recordkeeping
system, applicable elements from this model program should be included,
as well as any others which may apply to the specific situation.
5.3.1 Scrubber Operation Record Elements
There are certain data common to all scrubber types which should be
included in any scrubber recordkeeping system. These data elements are
listed in Table 5-1. These routinely measured parameter values, compared
to the baseline values obtained during a compliance test (or the initial
performance evaluation test), can provide a very good indication of the
performance of a scrubber. In addition, examination of these parameters
over time can aid in the detection of component deterioration in the
scrubber system.
It is recommended that whenever possible, the scrubber operation data
be obtained using portable instruments. Each tap hole through which a
measurement is made should be cleaned prior to every measurement. Because
the plugging of tap holes occurs so frequently, cleaning of the holes is
5-7
-------
TABLE 5-1. SCRUBBER OPERATION DATA
Inlet Gas Temperature
Outlet Gas Temperature
Total Static Pressure Drop
Static Pressure Drop of Mist Eliminator
Liquor Feed Rate
Liquor pH
Water Makeup Rate
Fan Current
Fan RPM
Fan Gas Inlet Temperature
Nozzle Pressure
Pump Discharge Pressure
Recycle Bleed Rate
Chemical Addition Rate
Liquor Solids Concentration
important to ensure that a partially or completely plugged hole does not
result in an erroneous measurement. This, in fact, is one of the reasons
that portable instruments should be used rather than fixed gauges. Too
often a reading from a fixed gauge will be recorded without checks to see
that the gauge's tap hole is not plugged. Regardless of whether the instru-
ments are fixed or portable, each must be calibrated at intervals which
are at least as frequent as the manufacturers' specifications.
The measurement of the inlet and outlet gas temperatures can provide
an indication of problems in the heat exchange mechanics of the scrubber.
For instance, abnormally high outlet gas temperatures suggest that the
scrubber liquor flow is below normal or the gas flow rate is well above
design value. In either case, the liquor-gas contact within the scrubber
is less than optimum, which results in a lower particulate removal effici-
ency. A low outlet gas temperature can indicate a low gas flow rate
through the scrubber or an inleakage of ambient air.
Pressure drop measurements can confirm problem areas identified
by the temperature measurements. They can also indicate other potential
problems. Low pressure drops across the scrubber are indicative of little
or no liquor flow to the scrubber and low gas flow rates through the
scrubber. In ventuni scrubbers, a low pressure drop can be caused by
the venturi throat being out of adjustment, while the cause in a tray-
type scrubber may be the collapse of the impingement tray. A high
5-8
-------
pressure drop on a packed bed scrubber is usually the result of the bed
becoming plugged. Pressure drop increases are associated with unexpected
increases in gas or liquor flow rate.
A determination of the gas flow rate through the scrubber can be
made using information collected from fan measurements. It is very
important to know if the scrubber is operating in the flow rate range
for which it was designed. If the necessary baseline data were collected
during a previous performance test, the fan current can be used to estimate
the present flow rate.
The measurement of the scrubbing liquor pH can provide insight
into operational problems, as well as indicate potential damage to the
scrubber structure. A low pH reading usually indicates loss of additive
feed or low liquor flow rates. Since these rate measurements are included
in the routinely collected scrubber data, the cause of the problem can
be easily determined. Low pH measurements should serve as a warning
that,corrosion damage to the scrubber shell may have occurred, and thus,
a visual, internal inspection of the scrubber should be conducted.
Low nozzle operating pressure is usually a result of low liquor
flow rate or erosion of the nozzles. If nozzle erosion has occurred,
the droplet size emitted by the nozzles will be larger than the design
size. Also, low liquor flow will produce altered droplet size distribu-
tions. Either situation will result in decreased collection efficiency.
Since the liquor flow rate is a routinely measured parameter, an increase
in the flow reading should prompt a visual inspection of the nozzles for
erosion; at this time, the valves should also be examined for erosion.
A decrease in the flow can indicate pump impeller wear or liquor line
pluggage. The recycle bleed rate measurements can show bleed line
pluggage when the flow is reduced and valve wear when the flow increases.
The various combinations of pump motor current and discharge pressure
can be used to identify several probable problems. For instance, a.de-
crease in both water pressure and amperage draw can mean that there are
nozzles missing, significant pump wear, or suction line plugging. Other
combinations suggest nozzle or -spray bar plugging and manifold or spray
bar leakage due to holes. The pump motor current can also be used to
determine if a decrease in scrubber liquid flow is caused by an equipment
5-9
-------
maintenance problem such as a plugged line, by a worn pump impeller, or
in fact, may only be a result of an improperly calibrated flow meter.
The remaining scrubber operation data which should be monitored are
liquor solids concentration and chemical addition rates. Monitoring the
solids concentration can ensure that the desired collection efficiency
is maintained and scaling is minimized. It also helps reduce abrasion
in the liquor contact points throughout the piping and scrubber system.
In scrubbers which use chemical additives to control the liquor pH or the
concentration of dissolved solids, measurement of the rate of addition
is important in reducing the problems created by corrosion and scaling.
5.3.2 Process Record Elements
In addition to the scrubber data, there are certain process data that
should also be routinely recorded. These data are shown in Table 5-2.
Since baseline scrubber data are used as standards on which to judge sub-
sequently measured scrubber parameters, it is equally important to know
how the daily process data compare to the baseline values. Process vari-
ations in feed types and rates can effect the collection efficiency of
scrubbers by altering the character of the inlet particulate gas stream.
Of particular importance in this regard is the particle size distribution
at the inlet, a parameter on which greatly affects scrubber performance.
TABLE 5-2. PROCESS CONDITION DATA
Process Feed Rate
Percent Load Capacity
Process Feed Descriptor
The ancillary equipment in a scrubber system is also often responsi-
ble for many of the operational problems. This equipment includes the
fans, pumps, motors, ductwork, piping, valves, clarifiers, and instrumen-
tation. Failure of any of these components can be disastrous not only to
the performance, but also to the longevity of the entire scrubber system.
In order to facilitate the recording of the scrubber parameters pre-
viously discussed, a suggested record format is presented in Table 5-3.
5-10
-------
The daily, monthly, and semi-annual performance logs for the scrubber's
ancillary equipment are provided in Tables 5-4, 5-5, and 5-6, respectively.
As with any other operation and maintenance record system, space is pro-
vided to explain any corrective actions which were initiated to resolve
any abnormal observed values. It is highly recommended that all data be
collected on a routine basis and compiled in a notebook used only for
recordkeeping purposes.
5-11
-------
LL.
O
GO
ou
O
0)
c_
o
o
tD
O
LU
1
s
Ul
Ou
O
LU
CO
CO
Ol
•a
0)
S_
CD
C/)
O
CO
OJ
CO
in
LU
CD
O)
fd
CO
ns
c:
Q.
O
£_
UJ
0)
UJ
t_
Ol
•»-»
Ol
ro
c_
CO
D_
c
o
CD
Q.
o
CL)
£_
3
•*->
ro
CL)
CL
CL)
CO
ro
CD
CD i—
CL)
£_
3
4-)
ro
t_
QJ
Q.
F
CL)
1—
CO
CO
CD
4_>
CD
i~~
3
0
0)
£_
3
in
to
CL)
<_
, CL.
O
•r-
40
re
-M
oo
r—
ro
o
1—
Q.
O
£_
Q
CL)
£_
3
CO
to
CL)
c_
Q-
C_
CL)
ja
ja
3
C_
O
00
•^
^v^
Q.
O
£_
Q
CL)
{_
3
to
CO
0)
£_
Q-
O
•n"
4->
ro
4J
OO
ry
C
O
•r—
ro
r^l
3
O
£_
• p—
O
CL)
Qi
J_
O
i 3
cr
_l
m
cu
e
O
3
cr
i
CL)
^
(O
CD
'r- •!- n3
Nozzle Pressure
0)
(O
o;
CL)
en
3
0-
O)
fO
o;
Add
Chemica
01
CD
c
•!"•
(O
o
occulant
•»->
C
ro
4-J
o
ro
4-
c_
3
oo
£_
3
£_
O
3
cr
5-12
-------
u_
1 — 1
IE
00
O3
O
1
_I
UJ
o
121
eC
s:
0
(i
rv 1 1 1
UJ b-'
Q. <:
Q
I—
2:
UJ
^Z"
•CL.
Q_
h-H
^3
o-
UJ
^~
— 1 O
_1 2:
1 — 1
0 h-
2: 1— 1
^f* ' —yr
^•J* <^—
UJ
CQ
CQ
o;
O
CO
UJ
UJ
•
1
ir>
UiJ
_i
CQ
1—
UJ
s:
e^
^
i—
^^
•r—
-M
O
t_
£_
O
o
(O
£_
o
c:
ja
£_
O
L«
O)
4^
QJ
E
(O
^.
fO
D_
C
o
•i—
03
L_
O)
OL
0
..
UJ
CO |—
CD CO
2: CO >^
i— i t/> t£3 CO
ct: CT) 2: . c
«£ c: I-H uj o
UJ OJ -i- C£. O3 -r-
CQ t- £_ eC Q; +J
rs ro co uj ^> to
2: 4-> Ol CD CQ Q. £_
«C to CQ c: 2: caj
u_ s_ o >— i co co ctoaiO-
Ci — £_O)£-'i- Qi i — _^C O •«- >— I O
Q o a) o Q- o -t-> «=c ojroo i — +-> =t
2:-i->i — E+J(OLU >d)T-O 03 <*-
•=C 4-»aJotuoocQ a) i +•> S <-> a; o
(O i c_5 ^— ^" (p— t fQ >r— 1 1 i
cos- c_o_ i— £_ a_ £_ a_ a>
2: JDI — i — i — cj2 s: i — ooou_ — i a. • o co > ci- i c^ uj
5-13
-------
CD
O
O
o
U.
a:
ui
a.
uj
LU
oi
«C
O
UI
OQ
CO
1
CO
in
in
ca
<:
u.
•— «
3C
CO
UI
a
o
\-
=
tJ-1
5
^
r-
_i
a.
0
•r™*
CJ
E
to
c_
(O
Ou
C
0
+j
(O
cu
o.
o
en en
c c
+J CO £
iZ =C iZ
2C
O) O O)
CO UI CO
O c o en c:
— 1 UI CU OJ 1 C O
>• in E T-T-
•>«•!- CCO«> X 4->
co o; o; £_ c: cncDco cu • CU CQ O CU 'i— CO *+— CU
CQco a; o c _l o oi wT (Ocoui ro
^liCS O.OT- OO-iil — JiiCUQ- CUjKi
*ZZ. CO ^ E t- CO r— C_ I — CO O CO O y CO CO
«=C CU Di CU O. x: -r— Q. O CU^> CUXej; COCU
Ll- _l Q t—COOOCO 2! —1 Q _IUI Q Ul_l
CU
13
c.
•r™
-p
c:
o
o
5-14
-------
1 1
U-
1 — 1
*~^ "T~
•o oo
CL)
3
f"»
+J
C
O
o
— '
o
_l
UJ
o
2: uj
s: «=c
fV I— I
O
.1 1
UJ
Q.
1—
^
UJ
Q-
1 — I
o- o
UJ Z
"^— |L_J
?- (—
Qi 1—4
«=c 2:
_l
1 — 1
•z.
«=c
C£
UJ
CO
CO
Qi
O
.00
>-
_i
rn
^^*
0
•
LO
I
LO
UJ
1
CO
•=C
LU
1—
•z.
. 1
o.
o
•r—
(J
eC
O)
O
0)
^_
0
o
(O
E
O
c.
-Q
ID
O
C
o
-!3
-o
o
C_5
£_
0
O)
1 ^
cu
(O
£_
(C
Q.
C
O
Q.
C3
-:
I
' o;, a.-
UJ 3
oo . ea • -a ' -
Q ' ' . 02 r— . . '
M ° 0£ '3 ' .
1 - C t — 1 1 i 1 (Sy
> — i Oi =C ^ n ni
Q- 01£_2: eC OO)
OO 3 IO CD i t
a; cc: i— cu en >- c
1-iJ OleC Q-3 'Z. CD UJ >> Q - i — CO
'-' =: CQ <: - c l 01 ' o; O0(0o-i-
U_ ••— O) CD -i- c: =3 ' rd ' ca • *i— >r- in
"— ' Ol>- i — i — - OO • Ol «r- OO i. t-(/lO
C^ Olc=C IMMUJ Ol_i£C/0 3S:O)(O£_
•^C 3 Q; Nl N] CL. H3 (TJ UJ O t— l J_> £_ £_
— 1 i — Q- OO i— i , — OJ QiO et (T3J2O
0 Q- 00 -Z. - -Z. Q- Q-_l Q- CC S S ^ -0
5-15
-------
t
t— 1
nr
CO
§
_j
UJ
o
z:
o-
UJ
>-
3
•-J *
I— I O
o 2:
CO
CD
cm
O
CO
|
2:
•a:
I
1— <
UJ
CO
•
CO
1
to
UJ
_J
CD
•a;
r-
UJ
«*T^
^"
h-
z:
^c
_j
cu
c:
o
•r—
O
•"•^
CU
'•p
0
cu
£_
£_
O
0
t
re
£_
o
c
jQ
r_
re
E
o
•z.
c
0
*r™
•p
•5
c
o
o
c_
o
c_
cu
4->
CD
fO
f^r
(0
o.
,_
o
•r-
+J
(O
c_
cu
o.
CD
*
V
CO
on
UJ
*~^
Q
in UJ co
en o; en
c: c: ;
•i- a; -r-
4J LU P
•r- CD T-
en LI_ en en u_ en
c: c: 03 c c
•i- CU •!- T- CU -r-
t- co £- CO t- co £-
O O O CD O • O O
o 002: o o • o
CO — 1 CO I— I CO — 1 CO
•> «> "•=£ •> "CO «
cncoco cnuj en COCD en
CO CCD-i^ CCQ C .^2: CCO
CD -i- 2= O -r- C T- O i— i T- _l
S cuts S £cup £i -?Jjcu£ < culpuj
Qi O •!- et CJ O -i- «C -P -r- O O LU O -r- CO
•^ C O- 1 1 1 c Q- "T" ro ri c CO C d.
UJ fG CQ •» fO O O re " re rv*
iv^ t— « CO t_ •* »r-- " £— CO rv* f A 1 1 t
re £— o. ^^ re c. CD £_ c_ re _^ CD re c_ cu c.
-^~ QJ re ^> re cu re ^^ ^> ^3 QJ
-------
-a
a>
oo
o
4J
O
<
O)
o
O
cs
o
o
O)
£_
£_
O
o
Qi
O
UJ
«=c
a
LU
ns
O
o-
UJ
C£.
UJ
CO
03
£_
O
o
oo
c:
O
ca
•=c
UJ
<;
O-
O
o
£_
O)
(O
(O
Q.
re
c_
O)
a.
o
00
I—1
o
UJ
UJ
>
>—I
o;
UJ
a.
O
_1
OQ
oo
UJ
Q
CO
CD
CXL
C -(-) UJ
o c en
•r- s • o;
"3 c: uj
(- o> o.
O) -I- S
CL I— <:
O =C Q
(O
(LI
o>
CD
(O
^
03
O)
00
o;
UJ
I—
UJ
(O
o
o
5-17
-------
-------
6.0 REFERENCES
1. Steiner, B. A. and Thompson, B. J. Wet Scrubbing Experience for
Steel Mill Applications. Second EPA Fine Particle Scrubber Symposium.
U.S. Environmental Protection Agency, Research Triangle Park, N.C.
Publication EPA 600/2-77-193. pages 5-24. September 1977.
2. Bete Fog Nozzle Inc. Catalog 81A. 1981.
3. Calvert, S., J. Goldschmid, D. Leith, and D. Mehta. Wet Scrubber
System Study, Volume I: Scrubber Handbook. U.S. Environmental Pro-
tection Agency; Research Triangle Park, N.C. Publication EPA-R2-72-118a,
pages 5-89 to 5-93. August 1972.
4. ' Calvert, S.; Barbarika, H. F.; and Monahan, G.M. Evaluation of Three
Industrial Particulate Scrubbers. U.S. Environmental Protection Agency,-
Research Triangle Park. Publication EPA-600/2-78-032, February 1978.
5. Calvert, S., H. Barbarika and 6. Monahan. American Air Filter Kinpactor
10x56 Venturi Scrubber Evaluation. U.S. Environmental Protection Agency,
Research Triangle Park, N.C. Publication EPA-600/277-209b. November 1977.
6. Calvert, S. "Engineering Design of Fine Particle Scrubbers." Journal
of the Air Pollution Control Associates, Vol. 24, No. 10. pages 929-934.
October 1974.
7. Yung, S., S. Calvert and H. F. Barbarika. Venturi Scrubber Performance
Model. U. S. Environmental Protection Agency. Publication 600/2-77-172.
August 1977.
8. Yung, S., Calvert, S., Barbarika, H., and Sparks, .L. "Venturi Scrubber
Performance Model" Environmental Science and Technology, Volume 12.
Number 4. pages 456-459. April 1978.
9. Yung, S., S. Calvert, and H. Barbarika. Venturi Scrubber Performance
Model. U .S. Environmental Protection Agency, Cincinnati, Ohio.
Publication No. EPA-600/2-77-172. August 1977.
10. Kalika, P. W. How Water Recirculation and Steam Plumes Influence
Scrubber Design." Chemical Engineering. July 28, 1969. pages 133-138.
11. McCabe, W. L. and J. C. Smith. Unit Operation of Chemical Engineering.
McGraw Hill, page 116. 1967.
12. Lapple, C. E. and H. J. Kemack. "Performance of Wet Dust-Scrubber".
Chemical Engineering Progress. Volume 51. pages 110-121. 1955.
13. Semrau, K. T. "Correlation of Dust Scrubber Efficiencies." Journal of
the Air Pollution Control. Volume 10. pages 200-207. 1961.
14. Semrau, K. T. "Dust Scrubber Design - A Critique on the State of the
Art." Journal of the Air Pollution Control Association. Volume 16.
pages 587-5-94. 1967.
6-1
-------
15.
16.
17.
18.
Semrau, K. T., C. L. Witham, and W. W. Kerlin. Energy Utilization by
Wet Scrubbers. U. S. Environmental Protection Agency, Cincinnati, Ohio.
Publication No. EPA-600/2-77-234. November 1977.
Engineering-Science, Inc. Scrubber Emissions Correlation Final Report
to U.S. Environmental Protection Agency. Contract No. 68-01-4146
Task Order 49. May 1979.
Perry, R. H. and C. H. Chilton, Editors.
Fifth Edition. 1973.
Chemical Engineers' Handbook,
Semrau, K. T., C. T. Witham, and W. W. Kerlin. Energy Utilization by
Wet Scrubbers. U. S. Environmental Protection Agency, Cincinnati, Ohio,
Publication No. EPA-600/2-77-234. November 1970.
19. Walker, A. B. and R. M. Hall. "Operating Experience with a Flooded
Disc Scrubber. A New Variable Throat Orifice Contactor." Journal of
the Air Pollution Control Association. Volume 18. pages 319-323.
May 1968.
20. Ranade, M. B. and E. R. Kashdan. Second Symposium on the Transfer
and Utilization of Particulate Control Technology. U. S. Environmental
Protection Agency, Cincinnati, Ohio. Publication 600/9-80-039a.
September 1980. pages 583-560.
21. Nixon, D., and C. Johnson. "Particulate Removal and Opacity Using a
Wet Venturi Scrubber - The Minnesota Power and Light Experience."
Second Symposium on the Transfer and Utilization of Particulate
Control Technology. U.S. Environmental Protection Agency, Reseach
Triangle Park, N.C. Publication EPA-600/9-80-039a. September 1980.
22. Kemner, W. F. and R. W. Mcllvaine. Review of Venturi Scrubber
Performance on Q-BOP Vessel C at the Fairfield Works of the United
States Steel Corporation, Birmingham, Alabama. Final Report to U.S.
Environmental Protection Agency. Contract 68-01-4147, Task 03.
February 1979.
23. Yocom, T. "Problems in Judging Plume Opacity - A Simple Device for
measuring Opacity of Wet Plumes." Journal,of the Air Pollution Control
Association. Volume 13. pages 36-39. 1963
24. Nadar, J. S. and Conner, W. D. "Impact of Sulfuric Acid on Plume
Opacity." Presented at the Symposium on Transfer and Utilization of
Particulate Control Technology. Denver, Colorado. July 23-28, 1978.
25. Hesketh, H. "Atomization and Cloud Behavior in Wet Scrubbers." Pre-
sented at the Symposium in Control of Fine Particulate Emissions
from Industrial Sources. San Francisco, California. January 15-18,
1974, pages 455-478.
26. Woffinden, C. J., Markawski, C. R. and D. S. Ensor. "Effects of
Surface Tension on Particle Removal." Symposium on the Transfer and
Utilization of Particulate Control Technology. U.S. Environmental
Protection Agency. Publication EPA 600/7-79-044c. pages 179-192
6-2
-------
27. Spraying Systems Co. Pollution Abatement Manual. ,No Date.
28. Referenced, page 6.
29. American Society of Testing Methods. Annual Book of ASTM Standards,
Part 31. Water. "Standard Test Method for Surface Tension of Water."
1982.' ••.;-'
30. Sparks, L. and M. Pilot. "Effect of Diffusiophoresis on Particle Collec-
tion by Wet Scrubbers." Atmospheric Environment - Volume 4. pages 651-
660. 1970.
31. Calvert, S. and N. Jhaveri. Flux Force/Condensation Scrubbing.
Journal of the Air Pollution Control Association. Volume 24. pages
946-951. October 1974.
32. Lemon, E. "Wet Scrubbing Experiencee with Fine Borax Dust." Second EPA
Fine Particle Scrubber Symposium. U.S. Environmental Protection Agency,
Research Triangle'Park, N.C. Publication 600/2-77-193 pages 25-34.
September 1977.
33. Czuchra, P. A. "Operation and Maintenance of a Particulate Scrubber
System's Ancillary Components." Presented at the U.S. EPA Environmental
Research Information Center Seminar on Operation and Maintenance of
Air Pollution Equipment for Particulate Control. Atlanta, Georgia.
April 1979.
34. Richards, J. Baseline Inspection Procedures (Draft) Final Report to
U.S. Environmental Protection Agency. Contract 68-01-6312, Task 30.
March 1983.
35. Richards, J and R. Segall. The Use of Portable Instruments. Final
Report to U. S. Environmental Protection Agency. Contract 68-02-6312,
Task 60. June 1982.
36. Schifftner, K. C. How to Check Scrubber Entrainment. Pollution
Engineering, July 1982, pages 38-39.
37. U.S. Environmental Protection Agency, Wet Scrubbers", Section included
in Control Techniques for Particulate Emissions from Stationary
Sources - Volume 1. Report No. EPA-450/3-81-005a, page 4.5-2.
September 1982.
38. Gilbert, W., "Troubleshooting Wet Scrubbers." Chemical Engineering,
October 24, 1977. pages 140-144.
39. Reference 37, page 4.5-32.
40. Reference 37, page 4.5-34.
41. W. W. Sly Inc. Bulletin lOM-2-66. No date.
42. Western Precipitation, Inc. Type "T" Tubulaire Scrubber. No Bulletin
Number. No Date,
6-3
-------
43. Carborundum Inc. Process Gas Scrubber System. Bulletin 900/12-74/2M.
44. Pangborn Corporation. The Pangborn Ventrijet. Bulletin 920 B. No Date.
45. Ducon. Venturi Scrubber Type WO. Bulletin W-9079. 1975.
46. Research-Cottrell. Flooded-Disc Wet Scrubbers, Bulletin RC-975. No Date.
47. Riley Engineering, Inc. A33 Venturi-Rod Scrubber. Bulletin 110.
October, 1975.
48. Ekono, Inc. Environmental Pollution Control, Pulp and Paper Industries,
Part I-Air. U.S. Environmental Protection Agency, Research Triangle
Park, N.C. Publication 625/7-76-01. October 1976.
49. Environmental Science and Engineering, Inc. Field Surveillance and
Enforcement Guide: Wood Pulping Industry. U.S. Environmental Protec-
tion Agency, Research Triangle Park, N.C. Publication 450/3-75-027.
March 1975.
50. U.S. Environmental Protection Agency. Atmospheric Emission for the
Pulp and Paper Manufacturing Industry. Publication EPA-450/1-73-002.
September 1973.
51. PEDCo Environmental, Inc. Operation and Maintenance of Particulate
Control Devices in Kraft Pulp Mill and Crushed Stone Industries.
U.S. Environmental Protection Agency, Research Triangle Park, N.C.
Publication 600/2-78-210. October 1978.
52. Scheroenan, J. A., and L. V. Binz. "Controlling Air Pollution While
Recycling Asphalt Pavements Through a Drum Mix Plant." Presented at
the Canadian Technical Asphalt Association Meeting. November 1979.
53. Patankar, W., and K. E. Foster. "Evaluation of the Drum-Mix Process
for Asphalt Concrete Manufacturing." Presented at the Seminar on
Asphalt Industries Environmental Solutions. January 1978.
54. JACA Corporation. Model Operation and Maintenance Guidelines for
Asphalt Concrete Plants. U. S. Environmental Protection Agency,
Washington, D.C. Final Report. Contract No. 68-01-4135, Task 44.
55. U. S. Environmental Protection Agency. "Asphaltic Concrete Plants."
In Compilation of Air Pollution Emission Factors. U.S. Publication
AP-42, Supplement 12. April 1981.
56. National Asphalt Paving Association. The Maintenance and Operation
of Exhaust Systems in the Hot Mix Plant. Information Series 52
(second edition) andd 52A (combined volumes). 1975.
57. American Foundryman's Society. Cupola Handbook. 4th Edition. 1975.
58. Reference 3, pages 7-44.
6-4
-------
59. Ranade, M. B. and-E. R. Kashdan. "Design Guidelines for an Optimum
Scrubber System." Second Symposium on the Transfer and Utilization
of Particulate Control Technology. U. S. Environmental Protection
Agency Cincinnati, Ohio. Publication No. EPA-600/9-80-039a. September
1980. pages 538-560.
60. U.S. Environmental Protection Agency. Control Techniques for Particu-
late Emission for Stationary Sources, Volume II. U.S. Environmental
Protection Agency, Research Triangle Park, N.C. Publication EPA
450/3-81-0056. page 9.2-8. September 1982.
61. Reference 24, page 3.
62. Reference 32, page 9.3-6.
63. Mcllvaine Co. The Mcllvaine Scrubber Manual. "Commercial and Industrial
Refuse Incineration". Chapter IX. pages 178.0-179.4. November 1975.
64. Hopper, J. G. Incinerator Enforcement Manual. Final Report to U.S.
Environmental Protection Agency. Contract No. 68-01-3173, Task 18,
January 1977.
65. Carpenter, B. H., D. W. Van Osdell, D. W. Coy; and R. Jablin.
"Pollution Effects of Abnormal Operations in Iron and Steel Making -
Volume II. Sintering, Manual of Practice." U. S. Environmental
Protection Agency. Publication 600/2-78-1186. June 1978.
66. Reference 60. page 9.8-56 to 9.8-67.
6-5
-------
-------
APPENDIX A
PENETRATION
Throughout this report the term "penetration" is used frequently.
This is a convenient means to express the quantity of emissions leaving
a scrubber. Penetration is related to the collection efficiency of the
scrubber as shown in Equation A-l.
X = 1 — E/100
Equation A-l
Where: X = penetration, dimensionless
E = collection efficiency (percent), dimensionless
Penetration is also related to the number of transfer units as shown in
Equation A-2.
Nt = Ln(l/X) Equation A-2
Conversion charts for converting from one to the other of these terms are
provided. The concept of penetration is illustrated in Figure A-l, below.
C Emissions
A Inlet
100 l/Hr.
B Solids, 99 #/Hr
Figure A-l. Penetration as related to control device inlet concentration,
solids collection, and resulting emissions; in this case,
the penetration equals 0.01.
A-l
-------
The penetration is simply the participate emissions divided by the inlet
loading of participate (C/A) or 0.01, while the efficiency is the inlet
loading minus the participate emissions divided by one hundred [(A-C)/100]
or 0.99. The penetration is a more descriptive parameter at the high effi-
ciencies typical of present scrubber systems and is a much more convenient
expression to use when referring to multiple sets of control systems, since
the penetration of a system of control devices is related to the individual
penetration values (see equations A-3 and A-4 below).
Parallel Collectors
Xt = Xj + $2 + ... +Xn
Series of Collectors
Xt = (Xi)(X2) ... (Xn)
Where: Xt = total penetration of the system
Xi = penetration of Unit 1
X£ = penetration of Unit 2
Xn = penetration of Unit n
Equation A-3
Equation A-4
A-2
-------
5 -
4 -
1
0
20
40 60
COLLECTION EFICIENCY
100
Figure A-2. Conversion from collection efficiency to transfer units.
A-3
-------
0.2 0.1 0.6 0.8 1.0
Figure A-3. Conversion from penetration to transfer units,
A-4
-------
T
PT = IT QQLLECIIQN_EFFICiENCY
0,0
40 ' 60 80
COLLECTION EFICIENCY
Figure A-4. Conversion from collection efficiency to penetrati
on.
A-5
-------
-------
APPENDIX B
BASIC STATISTICAL METHODS
1. Plotting Emission Correlation Lines
a. Step 1 - Plot Data in any one of the following forms:
(1) linear
(2) semi-log
(3) log-log
Select the curve with linear characteristics.
b. Compute statistical parameters using table below.
Point x x2 y y2
1
2
3
4
5
xy
c.
d.
IX'=
Sy=
m = z (x-x)2 = zx2 - (_JL*)2
n
p = z (x-7) (y-y) = zxy - _fex)fcy)
k = s(y-y)2 = zy2 - _(zj/)2
n
IT = zx/n
7 = zy/n
Calculate equation for linear regression line using form:
y = 7 + JP_ (x-x)
m
Calculate correlation coefficient:
r = P
V
B-l
-------
e. Calculate confidence interval for linear regression line.
(1) Calculate sum of squares of deviations of y from
regression line:
z = k -
m
(2) Calculate residual mean square:
So2=-i -
n-2
'n-2
(3) Select confidence interval level desired, then look up value
for t using the following table.
n-2
90%
95%
98%
.'99%
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
6.31
2.92
2.35
2.13
2.02
1.94
1.89
1.86
1.83
1.81
1.80-
1.78
1.77
1.76
1.75
1.74
1.74
1.73
1.73
1.72
12.71
4.30
3.18
2.78
2.57
2.45
2.37
2.31
2.26
2.23
2.20
2.18
2.16
2.15
2.13
2.12
2.11
2.10
2.09
2.09
'31.82
6.97
4.54
3.75
3.37
3.14
3.00
2.89
2.82
2.76
2.72
2.68
2.65
2.62
2.60
2.58
2.57
2.55
2.51
2.53
63.66
9.93
5.84
4.60
4.63
3.71
3.50
3.36
3.25
3.17
3.11
3.06
3.01
2.98
2.95
2.92
2.90
2.88
2.86
2.83
(4) Calculate interval values using the equations below.
yupper = yo + (t)(s) /H. + (x^E)2l
m
Blower = y - (t)(s) 1(1 + (x-x)2)
B-2
-------
APPENDIX C
DATA TABLES
TABLE C-l. VISCOSITIES OF AIR AT 1 ATM PRESSURE
Temp°F
Centipoise
Ib/ft-sec
120
140
160
180
200
250
300
350
400
450
500
600
700
800
0.0192
0.0195
0.0198
0.0201
0.0206
0.0218
0.0229
0.0238
0.0250
0.0260
0.0270
0.0285
0.0305
0.0320
1.29xlO-5
1.31xlO-5
1.33x10-5
1.35x10-5
1.38x10-5
1.46x10-5
1.54x10-5
1.60x10-5
1.68x10-5
1.75x10-5
1.81x10-5
1.92x10-5
2.05x10-5
2.15x10-5
Source: Chemical Engineers Handbook, page 3-211.
C-l
-------
TABLE C-2. GAS DENSITY AT SATURATED TEMPERATURE AND PRESSURE
Degrees, F
-0"
Pressure, inches w.c.
-20" -40" -60"
-80"
120
130
140
150
160
170
180
190
0.0655
0.0634
0.0613
0.0621
0.0601
0.0580
0.0556
0.0531
0.0502
0.0470
0.0434
0.0587
0.0568
0.0547
0.0524
0.0499
0.0471
0.0439
0.0404
0.0554
0.0535
0.0515
0.0492
0.0468
0.0440
0.0409
0.0374
0.0520
0.0502
0.0482
0.0461
0.0436
0.0409
0.0378
0.0344
Source: Environmental Elements, Psychrometric Tables for Wet
Scrubber System Design Information, Series DIS-10-003,
June 1979.
C-2
-------
0,0700
0,0600
0,0500
0,0400
0,0300 —
I
I
_L
120 130 140 150 160 170 180 190
SATURATION TEMPERATURE, °F
Figure C-l. Gas densities as a function of temperature and pressure.
C-3
-------
-------
APPENDIX D
TROUBLESHOOTING CHARTS
D-l
-------
r-
10
A
vo
in
CO
I—
1
LU
CO
LU
LU
l«
|
f*g*
<£
cu
J^
O
O
LU
CO
CO
H"
g
3:
o
(3
P»*|
HH
|— .
g
=c
CO
LU
oa
o
1—
1
a
LU
00
•^
i—
•
o at
c« 4.? cn
o at •> t-
«»- to re
•» at x:
_ii to > o
o at i — «/)
o r— re T-
r— N > -a
N 1
* * ^3 *^O ^^
C 4-> C Ol E
O 3 C 3
•r- Q. •> Ot O.
j ,> | ^ QT) *">
O 3 C: O T3
>i—
CL-r- i— 4-3
E CX-P 4->
3 00
Q.-O at t_ •
at s- j= to
NX cn e | ^ QJ
o cn o t- >
at 3 o at i—
x: r— c: > re
O CL-r- O >
to
at
4J
c_
3
o
^M
tf—
•o
•r—
3
cr
r—
at -i-
to t-
3 3
(O 4->
O C
at
5>
3
0
>J
c.
at
>
.
(0
at
c_
3
4->
re *r~ re
r** ^— ^.
o o at
X O CL
at o E
i at
re to
at 3 TJ
x: T-J c
TJ re
j*: re
o at
at -o 4->
x: c re
o re t-
-0*
at
3
4^
re
£_
at
CL
F
at
^_>
i ^
at
r—
CI
l^n
C_
at
re
i
cn
E=
^_
O
O
0
^-
cn
•i^
"T*
j^
I
to
t-
Ol
CL
O
j *
c—
at
F
CL
•r-
3
cr
at
E
O)
£_
CO
Q.
3
XX
o
O)
x:
o
o
f_
o
at
t.
3
[ ^
CO
t-
Ol
CL
E
at
4-»
i *
at
c
•i—
£_
O)
f*
f^V
3
C_
O
CO
^
cn
•r—
ar
o
•
£Z
O
•p
•
o
r—
<»-
to
re
cn
CO
CO
at
0
x
01
t-
Ol
o
r—
t~t
VX
O
at
x:
o
re
at
re
£_
^
o
i«_
*i
c.
•r—
re
^
o
re
CL
0
£_
T3
at
£_
3
CO
CO
at
a.
3
O
to
Ol
1^
N
N
O
C
CL
E
3
CL
VX
O
Ol
O
•°
t
QJ
4-^
CO
O
^~
<4—
T3
•r—
3
CT
i —
^
O
XI
c:
0
•r—
O
at
CO
£_
at
r~\
3
£_
0
CO
| ^
O
O)
CL
CO
C
M
o
c*~
o
•1—
1 ^
o
at
cn
c
•1—
c
to
Ol
r—
O
O)
•a
o
t_
LU
O
»
to
at
•i—
c
re
at
0
T3
-o
01
cn
cn
3
^_
CL
CO
C_
at
4->
Ol
s:
-a
e
at
o
i —
ri
NX
O
at
x:
o
re
ai
£
^
o
r—
t[
£_
•r-
re
r~
cn
«i —
ar
re
CL
0
t_
T3
Ol
e
3
to
CO
Ol
t_
CL
^*
cn
2C
[ ^
o
Ol
CL
CO
c:
t — i
ja
CO
o
3
-o (-
at
g~ r^
•r— o
3
cn £_
c o
•r- CO
cn
cn £_
3 0
51
ja
t_
at
o
^~
o
NX
O
at
0
re
Ol
4->
c.
^
o
P_
*i
t_
re
3
O
1
re
Q.
o
£_
-a
Ol
3
CO
CO
Ol
t
CL
3
O
to
at
i—
N
N
O
C
CL
E
3
CL
NX
O
at
o
•°
at
•4-*
fC
£.
O
^~
M —
T3J
•i—
rs
cr
•r-
r-™
3
o
i
X!
, — .
(_
0
re
c:
•r«
E
r—
O)
^_J
CO
•P—
F
^7^
j ^
O
0)
CL
CO
c:
i — i
0
-o
O)
re
o
o
r—
CO
•r—
T3
re
•i—
-a
O)
s:
o
t_
01
o
r— •
o
sx
O
01
x:
0
re
Ol
i *
re
£_
3
O
r~—
t)
£_
•r-
re
x:
cn
•r—
"T"
re
CL
o
t
^Q
Ol
£_
3
to
CO
Ol
£_
CL
t~
cn
'£
to
01
r—
IM
N
0
C
CL
E
3
CL
NX
O
Ol
0
XI
Ol
4_>
re
3
0
^—
T3
• r—
3
CT
•~
x:
cn
• r-
"T"
XI
r"
re
at
o
"••^
^_>
o
at
CL
to
cl
I-H
0
cn
c~
•1—
cn
cn
O
r—
O
o
•
-a
—
4_>
C
0
o
e .
•r-
re
•o
***^
p
o
at
CL
to
c
i— i
T3
cn
c
•a
o
o
r—
U_
•o
-------
r^.
vo
•V
^o
LO
CO
I—
2=
LU
s:
LU
tt.
CO
C i —
CU 03
O CU >
03 O
i — 3 C
CL T3 CU
CU CU Q.
0£. C£ O
03 jQ O
*
CO
CU i — C7>
,co ro C
3 CU •!- -O
03 «/) Cn
0 N i—
03 O 03
0. Z =»
03 J2 O
E CU
O £_
4-^ 3
Q. CO CO
E NX (/) CU
>, 03 CU CO
CO
r— CU
r— r— T3
CU NI CU
CL N CU
E 0 Q.
I-H -Z. CO
03 J3 O
^
[ ^
03 -a
t_ cu
S CU -C
o t. to
1 3 •!-
t- CO C
CO •!-
CL CU E
E t- -r-
3 CL T3
Q-
£_
CU
03
. £_
« CU
to i —
CU •—
Q. CU
•t- a.
cu a. £
O -1-
03 j^
i— o -a
CL CU C
CU JC 03
o; cj>
T3 CU
o^
c
•r—
CL
C^ 'I-
c: o_
\* ^-
U T-
03
CL C
o
CU •!-
•i— O
-4-5 3
o c_
CU 4->
<»- CO
Q 0
-a a>
-a
cu
O5
03
E -
03
-o
CU
O
03
i—
a.
cu
£_
T3
d
03
CO
^j^ _!_'.%
0 t-
O) 03
J= Q.
C_3
CU
C
•r™
r—
-o
O)
CU
r—
jQ
C.
03
CU
i—
O
-o
1=
03
4_>
O
CU
Q.
CO
C
1 — 1
CU
r—
03
•o
CU
CU
f—
o
CU
o
03
r-™
Q_
O)
£_
•a
c
03
vX
O
CU
JC
o
CU
o
03
r—
CL
CU
£_
£_
0
C
03
CU
O
-o
C CO
03 CU
r—
-^ N
O N
CU O
-G C
CJ>
CU
O
03
Q.
O)
^.
i_
0
g
•r-H
03
CL
CU
£_
T3
g~
03
^: CL
o E
CU 3
x: Q.
0
!_
O
03
>
-a
cu
c:
cu
CL
O
>^
^_
4->
O
CU
c_
t_
o
0
c
•r~
CU
c_
03
-C
o
to
•^5
CL
E
3
CL
Tjj
CU
[ %
4->
0
c_
J=
t_
CU
>
o
•
CU
r—
(O
c
o
o
03
03
03 03
03
CO
LU
o;
co
=c
LU
CO
o
t_
CU
CU
NJ
N
O
£_
O
cu
CD
03
CD
CU
in
3
03
O
CU
-Q CO
CU
O N
1-
o
CU
CO
03
O CO
CU Q-
jf^ £--
S Q-
o
<«-
03 i—
Q.
CU
a>
r— IM r— O)
'
C •!- S_
£- 4-> O
O 03 3
31- CT
O- CU -i-
Q. i—
£_
O T3
3 CU
O" CD
•r- O>
03 •!-
r— 3 r— O
r— C
o a. o T-
03 03 03
03 '
CU
^_3
03
£_
s
O
<^ CU
CO
CU 03
F— CU
O t_
>> o
o c:
O) ••-
o;
c:
0
^_>
^)
ZS
•o
OJ
^.
^
0
r— «
<4-
T3f
O)
CU
CO
CU
CO
03
CU
1_
O
c:
•r—
o
r—~
4-
T3
O)
CU
CO
CO
03
4->
CU
i —
4->
3
O
-a
C*"
03
1 ^
CU'
f~
1 — 1
03
3
cr
CU
CO
CU
1_
3
4->
03
5_
CU
Q.
E
CU
_l->
CU
t_
3
CO
CO
CU
£_
CL
CU
I—
tsl
N
o
a
o
D-3
-------
re
t-.
10
«l
10
in
co
UJ
o;
o
JQ
re
d)
O)
co o
o) a)
c 01 >
•i- c: t_
4-> •!- 3
4J 4J O
O) 4->
to O) C
to re
(_ lj_
O) 4->
' CL re •
E o to
re t— >
I 4-> C
c o
re -I- T-
O)
co
re a>
O 4-3
0)
0 O
t/) •!—
3 T3
•r-5-o
•a re
UJ
fO
UJ
UJ
o
o
UJ
•S
CO
a; QJ
«C to
:c 3
«j re
0
CO
z:
c
o
o
CO
f—
z
^
UJ
^^>
CO
UJ
0)
g £
•— 3:
<*- o
t- c
o
o
o
CO
UJ
CQ
o
cc
O v-
a>
r— t-
t- 0) ^
O •!- •—
3 o re
cr •!- o
•r-
•M
re
x
ai
a>
<_
I
O)
Q.
CD CU
O
£_
•i—
O
a>
c_
O
3
D-4
-------
r-.
IO
vo
tr>
UJ
o
e^
s:
a:
U_
Qi
UJ
a.
Q
•=c
o
1 — 1
r-
1 — 1
Q
O
a:
UJ
CO
CO
o;
0
CO
z:
o
Q
UJ
CO
CO
1—
CO
4-00 •> £-
CU 00 03
j>^ t— cu x:
O N > O
O N r— CO
_J O 03 -r-
c > -o
1
• -a -a Q.
c p c: cu E
O 3 03 C 3
•r- CL CU CL
P .p CO CL
O 3 C O -O
< o ••- cu
CL >>r—
CUT- r— p
E CL-P 4->
3 00
CL"O CU £-
cu t- x: •
.i*: co t_ .p cu
o co o t- >
CU 3 O CU r—
x: i — c > re)
O CL-I- O >
.
03 .
cu
-P
03
C_
3
O
l!
£_
O
cr
•i—
CU r- .
00 1
3 CO
03 C
O ••-
XI
3
£_
O
OO
5
O
1
03
CU
o
c
03
E
O
ll
*^—
C_
CU
CL
E CO
O C
-P -i-
Q. C
E 03
>, CU
OO r—
o
f_
o
0
Q-
t_
O
3
cr •-.
r— [ \
1 03
CO £_
c: 3
•r- 0
X5 4-
3
t- CO
O 03
OO CO
3 3
0 O
£_ ••>
o cu
4- -P
03
-i^ S-
o 3
CU 0
0
OO
00
o
£_
O
03
O.
0
£_
-pj
I
CU
3
CO
oo
cu
CL
3
0
_J
XI
•*^
cu
•o
"(_ J • ** CO
CU £- i—
•p cu x>
03 i —
t- i — -p
/^ O O3
•r- S_ O
>— -P t-
03 c x:
0 0 -P
c: 0 i
3 CU
•P t—
t- 03 XI
O O 03
C_ •!—
CU f* C_
> -P 03
•r- 1 >
-P CU
03 i— -a
t- XJ CU
cu 03 co •
CL-F- 03 J*
O t- E CO
C 03 03 -i-
•i- > -a -o
i— CU
•r- N
03 >,
4- r-
03 •
00 C >">
C 03 -P
O -r-
•r- • •> 4J
03 CU 03
O r- 3
•r— XI O"
4- O
•i- t_ -a
T3 CL C
O 03
E CU
x: cu
c •.-
•i— 4-> OO
-P 0
03 CU CU
£_ £_ i —
CU £- O
CL O T-
O O -P
4- O 03
i— i -p CL
.
O
1
•r—
C_
I ^
OO
•i- C
T3 CO
•r—
CU OO
IM CU
•r- T3
OO
t_
£- CU
O X2
CO 3
C £_
•r- O
T3 CO
03
O T3
r— C
o
•P ^i ^s
co a?-p
3 f^* «r—
' -o i—
c: -F-
] > ^5 Q
CU -r- 03
i— -P CL
C 3 03
i — i XI O
0
C
03
1 4-
c
cu E
> cu
| ^
CO OO
c >>
•i— OO
J *
•p "
CU CO
oo c:
. •!—
t_ 4^
CU -P
CL CU
E co
03
"O -P
1 03
C 0
03 t-
4- XT
•P
.*: i
O •!-
cu c-
x: 3
0 P
.
-a
3
O
^_
4-
oo
03
cn
O)
•r—
CO
OO
cu
o
X
UJ
"0
•
cu
^_
3 .
0
c~
03
4-
•
OO
>
c:
o
•r—
•P
03
£_
cu
CL
O
1 CO
c c: c
•r- 0-1-4-
•r- <— 0
4- +J 03
•r- -r- O CU •
oo oo oo t-
00 O 3 O
>> CL4- '3
03 E H-I CU cr
£_ O -P •!-
CL O 03 r—
OO « CO
^a t_ -r- co
c: o O -P c
3 CU 3 OO -r-
o x: cr cu x:
"O O T- > co
x: i — c: 3
00 -i- i —
03 • >, 4-
S -a 03 »
O) t- 00 3=
^i i— CL C- CL
O i — OO 3 1
CU 03 03
x: -P 4- o o
C_3 CO O O i —
.
cu
.p
c
cu
E
c
•1—
03
<
-p
cu
-a
cu
\>*
o
o
i —
o
^> f
r— O
i— ^^
03 03
•r- £-
p O3
t- CL
03 CU
O- CO
CU
E 0
CU -P
CU i—
x: xi t-
+-> o cu
t- P
P CL 03
3 3
0 4-
I-H D^
cu c
03 • x:
c: c: co
00 -r- 3
03 1—
-O £- 4-
C T3
03 T3
t- "O
C O 03
3 -P
O 03 •>
-a c oo
•i- C-
4-> E 3
3 T- 0
x: r— cu
OO CU t-
.
03
.^ — ,
•a >,
CU r-
CO C
CO O
3
i — OO
CL i-
O
c -P
•r- 03
03 c:
t- -r-
-0 E
•c—
£_ i—
O O
•P
03 CU
C C
•t- 03
E >
•i—
r— t—
CU 03
•r-
CU -P
C- £=
3 CU
-p CO
co c:
•r- 03
0 +J
^~ •**—'
03
•P
c:
cu
£
d
•r—
03
£_
-P
C
cu
o
3
CT
|n
CO
03
CO
1
•p
CO
03
x:
x
UJ
c
•r-
03
C_
•a
cu
•p
03
CO
•i—
t_
£_
•r—
>>
CO
3
0
3
C
•i— •
-P CU
c CL
0 ••-
O CL
t- co CU
03 c: N
OJ -i- >>
f^T ^y; r^
4- 0. 0 03
I-H 03 x: c: •
co O OO CO
CL O >> £= 03
E C 03 -r-
03 E 03 CO
oo oo t- c:
CU COT3 •!-
• t. o c E
O T3 T- T3 03
3 ECO
crx: 03 03 4-
•r- p O
i — O 4- oo t^.
£- £-0
03 4- •> CU 4-
>s E
CU £- r— 0 £-
CO O Jsri O O
I- 3 0 C 3
03 CT-i- 3 cr
CL-i- 3 O -r-
oo r- cr-O.r-
.
XI
>,
t—
x>
•r- *~^
CO t-
00 O
0 3
CL cr
•"•-^ .p-
r—
o>
C C
•r-* *i~~
^:
i ^ i %
"o c
t- a>
4- CO
03
t_
CU cn
•P !=
03 -r-
3 E
03
CU O
> 4-
•r~
CO O
00 -P
CU
O CU
X 3
LU T3
^
X)
•
13
-•
•P
C
O
o
D-5
-------
r-.
10
•*
to
in
UJ
o
z>
^c
J-
on
o
U.
cc
UJ
CL.
O
^^
«c
to cn
c c
O •!-
•r- _c 1 cn
4-> to to c:
•i— 3 re _i£ -i—
"O i — Or— O i —
C q- r- Q) re
0 C- -i- XT 0
0 Q) 0 5 O to
t- o
>, O Q- O> f_
18 E O • O
t. c 0.4-
e: CL o> • re 3
O to to t_ E "O >>
•i- 1 => 3 t- r— f_
•«J cn o o T- ^j
0 C -C 4- 3 t/)
«C 'I— • t_ _Q •!—
J= C C_ 0) E
to l_ <1) CL OJ OJ
3 0) CL +J J=
!—•»-> cn re o
4- 4-> to C C_ •
IO "O •!— OJ £_ to
JK! CL 0 C •— 0 ^->
o -i- re o> 3 c
Q) 73 C_ OL) O CT OJ
.G C 03 i— O T- Cn
o re CL o re i — re
i \
re
o
• c_
o -t->
•r~ 1
•r- t—
10 3
O 4->
CL c:
0)
t- >
O)
CL d)
E r—
ro ^^
TD re
•r—
C t-
re re •
c
O O £1
O) d) (L)
jc Jr CL
o o o
O
HH
H-
H-I
O
o
o
UJ
CO
CO
O
o
I
o
CD
z:
»— i
o
o
UJ
CO
o
r— CX OJ -(->
i— E •— re
•r- re .a o
3 -o re t-
O •!- JZ. i
t- 4-> C £_ ^J 3
a> re re o
CL= 4- > ^J t—
E a> o 4-
re j«j o jc D
•a o to 4-> jc t_
re = -o cr
4- re 73 • T-
^: o a) cnr—
^= 3 II o T- -o
+-> .o ID d) c: c
c > j= , O £_
i— O) +-> CU O
re T- a> re re
E > > -P o s-
t- o a> to t_ a>
O £_ t- 3 .C CL
z: CL CL E -(-» O
en
c
•r-
d
O) to
CL
0
I
O)
r—
N
N
O
C
"^^.
t.
O)
c:
• r—
re
£_
4->
to
^
»^-
•a
o
s:
. O)
r—
O
.£1
0)
n^
N
M
O
C
4->
re
g~
-t^
o
to
o
•r-
.{_3
re
£_
ameter
•r—
T3
ai
.c
4^
O)
O
•r—
3
4->
-l->
to
re
0)
r—
•»->
re
0)
i_
re
•
to
cn
c
•r—
C
0)
CL
O
C.
(U
C
•r—
re
£_
^_>
to
4 —
O
•
c:
O)
E
CL
•r—
3
cr
0)
c
O
•r—
P
re
£.
re
CL
(1)
to
SX
U
O)
J=
C_3
-a
re
o
•)->
to
3
•a
a>
•r—
to
to
0)
O
X
£_
4->
to
re
cn
c:
•r—
en
c:
•r—
C
O
•1—
I)
o
3
4-
r—
re
O)
•r-
P^
1
ai
cn
3
Q.
3
r—
4-
•a
c"
re
to
-a
re
O)
>,
(3
c_
CL
to
QJ
>
O
E
O)
OL
•
to
O)
M
N
o
C
-a
cr
re
cn
C
Q.
CL
>,
re
Q.
to
•I—
o
to
0
£_
£_
O
o
4-
cn
c:
•i—*
CL
•i—
CL
O)
O
re
r—
CL
O)
C£
•
t ^
C
£_
re
CL
CL
re
0)
to
=1
re
o
o
t_
o
re
c
0)
(O
to
ai
o
x
t- eu
JD -r-
JD t/)
3 tO
£_ <1)
O O
tO X
0)
§0
re
re
to
o
o
to
en
c:
CD
CL
O
a>
r—
N
N
O
en
o
o
cn
£_
O
3
cr
O
to £_
•a CL
•i— to
o c
OO •!-
>,
O) OJ
•a ja
t- E
o re
a>
<1) £-
i— -P
re
o
to -o
O) 3
Q. cr
CX J—
O
UJ
CO
o
4->
Ci
I
re
o
£_
a>
o
£_
O
re
u.
to
OJ
N
N
O
£_
Q.
(O
•a
a>
cn
cn
D-6
-------
r-.
10
10
in
UJ
o
p— en
C i-
O.
Q. 3
•r- -0 £_
3 O •
cr cu <»- en
c
C -r- -T3 -r-
o c to c c
•r- o in re o
P •!- CD •!-
o -p o E -P
E
^4 ^t to
o o en cu
cu cu e
O O •!- i—
re
o
o
•p
C i-
O 0
•i- 3
p cr
re T-
CU C- r-
3 C" >^
re cu re
o o t-
C Q.
O in
O
c
in T-
•o
•r— f~
.' i— en
O •!-
CO -C
re
£_
re
cu
cu
• r—
N
E N
0 0
•p cr
Q.
E cu
CO -r-
in
in
CU
O
X
UJ
£_
0 1
3 C
cr O s-
•r- -I- re
i— tO CU
re s
E t-
o jo c
t_ re T-
<4_
i — in
in i — en
cu re c:
> 1 * «r—
•r- in C
in C T-
re T- r—
£_
O f [ *
re o c
re
cu E -P
> re m •
§.
re
£_
Q.
to
C
•r—
in
cu
>
•i—
in
re
£_
-Q ,
•=C
JD
~^j
re
• 0
c:
cu -p
E in
Q. 3
3
cr 0)
cu >
•(•«
c: to
O in
•r- CU
4-> 0
re X
c- cu
re
CL I-
cu o
in <»-
XX NX
0 0
cu cu
0 0
o
^,
_E C
•P 0
•i— »r—
2 in
0
C 4-
0 CU
•^.
•P en
re c
C •!-
•i- in
-Q =5
E re
0 0
u
c
•i- cu
"^T", «^-
d. t/J
03
s t—
0 -Q
_i re
o
i cu E
cu > re
en o cu
t. E t-
3 CU -P
CXQ; m
t_ 5_
O • O
M- en 3
£= cr
^3 »^— »r™
C C i—
re o
•i- E
E -P o
re o «-
cu c <»-
t- 3
4-> <4- in
in r— cu
re >
in E-r-
re in
en cu re
'c t-
C: -r- J3
•i- r— re
*
c:
o
•r—
in
o
£_
£_
O
o
^_>
c:
as ~a •
•P -O C
«/» -si o
•r— »r-
IO 4J
cu • re
£_ in o
1 CU •!-
c c n-
O O -r-
•r- N T3
in o
re s- E
t- re
.a cu 3:
re s Q.
r- C £_
•— -r- O
re re en
•r- C
c: in E -r-
o in re c
•i- cu cu o
•P O £- •!-
re x -p 4->
t_ -P en
• a; c in t_
J= CU 3 3
o E -o ca-
re
c
o
•r—
) ^
re
3
o
•^
o
•i—
c to
O in
•r- O)
•P 0
re x
c- cu
re
CL £-
CU 0
in <*-
VX XX
0 0
cu cu
J= JZ
0 0
^
C. •
0
crt
O
c_
flj
""i«»^.
C
O
•r—
t/)
O
^.
L.
0
o
JD
•r—
re
VX p^
1 1— 1—
cu re re
en +j.
t- -o in
3 T3 C
Cu^C *"~*
£_
O • •
re o re
•r- O
E -p-i-
re o 4-
cu c: -r-
£- 3 T3
-P
CU 3 O
c cr re
r^ r— CL
p 4- p
c ••- o
re c:
-P in
in cu cu
•r- C «_
in o re
CU N
i- in
i c- c:
c re o
O CU "i—
•r~ 2 1 *
in I re
re j= o
t_ en-i-
JD T- <4-
re j= T-
TD
c
re
CL
O
t_ •
TD •
1 -P
CU C
t- •!-
3 O
in Q.
in
cu £=
t. en
Q.-I-
in
•p cu
re T3
o
£- 0
-P
NX O
O 3
CU T3
J= CU
O t- .
O
>^
1 ^
lr_
0
o
"cu
in
re
en
cu
>
• r—
0
O
D-7
-------
f^»
«£>
«i
«S
to
UJ
O
z
s:
o;
o
u_
cn
UJ
o.
/-*i
^r?
•=5
£_
0
! £_
•r-
03
Q.
•r- C_
4-> v^
O
C CD
o x:
<_>
CO
CL
03 •
[ ^ r>
3
t_ "XD
O i—
IO -r-
C 3
CD -Ct
CO
CO
\y -^3
0 T-
>-P O
i^ *^
f^ \s 40
CL O •!-
3 CD 10
co x: o
1 0 CL
«_
•I— \S
03 • O
CO CD
-P t- x:
C CD O
CD -P
E r—
3-r- •
£- •
i — CO C
x: i— o
en 03 -r-
3 c: P
0 t_ 03
t_ CD S-
O P CD
x: c a.
1— -r- 0
o
O
O
O
UJ
CO
CO
cc
o
CO
LU
CO
«c
CO
fe
£_ 03 03
T3
P
O
O
M- O)
03 i—i > CD 03 >
C
£_
CD
CD "O CL
(O
4-J C JQ
CD 03
£_
CD -P O
03 CD
E >
a) 03 en
c: i— •
•r- 03 .
cn4->
c x:
T- en
03
e: 03
03 CL >,
E r—
£_ CD
£- O
COOPOM--i—
-p S_
O 4J
CD 03 .
to
CO TD
COT303PCL3OCDC
CD
CO 4-» CD O 03
•i- 4- t- e cox: Xi
U_'O03 3-r- 03O CDCn
03
-------
o
•p
o
o;
o
Q-
Q
o
i.
0
o
o
cu
o
CU
cu
o
re
CL
cu
o;
co
i —
re
£_
cu
4->
re
cu
o
re
•—
CL
CU
OL
CD
-P
01
£_
CU
re
cu
CO
re
-P
cu
-P
re
cu
•P
co
re
JC
o
c:
re
cu
CO
co
CL
CL
cu
c
re
o
o
cu
cu
o
re
CL
cu
o a:
o
cu
O
CO
cu
T3
re
r—
O
C
re
i-
c:
re
cu
o
<4-
0
c:
0
•P
re
t_
cu
CL
O
j,,.
0
CU
o
t_
r_
re
CL
cu
o;
re
re
to .a
o
re
re
o
o
a;
CQ
CO
o
00
LU
CO
CQ
I—
C£.
3=
o
CD
O
O
CO
LU
re
o
c:
o
re
c: o
O ••-
•i- t-
•P cu
re -p
N cu
•i- -a
're co
t_ c
•p -i-
^ c=
CU -r-
C I—
•<->
C
CU
-p
c:
o
CO O
re
•r—
t_
CU
4->
re
cu
re
3
cr
cu
•a
re
cu
o
cu
ds
so
ed
I—I CL, I—I
+1 O
re cu
CL «-
o j=
o co
c: -r-
re
re
treatment
ca
cu
o
a>
CL
o
£_
CL
re
content
O
co
c:
o
•r-
•p
o
re
:_
+j
c
o
o
c:
o
•t—
to
c:
re
CL
x
cu
CL co
3 T3
s_ c:
f-i CU
=C -Q
CU
CO
^ re
c: E
cu re c:
E -o o
CO
CU
-o
re
c
o
CL
Z3
•a
stribution
ze
e
•r- o
CO CO
•i- . cr
re c- •!-
co re >
•r- cu re
s: CQ o
4-> re
re •!—
re
o
cu
re
re
c_
re
CL
cu
co
re
re
CO
CU
en
en
3
CO
CU
CL
re
to
cu
N
Nl
o
CU
re
cu
CL
o
c
to
CL
E
3
Q-
OJ
I
Q
CO
•=£
O
+J
Q.
CO
O
O
o
CO
O
LU
CO
CO
CO
3
CO r—
£= CL
r— CU
re CL
o -r-
CO Q-
re
cu
cu
CO
o
c:
CL
c:
o
CO T3
re o
•r- re
re
£_
re
£_
cu
CU J3
> JO
•r- 3
CO £-
CO O
CU CO
o
X c:
LU O
O
3
cr
*U.S. GOVERNMENT PRINTING OFFICE: 1984—739-002/13,503
D-9
-------
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA 340/1-83-022
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Scrubber Inspection and Evaluation Manual
5. REPORT DATE
September 1983
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO,
John R. Richards and Robin R. Segal!
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Engineering-Science
501 Wlllard Street
Durham, NC 27701
11. CONTRACT/GRANT NO.
68-01-6312
12. SPONSORING AGENCY NAME AND ADDRESS
US Environmental Protection Agency
Office of Air Quality Planning and Standards
Washington, D. C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
EPA Project Officer;
Kirk Foster, Stationary Source Compliance Division
16. ABSTRACT
This report concerns the inspection and evaluation of performance of paniculate
wet scrubbers installed at stationary sources to identify and rectify operating
conditions contributing to excessive emissions. It includes discussions of
gas-atomized scrubbers, plate-type scrubbers, packed tower scrubbers, and spray
tower scrubbers. The evaluation approach proposed utilizes comparisons of
present performance parameters and conditions with site-specific performance
established previously under a controlled set of conditions.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
Air Pollution
Wet Scrubber
Particulates
Operation and Maintenance
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
162
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------
-------