United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 277^ 1
EPA-450/3-79-020
June 1979
Air
Measurement of
Benzene Emissions
from a Floating Roof
Test Ta
-------
EPA-450/3-79-020
Measurement of Benzene Emissions
from a Floating Roof Test Tank
by
Royce J. Laverman and William N. Cherniwchan, CBI
Chicago Bridge & Iron Company
Research Department
Route No. 59
Plainfield, Illinois 60544
Contract No. 68-02-2608
Task No. 39
EPA Project Officer: Richard K. Burr
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
June 1979
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35) , U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711; er for a-nominal-fee-,
feom-the- National Technical Information Service, 523-S-Por-t Royal Road,
This report was furnished to the Environmental Protection Agency by
Chicago Bridge & Iron Company, Research Department, Route No. 59,
Plainfield, Illinois 60544, in fulfillment of Contract No. 68-02-2608, Task
No. 39. The contents of this report are reproduced herein as received
from Chicago Bridge & Iron Company. The opinions, findings, and conclusions
expressed are those of the author and not necessarily those of the
Environmental Protection Agency. Mention of company or product
names is not to be considered as an endorsement by the Environmental
Protection Agency.
Publication No. EPA-450/3-79-020
11
-------
TABLE OF -CONTENTS
Page
Chapter 1.0 Summary and Conclusions 1
1.1 References 6
Chapter 2.0 Test Description 7
2.1 Test Facility 7
2 .1 JL General Description 7
2.1.2 Principal Instrumentation 9
2.2 Test Method 10
2.2.1 Analyzer Calibration 10
2.2.2 Product Description 10
2.2.3 Seal Flow Tests 10
2.3 Test Description 11
2.3.1 Phase I, Pan Type Internal Floating
Roof 11
2.3.1.1 Description of Floating
Roof and Seals n
2.3.1.2 Description of Seal
Spacers 16
2.3.1.3 Description of Test
Conditions 16
2.3.2 Phase II, Bolted Cover Type In-
ternal Floating Roof 22
2.3.2.1 Description of Floating
Roof and Seals 22
2.3.2.2 Description of the Seal
Spacers 24
2.3.2.3 Description of Test Con-
ditions 25
iii
-------
Page
2.3.3
2.4
Chapter 3 . 0
3.1
3.2
3.3
3.4
3.5
Chapter
3.6
4.0
4.1
4.2
4.3
4.4
Appendix A
Appendix B
Phase III, External Double Deck
Floating Roof
2.3.3.1 Description of Floating
Roof and Seals
2.3.3.2
2.3.3.3
Description of the Seal
Spacers
31
Description of Test Con-
ditions 35
References
Test Results
Benzene Product Test Results
39
40
40
Propane/Octane Product Test Results 42
Vapor Pressure Function
Correlation of Test Results '
Deck Fitting Emissions for Internal
Floating Roofs
References
Benzene Emission Estimation
Summary of Estimation Method
Sample Problem
Relationship Between K_, Kp, and F
References
55
57
57
63
64
65
70
73
75
Summary of Emission Test Data A-l
Calendar of Events B-l
-------
LIST OF FIGURES
Page
Figure 1- 1 Roof and Seal Type Emission
Comparisons 4
Figure 2- 1 Simplified Process and Instrumentation
Schematic 8
Figure 2- 2 General Arrangement of an SR-8 Resilient
Foam Seal Mounted on a CBI Weathermaster
Floating Roof 12
Figure 2- 3 Position of the CBI Weathermaster Roof
Within the Emissions Test Tank 13
Figure 2-4 Rim Mounting of the BWB/CBI 1000 Flapper
Secondary Seal 14
Figure 2- 5 Pan Type Floating Roof Deck Penetrations
Schematic 15
Figure 2- 6 Typical Temporary Sealing of a Deck
Penetration Fitting 17
Figure 2- 7 Description of the Sheet Metal Seal
Spacers 19
Figure 2- 8 Installed Shingle Type Seal 23
Figure 2- 9 Position of the Ultraflote Internal
Bolted Cover Within the Emissions Test
Tank 25
Figure 2-10 Cross-Sectional View of the Shingle Type
Seal Installation 26
Figure 2-11 Sealed Deck Lap Seam for Test EPA-20.... 30
Figure 2-12 General Arrangement of an SR-1 Metallic
Shoe Seal Mounted on a CBI Double Deck
External Floating Roof 32
Figure 2-13 Position of the CBI Double Deck External
Floating Roof Within the Test Tank 33
Figure 2-14 Metallic Shoe Spacer Bar 34
Figure 3- 1 Emissions Vs. Wind Speed for an Internal
Pan Type Floating Roof, Primary Seal Only,
Product at Various Temperatures 43
V
-------
Page
Figure 3- 2 Emissions Vs. Wind Speed for an Internal
Pan Type Floating Roof, Primary Seal
Only 44
Figure 3- 3 Emissions Vs. Wind Speed for an Internal
Pan Type Floating Roof, Comparison with
Propane/Octane Test Data 45
Figure 3- 4 Emissions Vs. Wind Speed for an Internal
Pan Type Floating Roof with Primary and
Secondary Seal 46
Figure 3- 5 Emissions Vs. Wind Speed for an Internal
Pan Type Floating Roof with Primary and
Secondary Seal, Comparison with Propane/
Octane Test Data 47
Figure 3- 6 Emissions Vs. Wind Speed for a Bolted
Cover Type Internal Floating Roof with
Primary and Secondary Seal 48
Figure 3- 7 Emissions Vs. Wind Speed for a Bolted
Cover Type Internal Floating Roof with
Primary and Secondary Seal, Comparison
with Propane/Octane Test Data 49
Figure 3- 8 Emissions Vs. Wind Speed for an External
Double Deck Floating Roof with Primary
Seal Only 50
Figure 3- 9 Emissions Vs. Wind Speed for an External
Double Deck Floating Roof with Primary
Seal Only, Comparison with Propane/
Octane Test Data 51
Figure 3-10 Emissions Vs. Wind Speed for an External
Double Deck Floating Roof with Primary
and Secondary Seal 52
Figure 3-11 Emissions Vs. Wind Speed for an External
Double Deck Floating Roof with Primary
and Secondary Seal, comparison with
Propane/Octane Test Data 53
Figure 3-12 Emission Correlation. Versus Vapor
Pressure 58
Appendix B
Figure B-l Calendar of Events for July, 1978 B-2
Vi-
-------
Page
Appendix B (continued)
Figure B-2 Calendar of Events for August, 1978 B-3
Figure B-3 Calendar of Events for September, 1978... B-4
Figure B-4 Calendar of Events for October, 1978 B-5
Figure B-5 Calendar of Events for November, 1978.... B-6
Figure B-6 Calendar of Events for December, 1978.... B-7
. Vii
-------
LIST OF TABLES
2-1 Summary of Test Conditions for Phase I..
Page
18
Table
Table 2-2 Summary of Seal Gap Area for Phases I, II,
and III 20
Table 2-3 Summary of Test Conditions for Phase II... 27
Table 2-4 Summary of Test Conditions for Phase III.. 36
Table 3-1 Summary of Emission Factors F and n 59
Table 4-1 Summary of Emission Factors K and n 68
S
Table 4-2 Summary of Emission Factors K- and m for
Floating Roofs 69
Appendix A
Table A-l Summary of Emission Test Data A- 2
Table A-2 Summary of Emission Test Data for Tests
EPA-5, 8, and 9 Normalized to 1.75 psia
Vapor Pressure A-ll
Table A-3 Summary of Propane/Octane Emission Test
Data Normalized to a Benzene Basis at 1.75
psia Vapor Pressure A-13
Viii
-------
ABBREVIATIONS AND-CONVERSION FACTORS
Listed below are abbreviations and conversion factors for
metric equivalents of English units. Frequently used measure-
ments are also presented in dual units below for the reader's
convenience.
English Unit Alternate-Unit ' Conversion
pound (Ib) kilogram (kg) Ib x 0.454 - kg .
ton metric ton {r.i ton) ur ton v. 0.907 = r.i tovi
. jnegfagram (Mg) (or-Mg)
mile (mi) kilometer (km) mi x 1.C1 - kr.i
miles per hour (mph) kilometers per hour (kph) iwph x 1.61 - kph
foot (it) meter (m) ft x 0.305 = ra
inch (in) centimeter (cm) ' in x 2.5-1 = cm
gallon (gal) liter (1) • gal 3: 3.79 = 1
barrel (bbl) liter (1) bbl x 1I3S- = 1
pounds per square kiloPascals (kPa) psi x G.90 = kPa
inch (psi)
degrees Fahren- degrees Celsius (°C) (°r%-32) x 0.55G = °C
heit (°F)
FREQUENTLY USED MEASUREMENTS
5.00 psi = 34.5 kra 2 mph =3.22 kph
3.22 psi = 22.2 kpa 4 rnpli -- 6.44 kph
1.99 psi ^=13.7 kp.-. 6 mph =3.66 kph
1.75 psi =12.1 kpa 3 mph =12.9 kph
1.17 pui =8.07 kpa 10 mph = 16.1 kph
12 nph =19.3 kph
14 mph =22.5 kph
20 mph =32.2 kph
-------
1.0 SUMMARY AND CONCLUSIONS
This report presents the results of a pilot test tank
study of benzene emissions from floating roof storage tanks.
The study was conducted by Chicago Bridge and Iron Company
(CBI) at their research facility in Plainfield, Illinois. The
methodology used in the study was developed by CBI during
previous pilot test tank studies of emissions from a binary
product mixture (propane/octane).1 The results of these
emissions tests were used to develop equations to estimate
benzene emissions from commercial scale floating roof storage
tanks.
The testing was done in three phases, each using a dif-
ferent type of floating roof. Phase I used a pan type in-
ternal floating roof with a liquid-mounted, nonmetallic
flexible foam primary seal. Phase II used a bolted cover type
internal floating roof with a vapor-mounted, shingled flapper
primary seal. Phase III used a double deck external floating
roof with a metallic shoe primary seal.
A total of 29 tests were run during the three phases.
Conditions were varied in order to determine:
emissions from a tight primary seal,
emissions from a tight primary seal and secondary seal,
the effect of gaps in the primary and/or the secondary
seal,
the contribution of deck fittings (penetrations) to
emissions, and
the effect of vapor pressure (temperature) on emissions
1 -
-------
The details of the test methods and conditions are dis-
cussed in Sections 2.1 through 2.3. The test results are
presented graphically in Section 3.1, and the detailed re-
sults are tabulated in Appendix A.
Section 4.1 gives the details of the method for estimating
benzene emissions from commerical scale tanks. This method
breaks the total emissions down into the withdrawal loss, the
seal loss, and the fitting loss. The seal losses and fitting
losses are estimated by factors developed in this study. The
working loss estimate was derived from previous test work.2
Several important conclusions can be drawn from the re-
sults of this study, in regard to:
emission characteristics of a pure component versus a
mixture,
the relationship between emissions and benzene vapor
pressure,
the relative performance of the three types of floating
roofs,
the effectiveness of secondary seals, and
the effect of gaps on emissions.
Each of these will be discussed in the following paragraphs.
One significant result of this study is the determination
that benzene emissions (and presumably those of any pure com-
ponent) are higher than would have been predicted by previous
testing on liquid mixtures (e.g. propane/octane). This
phenomenon can best be explained by a "weathering" effect at
the vapor/liquid interface of a liquid mixture. The propane/
octane system is a simple example, since almost all of the
vapor pressure of the liquid is contributed by the propane
component. As propane is lost from the liquid surface by eva-
poration, the effective vapor pressure at the surface falls
below that of the bulk liquid. The vapor pressure at the sur-
face will continue to decrease until it reaches an equilibrium
-------
value where evaporative propane emissions are equal to the
rate of diffusion of propane from the bulk liquid to the sur-
face. Thus, the emissions from a mixture may be limited by
the mass transfer rate of the more volatile components from
the bulk liquid to the vapor/liquid interface at the seal.
This development is important for the following reasons.
This understanding of the emission mechanism may allow the use
of the benzene emission correlations to estimate the emissions
of other pure components. It also points out the dependence
of emissions from liquid mixtures upon the ability of the
volatile components to diffuse from the bulk liquid to the
vapor/liquid interface at the seal.
Three vapor pressure functions had been suggested in pre-
vious testing. All three were evaluated using the results
of the benzene tests and showed roughly the same degree of
accuracy. The "EPA" function was selected for use in the
benzene emission estimates based on a slightly lower standard
deviation.
The relative performance of the roof/seal combinations
tested is shown graphically in Figure 1.1. Several conclusions
can be drawn from this:
the pan type internal floating roof equipped with a
liquid-mounted, flexible foam primary seal and a
continuous flapper secondary seal, gave the lowest
emissions,
the bolted cover type internal floating roof and the
external double deck floating roof were roughly
equivalent in performance,
the use of continuous (rather than shingled) flapper
primary and secondary seals on the bolted cover type
internal roof would probably have reduced its emissions
significantly,
the use of a secondary seal is very effective in reducing
the emissions from floating roof tankage,
-------
100
10
o
N.
.O
VI
o
'w
V)
'§
UJ
0.1
FLOATING ROOF
TYPE
External
double deck
Internal
bolted cover
Internal
pan
SEAL TYPE AND CONDITION
Tight or gapped pri. seal,
Tests EPA-23,24
Gapped pri. and sec. seals,
Test EPA-26
Tight or gapped pri. and sec. seal,
Tests EPA-17,18
Tight or gapped pri. seal
with tight sec..seal,
Tests EPA-25,27
Gapped pri. seal,
Test EPA-16
Tight pri. seal,
Test EPA-5,9,15
Gapped pri. and sec. seals
Tight or gapped pri. seal with
tight sec. seal,
Tests EPA-12, 13 .
NOTE: Seal only,
not fittings
1O
Wind Speed
40 m p h
Figure 1-1 Roof and Seal Type Emission Comparisons
4.
-------
the condition (tight or gapped) of the primary seal
is not as significant if a tight secondary seal is
present, and
although somewhat sensitive to gaps, even a severely
gapped secondary seal performs better than a primary
seal alone. -j
The contribution of the deck fittings (or penetrations)
to the total emissions from internal floating roof tanks
deserves more study. The total emissions from the fittings
were determined in this study, but no attempt was made to
allocate these emissions to the various types of fittings.
This can be important in scaling emissions to commercial sized
tanks, because the number of some fittings (such as support
columns and guide poles) increases with tank diameter, while
others (such as sample wells and manways) remain constant.
It was assumed here that each fitting contributed equally to
the total fitting emissions. The Fitting Multipler, N, was
then based on the total number of fittings, regardless of
type.
-------
REFERENCES
1. SOHIO/CBI Floating Roof Emission Test Program
Final Report. Chicago Bridge & Iron Company.
November 18, $976.
2. SOHIO/CBI Floating Roof Emission Test Program.
Supplemental Report. Chicago Bridge & Iron Company.
February 15, 1977.
-------
2.0 'TEST DESCRIPTION
2.1 TEST FACILITY
2.1.1 General Description
The benzene emission test program was performed in
a covered floating roof test tank at CBI's Plainfield,
Illinois, Research Facility. The test tank was 20 feet in
diameter and had a 9 foot shell height (see Figure 2.1) . The
lower 5*-3" of the tank shell was provided with a heating/
cooling jacket through which a heated or cooled water/ethy-
lene glycol mixture was continuously circulated to control
the product temperature. The jacket was insulated externally,
and the tank rested on Foamglas load bearing insulating blocks.
Access into the interior of the test tank for the
purpose of interchanging the types of floating roofs was pro-
vided by means of a flanged and bolted exterior cone roof.
Personnel access into the test tank was provided by means of
a 30 inch diameter manhole in the cone roof.
The test tank had numerous fittings provided for in-
strumentation feed thru, sample withdrawal, product addition
or withdrawal, and product mixing. Since these tests were
concerned with a single component product however, no product
mixing was required, and the mixing system which was com-
prised of a recirculating pipeline and a transfer pump was
disconnected from the tank.
The effect of wind blowing across a floating roof
tank was simulated by means of a blower connected to the tank
by either a 30 inch or 12 inch diameter duct. An inlet plenum
with rectangular openings was used to distribute the air en-
-------
oo
INLET
CONCENTRATION
BLOWER
OUTLET
TEMP.
\
| FLOW RAT E|-
SHELL HEATING
SUPPLY TEMP
2O 01 A. x S HIGH
SHELL HEATING SUPPLY
Figure 2-1. SIMPLIFIED PROCESS AND INSTRUMENTATION SCHEMATIC
-------
tering the test tank shell. This air exited the tank through
a similar plenum into a 30 inch diameter exit duct. The
12 inch diameter air inlet duct was used for the internal
floating roof tests, and the 30 inch diameter inlet duct was
used for the external floating roof tests (which required
larger air flow rates). While one size of inlet duct was in
use, the other size was always blanked off.
Some instrumentation, pertaining to the product
heating/cooling system, was read locally, but the remainder
of the instrumentation and controls were panel mounted in an
equipment trailer located outside of the tank dike wall.
Figure 2-1 also shows a simplified schematic diagram of the
instrumentation relevant to the benzene emission tests.
2.1.2 Principal Instrumentation
The principal instrumentation consisted of the fol-
lowing :
1. The air speed in the inlet duct was measured
with a Flow Technology, Inc., Air velometer,
Model No. FTP-16H2000-GJS-12.
2. The total hydrocarbon concentrations were
measured with Beckman Instruments, Inc., Model
400, Total Hydrocarbon Analyzers. Two instru-
ments were used, one for the inlet and one for
the outlet air concentrations.
3. The airborne benzene concentration at the test
facility was measured with an HNU Systems, Inc.,
portable analyzer, Model PI 101.
4. The local barometric pressure was measured with
a Fortin, Model 453, Mercury barometer.
5. During unmanned periods (nights and weekends)
the barometric pressure was measured with a
Taylor Instruments, aneroid barometer, Weather-
Hawk Stormoscope Barometex No. 6450.
-------
6. The temperatures were sensed with copper/con-
stantan thermocouples and read with a multipoint
recording potentiometer, Doric Scientific Corp.,
Digitrend, Model 210.
2.2 TEST METHOD
2.2.1 Analyzer Calibration
Calibration gas mixtures were ordered from Matheson
Gas Products Company for the purpose of calibrating both the
Total Hydrocarbon Analyzers and the Portable Analyzer. Gas
mixtures of three different benzene concentrations in ultra
zero air were used:
0.894 ppmv
8.98 ppmv
88.6 ppmv
The inlet air analyzer and the portable analyzer
were routinely spanned on the 0.894 ppmv benzene calibration
gas. The outlet air analyzer was spanned on the gas mixture
corresponding to the range in use by the analyzer. Both total
hydrocarbon analyzers were spanned at the beginning of each
eight hour shift, and the portable analyzer was spanned at
least twice a week.
2.2.2 Product Description
The benzene used during the testing program was a
Nitration Grade, supplied by Union Chemicals of Schaumburg,
Illinois. The original shipment, for Phase I, consisted of
11,400 gallons. An additional 1,600 gallons (13,000 gallons
total) were purchased from the same supplier for the Phase II
tests. The Phase III tests required the least amount of
product. Therefore, approximately 4,300 gallons were sold,
leaving about 8,700 gallons in the test tank during Phase III.
2.2.3 Seal Flow Tests
Seal flow tests were performed to characterize the
tightness of the sealing ring on each of the types of roofs
used in this emissions testing program. The flow tests were
10
-------
conducted by measuring the air flow rate thru a sealing ring
at various pressure drops across the sealing ring. It was
concluded from the data that the installations of the re-
silient foam seal on the internal floating roof and the
metallic shoe seal on the external floating roof were typical
of previous installations in the pilot test tank.
With regard to the shingle type seal, this was the
first installation of this type seal on the internal bolted
cover, therefore no comparison could be made with previous
pilot test tank installations. However, no pressure drop
across the shingle type seal could be detected at the maximum
air flow available, so no data was actually noted for this seal
The condition of not being able to develope a pressure drop
across the seal indicated that the shingle type seal was the
least tight of any primary seal previously tested.
2.3 TEST DESCRIPTION
2.3.1 Phase I, Pan Type Internal Floating Roof
2.3.1.1 Description of Floating Roof and Seals
Figure 2-2 illustrates the SR-8 flexible foam
primary seal mounted on an internal pan type floating roof.
A cross-sectional view of the position of the floating roof
within the test tank is shown in Figure 2-3.
A flapper secondary seal was used during some of
the tests. The secondary seal was a type BWB/CBI 1000. This
seal was 15 inches wide, with internal stainless steel re-
inforcing fingers. A sketch of its installation on the rim
of the Weathermaster floating roof is shown in Figure 2-4.
The following is a list of the floating roof deck
penetrations:
(1) 8 inch diameter open gauge well
(2) 16 inch diameter center column well, with a
fabric sleeve sealed 6 inch diameter column
(3) 12 inch diameter guide pole well, with a fabric
sleeve sealed 2-1/2 inch diameter guide pole
11
-------
SEAL
ENVELOPE
RESILIENT
FOAM
TANK
SHELL
Figure 2-2.
GENERAL ARRANGEMENT OF AN
SR-8 RESILIENT FOAM SEAL
MOUNTED ON A CBI
WEATHERMASTER FLOATING ROOF
12-
-------
REMOVABLE EXTERNAL
CONE ROOF
3O*0 AIR DUCT
*
o
o
n
PAN TYPE INTERNAL
FLOATING ROOF
\
SR-8 RESILIENT
FOAM SEAL
RIM SPACE HEATING
& COOLING COILS
C
r>
u>
O U
z x
= o
1 o
J5
IU O
I O
_
PRbO
PRbOUCT LEVEL
1
Figure 2-3. POSITION OF THE CBI WEATHERMASTER ROOF
WITHIN THE EMISSIONS TEST TANK
13
-------
L
BOTTOM OF AIR OPENING
BWB/CBI 1OOO
SECONDARY SEAL
.CLIPS ON 3* CENTERS FASTENING SECONDARY
SEAL TO RIM OF ROOF
SR-6 PRIMARY SEAL IMMERSED IN BENZENE
^WEATHERM ASTER PAN TYPE
"INTERNAL FLOATING ROOF
Figure 2-4. RIM MOUNTING OF THE BWB/CBI 1000
FLAPPER SECONDARY SEAL
14
-------
(4) 10 inch diameter automatic bleeder vent
The guide pole and center column were actually only short
lengths of pipe supported 39 that they were in the proper
position within the floating roof penetration well. The open
ends of the pipes were sealed with polyethylene film to
minimize emissions through the pipes. More information on
the fittings and floating roof construction details is avail-
able in Reference 1.
Figure 2-5 is a cross-sectional sketch of the pan
type internal floating roof showing the deck penetrations
that could have contributed to the emissions.
ia* 9 GUIDE POLE WELL
16* « CENTER POLE WELL
ENDS SEALED WITH
PLASTIC FILM
PERMANENT FABRIC
SLEEVE SEAL
-------
During certain tests, the deck fittings were sealed
using two layers of 0.004 inch thick polyethylene plastic
bags taped and clamped over each fitting. A sketch of this
arrangement is shown in Figure 2-6.
2.3.1.2 Description of Seal Spacers
Certain tests (see Table 2-1 required the instal-
lation of metal seal spacers to create specified gaps between
the seal and the tank shell. Figure 2-7 details the spacers
used to create the seal gaps, and Table 2-2 summarizes the
seal gap areas created for the particular tests.
2.3.1.3 Description Of Test Conditions
The test conditions for Phase I are summarized in
Table 2-1. This table presents a brief overview of the
various temperatures, seal configurations, deck fitting
sealing and propane/octane reference tests, if any, for the
Phase I emission tests. Additional information on the test
conditions follows, where pertinent details of a particular
test are elaborated upon.
Tests EPA-1,2,3, and 4
These first four preliminary tests were conducted
over a limited windspeed range since they were used to learn
about the behavior of a single component product and to check
out the instrumentation.
During Test EPA-1, it was decided to seal the open
gage hatch, since it was typically sealed during the cor-
responding propane/octane tests.
It was noted during Test EPA-1 that the outlet con-
centration appeared to be affected by the ambient temperature.
As the ambient temperature increased with approaching mid-day,
the outlet concentration increased; as the ambient temperature
decreased after mid-day, the outlet concentration decreased.
During this behavior of cyclical concentration changes, the
bulk product temperature was below the ambient and blower out-
let temperatures at mid-day. It was thought that the increase
-------
Figure 2-6. TYPICAL TEMPORARY SEALING OP A
DECK PENETRATION FITTING
-------
TABLE 2-1. SUMMARY OF TEST CONDITIONS FOR PHASE I
oo
Test
No.
EPA-1
EPA-2
EPA- 3
EPA-4
EPA- 5
EPA-6
t
EPA-7
EPA-8
EPA- 9
EPA-10
EPA- 11
EPA-1 2
EPA- 13
EPA- 14
EPA-1 5
EPA- 16
rod.
Temp.
»F)
80
80
100
100
100
100
100
80
60
75
80
75
75
75
75
75
Primary
Seal
Gaps
None
None
None
None
None
4-l«sttx72"
None
None
None
None
t-lJs"x7211
None
l-l>il'x72"
1-1VX72"
None
Z-«j"x24"
Sec.
Seal
None
None
None
None
None
None
None
None
None
Yes
None
Yes
Yes
Yes
None
None
Sec.
Seal
Gaps
None
None
None
None
None
None
None
None
None
None
None
None
None
4-lJs"x72'
None
None
Gage
Hatch
Unsealed
Sealed
Sealed
Sealed
Sealed
Sealed
Unsealed
Sealed
Sealed
1 Sealed
Sealed
Sealed
Sealed
Sealed
Sealed
Sealed
Deck
Fittings
Unsealed
Unsealed
Unsealed
Sealed
Sealed
Sealed
Unsealed
Sealed
Sealed
Sealed
Sealed
Sealed
Sealed
Sealed
Sealed
Sealed
Reference
Test Nos.
(With Octane/Propane
(see Reference 1]
C19
C19
C19
C18R.
C18R
C19
C18R
C18R
C20R
C20R
C18R
Notes
i)
Partial Test
Partial Test
Partial Test
Partial Test
v
Void Test
-------
2j 2' 24*
mm
s,
8* 8-
• X
2' 2"
h 8"
^
^
MM
SEE DETAIL ABOVE
r T
NOTE (0). V FOR PHASE I AND PHASE tt TESTS, AND PHASE HI SECONDARY
SEAL GAPS; ("FOR PHASE m PRIMARY SEAL GAPS.
Figure 2-7. DESCRIPTION OF THE SHEET METAL SEAL SPACERS
-------
TABLE 2-2. SUMMARY OF SEAL GAP AREA FOR PHASES I, II AND III
Primary Seal
Phase
NO.
I . :
J
11
III
Test.
No.
EPA-6
EPA-11
EPA-13
EPA-14
EPA-16
EPA-18
EPA- 2 4
EPA- JSP
EPA- 25
EPA-26
EPA-28
EPA-29
Number and
Size of Gaps
4-1V x 72"
4-li»" x 72"
4-l"l" x 72"
4-1*" x 72"
2-"»" x 24"
2-1/2" x 24"
2-1" x 24"
2-1" x 24"
2-1" x 24"
2-1" x 24"
4-1V x 72"
4-1S" x 72
Total
Area of
Gaps
(in2)
420
420
420
420
26
26
68
68
68
68
288
288
/Gap Area \
(ft. Tank diaJ
(inV*t)
21
21
21
21
1.3
1.3
3.4
3.4
3.4
3.4
14.4
14.4
Percentage
of Tank
Circumfer-
ence Gapped
(»)
38.2
38.2
38.2
38.2
7.4
7.4
11-7
11.7
11.7
11.7
38.2
38.2
Secondary Seal '.
Number and
Size of Gaps
!
No secondary seal
No secondary seal
NO gaps
4.-1*" x 72"
No secondary seal
2-ij" x 24"
No secondary, seal
No gaps
No gaps
2-1," x 24"
2-l|" x 24"
No secondary seal
Total
Area of
Gaps
(in2)
_
-
0
420
—
26
-
0
0
26
26
—
/Gap Area \
\tt. Tank dl«J
(inVft)
—
-
0
21
™
1.3
-
0
0
1.3
1.3
~
Percentage
of Tank
Circumfer-
ence Gapped
<»)
^
-
0
38.2
-
7.4
-
0
. 0
7.4
7.4
-
to
o
-------
in outlet concentrations during the hottest part of the day
could be due to warming of the product which was in contact
with the floating roof. The increase in the product tem-
perature would result in an increase in the product vapor
pressure and hence the emission rate.
To eliminate the effect of the hot mid-day ambient
air on the cooler product, it was decided to raise the tem-
perature of the product to 100°F. Thus, the product was then
usually warmer than the blower outlet temperature. When this
was done, the mid-day increases in the outlet concentration
were not as evident. It was felt that the temperature
stratification of the product had been significantly decreased
by keeping the product temperature warmer than the air tem-
perature .
Tests EPA-2,3,4,5,6,7 and 8
During Test EPA-2, two (2) thermocouples were at-
tached to the top surface of the floating roof deck to de-
termine whether the deck temperature was approximately the
same as the bulk product temperature or whether the deck tem-
perature tended to rise and fall with the air temperature.
One thermocouple was located on the North-South centerline of
the floating roof deck near the air inlet; the other thermo-
couple was located on the East-West centerline of the floating
roof deck near the rim on the East side of the deck. These
thermocouples indicated that the deck temperature tended to
rise and fall with the blower outlet air temperature,thus
indicating that the temperature of the product just under the
floating roof deck could possibly be closer to the air tem-
perature than the average bulk product temperature.
Test EPA-10
The secondary seal used for Test EPA-10 was of a
neoprene construction. After Test EPA-10 was completed and
the tank was entered to install gaps in the primary SR-8 seal,
it was noted that the secondary seal had deteriorated due to
21
-------
exposure to the benzene vapors and was no longer in good
contact with the shell. Because of the poor contact of the
secondary seal with the shell due to the rapid effect of the
benzene vapors on the neoprene rubber, it was decided that
the data for Test EPA-10 was void, and a new secondary seal
was ordered to be fabricated from Viton. This new secondary
seal material did not show any deleterious effects of benzene
vapor exposure. This new secondary seal was then used for
the repeat of Test EPA-10, which was labeled Test EPA-12,
and was also used in subsequent Tests EPA-13 and EPA-14.
Tests EPA-13 and 14
In preparation for Test EPA-13, it was necessary
to install seal spacers between the primary seal and the
shell. When the secondary seal was pulled away from the tank
shell so that the seal spacers could be installed, it was
noted that the under surface of the secondary seal was damp.
This wetness was probably due to condensation of the benzene
vapors on the underaurface of the seal. The Viton construc-
tion of the secondary seal did not, however, show any de—-.
leterious effects of benzene contact with it. There was no
visible difference in appearance or feel of the Viton after
Test EPA-14 as compared with its new condition.
2.3.2 Phase II, Bolted Cover Type Internal Floating Roof
2.3.2.1 Description of Floating Roof and Seals
The internal floating roof for the Phase II tests
had been purchased from the Ultraflote Corporation of Houston,
Texas for use by CBI on earlier research work (Ref. 1). Prior
to the Phase II tests, an Ultraflote Corporation field crew
refurbished their roof, and a shingled flapper type of primary
and secondary seal were installed. A plan view sketch of a
portion of the seal is shown in Figure 2-8. Also, the dimen-
sions of a single piece, or shingle, of these seals is shown.
22
-------
4-
12
INDIVIDUAL PIECE OF SHINGLE TYPE SEAL
TANK SHELL
RIM PLATE
STEEL CLAMP BAR
PLAN VIEW
(the same detail was used for both
primary and secondary seals)
Figure 2-8. INSTALLED SHINGLE TYPE SEAL
23
-------
Additional details of the construction of the Ultraflote roof
are contained in Reference 1. Figures 2-9 and 2-10 describe
the details of the shingled flapper type of seal that was
installed in lieu of the single continuous flapper seal used
during the propane/octane tests. Figure 2-9 shows a cross-
sectional view of the position of the bolted cover type inter-
nal roof within the test tank.
The following is a list of the floating roof deck
penetrations:
(1) 32 in square access hatch well, with hinged
cover
(2) 32 in square column well, with a sliding, cover
plate and close fitting ladder penetration
seal
(3) 10 in diameter open sample well
(4) 10 in diameter vacuum breaker
(5) 1/4 in diameter stainless steel anti-rotation
cable housing
(6) Four (4), 1/4 in diameter open deck drains
Only fittings (1) through (4) are expected to have contributed
much to the emissions.
2.3.2.2 Description of the Seal Spacers
The seal spacers used in Phase II Test EPA-18,
were the two 1/2" x 24" spacers detailed on Figure 2-7.
They were located, one each, in the NE and SW quadrants of
the test tank.
2.3.2.3 Description of Test Conditions
The test conditions for Phase II are summarized
in Table 2-3.. This table presents a brief overview of the
various temperatures, seal configurations, deck fitting
sealing and propane/octane reference tests (if any) for the
Phase II emission tests. Additional information on the test
conditions follows, where pertinent details of a particular
test are elaborated upon.
24
-------
REMOVABLE EXTERNAL
CONE ROOF
AIR PLENUM
30V AIR DUCT
t
CO
O
r>
RIM SPACE HEATING
& COOLING COILS
T
«/>
0
o
o
4 t
u>
(0
• <
Figure 2-9. POSITION OP THE ULTRAFLOTE INTERNAL BOLTED COVER
WITHIN THE EMISSIONS TEST TANK
25
-------
SECONDARY SEAL
STEEL CLAMP BAR
BOLTED JOINT
PRIMARY SEAL
FOAM TAPE
FABRIC SEAL FOR MOUNTING BRA
MOUNTING BRACKET FOR SECONDARY
SEAL
FOAM TAPE
RIM PLATE
DECK SKIN-
DECK SKIN CLAMP BEAM ASSEMBLY
DECK SKIN
7* TO 8" ^
RIM SPACE
RIM PLATE
T
Figure 2-10
CROSS-SECTIONAL VIEW OF THE SHINGLE TYPE
SEAL INSTALLATION
-26
-------
Table 2-3. SUMMARY OF TEST CONDITIONS FOR PHASE II
Test
No.
EPA-17
EPA-18
EPA-19
EPA-20
EPA-21
EPA-22
Product
Temp.
(°P)
75
75
75
75
75
75
Primary
Seal
Gaps
None
2-1/2 "x24"
Nona
None
None
None
Sec.
Seal
Yes
Yes
Yes
Yes
Yes
Yes
Seo.
Seal
Gaps
None
2-1/2 "x24"
None
None
None
None
Deck
Fittings
Sealed
Sealed
Sealed
Sealed
Sealed
Unsealed
Reference
Test Nos.
With Octane/
Propane)
C7
[see Ref. 1]
• .
Notes
Rim space temporarily sealed
with plastic film.
Rim space temporarily sealed
with plastic film, and deck
seams also sealed.
•
Same conditions as EPA-20, but
with additional sealing of deck
seams .
Same conditions as EPA-21, but
with all the temporary seals
removed from the deck fittings.
to
-------
EPA-17 - The secondary seal was mounted in place before the
roof was installed into the test tank. All of the deck pene-
trations were temporarily sealed using Volara tape and two
(2) 0.006 thicknesses of polyethylene film. The deck seams
were left in the "as-built" condition, with no temporary
sealing. No sealing of the deck drains was done during this
testing program.
One of the deck drains was used as a passage for
the thermocouple wire used to monitor the bulk product tem-
perature at a depth of three feet. Another thermocouple was
attached to the top of the deck to monitor the deck skin tem-
perature. This thermocouple was labeled SW. No. 12 on the
Temperature Data Sheets. A third thermocouple was passed
through a fitting in the deck, and was used to monitor the
temperature of the vapor space between the deck and the liquid
surface. This thermocouple was labeled SW. No. 11 on the Tem-
perature Data Sheets.
EPA-18 - Two (2) - 1/2 inch wide by 24 inch long gaps were in-
stalled between the seal and the sheel at the NE and SW tank
locations. It was intended that this test would be performed
with only the primary seal being gapped. However, when the
spacers were being placed, it became evident that when the pri-
mary seal was being held away from the shell by the spacer, the
primary seal also held the secondary seal away from the shell
because of interference with the secondary seal. Therefore,
this test was actually performed with both seals gapped at the
two locations.
The seal spacers were the same ones used for Test
EPA-16 in Phase I, and they are detailed in Figure 2-7.
EPA-19 - The objective of Test EPA-19 was to determine the
emission rate through the bolted deck of the non-contact
type internal cover. To this end, Test EPA-19 was performed
with all of the deck penetrations sealed and rim space seals
temporarily covered with a sheet of plastic film. The plastic
film was glued and taped to the tank shell above the secondary
28
-------
seal and to the deck at the base of the rim seals. This film
was a nylon reinforced clear plastic, 0.006" thick, Type T-55,
manufactured by Griffolyn Co. Subsequent to the tank emission
tests, the Type T-55 film was determined by laboratory tests
to have a permeability of approximately 0.18 Ib/day ft2 when
the film was placed over benzene liquid at 70°F. There was
approximately 120 ft2 of film installed over the rim space
seals, which would have thus passed at most about 22 Ib/day
of benzene vapor. The actual level of benzene emissions
measured during Test EPA-19 was only about 6 Ib/day at 10 mph
wind speed. Since the measured value was less than the maxi-
mum possible expec.ted value, based on the laboratory permea-
bility tests, it is suspected that the primary and secondary
seal helped to reduce the benzene emission rate by reducing
the benzene partial pressure below the plastic film. It is
still, however, possible that some of the measured benzene
emissions came in part from other sources such as perhaps
leakage through the deck seams.
Test EPA-19 was similar in conditions to Test C7,
performed with a propane/octane.
EPA-20 - The rim space was left temporarily covered with the
plastic film from Test EPA-19. In addition to covering the
rim seals with plastic film to reduce the benzene emission
level, the deck lap seams were sealed as indicated on Figure
2-11. The plastic film was attached to the side of the
clamp beams and to the deck skin where the edge of the upper
sheet was exposed.
EPA-21 - After Test EPA-20 was completed, a survey for higher
local concentrations on the deck was made with the portable
analyzer. Higher local concentrations were noted along the
clamp beams on the opposite side from the previously taped
seams. Also, some higher readings were noted at the attach-
ment seams of the deck fittings to the deck. All of these
29
-------
TAPE
PLASTIC FILM GLUED
AND TAPED TO THE
CLAMP BEAM AND
DECK SKIN
TAPE
LL
D
— UPPER CLAMP BEAM
ALUMINUM DECK SKIN
LOJ
LOWER CLAMP SEAM
.Figure 2-11. SEALED DECK LAP SEAM FOR TEST EPA-20
30
-------
locations were then additionally sealed with plastic film and
tape. Then, data was recorded for a wind speed of approxi-
mately 21 MPH. When this was compared to the emission rate
at approximately 20 MPH from the previous test, it was noted
that there was no measurable reduction in emission rate.
Further test data was not therefore taken, and preparations
were made for the next test.
EPA-22 - Upon cessation of data taking for Test EPA-21, the
temporary seals on the deck penetrations were removed and
Tests EPA-22 was performed. This was the last test to be
performed under Phase II. The temporary plastic film seal
between the shell and the deck was kept in place for this
test. Also, all of the deck seams from Tests EPA-20 and 21
remained as they had been sealed for those tests.
2.3.3 Phase III> External Double Deck Floating Roof
2.3.3.1 Description of Floating Roof and Seals
Figure 2-12 illustrates an SR-1 metallic shoe seal
mounted on a double deck external floating roof. When a
secondary seal was required, the Viton flapper type secondary
seal from Phase I was reused. However, in order to fit it to
the double deck roof, the circumference of the secondary seal
had to be shortened because of the slightly smaller diameter
of the double deck roof.
A cross-sectional view of the position of the
double deck roof within the test tank is shown in Figure 2-13.
2.3.3.2 Description of the Seal Spacers
The specified gaps between the primary seal metallic
shoes and the tank shell were achieved by inserting spacer
bars, as detailed in Figure 2-14, between the shoes and the
tank shell. When gaps were required between the secondary
seal and the tank shell, the 1/2" x 24" long spacers detailed
in Figure 2-7 were used.
31
-------
FLEXURE
« FLEXURE !'
CLOSURE ''
PANTAGRAPH
HANGER
SEALING «
RING
TANK
SHELL
Figure 2-12
GENERAL ARRANGEMENT OF AN
SR-1 METALLIC SHOE SEAL
MOUNTED ON A CBI
'FLOATING ROOF
32
-------
REMOVABLE EXTERNAL
CONE ROOF
DOUBLE DECK EXTERNAL
FLOATING ROOF
RIM SPACE HEATING
& COOLING COILS
SHELL HEATING
COOLING
JACKET
Figure 2-13. POSITION OF THE CBI DOUBLE DECK
EXTERNAL FLOATING ROOF WITHIN
THE TEST TANK
, 33
-------
_
I
to
n
*
r»
i
t
o
1
*o
I
Vt
MMMB^W
•
(
f
5
)
)
ww
.Hr
NOTE: DIMENSION "a" WAS i" FOR TESTS EPA-24,25P,25 AND 26
DIMENSION "a" WAS l'«" FOR TESTS EPA-28 AND 29
Figure 2-14. METALLIC SHOE SPACER BAR
.34
-------
2.3.3.3 L Description of Test Conditions
The test conditions for Phase III are summarized in
Table 2-4. This table present a brief overview of the various
temperatures, seal configurations, deck fitting sealing and
propane/octane reference tests, if any, for the Phase III
emission tests. Additional information on the test conditions
follows, where pertinent details of a particular test are
elaborated upon.
Test EPA-23 - This first test of Phase III was performed on an
ungapped, primary, SR-1 metallic shoe seal and no secondary
seal.
Test EPA-24 - Following the completion of Test EPA-23, two (2)
1" x 24" gaps were created between the metallic shoes of the
primary seal and the tank shell by installing two metallic
shoe spacer bars. The locations of the gaps were the NE and
SW tank locations. The resultant gaps had the following
dimensions:
1
10"
>i, . i t y j i '//'/S//SSS
i
24"
10"
Test EPA-25? - A flapper type secondary seal was installed
after Test EPA-24 was completed. This seal was mounted with
clips on the rim plates of the double deck floating roof
analogous to the mounting of the secondary seal during the
Phase I tests. See Figure 2-12.
This secondary seal, as ordered for Phase I, was
designed and custom fabricated for the slightly larger diameter
35
-------
TABLE 2-4. SUMMARY OF TEST CONDITIONS FOR PHASE III
en
Test
NO.
EPA-23
EPA-24
EPA-25P
EPA-25
EPA-26
EPA-27
EPA-28
EPA- 2 9
t
Product
Temp.
CF)
75
75
75
75
75
75
75
75
Primary
Seal
Gaps
None
2-l»x24"
2-l"x24"
2-l"x24"
2-l"x24"
None
4-1 1/2 "x72"
4-1 1/2 "x72"
Sec.
Seal
None
None
Yes
Yes
Yes
Yes
Yes
None
Sec.
Seal
Gaps
None
None
1-1 l/4"x
377"
None
2-1/2 "x24"
None
2-1/2 "x24"
/ None
Reference
Test ivlos .
(filth/Octane
Propane)
W1,W1R,W2,V3
[see Ref. 2]
W5
(see Ref. 2]
W10
[see Ref. 2]
W24
[see Ref. 4]
W26
[see Ref. 4]
Notes
Deck fittings sealed for all
tests.
-------
Weathermaster roof. When this seal was cut shorter to fit
the double deck roof and installed, it was noted that there
was a gap approximately 1-1/4 inches at the widest point, ex-
tending from the North point of the tank, along the West side
to the South point of the tank. Thus, the secondary seal was
gapped approximately 1-1/4 inches for at least one half of the
tank circumference. The remainder of the secondary seal fit
tightly against the tank shell.
It was decided to take some preliminary data with
the secondary seal in this condition and this data was labeled
Test EPA-25P.
Test EPA-25 - After preliminary Test EPA-25P, the emissions
test tank was entered for the purpose of inspecting the sec-
ondary seal and installing some mechanical devices to hold the
secondary seal against the shell so as to eliminate the gap.
Upon inspecting the secondary seal, it was noted that the
large gap previously observed was no longer evident. Ap-
parently, during the course of the performance of Test EPA-25P,
the secondary seal material had stretched enough so that the
seal now was in contact with the shell of the tank. However,
there were still some areas where the secondary seal did not
fit properly, and mechanical adjustments were made to achieve
a proper fit. Then, Test EPA-25 was performed.
Test EPA-26 - Prior to performing this test, two sheet metal
spacers were inserted between the secondary seal and the tank
shell. They were located in the NE and SW tank locations,
above the gaps previously installed in the primary seal.
These sheet metal spacers were the same ones used in Phases
I and II, and are detailed in Figure 2-7.
Test EPA-27 - All of the primary and secondary seal spacers
were removed before this test was performed. Thus, both the
primary and secondary seals were continuously tight against
the tank shell.
Test EPA-28 - For this test, spacer bars were installed be-
37
-------
tween the metallic shoes and the tank shell so as to create
four (4) gaps 1-1/2 inches wide and 72 inches long in four
tank locations: NE, SE, SW, and NW. The spacer bars used
were similar to those shown in Figure 2-14, except that they
were wider, having two rows of bolts, and the bolts were lon-
ger so as to create a 1-1/2 inch wide gap. The resultant gaps
in the four tank locations had the following dimensions:
1
24"
1
24"
t^-.
24"
The secondary seal also was gapped, but only 5.n two locations:
the NE and the SW. The sheet metal spacers, detailed in
Figure 2-7, were again used to form two (2) gaps 1/2" x 24".
Test EPA-29 - The last test performed under Phase III was .
Test EPA-29. Prior to performing this test, the secondary
seal was removed, but the primary seal gaps from Test EPA-29
were left in place.
38
-------
2.4 REFERENCES
1. Cherniwchan, W. N. and R. J. Laverman. Hydrocarbon
Emission Loss Measurements on a 20 foot Diameter Pilot
Test Tank with an Ultraflote and a CBI Weathermaster
Internal Floating Roof. Chicago Bridge & Iron Company
Research Report. Research Contract R-0113/R-0191.
June 1978.
39
-------
3.0 TEST RESULTS
3.1 BENZENE PRODUCT TEST RESULTS
The test data is summarized in Table A-l. The raw test
data consists of:
(1) air flow rate,(scfm),(Column 2)
(2) product temperature,(°F),(Column 3)
(3) (outlet-inlet) concentration,(ppmv,
benzene basis),(Column 5)
For each of the product temperatures, the benzene vapor
pressure is listed in Column 4. The benzene vapor pressure
nay be estimated by using Equation 3.1.(Ref. 1).
./5019.312 \
P- 155,609e~V365-422+T/ . (3.1)
Between the temperatures of GOOF and 100°F,Equation (3.1)
gives the following benzene vapor pressure.
T
Temperature
(°F)
50
55
60
65
70
75
80
85
90
95
100
?
Vapor Pressure
(psia)
0.881
1.017
1.170
1.342
1.534
1.743
1.987
2.252
2.54b
2.B58
3.224
40
-------
The measured (outlet-inlet) benzene concentration changes
listed in Column 5 for each test were normalized to a common
vapor pressure for that test. The common vapor pressure
selected for each test is indicated in the footnotes of Table
A-l. Equation (3.2) was used to perform the normalization
calculations .
ACNORM. = ACMEAS.r 3 , NORM.] (3.2)
\*3fMEAS,/
Where *3 is defined by Equation (3.H) .
\
The emission rate listed in Column 7 of Table A-l was
calculated using Equation (3.3).
(E'iiy) - (S'MnO (AC NORM.'??1™ benzene basis V
U60 min\(24 hr \ f 1 Ibmole \ xl fract.x (78.114 Ib ]\
L hr '\ day/ V379.49 sftV\l°6 PPm / Ibmole /]
E,lb \ =^2.96xlO~lfWq,sft3\ /AC..-.-.. ,ppmv, benzene basis)
dTy/ V /\ Sin"/ V NORM' '
(3.3)
For the tests during Phases I and II on internal
floating roofs, the equivalent wind speeds listed in Column
8 of Table A-l were calculated using Equation (3.4) (see pg.
78 Ref . 2) .
V, mi\ = (O.OIS9] /q,sft3N (3.4)
hFJ V M iSin" J
41
-------
For the tests during Phase III on the external floating roof,
the equivalent wind speeds listed in Column 8 of Table A-l
were calculated using Equation (3.5) (see pg. 5 of Ref. 3).
15 ,- ,
(V ,mi W2. 96x10- M /a, sft3U .
V Kr7 V / \" 5In /
The calculated emission rate and wind speed results
for the tests are plotted on Figures 3-1 through 3-11. In
most cases, the measured concentrations were normalized to
a benzene vapor pressure of 1.75 psia (75°F) , but for Tests
EPA-5 and 9 they were also normalized to the vapor pressure
at the nominal product test temperatures of 100°F and 60°F,
respectively, to show the dependency of the emission rate
on vapor pressure. Figure 3-1 presents the results of Tests
EPA-5, 9 and 15, and these are further discussed in Section
3.2.
Table A-2 lists the results of Tests EPA-5, 8 and 9
normalized to a benzene vapor pressure of 1.75 psia (75°F)
instead of to the vapor pressure at the nominal product test
temperatures , as was listed in Table A-l .
It should be noted that Test EPA-8 was repeated in
Test EPA-15, and due to the scatter in the results of Test
EPA-8, only the results of Test EPA-15 were plotted on Figure
3-1.
3.2 PROPANE/OCTANE PRODUCT TEST RESULTS
Many of the tests performed with benzene product
had the same seal conditions that had been tested earlier with
a product which was a propane/octane mixture. The reference
test numbers are listed in Tables 2-1 , 2-3 , and 2-4 . The
measured test data are listed in Table A-3.
42
-------
10O
10-
o
T3
Ul
c
o
'55
(A
'i
Id
O.1
Product temperature 100*7,
Test EPA-5 O
Product temperature 75*P,
Test EPA-15 $
Product temperature SO*F,
[ ^ Ta*t EPA-9 fi^
SR-8 flexible foam primary seal
without gaps, results normalized
to the vapor pressure correspond-
ing to the stated temperature and
78.1 vapor molecular weight.
1O
Wind Speed
4O m ph
Figure 3-1.
EMISSIONS VS. WIND SPEED FOR AN
INTERNAL PAN TYPE FLOATING ROOF,
PRIMARY SEAL ONLY,
PRODUCT AT VARIOUS TEMPERATURES
43
-------
10O
4 Gaps, l>i" x 72" in primary seal,
deck fittings sealed,
Tests EPA-11 Q and EPA-6 Q
2 Gaps, h' x 24* in primary seal,
deck fittings sealed,
Test EPA-16 b
No gaps in primary seal,
deck fittings unsealed,
Test EPA- 7 O
No gaps in primary seal,
deck fittings sealed,
Test EPA- 15 0
SR-8 flexible foam primary seal,
results normalized to 1.75 psia
vapor pressure and 73.1
vapor molecular weight
1O
Wind Speed
4O m ph
Figure 3-2.
EMISSIONS VS. WIND SPEED FOR AN
INTERNAL PAN TYPE FLOATING ROOF,
PRIMARY SEAL ONLY
44_
-------
100
10
c
13
Benzene (Single Component) Tests
Symbol
Deck Fittii
0
Deck Fitti;
o
Test NO.
igs Sealed:
EPA- 5
EPA- 9
EPA- 15
igs Unsealed:
EPA-7
o
*5!
.2
UJ
Propane/Octane (Binary Mixture)
Tests
Symbol
Deck Pitti
' • '
D«ck Fitti
+
Test NO.
198 Sealed:
C18R
ngs Unsealed:
C19
SR-8 flexible foam primary seal
without gaps, no secondary seal,
results normalized to 1.75 psia
vapor pressure and
,_78.1 vapor molecular weight
1O
4O m ph
Wind Speed
Figure 3-3.
EMISSIONS VS. WIND SPEED FOR AN
INTERNAL PAN TYPE FLOATING ROOF,
COMPARISON WITH PROPANE/OCTANE TEST DATA
45-
-------
100
1O
o
•x.
a
c
O
°3>
.12
UJ
0.1
4 Gaps, 1-1/2" x 72" in primary seal,
4 gaps, 1-1/2" x 72 in secondary seal,
Test EPA-14 <3
No gaps in primary seal,
no gaps in secondary seal,
Test EPA-12 Q
4 Gaps, 1-1/2" x 72" in primary seal,
no gaps in secondary seal,
Test EPA-13 O
SR-8 flexible foam primary seal
and flapper secondary seal, deck
fittings sealed, results normalized
bo 1.75 psia vapor pressure and
78.1 vapor molecular weight
1O
4O m ph
Wind Speed
Figure 3-4.
EMISSIONS VS. WIND SPEED FOR AN
INTERNAL PAN TYPE FLOATING ROOF
WITH PRIMARY AND SECONDARY SEAL
4~6
-------
10
a
•a
0.1
vt
c
«»
in
'e
in
0.01
Benzene (single component),
Test EPA-12 Q
Propane/octane (binary mixture)
Test C20R«
SR-8 flexible foam primary seal
without gaps, flapper secondary
seal without gaps, deck fittings
sealed, results normalized to
1.75 psia vapor pressure and
78.1 vapor molecular weight
1O
4O m ph
Wind Speed
Figure 3-5.
EMISSIONS VS. WIND SPEED FOR AN
INTERNAL PAN TYPE FLOATING ROOF
WITH PRIMARY AND SECONDARY SEAL,
COMPARISON WITH PROPANE/OCTANE TEST DATA
47
-------
10O
2 Gaps, 1/2* x 24" in primary seal,
2 gaps, 1/2' x 24" in secondary seal,
deck fittings sealed.
Test EPA-18Q
No gaps in primary seal,
no gaps in secondary seal,
deck fittings sealed.
Test EPA-17Q
Plastic film over primary and
secondary seal,
deck fittings not sealed,
Tests EPA-22 Q
Plastic film over prinary and
secondary seal,
deck fittings sealed,
Tests EPA-19O, EPA-20Aand EPA-21O
0.1
Shingled flapper primary and
secondary seal, results normalized
to 1.75 psia vapor pressure and
78.1 vapor molecular weight
1O
Wind Speed
4O m ph
Figure 3-6.
EMISSIONS VS. WIND SPEED FOR A
BOLTED COVER TYPE INTERNAL FLOATING ROOF
WITH PRIMARY AND SECONDARY SEAL
48
-------
100
10
o
•o
V)
c
o
"35
,2
UJ
0.1
Benzene (single cor.nanent) tests,
shingled flapper primary seal,
shingled ;iacner_3oecndary
No gaps in prinary seal,
no gaps in secondary seal.
deck fittings sealed,
Test EPA-17 O
2 Gaps, 1/2" x 24' in primary seal,
2 gaps, 1/2" x 24" in secondary seal,
deck fittings sealed.
Test EPA-18 Q
Plastic film over primary and
secondary seal,
deck fittings sealed,
Teats SPA-19Q, EPA-20£and SPA-21Q
Propane/octane (binary mixture) tents
flapper primary seal without gaps,
no secondarv seal:
deck fittings not sealed,
Test C5+
deck Fittings sealed,
Test C6•
Plastic film over primary seal,
deck fittings sealed,
Test C7+
Results normalized to 1.75 psia
vapor pressure and
78.1 vapor molecular weight
10
Wind Speed
4O mph
Figure 3-7.
EMISSIONS VS. WIND SPEED FOR A
BOLTED COVER TYPE INTERNAL FLOATING ROOF
WITH PRIMARY AND SECONDARY SEAL,
COMPARISON WITH PROPANE/OCTANE TEST DATA
..49
-------
100
1O
a
•o
V)
c
o
'55
|