STANDARD SUPPORT
ENVIRONMENTAL IMPACT STATEMENT
FOR CONTROL OF
BENZENE FROM THE GASOLINE MARKETING INDUSTRY
Draft Report
June 21, 1978
U. S. Environmental Protection Agency
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
26803
TABLE OF CONTENTS
Page
Chapter 1.0 Health Effects
Chapter 2.0 Industry Description 2-1
2.1 Benzene Correlation to Hydrocarbon 2-3
2.2 Bulk Gasoline Terminals . 2-6
2.3 Bulk Gasoline Plants Z'-IO
2.4 Service Stations 2-14
2.5 Gasoline Tank Trucks 2-17
2.6 References for Chapter 2.0 2-21
Chapter 3.0 Emission Control Technology 3-1
3.1 Use of Control Methods 3-1
3.2 Bulk Terminals 3-2
3.3 Gasoline Bulk Plants 3-11
3.4 Service Stations 3-20
3.5 Gasoline Tank Trucks 3-27
3.6 Reduction of Benzene Content of Gasoline 3-27
3.7 Summary 3-32
3.8 References for Chapter 3.0 3-34
Chapter 4.0 Alternative Control Levels 4-1
4.1 Option 1 4-3
4.2 Option 2 4-4
n i
-------
Page
Chapter 4.3 Option 3 4-4
4.4 Option 4 4-5
4.5 References for Chapter 4.0 4-8
Chapter 5.0 Environmental Impacts of Applying the Technology . 5-1
5.1 Impact on Benzene Emissions 5-1
5.2 Other Air Impacts 5-5
5.3 Water Pollution Impact 5-10
5.4 Impact on Solid Waste 5-10
5.5 Impact on Energy 5-13
5.6 Air Quality Impact 5-13
5.7 Other Environmental Concerns 5-16
5.8 References T 18
Chapter 6.0 Economic Impact Analysis 3-1
6.1 Bulk Terminals ."-1
6.2 Bulk Plants 3-47
6.3 Service Stations :-99
6.4 Reduction of Benzene Content in Gasoline C-l 14
6.5 Total Costs of Gasoline Marketing Control Options . C-118
6.6 References C-122
Appendix B Index to Environmental Impact Considerations . . . . B-l
Appendix C Emission Source Test Data C-l
C.I Bulk Terminal Tests C-l
C.2 Bulk Plant Tests C-16
C.3 Service Station Tests C-21
C.4 Derivation of Benzene to Hydrocarbon Ratio for
Gasoline Vapor C-25
iv
-------
Page
C.5 References for Appendix C C-30
Appendix D Emission Monitoring and Compliance Testing
Technique D-l
D.I Emission Measurement Methods D-l
D.2 Performance Test Methods D-7
D.3 Continuous Monitoring D-9
Appendix E State and Local Hydrocarbon Regulations for
Gasoline Marketing E-l
Appendix F Description of OSHA Benzene Regulation F-l
-------
LIST OF TABLES
Page
Table 2-1 National Benzene Emissions From the Gasoline
Marketing Industry 2-11
Table 3-1 Summary of Benzene Emission Tests 3-3
3-2 Average Benzene Content of 1981 U.S. Gasoline Pool . . 3-30
3-3 Effect of Control Techniques on Emissions 3-33
Table 4-1 Gasoline Marketing Network Control Options 4-6
4-2 Gasoline Marketing Control Options - National
Emissions 4-7
Table 5-1 Estimated Impact on Benzene Emissions for Typical
Facilities 5-4
5-2 Gasoline Marketing Control Options - Estimated
National Emissions 5-6
5-3 Other Air Impacts for Typical Facilities 5-8
5-4 Estimated National Air Impacts Other Than Benzene . . 5-9
5-5 Water Impact.- Worst Case 5-11
5-6 Solid Waste Impact - Worst Case 5-12
5-7 Energy Impacts of Control Method 5-14
5-8 Air Quality Impact 5-15
5-9 Other Environmental Concerns 5-17
Table 6-1 Bulk Terminal Population 6-5
6-2 Regional Product Supply/Demand - 1976 6-7
6-3 Regional Gasoline Supply/Demand - 1976 6-8
6-4 Bulk Terminal Storage Distribution 6-10
6-5 Bulk Terminal Throughput Distribution 6-11
6-6 Bulk Terminal Ownership 6-13
VI
-------
Page
Table 6-7 Gasoline Terminal Distribution by Size and Ownership . . .6-14
6-8 Bulk Terminal Employment 6-16
6-9 Model Terminal Parameters 6-19
6-10 Cost Factors Used in Developing Annualized Cost Estimates
for Model Terminals 6-21
6-11 Estimated Control Costs for Model Existing Terminals . . .6-23
6-12 Estimated Control Costs for Model Existing Terminals . . .6-25
6-13 Cost Effectiveness for Model Existing Batteries 6-28
6-14 Summary of Hydrocarbon Control Systems Costs for
Actual Terminals 6-30
6-15 Estimated Control Costs and Cost Effectiveness for
Model New Terminals 6-33
6-16 Bulk Terminal Closures Due to Vapor Control
Economics 6-40
6-17 Vapor Control Employment and Costs Impacts at Bulk
Terminals -6-41
6-18 Vapor Control Costs at Bulk Terminals . . , 6-43
6-19 Gasoline Trailer Population 6-44
6-20 Total Cost to Install Vapor Control on the Gasoline
Trailer Fleet 6-46
6-21 Bulk Plant Population 6-49
6-22 Bulk Plant Storage Distribution 6-51
6-23 Bulk Plant Throughput Distribution 6-52
6-24 Bulk Plant Ownership . 6-53
6-25 Gasoline Bulk Plant Distribution by Size and Ownership . .6-55
6-26 Bulk Plant Employment 6-56
6-27 Parameters Used for Cost Estimates 6-62
6-28 Options 1 and 3 Capital and Annualized Cost Estimates . , 6-63
-------
Page
6-29 Option 4 Capital and Annualized Cost Estimates 6-64
6-30 Options 1 and 3 Cost Estimates 6-65
6-31 Option 4 Capital and Annualized Cost Estimates 6-66
6-32 Options 1 and 3 Capital and Annualized Cost Estimates . . 6-67
6-33 Option 4 Capital and Annualized Costs 6-68
6-34 Comparison of Capital and Annualized Costs 6-70
6-35 Cost Effectiveness 6-73
6-36 Bulk Plant Closures due to Inaccessibility of
Capital 6-84
6-37 Bulk Plant Closures due to Insufficient Profitability . . 6-85
6-38 Closure Summary at Bulk Plants 6-87
6-39 Closure Impact at Bulk Plants by Ownership 6-88
6-40 Employment Displaced at Bulk Plants 6-90
6-41 Estimated Employment Impact at Bulk Plants by
Ownership 6-91
6-42 Vapor Control Costs at Bulk Plants 6-93
6-43 Vapor Control Costs at Bulk Plants Based Upon
NOJC Costs 6-94
6-44 Vapor Control Costs at Bulk Plants Based Upon
Houston-Galveston Costs 6-95
6-45 Vapor Control Costs at Bulk Plants Based Upon
Colorado APCD Costs 6-96
6-46 Total Costs of Vapor Control at Gasoline Bulk
Plants by Ownership 6-97
6-47 Summary of Service Station Population 6-102
6-48 Distribution of Private Gasoline Dispensing
Outlets 6-104
6-49 Summary of Costs for Stage I Balance System 6-109
6-50 Cost-Effectiveness Estimates for Stage I Balance
Systems 6-110
vm
-------
Page
6-51 Total Service Station Industry Stage I Costs 6-112
6-52 National Cost of Benzene Removal From Reformate and
FCC Gasoline 6-116
6-53 Total Costs and Cost-Effectiveness of Gasoline
Marketing Control Options 6-121
Table C-l Summary of Bulk Terminal Gasoline Loading Vapor-
Control Devices C-l5
C-2 Service Station Bulk Drop Data C-26
C-3 Colonial Pipeline C-28
Table E-l State and Local Regulation of Hydrocarbons E-l
-------
LIST OF FIGURES
Page
Figure 2-1 The Gasoline Marketing Distribution System
in the United States 2-2
2-2 Data Summary Benzene/Hydrocarbon Vapor Relationship 2-5
2-3 Schematic of Bottom-Loading Tank Truck Terminal . . 2-7
2-4 Gasoline Tank Truck Loading Methods 2-8
2-5 Typical Bulk Plant 2-13
2-6 Tank Truck Unloading into an Underground Service
Station Storage Tank 2-16
Figure 3-1 Schematic of Refrigerated Vapor Recovery Unit . . . 3-5
3-2 Schematic of Compressor-Refrigeration-Absorption
System 3-7
3-3 Schematic of Adsorption-Absorption Vapor Recovery
System 3-9
3-4 Schematic of Thermal Oxidation System 3-10
3-5 Top Loading Systems 3-13
3-6 Top and Bottom Loading Systems 3-15
3-7 Vapor Balance System (Bottom Fill) 3-17
3-8 Vapor Balance System at a Service Station 3-21
3-9 Benzene Reduction by Component at 94.5 percent
Control 3-30
Figure 6-1 Product Flow Diagram 6-2
6-2 Comparison of Installed Capital Costs for Benzene
Control Systems 6-31
-------
Page
Figure 6-3 Cost Effectiveness for Most Expensive Equipment
(Top Loading) Bulk Plants 6-74
6-4 Cost Effectiveness for Most Expensive Equipment
(Bottom Loading) Bulk Plants 6-75
6-5 Cost Effectiveness for Less Expensive Equipment
(Top Loading) 6-76
6-6 Cost Effectiveness for Less Expensive Equipment
(Bottom Loading) 6-77
6-7 Cost Effectiveness for Least Expensive Equipment
(Top Loading) 6-78
6-8 Cost Effectiveness for Least Expensive Equipment
(Bottom Loading) 6-79
6-9 Cost of Benzene Removal vs. Gasoline Production
Using Refinery-Produced Hydrogen 6-119
XI
-------
CHAPTER 2. INDUSTRY DESCRIPTION
The gasoline marketing network consists of all storage and trans-
portation of gasoline from refinery to motor vehicle fuel tanks. It
includes pipelines, ships and barges, trucks and rail cars, and storage tanks.
Emissions occur as gasoline is stored in or loaded and unloaded from these
sources.
This document discusses four of the major benzene source categories
in this marketing chain: loading of trucks at bulk plants and terminals and
storage at bulk plants and service stations. Motor vehicle loading and bulk
terminal storage tanks will be examined in separate studies. Figure 2-1
illustrates the marketing network.
Gasoline is delivered to the terminal from the refinery via pipeline
or by ships and barges. Large transport trucks (30,000-36,000 liters or
8000 - 9500 gallon capacity for each cargo trailer) then deliver the gasoline
to service stations or intermediate bulk storage areas known as bulk plants.
Bulk plants, using 5700-11,000 liter (1500-3000 gallon) capacity delivery
trucks primarily service agricultural accounts and certain service stations
that are either long distances from terminals or inaccessible to the large
transports. In 1977 approximately 60 percent of gasoline delivered to service
stations came from terminals and 40 percent came from bulk plants. There
has been a trend in recent years for less bulk plant deliveries and more
terminal deliveries.
This document uses the term "service stations" to describe both the
familiar retail outlets and the non-retail and miscellaneous outlets such as
2-1
-------
FIGURE 2-1. THE GASOLINE MARKETING DISTRIBUTION SYSTEM IN THE
UNITED STATES
REFINERY STORAGE
SHIP, RAIL, BARGE
BULK TERMINAL
STORAGE
TANK TRUCK
AUTOMOBILES,
TRUCKS
i
PIPELINE
1
t
SERVICE STATIONS
STORAGE
SULK PL
STORAG
1
TRUCK
I
J
COMMERCIA
USERS, ST
2-2
-------
fleet services (rental car agencies and governmental agencies), parking
garages, and large agricultural accounts. (All non-retail stations receive
less than 50 percent of their revenue from the sale of gasoline.) It does
not include about 2.7 million small farms.
2.1 BENZENE CORRELATION TO HYDROCARBON
Data has been collected on the relationship of benzene emissions to
total hydrocarbon emissions from gasoline. Previous work in this area has
been done by H. E. Runion of Gulf Oil Corporation, and H. J. McDermott and
2 3
S. E. Killiany of Shell Oil Company. EPA has also done limited testing
to determine the relative concentration of benzene to hydrocarbon in
gasoline emissions.
In his study, Runion conducted laboratory tests on a premium leaded
gasoline and two regular leaded gasolines differing in octane levels. The
equilibrium vapor phase was formed by injecting 50 ml of fresh gasoline into
a 4 ounce bottle equipped with a septum cap and a small wire stirrer.
After a suitable period for vapor equilibration at 25°C (77°F), vapor samples
were withdrawn and analyzed by gas chromatography. This procedure was
repeated on U. S. and European gasolines, and on a series of gasolines spiked
with benzene to achieve a broad spectrum of benzene liquid concentrations.
These analyses resulted in an approximate linear relationship between liquid
volume percent benzene in gasoline and benzene volume percent in the vapors.
These data are shown in Figure 2-2 as gm benzene/gm hydrocarbon. The data
and interpretation are outlined in Appendix C, Section C.4.
Runion also conducted tests to determine if benzene emission levels
might increase during gasoline weathering or evaporation. His conclusion
was negative.
2-3
-------
Shell Oil Company addressed the benzene/hydrocarbon relationship in
their analysis of 86 gasoline samples of different brands collected for
marketing research and process control. The average liquid composition of
these samples was about 1 weight percent. The average benzene concentration
in the gasoline vapors was about 0.7 volume percent (shown on Figure 2-2.
as gm benzene/gm hydrocarbon).
In addition, samples collected from ten gasoline storage tanks at
Colonial Pipeline Company, Greensboro, N.C., by EPA have been evaluated.
Premium leaded, premium unleaded, regular, and unleaded gasolines were
analyzed for liquid and vapor benzene concentrations at 27 C to 31 C
(80 to 87°F). These data are presented in Figure 2-2, along with the
Shell and Runion data.
Temperature has a major influence on vapor-liquid equilibrium
concentrations. The data in Figure 2-2 were obtained at temperatures
varying from 25 to 31°C (77 to 87°F). This accounts for some of the
irregularities of data on the graph. Any attempt to adjust the data to
other temperatures would introduce an indeterminable degree of error.
Therefore, this document will use the least squares correlation for
27°C (80°F) without adjustment to determine a benzene/hydrocarbon emission
factor for gasoline.
The current national average of benzene content in gasoline is
5
1.3 liquid volume percent. Figure 2-2 shows about 0.008 gm benzene/
gm hydrocarbon in the vapors over gasoline containing 1.3 liquid volume
percent benzene at 27°C (80°F). Therefore, this document will use a factor
of 0.008 gm benzene/gm hydrocarbon to estimate benzene losses in known
amounts of hydrocarbon emissions. (See Appendix C for further details of
this correlation.)
2-4
-------
FIGURE 2-2. DATA SUMMARY
BENZENE/HYDROCARBON VAPOR RELATIONSHIP
o
+J
O)
c
tsl
c
d)
M- O
O CL
O >
ro -r-
S-
c
1/5 O
Benzene in Liquid Gasoline, Volume Percent
2-5
-------
2.2 BULK GASOLINE TERMINALS
In 1972, the Bureau of the Census totalled the United States terminals
as 1,925. They defined a terminal as any bulk gasoline marketing outlet
which receives product by pipeline, ship, or barge, or which has a total
product storage capacity of 7.95 million liters (2.1 million gallons) or
greater. A bulk plant was defined as a wholesale marketer of gasoline having
a total product storage capacity less than 7.95 million liters (2.1 million
gallons). Further, it was noted that the plant typically received product by
rail or truck. Estimates of gasoline throughput for terminals was 413 billion
liters (109 billion gallons) in 1977. Throughput is expected to increase until
1980, when there will be a slow decline in gasoline sales due to federally re-
o
quired increases in fuel economy. (For more detail of current industry sta-
tistics, see Chapter 6, "Economic Impact." The chapter estimates current bulk
terminals at 1511 and bulk plants at 17850.)
While throughput and storage capacities of terminals are subject to
considerable variability, a model existing terminal can be specified as
having 950,000 liters (250,000 gallon) per day throughput; three floating
roof gasoline storage tanks of 8.74 million liters (55,000 barrels) capacity
each; and two loading racks with three top loading, submerged fill arms per
rack. There is a trend toward bottom fill in the industry today.9 Figure 2-3
depicts a simplified schematic of bottom loading at bulk gasoline terminals.
While benzene is emitted from both loading operations and storage at
the terminal, the major source - loading operations - will be discussed in
this document. As noted above, storage tanks at terminals are generally
equipped with floating roofs. Storage tank losses of benzene are relatively
small (estimated to be about 20 kg/yr) compared to tank truck loading losses
(about 1300 kg/yr for the typical terminal).
Gasoline is pumped from the large above ground storage tanks at a rate
of 1500-2300 liters (400-600 gallons) per minute.10 (See Figure 2-4).
2-6
-------
FIGURE 2-3. SCHEMATIC OF BOTTOM-LOADING TANK TRUCK TERMINAL
Tank Trucks-
ro
I
Floating Roof
Storage Tanks
Vapor Line
Gasoline From
Pipeline, Ship
or Barge
Gasoline
Pumps
To Vapor
Control
or
Atmosphere
Bottom
Loading
Lines
-------
VAPOW EMI34>Ortt
V
Gasoline /
vapors
PIPE
—• - -MATCH
Gasoline
VAPORS
' Tank truck compartment
~3rl_ -.i. -:.'•._ PRODUCT =- -^-~~^_: - --.'- -\^=:^:.
1 . SPLASH LOADING METHOD
VAPOR EMISSIONS ^~^ x ~ FiUL !
" .-S^"-* MATCH COVER
O)
c
T—
O
VAPOSS ^ 10
Tank truck compartment
«t—*—=*r
Case 2. SUBMERGED FILL PIPE
\
VAPOR VENT
TO RECOVERY
OR ATMOSPHERE
HATCH CLOSED
A
VAPORS
Tank truck compartment
_'.Sf.~ ' ~ "~ "" ' •——— •• —
-Q.^'--~ PRODUCT ':l"-:"1i---
Case 3. BOTTOM LOADING
Gasoline
i FILL
Figure 2-4. Gasoline Tank Truck Loading Methods
2-8
-------
Gasoline is transferred through a 10 centimeter (4 inch) pipe to the top of
the truck. The truck contains 4 or 5 compartments, each having an access
hatch atop the truck. Gasoline is loaded through these hatches in pipes
(loading arms) which are extended to within 15 centimeters (6 inches)
of the bottom of the compartment. Assuming each tank truck compartment
has 5700-7500 liters (1500-2000 gallons) capacity and the pump rate averages
1900 liters (500 gallons) per minute, it takes 3 to 5 minutes to fill each
compartment after the liquid hose is lowered through the hatch into the
compartment. A measured amount of gasoline is loaded into the compartment
through a preset meter. A liquid level sensor in each compartment is
electrically connected to the pump and shuts the pump off should the
compartment be overfilled. As an example, a set of 3-5 loading arms,
3-5 pumps, and attendant piping may be collectively known as the loading
rack.
As the gasoline is loaded, vapors present in the tank truck are
displaced to atmosphere through the hatches. In the typical case of top
submerged filling, turbulence in the compartments is minimal. The
turbulence of the splash fill operation causes entrainment of gasoline
mist and droplets in the vaoor soace which are subsequently emitted
to the atmosphere through the hatches. Economics dictates that submerged
fill be installed at terminals that are now equipped with splash fill.
The trend in the industry is to convert trucks to bottom fill. In
bottom fill, gasoline is loaded through 10 centimeter (4 inch) diameter
couplings on the bottom of each compartment. Hatches remain closed during
filling of a truck so equipped. The tank truck compartments are manifolded
to a common vent pipe which directs displaced hydrocarbon vapors to atmosphere.
An estimated 25 percent of all terminals now have bottom fill.
2-9
-------
Emission factors for these three configurations are shown in Table 2-1.
Bottom fill and top submerged fill share the same emission factor (4.8 mg of
benzene per liter of gasoline loaded), since turbulence is minimal in both.
Splash fill has a higher emission factor because of the entrainment of
droplets of gasoline (11.2 mg of benzene per liter of gasoline loaded).
The term "balance service" in Table 2-1 refers to the situation in
which transport trucks return to the terminal with the vapor space nearly
saturated with hydrocarbons from "balanced" bulk plants or service stations.
In effect, the transport truck has exchanged the liquid gasoline for the
vapors displaced by filling the gasoline storage tanks at the service station
or bulk plant. The benzene emission factor for both splash and submerged
1 ?
loading in "balance service" is 8 mg/liter.
2.3 BULK GASOLINE PLANTS
Bulk gasoline plants are intermediate distributors which receive product
primarily by truck. Commonly the bulk plant will have a daily throughput of
15,000 liters (4000 gallons) and will have three above ground fixed roof
storage tanks of 38,000-76,000 liter (10,000-20,000 gallon) capacity each,
one unloading-loading rack with three overhead arms, and two delivery trucks.
In 1972 there were 23,367 bulk plants in the United States.14 (Current
estimates run closer to 18,000.) Gasoline throughput (bulk plants handle
other distillates and often agricultural supplies), was estimated to be
165 million liters (44 billion gallons) per year in 1977. The number
of plants is declining due to a trend toward the use of terminals
as opposed to plants for distribution. There is an economic
2-10
-------
TABLE 2-1. NATIONAL BENZENE EMISSIONS FROM THE GASOLINE MARKETING INDUSTRY
SOURCE
Bulk Terminal
Loading Trucks
Top Submerged
Bottom Fill
Splash Fill
Bulk Plants Storage
Loading
Service Station Underground
Storage Tank
Filling !/
Breathing — '
Emptying
Hydrocarbon
mg/1
600 17 (1000)27
600 (1000)
1400 (1000)
600 - Breathing
460 - Emptying
1150 - Filling
1400 - Splash
880 Submerged
1380 Splash
60
60
Benzene
mg/1
4.8 (8)
4.8 (8)
11.2 (8)
4.8
3.7
9.2
11.2
7.0
11.0
0.5
0.5
Throughput
liters/yr
(1 gallon = 3.8 liters)
413 X 109 2J
165 x 109 -1
165 x 109
413 x 109
National BZ Emissions
Metric Tons
per/yr - U.S.
1980
792
607
1518
1848
3734
207
207
no
i
TOTAL
10,893
I/ Model facility.
2/ Parentheses denote trucks are in balance service at stations. This factor has been rounded 9ff from 960 mg/1.
T/ Does not account for an undetermined amount of gasoline delivered to small farms. The quantity is expected to
be small.
4_/ About 50 percent of all stations have submerged fill, 50 percent splash.
$/ Breathing and emptying losses are generally estimated together for service stations. For this table it was
~ assumed the two are equal. In reality, breathing losses would likely be much lower than emptying losses.
-------
advantage in delivering gasoline directly to the service station where possible.
(For more detail of current industry statistics, see Chapter 6.0, Section 6.2.1,
"Bulk Plant Industry Characterization.")
There are two major source areas in bulk plants - storage tanks and
loading of delivery trucks at loading racks. Unlike bulk terminals, storage
tank losses are significant at the bulk plant. Figure 2-5 depicts the bulk
plant and its emission sources.
2.3.1 Gasoline Storage
Gasoline is stored in 38,000-76,000 liter (10,000-20,000 gallon)
capacity tanks at the bulk plant. The tanks are generally located above
ground and are loaded by pumping gasoline from large transport trucks to the
bottom of the 8 meter (26 foot) high storage tank. Ordinarily, a single pump
serves for both loading and unloading, but separate pumps are
provided for different tanks, especially where different grades of gasoline
are stored. Atop each tank is a pressure-vacuum relief valve which vents to
atmosphere when the pressure exceeds a preset limit (usually 2600 pascals
p
or 6 oz/in pressure).
Benzene can be emitted with other hydrocarbons during loading and
unloading of the tank (working losses) or during normal expansion of vapors
due to temperature changes during the day (breathing losses). Table 2-1
shows these emission rates.
Working losses occur during the filling and the emptying of liquid
in the tank. As gasoline liquid is pumped into the tank, vapor is displaced
to atmosphere (filling loss). As gasoline is pumped out, fresh air is brought
into the tank by the vacuum of diminishing liquid volume. This fresh air
gradually becomes saturated with vapors, expands, and a vapor laden portion
of the volume is emitted (emptying loss). Working losses account for about
2-12
-------
Unloading - Loading Rack
Storage Tanks
Figure 2-5 Typical Bulk Plant
(delivery truck splash loading)
-------
13 mg of benzene per liter of gasoline pumped. (3.7 mg/1 for emptying
loss and 9.2 mg/1 for filling loss.)
Breathing losses occur due to temperature changes during the day.
These diurnal fluctuations cause the vapor volume in the vapor space to
expand and contract. As it expands, a portion is vented to atmosphere
since the tank has a fixed volume. As it contracts, fresh air is brought
in and saturated. The vapor space is expanded and vented - in the same
fashion of unloading working losses. Breathing losses are affected by a
number of factors including ambient temperature and color and condition
of storage tanks. While breathing loss emission rates are difficult to
typify, this document uses 4.8 mg of benzene per liter of liquid pumped
to define breathing losses for a typical bulk plant having three storage
tanks. 17
2.3.2 Loading of Delivery Trucks
Deliveries of gasoline from the bulk plant are made in small
5700-11,000 liter (1500-3000 gallon) capacity tank trucks. These trucks
are generally loaded via the hatches by top splash fill at a pump rate
of 380-760 liters (100-200 gallons) per minute.18 (Top splash fill is
accomplished through open hatches atop the truck tank.) Clients of bulk
plants include agricultural interests, remote service stations, and
service stations in areas inaccessible to large trucks.
Loading losses are given in Table 2-1. Delivery trucks emit 11.2 mg
of benzene per liter of gasoline pumped during loading by splash fill.
2.4 SERVICE STATIONS
)
Service stations, as defined in this document, include all motor
vehicle refueling operations. This includes retail outlets, which
2-14
-------
1Q
numbered 178,000 in 1977. A retail outlet receives more than 50 percent
of revenue from sales of gasoline. The definition of service station also
includes the non-retail and miscellaneous outlets which numbered 243,000 in
20
1977. Non-retail stations include governmental, commercial, or industrial
fleet operations (e.g. the U. S. Post Office, rental car agencies, etc.)
Miscellaneous stations include large agricultural accounts, marinas, parking
garages and others which obtain less than 50 percent of revenue from gasoline
sales. The estimate does not include an estimated 2.7 million small farm
accounts.
Total national throughput in 1977 was 413 X 109 liters (109 X 109
pi
gallons) at service stations. Retail outlets pumped 77 percent of this or
9 9
318 X 10 liters (84 X 10 gallons). A typical retail outlet has a throughput
of about 150,000 liters (40,000 gallons) of gasoline per month. It has six
to nine nozzles for refueling (about half of retail stations are full service
and half have some self service). There are three underground storage tanks of
38,000 liter (10,000 gallon) capacity each.
Non-retail and miscellaneous outlets pump about 23 percent of total
gasoline consumed in the United States. Their throughput is generally less
than 38,000 liters (10,000 gallons) per month, per facility. 22
Emissions can occur from two major sources in service stations - the
loading of storage tanks (underground) and the refueling of motor vehicles.
This document will deal with the former only. Figure 2-6 illustrates the
loading of storage tanks at service stations. (Minor sources include breathing
and emptying losses from underground storage tanks and spillage.)
The loading of underground storage tanks is accomplished by gravity. The
tanks are coupled to the delivery tank truck by flexible ten centimeter (four
inch) diameter hoses. On-truck valves are opened and the liquid gasoline is dropped
2-15
-------
Vapor Vent Line
Submerged Fill Pipe
FIGURE 2-6. TANK TRUCK UNLOADING INTO AN UNDERGROUND SERVICE STATION
STORAGE TANK
2-16
-------
into the storage tank. Displaced vapors (filling losses) are vented to
atmosphere via a vent pipe usually located at the rear of the station.
Unloading losses (emptying losses) are generated as discussed in Section 2.3.
These emptying losses are generally very small. Storage tank losses are
shown in Table 2-1. The table shows the difference between splash loading
the tank and drop or submerged filling. About 50 percent of stations are
currently equipped with submerged fill and the other half have splash fill.
A typical emission factor falls in between, therefore, and is 9 mg of
23
benzene per liter of gasoline pumped.
Because a great majority of service station tanks are underground
(in compliance with safety regulations) diurnal temperature changes have
little effect on emissions. Breathing losses do occur, however, and as
Table 2-1 shows, these losses summed with emptying losses are 1.0 mg of
24
benzene per liter of gasoline pumped. Because control technology such
as vacuum assist and balance systems used for controlling emissions from the
refueling of automobiles also controls emptying and breathing losses, this
document shall omit discussion of these two relatively small sources.
Breathing and emptying losses will be discussed in an upcoming study
of benzene emissions from automobile refueling.
2.5 GASOLINE TANK TRUCKS
Losses from trucks can occur during loading and in transit. Loading
losses occur because of vapor displacement and are described below.
Transit losses are due to vapor breathing or vapor leaks during transit.
There is currently only a limited amount of data on the significance of
transit losses from trucks.
2-17
-------
2.5.1 Tank Truck Description
There are two basic types of tank trucks used for gasoline delivery;
tractor-semi tank trailers and straight tank trucks. Tractor-semi tank
trailers range in total capacity from 30,000 to 36,000 liters (8,000-9,500
gallons) with one to six compartments for different grades of gasoline or
other products. Straight tank trucks are smaller with a total capacity
of 5,700 to 11,000 liters (1500-3000 gallons) and one to six compartments.
Each type of tank truck may pull a full trailer in some states of equal
or less total storage capacity. Each compartment has a hatch opening, dome
cover, pressure-vacuum relief valves and vents. Because tank trucks usually
•
vary only in size and shape, no distinction will be made between the two
when discussing the emission sources.
The hatch opening on top of the truck tank is for access in cleaning
and maintaining the tank. A dome cover or lid is used to seal the hatch
opening during transport and loading-unloading operations. The hatch lid
also serves as a pressure relief valve. If extreme pressure or vacuum is
built up the hatch lid will lift (normally at 20,000 pascals or 3 psi) to
relieve this pressure.
The pressure-vacuum (P-V) valve is completely open at 6900 pascals--
(1 psig) pressure or 2600 pascals (6 ounces) vacuum (required by Department of
Transportation) for normal .venting duruig-loading^unloading operations and during
transfer operations. These P-V valves are normally a spring loaded type valve.
An emergency vent (high capacity vent) is another major relief system
for the compartment. These vents are mechanically or air actuated
when bottom loading or unloading the tank to relieve pressure or vacuum
during the loading-unloading operation. On vapor collection tank trucks,
the emergency vent is encased in metal or rubber (hoods) and the vapors are
2-18
-------
vented through piping to the bottom side of the tank. When top loading
without vapor recovery, these emergency vents are not normally used because the
hatches are open and vapor escapes around the loading arm.
2.5.2 Sources of Emissions
The major emission sources on tank trucks are the hatch covers, P-V
vents, valves and power vents. Losses from truck tanks occur during loading
and in transit. Loading losses occur because of vapor displacement and are
described in this chapter. Transit losses are due to vapor breathing and
vapor leaks during transit. During EPA testing of five terminals the average
tank truck leakage was found to range from 46 to 155 mg of hydrocarbons
per liter loaded """ (or 0.37 to 1.2 mg of benzene per liter loaded).
Hatch Covers
A dome or hatch cover is used to seal the hatch opening during transport
and bottom loading-unloading operations. The seal around the dome cover
and around the base ring where the cover attaches to the tank shell are the
most likely locations for leaks to occur when the dome cover is closed.
During top loading operations (without vapor collection) the hatch cover is open,
therefore these leaks occur only during transit. These leaks can be caused
by cracked or worn seals, warped or damaged hatch covers, and cracked or
improperly installed dome cover base rings.
P-V Relief Valves
Leaks can also occur at the P-V valves when the dome covers are closed
during bottom loading, unloading, and transfer. Emissions occur when the set
pressure or vacuum is exceeded, or when they are not properly maintained.
The valve seat may become dirty or damaged which would not allow the valve to
seal properly. The valve actuating device, such as a spring on a spring
2-19
-------
loaded valve, may become damaged also allowing improper sealing and cause
leakage. Also, many of the P-V vents partially open before the set pressure
or vacuum is reached and fully open at the set pressure or vacuum.
Emergency Vents
Emergency vents may leak when closed if the vent is not installed properly
or is not maintained properly. The vent seal may become dirty or damaged
which would not allow the valve to seal properly. In cases where the vent
is encased for vapor collection, seals, hoods, and rubber hoses may become
cracked or loose which would allow vapor leakage.
Miscellaneous Sources
Other emission sources may occur from various locations around the
tank truck. Improperly installed or damaged hose couplers can be emission
sources. The tank shell, if damaged, can produce emission sources from
cracks or failures in welds or failure of tank shell itself. These types
of leaks occur less frequently than those discussed previously, but may
be large emission sources on some truck tanks.
2-20
-------
2/6 REFERENCES
1. Department of Energy, "Monthly Petroleum Statement,"
December, 1977. (Suggested by Ed Crockett, API.)
2. Runion, Howard E., "Benzene in Gasoline," American Industrial
Hygiene Association Journal, May, 1975.
3. McDermott, H. 0., and Killiany, S. E., "Quest for a Gasoline TLV,"
AIHA Journal, February, 1978.
4. Analyses of Vapor Samples from Gasoline Storage Tanks, Colonial
Pipeline Company, Greensboro, North Carolina. Scott Environmental
Technology, November, 1977.
5. "Cost of Benzene Reduction in Gasoline to the Petroleum Refining
Industry," EPA 450/3-78-021. A. D. Little, Inc., Cambridge, Mass, for
EPA, April, 1978.
6. U. S. Department of Commerce, "1972 Census of Wholesale Trade-
Petroleum Bulk Stations and Terminals," October, 1975.
7. Arthur D. Little, Inc., "The Economic Impact of Vapor Recovery
Regulations on the Service Station Industry." Report to OSHA, March, 1978.
8. Reference 5, Op. Cit.
9. National Petroleum News Fact Book Issue, "The Move Toward
Bottom Loading," mid-May, 1975.
10. Kleeberg, Charles F., "Trip Report—Bulk Terminal Control
Techniques in Denver, Colorado," memo to James F. Durham, EPA,
September 13, 1977.
11. Reference 9, Op. Cit.
2-21
-------
12. Radian Corporation, "Revision of Evaporative Hydrocarbon Emission
Factors," EPA-450/3-76-039, August, 1976.
13. Pacific Environmental Services, Inc., "Study of Gasoline Vapor
Emission Controls at Small Bulk Plants," Final Report to EPA, Region VIII,
October, 1976.
14. Reference 6, Op. Cit.
15. Reference 1, Op. Cit.
16. "Supplement No. 7 for Compilation of Air Pollutant Emission
Factors, Second Edition," U.S. EPA, April, 1977.
17. Reference 13, Op. Cit.
18. Reference 13, Op. Cit.
19. Reference 7, Op. Cit.
20. Reference 7, Op. Cit.
21. Reference 7, Op. Cit.
22. Reference 7, Op. Cit.
23. Reference 16, Op. Cit.
24. Reference 16, Op. Cit.
25. "Control of Hydrocarbons From Tank Truck Gasoline Loading
Terminals," EPA-450/2-77-026, October, 1977, Table 3-2, page 3-5.
2-22
-------
3.0 EMISSION CONTROL TECHNOLOGY
The purpose of this chapter is to describe available control
techniques which can be used to reduce benzene emissions from the gasoline
marketing network.
3.1 USE OF CONTROL METHODS
With the exception of reduction of benzene in liquid gasoline at
the refinery, all control techniques discussed in this chapter have been
applied to hydrocarbon sources in bulk terminals, bulk plants, or service
stations. The techniques have been applied to comply with air pollution
regulations designed to minimize hydrocarbon emissions in certain Air
Quality Control Regions, not because of economic incentives. The source
test data developed to support control of hydrocarbon emissions can be used
to support control of benzene emissions also. As discussed in Chapter 2,
empirical correlations were developed to derive benzene emission factors
from hydrocarbon emission factors. Therefore, data derived from control of
hydrocarbon emissions are used,in part, as the basis for control of benzene
emissions in this study. In addition, EPA has collected data on the effect of
controls on benzene specifically. Chapter 2 indicates that for the conditions
under which equilibrium data were derived (27°C, 1.3 liquid volume percent
benzene in gasoline), 0.008 grams of benzene are emitted with every gram of
hydrocarbon emitted. This provides an emission factor for benzene in gasoline.
3-1
-------
Table 3-1 summarizes hydrocarbon emission data gathered by source
tests of various emission control techniques employed at bulk terminals,
bulk plants, and service stations. Estimated and measured benzene emissions
are included in the table.
3.2 BULK TERMINALS
About 300 vapor control systems have been installed and are in
commercial operation at tank truck gasoline loading terminals. Stage I
service station controls (balance systems between underground storage tanks
and tank trucks) have provided impetus for such installations in Air Quality
Control Regions with oxidant problems, since the vapor in trucks must be
controlled.
The benzene content of gasoline vapors vented to vapor control
systems source tested by EPA at tank truck loading terminals are approximately
4.8 mg/1 of gasoline loaded. It should be noted that many trucks in these
tests leaked and many were "lean" (only partially saturated), which are both
conditions which affect processor efficiency. Benzene test data indicate
outlet emissions are in the range of 0.003 to 0.33 mg/1 of gasoline loaded.
Table 3-1 (tests A through F) summarizes actual EPA hydrocarbon test data
(including total hydrocarbon and estimated benzene mass rates in grams per
liter of gasoline transferred). Tests G through K summarize actual EPA test
data for benzene at terminals.
A brief process description of the types of vapor control systems
installed at gasoline tank truck loading terminals and source tested by
EPA follows.
3-2
-------
TABLE 3-1. SUMMARY OF BENZENE EMISSION TESTS
SOURCE
Bulk Terminals
Truck Loading
See References
1-11)
Bulk Plants
Storage Tanks
Delivery Trucks
Service Stations
Filling Storage
Tanks 3/
Test
A
B
C
D
E
F
G
H
I
J
K
A
B
A
B
A
Date
12/10-12/74
12/16-19/76
9/20-22/76
9/23-25/76
11/18/73-5/2/74
11/10-12/76
5/25-27/77
12/16/77
1/10-12/78
3/7/78
3/l-b//8
7/76
8/76
7/76
8/76
6/12/74
B | 6/18/74
Control
Device
CRA
RF
RF
CRA
TO
RF
AA
CRA
TO
RF
CRA
Balance
Balance
Balance
Balance
Balance
Balance
Size of
Facility
600,000 I/day
380,000
1,430,000
1,190,000
1,100,000
810,000
284,000
600,000
1,000,000
810,000
1,000,000
64,000 I/day
13,000 I/day
64,000 I/day
13,000 I/day
^150,000 liters/mo
^ 75,000 liters/mo
Processor
Outlet
Total HC
mg/l
31.2
37
33.6
43.3
1.3
62.6
30
41.1
34.2
53.4
Processor
Outlet
Total Benzene
mg/1
N/A
N/A
N/A
N/A
N/A
iVA
.003 2/
.106 U
.330 i/
.052 !/
. D8U 2j
8.5
46
81
75
7.9
10.6
0.07 -'
0.37 I/
0.65V
0.60 I/
0.06 V
0.08 I/
CO
CO
V Estimated from hycrocarbon test data.
2/ Test data - preliminary results
CRA - Compression-Refrigeration-Absorption
RF - Refrigeration
TO - Thermal Oxidizer
AA - Adsorption-Absorption
3/ Other systems which incorporated treatment of auto refill losses in addition to storage tank filling were tested
by EPA. The systems had higher efficiency than the two tests shown here. (See Appendix C)
-------
3.2.1 Refrigeration Systems (RF)
The principle of the straight refrigeration system (RF) is based
on the condensation of gasoline vapors by refrigeration at atmospheric
pressure. It is estimated that 90 units of this type are in commercial
1 ?
operation. Vapors displaced from the trucks enter a double pass fin-tube
condenser where they are cooled to a temperature of about -73°C and condensed.
The remaining air containing 3 to 5 percent hydrocarbon is vented to the
atmosphere. Because vapors are treated as they are vented from the tank
trucks, no vapor holder is required. Condensed gasoline is withdrawn from
the condenser and separated from condensed water. Hydrocarbon condensate
is returned to premium gasoline storage tanks and water typically passes
to a slop tank or oil-water separator. A simplified schematic of a recent
model of this type of vapor recovery system is shown in Figure 3-1. A source
test for benzene was conducted on a refrigeration unit (Test J). Outlet
emissions of benzene from the unit averaged 0.052 mg/1. Inlet vapors to
I O
the unit contained an average of 0.99 mg of benzene per liter of gasoline.
3.2.2 Compress I'on-Refrigeration-Absorpti on Systems (CRA)
The compression-refrigeration-absorption vapor recovery system (CRA)
is based on the absorption of gasoline vapors under pressure with chilled
gasoline from storage. Incoming vapors are first passed through a saturator
where they are sprayed with fuel to ensure that the hydrocarbon concentration
is above the explosive level. This is done as a safety measure to reduce
the hazards of compressing hydrocarbon vapors.
3-4
-------
£.
CONDENSER
AIR
(PRECOOLER)
/ y
, \, \
PRECOOLER
REFRIGERATION
UNIT
SCHEMATIC DIAGRAM
CONDENSER
AIR
A L
LOW TEMPERATURE
REFRIGERATION
SYSTEM
PRECOOLER
COIL
DISCHARGE FROM UNIT
CONDENSED
GASOLINE
WATER
COOLING
-RECOVERY
SECTION
VAPOR
CONDENSER
SECTION
FIGURE 3-1. REFRIGERATION VAPOR RECOVERY UNIT
by Edwards Engineering Corporation
3-5
-------
The partially saturated vapors are then compressed and cooled prior
to entering the absorber. In the absorber, the cooled, compressed vapors
are contacted by chilled gasoline drawn from product storage and are absorbed.
The remaining air containing only a small amount of hydrocarbons is vented
from the top of the absorber and gasoline enriched with light ends is
withdrawn from the bottom of the absorber and returned to the gasoline
storage tanks. A schematic of a typical system is shown in Figure 3-2.
One CRA unit test by EPA at a tank truck loading facility averaged
benzene outlet emissions of .106 mg/1. The benzene content of the inlet
vapors to the unit was approximately 2.45 mg/1.14 (Test H on Table 3-1.)
3.2.3 Adsorption-Absorption (AA)
A recently developed vapor recovery system is carbon bed adsorption-
absorption (AA). This type of system commonly consists of two vertically
positioned carbon beds and a vacuum regeneration system. During normal
tank truck gasoline loading operations,one carbon bed is in the adsorbing
mode and the other carbon bed is in the regeneration mode.
Hydrocarbon vapors collected during the adsorbing mode are stripped
from the carbon bed by vacuum during the regeneration cycle. The vapors
pass through a gasoline condensing bath which is returned to the supply
tanks as liquid gasoline. Water is removed in a separator. The air and
any remaining hydrocarbons exiting from the condensing bath are then
passed through an absorber utilizing gasoline as the absorbent and exhausted
3-6
-------
Vapor Line
CO
..._ rzr
Tank Truck
D-
•D
Saturator
Vapor Holder
From Gasoline Storage
Compressor-Refrigerator
To Gasoline
Storage
Tank Gage
and Switch
To Compressor
Starter
FIGURE 3-2. SCHEMATIC OF COMPRESSOR-REFRIGERATION-ABSORPTION SYSTEM (CRA)
-------
to atmosphere. Thus, during regeneration even when no trucks are being
loaded with gasoline some hydrocarbon vapors are \Jpnted from the control
«
equipment. A schematic of a typical unit is shown in Figure 3-3.
During a source test of an adsorber-absorber (Test G on Table 3-1),
benzene emissions at the outlet of the vapor control equipment averaged
•
.003 mg/1. Inlet vapors to the unit contained an average of 2.5 mg/1 of
benzene. All tests were performed for a relatively short period of time
on a new carbon bed. No data are available on the bed life of the adsorber.
Insufficient data are available to determine if bed life is affected by vacuum
desorption of the carbon. During desorption heavier compounds cling to the
carbon creating a "heel" which eventually builds up on the bed, lowering
working capacity. There has been insufficient experience with the design
to determine how fast the heel builds up. Other modes of regeneration have
not been evaluated for this application.
3.2.4 Oxidation Systems
Table 3-1 indicates that there is not a significant difference
between oxidation and vapor recovery in terms of benzene control efficiency.
Gasoline vapors from loading operations at one terminal were displaced
to a vapor holder as they were generated. The vapors were kept above the
upper explosive limit in the vapor holder by injecting propane. When the
vapor holder reached its capacity, the gasoline vapors were released to
the oxidizer after mixing with a properly metered air stream and there
the vapors were combusted. The thermal oxidizer is not a true incinerator,
rather it operates in the manner of an enclosed flare. A simplified
schematic of the system is shown in Figure 3-4.
Twelve to fifteen oxidizers have reportedly been installed by terminal
operators. Later models of this type of control equipment are not equipped
with vapor holders; vapors from the tank trucks during loading operations are
vented directly to the thermal oxidizer. In a recent EPA test of this type
3-8
-------
co
INLET
VAPOR
AIR
VENT
"ARRESTCR
CARBON
ADSORPTION
BEDS
AIR RECYCLE
_L
SEPARATOR
LIQUID RING
VACUUM PUMP
COOLER
FIGURE 3-3. SCHEMATIC OF ADSORPTION-ABSORPTION VAPOR RECOVERY
SYSTEM
.GASOLINE
SUPPLY
PUMP
GASOLINE
RETUHN
PUMP
-------
Vapor
Line
co
o
Q
-x-
Vapor
A
i
_.
Stack
Burner
Air
/- Pilot
^r Line
Propane
Tank
FIGURE 3-4. SCHEMATIC OF THERMAL OXIDATION SYSTEM (TO)
Note: Dotted lines represent optional equipment.
-------
of unit (Test I on Table 3-1), benzene average outlet emissions of
.330 mg/1 were indicated. Inlet vapors to the unit contained approximately
1.68 mg/1. The system was tested during very cold conditions. Very small
amounts of hydrocarbon were vented to the oxidizer (most being condensed in
the truck). Consequently, the system did not operate as efficiently as
expected. This problem can be remedied with a vapor holder.
Environmental Protection Agency hydrocarbon and benzene source
tests for compression-refrigeration-absorption, refrigeration, thermal
oxidation, and adsorption-absorption are summarized in Appendix C.
3.3 BULK GASOLINE PLANTS
Control of gasoline working losses resulting from storage and handling
of gasoline at bulk plants can be accomplished through submerged fill and
balance systems. While vapor processing systems as discussed above for
terminals have not been applied to bulk plants, they could be used to
control both breathing and working losses from plant sources.
3.3.1 Submerged^Fill
One method for controlling emissions at bulk plants is to reduce the
vapors generated during filling of tank trucks and storage tanks by using
submerged fill. The reason for this reduction is that submerged fill
decreases turbulence and evaporation and eliminates liquid entrainment.
(Bulk plant storage is typically equipped with submerged fill.) Submerged
loading can be accomplished with a top submerged fill pipe or bottom filling.
In the top submerged fill pipe method, the fill pipe descends through an
open hatch to within 15 centimeters (6 inches) of the bottom of the
compartment. In the bottom filling method, the fixed fill pipe is
attached to the tank truck at the bottom of each compartment (on the side of
the tank). Changing from splash to submerged loading, benzene vapors generated
18
by filling of tank trucks can be reduced from 11.2 to 4.8 mg/liter transferred.
3-11
-------
The following discussion and figures describe three top-submerged
fill systems and two bottom loading systems presently being used at gasoline
bulk plants to load gasoline tank trucks.
Submerged drop tubes are the simplest type of top-submerged fill
system used to reduce generated emissions, but they do not collect vapors,
since the hatch remains open during loading operations. Figure 3-5 shows
a typical system. To convert the existing top-splash fill arm to a sub-
merged drop tube requires attaching a straight section of pipe or a
telescoping pipe onto the top-splash nozzle. The length of pipe required
is determined by measuring the distance from the top-splash nozzle to within
15 cm of the bottom of the truck tank. In order to properly align and
maneuver the drop tube into the open truck hatch, it may be necessary to
install extra swivel joints on the loading arm. No conversion of the tank
truck is needed.
The second type of top-submerged fill system is a dry break drop tube
system. Figure 3-5 shows a simple schematic of this system. Principal
features of the system include (1) minimal modifications to the existing loading
rack; (2) the use of dry break, quick-connect connections between the top
loading arm and new fill ports on the truck; (3) the use of a single vapor
return line which connects to the compartment vapor hoods on the truck; and
(4) the discontinued use of filling through existing truck hatches. The system
requires some modification to the truck and requires meter pumps or some other
overfill protection system.
Top loading vapor heads are a third type of top-submerged fill system.
Figure 3-6 shows a simple schematic of this system. Top vapor head arms
consist of a splash or submerged loading nozzle fitted with a head which
3-12
-------
TOP-SPLASH FILL
No vapor collection
TOP-SUBMERGED FILL
No vapor collection
Top Loading Arm
Flexible Hose—,
Compatible Dry Breaks
Permanent
Submerged
Drop-tube
Vapor
Line
Vapor Dry Break
DRY BREAK DROP TUBE SYSTEM
FIGURE 3-5. TOP-LOADING SYSTEMS AT BULK PLANTS
3-13
-------
seals tightly against the hatch opening. Liquid flow is possible as long
as a positive seal is maintained (pneumatically or mechanically) between
the vapor recovery head and the hatch opening. Liquid is loaded through
a central channel in the nozzle and the displaced vapors flow into an
annular vapor space surrounding the central channel. The vapors flow into
a hose on the loading arm. Since the vapor line is incapable of handling
liquid overflows and the liquid level is no longer visible through the open
hatch, a safety shutoff valve is included in the nozzle. Truck conversion
is not necessary for the loading operation at the plant. The principal
limitation to the use of this vapor recovery head at any existing top loading
r*ack arm is its weight. The existing loading arm and rack supports must be
modified to hold the vapor head. With a few types of vapor heads there must
be a supply of air pressure to operate the heavy loading arm.
Bottom loading is a ground-level facility, as opposed to the elevated
platform used for top loading. Here the truck is filled through adapters
at the bottom of the tank. Figure 3-6 shows a simple schematic of this
system. There are two major types of bottom load systems used at bulk
plants, the normal type used at bulk terminals and the Wiggins system adapted
specifically for use at bulk plants. Both types of bottom loading systems
use the same principles of operation. Both types of bottom fill systems have
several variations but a basic bottom loading system consists of; (1) an
adapter, the device which permits coupling of the loading rack liquid hose
to the tank truck piping; (2) liquid level sensors which prevent overfilling
by shutting down the rack pumps or closing the internal valve system; and
(3) a vapor collection system which collects vapors from the compartments and
routes them through a common vapor manifold that terminates at a dry break on
the side of the truck.
3-14
-------
Vapor Line
Hatch Cover
Liquid Line
Lift Cylinder
Truck Tank Shell
Drop Tube
TOP LOADING VAPOR HEAD SYSTEM
SENSOR
VAPOR RECOVERY HOOD
ISLAND WIRING
X
\
\
Bottom Loading
FIGURE 3-6. TOP AND BOTTOM LOADING SYSTEMS
3-15
-------
An overfill protection system is needed for loading of tank trucks
when the hatches are closed. The four basic types of overfill protection
systems are preset meters, meters, liquid level sensing devices, and
float rods.
Most vapor controlled facilities use a preset meter on the loading
rack to provide primary overfill control. The driver selects the amount of
product to be loaded and when the preset volume has passed through the meter,
the pump is automatically shut down. Meters without preset equipment are
also used. The driver simply loads the desired amount, and shuts off the
pump manually.
Liquid level sensing devices are commonly used with preset meters to
provide a secondary control system in the event of a meter failure or in-
correct meter setting. Liquid level sensing devices can also be used as
the primary overfill protection system. There are two basic types of
sensor systems commonly used for bottom loading. The most common type in
use today is an electrical system in which the tank level sensor sends an
electrical signal to the loading rack to shut down the pump when the tank
is full. Figure 3-6 (bottom loading) shows a simple schematic of this system.
The other type of system is completely self-contained on the truck and closes
the tank inlet valve when the sensor determines that the tank is full. The
loading rack pump is then shut off manually.
Floats and level rods are being used at bulk plants with the dry
break drop tube system discussed earlier. As the liquid level reaches the
float, the graduated rod rises and the liquid level is visually determined.
Rubber o-rings are installed to seal around the rod to eliminate the
escape of vapors. When not being used, a cap is placed over the fitting.
If the rod is in the full position, the rod is simply pushed down and the
cap installed.
3-16
-------
3.3.2 Balance System
The displacement, or vapor balance system, operates by transferring
vapors displaced from the receiving container to the container being
unloaded. A vapor line between the truck and storage tanks essentially
i
creates a closed system permitting the vapor spaces of the two vessels to
balance with each other. (See Figure 3-7). Balance systems are applicable
4>to both above and below ground facilities.
Vapor balancing of incoming transport trucks displaces vapor from
storage tanks to truck tank compartments; emissions can be ultimately treated
at the terminal with secondary recovery control systems. EPA-sponsored
source tests at two bulk plants have shown that a control efficiency greater than
90 percent for hydrocarbon filling losses is attainable with vapor balancing
I Q
of incoming trucks and storage tanks. Benzene reductions would be
equivalent. (The 10 percent loss is due to a small amount of vapor growth in
the returned vapors.)
Vapor balancing of storage tanks and delivery trucks also reduces
20
account truck hydrocarbon filling losses by greater than 90 percent.
Also, balance systems on delivery truck filling virtually eliminate emptying
^losses from storage tanks, since displaced air is saturated or nearly
saturated with hydrocarbons. The efficiency attainable in loading delivery
trucks is significantly affected by tightness of the truck compartments,
i.e. condition of hatches, pressure-vacuum relief valves and seals, and
the care exercised in making line connections.
Assuming the lost vapors from the vapor balance system are ideal
gases, the benzene vapors will be emitted in proportion to the hydrocarbon
vapors. Therefore, a benzene efficiency greater than 90 percent is also
3-17
-------
Loading-Unloading Rack
Emission
t
ission
Storage Tank
Emission
Storage Tank
Storage Tank
Vapor Return Line
FIGURE 3-7. VAPOR BALANCE SYSTEM
(Bottom Fill)
-------
attainable with a vapor balance system for filling and emptying losses.
Breathing losses are not controlled by the balance system. Accounting
for breathing losses, the balance system achieves about 70 percent efficiency
for the entire plant.
The following criteria should be met to attain 90 percent or greater
efficiency for all bulk plant sources except storage breathing loss.
(1) Storage Tanks
(a) Above and below ground storage tanks should have submerged
fill in order to reduce generated emissions from the loading of storage
tanks (this is typically done on plants at present).
(b) Pressure-vacuum relief valves should be set as high as possible
and in accordance with the current National Fire Protection Association
Pamphlet No. 30, "Flammable Combustion Liquids Code."
(c) Vapor return line piping and storage tank manifold piping
should be leak tight and of sufficient size to allow efficient transfer of
vapors to the tank trucks. The vapor return piping is generally 5 to 8 cm
(2-3 inches) in diameter.
(2) Loading-unloading rack
A dry break fitting is needed on the rack end of the vapor return
•
piping. A dry break is required to prevent ground level gasoline vapor
emissions when gasoline transfer is not being made. This fitting keeps
the storage tanks sealed until the vapor hose is connected.
(3) Tank Trucks
(a) Tank trucks should be submerge filled to reduce emissions
generated during loading operations.
3-19
-------
(b) Tank Trucks must be modified to recover all vapors during
loading and unloading at the bulk plant, and to recover vapors at balanced
customer tanks (service stations).
(c) A dry break closure is required on the end of the tank truck's
vapor return line to prevent ground-level gasoline vapor emissions. These
emissions would occur as a result of failure to connect the vapor return line
to the tank truck's vapor return line.
(d) Tank truck vapor tightness - if truck hatches or relief valves
leak during balancing, they either vent the recovered vapors or draw in air.
It is necessary to ensure that trucks are vapor tight during the loading and
unloading operation in order to assure proper balancing. Many plant owners
check the liquid level in the truck compartments before and after loading to
ensure they are receiving the desired volume of gasoline. This procedure is
acceptable as long as the hatches are secured during loading and unloading.
3.3.3 Vapor Recovery and Oxidation Processing Systems
Vapor recovery (CRA, RF or AA) and oxidation systems can be used to
process all the vapors displaced from the storage tanks and the tank trucks
during loading. Such systems have been applied to bulk terminal truck
loading losses, but have not been applied in bulk plants. These systems
will yield a higher control than vapor balance systems when applied to the
loading-unloading rack and storage tanks, since breathing losses are also
controlled by "add-on" equipment. See Section 3.2 for discussion of
these systems.
3.4 SERVICE STATIONS
As explained in Chapter 2, benzene is emitted from underground
storage tanks during loading and emptying of the tank (working losses) and
during the day as temperatures fluctuate (breathing losses). This document
3-20
-------
it
discusses only those systems applicable to the control of loading losses
(breathing and emptying loss controls are deferred to another study).
In gasoline service stations, balancing has been used to control
hydrocarbon emissions from both automobiles and storage tanks (the two
major sources). The technique is equally effective in reducing benzene
emissions from these sources.
In the service station balance system, vapors are vented by dis-
placement to the transport or delivery truck which unloads gasoline. The
truck transfers the vapors to the terminal or plant for ultimate treatment
at the terminal. The system for underground storage tanks is detailed
below. (Figure 3-8 illustrates balancing at service stations.)
3.4.1 Balance System Description
Gasoline is delivered in large (30,000-36,000 liter or 8000-9500 gallon
capacity) transport trucks. The gasoline is loaded by gravity into the
underground storage tanks via a flexible hose. Liquid gasoline displaces a
nearly equal volume of partially saturated gasoline vapors. The vapor is
vented through a pipe and flexible hose connected to a vapor collection system
(simply a manifolded pipe) on the transport truck. Liquid transfer creates
a slight pressure in the storage tank and a slight vacuum in the truck
compartment. These pressure differences effectively cause the transfer of
more than 95 percent of displaced vapor to the truck. Because of a phenomenon
known as vapor growth caused by liquid temperature differences, the truck
volume cannot always accomodate all of the vapors. Any excess vapor is
released through the vapor vent line shown in Figure 3-8.
The following scenario depicts how the whole process could take place:
(1) The tank truck arrives loaded with gasoline. The station operator
has ordered about 10,000 liters of premium and regular leaded gasolines
(2 compartments of the 4 compartment truck).
3-21
-------
MANIFOLD FOR RETURNING VAPORS
VAPOR VENT LINE
TRUCKSTORAGE\
COMPARTMENTS
(Usually at rear
of station.
Four meters high
with restrictive
orifice.)
/111 t\t1111111rrrrf
v u
v\Mtmtm\
UNDERGROUND
STORAGE TANK
FIGURE 3-8. VAPOR BALANCE SYSTEM AT A SERVICE STATION
3-22
-------
(2) As the station operator opens the storage tank liquid fill
cap, the truck driver unwinds and lays out the two flexible hoses (liquid
and vapor) which he carries on the truck.
(3) The station operator dips a pole into the tank, measures
the liquid level, and calculates the amount in the tank (to ensure against
overfill). He climbs atop the truck, opens the two compartment hatches,
checks that the compartments are full and closes the hatches.
(4) The driver connects the liquid fill and vapor hoses to
his truck and then to the storage tank. He opens the valve for one compart-
ment and the gasoline flows by gravity to the underground tank.
(5) As the first "drop" is completed, the truck driver "milks"
the liquid line, then disconnects the hoses at the tank.
(6) He disconnects the liquid line at the truck and puts the
liquid hose on the second compartment. He then repeats steps 4 and 5.
(7) The station operator climbs atop the truck, opens the
hatches, and assures that all gasoline has been delivered from
the compartments. He secures the hatches and the driver leaves
(after he has disconnected his two hoses). The driver may return to the
bulk terminal/bulk plant or may proceed to another station to empty his
other compartments of gasoline. The driver may also unload more
than one compartment at a time. Manifolded storage tank vapor return lines
or multiple vapor couplings on the truck are necessary to do this.
The effectiveness of the system is adversely affected by leaks. Truck
hatches should be closed and hose connections should be tight during loading.
Tests demonstrate balance systems to be greater than 95 percent efficient for
3-23
-------
21 22
reducing underground storage tank filling losses. ' 'Note that breathing
and emptying losses are not controlled by this method. These two losses
account for about five percent of total station losses. Certain controls
for automobile refueling emissions control these two sources.
3.4.2 Necessary Criteria for the Balance System
A November, 1975, EPA report entitled, "Design Criteria for Stage I
Vapor Control Systems - Gasoline Service Stations," specified the necessary
components of the vapor balance system as applied to underground storage tanks
at service stations.
As stated in the document there are at least four objectives of detailing
equipment for the system.
(a) Assure that the vapor return line will be connected during
tank filling,
(b) assure that there are no significant leaks in the system or
tank truck which reduce vacuum in the truck or otherwise inhibit vapor
transfer,
(c) assure that the vapor return line and connectors are of
sufficient size and sufficiently free of restrictions to allow transfer of
vapor to the truck tank and achieve the desired recovery,
and (d) assure that gasoline is discharged below the gasoline
surface in the storage tanks.
All test data submitted to EPA were obtained from systems which met
these four objectives. If the balance system's efficiency is to be duplicated
on other service stations, these objectives must also be met.
The following details specific equipment necessary:
3-24
-------
1. Drop Tube - a tube which extends from the tank fill neck to below
the liquid level in the tank is necessary. This tube is called a "drop tube"
and tanks so equipped are "submerge filled." Generally, if the tube extends
within 15 centimeters (6 inches) of the tank bottom, it will be submerged in
gasoline since tanks are not pumped dry.
2. Gauge well - operators gauge the amount of liquid in their tanks
by use of a long marked pole or "dip stick." The pole is generally inserted
through the fill neck and dropped to the bottom. The liquid level is indicated
by wetting of the pole. As long as the fill pipe is submerged (see 1), this
creates no problem. Some stations are equipped with a separate gauge well.
If left uncapped during filling, vapors are displaced through this opening
rather than to the tank truck. The gauge well should be equipped with a drop
tube to prevent this.
3. Vapor hose return - typically, gasoline is gravity fed into the
storage tank from the truck by the 10 cm (4 inches) diameter drop tube at a
rate around 1500 liters (400 gallons) per minute. An 8 cm vapor return
hose (3 inches) will accomodate the volume of vapor generated by such a "drop."
4. Vapor line connections - vapor lines from two or more tanks may be
manifolded to a common vapor hose connector. This can be advantageous to
the operator who fills more than one tank at a time. A general rule is to
provide a vapor return hose cross sectional area of at least half of the
cross sectional areas of all fill hoses which displace vapors to the hose.
5. Liquid fill connection - the liquid fill connector should be equipped
with a vapor tight cap. Gaskets and similar sealing devices can ensure this
closure as can "cam-lock" and "dry break" closures.
3-25
-------
6. Tank truck vapor tightness - If truck hatches or relief valves
leak during balancing, they either vent the recovered vapors or draw in air.
It is necessary to ensure that trucks are vapor tight during the unloading
operation in order to assure proper balancing. Many station owners check
the liquid level in the truck compartments before and after loading to
ensure they are receiving the desired volume of gasoline. This procedure
is acceptable as long as the hatches are secured during loading.
7. Closures or interlocks on underground tank vapor hose connectors
and on the tank truck - (optional to ensure 95 percent)
Closures and interlocks ensure that vapor hoses are connected to the
tank truck and to the underground tank. If the vapor hose is not connected,
no gasoline can be dropped into the storage tank. Further, they ensure that
the storage tank is sealed unless the vapor hose is connected.
8. Vent line restrictions - (optional) Vent line restrictions which reduce the
the vent line diameter from about 5 centimeters to about 2 cm (2 to 0.75 inches)
assure that gasoline vapor goes to the tank truck and not through the tank vent
pipe to atmosphere. Further the restriction helps assure that the vapor
hose is properly attached. If it is not, then the flow of gasoline into the
tank is significantly slowed because of the back pressure caused by the
vent pipe restriction.
A pressure vacuum relief valve set to open at 3450 Pascals (8 oz. per
square inch) or greater pressure and 1725 Pascals (4 oz. per square inch)
or greater vacuum will accomplish the same end. Fire regulations differ in
different areas of the country and more or less stringent settings may be
required by local fire marshal Is.
3-26
-------
3.5 GASOLINE TANK TRUCKS
As explained in Chapter 2, benzene vapors are emitted from the truck
tank's hatch seals, P-V vents, and emergency vent hoods. Limited data are
available at this time to quantify typical emission rates or potential
emission reductions. Many of these leaks can be found through visual
inspections or heard during the loading operations of the tank. These
leakage points can be controlled through good maintenance procedures and
schedules. In many instances, replacement of worn or damaged parts may be
the only logical and long term method for ensuring the truck tank will stay
leak tight. EPA will have more data on control methods by the end of
September, 1978.
3.6 REDUCTION OF BENZENE CONTENT OF GASOLINE
The purpose of this section is to discuss another option for controlling
emissions from the marketing industry by reducing the level of benzene
in motor gasoline.
3.6.1 Assessment of Benzene Content of 1981 Gasoline Pool
Results from an EPA contract study conducted by A. D. Little, Inc.,
(ADL) indicate that the average U.S. gasoline pool in 1981 will contain about
23
1.37 volume percent benzene. The average was based on determining the blending
component composition of the 1981 pool and the benzene contents of each of these
components as shown in Table 3-2.
In a similar manner the average benzene content of the 1977 U.S. gasoline
pool was determined to be 1.30 volume percent. This average is in good agree-
24 25
ment with the 1.24 and 1.25 volume percent reported by NIOSH and Gulf Oil '
respectively. It is somewhat higher than the 1.0 volume percent weighted average
?fi
of the samples reported in a DuPont, June, 1977, survey.
3-27
-------
3.6.2 Control Options for Removal of Benzene from Reformates and FCC Gasolines
Reformates and fluid catalytic cracked gasolines comprise 64.5 percent
of the gasoline pool and contribute 86 percent of the pool benzene. The study
focused on these two major contributors and determined it would be feasible
to remove 94.5 percent of their benzene content (82 percent removal of benzene
from the pool) using the following selected processing routes:
Reformates
s
1. Fractionate the total (full boiling range) reformate produced
in the gasoline reformers in a new tower (deisohexanizer) to remove isohexane
and lighter in the overhead stream and the benzene and heavier paraffins
and aromatics in the bottoms. The benzene free overhead from the deiso-
hexanizer is sent to gasoline blending.
2. Fractionate the bottoms stream in a second new distillation tower
(Cg fractionator) to remove a Cg overhead stream (Cg heart cut) which would
contain 95 percent of the benzene contained in the reformer gasoline and the
other Cg paraffins. The heavier aromatics and C7 paraffins bottoms are sent
to gasoline blending.
3. The Cg heart cut which contains 15 volume percent benzene is sent
to a benzene extraction plant where 99 percent of the benzene is removed as
commercial grade benzene and the raffinate (essentially free of benzene) sent
to gasoline blending. The sulfolane process is assumed used for benzene
extraction. As shown in Figure 3-9, reduction of benzene in reformate
would lower the average total U.S. gasoline pool level by 62 percent to a
pool level of 0.52 volume percent.
3-28
-------
FIGURE 3-9.
CUMULATIVE REDUCTION IN BENZENE CONTENT OF GASOLINE
(Extracting 94.5 percent of benzene from gasoline
blending components)
Benzene Content
of Gasoline: Vol %
1.5
1.0
ro
10
0.5
1.37
0.52
62%
Reduction
0.26
82%
Reduction
Jncontrolled Control Plus
1981 of Control
Pool Reformate of FCC
Gasoline
0.165
88%
Reduction
Plus
Control
of
S.R.
Gasoline
0.13
Plus
Control
of
Natural
Gasoline
0.08
1
Plus
Control
of
All Other
Streams
-------
OJ
o
Reformate
FCC Gasoline
S.R. Naphtha
Natural Gasoline
Hydrocrackate
Coker Gasoline
Isomerate
Raffinate
Alkylate
Butane
TABLE 3-2. AVERAGE BENZENE CONTENT OF 1981
U.S. GASOLINE POOL
(Volume %)
Pool Component Compositioji
Blending Component
Thousand
Barrels/Day
2235
2571
536
186
134
89
104
104
1014
477
%
30.0
34.5
7.2
2.5
1.8
1.2
1.4
1.4
13.6
6. .4
Benzene
Vol %
3.0
0.8
1.4
1.5
1.1
1.4
0.4
0.2
0
0
BZ Contribution
to Pool
Vol %
0.90
0.28
0.10
0.04
0.02
0.02
< 0.1
< 0.1
0
0
% of Pool
Benzene
65.7
20.4
7.3
2.9
1.5
1.5
0.4
0.32
0
0
Gasoline Pool
7450
100
1.37
100
-------
FCC Gasoline
A Cg heart cut is first fractionated from the full range FCC
gasoline in a manner identical to the two fractionation steps for reformates.
Two new additional towers are required.
1. FCC gasolines contain olefins and diolefins boiling in the benzene
range. Reformates are free of olefins. It has not been commercially demon-
strated that aromatics can be extracted from Cg heart cut containing these
olefins without causing operational problems in the extraction plant.
2. This requires that the Cg heart cut be hydrogenated in a new
hydrogenation plant to saturate the olefins to paraffins. Paraffins are
much lower in octane that the olefins and the octane of the Cg heart cut
which represents about 15 percent of total FCC gasoline is reduced by
20 octane numbers.
3. The hydrogenated Cg heart cut containing 15 volume percent benzene
and 95 percent of the benzene in the FCC gasoline is sent to a benzene extraction
plant where 99 percent of the benzene is removed as commercial grade benzene.
3.6.3 Benzene Removal From Other Gasoline Blending Components
Reduction of the benzene content of straight run naphtha is also feasible
and would further reduce the benzene content by a nominal 7 percent for a total
of 88 percent reduction from the pool. The process for benzene removal would
be similar to that for FCC gasoline, but requires only mild hydrogenation to
remove the sulfur in the S.R. naphtha. The benzene content of SR naphthas
are directly dependent on the benzene in each crude oil. A detailed analysis
of this variability to accurately determine removal costs was beyond the scope
of the ADL study.
3-31
-------
A variety of other gasoline blending components such as isomerate,
hydrocrackate, and natural gasoline contribute to 12 percent of the benzene
in the pool. Although it is probably feasible to reduce their benzene
content in a similar manner, the complexity of analysis of removal options
was not considered warranted in this study.
3.7 SUMMARY
This chapter has shown that controls can be applied to bulk terminals,
bulk plants and service stations which reduce benzene emissions significantly.
Recovery or oxidation systems at terminals reduce truck loading emissions
by as much as 97 percent. Balance systems at bulk plants can reduce total
plant emissions by about 70 percent while "add-on" equipment can reduce the
plant emissions by over 90 percent. Service stations employing balance
systems can cut benzene emissions from the loading of storage tanks by
95 percent. It was estimated that reduction of benzene from liquid gasoline
at the refinery could reduce benzene emissions from terminals, plants, and
service stations by over 80 percent.
Table 3-3 summarizes these reductions for the individual sources
within the facilities.
3-32
-------
TABLE 3-3. EFFECT OF CONTROL TECHNIQUES ON BENZENE EMISSIONS
SECTOR
Bulk Terminal
Bulk Plant
CO
1
OJ
Service Station
SOURCE
Loading of trucks
Storage Tanks
Loading of trucks
Storage Tank
Loading
Typical BZ Emission
Factor with Current
Controls (mg/1)
4.8
17.7
11.2
9.0
1
Control
Technique
RF
CRA
AA
TO
BZ reduction in gasoline 3/
Balance ST -/
Balance ST & T -/
Controlled
BZ Emission Factor
mg/1
0.3
0.3
0.3
0.3
0.96
9.4
5.7
Add-on Controls , ( 1.7
BZ reduction in gasoline -'
Submerged fill
Balance system w/submerged fi
Add-on Controls w/splash fill
BZ reduction in gasoline '
Balance
BZ reduction in gasoline —
3.54
4.8
11 .-48
1.1
2.24
.45
T.8
I/Balance ST = storage tanks only are balanced to transport trucks
I/Balance ST & T = storage tanks are balanced to transport trucks, delivery trucks are balanced to storage tanks
RF - Refrigeration CRA - Compression-Refrigeration-Absorption
AA - Adsorption-Absorption TO - Thermal Oxidizer
3/At 94.5% extraction from 86% of pool (reformate and FCC gasoline)
-------
3.8 REFERENCES
1. Test No. A, EMB Project No. 75-GAS-10, EPA Contract No. 68-02-1407,
Task No. 7, September, 1975.
2. Test No. B, EMB Project No. 75-GAS-8, EPA Contract No. 68-02-1407,
September, 1975.
3. Test No. C, EMB Project No. 76-GAS-16, EPA Contract No. 68-02-1407,
September, 1976.
4. Test No. D, EMB Project No. 76-GAS-17, EPA Contract No. 68-02-1407,
September, 1976.
5. Test No. E, EPA-650/2-75-042, June, 1975.
6. Test No. F, EMB Project No. 77-GAS-18, EPA Contract No. 68-02-1407,
November, 1976.
7. Test No. G, EMB Project No. 77-GAS-19, EPA Contract No. 68-02-1400
(Draft).
8. Test No. H, EMB Project No. 78-BEZ4, EPA Contract No. 68-02-2813,
(Draft).
9. Test No. I, EMB Project No. 78-BEZ-5 (Draft).
10. Test No. J, EMB Project No. 78-BEZ-3 (Draft).
11. Test No. K, EMB Project No. 78-BEZ-ll.
12. Edwards, Ray C., President, Edwards Engineering Corporation,
letter to Mr. George Walsh, U.S. EPA, January 27, 1978.
13. Reference 7, Op. Cit.
14. Reference 8, Op. Cit.
15. Reference 9, Op. Cit.
16. Scott Environmental Technology, "Control Characteristics of
Carbon Beds for Gasoline Vapor Emissions," EPA-600/2-77-057, February, 1977.
3-34
-------
17. Reference 9, Op. Cit.
18; Supplement No. 7 for Compilation of Air Pollutant Emission
Factors, Second Edition, U.S. EPA, April, 1977.
19. Pacific Environmental Services, "Compliance Analysis of Small
Bulk Plants," prepared for U.S. EPA Region VIII, October, 1976.
20. Ibid.
21. Hasselmann, D. E.. "Gasoline Transfer Vaoor Recovery Svstems-
San Diego County, California," TRW, Inc., Contract No. 68-02-0235, for EPA,
November, 1974.
22. Scott Research Labs, "Performance of Service Station Vapor
Control Concepts," API Project EF-14, Phase 2, Interim Report CEA-8,
San Bernardino, California, June 26, 1974.
23. "Cost of Benzene Reduction in Gasoline to the Petroleum Refining
Industry," Arthur D. Little, Inc., EPA-450/3-78-021, for EPA, April, 1978.
24. Hartle, R. and Young, R., "Occupational Exposure to Benzene at
Service Stations," National Institute for Occupational Safety and Health,
Cincinnati, Ohio, June 16, 1977.
25. Runion, H.E., "Benzene in Gasoline II," Gulf Science and
Technology Company, Pittsburgh, Pennsylvania, November, 1976.
26. Luskin, M. M., "Hydrocarbon Distribution in Commercial Gasolines-
Summer, 1976," Petrochemicals Division, E.I. duPont de Nemours Company, Inc.,
June, 1977.
3-35
-------
4.0 ALTERNATIVE CONTROL LEVELS
This chapter presents control alternatives for the gasoline marketing
industry and shows the relative impacts on national benzene emissions for
each. Table 4-1 summarizes the options and Table 4-2 outlines emission levels.
As described in Chapter 2 the United States marketing network consists
of bulk terminals,which typically have top submerged or bottom fill on transport
trucks; bulk plants,which have no controls on delivery trucks and are generally
equipped with bottom fill on storage tanks; and service stations, of which about half
have splash loaded underground storage tanks and the other half utilize sub-
merged f i 11.
Controls for terminals are well demonstrated in that 300 or so of the
1500 terminals currently employ some sort of VOC recovery or oxidation
device on loading facilities (described in Chapter 3.0). These devices
include refrigeration (RF), compression-refrigeration-absorption (CRA),
adsorption (AA), and incineration (TO). Tests indicate that a high benzene
12345
removal efficiency is expected from the use of each of the systems. ' * ' '
Bulk plants can install submerged fill on delivery trucks and balance
storage tanks to transport trucks. This is a relatively inexpensive method
of reducing total plant emissions by about 50 percent (from 28.9 mg/1 to 14.2 mg/1).
A balance system installed on the entire plant (trucks and tanks) can reduce
emissions by over 70 percent (from 28.9 mg/1 to 6.2 mg/1). Finally,
4-1
-------
"add-on" controls, similar to those described for terminals, applied to tanks
and trucks at plants can reduce benzene emissions by at least 90 percent.
(The difference between the efficiency of add-on controls and
the efficiency of the total plant balance system is that add-on controls reduce
breathing losses, which are unaffected by balancing.)
Filling losses from underground storage tanks at service stations can
be reduced over 95 percent by use of a balance system. The shortcoming
of this approach is that unless the truck delivering gasoline to the station
is controlled at the terminal or bulk pi ant,the truck's benzene vapors are emitted
to atmosphere anyway. As will be seen in the discussion, this is a possibility
with Option 1.
Reducing benzene content in gasoline at the refinery will not only
reduce benzene emissions at terminals, plants and service stations by 80 percent,
but may also reduce emissions from significant benzene sources such as storage
tanks, automobile refueling operations and auto tailpipes. Smaller sources
such as marine operations, spills and consumer equipment would also be con-
trolled. This added impact must be weighed into consideration when comparing
the options listed here, (note: EPA is still developing data on the effect of
benzene content in gasoline on auto tailpipe emissions. Because these studies
have not been completed as yet, this document does not estimate the total benzene
control, if any, attributable to reduction of benzene in gasoline.)
Four options are presented in this chapter which combine different control
strategies at each segment. For example, in Option 1, the least effective
alternative for reducing benzene, high efficiency add-on controls at terminals
are combined with balance of transport trucks and storage tanks at bulk plants
(with submerged fill for delivery trucks) and with balance systems at service
stations.
The options are presented in increasing effectiveness of benzene reduction.
(Refer to Tables 4-1 and 4-2.)
4-2
-------
4.1 OPTION 1
Option 1 reduces the benzene emissions the least of all the options.
Bulk terminal operators are required to install refrigeration, adsorption-
absorption, incineration or equivalent systems on loading facilities. All
of these devices are well demonstrated on operating terminals in the
United States. They are considered to be the most effective control methods
in current use and test data indicate several systems have the capability
of reducing benzene by 95 percent.
Bulk plants under Option 1 would be required to install submerged fill
on delivery trucks and to balance storage tanks with incoming transport
trucks. This effectively means that 50 percent of the plant is uncontrolled.
It also means that those service stations which are serviced by bulk plants
would be uncontrolled since the delivery trucks would not be equipped to
recover vapor.
Under Option 1, service stations serviced by bulk terminals would
be required to install balance systems for the filling of their under-
ground storage tanks. Those serviced by bulk plants would be exempted
from balancing (since the delivery trucks from bulk plants would not be
equipped to handle the vapor), but would still be required to install
submerged fill. It has been estimated that about 40 percent of all station
gasoline throughput comes from bulk plants.
The balance system is very effective in handling gasoline vapors.
Those stations installing balance could expect a 95 percent reduction in
benzene emissions from filling the underground storage tanks.
Overall efficiency of Option 1 is about 60 percent. National benzene
emissions from the marketing network sources discussed in this document
would drop from about 10,500 to 4050 metric tons per year.
4-3
-------
4.2 OPTION 2
Option 2 involves the reduction of benzene from liquid gasoline at
the refinery. Estimates have been made of possible reductions of benzene
o
from the gasoline pool (see Chapter 3). Reductions of 80 percent appear
to be possible. It is expected that an 80 percent reduction in benzene from
the liquid gasoline would mean an approximate 80 percent reduction in the benzene
in gasoline vapor. Using this factor, an 80 percent reduction in benzene
emissions can be expected with Option 2. National emissions from the sources
discussed here would drop from about 10,500 to 2100 metric tons per year.
This technology would also remove benzene from other significant
sources of the pollutant. Emissions from sources such as automobile
refueling operations, and gasoline storage may be reduced by as much as
80 percent.
4.3 OPTION 3
Option 3 represents a more effective alternative for the marketing
network. Bulk terminals would be required to apply the same effective
controls as listed in Option 1, (e.g. absorption, refrigeration, oxidation,
adsorption or equivalent). Reduction at bulk terminals is the same in
Option 3 as in Option 1 (about 95 percent).
The bulk plant, under Option 3, would be required to install a full
balance system on both delivery trucks and storage tanks. Note, in Table 4-2,
the effect of applying the balance system to delivery trucks on emptying losses
in the storage tank. As the tank is emptied, the increased volume of the vapor
space is taken up by nearly saturated vapors from the delivery trucks and not
4-4
-------
by fresh air. If the delivery truck vapors are close to saturation (as
would be expected from trucks returning from balanced service stations),
emptying losses would approach zero. For Table 4-2, it was assumed that the
trucks would return saturated and emptying losses are zero.
Service stations would be required to apply the highly effective
balance systems. In this case, all service stations would balance to in-
coming trucks. The filling losses would be ultimately carried by truck
back to the terminal.
Overall efficiency of Option 3 is about 86 percent. Benzene emissions
from the marketing network would decrease from about 10,500 to 1400 metric
tons per year.
4.4 OPTION 4
The fourth and last option represents the highest emission reduction
possible for the gasoline marketing network with current technology. It
differs from Options 1 and 3 only in that all of the significant losses from
bulk plants are controlled.
For bulk terminals, the add on controls (absorption, refrigeration, oxidation,
adsorption, etc.) are required. Bulk plants are required to install similar
controls on both storage tanks and loading racks. This would result in at
least 90 percent control of breathing, emptying and filling losses from the
plant. Losses may be reduced by as much as 95 percent. Since no "add-on"
controls have been applied to bulk plants, however, a conservative 90 percent has
been assumed. All service stations would be required to install balance systems.
The balanced vapors would be returned to the terminal or plant for disposal.
Overall efficiency of Option 4 is about 93 percent. National benzene
emissions from these sources would be reduced from about 10,500 to 760 metric
tons per year.
4-5
-------
TABLE 4-1. GASOLINE MARKETING NETWORK CONTROL OPTIONS
Source
A. Terminals
Loading Racks
B. Bulk Plants
Storage
Breathing
Emptyi ng
Filling
Loading Rack
Fill ing
C. Service Station
Underground /_]_
storage tank
loading only
Base Case
Top Submerged or Bottom
Fill on Trucks
No control
No control
Bottom fill
Splash fill
50% splash load
50% submerged
Option 1
Vapor Recovery
or Oxidation
No control
No control
Balance to transport
Submerged fill
Balance w/submerged
fill /2.
Option 2
Reduction of
Benzene in
Gasoline
Reduction of
Benzene in
in Gasoline
Reduction of
Benzene in
gasoline
Option 3
Vapor Recovery
or Oxidation
No control
100% control
Balance to trans-
port
Balance to
storage
Balance with
submerged fill
Option 4
Vapor
Recovery or
Oxidation
All
sources
Vapor
Recovery or
Oxidation
Balance
w/submerged
fill
/!_ Breathing & emptying losses to be discussed with refueling operations in a separate study.
^2 Those stations serviced by terminals would be balanced.
Those stations serviced by bulk plants would not be balanced (would have submerged fill).
-------
TABLE 4-2. GASOLINE MARKETING CONTROL OPTIONS - NATIONAL EMISSIONS
Terminals
Loading
Bulk Plants
Storage
Loading
Service
Stations
Underground
Storage Tank
Loading Only
/I
Throughput
liters/yr
413 x 109
165 x 109
165 x 109
413 x 109
TOTAL
PERCENT REDUf
HC Base
Emission
Factors
(mg/1)
600
600
Breathing
460
Emptying
1150
Fill ing
1400
1130
TION
gm BZ
gm HC
0.008
0.008
0.008
0.008
0.008
0.008
National
Base
metric
tons/yr BZ
1980
792
607
1518
1848
3734
10,479
0
Option 1
metric
tons/yr BZ
100
792
607
152
792
1605
4048
61
Option 2
metric
tons/yr BZ
396
158
120 ,
304
370
747
2095
80
Option 3
metric
tons/yr BZ
100
792
0
152
185
187
1416
86
Option 4
metric
tons/yr BZ
100
79
60
152
185
187
763
93
r\_ Breathing and emptying losses to be discussed with refueling operations.
-------
4.5 REFERENCES
1. Thermal Oxidizer Test I, EMB Project No. 78-BEZ-5.
2. CRA Test H, EMB Project No. 78-BEZ-4.
3. Refrigeration Test J, EMB Project No. 78-BEZ-3.
4. Adsorber Test G, Contract No. 68-02-1400, October, 1977.
5. CRA Test K, EMB Project No. 78-BEZ-ll.
6. U. S. Department of Energy, "Monthly Petroleum Statement,"
December, 1977.
7. Hasselmann, D. E., "Gasoline Transfer Vapor Recovery Systems -
San Diego County, California," TRW, Inc., Contract No. 68-02-0235, for EPA
November, 1974.
8. A.D. Little, Inc., "Cost of Benzene Reduction in Gasoline to the
Petroleum Refining Industry," EPA 450/3-78-021, April, 1978.
4-8
-------
5.0 ENVIRONMENTAL IMPACTS OF APPLYING THE TECHNOLOGY
This chapter will assess the environmental and energy impacts of
applying the control technology discussed in Chapter 3 and the control
options outlined in Chapter 4.
5.1 IMPACT ON BENZENE EMISSIONS
In order to determine emission reductions which would occur as a
result of using each technique, it is necessary to examine air pollution
control requirements of existing State and local regulations.
There are currently no State regulations which control benzene
emissions from the gasoline marketing network. There are, however, regulations
covering hydrocarbons in a few areas and these need to be examined in order
to develop the typically controlled network. Controls for benzene and
hydrocarbons are identical in terms of equipment.
Appendix E outlines the States which regulate sources under consideration
here and also indicates the hydrocarbon emissions standards which apply.
It can be seen that the typical terminal of 950,000 liters
(250,000 gallons) per day throughput is required to have top submerged
or bottom loading on loading racks. A few Air Quality Control Regions have
required vapor recovery, but it is estimated that less than 20 percent of
all United States terminals are affected by these regulations.
The typical bulk plant of 15,000 liters (4000 gallons) per day through-
put is generally required to have bottom loaded storage tanks, but there
5-1
-------
are virtually no controls required for loading operations.
It is assumed that about one-half of the retail service stations in
2
the United States are equipped with "drop tubes." The others use splash fill.
A few areas (about 15 ACQR's) require balance systems on retail stations.
These stations represent about 16 percent of all retail stations in the
3
United States (or about 12 percent of all retail station throughput). The
typical retail service station has a throughput of about 150,000 liters
(40,000 gallons) per month. Non-retail stations have a throughput of about
38,000 liters (10,000 gallons) per month (represent about 25 percent of
throughput and over 50 percent of stations).
5.1.1 Bulk Terminal Controls
It was shown in Chapter 3 that the four add-on controls source tested
for bulk terminal loading losses are approximately of equal efficiency in
reducing benzene emissions. Compression-refrigeration-absorption, refrigeration,
adsorption-absorption, and oxidation systems all achieve about 95 percent con-
trol of benzene. In terms of mass reduction, this means that the typical
terminal can reduce annual benzene emissions from 1300 to 65 kg by use of
these controls. (See Table 5-1. Data in the table have been derived in
Chapters 3 and 4.) Reducing the amount of benzene in liquid gasoline at the
refinery can reduce the terminal losses by 80 percent. The typical terminal
benzene emissions would go from 1300 kg/yr to 260 kg/yr by use of this method.
5-2
-------
5.1.2 Bulk Plant Controls
The bulk plant has two major source areas: the vent from storage
tanks emitting filling, emptying and breathing losses; and delivery trucks
which emit benzene along with other hydrocarbons during loading.
Table 5-1 shows that losses from storage tanks could be reduced
from 76 to 24 kg/yr with use of the balance system; losses are reduced
to 8 kg/yr with add-on controls; and losses drop to 15 kg/yr with benzene
reduction in gasoline. Note that balancing the total plant reduces losses
even further than balancing the storage tank only. This is because emptying
losses and delivery truck filling losses are controlled with total plant balancing.
Losses from filling delivery trucks can be reduced from 48 kg/yr to
5 kg/yr by use of add-on controls. Other options reduce the kg/yr emission
level to 2 for balancing, 10 for benzene reduction in gasoline, and 20 for
submerged fill only.
5.1.3 Service Station Controls
Service station underground tank filling losses can be controlled
by one of two ways. A balance system can be installed to vent filling
losses to the truck delivering liquid gasoline or benzene can be reduced
at the refinery. In the first case, Table 5-1 shows the effect as
reducing benzene from 16 to less than 1 kg/yr. The second case reduces
benzene to 3 kg/yr.
Station balance systems are only effective if the delivery or
transport truck is equipped to transfer the vapors ultimately to the
terminal for disposal.
5-3
-------
TABLE 5-1. ESTIMATED IMPACT ON BENZENE EMISSIONS FOR MODEL FACILITIES
SOURCE
Bulk Terminal Loading
Bulk Plant Storage
tn
Loading
Service Station Filling
Underground Storage Tank
Throughput
950,000 liters/day
15,000 liters/day
15,000 liters/day
150,000 liters/mo
Estimated BZ
Emission Rate
mg/1
4.8
17.7
11.2
9
Typical Annual
Emission Rate
kg/yr
1300
76
48
16
Control
Method
with CRA, R, Ad, OX
BZ reduction in gasoline
Balance 1 —
Balance 2 -/
R, OX
BZ reduction
Submerged f i 1 1
Balance 1 I/
Add-on (R, OX)
BZ reduction
Balance
BZ reduction
Controlled
Rate
kg/yr *
65
260
40
24
8
15
20
2
5
10
0.8
3.3
* 286 days/yr
]_/ Balance only incoming trucks
2/ Balance entire plant
-------
5.1.4 Control Options
Table 5-2 contains the same control options discussed in Chapter 4.
The table sums the individual source emissions into a national emission
reduction for each option. Option 1 reduces marketing benzene emissions nationally
from 10,500 to 4050 metric tons per year. Option 2 reduces the benzene emissions
to 2100, Option 3 to 1400, and Option 4 to 760 metric tons per year.
5.2 OTHER AIR IMPACTS
There are air impacts directly associated with some control technology.
Incinerators, for instance, emit small amounts of NO,,, CO and particulate.
All of the control options presented here, except for reduction of benzene
at the refinery, reduce hydrocarbon losses.
There are other components in gasoline vapor which have been
implicated in health problems. These include the additives ethylene
dichloride and ethylene dibromide which are suspected carcinogens and
xylene which is similar in structure to benzene and also suspect. All controls
in the marketing industry would control these suspected toxics except
reduction of benzene at the refinery.
This section will discuss direct air impacts for each individual method
(other than benzene removal) and then will sum the impacts for each control
option. Table 5-3 summarizes this section.
5.2.1 Bulk Terminals
The add-on controls discussed for bulk terminals generally have no
adverse impacts on air emissions. CRA, refrigeration, adsorption and
oxidation minimize emissions of benzene and other hydrocarbons to the
5-5
-------
TABLE 5-2. GASOLINE MARKETING CONTROL OPTIONS - ESTIMATED NATIONAL EMISSIONS
Terminals
Loading
Bulk Plants
Storage
n
Tt
Loading
Service
Stations
Underground
Storage Tank
Loadinq Only
I/
Throughput
liters/yr
413 x 109
165 x 109
165 x 109
413 x 109
TOTAL
HC Base
Emission
Factors
(mg/1)
600
600
Breathing
460
Emptying
1150
Filling
1400
Filling
1130
gm BZ
gm HC
0.008
0.008
0.008
0.008
0.008
0.008
National
Base
metric
tons/yr BZ
1980
792
607
1518
1848
3734
10479
Option 1
metric
tons/yr BZ
100
792
607
152
792
1605
4048
Option 2
metric
tons/yr BZ
396
158
120
304
370
747
2095
Option 3
metric
tons/yr BZ
100
792
0
152
185
187
1416
Option 4
metric
tons/yr BZ
100
79
60
152
185
187
763
I/ Breathing and emptying losses to be discussed with refueling operations.
27 Under Option 1, 40 percent of service station throughput is uncontrolled.
-------
atmosphere. Oxidizers, however, vent small amounts of hydrocarbon as
well as secondary pollutants as products of combustion. Table 5-3 shows
the estimated quantity of secondary pollutants emitted in the effluent
of the control device.
The table also shows the effect of each control technique on hydro-
carbon emissions from the typically uncontrolled source. As can be seen,
significant quantities of hydrocarbon are controlled by add-on controls.
Benzene reduction at the refinery has no effect on hydrocarbon losses.
5.2.2 Bulk Plants
The balance system does not increase other air contaminants to the
atmosphere. It does, however, significantly reduce hydrocarbon and
suspected toxic substance emissions.
Add-on controls provide the same adverse and positive impacts on
emissions to the atmosphere as shown for bulk terminals, except in smaller
quantities (because of a lower throughput).
5.2.3 Service Stations
The balance system reduces hydrocarbons and suspected toxic
substances to the atmosphere with no effect on other contaminants.
5.2.4 National Impacts for Options
Table 5-4 shows the national impact of Options 1-4 on other air
contaminants. It is shown that Option 4 would the largest negative impact
because of the widespread use of add-on controls.
5-7
-------
TABLE 5-3. OTHER AIR IMPACTS FOR MODEL FACILITIES ESTIMATED FROM TEST DATA - kg/yr
SOURCE
Bulk Terminal Loading
Bulk Plant Storage
n
o
Loading
Service Station
Control Technique
CRA
Ref
Ad
OX
Reduction at refinery
Balance - Incoming
Balance - Incoming/
Ref Outgoing
OX
BZ reduction at
refinery
Balance
Submerged fill
Ref
OX
BZ reduction at
refinery
Balance
BZ reduction at
refinery
Particulate
0
0
0
Negligible
0
0
0
Negligible
0
0
0
Negligible
0
CO
0
0
0
17,000
0
0
0
367
0
0
0
232
0
NOX
0
0
0
4800
0
0
0
108
0
0
0
68
0
HCI/
(140,000)
(140,000)
(140,000)
(140,000)
See 1
(4,500)
(6,500)
(7,600)
(7,600)
See 1
(5,700)
(3,500)
(5,400)
(5,400)
See
(1,865)
See T
EDC -1
(Unk)
(Unk)
(Unk)
(Unk)
able 54
(90%)
(90%)
(Unk)
(Unk)
able 54
(90%)
(57%)
(Unk)
(Unk)
able 54
(95%)
able 5-4
EDB -1
(Unk)
(Unk)
(Unk)
(Unk)
(90% reduction)
(90% reduction)
(Unk)
(Unk)
(90%)
(57%)
(Unk)
(Unk)
(95%)
!_/ Parentheses indicate reduction in pollutant from typical facility
2/ Unknown uncontrolled emission rate - controls will reduce toxics ethylene dichloride and etnyiene dibromide
-------
TABLE 5-4. ESTIMATED NATIONAL AIR IMPACTS OTHER THAN BENZENE - WORST CASES
(thousands of metric tons/yr)
POLLUTANT
Participate
NOX
CO
HC
EDC
EDB
OPTION 1
Negligible
*
8
*
30
(936) *
(Unknown)
(Unknown)
OPTION 2
4.5
31.8
2.45
.7
No effect
No effect
OPTION 3
Neg
*
8
*
30
(1200)
(Unknown)
(Unknown)
OPTION 4
Neg
*
12
*
40
(1300) *
(Unknown)
(Unknown)
en
* 1750 terminals and 17,850 bulk plants using oxidation systems - worst case
Parentheses indicate reduction of pollutant
-------
5.3 WATER POLLUTION IMPACT
No control option discussed here uses water (the adsorber is vacuum
regenerated). Water is present, however, in treated vapors and for all
add-on systems it is recovered with the gasoline, separated, and disposed of.
Table 5-5 estimates the impact that add-on controls have on wastewater.
The estimates are based on analysis of water samples taken during EPA tests.
National emissions are extrapolated for each control option.
The amount of water will vary, depending on the temperature and
relative humidity of the atmosphere. It is suspected that the removal of
benzene from liquid gasoline at the refinery will place an additional
burden on refinery waste water. This burden has not been quantified.
5.4. IMPACT ON SOLID WASTE
The disposal of discarded carbon is the only major source of solid
waste for the marketing network control methods. Table 5-6 estimates the
impact for a single terminal and bulk plant and extrapolates to national
impacts for the four control options.
Assumptions made include a conservative estimate of carbon life
(3-5 years), a mass of carbon necessary, and total industry use of
carbon. It is further conservatively assumed that the carbon cannot be
regenerated, but would be disposed of.
5-10
-------
TABLE 5-5. WATER IMPACT - WORST CASES
Control
Method
R, CRA, AA
OX
R
OX
Balance
m Balance
_j
Source
Bulk terminals
Bulk Plants
Service Station
Estimated
Quantity
of water
disposed
of
^ 20 I/day
0
1 I/day
0
0
0
ppm
HC
0-57
0
0-57
0
0
0
Total
Estimated National Mass Rate - kg/yr - HC in waste water
Option 1
570
0
0
570 U
Option 2
Unknown
Option 3
57U
0
0
570 U
Option 4
570
293
0
0
0
863 -/
V Trace benzene in water samples
2/ Assumes no terminal uses oxidation
!' All plants and terminals use refrigeration
-------
TABLE 5-6.
SOLID WASTE IMPACT - WORST CASE
Control
Method
Adsorption
Source
Bulk terminals
Bulk plants
Service Stations
tstimatea
Quantity
of
Carbon
4500 kg
0
0
TOTAL
Estimated National Mass Rate (106 kg/yr)
Option 1
2.2
0
0
2.2
Option 2
0
0
0
0
Option 3
2.2
0
0
2.2
Option 4
2.2
0
0
2.2
I
_J
ro
ASSUMPTION: All terminals use adsorption-absorption.
Bed life is 4 years.
Carbon cannot be regenerated.
-------
5.5 IMPACT ON ENERGY
All control methods discussed here, except balance systems and
submerged fill, require energy. The amounts of energy vary. This section
estimates the energy requirements for each method and then sums the impacts
for each option. Table 5-7 summarizes the estimates.
Add-on controls require energy to operate. Integral parts of
refrigeration, AA, and CRA units are the electrically powered pumps and
compressors. In the case of these controls, however, there is an energy
credit in the form of recovered gasoline. The recovery credit has been
added into the penalty for the net requirement. Oxidation units require
electrically powered blowers and auxiliary fuel in some cases. Recent model
oxidizers use auxiliary fuel for pilot flame only. The energy requirements
for all of these sources are small compared to the energy consumption of
removing benzene from gasoline to the 80 percent level discussed in this
document.
Table 5-7 shows the energy consumption for each technique. The
table also estimates the national consumption for each option.
5.6 AIR QUALITY IMPACT
Estimates are being made using dispersion analysis of the impact of
the controls on ambient air levels of benzene. The results will be
tabulated in similar fashion to Table 5-8.
The base case and subsequent optional controls consider the total
benzene emissions from the facility and not just the sources discussed in
this document. Thus, bulk terminal concentration estimates include
5-13
-------
TABLE 5-7
ENERGY IMPACTS OF CONTROL METHODS -'
SOURCE
Terminals
Bulk Plants
01
*"
Service Stations
Control
Method
CRA 2J
Ref 2J
Ad 21
OX 2J
Reduction of BZ
Ref 3/
OX 3/
Reduction of BZ
Submerged fill/balance
Balance/submerged fill
Reduction of benzene
Estimated
Energy
Required
per facility
10 Joules/day
(48,300)
(47,300)
(48,500)
1,860
(820)
150
_
TOTAL -
Estimated National Energy Required
Option 1
1012 J/yr
(5220)
(5220)
(5220)
200
NA
NA
(2720)
(5180)
(23,270)
Option 2
1012 J/yr
See
Total
See
Total
See
Total
330,000 or
54,100,000 barre
Option 3
1012 J/yr
Same as
Option 1
NA
NA
(3141)
(5180)
-
)23,700)
Is
1- . . ,-,--!
Option 4
1012 J/yr
(3220)
(3110)
(3250)
200
(2100)
383
(5180)
(16,277)
I/ Parentheses indicate energy credits from recovered gasoline.
21 Assumes each type of control method installed at 25 percent of terminals
3i/ Assumes each type control method installed at 50 percent of bulk plants
4/ One liter of gasoline equals 3.6 x 10 Joules. 6.1 billion Joules per barrel of oil
-------
TABLE 5-8. AIR QUALITY IMPACT
(ppb BZ - meters from fence!ine)
FACILITY
Terminal
Bulk Plant
Rural
Urban
Service
Y1 Station
ui
Base Case
Option 1
Option 2
Option 3
Option 4
-------
emissions from storage tanks and service station estimates include refueling
operations.
5.7 OTHER ENVIRONMENTAL CONCERNS
Other concerns to be considered include space requirements, availability
of resources, and noise. All of these are considered insignificant. Estimates
are made in Table 5-9 of the impacts of individual controls.
Reduction of benzene at the refinery will also reduce benzene emissions
from sources other than those discussed in this document. These sources are
ships and barges, storage tanks at terminals, and automobile refueling
operations. It is suspected that the reduction of benzene in gasoline will
also reduce automobile tailpipe emissions, but no data are currently available
to confirm this. (EPA is accumulating the data in a research project.)
Tailpipe benzene emissions may occur as a result of the combustion process
in the auto engine, thus the effect of reducing benzene in gasoline may not
necessarily be linear to tailpipe benzene emissions.
4
Auto tailpipe benzene emissions are significant. A recent study
estimated that benzene from automobiles totaled 169,500 metric tons in 1976.
This compares to 10,500 metric tons emitted from the marketing sources discussed
in this document. Tailpipe benzene losses account for about 65 percent of all
benzene sources. Because of decreasing gasoline consumption and lowered
hydrocarbon emissions from new cars, tailpipe benzene emissions in 1985 are
predicted to be much lower than in 1976--down to 27,730 metric tons per year.
The effect of reducing benzene content in liquid gasoline on these levels is
unknown at this time.
5-16
-------
TABLE 5-9.
OTHER ENVIRONMENTAL IMPACTS
SOURCE
Bulk Terminals
Bulk Plants
en
i
*vl
Service Stations
Control
Technique
CRA
Ref
Ad
OX
OX w/vapor holder
Ref
OX
Balance
Balance
Approximated
Space Requirements
Sq Meters
50
30
30
30
50
30
30
Neg
Neg
Noise Level - db I/
< 70 @ 7 meters
< 70 @ 7 meters
< 70 @ 7 meters
< 70 @ 7 meters
< 70 @ 7 meters
< 70 @ 7 meters
< 70 0 7 meters
0
0
Estimated
Availability of
Resources
(months)
6 - 12
6 - 12
6 - 12
6 - 12
6-12
: 6-12
6 - 12
6-12
6 - 12
I/ A CRA unit, which created significantly more noise to the unprotected ear than any
other system encountered, was tested for noise levels by a trained analyst. The
system registered less than 70 db at 7 meters from the compressor (the noise source).
-------
5.8 REFERENCES
1. A. D. Little, Inc., "The Economic Impact of Vapor Control on the
Bulk Storage Industry," prepared for EPA, July, 1978, Draft report.
2. Radian Corporation, "Control of Hydrocarbon Emissions From
Petroleum Liquids," EPA-600/2-75-042, September, 1975.
3. EPA, OAQPS, "Review and Analysis of Comments Received in Response
to EPA's November, 1976, Proposed Stage II Vapor Recovery Regulations,"
April 18, 1977.
4. PEDCo Environmental, Inc., "Atmospherip Benzene Emissions,"
EPA-450/3-77-029, October, 1977.
5-18
-------
6. ECONOMIC IMPACT ANALYSIS
6.1 BULK TERMINALS
6.1.1 Bulk Terminal Industry Characterization
6.1.1.1 Introduction
Bulk terminals are primary storage facilities which receive petroleum
products from domestic and offshore refineries for market distribution.
Output from domestic refiners moves to market via pipeline terminals and
marine terminals; imported product moves via marine terminals (Figure 6-1)
Most terminals load all of the product they receive into truck transports
at the terminal's loading racks. These truck transports have capacities
between 30 and 34 M3 (8,000 and 9,000 gallons) and deliver gasoline to
service stations and bulk plants for further distribution. Some large
terminals, however, distribute only a portion of their products at the
loading rack and move the remaining volumes to secondary storage facilities
via pipeline, barge or coastal tanker.
6.1.1.2 Operations and Market Environment
For more than a quarter of a century until about 1970, the production
of domestic and foreign crude contributed the most significant portion of
total corporate earnings at integrated oil companies. The function of
marketing then was to increase the demand for petroleum products thereby
generating greater profits from increased crude production. To assure a
high demand for products, prices were set at levels which encouraged
consumption but which did not fully recover the true costs associated with
refining and marketing. These activities were, in effect, subsidized by
the profitability of crude production.
6-1
-------
DOMESTIC
REFINERIES
IMPORTS
PIPELINE
TERMINAL
CT>
t\>
BULK
PLANTS
MARINE
TERMINAL
BULK
PLANTS
CONSUMER
FIGURE 6-1 PRODUCT FLOW DIAGRAM
-------
Beginning around 1970, oil companies began to view their refining
and marketing operations as separate profit centers to be judged on "stand
alone" economics. No longer would marketing activities, including bulk
storage operations, be subsidized by crude production. Terminals were
now expected to recover all operating expenses as well as provide an
acceptable return on capital. Facilities unable to operate profitably
were forced to close. This trend, which was well underway in 1973, was
accelerated by the Oil Embargo of 1973-1974.
"Stand alone" economics has caused petroleum marketers, both majors
and independents, to review their marketing strengths and to re-evaluate
overall strategies. This has led to discussions to close many uneconomic
or marginal facilities. Due to this "market rationalization," some marketers
are withdrawing from selected regions of the country as part of an overall
corporate strategy. Terminals in these markets will either be consolidated,
sold or closed.
The cost of transporting petroleum products by pipeline are significantly
less expensive than by either tanker or barge. Most product pipelines are
currently operating at full capacity thereby making pipeline terminals the most
financially attractive type of bulk storage. Pipeline terminals do not
compete directly with each other because of their well-defined locus of
operation. Marine terminals, however, transport the marginal barrel of
product and may compete among themselves whenever several facilities
operate within the same area. Marine terminals of equal size compete
with each other but none realize a competitive advantage if they are equally
efficient. If the competing terminals are not equal in size, the largest,
and presumably the most efficient, facility will gain a competitive edge
6-3
-------
over the smaller and less efficient marine terminals. This competitive
disadvantage may initiate or accelerate a marketer's decision to cease
marketing operations in selected areas.
6.1.1.3 Bulk Terminal Population
The total number of bulk terminals has declined from 1,925 in 1972
to 1,751 in 1978, a decrease of 9 percent (Table 6-1). This decline has
been the result of "stand alone" economics and the rationalization process
of petroleum marketers. An estimated 1,511 or 86 percent of all terminals,
store some amount of gasoline. Terminals not storing gasoline may specialize
in residual fuels, distillates, bunker fuels or chemicals. Many terminals
which only store home heating fuel are located in the Northern states.
Most terminals are located in PADD's I and II (Figure 6-2), PADD I
has 43 percent of all bulk terminals and 43 percent of the gasoline terminals,
i.e. those facilities having some gasoline storage. PADD II has 24 percent
of all terminals and 23 percent of all gasoline terminals. The large
number of terminals in these two PADD's reflects the regions' lack of
refinery self-sufficiency and their reliance on shipments from other parts
of the U.S. and from foreign countries in order to meet their local product
demand. PADD I received 88 percent of all petroleum products imported
into the U.S. in 1976 and 90 percent of all imported gasoline (Tables 6-2
and 6-3). Together, PADD's I and II received almost all of the inter PADD
shipments originating in PADD III.
While the total number of terminals in the U.S. decreased 9 percent
since 1972, total storage increased 30 percent to 122.5 million M3 (770.7
million barrels) (Table 6-1). Gasoline storage is estimated to be 47.1
o
million M (296.3 million barrels), or 38 percent of total terminal storage.
6-4
-------
TABLE 6-1
BULK TERMINAL POPULATION*
PADD
I
II
III
IV
V
Total
* Does
Source:
Terminals
745
429
276
39
262
1,751
Al 1
ALL
%
Total
43%
24%
16%
2%
15%
100%
not include product
Bureau of
Census, 1
National Petroleum
1978 Director - Bui
TERMINALS
Total Stc
Capacil
000 M3
64,172
25,155
20,068
1,151
11,988
122,534
storage at
972 Census
>rage
^y
(000 Bbl)
(403,633)
(158,219)
(126,223)
( 7,238)
( 75,403)
(770,716)
refineries.
of Wholesal
News, Factbook (1972-1
k Liquid Terminals and
i
o/
•k
Total Terminals
52%
21%
16%
1%
10%
100% 1,
e Trade; U.S. Army
657
343
234
39
238
511
FERMINALS
%
Total
43%
23%
15%
3%
16%
100%
STORING GASOLI
Gasoline Stora
Capacity^
000 M3
23,815
9,875
8,228
674
'4,517
47,109
Corps of Engineers, Port
(000
(149
( 62
( 51
( 4
( 28
(296
MC
INC.
ge
Bbl)
,792)
,115)
,753)
,240)
,408)
,308)
%
Total
51%
21%
17%
1%
10%
100%
Series;
978); Independent Liquid Terminals Association,
Storage Facilities; Industry contacts; ADL estimates.
-------
(Incl. Alaska
and Hawaii I
en
01
FIGURE 6-2
PETROLEUM ADMINISTRATION FOR DEFENSE DISTRICTS
-------
TABLE 6-2
PADD
I
II
III
IV
V
Total
DEMAND
1,032
768
517
84
375
2,776
(
(
(
(
(
6
4
3
2
(17
,488)
,828)
,253)
531)
,359)
,459)
REGIONAL PRODUCT SUPPLY/DEMAND - 1976
000 M3/Day (000 Bbl/Day)
TM-rrn nnnrv CUTDMITMTC
j.nii-i\ \r\u\j oiin rii_ii i o
REFINERY FROM FROM FROM FROM
OUTPUT I II III IV
270
597
927
73
344
2,212
(
(
(
(
(
(1
1,700) — 18 (111) 491 (3,089)
3,757) 29 (183) -- 111 ( 696) 4 (26)
5,832) -- 17 (110)
459) - 10 ( 63)
2,166) — — 12 ( 76) 10 (61)
3,914)
FROM
V IMPORTS
269
15
- ( D 6
3 (18) 3
14
306
(1,691)
( 94)
( 37)
( 16)
( 89)
(1,927)
13
57
180
10
2
257
OTHER
( 80)
( 356)
(1,134)
( 62)
( 14)
(1,618)
Source: U.S. Department of Energy, Monthly Petroleum Statement
-------
TABLE 6-3
REGIONAL GASOLINE SUPPLY/DEMAND - 1976
PADD
I
II
III
IV
V
Total
DEMAND
381 (2,396)
380 (2,388)
150 ( 945)
36 ( 227)
162 (1,022)
1,109 (6,978)
REFINERY
OUTPUT
119 ( 750)
300 (1,887)
371 (2,331)
32 ( 203)
144 ( 904)
966 (6,075)
000 M3/Day (0(
TMTf
il\ I I
FROM FROM
I II
7 (41)
22 (140)
9 (56)
6 (36)
—
)0 Bbl/Day)
[R PADD SHIPMENT
FROM
III
255 (1,606)
41 ( 260)
—
—
6 ( 37)
FROM
IV
__
2 (15)
—
--
6 (35)
FROM
V IMPORTS
19 (119)
- ( 1)
1(6)
1 (9) - ( 0)
1(4)
21 (130)
OTHER
3 ( 20)
35 (218)
72 (455)
5 ( 29)
8 ( 51)
123 (773)
Source: U.S. Department of Energy, Monthly Petroleum Statement
-------
Because of the large number of terminals in this area, PADD's I and II
account for most of the total terminal storage and most of the total gaso-
line storage. PADD I has 52 percent of all storage and 51 percent of the
gasoline storage while PADD II has 21 percent of total storage and 21
percent of the total gasoline storage.
6.1.1.4 Bulk Terminal Size
Small facilities comprise the largest portion of the bulk terminal
3
population. Almost half of all bulk terminals have less than 32,000 M
(200,000 barrels) of storage capacity (Table 6-4). Another 30 percent
have capacities between 32,000 and 95,000 M (200,000 and 600,000 barrels);
3
22 percent have storage greater than 95 M (600,000 barrels). Similarly,
50 percent of gasoline terminals have total storage capacity less than
-3
32,000 M (200,000 barrels); 28 percent have capacities between 32,000 and
95,000 M (200,000 and 600,000 barrels); and 22 percent have a storage
capacity greater than 95,000 M (600,000 barrels).
The distribution of terminals by throughput is fairly even across the
selected throughput ranges (Table 6-5). Approximately 36 percent of all
3
terminals have total product throughput less than 636 M /day (168,000 gallon/
•3
day); 27 percent have a throughput between 636 and 2544 M /day (168,000 and
672,000 gallon/day); and 37 percent have a total product throughput greater
3
than 2544 M /day (672,000 gallon/day). Almost half of the gasoline terminals,
3
48 percent, have a gasoline throughput less than 754 M /day (200,000 gallon/
day); 27 percent have a gasoline throughput between 759 and 1514 M /day
(200,000 and 400,000 gallons/day); and 25 percent have a gasoline throughput
3
greater than 1514 M /day (400,000 gallon/day).
6-9
-------
TABLE 6-4
BULK TERMINAL STORAGE DISTRIBUTION
en
i
o
TOTAL STORAGE CAPACITY
000 M3 (000 Bbl )
*32(200
32(200) - 95(600)
95(600) - 159(1000)
>159(1000)
Total
Source: Bureau of Census,
National Petroleum
Al 1 TFRMTIMfll ^
rtLL 1 LIU'I llmLo
NUMBER OF TERMINALS %
834
534
215
168
1,751
1972 Census of Wholesale Trade;
TOTAL
48%
30%
12%
10%
100%
U.S. Army
TFRMTNAI <; STORING RA9DI TNF
ICKrlllinLO olvmilNU UMJULIIiC.
NUMBER OF TERMINALS %
764
423
192
132
1,511
Corps of Engineers, Port Series;
TOTAL
50%
28%
13%
9%
100%
News, Factbook (1972-1978); Independent Liquid Terminals Association,
1978 Directory - Bulk Liquid Terminals and Storage
Facilities; Industry contacts; ADL estimates.
-------
TABLE 6-5
BULK TERMINAL THROUGHPUT DISTRIBUTION"
ALL TERMINALS
AVERAGE
PRODUCT THROUGHPUT
M3/Day (000 Gal/Day)
NUMBER OF TERMINALS
4636(168)
636(168)-2,544(672)
en
± 2,544(672)-6,995(l,848)
* 6,995(1,848)
Total
% TOTAL
626
475
375
275
36%
27%
21%
16%
TERMINALS STORING GASOLINE
AVERAGE
GASOLINE THROUGHPUT
NUMBER OF TERMINALS
M3/Day (000 Gal/Day)
4754(200)
754(200)-!,514(400)
1,514(400)-2,271(600)
>2,271(600)
% TOTAL
728
401
312
70
48%
27%
21%
5%
100%
100%
Source: Bureau of Census, 1972 Census of Wholesale Trade; Industry contacts; ADL estimates.
-------
6.1.1.5 Ownership
Major oil companies* own most of the bulk terminals. Major oil
companies own 67 percent of all terminals and 72 percent of the gasoline
terminals (Table 6-6). Independents, which includes wholesale/marketers,
jobbers** and bulk liquid warehousers,*** own 33 percent of all facilities
and 28 percent of those handling gasoline.
The majors own the greatest number of bulk terminals within each
gasoline throughput range (Table 6-7). The majors also own a disproportion-
ately greater number of the largest bulk terminals. While the majors own
72 percent of all gasoline terminals, they own 77 percent of the terminals
having a storage capacity between 95,000 and 159,000 M3 (600,000 and 1
million barrels) and 78 percent of the facilities with greater than 159,000
o
M (1 million barrels) of storage capacity but only 60 percent of the
smallest terminals having less than 32,000 M (200,000 barrels). The
independents, which own 28 percent of all gasoline bulk terminals, own
42 percent of the smallest terminals, i.e., total storage less than
32,000 M3 (200,000 barrels), and only 22 percent of the largest facilities,
i.e., storage greater than 95,000 M (600,000 barrels).
*
Includes regional refiner/marketers. Majors are defined as a fully-
integrated company which markets in at least 21 states. A regional
refiner/marketer is a semi-integrated comany with at least one refinery
which generally markets in less than 21 states.
**
A jobber is a petroleum distributor who purchases refined product from
a refiner or terminal operator for the purpose of reselling to retail
outlets, commercial accounts or reselling through his own retail outlets.
***
Bulk liquid warehousers only store products at their facilities for a fee
($/gallon) and do not engage in any marketing activity.
6-12
-------
TABLE 6-6
en
i
BULK TERMINAL OWNERSHIP4
ALL TERMINALS TERMINALS STORING GASOLINE-
OWNERSHIP SEGMENT NUMBER OF TERMINALS % TOTAL NUMBER OF TERMINALS % TOTAL
Majors* 1,170 67% 1,086 72%
Independents 581 33% 425 28%
Total 1,751 100% 1,511 100%
*Includes Regional Refiner/Marketers
Source: U.S. Army Corps of Engineers, Port Series; National Petroleum News, Factbook (1972-1978);
Independent Liquid Terminals Association, 1978 Directory - Bulk Liquid Terminals and
Storage Facilities; Industry contacts; ADL estimates.
-------
TABLE 6-7
en
i
TOTAL STORAGE CAPACITY
000 M3 (000 Bbl)
432(200)
32(200 - 95(600)
96(600) - 159(1,000)
7159(1,000)
% Total
GASOLINE TERMINAL DISTRIBUTION
BY SIZE AND OWNERSHIP 5
MAJORS*
30%
25%
10%
7%
% OF TOTAL TERMINALS
INDEPENDENTS %
21%
3%
3%
2%
STORING
TOTAL
50%
28%
13%
9%
r.Acni TWC
UnoUL inc.
TOTAL NUMBER OF
TERMINALS STORING GASOLINE
764
423
192
132
72%
28%
100%
Total Number of
Gasoline Terminals
1,086
425
1,511
* Includes Regional Refiner/Marketers
Source: Bureau of Census, 1972 Census of Wholesale Trade; U.S. Army Corps of Engineers,
Port Series; National Petroleum News, Factbook Q972-1978); Independent Liquid
Terminals Association, 1978 Directory - Bulk Liquid Terminals and Storage Facilities;
Industry contacts; ADL estimates.'
-------
6.1.1.6 Employment
Employment at all bulk terminals declined from 40,222 in 1972 to
35,700 in 1978, a decrease of 11 percent (Table 6-8). Employment at gaso-
line terminals was estimated to be 30,830 in 1978. PADD I accounts for
55 percent of the total employment at all terminals and 56 percent of the
employment at gasoline facilities. PADD II accounts for 22 percent of
employment at all terminals and 20 percent of the employment at terminals
storing gasoline. Overall, employment is expected to decline as non-
competitive facilities close and more terminals install more automated
equipment in order to reduce labor costs and to increase plant
efficiencies.
6.1.1.7 Future Trends
Recent demand forecasts have indicated that only a modest growth
in gasoline consumption is likely through 1979. Demand is then expected
to level off or even begin to decline in the early 1980's. These forecasts
indicate that no significant increase in additional gasoline storage will
be necessary. No new gasoline terminals are expected to be built in the
near future.
"Stand alone" economics and the rationalization of marketers will continue
to exert closure pressure on marginal facilities. Most bulk terminal closures
have already occurred in the bulk terminal market and only 3 percent or
20 of the smallest gasoline terminals, i.e. gasoline throughput less than
200,000 gallons/day, are expected to close by 1983. These closures
represent less than 1 percent of the 1978 bulk terminal population.
6-15
-------
TABLE 6-8
PADD
I
II
III
IV
v
en
ALL TERMINALS-
EMPLOYMENT
19,280
7,850
4,460
440
3,670
BULK TERMINAL
% TOTAL
55%
.22%
12%
1%
10%
EMPLOYMENT5
-TERMINALS STORING
EMPLOYMENT
17,000
6,280
3,770
440
3,340
GASOLINE-
% TOTAL
56%
20%
12%
1%
11%
Total
35,700
100%
30,830
100%
Source: Bureau of Census, 1972 Census of Wholesale Trade; U.S. Army Corps of Engineers,
Port Series; National Petroleum News, Factbook, (1972-1978); Industry contacts;
ADL estimates.
-------
6.1.2 Bulk Terminal Costs
6.1.2.1 Introduction
Estimates of total installed cost and total annualized costs are
developed for the bulk terminal vapor control systems discussed earlier
in Section 3.2. These systems include refrigeration (RF), compression-
refrigeration-absorption (CRA), and adsorption-absorption (AA) vapor
recovery and incineration by thermal oxidation (OX). Costs are included
for the option of providing both a primary and back-up control system
at terminals.
The cost analysis relies upon the use of model terminals and considers
those costs associated with the control of benzene emissions directly
resulting from the loading of gasoline into tank trucks. Model terminal
sizes analyzed are gasoline loading rates of 950 m /day (250,000 gallons/day)
and 1900 m /city (500,000 gallons/day). Cost estimates are provided for
both existing and new terminals and are based upon a combination of vendor
equipment prices and design information, and installation and operating cost
information supplied by actual terminals that have installed vapor control
systems. Wherever possible model terminal costs estimated by EPA are
compared to actual costs and possible reasons are cited for significant
discrepancies.
Control costs incurred to comply with the proposed NESHAP standard
are calculated assuming no controls are required due to SIP requirements
Terminals located in states requiring vapor recovery (Appendix E) are
expected to incur costs for monitoring and possibly a stand-by control
system.
6-17
-------
A portion of the analysis focuses upon the estimated cost-effective-
ness of the various control systems and options considered. This cost-
effectiveness is determined-in each case by dividing the total annualized
control cost by the estimated annual reduction in emissions.
6.1.2.2 Model Terminal Parameters
The model plant approach utilized in this cost analysis required
that various technical assumptions be made once the average daily gasoline
loading rate was established for a particular model. As mentioned earlier,
3
the two sizes considered were gasoline loading rates of 950 m /day
(250,000 gallons/day) and 1900 m3/day (500,000 gallons/day). Table 6-9
summarizes technical parameters and assumptions which served as bases for
sizing vapor control systems and analyzing costs and cost-effectiveness.
In sizing vapor control systems it appears that two critical design factors
include peak hour and maximum instantaneous loading rates at the terminal.
As evidenced in Table 6-9, these design\factors are hot directly linked
to the average daily loading rates. This point should be kept in mind
later when comparing capital costs for vapor control systems at various
daily loading rates.
6.1.2.3 Bases for Capital and Annualized Cost Estimates
The installed capital cost estimates are intended to represent the
total investment required to purchase and install a particular control
system. All capital costs are intended to reflect first quarter 1978
dollars. Purchase costs for the control systems considered were obtained
from vendors for the design factors provided in Table 6-9. Total
installed costs were developed from major equipment purchase costs by
6-18
-------
Table 6-9.
MODEL TERMINAL PARAMETERS
cr>
t
Average Daily Loading Rate:
m /day
gallons/day
DESIGN FACTORS
(a) Number of rack positions
(b) Number of loading arms per position
(c) Method of loading
(d) Pumps (each)
(e) Tank truck capacities
(f) Tank truck loading time (total)
(g) Peak hour loading
(e)_+Jf)._x_6Q_xJa.L
(h) Maximum instantaneous loading
__ U) x_(b) x (d)
EMISSION FACTORS
Uncontrolled:
Total hydrocarbon
Benzene
Controlled3:
Total hydrocarbon
Benzene (95% reduction)
TERMINAL OPERATING SCHEDULE
950
250,000
2
3
Submerged (top or
bottom)
1.9 m3/min
(500'gpmy
30 m3
(S.OOlTgaTTonsJ
20 minutes/truck
18p_jn?/hr.
(48.000 gph)
Y\ rrr/min
{3", 000" gpm)
960 mg/liter
8 mg/liter
80 mg/liter
0.4 mg/liter
300 days/year
1,900
500,000
4
3
Submerged (top or bottom)
1.9 m3/min (500 gpm)
3.P_nr_..(8tOpq gallons)
20 minutes/truck
.360 m3/hr_ (96,000 gph)
22 m3/min (6,000 gpm)
960 mg/liter
8 mg/liter
80 mg/liter
0.4 mg/liter
300 days/year)
aAissumes 100 percent vapor collection at rack during loading and no losses in vapor collection system,
-------
estimating total system installation costs as a percentage of purchase
cost. A factor of 100 percent was used for retrofitted terminals while a
factor of 70 percent was considered for new termirials. These total
installation costs are intended to include sales tax, freight, engineering,
unit installation, ancillary equipment and piping and contingencies.
Installation cost factors were estimated on the basis of actual installed
cost information available to EPA. In most cases these actual installations
converted existing top loading racks to bottom loading prior to or in
conjunction with hydrocarbon control system installation. EPA model
terminal costs do not include this conversion for top loading terminals
since EPA feels that this modification is more directly related to
operational and personnel safety considerations. Capital costs, however,
should adequately reflect higher vapor collection costs for these
terminals.
Capital costs for monitoring are not included in the model terminal
costs developed. It is estimated that a gas chromatograph monitoring
system would cost about $20,000 installed.
Annualized cost estimates include utilities, operating labor,
maintenance labor and materials, credits for gasoline recovery, and capital
charges for interest, depreciation, administrative overhead, property
taxes and insurance. Table 6-10 summarizes all annualized cost factors
used in this analysis. All annualized costs are intended to reflect
current estimates.
6-20
-------
Table 6-10.
COST FACTORS USED IN DEVELOPING ANNUALIZED
COST ESTIMATES FOR MODEL TERMINALS
Utilities:
- Electricity
- Propane
Operating Labor
Maintenance (percent of equipment cost)
- RF vapor recovery
- CRA vapor recovery
- AA vapor recovery
- Oxidizer
Capital charges (percent of capital cost):
- Interest and depreciation, plus
- Property taxes, insurance and
administrative overhead
Gasoline value (recovered) FOB
terminal before tax:
Carbon for AA unit (replacement cost)
Reference 7
Reference 7
cBased upon actual maintenance costs reported to EPA
Assumed to be comparable to CRA
Reference 8
Calculated using capital recovery factor formula assuming 10 year
equipment life and 10 percent interest rate.
90il Daily - March 1978.
$.017/105 joules ($.06/Kw-hr)a
$.10/liter ($.40/gallon)
$10/man-hour
percent
percent1*
4 percent (carbon replacement
additional)
percent6
16 percent
4 percent
$.10/liter ($.40/gallon)9
$21/Kg ($.90/lb)
6-21
-------
6.1.2.4 Cost Estimates for Emission Controls at Model Existing Terminals
6.1.2.4.1 Single System
As discussed earlier,~control options conside'red in Chapter Four for
the gasoline marketing network will require the use of a vapor control
system at terminals during tank truck loading operations. Table 6-11
presents estimates of capital and annualized costs for the individual
control systems at two daily loading throughputs.
Regarding capital costs presented in Table 6-11 it appears that
when compared to vapor recovery systems (RF, CRA, AA) incineration
control systems (OX) require the lowest capital investment. For the recovery
systems considered refrigeration (RF) capital costs appear to be the lowest.
The CRA system cost includes a vapor holder sized to accomodate peak
hour loading considering the design vapor flow rate of the CRA unit.
Costs can vary slightly depending on the vapor holder and CRA unit size
combination selected. The AA system costs are recent vendor budget
quotes where installation costs were estimated to be comparable to other
vapor recovery systems.
Reviewing annualized costs in Table 6-11 indicates that utilities !
and maintenance costs appear to be the significant operating costs for
all control systems analyzed. Overall, operating costs appear lowest
for incineration (OX) and highest for recovery systems if gasoline
recovery credits are not included. The capital charges included
6-22
-------
Table.6-11._ ESTIMATED CONTROL COSTS FOR MODEL EXISTING TERMINALS
*Single Vapor Control System Alternative*
Gasoline Loaded:
CT1
CJ
950 nT/day
(250.000 gallons/day)
1900 rrr/day
(500.000 gallons/day)
Vapor Control System:
Investment ($000)
Purchase Cost (FOB factory}3
Total Installed Cost
Annualized Cost(credit)($000/yr^
Electricity0
Propane(pilot)
Maintenance
Operating labor6
Carbon Replacement
Subtotal (Direct operating costs)
Capital Charges
Gasoline Recovery(credit)^
Net Annualized Cost(credit)
AA
120
240
3.9
—
4.8
1.5
2.4
12.6
48.0
(39.2)
21.4
CRA
128b
256
5.1
—
' 5.1
1.5
__.
11.7
51.2
139.2)
23.7
OX
72
144
2.9
1.0
2.9
1.5
__
8.3
28.8
**_**_
37.1
RF
102
204
9.9
—
6.1
1.5
__
17.5
40.8
(39.2)
19.1
AA
155
310
7.8
—
6.2
1.5
4.7
20.2
62.0
(78 A^
3.8
CRA
164b
328
8.3
—
6.6
1.5
16.4
65.6
(7BA)
3.6
OX
95
190
5.8
1.0
3.8
1.5
12.1
38.0
.. ^^__._
50.1
RF
, 153
306
19.8
—
9.2
1.5
__
30.5
61.2
(78.4J
13.3
aVendor quotes (see references 9, 10, 11, 12)
Includes vapor holder
CA11 systems except CRA calculated at 12 hours/day of vendor estimated nominal Kw draw - CRA hours based upon design flow rate.
Estimated at .72 gal/hour operation (Reference 11)
Inspections at .5 man-hr/day.
Estimated based upon three year carbon life (Reference 9)
Calculated at 16°C (60°F) and 100% vapor collection at rack.
-------
in annualized costs for the control systems analyzed have been defined
earlier in Table 6-10. These charges appear to be three to four times
greater than average operating costs for the control systems. Hence, their
impact on annualized costs is significant. Equally significant, however,
appears to be the effect of gasoline credits on the net annualized cost
of vapor recovery units and the relative impact these recovery credits
have when comparing net annualized costs for vapor recovery and incineration
systems. Based upon the technical assumptions used to estimate gasoline
recovery credits, i.e., 100 percent vapor collection at the rack and tank
truck compartments saturated with hydrocarbon vapors prior to loading,
gasoline recoveries appear to be substantial. However, these annual
recovery credit estimates have not been supported by actual terminal data
submitted to EPA. Actual recovery information is provided later in the
discussion of actual cost information.
6.1.2.4.2 Back-up Controls
An additional consideration analyzed for terminals is the requirement that
no tank truck loading be performed unless control units are operating
continuously and effectively. In the event the primary control system
could not provide the required level of control the terminal would have
to either switch to a back-up system or cease loading operations.
For purposes of cost analysis three alternatives were considered: (1)
Back-up vapor control system, (2) Vapor storage capacity (minimum five day
vapor generation capacity), (3) Shutdown of loading operations during
6-24
-------
Table 6-12. ESTIMATED CONTROL COSTS FOR MODEL EXISTING TERMINALS
*Stand-by Control System Alternative*
Gasoline Loaded:
no
ui
950 nT/day
(250.000 gallons/day)
1900 mj/day
(500,000 gallons/day)
Stand-by (OX)a
Total Installed Capital Cost ($000) 95
Direct Operating Costs ($000/yr):
Utilities Footnote b
Maintenance and Labor and materials 2.9
Capital Charges ($000/yr) 19.0
Gasoline (credit )($000/yr)
Net Annual ized Cost(credit)($000/yr) 21.9
(Primary/Stand-by)
(RF/OX) (OX/OX)
299 239
9.9 3.9
10.5 7.3
59.8 47.8
(37.2)c
43 59.0
(Primary/Stand-by)
Stand-by (OX)a (AA/OX)
126 436
Footnote b 17.8
3.8 16.2
25.2 87.2
(74.5)c
29.0 36.7
(OX/OX)
316
6.8
9.1
63.2
__
79.1
V
Stand-by system costs are shown separately for those terminals that have already installed vapor controls to comply with existing SIP
requirements for hydrocarbons.
These will vary but should not significantly effect net operating costs of primary/stand-by combination.
cRecovery reductions will vary but are estimated at 5 percent or 15 days down time on primary system.
-------
malfunction. Costs were estimated for Alternatives (1) and (2) at the
two model gasoline loading throughputs. Developing costs for Alternative
(3) above was considered beyond the scope of this analysis although it
may be a viable option for the terminal.
Since a back-up control system at terminals (Alternative (1) ) would
hopefully see minimal service, additional gasoline recovery potential is
expected to be negligible. Based upon Table 6-11 costs it appears that a
stand-by incinerator would represent the lowest capital investment and
operating costs. Additionally, because the oxidizer could be linked to
the primary system at the knockout tank for the latter, total installed
costs would be slightly lower than those included in Table 6-11.
Costs are presented separately in Table 6-12 for a stand-by incineration
system. This is an approximate cost incurred by terminals that presently
operate control systems because of SIP requirements for gasoline loading.
This incremental cost is then added to primary system costs for RF, AA
and OX systems to depict an expected range of cost impacts resulting
from this dual-system alternative.
Although costs were analyzed for Alternative (2) i.e., vapor storage
capacity for five day vapor generation at the loading throughput rate, these
costs appear extremely high when compared to the stand-by OX system costs.
3
As an example, five-day vapor storage capacity for the 950 m /day terminal
is estimated to cost $241,000 installed with annualized costs of approximately
3
$45,000. This is based upon installation of a new vapor holder with 4730 m
(170,000 ft ) vapor storage capacity. Vapor holders sized to handle reduced
6-26
-------
loading rates during primary system malfunction may be a more viable
alternative from a cost standpoint at terminals.
6.1.2.4.3 . Cost-Effectiveness
Based upon the annualized control costs presented in Tables 6-11
and 6-12 and estimates of annual reductions in total hydrocarbon and
benzene emissions, cost-effectiveness (C/E) is summarized in Table 6-13
for both the single and dual systems analyzed. The emission reduction
estimates were developed from parameters included in Table 6-9 and
reflect control levels that EPA considers attainable by all control systems
considered.
Cost-effectiveness estimates appear to indicate that recovery is more
cost-effective than incineration for the single system. This is basically
a result of the substantial gasoline recovery credits estimated for the
vapor recovery units.
For terminals installing both a primary and back-up control system, a
recovery unit and oxidizer unit, respectively, appears to be the most
cost-effective combination. This may change as the annual operating
time for the back-up oxidizer unit increases. Finally, when compared
to a single control unit, installing dual systems will generally double
costs per kilogram of benzene controlled. This assumes that stand-by
units are sized to handle the same loading rate as primary systems.
6.1.2.4.4 Comparison to Actual Costs
As mentioned earlier in Section 3.2, as a result of SIP hydrocarbon
controls, an estimated 300!control systems are presently installed at
6-27
-------
Table 6-13. COST EFFECTIVENESS FOR MODEL EXISTING BATTERIES
Gasoline Loaded:
950 m /day 1900 m /aay
(250,000 gallons/day) (500,000 gallons/day)
Dual System Dual •Systegt- — -
Single System (Primary/Standby} . .. S.innle .System.. .-....- (Primary/Standby)
(AA) (CRA) (OX) (RF) (RF/OX) (OX/OX) (AA) (CRA) (OX) (RF) (AA/OX) (OX/OX)
Net Annual ized Cost (credit)
Total Hydrocarbon Control leda'b>c
(Mg/yr)d
Benzene Control leda>c
(103 Kg/yr)'
Cost-Effectiveness HC
($/Mg)
Cost-Effectiveness Benzene
($/Kg)
Assuming control equipment always
<^> Equivalent to 92 percent reduction
££ Equivalent to 95 percent reduction
21,400 23
250
2
90
11
,700 '37,100 19,100 43,000 59,000 3,800 3,600 50,100 13,300 36,700 79,100
250 250 250 250 250 500 500 500 500 . 500 500
222 22 44444 4
90 150 80 170 240 10 10 100 30 70 160
12 19 10 22 30 , 1 1 13 3 >y 20
;
operating during gasoline loading
in hydrocarbon emissions.
in benzene emissions
Mg'= megagram = 1000 Kg
-------
bulk terminals to control loading emissions from tank trucks. Costs
associated with the installation, operation and maintenance of RF, CRA
and OX units were providedlby several terminal operators. Since many
installations were completed as much as four or five years ago, equipment
costs were escalated to current estimates by using the Chemical Engineering
Index for Fabricated Equipment. Reported costs and escalated estimates
are summarized in Table 6-14. EPA and actual capital costs are then
compared graphically in Figure 6-2.
When comparing actual capital costs to the model estimates, it is
important to consider such factors as the number of racks, vapor storage
capabilities (where applicable) and control unit design rate. An additional
consideration for actual terminals is that almost all control system
installations were done in conjunction with conversions at racks from
top to bottom loading. This work is not always^done concurrently. Hence,
it appears that some vapor piping and installation costs closely linked
with the rack conversion work could not be broken out by actual terminals.
Taking all the above into consideration, it is felt that the installed
capital costs for model and actual facilities compare reasonably well.
Although actual operating cost information was reported by terminals
operating control systems, these costs are reported for hydrocarbon concen-
trations to the control unit that are generally lower than those assumed for
model plants. Additionally, unit costs for utilities were not provided
with monthly estimates. For these reasons direct comparisons of actual
and model annual operating costs and gasoline recovery credits appear
-.6-29
-------
Table 6-14.
SUMMARY OF HYDROCARBON CONTROL SYSTEM COSTS FOR ACTUAL TERMINALS
Average
Loading
m3/day
450
630
1010
490
600
1930
NA
NA
Y1 NA
OJ
o
1230
1700
Gasoline
Throughput
lp3 gpd
119
167
267
129
159
510
—
. —
325
450
No. of Rack
Positions
2
2
t
3 '
2;
1
3
NA
NA
NA
4
3
Control
Unit
OX
OX .
OX
RF
RF
RF
RF
RF
RF
CRA
CRA
Design Rate
300 cfma
250 cfma
- 800 cfma
250,000 gpd
250,000 gpd
800,000 gpd
800,000 gpd
800,000 gpd
800,000 gpd
'. 160 cfm
. 225 cfm
Installation
Date
6/75
5/75
3/75
7/75
12/74
4/76
1/74
5/74
5/74
10/75
12/72
Purchase
Cost (PC)
($000)
$5
65
NA
NA
45b
128
107
119
119
115C
115C
1
Total Inst
($000)
95
99
200.
116
112
250
173
190
164
165
195
ailed Cost
% of PC
173
170
NA
NA
248
195
161
160
137
143
170
Escalated Installed Cost
. (1st Quarter 1978)
($000)
115
119
240
140
150
287
280
285
250
215
317
NA - Not available
For comparison to design bases in Table 6.1.2-1, an approximate conversion from cfm to gpm gasoline loaded is 7.5 gal/cf vapor
(no vapor growth)
Prototype unit
cCost includes vapor holder
-------
Figure 6-2. Comparison of Installed Capital Costs
for Benzene Control Systems
400
o
o
300
"O
O)
200
EPA estimates
CRA (actual)
RF (actual)
Ox (actual
efrigeration
100
1000
m /day gasoline loaded
Note: 1 m3 = 1000 liters = 264 gallons
2000
3000
6-31
-------
inconclusive. Nevertheless, wherever possible actual operating cost
information was factored into the model estimates. This is reflected in
several of the maintenance-factors, operating labor requirements and
electrical requirements used in calculating the respective costs associated
with these operating cost factors.
Gasoline recoveries reported by terminals were significantly lower than
those estimated for model terminals. Specifically, recoveries ranges from
about 0.03 percent to 0.07 percent of the volume of gasoline loaded for
actual facilities while EPA estimates correspond to a 0.13 percent recovery.
Low recoveries at actual terminals may be attributable to: (a) loading
into "vapor lean" tank trucks that return from areas not requiring Stage I
vapor recovery; (b) inefficient vapor collection or control unit operation;
(c) the affect of seasonal influences on loading emissions and gasoline
recoveries; (d) some combination of these factors.
6.1.2.5 Cost Estimates for Emission Controls at Model New Terminals
Installed capital and annualized costs for the control systems and
options considered for model existing terminals have been provided in the
preceeding section. For the purpose of estimating costs for new terminals
it has been assumed that installation costs will be slightly lower than
retrofit situations. This is based upon information indicating that
vapor piping installations at existing rack positions often require
concrete and structural modifications that could more economically be
included in the design of the rack during initial installation time.-^
Other physical constraints at existing terminals often affect vapor piping
6-32
-------
Table 6-15. ESTIMATED CONTROL COSTS AND COST EFFECTIVENESS FOR MODEL NEW TERMINALS
cr>
i
co
CO
Gasoline Loaded:
Control Unit(s)
Total Installed Cost ($000)
Direct Operating Costs ($000/yr)
Capital Charges ($000/yr)
Gasoline Recovery(credit)($000/yr)
Net Annualized Cost(credit)($000/yr)
Total Hydrocarbon Controlled (Mg/yr)
Benzene Controlled (Mg/yr)
• Cost-Effectiveness HC ($/Mg)
Cost-Effectivensss Benzene ($/Mg)
950 m°/day
(250.000 gallons/day)
1900 nT/day
(500.000 gallons/day)
Single
AA
204
12.6
40.8
(39.2)
14.2
250
2
60
CRA
218
OX.
122
i
11.7 8.3
43.5 24.5
(39.2) --
16.0
250
2
.
.
32.8
'
250
60 130
RF
173
17.5
34.7
(39.2)
13.0
250
2
50
Back-up
OX
88
2.9
17.6
—
20.5
*
*
*
Dual
RF/OX
261
20.4
52.3
(37.2)
35.5
250
2
140
OX/OX
210
11.2
42.1
—
53.3
250
2
210
AA
264
20.2
52.8
(78.4)
(5.4)
500
4
(10)
Single
CRA OX
279 161
16.4 12.1
55.8 32.2
(78.4) --
(6.2) 44.3
500 500
4 4
(10) 90
, Back-up
I
, RF OX
260 114
30.5 3.8
52.0 22.8
(78.4) --
4.1 26.6
500 *
4 *
10 *
Dual
. AA/OX
378
24.0
75.6
^•5)
25.1
500
4
50
OX/OX
275
15.9
55.0
—
70.9.
500
4
140
7,100 8,00016,400 6,500 * .. 17,700 26,600 (1350) (1550)11,1001,000* 6,300
17,700
-------
runs and other ancillary costs. Net annualized .costs for new terminals
are assumed to be impacted only by the slightly lower capital charges
associated with lower investments for control system installations.
Considering the foregoing, model costs for new terminals will exhibit
cost and cost-effectiveness results that are relatively consistent with
those presented earlier for existing terminals. For this reason costs
are summarized only in Table 6-15 for new terminals and the reader is
advised to refer back to the discussions of existing cost tables for
detailed coverage of cost considerations and analyses.
6.1.2.6 Cost Estimates for Tank Truck Vapor Recovery
As discussed earlier in this section, costs for converting
terminals to bottom loading are not assumed to be attributable to the
proposed EPA regulations. Hence, only those costs associated with
providing vapor recovery equipment on tank trucks are considered
here. For existing four compartment transports, retrofit capital
costs are estimated at $21,000/transport and total annualized costs
(including maintenance) are estimated at $780/year. A slight savings
(less than 10 percent) is projected when installing vapor recovery on
new transports. These estimates were developed from information reported
by Reference 7.
6-34
-------
6.1.3 Bulk Terminal Impacts
6.1.3.1 Introduction
The principal economic impacts of the proposed vapor control strategy
which would reduce the amount of benzene emitted into the atmosphere by
bulk terminals are:
a the number of potential bulk terminal closures;*
• the employment displaced by these closures;*
• the total cost of installing vapor control at terminals;
• the total cost of installing vapor control on gasoline
tank trailers.
Most terminal operators own gasoline tank trailers, however, a significant
number of trailers are also owned by common carrier. Because the cost of
converting the trailer fleet to vapor control will not be borne entirely
by the bulk terminal industry, its cost will be treated separately.
6.1.3.2 Closure Methodology
Bulk terminals may be forced to close due to vapor control economics
because of either of the following reasons:
• Terminals operators are unable to obtain the capital
necessary to install vapor control equipment.
• Terminals would fail to achieve a sufficient or acceptible
level of profitability if vapor control were installed.
*
The monetary costs of these impacts have not been calculated.
6-35
-------
Terminals having no gasoline throughput would be exempt from proposed
vapor control program and hence, would not be subject to possible closure.
Similarly, the gasoline terminals which are expected to close anyway within
the next five years due to competitive economics or market rationalization
are not included with the closures caused by vapor control.
The control technologies which will be analyzed in determining bulk
terminal impacts are a refrigeration system with an incineration stand-by
unit (refrigeration/incineration) and an incineration system with an .
incineration stand-by unit (incineration/incineration). These two technologies
were selected from the various control systems described in Section 6.1.2
because they have the least capital requirement of any of the model systems
and they are also the most common technologies currently used by the bulk
terminal industry. Stand-by units are required to assure the continuous
and efficient control of hydrocarbon vapors during gasoline handling
operations. The incineration stand-by unit has the least capital cost of
any of the back-up systems evaluated.
Because it would be impossible to determine the number of terminals
which would close due to vapor control by examining the entire terminal
population on an individual basis, several bulk terminal prototypes were
developed to facilitate this analysis. These prototypes, taken collectively,
are representative of the bulk terminal industry. Changes in the operational
and financial profiles of these facilities which are caused by either of the
above two control systems will be translated into potential closures in the
actual terminal population. Separate marine and pipeline prototypes were
developed for this analysis since these are the two primary modes of product
receipt at bulk terminals. For both the marine and pipeline terminals, a
6-36
-------
large and small prototype was created having the same gasoline throughput
as the model vapor control systems discussed in Section 6.1.2, i.e. 950
and 1900 M4/day (250,000 and 500,000 gallons/day).
The incremental cost of vapor control will impact bulk terminals to
varying degrees depending upon the terminal's ability to pass through the
costs of vapor control. The most efficient terminals will be able to pass
through the full cost of vapor control to their customers in the form of
tariff increases. The less efficient terminals, however, will be limited
to only passing through, at most, the same unit cost as the more efficient
facilities or the market price-setters. Because pipeline terminals are
the most financially attractive type of bulk storage, both large and small
facilities are assumed to be able to pass through the full cost of vapor
control. In the case of marine terminals, which handle the marginal barrel
of product, full pass through is limited to the large terminals, while the
smaller marine terminals are limited to the same unit cost as the larger
facilities.
6.1.3.2.1 Availability of Capital
While over two-thirds of the bulk terminals are owned by major oil
companies and regional refiners, who have very good access to capital,
each terminal was considered as a separate profit center in order to
determine its ability to secure the necessary funds for vapor control
equipment. Since no financial assistance was available from a parent
corporation or from ancillary marketing operations, the necessary capital
would probably be obtained from a commercial lender. A commercial lender
is most interested in the terminal's ability to repay the full amount
of the loan, i.e. principal as well as interest. If the terminal operator
can demonstrate satisfactorily that the loan can be comfortably repaid
6-37
-------
while still meeting all other current liabilities, e.g., salaries, operating
expenses, other loan payments, the capital will most probably be made
available. If, however, the proposed loan strains the terminal's debt
capacity and hence, jeopordizes the terminal's ability to repay the entire
obligation, the capital may not be available. Such a decision would depend
upon the lender's perception of the risks and his risk threshold. Cash
flows were projected for the bulk terminal prototypes assuming that the
vapor control was installed. Based on the cash flow available to service
the incremental debt obligation and operating expense, in addition to all
pre-control expenses, potential closures in the bulk terminal population
were calculated.
6.1.3.2.2 Insufficient Profitability
Bulk terminals unable to pass through the full cost of vapor control
would be forced to absorb all remaining control costs. This could cause
some facilities, which were just breaking even or marginally profitable
in a pre-control case, to now operate at a loss. For each of the bulk
terminal prototypes, the gasoline throughput necessary to meet all current
liabilities was calculated. Vapor control costs, i.e. operating expense
and debt obligations, would increase this breakeven throughput in facilities
unable to pass through the full costs of vapor control. Using the increase
in breakeven throughput caused by vapor control costs, the number of terminals
which once operated above the pre-control breakeven throughput but which
now operate below the adjusted breakeven throughput was calculated.
6-38
-------
6.1.3.3 Bulk Terminal Closures
An estimated 45-50 bulk terminals are expected to close if refrigeration/
incineration or incineration/incineration systems are installed at all
gasoline terminals (Table 6-16). Approximately 46 closures would occur
due to the cost of an incineration/incineration system while 61 closures are
expected because of the cost of a refrigeration/incineration system. No
closures are likely due to an inability to obtain the necessary capital;
all are expected to be the results of terminals failing to achieve a
sufficient or acceptable profitability.
Using the less expensive incineration/incineration system, approxi-
mately 30 of the terminal closures are expected at facilities owned by
majors or regional refiners while the remaining 15 will be at independents'
facilities. All of these closures are expected to be small marine terminals
that are unable to pass through the full cost of vapor control. The impact
of these closures upon the U.S. gasoline marketing network would be minimal
as each of these terminals has an average gasoline throughput which is less
than 150,000 gallons/day.
6.1.3.4 Employment Displaced by Terminal Closures
Between 640 and 710 workers are employed at the terminals which are
assumed to close due to vapor control (Table 6-17). These workers represent
approximately 2 percent of all bulk terminal employees, excluding drivers.
Two-thirds of the impacted work force or 430 workers, are employed at
terminals owned by majors and regional refiners. The remaining 210 workers
are employed at independents' facilities.
6-39
-------
TABLE 6-16
16
BULK TERMINAL CLOSURES DUE TO VAPOR CONTROL ECONOMICS
REFRIGERATION/ INCINERATION/
INCINERATION INCINERATION
Terminal Population Subject
to Vapor Control 1131 1131
Terminal Closures Due to
Inaccessibility of Capital 0 0
cr>
i
o Terminal Closures Due to
Insufficient Profitability 46 61
Terminals Installing Vapor Control 1085 1080
-------
TABLE 6-17
VAPOR CONTROL EMPLOYMENT AND COSTS
IMPACTS AT BULK TERMINALS ]'
REFRIGERATION/ INCINERATION/
INCINERATION INCINERATION
Terminals Closed Due to Vapor
Control Economics 46 51
Estimated Employment at
Closed Terminal 640 710
Terminals Installing
Vapor Control 1085 1080
Total Vapor Control Cost
(Million 1978 Dollars) 473.2 580.4
-------
6.1.3.5 Vapor Control Costs - Bulk Terminals
The total cost of installing vapor control at all gasoline bulk
terminals is $580.4 million using the cost of a refrigeration/incineration
system and $473.2 million using the cost of an incineration/incineration
system. These figures include the cost of installing incineration stand-
by units at all terminals which presently have a primary vapor control
system. The total figure is the sum of the capital charges, financing costs
and operating expenses less any applicable recovery credits over the 10
year life of the control equipment (Table 6-18). All costs are expressed
in constant 1978 dollars; future cash streams have been discounted to present
value using a discount rate of 10%.
Majors and regional refiners will bear most of the dollar cost of
vapor control. Using the total cost of the incineration/incineration system which
has the smaller capital requirement of the two model control systems
evaluated, the cost of vapor control to the majors is calculated to
be $369.1 million or 78 percent of the total cost of $473.2 million.
Independents' are expected to bear the remaining $104.1 million cost, or
28 percent of the total.
6.1.3.6 Vapor Control Costs - Tank Trailer Fleet
There are an estimated 24,800 gasoline tank trailers in operation
today, of which 7,400 or 30% are already equipped with vapor control
(Table 6-19). Of the remaining trailers, 7300 are expected to be
retrofitted while the remaining 10,000 would have vapor control installed
as they are replaced. Because gasoline demand is not expected to increase
significantly during the next 5 years, no new tank trailers are expected
to be built other than those needed to replace existing trailers. Based
6-42
-------
TABLE 6-18
VAPOR CONTROL COSTS* AT BULK TERMINALS18
(Million 1978 Dollars)
REFRIGERATION/ INCINERATION/
INCINERATION INCINERATION
Capital Investment 364.3 280.5
Financing (8 years) 110.8 85.3
Operating Expense (10 years) 250.0 151.8
Recovery Credit (10 years) (314.7)
Capital Investment** 37.0 37,0
Financing (8 years) 9,8 9.8
Operating Expense (10 years) 16,0 16.0
Total Vapor Control Cost 473,2 580.4
* Future cash streams discounted to present value. Discount rate = 10%.
** Cost of incineration stand-by unit for terminals with existing vapor control.
-------
TABLE 6-19
GASOLINE TRAILER POPULATION 19
Estimated 1978 Gasoline Tank Trailer Fleet 24,760
Gasoline Trailers Vapor Control ( 7,440)
Trailers to be Replaced over Next 5 Years* (10,060)
Retrofit Market 7.260
New Trailers to Replace Existing Fleet 10,060
New Trailers Necessary Due to Increased
Gasoline Demand 0
New Trailer Market 10,060
Total Number of Trailers Installing
Vapor Control 17,320
* Estimated trailer lifetime of 12.3 years.
6-44
-------
on the above population estimate and the capital requirement and operating
expens.e described in Section 6.1.2, the total cost of installing vapor
control on the tank trailer fleet is $79.5 million (Table 6-20). This
cost includes the capital requirement, financing costs and operating
expenses expressed in constant 1978 dollars.
6-45
-------
TABLE 6-20
TOTAL COST TO INSTALL VAPOR CONTROL ON
THE GASOLINE TRAILER FLEET20
(Million 1978 Dollars*)
Capital Cost (Retrofit Market) 15.2
Capital Cost (New Market) 19.1
Financing** (3 years) 3.9
Operating Expense (12 years)*** 41.3
TOTAL CONVERSION COST 79.5
* Future cash streams discounted to present value.
Discount rate = 10%.
** 100% debt financing for 3 years @ 9%.
*** Estimated trailer lifetime of 12.3 years.
6-46
-------
6.2 BULK PLANTS
6.2.1 Bulk Plant Industry Characterization
6.2.1.1 Introduction
Bulk plants are secondary storage facilities which operate as
satellite distribution centers receiving petroleum products from primary
terminals. Most bulk plants receive product from primary terminals via
truck transport. These vehicles deliver 30-34M (8,000-9,000 gallons) of
product and are usually owned by terminal operators or by common carriers.
Some bulk plants receive product by rail and a few are supplied by small
tanker, barge or pipeline. Bulk plants supplied by rail are most common
in the Rocky Mountains and along the West Coast. Delivery by barge is
most common on the East and West Coasts and along the Mississippi River.
6.2.1.2 Operations and Market Environment
Because bulk plants service agricultural, commercial and residential
accounts as well as retail gasoline outlets, most facilities store a
variety of products, e.g., kerosene, gasoline, diesel and distillate. In
the Northeast, however, bulk plants tend to specialize in either gasoline
or distillate sales.
Bulk plants distribute petroleum products to accounts requiring small
and infrequent deliveries; however, bulk plant operators may also supply
a number of high volume accounts. Products are delivered by truck trans-
port to high volume customers directly from primary terminals, thus by-
passing storage at the bulk plant. Smaller tank wagons, 8-16 M (2,000-
4,000 gallons), are used if customers do not have sufficient storage to
permit transport deliveries or if roads are impossible to transport traffic.
6-47
-------
Approximately two-thirds of all petroleum products sold by bulk plant
operators are delivered by tank wagons which are usually owned by the bulk
plant operator.
Bulk plants are also subject to the same "stand alone" economics
under which bulk terminals now operate. A substantial number of plants
have already closed because of their marginal profitability. More closures
are expected in the future; however, some rural and semi-rural bulk plants
are more secure than urban facilities because they are partially shielded
from competitive market forces by transportation economics.
6.2.1.3 Bulk Plant Population
There are presently 18,640 bulk plants of which 17,850, or 96 percent,
store gasoline (Table 6-2l). The total number of bulk plants has declined
20 percent from the 23,370 facilities reported in 1972. Because bulk plants
have been subject to the same "stand alone" economics as terminals, shrinking
margins and increasing operating costs have forced the less efficient
facilities to close. Furthermore, the withdrawal of the major oil companies
from bulk plant operations has removed the financial subsidy required by
many marginal facilities. Almost half of all bulk plants and half of the
gasoline bulk plants are located in PADD II where distribution logistics
and a large concentration of rural accounts warrant secondary storage.
3
Total storage of bulk plants was estimated to be 6.8 million M
(1.8 billion gallons) in 1978. Storage capacity has been declining due to
the number of facilities going out of business. Gasoline storage is
estimated to be 4.0 million M (1.1 billion gallons), or 60 percent of
the total storage. Gasoline storage has also been declining because of the
number of bulk plant closures and because an increasing portion of national
6-48
-------
TABLE 6.21
UD
21
BULK PLANT POPULATION
PADD
fll 1 Dill I/ m AMTC
PIM v r>i nMTf crnoTMr rncni IMF
Number of % Total Storage % Number of % Total Storage
Bulk Plants Total Capacity Total Bulk Plants Total Capacity
000 M3 (000 Gal)
I
II
III
IV
V
Total
3,510 19% 1,641 (
8,850 47% 2,691 (
3,320 18% 958 (
990 5% 323 (
1,970 11% 1,144 (
18,640 100% 6,757 (1
433,290) 24%
710,670) 40%
253,380) 14%
85,490) 5%
302,270) 17%
,785,100) 100%
source: Bureau of Census, 1972 Census of Wholesale Trade; National
000 M3 (000 Gal)
3,190 18% 947 ( 250,270)
8,540 48% 1,521 ( 401,830)
3,320 19% 709 ( 187,190)
990 5% 221 ( 58,490)
1,810 10% 623 ( 164,600)
17,850 100% 4,021 (1,062,380)
%
Total
24%
38%
18%
5%
15%
100%
Oil Jobbers Council; National Petroleum News,
Factbook, (1972-1978); Industry contacts; ADL estimates.
-------
gasoline throughput is bypassing storage at the bulk plant and is being
delivered directly to end-users, Over 60 percent of the gasoline storage
capacity at bulk plants is located in PADD's I and II.
6.2.1.4 Size Distribution
3
The average storage capacity of bulk plants is approximately 300 M
(80,000 gallons). Almost 80 percent of all bulk plants and all gasoline
3
bulk plants have total storage between 151 and 568 M (40,000 and 150,000
3
gallons); approximately 13 percent have storage capacities less than 151 M
3
(40,000 gallons); and 8 percent have capacities greater than 568 M
(150,000 gallons) (Table 6-22).
Almost 60 percent of all bulk plant have a total product throughput
between 11 and 30 M3/day (3,000 and 8,000 gallon/day) (Table 6-23).
q
Almost 25 percent have a throughput less than 11 M /day (3,000 gallon/day)
3
and 18 percent have a throughput greater than 30 M /day (8,000 gallon/day).
Similarly, 63 percent of all gasoline bulk plants have a daily gasoline
throughput between 11 and 30 M3/day (3,000 and 8,000 gallon/day); 29 percent
3
have a gasoline throughput less than 11 M /day (3,000 gallon/day); and 8
J 3
percent have a throughput greater than 30 M /day (8,000 gallon/day).
6.2.1.5 Ownership
Jobbers own the greatest number of bulk plants. Jobbers own 74 percent
of all bulk plants and 76 percent of the gasoline bulk plants (Table 6-24).
The majors' share is 22 percent and 20 percent, respectively, while
independent wholesale/marketers own approximately 4 percent of each group.
The jobbers' share of the market has been increasing steadily in recent
years as the majors have been pulling out of secondary storage operations
as part of an overall marketing strategy.
6-50
-------
TABLE 6-22
22
BULK PLANT STORAGE DISTRIBUTION
Al 1 DIM I/ m AMTC
cr>
en
TOTAL STORAGE CAPACITY
M3 (000 Gal)
<150(40)
150(40) - 568(150)
568(150) - 1,136(300)
r 1,136(300)
NUMBER OF BULK PLANTS % TOTAL
2,380
14,800
1,180
280
13%
79%
6%
2%
BULK PLANTS STORING (
NUMBER OF BULK PLANT!
2,380
14,100
1,100
260
5ASOLINE-
5 % TOTAL
13%
79%
6%
2%
Total
18,640
100%
17,850
100%
Source: Bureau of Census, 1972 Census of Wholesale Trade; National Oil Jobbers Council;
National Petroleum News, Factbook (1972-1978); Industry contacts; ADL estimates.
-------
cn
TABLE 6-23
Al 1 pill I/
MLL DULI\
AVERAGE
PRODUCT THROUGHPUT
M3/Day (000 Gal /Day)
£11(3)
11(3) - 30(8)
30(8) - 65(17)
765(17)
ni ftMTC
1 LHIN 1 o •
NUMBER OF
4,400
10,760
2,650
830
BULK PLANT THROUGHPUT
PLANTS % TOTAL
24%
58%
14%
4%
23
DISTRIBUTION
PHI V Dl flMTC
liULN r LHIN 1 o
AVERAGE
GASOLINE THROUGHPUT
M3/Day (000 Gal /Day)
*11(3)
11(3) - 30(8)
30(8) - 65(17)
>65(17)
CTODTNT (iAQfil TNIF
jlUKllNb uMoULilNL
NUMBER OF PLANTS
5,210
11,210
1,170
260
% TOTAL
29%
63%
7%
1%
Total
18,640
100%
Total
17,850
100%
Source: Bureau of Census, 1972 Census of Wholesale Trade; National Oil Jobbers Council;
National Petroleum News, Factbook, (1972-1978); Industry contacts; ADL estimates.
-------
TABLE 6-24
24
BULK PLANT OWNERSHIP
ALL BULK PLANTS
- BULK PLANTS STORING GASOLINE -
OWNERSHIP SEGMENT NUMBER OF BULK PLANTS % TOTAL NUMBER OF BULK PLANTS % TOTAL
Majors
4,110
22%
3,610
20%
cr>
i
en
Co
Independent
Marketers/Wholesalers 770
Jobbers 13,760
Total 18,640
4%
74%
770
13,470
4%
76%
100%
17,850
100%
Source: National Oil Jobbers Council; National Petroleum News, Factbook, (1972-1978);
Industry contacts; ADL estimates.
-------
Jobbers tend to own a greater portion of the small gasoline bulk plants
and a smaller portion of the large bulk plants than either the majors or
the independent wholesale/marketers. Jobbers, who own 76 percent of all
gasoline bulk plants, own 82 percent of the smallest bulk plants, i.e. less
3
than 150 M (40,000 gallons) of storage capacity, and only 36 percent of
3
the largest facilities, i.e. storage greater than 1,136 M (300,000
gallons) (Table 6-25). The majors, who own 20 percent of the gasoline
bulk plants, own 75 percent of the largest bulk plants and only 18 percent
of the smallest facilities.
6.2.1.6 Employment
Total employment at bulk plants decreased from 105,525 in 1972 to
75,010 in 1978, a decline of 24 percent (Table 6-26). Employment at
gasoline bulk plants was estimated to be 72,130 in 1978 or 96 percent
of the total bulk plant employment. PADD's I and II account for almost
75 percent of the total bulk plant employment and employment at gasoline
facilities.
6.2.1.7 Future Trends
Because gasoline demand is not expected to increase substantially
from its present level and because more gasoline throughput will bypass
storage at bulk plants, no new gasoline storage at bulk plants is expected
to be built. Furthermore, no new bulk plants are expected.
Additional bulk plant closures are anticipated due to increasing
market competition and the ongoing rationalization of petroleum marketing
facilities. Increasing operating costs will continue to favor the larger,
more efficient operators. Because of these factors, an estimated 3,480
27
gasoline bulk plants will close over the next 5 years. All of the
3
closures are expected to be in bulk plants having less than 30 M /day
(8,000 gallons/day) of gasoline throughput.
6-54
-------
TABLE 6-25
GASOLINE BULK PLANT DISTRIBUTION BY SIZE AND OWNERSHIP
25
en
en
TOTAL STORAGE CAPACITY
M3 (000 Gal)
^150(40)
150(40)-568(150)
568(150)-!, 136(300)
Tl, 136(300)
% Total
Total Number of
Bulk Plants Storing
Gasoline
Source: Bureau of Census
MAJORS
2.0
16.2
1.2
0.8
20.2
3,610
, 1972
INDEPENDENT/
MARKETERS TOTAL NUMBER OF BULK PLANTS
WHOLESALERS JOBBERS % TOTAL STORING GASOLINE
0.4 11.0 13.4 2,380
3.5 59.3 79.0 14,100
0.3 4.7 6.2 1,110
0.1 0.5 1.4 260
4.3 75.5 100.0
770 13,470 17,850
Census of Wholesale Trade; National Oil Jobbers Council;
National Petroleum News, Factbook (1972-1978); Industry contacts; ADL estimates.
-------
TABLE 6-26
CM
I
cn
CD
PADD
I
II
III
IV
V
Total
BULK PLANT EMPLOYMENT
AM PI 1 1 y
ALL t>ULI\
EMPLOYMENT
24,210
31,220
9,780
3,520
6,280
75,010
u of Census,
il ; National
PI AIMTr
1 LMIM 1 o
% TOTAL
32%
42%
13%
5%
8%
100%
1972 Census of
Petroleum News
-BULK PLAN'
EMPLOYMENT
22,850
30,180
9,780
3,520
5,800
72,130
Wholesale Trade; Nationa'
, Factbook (1972-1978); Ii
% TOTAL
32%
42%
13%
5%
8%
100%
ADL estimates.
-------
Major oil companies will continue to withdraw from bulk plant
operations in rural and semi-rural areas. An estimated 1,540 bulk plants
28
will be offered for sale by the majors over the next 5 years. Most of
these bulk plants will be bought by jobbers who will consolidate these
facilities with their existing operations. Some attrition, however, will
take place in the total number of facilities.
6-57
-------
6.2.2 COST ANALYSIS FOR BULK PLANTS
6.2.2.1 Introduction
Estimates of the costs for the control of benzene emissions from
the transfer and storage of gasoline at bulk plants are presented for
each of the control options described in Chapter 3: Option 1 is vapor
balancing of incoming transport trucks and either top-submerged or bottom-
loading of delivery trucks (tank wagons); Option 3 is vapor balancing of
both incoming transport trucks and delivery trucks with either bottom
or top-submerged loading; Option 4 is vapor processing, either by
refrigeration or incineration,-in addition to the vapor balancing
specified for Option 3. Both installed capital and total annualized
costs, in January 1978 dollars, are presented for each of the three
options. The control options apply to gasoline bulk plants, which are .
less than 76,000 liters per day of throughput.
The estimates were developed from a combination of costs incurred by
owners of bulk plants and prices quoted from suppliers of control equipment.
The considerable variation in vapor balance equipment costs which results
from the wide variety of equipment already installed and/or differences in
availability is addressed by the presentation of alternative costs for
three systems. The first vapor balance system is the one described by
National Oil Jobbers Council members McCormack and Shuster on February 28,
1978. This system includes all the features such as check valves, flame
arrestors, and high-quality supports and piping, which would be necessary
to meet the strictest local fire and safety regulations.
6-58
-------
The second vapor balance system is the one commonly known as the
"Wiggins system" for bottom loading. For top-loading, the system
commonly known as the "Hous,ton-Galveston" system is presented as the
top-loading alternative to the Wiggins system and as a part of this
second vapor balance system. " The cost estimates for the Wiggins system
have been corroborated by recent estimates by the National Oil Jobbers
Council.31
The third type of vapor balance systems is the one reported by
the Colorado Air Pollution Control Division in October, 1976.32 This
system is an adaptation of the Wiggins system, for bottom-loading, and
various combinations of less expensive piping and supports for top-loading.
Detailed lists of the existing and installed equipment used are not
available, because the permit applications show total costs, but these
combinations of equipment have been judged adequate by the Colorado
Department of Health.
Two other large variations exist in the estimates. Variation in the
labor rate, age$sexisting equipment, and physical configuration of bulk
plants, have not been addressed in the estimates. Therefore, where a
specific facility has conditions which vary substantially from the
assumed parameters of the model plants, considerable variation in cost
should be expected.
Estimates relate to application of the control options to existing
bulk plants and do not include application to new facilities. The bulk
plant portion of the gasoline marketing industry has a negative growth
rate, as mentioned in Section 2.3, and industry members do not foresee
6-59
-------
33
building new bulk plants. Knowledgeable estimates from the industry
indicate that, for the few new bulk plants which might be built, the
only significant difference, in the cost of control equipment would be
a reduction of the installation costs by at least 75 percent for a new
facility compared with installation costs for an existing bulk plant.34
Monitoring costs are not included for any of the three control
options. For Option 4, monitoring of the vapor processing portion of
the control equipment can be accomplished by use of a gas chromatograph.
Since the application of continuous monitoring is a separate decision
from selection of one of the control options, the costs of such monitoring
are not included in the capital and annualized costs estimated for the
control options. The control options do not require monitoring.
State Implementation Plan (SIP) emission levels are the uncontrolled
levels shown in Table 4-2, because there are virtually no SIP requirements
for control during loading and unloading of storage tanks, as mentioned
in Section 5.1. The cost estimates in this chapter represent the cost of
increasing control from the uncontrolled (SIP) level to the options
described. All of these costs, therefore, are attributable to the
proposed control options, and none of the costs is attributable to SIP
requirements.
Cost-effectiveness comparisons among the three control options are
presented for each control option. Although the installed capital control
costs are not significantly related to throughput, the recovery credit
varies directly with throughput. Some conclusions on the cost-effectiveness
of options for various volumes of operation are presented.
-------
Two model plants are used to illustrate the range of cost estimates.
The 15,000 liters per day (4,000 gallons per day) model consists of three
above-ground storage tanks^one loading rack with three arms, and two
account delivery trucks (tank wagons) each with four compartments. The
76,000 liters per day (20,000 gallons per day) model consists of the
same equipment as the smaller model, with two additional account trucks.
The parameters which serve as the basis for the cost estimates are
summarized in Table 6-22.
6.2.2.2 Capital and Annualized Cost Estimates
Shown in Tables 6-23 and 6-24 are cost estimates for the three
control options, based on the assumption that the model bulk plant uses
the most complete and expensive vapor balance equipment, including check
valves, pre-set meters, and high quality support materials and piping.
Shown in Tables 6-25 and 6-26 are cost estimates for the three
control options, based on the assumption that the bulk plant uses less
expensive vapor balance equipment; i.e., the Wiggins system or the Houston-
Galveston system. Shown in Tables 6-27 and 6-28 are cost estimates
for the three control options based on use of the third least expensive
type of vapor balance equipment as described in reference 10. Applications
of these less expensive systems are dependent on acceptances by local fire
and safety officials. General descriptions of the items included in the
estimates are presented in the following paragraphs.
Capital costs include hardware, transportation, installation and sales
tax. Annualized costs include (1) operating costs, such as labor, utilities.
6-61
-------
Table 6-27. _ PARAMETERS USED FOR COST ESTIMATES
Small Model Large Model
1. Throughput,(galIons per day) 15,aOO_(4,OOQ_qalIons/day) 76,000 .(20,000 gallons/day)
2. Loading Racks 1 1
3. Loading Arms per Rack_ 3 3
4. Storage Tanks (above-ground) 3 3
5. Account Trucks (Tank Wagons) 2 4
6. Account Trucks Converted to Bottom Loading 1 2
7. Compartments per account truck 4 4
8. Density of gasoline (Ib/gal) __
9. Emissions of HC prevented (mg/liter)
Option 1 800 800
w Option 3 1260 1260
05 Option 4 3429 3429
10. Working Days per Year 286 286
11. Working hours per day 8 8
12. Peak Loading Rate (Jltersjper min.) _490_(130 .gallons/mlri) 490 (130 gallons/min)
13. Liquid to Vapor Ratio 7.5 7.5
14. Operating Labor Cost ($/hour) 10.0 10.0
15. Propane for Oxidizer (gallons/hour) 0.72 0.72
16. Price of propane ($/gallon) 0.40 0.40
17. Price of electricity ($/KWH) 0.05 0.05
18. Capital Recovery Factors (interest)
a. Vapor Balance Equipment at 20-year life,
10% interest 0.118 0.118
b. Refrigeration or Oxidation equipment at
10-year life, 10% interest 0.163 0.163
c. Taxes, insurance, administration on
capital (all equipment) 0.04 0.04
-------
-Table 6-28.
OPTIONS 1 AND 3 CAPITAL AND ANNUALIZED COST ESTIMATES
(In thousands of January 1978 dollars)
Option 1
Bottom or Top-Submerged
Loading with Incoming Vapor Balance
1.
Truck (Tank Wagon) Conversion, including
Labor
Z. Rack Conversion, including labor
cr>
i
u>
3.
4.
5.
6.
7.
8.
9.
10.
11.
Installation, excluding labor
TOTAL INSTALLED CAPITAL
Operating Labor
Utilities
Maintenance Labor and Materials
Capital Charges
TOTAL ANNUAL I ZED COST
Less Recovery Credit
NET ANNUAL I ZED COST
Bottom
Loading
15,000 76,000
^6.21 12.5A_
v. 35.45
5.31
4/.U3
35.45
5.82
53-.ol
Top-Submersed
15,000
TpcT
N/A
3.54
0.71
4.25
NONE .
NONE
1.41
7.41
8,82.
0.51
Ml _
1.61
8.48
10.09
2.59
0.13
0.67
0.80
0.51
.0.29
_-?*•
000
pd~
N/A
3.
0.
..._«•
NONE
NONE
0.
0.
0.
2.
__u.
54
71
- ;
13
67
80
59
79) ..
Option 3
Bottom or Top-Submerged Loading With
Incoming and Outgoing Vapor Balance
Bottom Loading
15,000 76,000
lpj[ Ipd
.7.02.. 14
35.45 35
5.52 6
47. yg 55
NONE
NONE
1.43 1
7.56 8
B.99 10
0.81 4
8.18 6
,05 .
.45 .
.22
.72
Top-Submerged
15,000^
2.38
18.30
2.3b
23. u3
, 76,000
'Ipd "
4.76
18.30
2.bX
25,. /3
NONE
NONE
.67
.76
.43
.08
•35...
0.69
3.63
4.32
0.81
-3.51
0.77
4.05
4.82
4.08
0.74 .
-------
Table 6-29 OPTION 4 (vapor Processing)CAPITAL AND ANNUALIZED COST ESTIMATES
(In thousands of January, 1978 Dollars)
SINGLE SYSTEMS
Refrigeration
Oxidation
0>
cr>
Recovery Equipment
. Processing Equipment
Recovery Installation
Processing INstallation
TOTAL INSTALLED CAPITAL
Recovery Operating Labor
Processing" Operating Labor
Recovery Utilities
Processing Utilities
Recovery Maintenance
Processing Maintenance
Recovery Capital Charges
Processing CapitaTCfiarges
TOTAL ANNUALIZED COST
Less: Processing Recovery Credit'
NET ANNUALIZED COST
Recovery tquipment
Pr6cessirig~'Equipment
Recovery Installation
Processing'lnTtallation
TOTAL INSTALLED CAPITAL
Recovery Operating Labor
: Processing Operating Labor
Recovery Utilities
: PrOcesSlncpJflTlties
Recovery Maintenance
" Processing Maintenance
Recovery Capital Charge
Procession Capital Charges
TOTAL ANNUALIZED COST
Less: Processing Recovery Credit
NET ANNUALIZED COST
Bottom Loading
15,000
Ipd
42.47"
43.22
5.52
25.93
117.14
NONE
1.43
NONE
2.17
T.43~
2.59
"V.S6
14.02
29.20
2.19
76,000
Jpd
49.50
43.22
B.22
25.93
'124.87
NONE
1.43
NONE
2.17
1.67 1
2.59
o./D
14.02
30.64
11.11
27.01 19. W
Refrigeration Plus
Bottom
15,000
Ipd
42.47
58.70
5.52
35.69
142.38
NONE
1742~
NONE
2.17
1.43
2~.?(T~
7.56
19.54
35.03
2.19
32.84
Loading
76,000
Ipd
49.50
58.70
5.52
35.69
149.41
NONBl
1.43
NONE
2.17
1.67
2.90
8 .76
19.54
36.47
11.11
25.36
Top Submerged
15,000
_j£d_
20.68
43.22
2.3S "
25.93
-$J2Tt8 -
NONE
1.43
NONE
2.17
• 0.69
2.59
-3763-
14.02
24. S3
2.19
76,000
Ipd
~23.06~~
43.22.
— -TXT —
25.93
94.88"
NONE
1.43
NONE
2.17
0.77
2.59
4VD5
14.02
2S.UJ
11.11
22.34 13.92
DUAL
Oxidation
Top Submerged
15,000
Ipd
20.68
58.70
2.35
35.69
117.42
NONE
1.43
NONE
2.17
0.69
" 2.90~
3753""
19.54
30.36
2.19
28.17
76,000
Ipd
22.06
58.70
2.67
35.69
120.12
NONE
1.43
NONE
2.17 ','•
0.77
2.90
4.05 ~
19.54
30.86
11.11
19.75
Bottom
15,000
Ipj^
."42747
15.50
/5.5r-
9.76
73;zr
NONE •
1.43
NONE
0.16
1.43 '
0.62
7-56 .
- 5.12
16;32
NONE
1b.32
SYSTEMS
Bottom
15,000
Ipd
T2T47
31.00
5.52
19.52
98.51
NONE
1.43
NONE
0.16
1.43
0.93
" ' !
/.bb
10.24
,21.75
NONE'""
21.75
Loading
76,000
Ipd
49.50"
15.50
6.22
.9.76
80.96 •
NONE .
1.43
NONE
0.16
1.67
0.62
8.76
s.:2 :
17:76
NONE
' • • 1 U
17. 76
Oxidation Plus
Loading
76,000
Ipd
49.50 :
31.00
5.52
19.52
105.54
NONE
1.43
NONE
0.16
1.67
U.93
8:76
10.24
23.19
.NONE •*•*•'
23.19
Top Submerged
15,000
Ipd
;i0.68
15.50
2Y35
9.76
43:32-
NONE
1.43
NONE
0.16
-tf.69'
0.62
3.63
5.12
76,000
Ipd
23.06
15.. 50
Z.'67
.9.76
td.w
NONE
1.43
NONE
0.16
._..._,__-
0.62
4.0b '
5.1.'
flT.eS 12.15
NONE NONE
TTtes I2H5~
Oxidation
Top Submerged
15,000
Ipd
.20.68"
31.00
"7.35
19.52
•~?J:BS-
NONE
~1.43
NONE
0.16
"0.69"
~0.93
,3;63
.10.24
T7'.08
:;if NONE
17.08
76,000
. Jpd
"23.06-
31.00
2.67
. 19.5?
>6.2"5"
NONE
"1.43""
'NONE
0.16
U.93
"' 4.05
.10.24
17.58
! i NONE
17.58
-------
Table 6-30 ~" OPTIONS i AND 3 (VAPOR BALANCE WITH LESS EXPENSIVE EQUIPMENT)
COST ESTIMATES5(1n thousands of January 1978 dollars)
Option 1
Option 3
i
cr>
en
Truck (tank wagon) conversion, Including
labor
Rack conversion, including labor
Piping jack to storage, including
labor1
Installation, excluding labor
TOTAL INSTALLED CAPITAL
Operating Labor
Utilities
Maintenance Labor and Material
Capital charges
TOTAL ANNUALIZED COST
Less Recovery Credit
NET ANNUALIZED COST (credit)
B<
Loading
ottom or Top-Submerged
with Incoming Vapor Balance ..
Bottom Loading
15,000
Ipd
0.97
7.47
1.58
2.29
12.31
NONE
NONE
0.37
1.94
2.31
0.51
1.70
76,000
Ipd
1.95
7.47
1.58
2.34
13.34
NONE
NONE
0.40
2.10
2.50
2.59
(0.09)
Top-Submerged
T5.000
N/A
. 3.54
N/A
0.71
4.25
NONE
NONE
0.13
0.67
0.80
1
. 0.51
0.29
76,000
Ipd
N/A
3.54
N/A
0.71
4.25
NONE
NONE
0.13.
0.67
0.80
2.59
(1.79)
Bottom or Top-Submerged Loading With
Incoming and Outgoing Vapor Balance
Bottom Loading
15,000
Ipd
. 1.95
7.47
1.58
2.34
13.34
NONE '
NONE
0.40
2.10
2.50
0.81-
1.69
76,000
Ipd
3.90
7.47
1.58
2.45
15.40-
NONE
NONE
0.46
2.43
2.89
4.08
(1.19)
Top-Submerged
15.UUU
2.16
6.71
N/A
1.83
10.70
• NONE
NONE
0.32
1.69
2.01
0.81
1.20
/b.UUU
4.33
6.71
N/A
1.94
12.98
NONE
NONE
0.39
2.04
2.43
4.08
(1.85)
-------
Table 6-TTr
OPTION 4 (Vapor Processing with less expensive equiDmenU
CAPITAL AND ANNUALIZED COST ESTIMATES (In thousands of '
January 1978 dollars)
cr»
Recovery Equipment
Processing Equipment
Recovery Installation
Processing Installation
TOTAL INSTALLED CAPITAL
Recovery Operating Labor
Processing Operating Labor'
Recovery Utilities
Processing Utilities
Recovery Maintenance
Processing Maintenance'
Recovery Capital Charges
Processing Capital Charges
TOTAL ANNUALIZED COST
Less: Processing Recovery Credit'
NET ANNUALIZED COST
Recovery Equipment
Processing Equipment
Recovery Installation
Processing Installation
TOTAL INSTALLED CAPITAL
Recovery Operating Labor
Processing Operating Labor'
Recovery Utilities
Processing Utilities
Recovery Maintenance
.Processing Maintenance
Recovery Capital Charge
Processing Capital Charges
TOTAL ANNUALIZED COST
Less: Processing Recovery Credit
NET ANNUALIZED COST
Refrigeration
Oxidation
Bottom
15,000
Ipd
11.00
43.22
2.34
25.93
82.49
NONE
1.43
NONE
2.17
0.40
2.59
2.10
14.02
22.71
2.19
20.52
Loading
76,000
.IP.d-
12.95
43.22
2.45
25.93
84.55
NONE
1.43
NONE
2.17
0.46
2.59
2.43
14.02
23.10
11.11
11.99
Refrigeration
Bottom
15,000
_ Ipd
4.00
58.70
2.34
35.69
107.73
NONE
1.42
NONE
2.17
0.40
2.90
2.10
19.54 "
28.53
2.19
26.34
Loading
76,000
Ipd
12.95
58.70
2.45
35.69
109.79
\l NONE
1.43
NONE
2.17
0.46
2.90
2.43
19.54
28.93
11.11
17.82
Top Submerged
16,000
Ipd '
8.87
43.22
1.83
25.93
79.85
NONE
1.43
NONE
2.17
0.32
2.59
1.69
14.02
22.22
2.19
20.03
76,000
Ipd
11.04
43.22
1.94
25.93
82.13
NONE
1.43
NONE
2.17
0.39
2.59
2.04
14.02
22.64
11.11
11.53
Bottom
15,000
Ipd
11.00
15.50
2.34
9.76
38.60
NONE
1.43 .
NONE
0.16
0.40
0.62
2.10
5.12
9.83
NONE
9.83
Plus Oxidation
Top Submerged
15,000
JPd
8.87
58.70
1.83
35.69
105.09
NONE
1.43
NONE
2.17
0.32
2.90
1.69
19.54 ~~~
22.05
2.19
25.86
76,000
1Pd
11.04
58.70
1.94
35.69
102.37
NONE
1.43
NONE
','2.17
0.39
2.90
2.04
IF. 54
28.47
11.11
17.36
Bottom
15,000
Ipd
11.00
31.00
2.34
19.52
63.86
NONE
1.43
NONE
0.16
0.40
0.93
2.10
10.24 —
15.26
NONE ( .
15.26
Loading
76,000
Ipd
12.95
15.50
2.45
9.76
40.66
NONE
1.43
NONE
0.16
0.46
0.62
2.43
5.12
10.22
NONE
10.22
Oxidation
Loading
76,000
Ipd ...
12.95
31.00
2.45
19.52
65.92
NONE
1.43
NONE
0.16
0.46
0.93
2.43
"- 10.24
15.65
NONE
15.65
Top Submerged
15,000
1Pd
8.87
15.50
1.83
9.76
35.96
NONE
1.43
NONE
0.16
0.32
0.62
1.69
5.12
9.34
NONE
9.34
Plus Oxidation
76,000
Ipd
11.04
15.50
1.94
9.76
38.24
NONE
1.43
NONE
0.16
0.39
0.62
2.04
5.12
9.76
NONE
9.76
Top Submerged
15,000
Ipd
8.87
31.00
1.83
19.52
61.22 l
NONE
1.43
NONE
0.16
0.32
0.93
1.69
10.24
14.77
NONE
14.77
76,000
Ipd
11.04
31.00
1.94
19.52
63.50
NONE
1.43
NONE
0.16
0.39
0.93
2.04
10.24-
15.19
NONE
15.19
-------
en
Table 6-32 OPTIONS i AND 3 (VAPOR BALANCE WITH LEAST JXPENSIVE EQUIPMENT CAPITAL AND
•ANNUALIZEO COST ESTIMATES (in thousands of January 1978"dollars)
Option 1
Option 3
1. Truck (Tank Wagon) Conversion, including
laborb
2. Rack Conversion, including labor
3. Installation, excluding labor
4. TOTAL INSTALLED CAPITAL
5. Operating Labor
6. Utilities
7. Maintenance Labor and Materials
8. Capital Charges
9. TOTAL ANNUALIZEO COST
10. Less Recovery Credit
11. NET ANNUALIZED COST
Bottom or Top-Submerged
Loading with Incoming Vapor Balance
Bottom Loading
15,000 76,000
0.97 1.94
1.08 1.08
0.28 0.41
2.33 3.43
NONE
NONE
0.07 0.10
0.37 0.54
0.44 0.64
0.51 2.59
(0.07) (1.95)
Top-Submerged
15,000 76,000
0.75 0.75
0.75 0.75
0.20 0.20
1.70 1.70
NONE
NONE
0.05 0.05
0.27 0.27
0.32 0.32
0.51 2.59
(0.19) (2.27)
Bottom or Top-Submerged Loading With
Incoming and Outgoing Vapor Balance
Bottom Loading
15,000 76,00(3
_lpd Ipd
1.61 3.23
1.08 1.08
0.36 0.58
3.05 4.89
NONE
NONE
0.09 0.15
0.48 0.77
0.57 0.92
0.81 4.08
(0.24) (3.16)
Top-Submerged
15,0(JU 76.0UU
1.69 2.15
1.69 2.15
0.46 0.58
3.84 4.88
NONE
NONE
0.12 0.15
0.60 0.77
0.72 0.92
0.81 4.08
(0.09) (3.16)
-------
Table 6-33. OPTION 4 (VAPOR PROCESSING WITH THE LEAST EXPENSIVE VAPOR RECOVERY
EQUIPMENT) CAPITAL AND ANNUALIZED COSTS
(in thousands of January 1978 dollars)
Refrigeration
Oxidation
en
i
oo
Recovery Equipment
Processing Equipment
Recovery Installation _
Processing Installation
TOTAL INSTALLED CAPITAL
Recovery Operating Labor
Processing Operating Labor
Recovery Utilities
Processing Utilities
Recovery Maintenance
Processing Maintenance
Recovery Capital Charges
Processing Capital Charges
TOTAL ANNUALIZED COST
Less: Processing Recovery Credit'
NET ANNUALIZED COST
Bottom
15,000
lod
^ ^__j.^_— „
2.69
43.22
0.36
25.93
72.20
NONE-
1.43
NONE
2.17
0.09
2.59
0.48
14.02
20.78
2.19
18.59
Loading
76,000
Ipd
4.31
43.22
0.58
25.93
74.04
NONE
1.43
NONE
2.17
0.15
2.59
0.77
14.02
21.13
11.11
10.02
Top-Submerged
15,000
lod
3.38
43.22
0.46
25.93
72.99
NONE
1.43
NONE
2.17
0.12
2.59
0.60
14.02
20.93
2.19
18.74
76,000
Ipd
4.30
43.22
0.58
25.93
74.03
NONE
1.43
NONE
2.17
0.15
o!?7
14.02
21.13
11.11
10.02
Refrigeration Plus Oxidation
Recovery Equipment
Processing Equipment
Recovery Installation
~~ "Processing InstaTlat'lorT"
TOTAL INSTALLED CAPITAL
Recovery Operating Labor
Processing OperatingTtabor
Recovery Utilities
' Processing Utilities
Recovery Maintenance
Processing Maintenance
Recovery Capital Charge
Processing Capital tnarges
TOTAL ANNUALIZED COST
Less: Processing Recovery Credit
NET. ANNUALIZED COST
Bottom
15,000
Ipd
2.69
58.70
0.36
35.69
97.68
NONE
1.43
NONE
2.17
0.09
2.90
0.48
19.54
26.61
2.19
™.42
Loading
76,000
Ipd
4.31
58.70
0.58
35.69
99.28
NONE
1.43
NONE
2.17
0.15
2.90
0.77
19.54
26.96
11.11
15.85
Top-Submerged
15,000
Ipd
3.38
58.70
0.46
35.69
98.23
NONE
1.43
NONE
2.17
0.12
2.90
0.60
19.54
26.76
2.19
24.57
76,000
Ipd
4.30
58.70
0.58
35.69
99.27
NONE
1.43
NONE
,,2.17
• f\ 1C
2!go
0.77
19.54
26.96
11.11
15.85
Bottom
15,000
Ipd
2.69
15.50
0.36
9.76
28.31
NONE
1.43
NONE
0.16
0.09
0.62
0.48
5.12
7.90
NONE
7.90
Loading
76,000
_l£d__
4.31
15.50
0.58
9.76
30.15
NONE
1.43
NONE
0.16
0.15
0.62
0.77
5.12
8.25
NONE
8.25
Oxidation Plus
Bottom
15,000
Ipd
2.69
31.00
0.36
19.52
53.57
NONE
1.43
NONE
0.16
0.09
0.93
0.48
10.24
13.33
NONE
13.33
Loading
76,000
Ipd
4.31
31.00
0.58
19.52
55.41
NONE
1.43
NONE
0.16
0.15
0.93
0.77
10.24
13.68
NONE
13.68
Top Submerged
15,000
— IPA.
3.38
15.50
0.46
9.76
29.10
NONE
1.43
NONE
0.16
0.12
0.62
0.60
5.12
8.05
NONE
8.05
Oxidation
76,000
]Pd
4.30
15.50
0.58
9.76
30.14
NONE
1.43
NONE
0.16
0.15
0.62
0.77
5.12
8.25
NONE
8.25
Top Submerged
15,000
Ipd
3.38
34:«
19.52
54.36
NONE
1.43
NONE '
0.16
0.12
0.93
0.60
10.24
13.36
NONE
13.36
76,000
Ipd
4.30
3H8
19'.52
55.40
NONE
1.43
NONE
0.16
0.15
0.93
0.77
10.24
13.58
NONE
13.58
-------
and maintenance (2) capital charges, such as interest, insurance and taxes,
and (3) credit for recovery of gasoline as a salable product. Among the
three control options there" are some differences in the rates used to
compute interest charges, recovery credits and indirect installation
charges, as a result of differences in equipment. For example, the
refrigeration equipment has an expected life and expected installation
contingency factor different from those factors which apply to the vapor
recovery equipment. Similarly, the refrigeration unit recovery credit
factor, based on prevented emissions, differs from the factor used for
the vapor recovery equipment. Operating costs for the vapor balance
equipment in Options 1 and 3 are limited to maintenance costs, while the
vapor processing equipment in Option 4 has labor, utilities and maintenance
costs.
6.2.2.3 Comparisons of Costs
Capital and annualized costs for all three control options and for all
three types of vapor balance equipment are presented in Table 6-29.
For Option 1, with most expensive equipment, installed capital
costs range from $4,250 for top-submerged loading for the small model
plant to $53,810 for bottom-loading, and annualized costs range from an
$1,790 credit for top-submerged loading to $8,310 for bottom-loading.
Using less expensive equipment, the top-loading installed capital costs
are the same (same equipment) but the bottom-loading installed capital
costs are $13,340 for use of the Wiggins system. Annualized costs for
the less expensive equipment range from $1,790 credit for top-submerged
loading to $1,700 for bottom-loading.
6-69
-------
Table 6-34.
COMPARISON OF CAPITAL AND ANNUALIZED COSTS
(In thousands of January 1978 dollars)
i
•vj
o
Most Expensive. Eauipment
Control Alternative
Option 1
Installed Capital
Net Annual ized
Option 3
Installed Capital
Net Annual ized
Option 4
Single Systems
(1) Refrigeration
Installed Capital
Net Annual ized
(2) Oxidation
Installed Capital
Net Annual i zed)
Dual Systems
(1) Refrigeration plus
oxidation
Installed Capital
Net Annual ized
(2) Oxidation plus
oxidation
Installed Capital
Net Annual ized
Bottom
16,000
lod -^
47.03
8.31
47.99
8.18
117.14
27.01
73.25
16.32
142.38
32.84
98.51
21.75
Loading
"76,000 "
^ lod
53.81
7.50
55.72
6.35
124.87
19.53
80.98
17.76
149.41
25.36
105.54
23.19
Top-Submerged
15,000
lod
4.25
0.29
23.03
3.51
92.18
22.34
48.32
11.65
117.42
28.17
73.55
17.08
76,000
lod
4.25
(1.79)
25.73
0.74
94.88
13.92
50.99
12.15
120.12
19.75
76.25
17.58
Less Expensive Equipment
Bottom
15,000
Inci
12.31
1
13
,' 1
82
20
38
9
107
26
63
15
.70
.34
.69
.49
.52
.60
.83
.73
.34
.86
.26
Loading
76,000
lod
13.34
(0.09)
15.40
(1.19)
84.55
11.99
40.66
10.22
109.79
17.82
65.92
15.65
Top-Submerged
15,000
lod
4.25
0.29
10.70
1.20
79.85
20.03
35.96
9.34
105.09
25.86
61.22
14.77
76,000
lod
4.25
(1.79)
12.98
(1.85)
82.13
11.53
38.24
9.76
102.37
17.36
63.50
15.19
Least Expensive Equipment
Bottom
15,000
Ipd
2.33
(0.07)
3.05
(0.24)
72.20
18.59
28.31
7.90
97.68
24.42
E3.57
13.33
Loading
76,000
Ipd
3.43
(1.95)
4.89
(3.16)
74.04
10.02
30.15
8.25
99.28 .
15.85
55.41
13.68
Top-Submerged
15,000 76,000
Ipd Ipd
;»
1.70
(0.19))
3.84
(0.09)
72.99
18.74
29.10
' 8.05
98.23
24.57
54.36.
13.36
• •' ".t
1.70
(2.27)
•* \\
4.88
(3.16)
74.03
10.02
30.14
8.25
99.27
15.85
55.40
13.58
6-55
-------
Because of this lack of experience in application of these control
devices, installation costs include a 20 percent allowance for contin-
gencies, compared with the~10 percent used for vapor recovery equipment.
Annualized costs range from $11,650 for the oxidation system with
top-loading vapor recovery to $32,840 for the refrigeration system with
bottom-loading vapor recovery. The refrigeration system and the refrigera-
tion-plus-oxidation dual system show the effect of recovery credits,
which cause the larger model plant to have lower net annualized costs.
The oxidation system and the oxidation-plus-oxidation dual systems,
which lack recovery credits, have higher annualized costs for the larger
model plant.
Using the less expensive vapor recovery equipment installed capital
costs for Option 4 range from $35,960 for single system oxidation using
top-submerged vapor recovery to $109,730 for dual system refrigeration-
plus-oxidation using bottom-loading. Annualized costs range from $9,340
for single-system oxidation to $26,340 for dual system refrigeration-
Pius-oxidation.
Using the least expensive vapor recovery equipment, installed capital
costs for Option 1 range from $1,700 for the top-submerged loading to
$3,430 for bottom-loading for the larger model plant. Annulaized costs
range from the $2,270 credit for the larger plant with top loading to
the $70 credit for the smaller plant with bottom-loading. For Option 3
6-71
-------
installed capital costs range from $4,890 for the larger plant with
bottom-loading to $3,050 for the smaller plant with bottom loading.
Annualized costs range from the $3,160 credit for the larger plant with
either bottom or top-submerged loading to the $90 credit for the smaller
plant with top-loading.
Applying the least expensive vapor recovery equipment to the control
systems required for Option 4 results in installed capital costs which
range from $28,310 for the single oxidation system for the smaller plant
using bottom loading to $99,280 for the refrigeration plus oxidation
system for the larger plant using bottom loading. Annualized costs
range from $7,900 for the single oxidation system for the smaller plant
using bottom loading to $24,570 for the refrigeration plus oxidation
system for the smaller plant using top-submerged loading.
The summary of capital and annualized costs-for the three control
alternatives shown in Table 6-29 presents several comparisons. For both
all three categories of equipment, the highest installed capital and
annualized costs result from using the dual system refrigeration-plus-
oxidation. Also, for all three categories of equipment, the lowest
installed capital and annualized costs result from using the single
system oxidation with top-submerged loading.
6.2.2.4 Cost-Effectiveness
Comparisons of the control options cost-effectiveness ratios, in
dollars per kilogram of hydrocarbon removed, are shown in Table 6-30
and graphically in Figures 6-3 through 6-8. The conversion of the
6-72
-------
Table 6-35. COST-EFFECTIVENESS (1n January 1978 dollars per kilogram of Benzene controlled)
co
Most Expensive Equipment
Control Alternative
Option 1
Option 3
Option 4
(1) Single Systems
Refrigeration
Oxidation
(2) Dual Systems
Refrigeration plus
oxidation
Oxidation plus
oxidation
. Bottom Loading
15,000 76,000
lod Ipd
303 ,
189
230
139
279
185
54
29
33
30
43
39
Top-Submerged
~ 15,000 76,000
Ipd Ipd
10
81
190
99
240'
145
(13)
4
24
20
34
29
Less Expensive Equipment
Bottom
15,000
Ipd
63
39
175
84
224
130
Loading •'!
"76,000""
Ipd
(1)
(5)
20
:18
30
26
-------
LEGEND
SYMBOL OPTION CONTROL DEVICE LOADING
VAPOR BALANCE
VAPOR BALANCE
REFRIGERATION
OXIDATION
REFRIGERATION
PLUS OXIDATION
OXIDATION
PLUS OXIDATION
15 20 25 30 35 40 45 50 55 60 65 70 75
THROUGHPUT. 103 liter/day
Figure 6-3. , Cost-effectiveness for most expensive equipment (top-loading) in January, 1978 dollars.
6-74
-------
CONTROL DEVICE
SYMBOL OPTION
VAPOR BALANCE
VAPOR BALANCE
REFRIGERATION
OXIDATION
REFRIGERATION
PLUS OXIDATION
OXIDATION
PLUS OXIDATION
BOTTOM
BOTTOM
BOTTOM
BOTTOM
BOTTOM
15 20 25 30 35 40 45 50 55 60 65 70 75
THROUGHPUT. 10'liter/day
Figure 6-4. : Cost-effectiveness for most expensive equipment (bottom-loading) in January, 1978 dollars.
6-75
-------
300
280
260
240
220
200
»180
JJ
1 160
o
oc
140
120
100
80
60
40
20
20
iiii,;
LEGEND
SYMBOL OPTION CONTROL DEVICE LOADING
1T 1 VAPOR BALANCE TOP
3T 3 VAPOR BALANCE TOP
4RT 4 REFRIGERATION TOP
40T 4 OXIDATION TOP
4ROT 4 REFRIGERATION TOP
PLUS OXIDATION
400T 4 OXIDATION TOP
PLUS OXIDATION
15 20 25 30 35 40 45 50 55 60 65 70 75
THROUGHPUT. 103 liter/day
Figure 6-5. . Cost-effectiveness for less expensive equipment (top-loading) in January", 1978 dollars.
6-76
-------
300
280
260
240
220
200
180
£ 160
140
120
100
80
60
40
20
20
LEGEND
SYMBOL OPTION CONTROL DEVICE LOADING
1B 1 VAPOR BALANCE BOTTOM
3B 3 VAPOR BALANCE BOTTOM
4RB 4 REFRIGERATION BOTTOM
40B 4 OXIDATION BOTTOM
4ROB 4 REFRIGERATION BOTTOM
PLUS OXIDATION
400B 4 OXIDATION BOTTOM
PLUS OXIDATION
15 20 25 30 35 40 45 50 55 60 65 70 75
THROUGHPUT. 103 liter/day
Figure 6-6. Cost-effectiveness for less expensive equipment (bottom-loading) in January, 1978 dollars.
6-77
-------
300
280
260
240
220
200
180
160
140
120
100
80
60
40
20
20
LEGEND
SYMBOL OPTION CONTROL DEVICE LOADING
IT 1 VAPOR BALANCE TOP
3T 3 VAPOR BALANCE TOP
ART 4 REFRIGERATION TOP
40T 4 OXIDATION TOP
4ROT 4 REFRIGERATION TOP
PLUS OXIDATION
400T 4 OXIDATION TOP
PLUS OXIDATION
15 20 25 38 35 40 45 50 55 60 65 70 75
THROUGHPUT. 103 liter/day
Figure 6-7. [Cost-effectiveness for least expensive equipment (top-loading) in January, 1978 dollars.
6-78
-------
SYMBOL OPTION
CONTROL DEVICE
VAPOR BALANCE
VAPOR BALANCE
REFRIGERATION
OXIDATION
REFRIGERATION
PLUS OXIDATION
OXIDATION
PLUS OXIDATION
BOTTOM
BOTTOM
BOTTOM
BOTTOM
BOTTOM
15 20 25 30 35 40 45 50 55 60 65 70 75
20
THROUGHPUT. 10J liter/day
Figure 6-8 Cost-effectiveness for least expensive equipment (bottom-loading) in January, 1978 dollars.
6-79
-------
second delivery truck to either bottom or top-submerged loading vapor
balance (which is the only difference between the parameters of the
small and the large model plants) would cause an abrupt step in the cost
curves. Since the point at which this conversion would be made cannot
be accurately estimated for an actual plant, a straight line is used to
represent the change. Costs do not vary smoothly with throughput, but
the recovery credit does vary directly with throughput. Thus, the cost-
effectivenesses for each of the three control options, including single
and dual systems, for all three categories of vapor balance equipment are
presented in Figures 6-3 through 6-8 in such a way that comparisons
among control options and control equipment can be made.
6-80
-------
6.2.3 Bulk Plant Impacts
6.2.3.1 Introduction
The principal economic impacts of the three proposed vapor control
options which would reduce the amount of benzene emitted into the
atmosphere by bulk plants are:
• the number of potential bulk plant closures;*
t the employment displaced by these closures;*
• the total cost of installing vapor control at bulk plants.
Since all tank wagons (or account trucks) are owned by bulk plant operators,
the cost of modifying the tank wagon fleet is included in the total cost
of installing vapor control at bulk plants.
All of the control systems discussed in this section will be top-
loading systems. It was assumed that bulk plant operators would choose
a top-loading system in order to comply with the various control options
because it is a less expensive modification than converting their
operations to bottom-loading. Some bulk plant operators, however, may
choose bottom-loading for reasons of greater efficiency and safety.
But the operators that would choose a bottom-loading system are ones that
are in a stronger financial position than the rest of the industry and
their decision to bottom-load would not significantly affect the results
of the closure analysis.
*
The monetary costs of these impacts have not been calculated.
6-81
-------
6.2.3.2 Closure Methodology
The approach used to calculate bulk plant closures is the same as
that which was used for bulk terminals in Section 6.1.3. Bulk plants may
close because of vapor control economics for either of the following reasons:
• Bulk plant operators are unable to obtain the capital necessary
to install vapor control equipment.
• Bulk plants would fail to achieve a sufficient or acceptible
level of profitability if vapor control were installed.
However, in the closure analysis of bulk plants, three distinct cost
scenarios were evaluated for each of the three proposed control options.
These control costs were discussed in Section 6.2.3 and will be referred
to here as NOJC (most expensive cost scenario ), Houston-Galveston (less
expensive) and Colorado APCD (least expensive). It is important to note
that all of control systems represented by these cost scenarios were assumed
to be equally efficient in controlling gasoline vapors for each one of the
three control options.
Large and small bulk plant prototypes were developed to facilitate
the bulk plant closure analysis. The gasoline throughput characteristic
of these prototypes corresponds with the gasoline throughput capacities of
the model vapor control systems described in Section 6.2.2. Because almost all
bulk plants receive petroleum products by truck transport, no differentiation
was made based upon mode of gasoline receipt.
For Option 3 compliance, refrigeration and incineration were the two
control technologies chosen for analysis. These systems have the least
capital requirement of the model control systems examined. Because the
continuous and efficient control of hydrocarbon emissions must be assured,
the above control systems will include an incineration stand-by unit,
6-82
-------
e.g. refrigeration/incineration and incineration/incineration. The
incineration unit also had the least capital requirement of any of the
stand-by systems examined.
6.2.3.3 Bulk Plant Closures
Depending upon the control option, cost scenario and control
technology selected the number of bulk plant closures due to an inability
to access adequate capital ranges from 0 to almost 9,000 (Table 6-36).
All of these facilities are assumed to be operated by jobbers and
independent marketers. No closures are expected under Option 1 compliance
for any of the three cost scenarios. Under Option 3 an estimated 1,700
closures are expected for the NOJC cost scenario; no closures are likely
for either of the other two cost scenarios. Bulk plant closures under
Option 4 compliance are approximately the same for the various cost scenarios
and control technologies, i.e. between 8,000 and 9,000 facilities or up
to 48 percent of the 1978 bulk plant population.
The closures caused by capital constraints were then subtracted from
the bulk plant population to avoid possible double counting. Not all of
these closures, however, would have resulted from this factor exclusively.
Some would also close because they were unable to achieve an an acceptable
level of profitability.
The number of bulk plant closures due to insufficient profitability
also depends upon the control option, cost scenario and control technology
chosen. Up to 130 closures are expected under Option 1 for the NOJC and
Houston-Galveston scenarios; no closures are likely for the Colorado APCD
cost scenario (Table 6- 37). Option 3 compliance is expected to cause
between 50 (Colorado APCD) and 530 (NOJC) closures while Option 4 compliance
will cause roughly the same number of closures for each cost scenario.
6-83
-------
TABLE 6-36
35
BULK PLANT CLOSURES DUE TO INACCESSIBILITY OF CAPITAL
OPTION 1 OPTION 3 OPTION 4
BALANCE INCOMING BALANCE INCOMING REFRIGERATION/ INCINERATION/
COST SCENARIO TRANSPORTS ONLY & OUTGOING TRUCKS INCINERATION INCINERATION
NOJC Costs 0 1,690 8,990 8,880
:T>
Houston-Galveston Costs 0 0 8,960 8,820
Colorado APCD Costs 0 0 8,950 8,820
-------
TABLE 6-37
36
BULK PLANT CLOSURES DUE TO INSUFFICIENT PROFITABILITY
COST SCENARIO
OPTION 1
BALANCE INCOMING
TRANSPORTS ONLY
OPTION 3
BALANCE INCOMING
& OUTGOING TRUCKS
OPTION 4
REFRIGERATION/
INCINERATION
INCINERATION/
INCINERATION
NOJC Costs
130
530
1,300
800
Houston-Galveston Costs
I
00
cn
130
240
1,180
690
Colorado APCD Costs
50
1,100
610
-------
Option 4 closures, however, will vary somewhat by control technology.
Between 600 and 800 closures are expected due to the costs of and
incineration/incineration system, which has the smaller capital require-
ment of the two technologies, while 1,100 to 1,300 bulk plant closures are
expected due to the more expensive refrigeration/incineration costs. A
summary of the bulk plant closures for each cost scenario appears in
Table 6-38 .
Bulk plant closures will not significantly impact the national gasoline
marketing network under Options 1 and 3 since most closures will be low
throughput facilities. A large portion of these closures will occur in
metropolitan areas where other bulk storage facilities, i.e. terminals
and larger, more efficient bulk plants, will subsequently handle the
product throughput of the closed facilities. Option 4, however, would
impact a significant portion of the gasoline marketing network both in
terms of number of facilities and the amount of product throughput. The
product throughput of the closed facilities is assumed to continue to
flow to end-users (at a higher price, however) but a major re-structuring
of the bulk plant market would be likely.
Using the high and low closure estimates for Options 1 and 3 and the
high and low closure estimates of the less expensive control, i.e.
incineration/incineration, technology for Option 4, the number of bulk
plant closures by ownership was calculated. Approximately 85 percent of
the closures, or 110 bulk plants, which occur because of Option 1 compliance
will be at jobber operated facilities (Table 6-39 ). The remaining 15
percent will be bulk plants owned by independent marketer/wholesalers; no
closures are anticipated at any majors' facilities. Jobber closures under
6-86
-------
TABLE 6-38
CLOSURE
OPTION 1
BALANCE INCOMING
COST SCENARIO TRANSPORTS ONLY
NOJC Costs 130
N
Houston-Gal veston Costs 130
£ Colorado APCD Costs 0
37
SUMMARY AT BULK PLANTS
OPTION 3
BALANCE INCOMING
& OUTGOING TANKS
2,220
240
50
OPTION 4
REFRIGERATION/
INCINERATION
10,290
10.140
10,050
INCINERATION/
INCINERATION
9,680
9,510
9,430
-------
TABLE 6-39
CLOSURE IMPACT AT BULK PLANTS BY OWNERSHIP
38
COST SCENARIO*
OPTION 1
HIGH
LOW
OPTION 3
HIGH
LOW
OPTION 4
HIGH
LOW
Majors
**
40
540
340
Independent
Marketer/Wholesalers
20
80
10
180
130
Jobbers
110
2,100
40
8,960 8,960
TOTAL
130
2,220
50
9,680
9,430
* High impact = NOJC cost scenario
Low impact = Colorado APCD cost scenario
** Includes regional refineries
-------
Option 3 range from 2100, or 95.percent of the total, in the high cost
scenario to 40, or 80 percent of the total, in the low cost scenario.
Most of the other closures will be at independents' facilities. Jobber
closures in Option 4 are the same in the high and the low scenarios,
representing 93 and 95 percent of the total closures in each respective
case.
6.2.3.4 Employment Displaced by Closures
The number of workers employed at the bulk plants which are closed
because of vapor control ranges from 0 to 43,000 (Table 6- 40). As many
as 550 workers, less than 1 percent of the 72,000 workers at gasoline
bulk plants, will be displaced by closures attributable to Option 1. Up
to 9,400 workers, or 13 percent of bulk plant employment, are impacted by
closures caused by Option 3 compliance while as many as 43,700 workers,
or 61 percent of those employed at gasoline bulk plants, may be displaced
by Option 4.
Because, on average, the labor force is not significantly different
by ownership classification for facilities of the same size, the employment
displaced by ownership will be proportional to the number of bulk plant
closures. Since the overwhelming majority of bulk plant closures will be
jobber operated, most of the employment impacts will also be jobber related.
The jobber employment displaced by the proposed control options is between
0 and 490 under Option 1, between 170 and 8,930 under Option 3, and
approximately 38,080 under Option 4.(using the less expensive incineration/
incineration costs) (Table 6-41 ).
6-89
-------
TABLE 6-40
EMPLOYMENT DISPLACED AT BULK PLANTS39
OPTION 1 OPTION 2 OPTION 4
BALANCE INCOMING BALANCE INCOMING REFRIGERATION/ INCINERATION/
COST SCENARIO TRANSPORTS ONLY & OUTGOING TRUCKS INCINERATION INCINERATION
NOJC Costs 550 9,440 43,730 41,140
Houston-Galveston Costs 550 1,020 43,100 40,420
10
° Colorado APCO Costs 0 210 42,710 40,080
-------
TABLE 6-41
ESTIMATED EMPLOYMENT IMPACT AT BULK PLANTS BY OWNERSHIP
40
OPTION 1
COST SCENARIO* HIGH LOW
Majors** - .
Independent
Marketer/Wholesalers 80
Jobbers 470
TOTAL 550 0
OPTION 3
HIGH LOW
170
340 40
8,930 170
9,440 210
OPTION
HIGH
2,300
760
38,080
41,140
4
LOW
1,450
550
38,080
40,080
* High impact = NOJC cost scenario
Low impact = Colorado APCD cost scenario
** Includes regional refiners
-------
6.2.3.5 Vapor Control Costs at Bulk Plants
The total cost of vapor control systems at bulk plants may cost up
to $750 million or produce a savings of $23 million depending upon the
control option, cost scenario and control technology chosen (Table 6- 42).
These costs include capital, financing and operating costs less any recovery
credits over the 10 year life of the vapor control equipment. The cost of
Option 1 compliance ranges from $37 million down to a savings of $23 million.
A cost savings is possible because the Colorado APCD cost scenario requires
less than half the capital of the other scenarios, but it produces the same
recovery credits as the more expensive scenarios. The cost of Option 3
compliances also varies from $376 million down to a savings of $6 million
while Option 4 compliance costs between $465 and $750 million. The more
expensive technology under Option 4 for each of the three cost scenarios
is the refrigeration/incineration system. The individual capital, financing
and operating costs, as well as any applicable recovery credits, are
presented in Tables 6-43 through 6-45 for each cost scenario.
Jobbers will bear most of the cost of vapor control, not only because
they would own most of the post-control bulk plants, but also because they
would generally own most of the smaller facilities. The vapor recovery
savings, via a recovery credit, is substantially less in small bulk plants
than in the larger facilities. The jobbers' share of the vapor control
costs will be $35.3 million, or 96 percent of the total, under the high
cost scenario of Option 1 (Table 6-46 ). Similarly, jobbers will account
for $14.2 million, or 63 percent, of the total savings produced by the low
scenario. For Option 3, the jobbers' share of the cost could be as high as $301.9
million, or 80 percent of the total cost. Under the low cost scenario
6-92
-------
TABLE 6-42
VAPOR CONTROL COSTS AT BULK PLANTS
(Million 1978 Dollars)
41
OPTION 1
OPTION 3
OPTION 4
CO
COST SCENARIO
NOJC COSTS
Bulk Plants Installing
Vapor Control
Total Vapor Control Cost
HOUSTON-GALVESTON COSTS
Bulk Plants Installing
Vapor Control
Total Vapor Control Cost
COLORADO APCD COSTS
Bulk Plants Installing
Vapor Control
Total Vapor Control Cost
BALANCE INCOMING
TRANSPORTS ONLY
14,120
36.9
14,120
36.9
14,250
(22.7)*
BALANCE INCOMING
& OUTGOING TRUCKS
12,030
375.5
14,010
154.8
14,200
(6.5)*
REFRIGERATION/
INCINERATION
3,960
747.3
4,110
698.3
4,190
656.0
INCINERATION/
INCINERATION
4,570
589.5
4,740
514.0
4,820
465.2
* Negative cost
-------
TABLE 6-43
en
i
10
Capital Investment
Financing (5 years)*
Operating Expense (10 years)*
Recovery Credit (10 years)*
Total Vapor Control Costs
VAPOR CONTROL COSTS AT
(Mill
OPTION 1
BALANCE INCOMING
TRANSPORTS ONLY
60.0
13.3
)* 26.0
(62.4)
BULK PLANTS BASED UPON
ion 1978 Dollars)
OPTION 3
BALANCE INCOMING
& OUTGOING TRUCKS
280.9
62.3
120.7
(88.4)
42
NOJC COSTS
OPTION
REFRIGERATION/
INCINERATION
468.4
103.9
290.7
(115.7)
4
INCINERATION/
INCINERATION
339.8
75.4
174.3
—
36.9
375.5
747.3
589.5
* Future cash streams discounted to present value. Discount rate = 10%
-------
TABLE 6-44
en
i
en
Capital Investment
Financing (5 years)*
Operating Expense (1
Recovery Credit (10 Years)*
Total Vapor Control Costs
43
VAPOR CONTROL COSTS AT BULK PLANTS BASED UPON HOUSTON-GALVESTON COSTS
(Million
OPTION 1
BALANCE INCOMING
TRANSPORTS ONLY
60.0
13.3
i Years)* 26.0
ears)* (62.4)
1978 Dollars)
OPTION 2
BALANCE INCOMING
& OUTGOING TRUCKS
153.1
34.0
66.0
(98.3)
OPTION
REFRIGERATION/
INCINERATION
434.8
96.5
286.2
(119.2)
4
INCINERATION/
INCINERATION
293.4
65.1
155.5
—
36.9
154.8
698.3
514.0
* Future cash streams discounted to present value. Discount rate = 10%.
-------
TABLE 6-45
CT>
I
CTl
Capital Investment
Financing (5 Years)*
Operating Expense (10 Years)*
Recovery Credit (10 Years)*
Total Vapor Control Costs
44
VAPOR CONTROL COSTS AT BULK PLANTS BASED UPON COLORADO APCD COSTS
(Million
OPTION 1
BALANCE INCOMING
TRANSPORTS ONLY
24.2
5.4
ears)* 10.5
rs)* (62.8)
1978 Dollars)
OPTION 3
BALANCE INCOMING
& OUTGOING TRUCKS
56.0
12.4
24.3
(99.2)
OPTION 4
REFRIGERATION/
INCINERATION
412.9
91.6
272.2
(120.7)
INCINERATION/
INCINERATION
263.5
58.5
143.2
(22.7)
( 6.5)
656.0
465.2
* Future cash streams discounted to present value. Discount rate = 10%
-------
TABLE 6-46
45
TOTAL COSTS* OF VAPOR CONTROL AT GASOLINE BULK PLANTS BY OWNERSHIP
(Million 1978 Dollars)
OPTION 1 OPTION 3 OPTION 4
COST SCENARIO** HIGH LOW HIGH LOW HIGH LOW
Majors*** 1.1 (7.1) 58.0 (7.5) 192.8 163.2
Independent
Marketer/Wholesalers 0.5 (1.4) 15.6 (1.1) 52.6 44.1
Jobbers 35.3 (14.2) 301.9 2.1 344,1 257.9
TOTAL 36.9 (22.7) 375.5 (6.5) 589.5 465,2
* Includes capital charge, financing cost and operating expense over life of control system
expressed in constant 1978 dollars.
** High impact = NOJC cost scenario
Low impact = Colorado APCD cost scenario
*** Includes regional refineries.
-------
for Option 3, however, the cost to jobbers will be $2.1 million while the
majors and independents realize a $7.5 and $1.1 million savings, respectively.
The total cost to jobbers under Option 4 is calculated to be $344.1 million,
58 percent of the total, in the high scenario and $257.9 million, 55 percent
of the total, in the low scenario. Both of these cost figures assume that
the less expensive control technology, i.e. incineration/incineration,
will be installed in order to comply with the control option.
6-98
-------
6.3 SERVICE STATIONS
6.3.1 Industry Characterization
6.3.1.1 Retail Service Stations
In 1977, there were approximately 178,400 retail service stations
in the U.S. which dispensed nearly 84.5 billion gallons of gasoline.
Over 48,000 service stations have closed in the U.S. since the population
peak of 226,000 in 1972. This attrition is expected to continue at least
through the early 1980's to a leveling off point of anywhere from
125,000 to 150,000 outlets. The economies of scale of high volume
stations and the shift to self-service operations are a prime factor
in shrinking retail margins. Consequently, the closure of outlets
due to market rationalization processes will be most severe for those
outlets which have relatively low sales volume coupled with high unit
expenses.
Retail service stations are supplied by various classes of suppliers.
The largest suppliers are the major oil companies, which directly supply
nearly 48 percent of the stations. These firms are the 17 largest oil
companies which are fully integrated and market gasoline in 21 or more
states. The next 21 largest oil companies are considered to be regional
refiner/marketers which tend to be partially integrated, operate at least
one refinery, and generally market gasoline in less than 21 states. These
companies supply about nine percent of the retail outlets. Another group
of suppliers is the independent marketer/wholesaler group, including
gasoline-oriented super jobbers. These suppliers, which are multi-state
6-99
-------
retailers but lack their own refining capability, furnish gasoline to
about 17 percent of the stations. The la.st direct supplier category is
the small jobber which generally markets gasoline under major oil company
brands through 6 to 12 service stations within a single state. There
are approximately 9,000 small gasoline jobbers in the U.S. which supply
almost 27 percent of the retail stations. 4^
A summary of the U.S. service station population by direct supplier
as well as by type of operation in various throughput ranges is presented
in Table 6-47.48
Service stations in the U.S. can broadly be classified into the
following four operational groups:
• Direct outlets (supplier operated)
• Convenience stores
• Lessee dealers
• Open dealers (dealer owned/dealer operated)
The traditional retail marketing strategy of the major oil companies
has been to operate stations through lessee dealers. These lessee outlets
represent approximately two-thirds of the major oil company stations and
about 47 percent of all the stations in the country. However, the proportion
of these types of stations is expected to decline as the marketing
strategy moves toward direct outlets, which are low expense, high volume
operations. Currently, direct outlets represent 18 percent of the total
U.S. outlets, with more than half of independent marketer/wholesaler and
super jobber outlets being directly operated and about 26 percent of the
49
regional refiner outlets being direct salary operation.
6-100
-------
Convenience store outlets have grown rapidly in the last few years and
represent aggressive gasoline competitors. While such outlets currently
account for only five percent of the retail station population, their
50
proportion is expected to increase significantly.
The second largest group of outlets is known as open dealers. In
these operations, the onsite dealer actually owns or controls the invest-
ment in his station where he is physically employed. The dealer is not
permanently tied to any particular brand, but "flies the flag" of the
supplier from which he can extract the best deal. Open dealer sites,
which tend to be older and more depreciated, represent about 30 percent
of the total stations in the country but have less than the national
51
average sales volume per outlet.
Retail service stations dispense an average of about 40,000 gallons
per month. In recent years, marketing economics have resulted in a trend
toward stations with larger volumes, with small volume operations being
marginal operations that have to rely on other parts of the retail trade,
such as mechanical work and sales of accessories, in order to remain in
business. The high volume stations tend to be mostly direct operations
which are controlled and operated by the supplier and operate on relatively
low margins. Low volume stations, those dispensing less than 25,000 gallons
per month, are mostly lessee dealers and open dealers supplied by all
classes of suppliers. These low volume stations, which comprise close
to 50 percent of the total number of stations, are the segment of the retail
industry that is most vulnerable to changes in marketing economics as well
as external costs such as vapor recovery costs.
6-101
-------
Table 6-47. SUMMARY OF SERVICE STATION POPULATION
,48
THROUGHPUT (000 gal/mo)
DIRECT SUPPLIER
MAJOR
Direct
"C" Store
Lessee
Open
SUBTOTAL
REGIONAL REFINER
Direct
"C" Store
Lessee
Open
SUBTOTAL
INDEP. MARKETERAVHOLESALEE
"SUPER JOBBER"
Direct
"C" Store
Lessee
Open
SUBTOTAL
SMALL JOBBER
Direct
"C" Store
Lessee
Open
SUBTOTAL
% Total Outlets
Total No. Outlets
% Total Annucl Volume
Total Annual Volume (MM gal/yr)
% OF TOTAL OUTLETS
<10
0.4
-
2.3
.
2.7%
_
-
0.6
-
0.6%
_
-
0.2
.
0.2%
_
.
0.6
0.4
1.0%
4.5%
8,100
1%
777.6
11-24
0.1
0.4
14.9
9.0
24.4%
0.1
0.1
1.3
0.4
1.9%
0.3
4.3
0.6
0.4
5.6%
0.5
0.6
4.3
3.4
8.8%
40.7%
72,650
22%
18,602.4
25-49
0.9
-
6.6
5.7
13.2%
0.5
-
1.9
0.6
3.0%
1.1
-
0.8
0.1
2.0%
1.0
.
4.7
7.3
13.0%
31.2%
55,740
30%
24,748.5
50-99
1.4
-
4.0
0.9
6.3%
1.1
.
1.3
0.1
2.5%
5.5
-
0.6
0.1
6.2%
1.1
.
1.4
1.2
3.7%
18.7%
33, 270
33%
28,252.8
>100
0.8
-
0.4
-
1.2%
0.6
-
0.2
-
0.8%
2.4
-
0.3
-
2.7%
0.2
.
-
-
0.2%
4.9%
8,630
14%
12,030.7
%
Total
3.6
0.4
28.2
15.6
47.8
2.3
0.1
5.3
1.1
8.8
9.3
4.3
2.5
O.G
16.7
2.8
0.6
10.9
12.3
26.7
100%
100%
Total
6,320
800
50, 260
27,890
85,270
4,010
200
9,420
2,030
15,660
16, 630
7,560
4,510
1,100
29,800
5,110
1,040
19,500
22,010
47,660
178,390
84,412.0
a) Direct: Company-controlled/company-operated
"C" STORES: Convenience stores
Lessee: Company-controlled/dealer-operated
Open: Dealer-controlled/dealer-operated
6-102
-------
6.3.1.2 Private Gasoline Dispensing Facilities
In addition to the retail service stations, there are a significant
number of facilities other than conventional retail stations which
dispense gasoline. The number and geographical distribution of private
dispensing facilities in the U.S. closely follows the pattern of service
stations. Private facilities are maintained by governmental,
commercial, and industrial consumers for their own fleet operations.
Miscellaneous retail outlets not classified as service stations include
marinas, parking garages, and rural businesses which sell gasoline as
a convenience to their customers rather than as a major source of income.
In 1977, there were an estimated 243,000 private locations in the country
co
which dispensed over 25 billion gallons of gasoline. However, only one
percent of these facilities dispense more than 20,000 gallons per month
since most have only one or two pumps. While these private facilities
account for 58 percent of the total gasoline dispensing outlets in the
country, they dispense only 23 percent of the total gasoline volume.
Table 6-48 indicates the breakdown of private dispensing facilities
CO
by end-use sector. The largest group in terms of gasoline consumed
is the trucking sector, which includes all non-government gasoline-powered
vehicles used in wholesale/retail delivery operations, as well as
miscellaneous services, construction, manufacturing, and extractive
industries. This segment consumes approximately five percent of the
total gasoline in the country and 21 percent of the total private
n . , 54
gasoline volume.
6-103
-------
Table 6-48. DISTRIBUTION OF PRIVATE GASOLINE DISPENSING OUTLETS 53
End-Use Sector
Agriculture
Trucking and
local service
Government
- Federal
- Military
- Other*
Taxis
School Busses
Miscellaneous**
Total Non-Service
Station Segment
Retail Service
Station Segment
All Segments —
Number of
"Private" Gasoline-
Dispensing Outlets
32,600
21,900
85,450
5,380
3,070
94, 530
242,930
178,390
421,320
Annual Gasoline
Consumption
(Million Gal)
3,801.3
5,241.6
227.6
174.1
2,266.4
882.1
144.7
12,497.2
25,235.0
84,412.0
109,647.0
% Total U. S.
Private
Gasoline
Volume
15%
21%
11%
0.
% Total U.S.
Gasoline
Volume
3%
5%
2%
3%
0.6%
9.0%
3%
1%
49%
100%
0.8%
0.1%
11%
23%
77%
100%
•State and municipal governments.
**Auto rental, utilities, and other.
6-104
-------
Another significant sector is agriculture related businesses. The
estimate of nearly 33,000 outlets nationwide for the agricultural sector
represents those outlets which have relatively large size tanks (greater
than 1,000 gallon capacity) on the farm and an average of three to five
trucks per farm. This would include all major farms and irrigation sites,
nurseries, and landscaping firms. Approximately 2.7 million farms in the
U.S. are not included in this estimate as they would typically have small,
above-ground tanks (e.g., 275-500 gallons) and would have a higher propor-
tion of diesel-fueled vehicles than of gasoline-powered equipment. In
general, all agriculture outlets would have less than 10,000 gallons
per month. ^
Government agencies with central garages are typically regional
locations for the postal service, Federal government agencies, and state
and county agencies. The central facilities typically dispense more than
10,000 gallons per month. There are over 85,000 of these facilities but
they dispense only two percent of the total nationwide volume of gasoline.
Other miscellaneous facilities include utility companies, taxi fleets,
rental car fleets, school buses, and corporate fleets. These sectors
combine for over 94,000 outlets that dispense around 11 percent of the
nationwide gasoline volume.
6-105
-------
6.3.2 Cost Analysis
6.3.2.1 Capital Costs
Little data are available on capital and installation costs of a
Stage I balance system alone since the system is normally installed
in conjunction with Stage II systems and may share some of the piping.
The earthwork and asphalt patching, if needed for Stage I, is also usually
all done at once at the service station. Due to the limited cost data, costs
in this section are presented as a range of costs which have been reported
by a number of sources. Costs are based on limited information from
vendors, oil companies, state and local agencies and other sources.
The capital cost of Stage I systems is dependent upon whether the
station can use the coaxial fitting that combines the filling tube and
the vapor return line into one piece of equipment. Some stations may have
problems with small openings in the tanks preventing use of the coaxial
fitting. In addition, the coaxial fitting prevents simultaneous filling
of tanks, so large throughput stations may want to remedy this problem by
manifolding the vapor return lines and not using the coaxial fitting.
If the coaxial fitting can be used, the installed cost per tank is $150 to
$250. No earthwork is needed since the fitting utilizes the existing
tank opening.
If a station cannot or choses not to use the coaxial fitting, then
earthwork and additional piping are needed. Separate vapor return lines
have to be added to each tank and the lines manifolded into one return
line at the surface. This cost is highly dependent upon the number and
5-106
-------
configuration of the underground tanks. Due to the fact that actual data
usually contain overlap between Stage I and Stage II installation, the
precise cost is not known. However, an estimated cost of additional
piping, trenching backfilling, and paving is estimated to range from
$1,000 to $1,500 per station. Thus, total capital cost for a manifolded
Stage I system, including tank hardware, is expected to range from $1300 to
$2000 per station.
These costs are consistent with other estimates of Stage I capital
costs that have been made. One source indicates that experience in
California has indicated that Stage I costs will range from $300 per tank
to $2000 per station, while another source reports that the cost sub-
mitted by two contractors for installing Stage I alone was $1,350.5?
Based on costs quoted by a Los Angeles contractor, another source reports
a Stage I capital cost of $1955, based on separate new fill and vapor
return.risers and associated hardware.58 An oil company reports that
hardware costs for Stage I will be over $200 per tank, with an installation
cost approaching $3000. This cost appears high and probably contains
much overlap with Stage II. The experience of another oil company at its
installations indicates that the cost of Stage I hardware was $526 with
contractor installation averaging $1411 for a total cost of $1937. Finally,
a consultant's report shows the installed cost of a coaxial system to be
$150 per tank based on conversations with an equipment vendor.
6.3.2.2 Operating and Maintenance Costs
There are no operating and maintenance costs associated with a Stage I
system since there are no mechanical or moving parts involved with the system.
6-107
-------
6.3»2.3 Annualized Costs of Control
Since there are no operating or maintenance costs involved with the
Stage I systems, the annualized costs represent the annualized capital
charges associated with the investment in the system. For purposes
of this analysis, the costs are annualized over a 10-year period based
on an interest rate of percent.
Table 6-49 summarizes the capital costs and annualized costs for
small, medium, and large service stations. Naturally, the costs per
gallon of throughput are higher for the small station, but the costs
are at most 0.10 cents per gallon for the most expensive manifolded
Stage I balance system.
6.3.2.4 Cost-Effectiveness
Based on the costs presented in Table 6-49 and estimates of annual
reductions in total benzene emissions in Table 2-1, the cost-effectiveness
of Stage I controls at service stations is presented in Table 6-50. For
the coaxial balance systems, the cost-effectiveness ranges from $2-4 per
kilogram of benzene controlled at large stations to $9-14 per kilogram
at small stations. For the manifolded systems, the cost-effectiveness
ranges from $5-6 per kilogram at large stations to $24-28 per kilogram
at small stations.
6*3.3 Service Station Impacts
6.3.3.1 Total Costs of Control
Based on the station costs summarized in Section 6.3.2 and the
estimates of the service station population presented in Section 6.3.1, the
6-108
-------
Table 6-49. SUMMARY OF COSTS FOR STAGE I BALANCE SYSTEMS
Monthly Throuahout
(liters)
75,700
227,100
454,200
(gallons)
20,000
60,000
120,000
No.
of
Tanks
3
4
5
Capital Cost
$450-750
$600-1000
$750-1250
Coaxial System
Annual ized Cost*
$75-125
$100-165
$125-205
-------
Table 6-50. Cost-Effectiveness Estimates for
Stage I Balance Systems
Monthly Throughput Coaxial System Manifolded System
(liters) ($/Kg Bz) ($/Kg Bz)
75,700 9-14 24-28
227,100 4-6 9-12
454,000 2-4 5-6
6-110
-------
total industry costs of installing Stage I equipment have been estimated.
The costs are summarized in Table 6-51. The costs are dependent upon
whether the industry uses a coaxial fitting for the system or whether the
storage tanks are instead manifolded. It is likely that the portions of
the industry will employ both techniques so that the actual costs will
fall between the costs of installing either system industry-wide,,
For the coaxial system, the total capital investment would
range from $213.6 million to $356.0 million, with the costs almost
evenly divided between retail outlets and non-retail facilities. While
there are more non-retail facilities than retail outlets, retail stations'
tend to have more underground storage tanks and thus higher investment costs.
Taking into account the financing costs, the total cost of control, expressed
as the discounted present value, would range from $257.4 million to $429.1
million.
The total costs for total installation of manifolded systems are two
to three times greater than the costs for the coaxial system. The capital
investment for the system would range from $563.7 million to $698.7 million.
From 60 to 70 percent of the investment would be incurred by the non-retail
sector. The total cost of control(discounted present value)of the manifolded
system would range from $679.4 million to $842.1 million.
In the retail sector, the distribution of costs by ownership class
closely parallels the distribuiton of service station ownership in the
industry. Open dealers and major oil companies will each incur about 30
percent of the total costs, while small jobbers will incur about 15 percent
of the costs. Other large independent marketers will account for the
remainder of the costs.
6-111
-------
Table 6-51. TOTAL SERVICE STATION INDUSTRY STAGE I COSTS
($ Millions)
Coaxial System Manifolded System
Retail Outlets
Capital Investment 104.3-173.8 247.9-334-3
Financing (5 years)* 21.4- 35.7 50.9- 68.6
Operating Expense 0 0
TOTAL COST* ' 125.7-209.5 298.8-402.9
Non-retail Outlets
Capital Investment 109.3-182.2 315.8-364.4
Financing (5 years)* 22.4- 37.4 64.8- 74.8
Operating Expense 0 0
TOTAL COST* 131.7-219.6 380.6-439.2
All Outlets
Capital Investment 213.6-356.0 563.7-698.7
TOTAL COST* 4 257.4-429.1 679.4-842.1
Future cash streams discounted to present value. Discount rate = 10%
6-112
-------
6.3.3.2 Potential Service Station Closures
A Stage I only control program is not expected to have a significant
impact upon incremental service station closures above those closed by
"normal" market factors without vapor recovery. The magnitude of the capital
investment is such that capital availability constraints for station owners
do not appear likely. As a worst case, the costs could reduce the
profitability of exceptionally marginal stations to the point that some
could not justify making even the limited investment and thus would
choose to close the station. One analysis estimates that at most 500
marginal stations could close as a result of Stage I over and above
those expecting to close due to market rationalization. These closures
would be concentrated in small leasse and open dealer stations. The
potential closures represent 0.4 percent of the estimated 1981 service
CO
station population.
It is also unlikely that Stage I costs would appreciably impact the
non-retail sector. Most firms in this sector have a large enough financial
base to be able to afford the equipment, which for these outlets will most
likely be the less expensive coaxial system. For marginal operations that
find investment in Stage I equipment to be unprofitable, the firms have
the option of purchasing gasoline at commercial service stations.
Furthermore, small agricultural outlets will not be affected by the control
requirements since nearly all have tanks less than 2000 gallons in size.
6.3.3.3 Potential Employment Displaced by Service Station Closures
If 500 stations were closed due to Stage I requirements, from 1,000
to 1,500 service station workers would be displaced. This is based on an
estimate of two to three employees, including the dealer, at small open
dealer and leasse dealer stations.
6-113
-------
6.4 REDUCTION OF BENZENE CONTENT IN GASOLINE
6.4.1 Petroleum Refining Industry Characterization
Crude petroleum is refined by 150 companies at 266 refineries located
in 40 different states. Production of refined products in the U.S. totalled
over 15 million barrels per day in 1976, or 93 percent of nameplate capacity.
The industry employs 100,000 workers and is heavily concentrated in the
West South Central region of Arkansas, Oklahoma, Texas, and Louisiana.
These four states employ 44 percent of all industry workers and supply
43 percent of all refined products.
The petroleum refining industry is somewhat concentrated. The five
leading producers own 36.5 percent of all industry capacity; the top ten,
58.5 percent. These leading producers are integrated, major oil 'companies
that engage in exploration, production, refining, distribution, and
marketing on the retail level. Other refiners are independent companies
that are typically not integrated into more than one other segment of
the industry. Prices vary little among companies, although there are
occasional examples of price cutting when there is weak demand and an
excess of supply.
6.4.2 National Costs of Benzene Removal
The costs presented in this section come from an analysis by Arthur
C"3
D. Little, Inc. The costs of benzene removal from reformates and FCC
gasoline were developed on a 1977 Gulf Coast basis. The main variable
affecting the costs of benzene removal from reformates and FCC gasoline
6-114
-------
was determined to be the total volume to fractionation, hydrogenation,
and extraction. Although extraction costs are somewhat dependent on
aromatics content, because of the greater dependence on total volume
to extraction, the costs were assumed to be independent of aromatics
content. The base case costs were scaled up on a regional basis by
capacity in order to get the national cost impact of benzene removal
in 1977 Gulf Coast dollars.
The national costs of benzene removal from reformates and FCC gasoline
is shown in Table 6-52. The capital requirement in 1977 dollars for
benzene removal from reformates is $2.0 billion, while the capital require-
ment for removal of benzene from FCC gasoline is $3.3 billion. The total
investment required to remove benzene from both reformate and FCC gasoline
is $5.3 billion. There would be some potential savings from economies
of scale through combining the reformates and FCC gasoline streams prior
to extraction.
The manufacturing costs to remove benzene from both reformates and FCC
gasoline are over $2.0 billion per year.* About 42 percent of these costs
are capital charges, 45 percent variable costs, and 13 percent for labor
and maintenance. The main component of variable operating costs is energy
*The annualized cost differs from that presented in the Arthur D. Little
report, which used a before tax capital recovery factor of 0.28. The
capital charges have been changed to reflect an after tax capital recovery
factor of 0.16 in order to be consistent with other costs presented in this
document.
6-115
-------
Table 6-52. National Cost of Benzene Removal From
Reformate & FCC Gasoline
Investment Costs: $ Billion
Process
Offsites
Total Plant
Other Capital
TOTAL CAPITAL
Manufacturing Costs: ($M/SD)
(345 SD/Yr)
Reformates
FCC
Gasoline Total
1.009
0.404
1.413
0.584
1.746
0.699
2.445
0.845
2.755
1.103
3.858
1.429
1.997
3.290
5.287
Variable Costs
Labor & Maintenance
Fixed Costs^
Total Manufacturing ($M/SD)
Total Manufacturing ($MM/Yr)^2^
Total Manufacturing (
-------
requirements for steam, fuel, and utilities. The total energy requirements are
54 million Crude Oil Equivalent (COE) barrels per year of $648 million per
year. Energy requirements amount to 70 percent of variable costs or 26 percent
of total operating costs.
The costs of removing benzene from gasoline were converted to costs per
barrel of gasoline using the 1981 estimated gasoline production of 7.45 million
barrels per day. The cost of removing benzene from reformates is 0.63 cents
per gallon of U.S. gasoline, and the cost of removing benzene from FCC gaso-
line is 1.17 cents per gallon of U.S. gasoline. The cost of benzene removal
from these two streams is 1.80 cents per gallon.
These costs are only for removal of benzene from reformates and FCC
gasoline, and do not include the costs of removing benzene from other
streams, or the costs associated with replacing lost octane, gasoline volume,
and benzene disposal.
The national cost of benzene removal from FCC gasoline was based on
producing hydrogen plant hydrogen at all locations with FCC unit capacity.
Some locations may have sufficient reformer hydrogen available at fuel
value. Since a detailed hydrogen balance at each location was beyond the
scope of this study, the sensitivity to hydrogen cost was developed. If
all locations were able to use refinery produced hydrogen at fuel value, the
total cost of benzene removal would drop from 1.8 to 1.6 cents per gallon
of U.S. gasoline. If the hydrogenation step were not required in the
removal of benzene from gasoline, the total cost of benzene removal would
drop from 1.8 to 1.2 cents per gallon of U.S. gasoline.
6-117
-------
The most important variable affecting the economics of benzene removal
is the unit capacity. The effect of capacity on benzene removal costs from
reformates and FCC gasoline is shown in Figure 6-9. The increased costs with
decreasing size results in a cost of benzene removal of up to 7 cents per
gallon of gasoline produced for the small refiner, as compared with the U.S.
average of 2.19 cents per gallon. In addition, the removal of benzene from
gasoline would have a greater affect on the small refiner's ability to blend
gasoline because of less operational flexibility and fewer blending stocks.
It is likely that some small refiners may not be able to remain in business
because of their significant cost differential and due to the high costs
associated with meeting gasoline lead phasedown regulations.
6.5 TOTAL COSTS OF GASOLINE MARKETING CONTROL OPTIONS
Table 6-53 presents a summary of the total costs for the four gasoline
marketing control options. The table indicates the total capital investment
costs, the annualized costs, and the total discounted costs that will be
incurred by the gasoline marketing petroleum industries.
Option 1 is the least costly option since there is less control at
bulk plants than with the other options. The difference between options 1 and
3 depends on the cost scenario assumption used for bulk plants. Assuming
use of the "least expensive equipment," capital costs between the two options
differ by only $32 million and annualized costs by $1.5 million. On the
other hand, with the "most expensive equipment," the differences are
$221 million in capital costs and $47 million in annualized costs.
Option 4 is the most expensive vapor recovery option, with capital costs
$60 million to $200 million greater than those of option 3.
6-118
-------
en
i
Reformates
FCC Gasoline
Reformates and FCC
Gasoline
40 50 60
Gasoline Capacity: MB'SD
70
80
90
100
0 10 20 30
Source: Arthur D. Little, Calculations.
Figure 6-9. Cost of Benzene Removal vs. Gasoline Production Using Refinery-Produced Hydrogen
-------
Option 2 is by far the most expensive control option, with capital
costs five to seven times greater than the other options and annualized
costs 10 to 20 times greater. The cost per gallon for option 2 would
be at least 1.8 cents, while the other options exhibit unit costs ranging
from 0.10 to 0.17 cent per gallon. Likewise, option 2 is not as cost-
effective as the other options, with a cost of $245 per kilogram of
benzene removed while the other options have a cost of $13 to $21 per kilogram
of benzene controlled.
The total costs do not give a complete indication of the differences
in economic impact between the options. The impact on closures of bulk
plants varies significantly between the options. As already discussed
in section 6.2.3.3, option 1 would result in at most 130 bulk plants going
out of business. The impact resulting from option 3 depends on the cost
scenario assumed, with only 50 bulk plants projected to close with the
"least expensive equipment" and up to 2,200 closures with the "most
expensive equipment." For option 4, the closures could amount to 9,000
to 10,000 bulk plants, or close to 50 percent of the population. Thus,
option 4 will have a..much more significant impact, which is not entirely
reflected in the total cost numbers for two reasons. First, the total
costs for option 4 only reflect the costs incurred for vapor recovery by
the bulk plants that remain in business. Secondly, the costs do not reflect
the monetary costs of the closures of bulk plants since it is difficult
to place a monetary value on the continued existence of a bulk plant.
6-120
-------
Table 6-53. TOTAL COSTS AND COST-EFFECTIVENESS OF GASOLINE MARKETING CONTROL OPTIONS
Bulk Terminals
Bulk Plants1
2
Service Stations
Tank Trucks
Refineries
TOTAL
*/Gal3
S/Kg of Benzene
controlled4
Option 1
Capital
Invest-
ment
($MM)
401.3
24.2-
60.0
284.8
34.3
0
. 744.6-
780.4
—
Annual ized
CostsS
($MM/Yr)
55.8
'(4.6)-
3.9
46.4
17.7
0
115.3
123.8
0.10-
0.11
18-19
Total Costs
Discounted
To Present
($MM)
473.2
(22.7J-36.
343.3
79.5
0
873.3-932.9
—
Option 2
Capital
Invest-
ment
(SMM)
0
J 0
. 0
0
5287.0
5287.0
--
Annual Ized
Costs
($MH/Yr)
0
0
0
0
2052.0
2052.0
1.8
245
Total Costs
Discounted
To Present
($HM)
0
0
0
0
13,106.0
13,106.0
1
Capital
Invest-
ment
(SMM)
401.3
56.0
280.9
284.8
34.3
0
776.4-
1,001.3
--
Option 3
Annual ized
Costs
($MM/Yr)
55.8
(3.1)-
51.0
46.4
17.7
0
116.8-
170.9
0.10
13-19
Total Costs
Discounted
To Present
(SMM)
473.2
(6.5J-375.5
343.3
79.5
0
889.5-1271
—
Option 4
Capital
Invest-
ment
(SMM)
401.3
263.5-
339.8
284.8
34.0
0
.5 983.9-
1060.2
—
Annua'iized
rCbsts
($MM/Yr)
55.8
66.3-
83.8
46.4
17.7
0
186.2-
203.7
0.16-
0.17
19-21
Total Costs
Discounted
To Present
(SMM)
473.2
465.2-
589.5
343.4
79.5
0
1361.2-
1485.5
—
5,
Range of bulk plant costs represent "least expensive equipment" and "most expensive equipment"
o
Service station costs represent average of costs for coaxial system for all outlets
3Based on estimated 1981 volume of 115 billion gallons.
Emission reduction estimates come from Table 4-2.
'Parentheses indicate net cost savings
-------
6.6 REFERENCES
1. Arthur D. Little, Inc. The Economic Impact of Vapor on the Bulk
Storage Industry, Report to Environmental Protection Agency,
Contract Mo. C-79911, July, 1978, Draft, page II.4.
2. Ibid., p. II.9.
3. Ibid., p. 11.10.
4. Ibid., p. 11.12.
5. Ibid., p. 11.13.
6. Ibid., p. 11.14.
7. Letter from J.A. Petrelli, Mobil Oil Corp., New York, N.Y., to
R. Southers, API, New York, N.Y. dated August 31, 1977.
8. Report of Fuel Requirements, Capital Cost and Operating Expense for
Catalytic and Thermal Afterburners CE Air Preheater, Stamford,
Conn. EPA-450/3-76-031 September 1976.
10. Personal communication between K. Sasseen, Trico-Superior, Los Angeles,
Calif, and John Pratapas, OAQPS, U.S. EPA on February 28, 1978 and
March 10, 1978 (CRA control systems and vapor holders).
11. Letter from T.G. Psyhojos, AER Corp., Ramsey, N.J. to John Pratapas,
OAQPS, U.S. EPA dated March 22, 1977 (Vapor incinerators (OX)).
12. Edwards Engineering Corp., Pompton Plains, N.J. Technical and Pricing
Information, May 1977 (Refrigeration (RF) control systems).
13. Memorandum from W. Polglase dated March 8, 1978 to John Pratapas
(OAQPS, U.S. EPA).
14. Sasseen, op.cit.
15. Investment itemizations for vapor control system installations supplied
by J.R. Sexton, Texaco, Inc., White Plains, N.Y. to John Pratapas,
OAQPS, U.S. EPA, January 19, 1978.
16. Arthur D. Little, Bulk Storage, p. VI-19.
17. Ibid., p. VI-21.
18. Ibid., p. VI-22.
6-122
-------
19. Ibid., p. VI-25.
20. Ibid., p. VI-26.
21. Ibid., p. III-4.
22. Ibid., p. 111-16.
23. Ibid., p. III-7.
24. Ibid., p. III-8.
25. Ibid., p. 111-10.
26. Ibid., p. III-ll.
27. Ibid., p. 111-15.
28. Ibid., p. 111-12.
29. McCormack, Bert W. and Bob L. Shuster, California Independent Oil
Marketers Association, letter to William F.Hamilton, Economic
Analysis Branch, Strategies and Air Standards Division, Office of
Air Quality Planning and Standards, U.S. Environmental Protection
Agency, February 28, 1978.
30. Evaluation of Top-Loading Vapor Balance Systems for Small Bulk
Plants, EPA-340/1-77-014, April 1977, pp. IV-19, IV-23, V-2.
31. Bassman, Bob, National Oil Jobbers Council, message to Bill Hamilton
and Bob Walsh, U.S. EPA, Research Triangle Park, N.C., March 20,
1978, attachment.
32. Joseph, David, U.S. EPA, Regional Office VIII, and Mark Parsons, Air
Pollution Control Division, Colorado Department of Health, Records of
Permit Applications, Air Pollution Control Division, Colorado
Department of Health, October 17, 1977.
33. Houghton, Clark F. Professional Engineer, Mid-Missouri Oil Company,
Eldon, Missouri, and "Marketing Trends", National Petroleum News
Issue, Mid-May 1976, p. 15.
34. Houghton, Clark F., op_ cjjt., March 1, 1978.
6-123
-------
REFERENCES
35. Ibid., p. VII-11.
36. Ibid., p. VII-16.
37. Ibid., p. VII-21.
38. Ibid.
39. Ibid., p. VII-21.
40. Ibid.
41. Ibid., p. VII-22.
42. Ibid., p. VII-23.
43. Ibid., p. VII-24.
44. Ibid., p. VII-25.
45. Ibid.
46. Arthur D. Little, Inc., The Economic Impact of Vapor Recovery
Regulations on the Service Station Industry. Report to Occupational
Safety and Health Administration, Contract No. C-79911, March, 1978,
p.32.
47. Ibid., pp. 31-32.
48. Ibid., p. 32.
49. Ibid., p. 34. .
50. Ibid., p. 36.
51. Ibid., pp. 36-37.
52. Ibid., p. 47.
53. Ibid.
54. Ibid., p. 45.
55. Ibid., p. 48.
6-124
-------
REFERENCES
56. Perry, Francis, California Air Resources Board, personal communication
to Kenneth Lloyd, U.S. EPA, Office of Air Quality Planning and Standards,
October 31, 1977.
57. Pacific Environmental Service, Inc.5 Hydrocarbon Control Strategies
for Gasoline Marketing Operations, Draft Final Report, EPA Contract
No. 68-02-2606, April 1978, p. 6-4.
58. Hasselman, D.E., Hasstech, Inc., letter to Peter Principe, EPA Mobile
Source Enforcement Division, undated (approximately summer 1977).
59. Coppoc, W.J., Texaco, Inc., letter to Roger Strelow, U.S. EPA,
Office of Air and Waste Management, January 31, 1977.
60. Snider, H.T., Sunmark Industries, letter to Roger Strelow, U.S. EPA,
Office of Air and Waste Management, January 31, 1977.
61. Arthur D. Little, Inc., Service Stations, Appendix Q, p. Q-8.
62. Ibid., p. 135.
63. Arthur D. Little, Inc. Cost of Benzene Reduction in Gasoline to the
Petroleum Refining Industry, April 1978, EPA-450/2-78-021.
6-125
-------
APPENDIX B
-------
MATRIX OF ENVIRONMENTAL IMPACTS OF ALTERNATIVE CONTROL TECHNIQUE CONFIGURATIONS
Alternative
I
II
III
IV
Delayed
Standard
No Standard
Impact on
Benzene
+2
+2
+3
+4
0
Impact
on HC
+3
0
+3
+4
0
Other Air
Impacts
-1
-3
-1
-2
Water
Impact
-1
-3
-1
-2
Solid Waste
Impact
-1
0
-1
-1
Energy
Impact
+4
-4
+4
+4
Air Quality
Impact
-
-
-
-
Space
Impact
-1
0
-1
-1
. . i
Noise
Impact
-1
0
-1
-1
DO
I
KEY + Beneficial Impact
- Adverse Impact
0 No impact
1 Negligible impact
2 Small impact
3 Moderate impact
4 Large impact
-------
APPENDIX C
Much of the data used throughout this document was obtained from
tests. This appendix briefly describes the test sites, the test methods,
and the results of those tests.
C.I BULK TERMINAL TESTS
This section of Appendix C summarizes and discusses bulk terminal
source tests that were conducted by EPA during the period from
November, 1973, to March, 1978. The purpose of the earlier tests (A-F)
was to evaluate the effectiveness of bulk terminal gasoline loading vapor
control systems in controlling total hydrocarbons. Later tests (G-K)
evaluated the effectiveness of the control equipment in controlling
total hydrocarbons and benzene.
The types of control systems tested included thermal oxidizer systems (TO);
compression-refrigeration-absorption systems (CRA); refrigeration
systems (RF); and an adsorption-absorption system (AA).
The various types of control systems tested were considered to
be representative of best available control for hydrocarbons in the
bulk terminal gasoline loading industry.
A brief discussion of each bulk terminal tested and conditions
during the test periods follows.
C-l
-------
C.I.I Bulk Terminal Test A
Test No. A was conducted at a bulk terminal that had an average
gasoline throughput of approximately 600,000 liters (160,000 gallons) per
day. The terminal was source tested by EPA from December 10-12, 1974.
The terminal has eight loading racks for various fuels. Gasoline is
dispensed from three of the racks. Each of the gasoline loading racks
are equipped for bottom loading of premium, regular, and unleaded gasoline.
Also, on one of the gasoline racks, two grades of aviation fuel are dis-
pensed and vapors are vented to the vapor control system.
Hydrocarbon vapors and air in the tank truck are displaced by the
gasoline loaded. The vapor air mixture vents to vapor return hoses at each
end of the racks. The vapor hoses are manifolded to a common header
venting to a saturator. Saturated vapors pass to a vapor holder. At a
preset volume the vapor holder automatically discharges to a 85,000 liters
per minute (300 cfm) compression-refrigeration-absorption (CRA) system.
The purpose of the saturator is to ensure that the hydrocarbon vapors
vented to the vapor holder are saturated with hydrocarbons and are above
the upper explosive limit.
Testing was performed during 39 truck loadings to determine potential
hydrocarbon emissions, actual hydrocarbon emissions and vapor recovery
efficiency of the system. Only two loading racks were tested. The other
rack was not used for loading purposes because insufficient test equipment
was available. Hydrocarbon emissions from the vapor recovery unit were
determined to be 31.2 milligrams per liter (0.118 grams per gallon) of gasoline
loaded into the tank trucks.
C-2
-------
The only difficulties in testing encountered in the loading of
gasoline into the tank trucks were vapor leakage and spillage from the
tank trucks. Vapor losses occurred at almost all hatches and pressure vents
at the top of the trucks. Leakage of emissions from the trucks were
estimated to be 115.2 milligrams per liter (0.554 grams per gallon).
Liquid spillage occurred on occasion because of improper seating of the
shut-off valve at the liquid connection to the tanker, and also from
buckets used to catch a small amount of unleaded gasoline left in the
tank truck compartments from its previous load. The loss due to leakage
can be estimated; but, the loss due to liquid spillage cannot. This
test was conducted only for total hydrocarbons.
Further details are presented in the emission test report.
C.I.2 Bulk Terminal Test B
Test No. B was conducted at a relatively small bulk terminal since
the facility has only one gasoline loading rack; however, the throughput
of the bottom-loading rack is approximately 380,000 liters (100,000 gallons)
per day. Three grades of gasoline (premium, regular, and unleaded) are
dispensed at the loading facility.
Vapors displaced from the gasoline tank trucks are vented to a
refrigeration-type vapor recovery system. During the test period by EPA,
which ran from December 17-19, 1974, twenty-four trucks were loaded with
gasoline to determine the potential hydrocarbon emissions, actual hydro-
carbon emissions and the vapor recovery efficiency of the vapor recovery
unit.
C-3
-------
In the refrigeration-type system, hydrocarbon vapors and air from the
tank trucks are directly processed and condensed in a double-pass finned
tube condenser associated with the vapor recovery unit. There are no
saturators or vapor holders utilized in the system. The efficiency of the
condenser is directly related to the temperature of the condensing unit.
In normal operation, a condenser of -73 C (-100°F) would be anticipated.
Moisture in the vent gases condense and collect as frost on the
finned-tube vapor condenser. Defrosting of the condenser is conducted
at periodic intervals; usually, once or twice a day. Defrosting is com-
pleted in 10 to 30 minutes depending on the amount of frost collected on
the finned-tubes.
During the test period there were no difficulties encountered in
the actual loading of the tankers; however, there was significant leakage
from the hatches and pressure vents on top of the tankers. The majority
of the fuel loaded during the test period was on independent carrier trucks.
Only two company owned trucks were loaded.
Operational problems associated with the vapor processing unit were
encountered however. A leak had developed in the high pressure portion of
the refrigeration system resulting in refrigerant loss. This resulted in
higher than design temperatures in the condenser. After repairs, testing
was conducted at the time the temperature of the condenser was approximately
-31°C (-60 F). As previously noted, the design operating temperature
of the condenser is -73°C (-100°F).
Recovered gasoline is separated from the condensed water and pumped to
storage. Condensed water vapor passes to a slop tank. Hydrocarbon emissions to
C-4
-------
the atmosphere from the recovery unit were determined to be 37.0 milligrams
per liter (0.140 grams per gallon) of gasoline loaded into the tank trucks.
Leakage from the trucks was estimated to be 100.9 milligrams per liter
(0.382 grams per gallon). This test was conducted only for total hydrocarbons,
2
Further details are presented in the test report.
C.I.3 Bulk Terminal Test C
This test was conducted at a medium size bulk terminal. The facility
consists of two loading racks. The bottom loading arms are situated on
a concrete island so that the trucks load countercurrently to each other.
Trained operators load the trucks. Throughput in the terminal is about
1,430,000 liters (378,000 gallons) of gasoline per day. (The plant operates
from 6'a.m. to 3. p.m., Monday through Saturday.) Trucks servicing both
Stage I and non-Stage I service stations are loaded at the terminal.
Trucks to be loaded carry gasoline vapor laden air. (The trucks have
capacities of 30,300-36,000 liters (8,000-9,500 gallons) each. As gasoline
is loaded, these vapors are displaced. A flexible hose is attached to the
vapor vent on the trucks and the vapors are vented to a control device--
in this case a refrigeration unit. The operation of this unit was described
under C.I.2.
The facility and refrigeration unit were tested for three days
(September 20-22, 1976). During all three days the refrigeration unit was
operating below capacity due to refrigerant loss which resulted from a
leaking pump seal. As a result the actual refrigeration temperature was
-44 to -52°C (-47 to -61°F) rather than the -73°C (-100°F) design temperature.
Hydrocarbon emissions from the vapor recovery unit were determined to be 33.6
milligrams per liter (0.127 grams per gallon) of gasoline loaded into the
tank trucks. Emissions due to leakage were estimated to be 86.7 milligrams
C-5
-------
per liter "(0.328 grams per gallon). This test was conducted only
for total hydrocarbons. Further details are presented in the emission
3
test report.
C.I.4 Bulk Terminal Test D
This tank truck gasoline loading terminal consists of four loading
racks loading 1,190,000 liters (315,000 gallons) of gasoline product per
day and numerous product storage tanks. The facility is attended for about
10 hours per day, but drivers have pass keys which permit loading 24 hours
per day, 7 days per week. Trucks servicing both Stage I and non-Stage I
service stations are located at the terminal. Testing was conducted
from September 23-25, 1976.
The vapor recovery system is a compression-refrigeration-absorption
unit. The system handles emissions from the loading rack and from storage
tank loading operations.
Gasoline vapors, collected from tank truck loading operations, are
first sprayed with gasoline to ensure that they are saturated (above the
explosive range). The vapors are then vented to a regular gasoline
product storage tank equipped with a lifter roof. When the roof reaches a pre-
determined level the vapors are vented to the CRA unit where the vapors
are sprayed with gasoline again (to saturate) and then compressed and cooled.
The vapors are then vented to an absorber where they are absorbed in fresh
gasoline and vented to atmosphere.
Throughout the test period, the unit operated with no apparent problems.
In addition to truck and CRA outlets being monitored, the liquid levels in
the storage tanks, the flow to the pipeline, and the liquid volumes into
C-6
-------
and out. of the CRA were monitored.
One problem seen was that drivers frequently drained trucks of remaining
gasoline into a sump before loading. This caused several liters of gasoline
to evaporate to atmosphere during the course of the test period. This loss
cannot be quantified. Hydrocarbon emissions from the vapor recovery unit were
determined to be 43.3 milligrams per liter (0.164 grams per gal Ion).of gasoline
loaded into the tank trucks. Leakage from the tank trucks was estimated to be
154.6 milligrams per liter (0.585 grams per gallon).
Trucks loading diesel fuel also hooked up to the vapor return line and
vented emissions to the saturator of the CRA. The test was conducted only
for total hydrocarbons. Further details are presented in the emission test
4
report.
C.I.5 Bulk Terminal Test E
Test No. E was conducted at a bulk terminal with a throughput of
approximately 1,100,000 liters (291,000 gallons) per day. The terminal has
two bottom-loading racks and one top-loading rack. Hydrocarbon vapors from
the-tank truck are vented through flexible connections to a common header
venting to a vapor holder and to the thermal oxidizer.
An EPA contractor conducted extensive tests on a thermal oxidizer
system at tank truck gasoline loading terminal E during the period
November 18, 1973, to May 2, 1974. Hydrocarbon vapors from tank
truck loading operations were vented to a vapor holder. The hydro-
carbon vapors were enriched with propane to ensure they were above the upper
explosive limit. The hydrocarbon vapors from the vaporsphere were then
vented to the thermal oxidizer for incineration.
The oxidizer is a simple, reliable gas furnace which turns on and
operates as needed; however, if it is necessary to shut down the oxidizer
during tank truck loadings and if the vaporsphere fills beyond its capacity
C-7
-------
of 283 cubic meters (10,000 cubic feet) or about 8 truck loads, excess
vapors would vent to the atmosphere.
Tests at the terminal during the test period indicated that the oxidizer
disposes of 99+ percent of the hydrocarbon vapor collected, even in extremely
cold weather when the air-gasoline vapor mixture is in the flammable range."
Although the oxidizer disposed of 99 percent of the gasoline vapor it re-
ceived, only about 70 percent of the air-vapor mixture displaced from the
2
truck loading reached the oxidizer. Unusually high pressures 53.3 g/cm
(21 inches of water) produced in the truck during loading were responsible
for the vapor loss through maladjusted hatch covers and faulty pressure-
vacuum relief valves on the trucks. A problem also existed causing low
vapor transfer and pressure build-up due to blockage of the vapor collection
line by a column of gasoline. These problems were partly corrected and the
overall disposal efficiency of the entire system now exceeds 90 percent.
Hydrocarbon emissions to the atmosphere from the thermal oxidizer are estimated
to be less than 1.32 milligrams per liter (0.1 grams per gallon) of gasoline
loaded into the tank trucks. Leakage from the truck was not quantified, but is
5
estimated to be 30 percent. Further details are presented in the test report.
C.I.6 Bulk Terminal Test F
This tank truck gasoline loading terminal consists of three bottom
loading racks. Throughput in the terminal is about 810,000 liters
(220,000 gallons) per day. Trucks servicing both Stage I and non-Stage I
service stations are loaded at the terminal. Testing was conducted from
November 10-12, 1976.
C-8
-------
Trucks to be loaded carry gasoline vapor laden air. The trucks have
capacities of 30,300-36,000 liters (8,000-9,500 gallons) each. As gasoline is
unloaded, these vapors are displaced. A flexible hose is attached to the vapor
vent on the trucks and the vapors are vented to a control device—in this case
a refrigeration unit. The operation of this unit has been described previously
under C.I.2.
The facility and refrigeration unit were tested for three days. During
all three days the refrigeration unit was operating at capacity. A valve on
the coolant return line (from the coils) was not opening properly and thus
return temperatures were higher than expected, but the problem was not significant.
Icing at the decanter, caused by ambient air leaking into the separator occurred
but did not cause any problems. Hydrocarbon emissions from the vapor recovery
unit were determined to be 62.6 milligrams per liter (0.237 grams per gallon)
Hydrocarbon leakage from the trucks was estimated to be 46.0 milligrams per liter
(0.174 grams per gallon). The test was conducted only for total hydrocarbons.
Further details are presented in the emission test report.
C.I.7 Bulk Terminal Test G
This company operates a small tank truck gasoline loading terminal
with a storage capacity of 3,600,000 liters (950,000 gallons) of gasoline and
a daily throughput of 284,000 liters (75,000 gallons) of gasoline. Barges
deliver the supply of gasoline to the terminal. There is no vapor recovery
system for the barge unloading operations other than the vapors retained
under floating roof storage tanks. Two truck racks employ five (5) bottom
loading positions, with vapor recovery lines leading to a carbon
C-9
-------
adsorption type vapor recovery unit. The vapor recovery system was in
good working order and appeared free from leaks.
Testing was performed May 25-27, 1977, during 33 tank truck loadings
to determine actual hydrocarbon emissions, potential hydrocarbon emissions,
and the vapor recovery efficiency of the system.
Hydrocarbons generated during bottom loading of tank trucks at the
terminal are collected by a vapor line collection system and vented to a
carbon adsorption and gasoline absorption vapor recovery system. Hydro-
carbons broke through the carbon beds on the first two days of testing.
Outlet concentrations from the unit were observed during these break-
throughs to be greater than 10 percent. The problems causing hydrocarbon
bed breakthrough were found and corrected before the third (final) day of
source testing. Hydrocarbon breakthroughs of the carbon beds were caused
by incorrect settings in electrical timer switching of the dual bed system.
In the system, one charcoal bed will remove gasoline vapors while the other
bed is being vacuum regenerated. After a period of time, the beds will switch.
The first day of testing, it was noted that the same bed was on line to
absorb vapors whenever a truck started loading. This improper setting of
the bed switching system caused an overload on one bed. The setting of the
bed switching system was corrected before the second test day. However,
some breakthrough was noted on the second day while the system was catching
up. No hydrocarbon breakthrough was noted on the third day. The improper
setting was due to the fact that the system was previously adjusted for
processing a low volume lean stream and during the test had to be
readjusted to operate on a high volume rich stream. Further details are
C-10
-------
presented in the emission test report. Hydrocarbon and benzene emissions
.•
from the vapor recovery unit were determined to be 30 and .003 milligrams
per liter of gasoline loaded into the tank trucks, respectively.
C.I.8 Bulk Terminal Test H
This tank truck gasoline loading terminal vapor control system was
previously source tested by EPA on December 10-12, 1974 (see C.I.I). The
vapor control unit, a CRA unit, was retested to determine the efficiency
of the unit in removing benzene from tank truck gasoline inlet vapors.
Testing was conducted on December 16, 1977. Integrated bag samples
were taken; 1) from the line between the vapor holder and the vapor control
unit, and 2) from the outlet of the vapor control unit. In addition, liquid
samples of regular, premium unleaded, AV gas-80 and AV gas-100 were obtained
for benzene analysis.
The integrated bag samples were drawn during the period when the vapor
recovery unit was in operation. During the test period the vapor holder
vented to the vapor control system six times. A total of 24 tank trucks
were loaded during this period and meter readings were taken at the loading
rack for each gasoline product loaded. A turbine meter was utilized to
measure the exhaust volume from the control unit.
During the test period the three loading racks as well as the vapor
control system appeared to be in normal operation. The vapor holder would
fill with vapors until the height of the diaphragm reached approximately
3.18 meters (10 feet). This height would actuate the vapor
control unit and hydrocarbon vapors would vent to the system until the
vapor holder diaphragm was drawn down to approximately 1.55 meters
(5 feet). In some instances, trucks were loaded while the vapor
recovery unit was in operation.
C-ll
-------
During the vapor recovery unit operation the absorber pressure was
3.52 kg/cm (50 psig) and the temperature was -16.6°C (2°F). This is normal
operation for the unit. Liquid gasoline temperature at the loading rack
varied from an estimated 1.1°C to 9.9°C (34°F to 50°F)jduring the testing
period.
During the fourth cycle of the vapor holder, it was noted that the
vapor recovery unit inlet sampling line had a small hole in it. Testing
was stopped, the small hole was repaired, and testing was conducted during
two additional vapor holder cycles. All bag samples collected were pro-
cessed within a short time in the testing contractor's mobile van which
was parked at the site.
Testing of this facility gave the efficiency of the vapor
recovery unit in removing hydrocarbon and benzene from vapors vented from
the vapor holder at the site. No relation can be made to tank truck emissions
since a saturator is included in the system between the tank trucks and the
vapor holder. Hydrocarbon and benzene emissions from the vapor recovery
unit were determined to be 41.1 and .106 milligrams per liter of gasoline
loaded into the tank trucks, respectively. Further details are presented in
o
the test report.
C.I.9 Bulk Terminal Test I
This tank truck gasoline loading terminal was selected for source testing
because the loading facilities are vented directly to a thermal oxidizer. The
other thermal oxidizer unit source tested by EPA was equipped with a vapor
holder that allowed only vapors above the vapor explosive limit to be vented
to the unit. (See Report No. EPA-650/2-75-042, June, 1975.)
The terminal is equipped with three gasoline loading rack positions
(No. 9, 7, and 5). Regular, premium, and unleaded gasolines are loaded at
each of these racks. At the No.9 loading rack, the tank truck vapor vent
C-12
-------
line was connected to a turbine meter to quantitatively measure the volume
of vapors vented from the tank truck to the vapor control system. Integrated
bag samples.of vent gases from the trucks were taken at this point and tank
trucks loaded were monitored for leaks. A liquid sample for each type of
gasoline loaded was also obtained for analysis.
During the test period the terminal appeared to be in normal operation
and the thermal oxidizer appeared to be operating properly. It was stated
that the daily throughput of gasoline approximated 757,000 to 1,135,500 liters
(200,000 -300,000 gallons) of gasoline. To ensure that a sufficient number nf
tank trucks were monitored, most of the trucks were loaded at the No. 9 rack.
The gallons loaded for each rack, temperature of product and the date
are continuously recorded in the terminal office. Six trucks were monitored
the first day, fifteen the second, and ten on the third day.
The test appeared to have been conducted in a satisfactory manner.
The possibility exists that due to low temperature conditions, the vapors
vented to the thermal oxidizer unit in some instances may have been below
the lower explosive limit and could have passed through the thermal oxidizer
without being incinerated. Hydrocarbon and benzene emissions to the atmosphere
from the thermal oxidizer were determined to be 34.2 and .330 milligrams
per liter of gasoline loaded into the tank trucks, respectively. Further details
Q
are presented in the emission test report.
C.I.10 Bulk Terminal Test J
This tank truck gasoline loading terminal vapor control system was
previously source tested by EPA on November 10-12, 1976 (see C.I.6). The
vapor control unit was retested to determine the efficiency of the unit in
removing benzene from tank truck gasoline loading inlet vapors.
C-13
-------
Testing was conducted on March 7, 1978. Integrated bag samples were
taken; (1) from the vapor line from the tank trucks, and (2) from the outlet
of the vapor control unit. In addition, liquid samples of the gas'oline
from the vapor recovery unit were determined to be 53.4 and 0.052 milligrams
of gasoline loaded into the tank trucks, respectively. The operation of the
terminal is discussed in C.I.6. Further details are presented in the emission
test report.
C.I.11 Bulk Terminal Test K
This tank truck gasoline loading terminal vapor control system was
source tested by EPA on May 1-5, 1978. The terminal has a gasoline
throughput that approximates 1,000,000 liters per day.
The vapor control system at this plant is similar to that described
in Test No. H. The unit was tested to determine the efficiency of the unit
in removing benzene from tank truck gasoline loading vapors.
The total hydrocarbon concentration, at both the inlet and outlet
of the vapor recovery unit, was continuously monitored, the vapor volumes
were determined at these two sampling points and bag samples were collected
at each sampling point for analysis of benzene using gas chromatography.
Measurement of the li.quid volume percent of benzene in the different grades
of gasoline was also performed during this test.
Test results reported are based on preliminary data and indicate that the
benzene concentration in the control system vent averages 18.5 ppm. Inlet
concentrations average 920 ppm. The benzene removal efficiency averages
98.5 percent. It would appear that this type of compression-refrigeration-
absorption (CRA) unit will effectively remove benzene from gasoline vapors
generated during tank truck loading operations.
C-14
-------
C-l. SUMMARY OF BULK TERMINAL GASOLINE LOADING VAPOR CONTROL DEVICES SOURCE TESTED BY EPA
f
TYPE OF VAPOR
CONTROL SYSTEM (VRS)*
1. VRS Inlet-HC, mg/1
2. VRS Outlet HC, mg/1
3. VRS Inlet, BZ, mg/1
4. VRS Outlet BZ, mg/1
5. VRS HC Effic. (%)
6. VRS BZ Effic. (%)
7. VRS Outlet HC Cone.
(Vol %)
8. VRS Outlet BZ Cone.
(Vol X)
9. Benzene Content of
Liquid Gasoline
REGULAR
PREMIUM
UNLEADED
AV-80
AV-100
A
CRA
107.3
31.2
-
-
70.9
-
-
B
RF
236.7
37.0
-
-
84.4
-
-
C
RF
486.9
33.6
-
-
93.1
-
-
D
CRA
554.0
43.3
-
-
92.1
-
-
E
**
TO
_
-
-
-
99+
-
.0001 to
.0045
F
RF
318.9
62.6
-
-
80.4
-
G
AA
684
30
2.51
0.003
95.9
99+
0.21
0.000^
H
CRA
447
41.1
2.45
0.106
91.0
96.0
4.16
0.006
1.81
1.28
2.49
0.36
1.06
I
***
TO
368
34.2
1.68
0.33
91.0
81.0
1.64
1.87
1.92
-
~*
J
RF
203
53.4
0.992
0.052
73.0
95.0
4.0
0.001
1.00
0.68
1.04
•
~
K
CRA
998
53.6
4.51
0.07
93
98.5
2.6
.002
0.64
0.59
0.39
-
"
* CRA - Compression-Refrigeration-Absorption
RF - Refrigeration
TO - Thermal Oxidizer
AA - Adsorption-Absorption
**
***
See EPA Test Report, EPA-650/2-75-042, June, 1975.
Thermal Oxidizer with Vapor Holder
Thermal Oxidizer without Vapor Holder
-------
C.2 BULK PLANT TESTS
Pacific Environmental Services (PES), under EPA contract, conducted
12
hydrocarbon efficiency testing of vapor recovery systems installed at
bulk plants. Two installations were studied; one (Plant A) employed a
vapor balance system modified by refrigeration to maintain a reduced
temperature in the storage tanks, and the other (Plant B) employed a vapor
balance system without secondary vapor recovery.
Efficiency testing was done by measuring amounts of liquid gasoline
transferred and of gasoline vapor retrieved during transfer of gasoline into
and out of the storage tanks. Efficiency was defined as the ratio of vapor
retrieved to a theoretical estimate of the amount which would be lost during
transfer if emissions were uncontrolled.
C.2.1 Plant A - Description and Operation
The vapor recovery system installed at Plant A employs a refrigeration
unit to reduce pressure in the storage tanks and thereby to minimize venting.
In this system, vapors are drawn from the storage tanks by a blower, pass
over cooling coils in the refrigeration unit and exhaust back to the storage
tanks through an insulated return line. The system makes no effort to condense
vapors but is designed strictly to maintain a constant temperature in the
storage tanks (in this case 16°C) and thereby maintain a pressure below the
venting level. The system is actuated when the storage tank pressure reaches
2
748 N/m (3 in. H20) and continues to operate until the pressure falls below
C-16
-------
p
374 N/m (1.5 in. hLO) or until 20 minutes have elapsed. If after 20 minutes
o
the pressure has not decreased to 374 N/m the system is actuated again and
runs for another 20 minutes. This cycle continues until the pressure falls
p
below the set level of 374 N/m .
Plant A incorporates a 7.6 cm vapor return line manifolded to all
tanks handling gasoline which includes the insulated line that runs from
the refrigeration unit back to the storage tanks. Separate vapor return
connections are used for the delivery of gasoline to the bulk plant and the
loading racks for dispensing gasoline. At each location the vapor return
connection was sealed with a spring activated valve. A series of safety
vents similar to those described for Plant B are in the storage
tank system. The bulk plant also has four gasoline service station type pumps
connected into the vapor recovery system. The same blower which is used
for the storage tank refrigeration system is also used to supply vacuum .
assist at the nozzle of these pumps and is activated when the dispensing
pump is started.
C.2.2 Testing of Plant A
The testing of Plant A was treated as a vapor balance system. The
refrigeration system at this facility does not condense vapors but is
designed strictly to maintain a constant temperature (16 C) in the storage
tanks and thereby maintain a pressure below the vent level, thus decreasing
both breathing and working losses. This design made a direct evaluation of the
refrigeration system impossible since it was difficult to relate its
operation as being independent of the vapor balance system.
C-17
-------
Seven transport deliveries were tested at Plant A and an average
volumetric efficiency of 97.0 was obtained, based on five deliveries. The
average loss in volume (theoretical minus standard) was estimated as one
cubic meter. For the account truck tests, the average efficiency would
be. a rather misleading number since various factors have to be considered,
such as the type of account truck, system pressure and ambient temperature.
Each account truck has its own characteristics (i.e., hatch leakage,
capacity, etc.) and these all vary from one truck to another. Average
volumetric efficiencies for the trucks ranged from 58 to 94 percent.
An average concentration of 30 percent by volume as propane (20 percent by
volume as hydrocarbons) was found in the vapor return line during delivery
of gasoline by transport truck. Twenty-five percent by volume as propane
(17 percent by volume as hydrocarbons) was found during loadings of the
account trucks.
C. 2.3 Plant B - Description and Operation
The vapor recovery system installed at Plant B employs a vaoor balance
system which operates on the principle of a simple exchange of vapors
between the truck tank and the storage tanks. The liquid gasoline is pumped
from the incoming tank truck into the storage tanks and displaces an equivalent
volume of vapor-laden air which is routed back to the truck tank through the
vapor line. When loading delivery tank trucks with gasoline, the vapor-laden
air in the delivery tank trucks is displaced back into the storage tanks through
the vapor return lines.
The vapor balance system incorporates a 5 cm vapor return line
which is manifolded to each of the five storage tanks handling gasoline. A
spring actuated poppet valve is at the loading rack outlet of the
vaoor return line to eliminate hydrocarbon losses when the plant is idle.
C-18
-------
Plant B is also designed to utilize one vapor return hook-up for both
transport deliveries and dispensing product into delivery trucks. This is
accomplished through a series of valves enabling the loading rack pumps
to be used for pumping in either direction. The system also incorporates a
series of pressure relief vents installed in the storage tank system. A,
2
pressure vacuum (PV) vent is located in the vapor line with a 2586 N/m
2
(10 in. of water) pressure setting and a 2.5 N/m (0.9 in. of water) vacuum
setting to allow the release of vapors or the entrance of air into the
system during severe pressure changes in the loading or unloading operations.
If the system pressure continues to increase and the PV vent cannot
allow the escape of vapors quickly enough, each storage tank has an emergency
O
vent to open at 5309 N/m (17 in. of water) of pressure. If under extreme
conditions these vents could not relieve the system pressure, a series of
2
carbon shear pins and hatch covers will open between 13,790 N/m and
n
20,680 N/m (55 and 83 inches of water) pressure. These final two steps
in pressure relief are installed primarily as safety features if the PV
vent cannot handle the pressure load.
C.2.4 Testing of Plant B
The testing at Plant B was less complex, mainly because Plant B
had a much smaller throughput (13,000 liters/day for Plant B as compared to
50,000 liters/day for Plant A). Five tank truck deliveries and eleven
delivery truck loadings were tested. The average volumetric efficiencies
for truck deliveries were found to be 95 percent. Consistent readings
were obtained indicating that leaks were minimal. An average concentration
of 45 percent by volume propane (29 percent by volume as hydrocarbons) was
found in the vapor return line.
C-19
-------
The efficiencies for the account trucks ranged from 79 to 97 percent.
This wide range of efficiencies was due to the leaks present in the account
trucks and bulk storage tanks. An average concentration of 38 percent by
volume as propane (26 percent by volume as hydrocarbons) was found in the
vapor return lines.
C.2.5 Conclusions
As a result of tests performed on vapor recovery installations at two
gasoline bulk plants, the following conclusions are reached:
1. Vapor balance systems, with or without associated refrigeration
for cooling storage tanks, can control vapor emissions during delivery of
gasoline by tank trucks with efficiency greater than 90 percent. In
all of ten such transfers observed in this study, the volume efficiency
observed ranged from 90 to 100 percent.
2. Vapor balance systems, with or without associated refrigeration,
can control vapor emissions during loading of delivery trucks with overall
volumtetric efficiency greater than 90 percent. In twelve of thirty such
transfers observed in this study, the volumetric efficiency observed ranged
from 90 to 100 percent.
3. The tests performed yielded no evidence that the secondary system
employed at one bulk plant provided better emission control than the
unassisted system employed at the other plant. Minimum observed volumetric
efficiencies in loading of delivery trucks were 43 percent with the refrigeration
system as compared with 79 percent for the unassisted vapor balance system.
4. The efficiency attainable in loading account trucks appears to
depend markedly on the condition of hatches and seals, and on the degree
of care exercised in making connections. At Plant A, four delivery trucks
COf^
— I--.'
-------
were used; during loading, the two newer trucks had consistently lower
emissions than the oldest truck, but none of the four consistently showed
control volume efficiency as high as 90 percent.
5. Venting of a storage tank occurred once at each of the plants
during the period of testing. The venting of the tank at Plant A, with
the refrigeration system, released only a negligible amount of vapor,
unmeasurable with the study equipment. The venting from the tank at
Plant B, however, continued for about an hour and released an estimated
7 cubic meters (250 cubic feet) of gas containing about 25 percent hydro-
carbons. (This would be roughly equivalent to about 7 liters of liquid
gasoline, or about two gallons.)
6. The average molecular weight of hydrocarbon vapors recovered,
as indicated by gas chromatographic analysis, was about 64 (intermediate
between butane and pentane).
7. Reid Vapor Pressure measurements of the liquid gasoline transferred
indicated that gases emitted during liquid transfer at Plant A were typically
not saturated with gasoline vapor, whereas those emitted during transfer at
Plant B were near saturation. This difference is possibly attributable to
the effect of the refrigeration unit at Plant A.
C.3 SERVICE STATION TESTS
In June of 1974, EPA tested five service stations during bulk deliveries.
Two of these stations were equipped with balance systems for both refueling
C-21
-------
of automobiles and bulk drops. Three of the stations had balance systems
for bulk drops with excess vapors being treated in secondary processors
on the underground storage tank vents. The secondary processors were part
of systems used to control vapors from automobile refueling. All systems
were installed to comply with local hydrocarbon air pollution control
regulations.
As previously discussed in this document, balance system control
efficiency is equivalent for hydrocarbons and benzene emissions. Thus, the
performance of these five systems for hydrocarbon reduction demonstrates
the performance of the systems for benzene reduction. (The effect
of the secondary processors on benzene emissions was not established.
However, since the processors handled only excess vapors from the system,
it is expected that the three systems performed as well, if not better,
than the two straight balance systems. Further, while the two straight
balance systems were used in conjunction with balance systems on refueling,
the system efficiency is not expected to be any different than balance systems
unassociated with refueling controls.)
C.3.1 Service Station A
Station A was tested during a bulk drop of 33,000 liters (8250 gallons)
on June 12, 1974. The station employed a balance system on both vehicle
refueling operations and bulk deliveries. The station pumped about
115,000 liters (30,000 gallons) of gasoline per month based on the average
C-22
-------
gasoline pumped for three grades of gasoline during the test period.
Underground storage tank vents were manifolded by underground piping
to a common connection to which the delivery truck attached a single vapor
return hose. The hose had been attached to the truck vapor connector.
With such manifolded underground piping, it is possible to load more than
one storage tank at a time.
The truck unloaded four compartments of gasoline simultaneously
into two tanks holding regular and premium gasoline. The drop took
40 minutes from arrival at station to completion of the drop. The actual
unloading took 20 minutes (1220-1240). Based on a comparison of the
volume of air/vapor vented to the volume of air/vapor displaced, the
system achieved 97.6 volume percent efficiency. The mass rate was 8 mg
HC/liter of gasoline dropped. The benzene rate would approximate
0.07 mg/liter. Table C-2 summarizes these data.
C.3.2 Service Station B
Station B also employed a balance system for vehicle and bulk drop
sources. The system was tested on June 18, 1974, and the data indicated
that throughput approximated 77,000 liters (20,000 gallons) per month.
The underground storage tank vapor lines were not manifolded, so only a
single drop could be made at a time. The test took place during the loading
of 18,000 liters (4665 gallons) of gasoline around 1030. The volume
efficiency was 96.2 percent and hydrocarbon mass rate was 10 mg/liter
of gasoline dropped. Benzene mass rate would approximate 0.08 mg/liter.
C.3.3 Service Station C
Station C was tested during a bulk drop on June 21, 1974. The
station was equipped with a balance system and secondary processor
C-23
-------
(in this case an incinerator). Station throughput was estimated as
135,000 liters (35,000 gallons) per month. Underground storage tank vent
lines were manifolded and the 30,000 liter (7800 gallon) delivery took
about 35 minutes to complete for the two tanks which were loaded. Volume
efficiency was shown to be over 99 percent and the hydrocarbon mass emission
rate was 0.5 mg HC/liter of gasoline dropped. This would convert to
approximately 0.004 mg BZ/liter. Efficiency is high and mass rate is low
because excess vapors were incinerated in the system (installed to control
vehicle refueling losses).
C.3.4 Service Station D
Station D employed a balance system with a refrigeration secondary
processor. The station, which was tested on June 7, 1974, had an estimated
throughput of 289,000 liters (75,000 gallons) per month.
Venting of the storage tanks was manifolded and thus the truck
driver unloaded two compartments at a time. The total delivery totalled
33,000 liters (8600 gallons) and took 20 minutes. The total time from
arrival at the station to completion of load was about 45 minutes.
Volume percent efficiency was over 99 percent. Hydrocarbon emissions
were about 0.9 mg/liter. This converts to about 0.007 mg BZ/liter
of gasoline.
C.3.5 Service Station E
Station E was tested on June 25, 1974. The station employed a
balance system with a refrigeration/adsorption secondary processor.
Station throughput approximated 115,000 liters (30,000 gallons) per
month.
C-24
-------
Storage tank vents were manifolded so that two compartments were
unloaded simultaneously. (Four compartments totalling 34,000 liters
were unloaded during the test period which lasted about 10 minutes--
1015 to 1025.)
The processor vent did not exhaust during the load, indicating that
the system approached 100 percent efficiency. However, the test report
notes that a leak in an underground pipe was located and may have vented
during the test. Thus hydrocarbon and benzene emission rates are uncertain
in this test.
C.4 DETERMINATION OF BENZENE TO HYDROCARBON RATIO FOR GASOLINE VAPOR
Many attempts were made by EPA to theoretically predict benzene/
gasoline vapor-liquid equilibrium. None of these efforts conclusively
collaborated the test data on hand, presumably because of the number of
complicating influential parameters. Thus we have decided to rely entirely
upon the actual test data to supply the necessary information. Test data
were available from three sources; Colonial Pipeline, Gulf Oil (Runion), and
Shell Oil Company, as discussed below.
Field sampling and analyses were conducted at the Colonial Pipeline
14
Company tank farm in Greensboro, North Carolina, in September, 1977. These
tests were performed to determine the extent of saturation in the vapors
under floating roof tank seals, and to help establish the relationship
between benzene liquid and vapor concentrations for gasoline.
In the Colonial Pipeline test, three samples were drawn from the vapor
space on each of nine gasoline tanks. Two liquid samples were taken from
each tank. Of the nine tanks tested, one contained unleaded premium
gasoline, three contained leaded premium gasoline, two held unleaded regular
gasoline, and the remaining three tanks stored leaded regular gasoline.
C-25
-------
TABLE C-2. SERVICE STATION BULK DROP RATE
STATION
A
B
C
D
E
Estimated
Monthly
Throughput
(liters)
115,000
(30,000 gal)
77,000
(20,000 gal)
135,000
(35,000 gal)
289,000
(75,000 gal)
115,000
(30,000 gal)
Drop
(Liters)
32,00.0 liter
(8250 gal)
18,000
(4665 gal)
30,000
(7800 gal)
33,000
(8685 gal)
34,000
(8800 gal)
Volume
Efficiency
s 97.6
96.2
99+
99+
—
Mass
HC
Emissions
(gin)
260
183
14.5
28.5
—
Mass
HC
(mg/liter)
8
10
0.5
0.9
—
Estimated
BZ
(mg/liter)
0.07
0.08
0.004
0.007 '
—
-------
The results of these analyses are presented in Table C-3, along with
previous data obtained by Gulf Oil Corporation and Shell Oil Company.
The benzene vapor concentrations (ppm and gm benzene/gm hydrocarbon) were
calculated directly from gas chromatography analyses and averaged for each
tank in the Colonial Pipeline test.
The paper by Runion (Gulf Oil) presented benzene vapor concentrations
as air-free vapor volume percent benzene. Thus, the ppm quoted in Table C-3
for the Runion work is converted to benzene in gasoline/air vapor by assuming
46 percent hydrocarbon in the vapor. (Forty-six percent hydrocarbon in the
vapor is the average of all the Colonial Pipeline tests.) The gm benzene/.
gm hydrocarbon for the Runion tests were estimated by:
(Vapor volume % benzene) (MWB) = gm benzene
[1 - (Vapor volume % benzene)J (MWV) gm hydrocarbon
Where the average molecular weight of the vapor(MWV) was also averaged
from the Colonial Pipeline test data.
Similarly, assumptions were made in estimating the gm benzene/gm
hydrocarbon from Shell Oil Company work. Because the Shell work only
gave an average vapor benzene concentration, it was necessary to assume
that the average liquid gasoline in the 86 Shell tests was about one liquid
volume percent.
Data from Colonial Pipeline, Gulf, and Shell were then plotted as gram
benzene/gram hydrocarbon versus gasoline liquid volume percent benzene to
yield Figure 2-2. A least squares analysis of the data provided the best
linear fit. At the current national average of 1.3 liquid volume percent
benzene, the least square analysis predicts 0.0078 (rounded to 0.008)
grams benzene/gram hydrocarbon, which is used throughout this document.
C-27
-------
TABLE C-3. COLONIAL PIPELINE
TANK NO. GASOLINE
LIQUID VOLUME
PERCENT BENZENE
LIQUID VAPOR
TEMPERATURE(°F)
VAPOR ANALYSIS
PPM BENZENE
gm BENZENE / gm HYDROCARBON
310
818
819
821
822
824
837
840
844
UNLEADED
PREMIUM
UNLEADED
REGULAR
LEADED
PREMIUM
UNLEADED
REGULAR
LEADED
PREMIUM
LEADED
PREMIUM
LEADED
REGULAR
LEADED .
REGULAR
LEADED
REGULAR
1.23
1.33
1.41
1.19
1.02
. 0.819
1.48
1.07
1.23 i
82 / •"
A
87^
84...
81 _
84_
83. '
86 -•
i
83
82 ,-
3600
3900
3900
3000
4100
2600
4200
3000
2900
0.007
0.009
o.on
0.007
0.009
0.006
0.010
c,:.o '
0.008
ro
oo
-------
C-3. COLONIAL PIPELINE (Cont'd)
SOURCE
RUNION
Low Octane Regular
Leaded Regular
Unleaded Regular
SHELL
(Average of 86 samples)
EPA Hackensack Test
LIQUID VOLUME
PERCENT BENZENE
0.85
1.22
1.10
1.0
TEMPERATURE
78
78
78
_
VAPOR
PPM
4600
4400
3700
7000
ANALYSIS
gm/BENZENE / gm HYDROCARBON
0.006
0.005
0.005
0.009 I
0.009
o
ro
10
-------
Additional testing by EPA at a bulk loading terminal in Hackensack,
New Jersey, confirmed that the average weight fraction of benzene in
the vapors displaced during gasoline loading was about 0.009.
C.5 REFERENCES
1. Test No. A, EMB Project No. 75-GAS-10, EPA Contract No. 68-02-1407,
Task No. 7, September, 1975.
2. Test No. B, EMB Project No. 75-GAS-8, EPA Contract No. 68-02-1407,
September, 1975.
3. Test No. C, EMB Project No. 76-GAS-16, EPA Contract No. 68-02-1407,
September, 1976.
4. Test No. D, EMB Project No. 76-GAS-17, EPA Contract No. 68-02-1407,
September, 1976.
5. Test No. E, EPA-650/2-75-042, June, 1975.
6. Test No. F, EMB Project No. 77-GAS-18, EPA Contract No. 68-02-1407,
November, 1976.
7. Test No. G, EMB Project No. 77-GAS-19, EPA Contract No. 68-02-1400,
October, 1977.
8. Test No. H, EMB Project No. 78-BEZ-4, EPA Contract No. 68-02-2813.
9. Test No. I, EMB Project No. 78-BEZ-5.
10. Test No. J, EMB Project No. 78-BEZ-3.
11. Test No. K, EMB Project No. 78-BEZ-ll.
12. "Compliance Testing Analysis of Small Bulk Plants," Contract
No. 68-01-3156, Task Order No.17, prepared by Pacific Environmental
Services, Inc., for the U.S. EPA, Region VIII, Final Report, October, 1976.
C-30
-------
13. Hasselmann, D.E., "Gasoline Transfer Vapor Recovery Systems -
San Diego County, California," TRW Inc., Contract No. 68-02-0235, for
EPA, November, 1974.
14. Analyses of Vapor Samples from Gasoline Storage Tanks
jColom'al Pipeline Company, Greensboro, North Carolina), Scott
Environmental Technology, EPA Set 1656 01 1177, November, 1977.
15. Runion, H.E., "Benzene in Gasoline," AIHA Journal, May, 1975.
16. McDermott, H.J., "Quest for a Gasoline TLV," AIHA Journal,
February, 1978.
17. Reference 7, Op. Cit.
C-31
-------
APPENDIX D
D.I Emission Measurement Methods
D.I.I General Background
For stack sampling purposes, benzene will, except in the
case of systems handling pure benzene, exist in the presence
of other organics. Accordingly, methods for benzene analysis
consist of first separating the benzene from other organics,
followed by measuring the quantity of benzene with a flame
ionization detector. However, among various stack testing groups
concerned with measuring benzene, non-uniformity in procedures
could exist in the following areas: (1) sample collection,
(2) introduction of sample to gas chromatograph, (3) chromato-
graphic column and associated operating parameters, and (4) chro-
matograph calibration.
Two of the possible approaches for benzene sample collec-
tion are grab samples and integrated samples. Since emission con-
tration may vary considerably during a relatively short period of
time, the integrated sample approach offers a greater advantage
over the grab sample approach because emission fluctuations due to
process variations are automatically averaged. In addition, the
integrated approach minimizes the number of samples that need to be
analyzed. For integrated samples, both tubes containing
activated charcoal and Tedlar bags have been used. However,
charcoal sampling tubes were basically designed for sampling ambient
concentration levels of organics. Since source effluent concen-
trations are expected to be higher (particularly since organics other
than benzene could be present) there would be uncertainty
-------
2
involved with predicting sample breakthrough, or when sampling
should be terminated. Bag samples would also offer the potential
for the best precision, since no intermediate sample recovery
step would be involved.
Based on the above considerations, collection of an integrated
sample in Tedlar bags appears to be the best alternative. This
conclusion is in agreement with an EPA funded report whose purpose
was to propose a general measurement technique for gaseous organic
emissions. Another study of benzene stability, or deterioration
in Tedlar bags was undertaken to confirm the soundness of this
2
approach . This study showed no significant deterioration of benzene
over a period of 4 days. Consequently the integrated bag technique
was deemed suitable; however, anyone preferring to use activated
charcoal tubes has this option, provided that efficiency at least
equal to the bag technique can be demonstrated, and procedures to
protect the integrity of the sampling technique are followed.
A collected gas sample can be introduced to a gas chromato-
graph either through use of a gas-tight syringe or an automated
sample loop. The latter approach was selected for the reference
method since it has a lower potential for leakage and provides a
more reproducible sample volume.
Several columns are mentioned in the literature which can
3 4
be suitable for the separation of benzene from other gases; '
most notable among them have been 1, 2, 3 - tris (2-cyanoethoxy)
propane for the separation of aromatics from aliphatics; and
Bentone 34 for separation of aromatics. A program was undertaken
-------
3
to establish whether various organics that were known to be
associated with benzene in stack emissions interfered with the
benzene peaks from the two columns. The study revealed the
former column to be suitable for analysis of benzene in gasoline
vapors, and the latter column to be suitable for analysis of
benzene emissions from maleic anhydride plants. ' It should
be noted that selection of these two columns for inclusion in
Method 111 does not mean that some other column(s) may not work
equally well. In fact, the method has a conditional provision
for use of other columns.
Calibration has been accomplished by two techniques, the
most common being the use of cylinder standards. The second
technique involves injecting known quantities of 99 Mol percent
pure benzene into Tedlar bags as they are being filled with known
volumes of nitrogen. The second technique has been found to
produce equally acceptable results; both are included in Method 111.
D.I.2 Field Testing Experience
Based on the study of benzene stability in Tedlar bags,
possible interferences by various process associated gases, and
calibration methods, and as a result of a field study and tests
conducted at sources of benzene emissions, a new draft of Method 111
was prepared for determining compliance with benzene standards or
NESHAPS. This method is the same as the originally investigated
method, except that the audit procedure has been refined, and an
appendix has been added to aid in the verification of benzene peak
resolution.
-------
4
Four terminals have been tested during the development test.
program. Two of these terminals were originally tested to deter-
mine overall hydrocarbon emissions (including leaks at the trucks)
without specific determination of benzene emissions. These two
terminals were subsequently tested to specifically determine the
control device's efficiency in controlling benzene. The other two
terminals were tested to simultaneously determine total hydrocarbon
emissions (including leaks at the truck) and benzene emissions.
Each of the four terminals employed a different type of control device.
Of two terminals tested for benzene which had been previously
tested for overall hydrocarbon emissions, one employed a compres-
sion-refrigeration, absorption system (CRA) and the other a
refrigeration system. One of the terminals which was tested
simultaneously for overall hydrocarbons and benzene employed a
carbon adsorption system and the other employed a thermal oxidizer.
Emission test procedures used to collect the development test
program data exceeded the procedures required for compliance testing
of the proposed concentration standard. Data was collected to
determine emissions in terms of pollutant concentration, (2) mass
rate, (3) mass per mass of product dispensed, and (4) mass control
efficiency. Additional data was collected to assess the impact of
leaking trucks and unsaturated air-vapor mixtures in the trucks
returning from uncontrolled service stations. These latter data
were necessary because existing terminals typically have- some true!s
servicing stations without stage one control systems. Hov.'ever, it
is anticipated that in the future all service stations will
stage one controls.
-------
5
The sampling procedure differed slightly from the proposed
procedure. Instead of indirectly pulling the sample into a 100
liter Tedlar bag by means of a vacuum inside a rigid container
housing the Tedlar bag, a sample stream was removed from the
sample site by means of a stainless steel bellows sampling pump.
A portion (5-10 percent) of this sample stream was collected in
a smaller bag (10 liter) through a limiting orifice. The use of
the smaller bags was verified in the laboratory by first filling
a large bag with a known concentration of gasoline vapor and
analyzing it and then filling the small bags from the large
bag and analyzing the smaller bags.
The use of the sampling system was similarly verified by
introducing a known concentration of vapor into the sampling
system, analyzing the sample collected, and comparing with the
known concentration.
Analyses of all the samples were performed using the follow-
ing technique:
The Tedlar bag samples were analyzed for individual hydro-
carbons and benzene using a Shimadzu - GC - Mini 1 gas chromato-
graph equipped with dual flame ionization detectors. A Chroma-
topac El A Shimadzu Data Processor was used to measure peak areas.
The column used was a Supelco 20 percent SP 2100/0.1 percent
Carbowax 1500 on 100/120 mesh Supelcoport (D-4536) packed in
10 feet of 1/8 inch stainless steel tubing. This column was
evaluated and shown to provide adequate results for this pro-
7 8
gram ' . The chromatograph was programmed from 40°C to 160°C
-------
6
initially at a rate of 4°C/minute for ten minutes; then the pro-
gram rate was increased to 20°C/minute. Upon reaching 160°C,
it was held isothermal!./ until no more peaks eluted. The total
analysis time was twenty minutes. The calibration gases were
1.02 percent propane in nitrogen and 152 ppm benzene in air.
Samples for injection into the chromatograph were extracted
from theTedlar bags through a rubber septum into a 100 cc gas
sampling syringe. The inlet samples were diluted 50 percent with
room air before injection into the chromatograph. The outlet
samples were analyzed without dilution. Approximately 42 hydro-
carbon species were identified and measured by chromatograhic
separation.
Because two of the terminals had been previously tested for
total hydrocarbons, the benzene test was conducted over a period
of a single day at each of these two terminals. Daily varia-
tions in emissions had been adequately characterized in the
earlier hydrocarbon testing. For the two terminals where total
hydrocarbon and benzene emissions were simultaneously determined,
the data were collected over a period of three days.
Testing of the thermal oxidizer also included analysis of
the exhaust gas for carbon dioxide and carbon monoxide. Because
of the inherent dilution effect of this type of control device,
it was necessary to adjust the concentration for the dilution
effect of the excess air and products of combustion.
Three additional terminals are to be tested to determine total
-------
7
hydrocarbon and benzene emissions. Two of these terminals employ
a lean oil absorption control system and the third terminal uses
a CRA unit. The results of these tests will be included at a
later time.
Testing at three bulk plants is scheduled. During these
tests, the recommended test procedure will be used. Also, a
test program to collect leak test data on gasoline cargo com-
partments using the recommended test procedure is scheduled.
The results of these studies will be included at a later
time.
D.2 Performance Test Methods
The generally recommended performance test method for benzene
is Method 111. The method uses the Method 106 train for sampling,
and a gas chromatograph/flame ionization detector equipped with a
column selected for separation of benzene from the other organics
present, for analysis.
If dilution air is present, Method 3 must also be used.
The recommended field test procedures for determining benzene
emission concentrations at gasoline terminals incorporate Method 111
for benzene analysis. In addition, potential leak sources are
surveyed with a combustible gas indicator to detect any incidence
of direct leaks to the atmosphere.
The recommended field test procedure for bulk gasoline plants
is Method 110. This procedure incorporates measurement of the
volume of vapors vented during gasoline transfers. The vented
-------
8
volume is compared to the volume of liquid transferred to determine
a recovery efficiency. In addition, all potential sources of
direct leakage are monitored with a combustible gas detector.
The recommended field test procedure for gasoline cargo compartments
is Method 112. This is a pressure and vacuum tightness test. The
criteria used to determine vapor tightness is the pressure change
over a five-minute interval after the compartment has been initially
pressurized or evacuated to a specified level.
Subpart A of 40 CFR 61 requires that facilities subject to
Standards of Performance for Mew Stationary Sources be constructed
so as to provide sampling ports adequate for the applicable test
methods, and platforms, access, and utilities necessary to perform
testing at those ports.
Assuming that the test location is near the analytical labora-
tory, and that sample collection and analytical equipment is on
hand, the cost of field collection, laboratory analysis, and
reporting of benzene emissions from a single stack is estimated
to be $2500 to $3500 for a compliance test effort. This figure
assumes a cost of $25/man-hour. While this amount would be
reduced approximately 50 percent per stack if several stacks are
tested, it does presume that all benzene samples would be col-
lected and analyzed in triplicate.
If the plant has established in-house sampling capabilities
and were to conduct their own tests and/or do their own analyses,
the cost per man-hour could be less.
-------
9
D.3 Continuous Monitoring
No emission monitoring fnstrumentation, data acquisition,
and data processing equipment for measuring benzene from bulk
terminal emission gases that are readily available (on an
"as complete systems" basis) have been determined to date.
However, EPA has only recently begun to explore the development
of specifications for benzene monitoring, and it is felt that
such specifications, which would employ a package of individually
comrnercially available items, are feasible.
For a chromatographic system that reports benzene concentra-
tion, the installed cost of the chromatograph and its auxiliaries
is $30,000.a This figure would increase by approximately
$10,000 for the additional hardware necessary to report a benzene
mass emissions rate in terms of benzene feedstock. Depending
on the operating factor, the direct operating cost varies from
about $1,200 to $l,400/year.
References
1. Feairheller, W. R.; Kemmer, A. M.; Warner, B. J.; and
Douglas, D. Q. "Measurement of Gaseous Organic Compound Emis-
sions by Gas Chromatography," EPA Contract No. 68-02-1404,
Task 33 and 68-02-2818, Work Assignment 3. Jan., 1978.
2. Knoll, Joseph E.; Penny, Wade H.; Midgett, Rodney M.;
Environmental Monitoring Series Publication in preparation.
Includes: gas chromatograph with dual flame detector, automatic
gas sampling valve, air sampler, post run calculator, and gas
regulators.
-------
10
Stability of Benzene Containing Gases in Tedlar Bags. QAB/EMSL
U. S. Environmental Protection Agency.
3. Bulletins 743A, 740C, and D. "Separation of Hydro-
carbons" 1974. Supelco, Inc. Bellefonte, Pennsylvania 16823.
4. Volume 10, No. 1 "Current Peaks," 1977. Carle Instru-
ments, Inc., Fullerton, California 92631.
5. Communication from Joseph E. Knoll. Chromatographic
Columns for Benzene Analysis. October 18, 1977.
6. Communication from Joseph E. Knoll. Gas Chromatographic
Columns for Separating Benzene from Other Organics in Cumene and
Maleic Anhydride Process Effluents. November 10, 1977.
7. Communication from Joseph E. Knoll. Gas Chromatographic
Columns for Benzene Analysis - 20 percent SP 2100/0.1 percent
Carbowax 1500 on 100/200 Mesh Supelcoport in 15" by 18" Stainless
Steel. November 25, 1977.
8. Communication from Joseph E. Knoll. Analysis Methods for
Determining Benzene in Liquid Gasoline. January 11, 1978.
-------
APPENDIX E
STATE AND LOCAL
HYDROCARBON REGULATIONS FOR
GASOLINE MARKETING
-------
STATE AND LOCAL REGULATION OF HYDROCARBONS
State
Alabama
Alaska
Arkansas
Arizona
California *
e.g. Bay Area
m San Diego
South Coast
Colorado
Connecticut
Washington, D.C.
Delaware
Florida
* Regulated by Rec
Terminal
Loading Process
Submerged Fill
None
None
Submerged Fill
Vapor Recovery 90%
Vapor Recovery
Vapor Recovery
Vapor Collection &
Disposal = 90%
Vapor Collection &
Disposal
Vapor Collection S
Disposal = 90%
None
None
ional Agencies
Bulk Plant
Storage
!i
Submerged Fill
None
None
Submerged Fill
Balance & Submerged
Fill
Submerged Fill/
Balance
Submerged Fill/
Balance
Submerged Fill &
Collection = 1.15 lb/
1000 gal
Submerged Fill
Submerged Fill
& 90% Collection
None
None
Loading Rack
None
None
None
Submerged Fill
Balance & Submerged
Fill
Submerged Fill /Balance
Submerged Fill/Balance
Vapor Collection &
Disposal = 90%
'•'10,000 gal /day
exempted
Submerged Fill
& 90% Collection
None
None
Service Stations
Underground Storage Tank
Loading
Submerged Fill
None
None
Submerged Fill
•
90% Collection
90% Collection
90% Collection
Submerged Fill & Collection
Equivalent to 1.15 lb/1000 gal
Submerged Fill
Submerged Fill & 90% Collection
None
None
-------
STATE AND LOCAL REGULATION OF HYDROCARBONS
State
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
m Kansas
i
f\3
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missojri
Montana
Terminal
Loading Process
None
None
None
Submerged Fill
Submerged Fill
None
None
90% Control
Submerged Fill
None
None
None
None
None
None
None
None
Bulk Plant
Storage
None
Submerged Fill
None
Submerged Fill
Submerged Fill
None
None
Submerged Fill
Submerged Fill
None
None
None
None
Submerged Fill
None
None
None
Loading Rack
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
Service Stations
Underground Storage Tank
Loading
None
Submerged Fill
None
Submerged Fill
Submerged Fill
None
None
Submerged Fill
Submerged Fill
None
None
None
None
Submerged Fill
None
None
None
-------
STATE AND LOCAL REGULATION OF HYDROCARBONS
State
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
m New York
CO
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Terminal
Loading Process
None
Submerged Fill
None
Submerged Fill
( Reg ioai Requires 90%
control )
None
None
Submerged Fill
Submerged Fill
Vapor Collection & Recov.
Bottom Loading
None
Vapor Collection
Submerged Fill
None
None
None
Bulk Plant
Storage
None
Submerged Fill
None
Submerged Fill
None
None
None
Submerged Fill
Submerged Fill
Submerged Fill
None
Submerged Fill
Submerged Fill
None
None
None
Loading Rack
None
Submerged Fill
None
Submerged Fill
None
None
None
None
None
Submerged Fill
None
None
None
None
None
None
Service Stations
Underground Storage Tank
Loading
. None
Submerged Fill
None
90% Collection
None
None
None
Submerged Fill
Submerged Fill
Submerged Fill
None
Submerged Fill
Submerged Fill
None
None
None
-------
STATE AND LOCAL REGULATION OF HYDROCARBONS
State
Terminal
Loading Process
Bulk Plant
Storage Loading Rack
Service Stations
Underground Storage Tank
Loading
Texas
Vapor Recovery
Submerged Fill
None
Submerged Fill
Utah
None
None
None
None
Virginia
Vapor Control
None
None
None
Vermont
None
None
None
None
Washington
None
None
None
None
West Virginia
None
None
None
None
Wisconsin
None
None
None
None
Wyoming
None
Submerged Fill
None
Submerged Fill
-------
APPENDIX F
DESCRIPTION OF OSHA BENZENE REGULATION
-------
On Friday, February 19, 1978, the Occupational Safety and
Health Administration promulgated a permanent standard for benzene
exposure at workplaces. The standard, which was scheduled to become
effective on March 13, 1978, provided for the measurement of
employee exposure, engineering controls, work practices, personal
protective clothing and equipment, signs and labels, employee
training, medical surveillance, and recordkeeping.
In accordance with OSHA's regulatory approach to the control of
employee exposure to carcinogens, the standard was set at the lowest
feasible level, 1 ppm as an 8 hour time-weighted average and with a
ceiling level of 5 ppm for any 15 minute period during an 8 hour day.
Eye and skin contact with benzene are prohibited. The standard applies
V
to occupational exposure to benzene in all workplaces in all industries
where benzene is produced, reacted, released, packaged, transported,
handled, or otherwise occupationally used, except for the agriculture
industry. The standard does not apply to the sale, discharge, storage,
transportation distribution, or use as a fuel of gasoline and other
fuels, subsequent to discharge from bulk terminals. This means that
bulk plant operators and service station attendants are not covered
by the standard.
Each employer must determine airborne exposure levels from air
samples that are representative of each employee's exposure to
benzene over an 8 hour period. Initial monitoring must be conducted
within 30 days of the effective date of the regulation and frequency
of additional monitoring depends upon whether exposure is above or
-------
below the "action level" of 0.5 ppm, averaged over an 8 hour work
day. If exposure levels are found to be below the action level, no
I
further monitoring is required unless some change occurs which would
lead the employer to believe that benzene levels may be increased. If
exposure levels are above the action level, the employer must repeat
monitoring at least quarterly. Employees must be notified of the
exposure measurements and if exposure levels exceed permissible limits
the employer must include in his report the corrective action being
taken to reduce exposure levels.
The employer is required to use engineering and work practice
controls to reduce exposure levels. If feasible engineering and
work practice controls are not adequate to reduce exposure to
permissible levels, then these controls must be used to reduce exposure
L.
to the lowest possible level. The employer is then required to supply
respirators to reduce worker exposure to a permissible level. Where
eye or dermal contact may occur, the'employer is required to supply
and assure that the employee wears impermeable clothing and equipment
to protect the part of the body which may come in contact with benzene.
The employer must post signs in areas where the use of a respirator
is necessary and affix caution labels to all containers of benzene.
Labels and signs must contain the warning that benzene exposure presents
a potential cancer hazard. The employer is also required to institute a
training program to instruct; employees on the contents of the standard, to
medical surveillance program, the nature of operations which could
result in exposures above permissible levels, and the proper use of
personal protective equipment and clothing.
-------
The medical surveillance program required by the standard
includes the following elements for each employee: a medical
history which includes past work exposure to benzene and other
factors which could influence the effects of benzene on the worker;
and laboratory tests, including a complete blood cell count with red
cell count, white cell count with differential, platelet count,
hematocrit, hemoglobin and red cell indices, serum bilirubin and
reticulocyte count, and additional tests where, in the opinion
of the examining physician, alterations in the components of the
blood are related to benzene exposure. The employer is required
to maintain a record of each employee's exposure to benzene
and medical records for 40 years or the duration of employment plus
20 years, whichever is longer.
This standard has not yet gone into effect, however. Shortly
after the standard was promulgated, OSHA was sued by DuPont Company
and the American Petroleum Institute. In response to these
petitions, the U.S. Court of Appeals for the Third and Fifth Circuits
issued temporary stays on March 12, 1978.
Since the OSHA standard has been stayed, there are no regulations
which require industry to use engineering controls to reduce benzene
levels. Even if the standard had become effective on March 13, 1978,
there is no guarantee that engineering controls will be installed
in the near future since the standard does not specify a date by
which controls must be implemented. Before engineering controls are
installed, respirators must be used. Respirators offer protection
«
only to the workers who weac, them and other persons in the vicinity
-------
of the plant are not affected. Also, if the employer chooses to ventilate
the work area, no beneficial environmental impact will result. For these
reasons, EPA must develop and implement standards to protect the general
public.
------- |