PA/450/3-78/016
EPA-450/3-78-016
April 1978
27711
EMISSION CONTROL
TECHNOLOGY FOR TWO
MODEL MARINE TERMINALS
HANDLING CRUDE OIL
AND GASOLINE
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research'Triangle Park, North Carolina 27711
-------
EPA-450/3-78-016
X
EMISSION CONTROL TECHNOLOGY
FOR TWO MODEL MARINE TERMINALS
HANDLING CRUDE OIL AND GASOLINE
I
v bv
Don M. Gamrnell
Robert Brown Associates
500 East Carson Plaza Drive
Carson. California 90745
Contract No. 68-02-2838
0> EPA Project Officer: David W. Markwordt
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
April 1978
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35) , U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711; or, for a fee, from the
National Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22161.
This report was furnished to the Environmental Protection Agency by
Robert Brown Associates, 500 East Carson Plaza Drive, Carson,
California 90745, in fulfillment of Contract No. 68-02-2838. The contents
of this report are reproduced herein as received from Robert Brown
Associates. The opinions, findings, and conclusions expressed are those
of the author and not necessarily those of the Environmental Protection
Agency. Mention of company or product names is not to be considered
as an endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-78-016
11
-------
TABLE OF CONTENTS
1.0 INTRODUCTION 1
2.0 SUMMARY 3
3.0 CONCLUSIONS 9
4.0 BASE CASE DESCRIPTIONS 15
4.1 General Operations 16
4.2 Facility No. 1 19
4.3 Facility No. 2 25
5.0 ALTERNATIVE VAPOR CONTROL SYSTEMS 31
5.1 Unselected Systems 34
5.2 Selected Systems 36
5.3 Case I Description « 38
5.4 Case II Description 44
5.5 Case III Description 52
5.6 Case IV Description 57
5.7 Case V Description 64
5.8 Case VI Description 73
5.9 Case VII Description 77
6.0 VAPOR CONTROL ECONOMICS 82
6.1 Cost Effectiveness 83
6.2 Total Installed Cost 86
6.3 Direct Operating Cost 87
6.4 Uncontrolled Emissions 88
6.5 Controlled and Secondary Emissions 92
-------
6.6 Marine Retrofitting Costs 96
7.0 ENERGY RESOURCE CONSIDERATIONS 97
8.0 SAFETY and RELIABILITY CONSIDERATIONS . 98
9.0 VAPOR CONTROL DESIGN CRITERIA , 100
9.1 System Safety Features 101
9.1.1 Control Instrumentation 102
9.1.2 Operating Feature 105
9.1.3 Maintenance Features 106
9.1.4 Fire Protection 108
9.1.4.1 Flame Arresting 109
9.1.4.2 Detonation Barriers 110
9.2 Equipment Reliability 113
9.2.1 External Floating Roofs 113
9.2.2 Internal Floating Roofs 114
9.2.3 Cone Roofs 115
9.2.4 Flexible Diaphram Gas Holder 116
9.2.5 Floating Roof Gas Holders 117
9.2.6 Blowers 117
9.2.7 Refrigerator Units 118
9.2.8 Flares and Incinerators 120
9.2.9 Inert Gas Generators 121
9.2.10 Treating Units 122
9.3 Piping Arrangements 126
10.0 REFERENCES 128
11.0 ABBREVIATIONS AND CONVERSIONS 130
-------
1.0 INTRODUCTION
The purpose of this study is to develop basic background information
on emission control systems for a hypothetical deep water marine terminal
handling crude oil, and an inland marine terminal handling crude oil and
gasoline. Terminal models represent maximum design cases as though
for new installations, in order to verify physical feasibilities with
commercially available technology, and therefore are not "typical"
marine terminals as might exist today. These hypothetical terminals do
not provide a suitable basis for retrofit consideration in existing terminals.
The study includes a comparative cost analysis for alternative emission
control systems together with comparable safety and reliability analysis
for both marine terminal modules. A wide variety of collection systems
has been considered. Seven alternative systems were ultimately selected as
being the most comparable and/or feasible for the two marine terminal
facilities. Each emission control system for marine vessel and shore
tankage at both hypothetical terminals has been compared to a base case
arrangement having no emission controls for marine vessels and having
external (open top) floating roof storage tanks on shore.
The collection and disposal of hydrocarbon emissions from marine
vessels at berth and shore tankage has been considered both separately and
in combined displacement systems. Natural gas, nitrogen, and flue gas
have been evaluated as "blanketing" media. Facility design and cost
analysis has been developed for relative cost effectiveness j.n the limited
1
-------
time available for this analysis. Systems have been process-developed
and equipment-rated with redundancy for reliability. Major pipelines
have been sized, terminal layouts established, vendors have been contacted
for current pricing, related utilities and manpower costs established,
load factors have been developed for direct operating costs, and secondary
hydrocarbon emissions have been estimated along with total control methods
to enhance safety.
The use of fixed roof tankage with any vapor recovery system in
lieu of external floating roof (open top) tankage is a sacrifice of safety,
and represents additional capital investment, and increased operating
expenses, for the sake of reducing hydrocarbon emissions. Consequently,
developments and conclusions established by this report are intended to
compare these burdens on private industry in dollars per ton of hydro-
carbons reduced, and to qualify related safety aspects, for various methods
of emission control. Adsorption, absorption and catalytic conversion
methods of vapor control are beyond the scope of this study.
-------
2.0 SUMMARY
Seven vapor control system cases have been selected for comparing
cost versus hydrocarbon emission reduction against two maximum sized
base case marine terminal models, both having a set of fixed conditions.
Emissions other than hydrocarbons to atmosphere are not germane to the
emission control economics of this study, however, other atmospheric
pollutants, including marine stack emissions, have also been developed.
Neither liquid nor solid wastes occur in definable quantities.
Each vapor control system for tankage is sacrificing safety,
by virtue of the complexity involved, for the sake of reducing hydro-
carbon emissions beyond that obtainable by passive external floating
roof tank seals. Floating roof tanks are intrinsically safer than vapor
control systems because their vapor spaces are negligible. Furthermore,
each such system contains an enclosed gas volume which provides a potential
source of atmospheric emissions that can invalidate the purpose of the total
vapor recovery system by any one of several possible mal-operations.
Facility No. 1 is a deep water terminal servicing 125,000 dead
weight ton oil tankers which discharge 2,250,000 B/CD of crude oil,
which does not require heating for fluid properties. Facility No. 2
is a shallower inland port servicing 35,000 dead weight ton oil tankers
which discharge 525,000 B/CD of the same crude oil and 175,000 B/CD of
-------
gasoline. Both terminals discharge their commodities to pipelines
except that Facility No. 2 also discharges some gasoline into unballasted
dedicated barges. Ships are ballasted at berth to 20% of their cargo
capacity in unsegregated compartments. Base case conditions are without
vapor control on ships or barges, and terminal storage is exclusively
with open top, external floating roof tankage.
Base case hydrocarbon emissions have been computed from Supplement
7 of AP-42 "Compilation of Air Pollutant Emission Factors" 2nd Edition.
With these factors three quarters of the base case hydrocarbon emissions
from Facility No. 1 and more than three-quarters from Facility No. 2 result
from ship ballasting and barge loading operations, and essentially all
of the rest come from floating roof tank seals. Both external and internal
floating roof seal factors assume a value of 1.0 in the conventional API
vapor loss formulas. There is substantial evidence, however, that this
o
factor is inordinately high. To illustrate the sensitivity of this
feature, total hydrocarbon emissions are reduced about 21% in Facility
No. 1 and 8% in Facility No. 2 with a seal factor of 0.1. The only
relative distinction between emissions from internal and external floating
roof tankage is that a 4 mph wind is assumed in the internal floating roof,
while both terminal conditions assume ambient wind to average 6 mph for the
external configurations.
Vapor control systems which have been evaluated for both facilities
can be summarized using the following abbreviations:
XFR - external (open top) floating roof tanks
IFR » internal floating roof tanks
CR » cone roof tanks
-------
Case
Base
I
I A
II
II A
III
III A
IV
IV A
V
V A
VI
VI A
VII
VII A
R
I
C
NG
= refrigeration
= incineration or flare
= compression
= natural gas
Marine
Emission Control
none
R
R
I
I
R
R
R
R
R (combined)
R (combined)
R
R
R (combined)
R (combined)
Tankage
Emission Control
XFR
XFR
IFR
CR + I
IFR + I
CR + R + I
IFR + I
CR + C
IFR + C
CR + C
IFR + C
CR + R + I
IFR + I
CR + R
IFR + R
Blanket
Media
none
NG
NG
NG
NG
NG (w/returns)
NG (w/returns)
NG (w/returns)
NG (w/returns)
nitrogen
nitrogen
flue gas
flue gas
The word "returns" is meant to mean natural gas returns to the gas utility
supply system, and the word "combined" to mean that marine and tankage
emission control is with balanced displacements between marine vessels at
berth and shore tankage. In these Cases V and VII, fixed tank roofs are
1/2" thick, instead of 3/16", to allow more operating pressure margin, and
terminal staffs are increased to coordinate ship and shore activities.
Another 1/8" corrosion allowance is provided to all interior surfaces in
Case VII that are exposed to cold flue gas. Case VII receives flue gas
-------
primarily from ships' stacks at berth.
Cases IV, V, VI and VII have gas holders that receive and recycle
blanket gas. Holder capacities vary up to about 25% of total terminal
tankage volumes.
In all but Case I, tankage is blanketed with a gas media, and marine
vessels are also blanketed with the same media in Cases V and VII. Hydro-
carbon vapors will become saturated in these enclosed blanket medias, and
the referenced emission factors, which are intended for vapor losses
into unsaturated air movements, are not applicable to this enclosed environ-
ment. Net effluent gases from enclosed blanket systems are assumed herein
to be saturated with hydrocarbons.
Design capacities for equipment in each vapor recovery case have
been process-rated for the most extenuating set of circumstances con-
sistent with a selected schedule of liquid transfers at each terminal.
These capacities conservatively include breathing volumes as defined by
AP-42, although such volumes would be much less by that amount which
is not vaporized into saturated gas blankets. Schedules have been
selected to reflect each coincident occurance possible with the average
daily, terminal thru-puts required of each facility. This operating approach
is considered more objective than employing a random arrangement of operation
histograms. Four days of varying occurances resulted from these scenarios,
and each day was calculated for net expulsions and impulsions, including
breathing inhalations and exhalations during the night and day hours, respec-
tively, for cone roof tankage. Ships' ballasting has been assumed to occur
during ship unloading operations. A complete scenario of ship and tug-boat
-------
movements has also been developed for both terminals.
The size of floating roof blanket gas holders has been based on the
blanketing capacity needed to supply 1 1/2 days of maximum pump-out
replacements. Flexible diaphram gas holders have been sized to store at
lease one hour of maximum total tankage expulsion rates. Inert gas
generators have been sized to reduce the oxygen concentration in one
empty storage tank from 21% to 4% in about 48 hours, and while pro-
ducing an inert gas containing 1% oxygen. This purging is necessary
for safe start-up operations after tank turn-arounds. The number of units
of equipment to service a process function have been selected to conform
to commercially available sizes, and to provide adequate redundancy so
that one unit can be shut down for maintenance without shutting down
the total vapor control system. Exceptions to this redundancy occur
in large, expensive refrigeration units and in facilities requiring
little if any maintenance, such as ground level flares and flame arresters.
Small rotating machinery has been spared with two 100% units and larger
machines with three 50% units. Control systems for each vapor recovery
case have been defined in a manner which optimizes safety aspects with
the best available technology.
Total installed costs reflect the difference between base case
facilities and each vapor recovery system case selected. Total system
costs have been derived by summarizing individual modulated equipment
function costs. Fourth quarter 1977 and first quarter 1978 costs have
been obtained for equipment and materials delivered to the Los Angeles
area. Equipment modules include estimated equipment erection costs
-------
and all associated piping, instrumentation, electrical, structural,
civil, painting, insulation, start-up, engineering pro-rate, con-
struction indirects, spare parts and expendable costs, in addition
to the delivered equipment cost. Vapor collection and blanket gas
distribution piping have been treated as separate modules to illustrate the
cost magnitude of these features. Direct operating costs are also
modulated with equipment. In this case, design utility loads, associated
operating and maintenance labor, and associated secondary emissions,
are reduced by a load factor obtained from the aforementioned schedule
of liquid transfers at each terminal. Direct operating costs are
further divided into high and low gas and electric utility rates,
consistant with current maximum and minimum rates in coastal regions of
the United States. Thus, any equipment module can be extracted from
or added to a total facility cost difference.
Cost effectiveness of each vapor control case has been evaluated
from the ratio of the annualized costs of the system to the net reduction
of hydrocarbon emissions obtained by the system. Total installed costs
have been annualized by a return of capital constant which reflects
interest, taxes, and depreciation on the cost of each installed system.
Cost effective values are somewhat insensitive to the removal of the last
vestiges of contaminents because of the relative cost of capital in-
vestments to that of the difference in hydrocarbons removed.
-------
3.0 CONCLUSIONS
Vapor control facilities have been sized on the basis that net
effluent gases from each vapor control system can be handled at peak
tanker unloadings while no commodities are being transfered from the
terminal, and that blanket gas supplies can be provided at peak terminal
discharges while no tankers are being unloaded. Internal vapor control
piping (or ducting) has been sized for maximum vapor expulsions and im-
pulsions from any one or more shore tanks and/or marine vessels. Duct
sizes have been calculated for safe operating pressure margins below tank
relief value settings. The resulting size of vapor collection and dis-
tribution headers in these large, atypical terminals consequently vary up
to 80" in diameter, and are supported overhead for appropriate drainage.
Multiple equipment units are employed where design capacities exceed that
for commercially proven equipment sizes. However, the magnitude of these
total system applications exceeds any known real world installation, and
their feasibilities in this context therefore, have not been proven
commercially.
The cost of any vapor recovery system of the magnitude considered
in this study gives cause to consider the relative cost effectiveness of
other methods of mitigating hydrocarbon emissions. Requiring all oil
tankers to have total segregated ballast, for instance, would in itself
reduce total terminal hydrocarbon emissions from the deepwater port
-------
facility in this study by 75%. The verification and monitoring of
realistic double seal floating roof factors could conceivably reduce total
estimated terminal hydrocarbon emissions in this study by another 20%,
considering an ideal seal factor of 0.1. These two developments alone,
therefore, would collectively reduce calculated emissions by some 95%.
A vapor control system for the remaining 5% would increase the cost per
ton of hydrocarbons removed by more than 20-fold.
Discounting the benefits of segregated ballasts, and allowing a 1.0
seal factor for floating roof double seals, Case I has been shown to
be the most cost effective arrangement. Here ballast and barge emissions
only are refrigerated and tankage remains with external floating roofs.
Since hydrocarbon emissions from marine vessel loading and unloading
operations comprise 75% of all such emissions from Facility No. 1 and
89% of those from Facility No. 2, investments to reduce these emissions
would obviously be most cost effective. Conversely, investments for
tankage vapor control systems become relatively less cost effective,
especially in those systems where large blanket gas storages are used.
Furthermore, since these massive volumes of blanket gas are saturated
with hydrocarbon vapors, there prevails the potential ability for such
systems to defeat their purpose by leaking these gases into the
atmosphere through maloperation from time to time.
The cost effectiveness of Case IA is the next best evaluated. This
case also refrigerates ballast and barge emissions, but in addition uses
air-vented internal floating roof tankage instead of external floating
roof tankage. Installed costs are significantly more than Case I because
of internal versus external floating roofs. Although tankage emissions
are passively restrained by both Case I and IA, the latter case presents
10
-------
the need to maintain a very lean gas mixture in the enclosed tank ullages.
Explosive ranges begin under 98 vol % air. Being a passive system, natural
air movement is needed in these ullages which, in itself, stimulates seal
leakage.
Cases II and IIA waste natural gas-blanketed tank emissions and marine
vessel emissions to flare, a feature which makes Case IIA, with internal
floating roofs, the next most cost effective control system. Case II, with
cone roof tanks, on the other hand, becomes the least cost effective case
because of the hydrocarbon vapor from free liquid surfaces that is wasted
to flare. This case illustrates that emissions from cone roof tanks can
be reduced almost 99% by the use of more expensive internal floating roof
tanks, Case III therefore compares the value of recovering these marine
and cone roof tankage' emissions by refrigeration before incineration.
Thus, the high utility rate example for both terminals in Case III
converts the least cost effective vapor control system (Case II) to an
average cost effective system, and the low utility rate example
converts Case II to one of the better cost effective systems, by refrigera-
ting cone roof tank emissions.
Cases IV, V, VI and VII (and their alternates), are much more cost
intensive, primarily because of their ability to store and recycle blanket gas.
Capacities are provided to supply terminal pipeline pumpouts in the absence
of tanker unloadings for 1 1/2 days. Blanket gas make-up and disposal
means are not necessary with blanket storage capacities equal to total
terminal volumes. These case studies have blanket gas storages under
25% of total terminal volumes in order to reduce capital costs with
reasonable inventory management. Cases IV and VI blanket tankage with
natural gas and nitrogen, respectively, while V and VII, which totally
11
-------
recover all hydrocarbon emissions by combining marine vessel emissions
at berth with shore tankage in balanced displacement system, use a
natural gas and flue gas blankets respectively. These two cases are
expensive primarily because of additional tank roof thicknesses that are
used for improving the safety of the operations. The most cost
effective control system of these recycled blanket gas cases is Case V
and the next is IV, both having cone roof tankage and natural gas
blankets. Their alternate cases with internal floating roof tanks are
substantially less cost effective because tank emissions restrained by
internal floating roofs are not an important feature with these totally
enclosed and saturated vapor recovery systems. Treating units are
included in Cases IV and V to return excess blanket natural gas to commer-
cial pipeline service, the recovery of hydrocarbons being incidental.
Case V, which balances a natural gas blanket between marine vessels and
shore tankage, requites tankers and barges to be natural gas blanketed.
The economics of Cases IV and V, consequently, are dependent upon the value
of net excess blanket gas as a fuel elsewhere, and Case V further depends
upon the acceptance of marine vessels containing cargo ullages of natural
gas. A Btu import/export ratio of about 1100 to 1 would result at each
crude receiving terminal in Case V. Case VII, which is like Case V, but
utilizes flue gas as an inerting media instead of natural gas, is the
lease cost effective control system evaluated. Although tankers are being
inerted with flue gas, the cost of their supplying their stack gas to a
terminal for blanket gas usage instead of to their cargo compartments has
been found in this study to be very cost intensive. Costs are attributed
not only to increased tank roof thicknesses for operating safety, but for
12
-------
an additional 1/8" corrosion allowance on all steel surfaces exposed to
the flue gas. Adding further to the costs of Cases V and VII is the need
for additional manpower to coordinate marine and terminal operations
safely.
While cost effectiveness is an objective measurement for the cost of
removing hydrocarbon emissions from the atmosphere, more subjective con-
siderations also warrant concern. The margin of operating safety should
perhaps head the list. The potential impact of human error and of faulty
equipment operations, the consequence of tank fires, the reliability of
automatic control equipment, and the ease of operating and maintaining
vapor control systems can easily overweigh marginal cost effectiveness
differences. Additional fire control investments have been added to
each of these cases utilizing cone roof tanks. It has been noted that
internal floating roof tanks of the sizes in the two terminal facilities
studied may have difficulty retaining reliable floating roof performance.
Control instrumentation has been defined and estimated with the intent of
automatically purging blanket gas mixtures when preselected unsafe oxygen
levels are recorded by oxygen analyzing monitors in strategic blanket gas
locations. Additional staffing has been added for complicated operations.
Installation and operating designs have been estimated that treat vapor
control systems in each case as an auxiliary device which must be as reliable
and safe as the best technology can afford because the primary purpose and
human attention in any terminal facility is in the transfer of bulk hydro-
carbons, and not in the operation of a vapor control system.
It has been noted that emission control effectiveness and cost
effectiveness are not necessarily compatible in view of the gas and
13
-------
electric energy consumed for the Btu value recovered. Similarily,
it has been noted that all vapor control systems represent a compromise
with ultimate terminal safety for the sake of recovering hydrocarbons.
1A
-------
4.0 BASE CASE
Base case conditions have been established for two marine terminal
models in order to compare the relative costs and emission reductions
affected by each vapor control system at each terminal. Refinery operations
are not associated with either terminal operation. Terminal sizes and
throughputs are maximum in order to maximize the utility of' commercially
available technology in controlling emissions. No attempt has been made
to establish a United States average set of crude oil and gasoline properties,
tanker configurations, or marine procedures. The wide variations in the
current world tanker fleet and in U.S. terminal operations would lend no
credence to an "average" terminal module in any case. These features are
believed to be reasonable for such large terminal facilities, however, and
satisfactory to compare the effectiveness of alternative vapor control
systems.
Facility No. 1 is a deepwater terminal which receives 2,250,000
B/CD of crude oil by tanker and dispatches it to pipeline. Facility
No. 2 is an inland port terminal which receives 525,000 B/CD of crude
oil and 175,000 B/CD of gasoline, by tankers, and dispatches all of
it to pipeline except 5,000 B/CD of gasoline, which is removed .by
barge.
Vapor control system cost and emission removals require the
definitions of these two facilities, which are outlined below.
15
-------
4.1 GENERAL
Both models are considered essentially sea level terminals with
ship and barge berthing facilities at docks which run to shore-side
tankage. Ship and barge unloading pumps, therefore, discharge directly
into tankage without the aid of intermediate booster pumps. Ships are
ballasted at berth to 20% of their cargo capacity in unsegregated
compartments. Neither tankers nor barges are necessarily v-erted. Out-
going crude oil and gasoline from both terminals are pumped into pipe-
lines at high pressures. Gasoline is also barged out from Facility No.
2 in dedicated, unballasted barges. Ballast water treating, therefore,
is not normally necessary. However, standby water treating facilities are
provided in both models, but emissions therefrom are not counted.
Annual average diurnal temperature change is 18°F and wind velocity averages
6 mph at both terminals. Crude oil is 34.5° API with an RVP of 6.0 psia.
Average annual storage and transportation temperatures are 75°F. Crude
oil heating is not necessary. Vapor from crude oil contains essentially
no l^S, however, the effect of H_S on vapor recovery systems has
been addressed. Average tank outage is considered to be 50% at both
terminals. Tanks are of welded construction and exterior surfaces are
painted white. The sides of the tanks are clean white and the roofs a
dirty white.
Both terminal facilities are without steam generation capability,
since crude oil heating is not necessary, and electric power and natural
gas is available in the quantities needed for the various vapor control
systems. Firewater is assumed to be provided from sea water at both
facilities and major fires require manpower from local fire districts.
16
-------
A simplified base case emission flow diagram is shown for both
facilities on Drawing 153-1-5. This diagram is used to overlay various
vapor control systems.
17
-------
Q
2
LU
(31
[11
_l
cf.
UJ
Q
_J
O
I
UJ
±
2
<
J
-
J
i
-NO 1_ O
ROBERT MOWN ASSOCIATES
CARSON, CALIFORNIA
CUSTOMH
Snvironmental Protection Agency
PtANT
-oncracc ff68-02-2838
LOCATION Base Case Flow
KLINCLEM VELLUM IO2I-*
18
-------
4.2 FACILITY NO. 1
Terminal thru-put rates average 2,250,000 B/CD. Consequently
this facility berths three 120,000 DWT tankers, each having the following
characteristics:
Overall length 886 feet
Maximum beam 140 feet
Draft (loaded) 52 feet
Freeboard (loaded) 16 feet
Cargo capacity 840,000 bbls
Off-loading rate 74,000 BPH
Ave. ballasting rate 50,000 BPH
Refer to drawing 153-1-1 for Facility No. 1 plot plan and drawing
153-1-2 for a simplified flow sheet. Facility No. 1 contains 24 770M
barrel crude oil tanks, each 340' in diameter by 48' high. Bunker fuel
oil tanks and potable water tanks are shown on the plot plan, but since
these are not related to emissions sources they are not relevant to this
study. Incoming crude oil rates vary up to 222,000 BPH from three tankers
simultaneously. Outgoing crude oil is discharged into pipelines at rates
up to 150,000 BPH maximum. Drawing 153-1-17 illustrates the transfer
schedules for Facility No. 1, which encompasses all varieties of opera-
tion that are consistent with daily thru-puts. A scenario of ship and
tugboat operations within 5 miles of this deepwater terminal has been
developed to illustrate the relative impact of marine stack emissions
to the alternative vapor recovery systems. Tankers burn No. 6 fuel oil
for their boilers in this model, and sea-going tugs run on diesel oil.
19
-------
Six tug-hours are used for each full ship arrival and two tug-hours
are used for empty ship departures. Ship boiler loads have been
reduced for arrival, pumping, stripping and departure operations.
20
-------
i
$
xU
2^
------- An error occurred while trying to OCR this image.
-------
FACILITY MS I
TRANSFER SCHEDULE
AM PM AM
9 IO II 12 I 234-5^,739 10 II 1-2 I Z34-5-7
I _ i i i i i _i _ i _j i I i i i I I _ I I i I i I i
I i
DAY I
CRUDE TAMKER I I.BS HR (TVP)
PIPELJME PUMPS
CRUDE TAMKER 11/35 HR(TYp)
CRUDE TAK1KER II/3g> HR (TYP)
DAY
DAY 2»
DAY 4
r-
CRUDE TAKJKER
CRUDE TAN1KER
PIPE-LlkJE PUMPS
CRUDE TAMKER
CRUDE TANJKER
CRUDE TAMKER
CRUDE TAKJKER
PIPEUKJE PUMPS
CRUDE TAMKER
CRUDE TAMKE.R
CRUDE
PIPELIME PUMPS ZC
H-UTI
RELOCATED P/L POMPIW<=,
ROBERT BROWN ASSOCIATES
CARSON, CALIFORNIA
CUSTOMER Environmental Protection Agency
PLANT Contract No. 68-02-2838
LOCATION FarlHt-v //I. Transfpr Schedule
DRAWING NUMBER
155-1-17
REV.
KI_INCt_ER VEULUM 1031 -
23
-------
o
2
1 1/25/75 Ifc-S.zCD LINK ft*>ce. R.ON
.NO.. T--T ^'^-^.giA
IX.
ROBERT BROWN ASSOCIATES
CARSON, CALIFORNIA
CUSTOMER Environmental Protection Agency
PLANT Contract No. 68-02-2838
LOCATIONFac. VI, Dock VCS Collection System
DRAWING NUMBER
REV.
153-1-13
KV.INSLER VELLUM I021-*
24
-------
4.3 FACILITY NO. 2
This shallower inland port has a terminal thru-put averaging
525,000 B/CD of crude oil and 175,000 of gasoline, totaling 700,000
B/CD. Docking facilities berth two typical 35,000 DWT crude oil
tankers and one similar gasoline tanker and one barge for loading
gasoline. Ships and barges have the following characteristics:
Ships Barges
Overall length 690 feet 220 feet
Maximum beam 92 feet 62 feet
Draft (unloaded) 35 feet 17 feet
Freeboard (loaded) 12 feet 5 feet
Cargo capacity 235,000 bbls 50,000 bbls
Off-loading rate 30,000 BPH
On-loading rate - 5,000 BPH
Ave. Ballasting rate 20,000 BPH
Refer to drawing 153-1-3 and 153-1-4 for Facility No. 2 layout
and simplified flow sheet, respectively. This facility contains six
615M barrel tanks for crude oil and two 615M barrel tanks for gasoline,
each tank being 262' in diameter and 64" high. Incoming crude oil varies
in rates up to 60,000 BPH when two oil tankers are discharging simulta-
neously, and up to 30,000 BPH of gasoline from one tanker. Outgoing
crude rates to pipeline vary up to 35,000 BPH maximum and outgoing gas-
oline to pipelines up to 12,000 BPH. Gasoline is 55° API and 10 RVP
and is loaded onto barges at rates up to 5,000 BPH. The arrival condition
of the barge is uncleaned and unballasted, such that the uncontrolled
25
-------
emission factor from gasoline loading is 4.0 Ibs of hydrocarbons per
1,000 gallons loaded. The transfer schedule for Facility No 2 is
shown on drawing 153-1-18.
A scenario for ship and tugboat movements has also been developed
within 5 miles of this port. These smaller tanks and harbor tugs run
exclusively on diesel oil. Four tug-hours are used for each full
ship arrival and two tug-hours for each empty departure. Two tug-
hours are used for each empty barge arrival and four tug-hours for
each full barge departure. Ships' diesel loads have been reduced for
arrival, pumping, shipping and departure operations.
26
------- An error occurred while trying to OCR this image.
------- An error occurred while trying to OCR this image.
-------
DAV I
PAV
DAV3
DAY 4
FACILITY
TRANSFER SCHEDULE
AM PM AM
9 1011 12 1 23 4 5 fe ~7 a -9 10 II 12 I 22>45&>-7e>9
i i i i I i i i i i i i i i i i i i i i i I i i I
GASOLINE TAKJKER 7.Q3
BARGE LOADIMQ IOHR(TVP;
CRUDe TANKER 7.S3HR.CTVP)
CRUDE TAK1KER T.
CRUDE PIpeUKJE PUMPS ISMR.XTVP)
PIPEUNE PUMPS
CRUDE TAKJK&R
CRUDE TANKER
BARG.E LOADING
C5ASOLI ME TANKER
CRUDE TAMKER
CRUDE- TAMKER
G,ASOLI
CRUDE PIPEUME PUMPS
GiASOLlME PIPELINE PUMPS
CRUDE PIPELINE: PUMPS
QASOUME: PIPELIME PUMPS
QASOUME TAMKER
CRUDE TAEJKER
CRUDE TAMKER
BARGE
CRUDE1 PIPELIKJE PUMF*S»
C5ASOL.IME; PIPEUKJE PUMP'S
NO » OATI.
J«
\
ROBOT BROWN ASSOCIATES
CARSON. CALIFORNIA
KLINGLCM VCLLUM I03I-*
CUSTOMER Environmental Protection Agency
FtANT Contract No. 68-02-2838
LOCATION Facility #2. Transfer Schedule
29
DRAWING NUMBER
153-1-18
REV.
0
-------
fyb
//
// /
vW -/
\ )
II/I4.J77
1
_«0 - OATI.
1/25/78 J?e"Si:i:UMeS foe
r.c
»" -I
ROBERT BROWN ASSOCIATES
CARSON.CALIFORNIA
KLINCLCR VELUUM 1021-9
CUSTOMER Environmental Protection Agency
PLANT
Contract No. 68-02-2838
LOCATION Fac. //2, Dock VCS Collection System
DRAWING NUMBER
153-1-15
REV.
30
-------
5.0 ALTERNATIVE VAPOR CONTROL SYSTEMS
Vapor control is manifested by the collection and disposition or
containment of a non-explosive mixture of gases in order to prevent pol-
lutant components from escaping to atmosphere. A mixture of gases results
from the effect of the partial pressure of pollutants in this contiguous
gas phase over the parent liquid. The mixture of gases is referred to
herein as "blanket gas", which includes "hydrocarbons", the latter being
a pollutant. Blanket gas, therefore, is a media for conveying the pol-
lutants to some disposition, without causing structural pressure damage
to the parent liquid containers. Thus, excessive physical vacuums or
pressures and chemical explosive or implosive mixtures must be avoided.
Means are provided to maintain oxygen concentrations below safe levels,
or to warn operators to take necessary precautions when certain levels
are reached.
The saturation composition of crude oil vapors and gasoline
vapors in air at 75°F and atmospheric pressure is about (4.6/14.7) (100)
and (6.8/14.7) (100) percent, respectively, in the models herein.
These compositions are approximately one third and one half of their upper
explosive limit in air (UEL=84 vol % H-C), but air has not been con-
sidered as an alternative blanket gas media because of safety reasons. Air is
saturated by gasoline in some commercial truck loading rack vapor recovery
31
-------
packages in order to render the mixture of gases too rich to burn. This
has been discounted, in this study, primarily because the margin of safety
from explosive conditions is considered inadequate in view of the size
of these terminal volumes and the consequences of explosions. Secondly,
the vaporizing process itself must pass through the explosive range. If
the true vapor pressure of the parent liquid would drop from 4.6 to 2.2 psia
during vaporization, for instance, that equilibrium vapor-air mixture
produced would be explosive.
Gas blanketed systems in this study are assumed to be saturated at
the abovementioned compositions, as defined from Raoults' and Dalton's
Law for ideal gases. Cone roof tank breathing and working losses, float-
ing roof tank standing and withdrawal losses, ballast emissions, and barge
loading emissions, also have been assumed in this comparative study to
produce hydrocarbon emissions of the magnitudes defined by Supplement 7
of AP 42. Hydrocarbons recovered by refrigeration and returned to ter-
minal throughputs, or those lost to incineration, have conservatively in-
cluded breathing loss volumes along with displacement volumes. Actual
hydrocarbon vaporizations into saturated hydrocarbon gas phases, however,
would be substantially less than those defined by Supplement 7 of AP-42
because mass transfer diffusion potentials would disappear. The
continuous removal of an equal amount of vaporized hydrocarbons from the
enclosed blanket media is assumed in this study to off-set saturation
equilibria to an extent that these published emission factors prevail.
This assumption, while being expedient, also helps to magnify comparative
differences between alternative vapor control systems. It should be kept
in mind, therefore, that actual breathing losses into closed blanket gas systems
32
-------
would be substantially less than those in this study, and that refrigera-
tion power requirements and incineration losses would also be less by
this difference in volumes of saturated hydrocarbons handled.
33
-------
5.1 UNSELECTED SYSTEMS
A number of vapor recovery systems have been proposed for terminal
applications in recent years. Many have been considered, but not
selected, for this study.
Refrigerated cargoes have not been selected for study because of
excessive refrigeration and cargo pumping cost in both the terminal
and the tankers, plus capital investments at the terminal to main-
tain cold temperatures.
Burning waste blanket gas in tanker boilers has not been selected
for study in view of a limited application and the danger of positive
fire box pressures in boilers back-flashing into cargo volumes.
Venting secondary seals on external floating roof tanks has npt been
selected for study because negative pressures in the long narrow
spaces between the two seals would stimulate both the wind effect
on the primary seal increasing hydrocarbon emissions and air leakage
through the secondary seal causing explosive gas mixtures. Positive
pressure with inert gas on the other hand, would also increase the
wind effect on the primary seal and emissions through the secondary
seal, increasing hydrocarbon emissions and causing excessive inert
gas demands.
Water displacement storages have not been selected for study because
Sea Dock studies showed these systems to be excessively capital
intensive.
Variable vapor volume storages equal to total terminal volumes have
not been selected for study because other (unpublished) studies
have shown these systems to be too costly, and too vulnerable to
atmospheric leakages.
34
-------
High pressure compression and storage of excess blanket gas has
not been utilized because of high energy and pressure vessel storage
costs.
Commercial vapor recovery packages have not been selected because
these units, which are utilized at truck loading racks, have not
been designed for such large capacities. In effect, their unit
features, namely saturation, compression, absorption and refriger-
ation have been utilized where applicable to the scope of this effort.
Hydrocarbon/air saturation methods to render blanket gas too rich
to burn, as in some commercial truck loading vapor recovery units
have not been selected, as explained in section 5.0, p. 32.
It has been decided that evaluations of the following three methods of vapor
collection and/or disposal fall outside the scope of this work effort:
Absorption systems, which may well have feasible application in
these models.
Catalytic conversion methods, which may feasibly reduce fuel gas
consumption in the oxidation of hydrocarbon emissions to atmosphere.
Adsorption systems, which may have a limited role also in these models.
35
-------
5.2 SELECTED SYSTEMS
The systems selected fall into three basic categories regarding
tankage emission controls:
1. Minimum investment (Cases I and II)
2. Natural gas blanket media (Cases III, IV, and V)
3. Non-combustible gas blanket media (Cases VI and VII)
Ballast emissions are either refrigerated or incinerated. Refrigeration
vapor emissions are either vented to atmosphere or returned to the
blanket gas system (Cases V and VII). Some of these cases were selected
with the understanding that, while their economics may be obviously in
disfavor, their economic evaluations were necessary to illustrate the
comparative value of other cases. Each case has an alternative system
wherein internal floating roof tankage is used instead of cone roof
tankage or, as in Case I, external floating roof tankage.
These hypothetical terminal facilities do not address their vapor
control systems to the handling of sour crudes. The basic change to the
systems selected, if crudes were to be handled whose vapors would be sour
enough to contaminate gasoline storages, would be to separate crude and
gasoline tankage vapor collections and recycled blanket gas distributions.
Marine contaminations would not occur in the scenarios chosen for this study.
Separate flexible diaphram gas receivers and floating roof gas holders
would, therefore, be used for crude and gasoline tankage vapors in Facility
No. 2, and separate blanket gas piping would be needed in Cases IV, V, VI,
and VII. In these cases where blanket gas is recycled, sour vapors should
be treated as proposed for Cases IV and V regarding sulfur removal. Here
36
-------
sponge iron replacements would increase in direct proportion to the amount
of H_S that is removed. The effect on vapor control arrangements for
handling sour crudes, meaning those whose vapors contain hydrogen sulfides,
is described for each vapor control case below.
37
-------
5.3 CASE I
Case I differs from the Base Case only by refrigerating ballast
emissions. Case IA differs further by having internal instead of exter-
nal floating roof tanks. Refer to drawings 153-1-6 and -6A for both
facilities, and to Tables 1 and 2 for relative economic values.
Ballast emissions are explosive, according to the latest EPA emission
factors for ballasting into air-ladened crude or gasoline tanker compart-
ments . Emissions from loading dedicated unballasted air-ladened
gasoline barges are almost too rich to burn. A single collection header
is used to collect these emissions from tankers and barges to an enclosed 3"
water seal and refrigeration unit on shore. Collection piping, which is about
1700 ft. long in Facility No. 1, and some 500 ft. further in Facility No. 2
has been sized for 2 psig working pressures in tanker and barge compartments^' .
Marine vessel cargo flanges are connected to dock piping with hose con-
nections. Flame arresters are located between the 3" water seal and the
refrigerator on shore, at each valved hose connection on the docks, and at each
valved ship-board hose connection. One-quarter inch thick collection
piping is located on the wharfs so as to cause the least amount of detona-
tion damage. More importantly, detonation is avoided by purging from the
extremities of each collection header with recycled normal butane gas
to and from the refrigerator, in order to render the pipe contents too
rich to burn. Heat exchange within the refrigerator is used to vaporize
these condensed hydrocarbons at 30 psig. Oxygen analyzing recorder
controllers regulate the amount of purge gas into each vapor collection
header. Appropriate alarms are sounded if oxygen concentrations above
safe levels are recorded. This slave butane circulating system consumes
38
-------
only 3.9 MM Btu/Hr at the worst conditions, (i.e. having 3 tankers ballast-
ing simultaneously in Facility No. 1) in order to vaporize enough butane
to increase hydrocarbon pipeline concentration from 3.37 vol % to 20 vol %.
Revaporization is at -32°F. About 200 operating horsepower is, therefore,
required for this recycle purge at design conditions in Facility No. 1,
and about 75 horsepower at most in Facility No. 2.
Refrigerator horsepower and piping costs could be reduced by placing
the refrigeration package on the docks, adjacent to the ships and barges.
The enrichening butane recycle might then be omitted by reason of having
smaller potentially explosive volumes. However, this study has considered
only remote refrigeration, in a pressurized housing, with gas blanketing
throughout all collection branches, because it is the safer arrangement.
Accidents do occur at wharfs, expecially while docking during inclement
weather. A multi-million dollar refrigeration system, containing non-
explosion proof electrical gear in a pressurized housing on the dock could
become hazardous. The elevated housing air intake, for instance, could
inhale a vapor blanket expelled by a tanker maloperation.
Refrigeration installation and operating costs are for temperatures
reduced to -170°F in order to minimize propane emissions from virgin crudes.
Calculated hydrocarbon emissions amount to 925 short tons per year none-the-
less, before applying loading factors, for Facility No. 1, and 396 short
tons per year for Facility No. 2. H^S from sour crude vapors would also be
emitted.
Hydrate formation problems have been considered minimal at these
low pressures with proper precooling to the hydrate point, and heat
recovery sections that conduct defrost-cycle vapors back to the
39
-------
refrigerating section. See paragraph 9.2.7. The above emissions calcu-
lated from vapor pressure equilibrium are expected to include any losses
resulting from defrosting hydrates.
Controlled versus uncontrolled Base Case emission in short tons/year
and installed vapor control costs are:
EMISSIONS ST/Y COST
Facility No. 1
Base Case
Case I
Case IA
H-C
2890
1387
1158
NO
X
148
148
148
SO
X
111
111
111
CO
40
40
40
Part.
11
11
11
$ M
-
2,630
6,355
Facility No. 2
Base Case 2578 280 20 136 0
Case I 515 280 20 136 0 1,392
Case IA 441 280 20 136 0 2,260
Case I is the most cost effective, and Case IA the next most cost
effective, control system evaluated. "Cost effectiveness" is a measure
of those investment and operating costs required to reduce a ton of hydro-
carbon emissions from base case conditions. The best cost effectiveness
is the lowest cost value. Refer to Tables 1 and 2 for Facilities No. 1
and 2 respectively. The primary reason for these cases being the most
cost effective for both facilities is that they direct their efforts to
emissions from marine vessels, from which 75% of all hydrocarbon emissions
from Facility No. 1 and 89% of all those from Facility No. 2 come. Vapor
control systems to remove tankage emissions are, therefore, relatively
costly.
40
-------
Case IA clearly shows the cost effectiveness of internal versus
external floating roof tankage, based upon the arbitrary criteria only
that an average 4 mph wind effect prevails in the former and a 6 mph wind
effect prevails for the latter. With these ground rules, the cost effec-
tiveness is less than half as good by having internal floating roofs in
the larger Facility No. 1 and somewhat better in Facility No. 2. Case IA
is the only example of such tankage with conventional air vents on the
upper shell. Other alternate cases utilize gas blankets with internal
floating roof tankage. Air-vapor mixtures in these ullages can become
explosive, especially after fast withdrawals from small tanks, but both
costs and explosive conditions are related to tank sizes, amongst other
things. Refer to paragraph 9.2.2.
41
-------
Environmental Protection Agency
DRAWING NUMBER
ROBERT BROWN ASSOCIATES
CARSON.CALIFORNIA
Contract //68-02-2838
153-l-fc
Case I Flow
KL1NGL.EB VELLUM 1021 -
42
-------
ROBERT BROWN ASSOCIATES
CARSON, CALIFORNIA
CUSTOMER Environmental Protection Agency
PLANT
Contract #68-02-2838
LOCATION
Case I Flow
DRAWING NUMBER
REV.
153-1-foA
Kt.iNci.ER VELLUM 1021 -*
43
-------
5.4 CASE II
This case incinerates ballast and gasoline barge loading emissions
directly and flares net excess blanket natural gas from tankage directly.
No refrigeration is used. Drawings 153-1-7 and 7A describe this application
to both facilities. Tables 1 and 2 present relative economic values.
The same safety precautions are taken for explosive ballast and
barge emissions here as in Case I, except that fuel gas for incineration
is used instead of the slave vaporization system used in Case I, and
aluminum emission blowers are used in this case to provide pressures that
will assure smokeless burning characteristics.
Fuel gas is purged into the extremities of the dockside vapor col~
lection headers, rendering the pipe contents slightly too rich to burn,
at about 20 vol % hydrocarbons. At worst conditions, (i.e.; ballasting
three tankers at once in Facility No. 1) this requires 169,7 MSCFH of
gas. Gas purge flows are regulated by oxygen analyzing recorder-
controllers. The enclosed 3" water seal and flame arrester preceding
the incinerator in this case is more critical in preventing hackflashes
than they were in Case I because constant ignition prevails at the
incinerator. Consequently, automatic controls must maintain the
water seal in the vessel with fail safe redundancy and alarms. A small gas
purge after the seal and through the flame arrester is used to denote
a forward flow of gases. Any failure in the forward flow automatically
shuts off all fuel gas to the incinerator, including pilot gas, and
sounds appropriate alarms.
Natural gas under about 60 psig is distributed directly to storage
tankage whenever tank ullage pressures reach 0.5 ounces per square inch
44
-------
(osi) vacuum. Gas valves are set to start opening at 0.5 osi vacuum and
to be fully open at 1.3 osi vacuum. Refer to Table 3, paragraph 9 .1.1.
Cone roof tanks in Case II (both facilities) and internal floating roof
tanks in Case HA (both facilities) both have 3/16" fixed steel roofs with
vacuum relief valve settings of 2.78 osi vacuum. Multiple PSV's are needed
for each major storage tank in both facilities, and a blanket gas tank
inlet is located near .each PSV to prevent any pressure lag from opening the
vacuum PSV to atmosphere. One common pressure control valve (PCV), however,
is used for blanket gas per tank in both facilities. See paragraphs.3.
Net excess blanket gas expulsions are collected in flexible diaphram
gas holders at 0.2 osi. Thus only 0.3 osi of pressure drop motivates
blanket gas flow through a flame arrestor and into these holders. Vapor
collection ducts have been sized accordingly. Tankage vapor collection
piping is shown on drawings 153-1-14 and 153-1-16 for Facilities No. 1 and
2. Four holders are used in Facility No. 1 and two in Facility
No. 2. If crudes with sour vapors are to be handled in Facility No. 2,
contamination of gasoline from crude vapors can be avoided by collecting
crude and gasoline vapors separately into separate gas holders. When the
bags in these groups of holders are full, a bag level switcn activates a
blower that transfers the gas first from one gas holder, then from another,
into a water-sealed ground flare. Storage tank impulsions, due to breathing
inhalations and/or pumpout replacement volumes, will back-flow blanket
gas from these gas holders before new purchased natural gas is admitted
into storage tankage because 0.7 osi pressure drop is available in that
direction (versus 0.3 osi for expulsions) before the PCV's begin to open.
-------
ROBERT BROWN ASSOCIATES
CARSON, CALIFORNIA
CUSTOMER Environmental Protection Agency
PLANT Contract No. 68-02-2838
LOCATION
yannr Hnllprflnn Layout
DRAWING NUMBER
REV.
153-1-14
VCLLUM 1021-
46
-------
ROBOT BROWN ASSOCIATES
CARSON,CALIFORNIA
CUSTOMER Environmental Protection Agency
PIANT
Contract No. 68-02-2838
LOCATION Facility #2 Vapor Collection Layout
DRAWING NUMBER
REV.
I53-H6
KLINGLEM VELLUM 102 t-«
47
-------
In more capital intensive vapor control systems, Cases IV, V, VI and
VII, the reuse of net excess blanket gas, before new gas sources are used,
is assured by boosting excess blanket gas from the flexible diaphram
receiver tanks into floating roof gas holders at 7.0 osi pressure with
blowers. From there blanket gas needs are supplied by a separate distri-
bution and PCV system to each storage tank. Floating roof gas holders are
relatively expensive, costing about $2.00/cubic foot. With natural gas
costing roughly $2.00/1000 cubic feet, the direct payout in natural gas
by such storages requires more than 8 years, excluding the cost of separate
distribution systems.
Controlled versus uncontrolled Base Case emissions in short tons/year
and installed vapor control costs are:
EMISSIONS ST/Y COST
Facility No. 1
Base Case
Case II
Case IIA
H-C
2890
78
77
NO
X
148
267
228
SO
X
111
114
111
CO
40
57
51
Part.
11
29
23
$ M
-
2,720
11,290
Facility No. 2
Base Case
Case II
Case IIA
2578
61
61
280
335
337
20
21
20
136
144
144
0
8.3
8.5
-
1,940
3, 680
Although it is not evident in the tabulation above, the value of
restraining vaporizations in fixed roof tankage by internal floating pans
48
-------
is dramatically quantified by Case II control systems. This is the only
case where tankage emissions are completely wasted. Those from cone roof
tankage total $7,922,000 annually in lost crude from Facility No. 1, and
$3,423,000 annually in products from Facility No. 2. Losses from internal
floating roof tankage in Case IIA, however, amount to only $98,200 and
$38,000, respectively. On the other hand, internal floating roof tankage
costs much more and requires p.reater blanket gas demands because tank with-
drawals therefrom are not partially replaced by surface vaporizations to
the extent they are in cone roof tankage. Once the blanket gas is drawn
into the system from outside, in this case, it is ultimately incinerated.
These off-setting effects do not manifest themselves in the above tabula-
tion, and the cost effectiveness of those negative credits to Case II become
relatively inconsequential. Refer to Tables 1 and 2.
Sour crude vapors would convert stoichiometrically to SO from the
A
incinerators in this case. Since this study is not based upon a specific
crude source, and sulfur levels have not been defined, the SO
x
emissions tabulated above do not include sulfur sources from crude
or gasoline vapors.
-------
r
d?
:
«.:« JA»"V* _
ROBERT BROWN ASSOCIATES
CARSON, CALIFORNIA
KV.INCLCH VELUUM 1011-1
CUSTOMER Environmental Protection Agency
PUNT
Contract #68-02-2838
LOCATION CaseH Flow
DRAWING NUMBER
REV.
153-1-7
50
-------
CUSTOMER
PUNT
DRAWING NUMBER
Environmental Protection Aeency
ROB£XT EROWN ASSOCIATES
CARSON.CALIFORNIA
Contract j-68-02-2838
I53-I-7A
LOCATION Case IIA Flow
«l_INCLC» VE L CU" 'CI1
-------
5.5 CASE III
This case uses refrigeration for both direct marine emissions and
net excess blanket natural gas from cone roof tankage. Emissions from
tanker ballasting and gasoline barge loadings are handled as outlined for
Case I. Refer to drawings 153-1-8 and 8A for this application to both
facilities, and to Tables 1 and 2 for relative economic values.
A study was made for this case to burn refrigeration vapor effluent
from marine emissions by only condensing those hydrocarbons in excess of
that needed to support combustion. The only substantial amount of marine
hydrocarbons condensed, however, became those relatively few from gasoline
barge loadings, and a great quantity of fuel was consumed, and combustion
pollutants emitted, by simply heating large amounts of air. The combina-
tion of incineration and refrigeration was found thereby to be self defeat-
ing without a combustible gas blanket. Case 111 quantifies the value of
this combination where relatively large vaporizations occur from cone roof
tankage.
Net blanket gas expulsions from tankage for Case III are refrigerated
from the same flexible diaphram gas holders as those in Case II. All
hydrocarbons vaporized from crude oil and gasoline net breathing and
working losses in cone roof tanks are condensed by refrigeration. Pilot
gas is supplied from a more reliable source. Emissions from floating roof
tanks in Case IIIA, however, are too few to be worth refrigerating.
The basic difference between this case and Case II regarding tankage vapor
control, therefore, is only the return of tankage vapor losses to terminal
through-puts for the cone roof tanks in the primary Case III. Case IIIA
is the same as Case IIA regarding tankage vapor control.
52
-------
Controlled emissions versus uncontrolled Base Case emissions in short
tons per year and installed vapor control costs are:
EMISSIONS ST/Y
COST
Facility No. 1
Base Case
Case III
Case IIIA
H-C
2890
341
340
NO
X
148
195
216
SO
X
111
114
111
CO
40
47
50
Part.
11
18
21
$ M
15,950
13,220
Facility No. 2
Base Case 2578
Case III 160
Case IIIA 159
Cases II and III have avoided the cost of significant blanket gas
storages at the expense of consuming larger amounts of commercial natural
gas. Gas demands for these cases in MSCFD are:
280
305
332
20
21
20
136
139
143
0
3.8
7.9
-
7,510
4,510
Case II
IIA
Case III
IIIA
Facility No. 1
2,723
3,613
2,163
3,062
Facility No. 2
1,347
2,573
1,156
2,382
These huge commercial gas costs are reflected in Tables 1 and 2. Refrig-
eration electrical costs obscure the relationships somewhat in Case III.
Here, in an effort to recover essentially all cone roof hydrocarbon emis-
53
-------
sions, refrigeration to -170 F has been estimated. Reliable availability
of the above commercial gas supplies would greatly limit, if not inhibit,
the application of these control systems.
Crudes with sour vapors would cause the release of HLS to atmosphere
from refrigeration units handling ballast emissions, as in Case I, and
they would cause the release of SO from incinerators handling tankage
2v
emissions.
-------
REMOVED BLOWER
(VKIOH
CUSTOMER Environmental Protection Agency
DRAWING NUMBER
ROBERT BROWN ASSOCIATES
CARSON,CALIFORNIA
153-1-8
Case III How
KLINGLEM VELLUM 1021-i
55
-------
ROBERT BROWN ASSOCIATES
CARSON.CALIFORNIA
CUSTOMER Environmental Protection Agency
PLANT
Contract f-68-02-2838
LOCATION Case I1IA Flow
DRAWING NUVBER
REV
I53-I-8A
VELLUM 1C2 1 4
56
-------
5.6- CASE IV
This case differs from previous cases by storing and recycling blanket
natural gas to tankage needs, instead of to flare, at the expense of conven-
tional low-pressure gas storage facilities. Condensed hydrocarbons are
returned from the treating process by refrigeration where cone roof tankage
is used, and by simple compression where internal floating roof tankage is
used. In both cases these net gas expulsions are compressed to pipeline
pressures in order to economically condense tankage vaporizations, and
are normally recycled back to blanket gas storage for power recovery.
Marine emissions are treated by refrigeration as in Case I.
This case has been charged with the cost of floating roof gas storage
and associated blanket gas distribution piping necessary to provide blanket
gas for 1 1/2 days of maximum daily pumpout rates without a tanker being
unloaded. Gas holder volumes amount to about 13% of the total cone roof
terminal volumes and 18% of the total internal floating roof terminal
volumes, the latter requiring higher blanket gas demands. With proper
inventory management, large amounts of blanket gas purchases should only
be limited to occasional storage tank turn-around operations, and this
can be minimized by purging the tank with water fillings. Annual replenish-
ment charges have been arbitrarily assumed to average that required for
1 1/2 major storage tank fills per terminal, however. Other fresh make-up
blanket gas is eventually returned to sales in a reasonably scheduled and
treated manner, and other direct operating costs, therefore, cancel out
after the initial charge is capitalized. Refer to drawings 153-1-9 and
9A for this application to both facilities. Net blanket gas expulsions
from tankage are transferred from small flexible diaphram surge tanks by
57
-------
the same blower arrangement as in Case II, but into floating roof gas
holders at 7.0 osi pressure instead of into a water-sealed flare. Four
such gas holders are needed for Facility No. 1 and two for Facility No. 2.
Separate distribution piping and PCV's are needed to recycle this gas
from these gas holders to storage tanks for blanketing before natural gas
make-up is used. Drawing 153-1-23 illustrates a typical piping arrangement
where a separate high and low (recycle) pressure supply source provides blanket
gas to tankage. One oxygen analyzer sequentially records 09 concentrations
near each pressure-vacuum relief valve. An alarm is sounded at a central
control location whenever excessive oxygen concentrations are reached.
Refer to Table 3 paragraph 9.1.1 for pressure control valve settings.
Only when all floating roof gas holders and flexible diaphram receiving
holders are full, or when operating schedules demand, is blanket gas from
storage tankage returned to sales. A standby flare has been estimated for
disposing of excess blanket gas when operations cannot be accommodated by
sales. Normally the treated blanket gas is returned to gas storage and
compression power is largely recovered. Treating the gas removes any
accumulation of inorganic sulfur compounds, of water vapor, and of hydro-
carbon vaporizations from tankage. The gas product is compressed to a
pipeline pressure of 350 psig in order to accommodate the latter, and it
is, thereby, suitable for returning to commercial gas. Commingling this
treated gas with commercial natural gas supplies should cause no problems
in heating value or flame control since essentially all of it was commer-
cial gas to begin with. Air-propane blending for Btu control is not
considered necessary. Two treating systems have been rated and estimated,
one for condensing larger amounts of vaporizations from cone roof tankage
58
-------
111
N
>U
m
c
1 (1
1
' Ul
\
(
v&x
0
)
1
Q>
. IV»IOH-
J_Af'VO _J
ROBERT BROWN ASSOCIATES
CARSON, CALIFORNIA
CUSTOMER Environmental Protection Agency
PLANT
Contract //68-02-2838
LOCATION Blanket Gas Pipine
DRAWING NUMBER
153-1-23
REV.
o
KLINCLER VELLUM 102 I-
59
-------
in Case IV and another for condensing smaller amounts from internal
floating roof tankage in Case IVA. These have been considered for comparable
reasons and do not necessarily constitute an optimum design application.
Refer to drawings 153-1-21A and -21B. Treating facilities consist of a
sponge iron guard chamber for the removal of stray inorganic sulfur compounds
at 75 psig. Water removal is then accomplished by indirect refrigeration
to 10 F at 350 psig or glycol absorption at 90°F and 350 psig. Refrigeration
is also used where larger amounts of hydrocarbons are to be removed, such
as emissions from cone roof tanks. Only moderate cooling has been estimated
for relatively small standing losses from internal floating roof tanks.
Both refrigeration and glycol treating are basically provided to reduce
dew points to 10 F in the event that blanket gas is returned to sales.
Hydrocarbons condensed either by sea water in shell and tube exchanges,
or by refrigeration, are returned to terminal throughputs. Returning
condensed hydrocarbons is thereby incidental to the need for returning
pipeline quality gas. The disadvantage of hydrocarbon build-ups
to saturation levels in a combustible blanket gas media is in
the collection and disposition of random condensations whenever temperature
drops occur.
Controlled emissions versus uncontrolled Base Case emissions in
short tons per year and installed vapor control costs are:
EMISSIONS ST/Y COST
H-C NO^
Facility No. 1
Base Case
Case IV
Case IVA
60
H-C
2890
344
339
NO
X
148
148
149
SO
X
111
111
111
CO
40
40
40
Part.
11
11
11
$ M
-
34,300
50,530
-------
H-C NO SO CO Part.
Facility No. 2. ^_ $ M
Base Case 2578 280 20 136 0
Case IV 164 280 20 136 0 13,350
Case IVA 160 281 20 136 0.1 18,530
Although lower operating costs result in Case IVA, the added cost
over Case IV for internal floating pans lessens that control system's
overall cost effectiveness. Also, the benefits of almost total hydro-
carbon recovery to terminal throughputs is not sensitive to the cost
effectiveness. Refer to Tables 1 and 2 for related economic values.
Crudes with sour vapors would cause the release of H~S to atmosphere
from refrigeration units handling ballast emissions as in Case I. They
would also cause more frequent replacement of sponge iron in the guard
chambers of treater units. Very sour crudes may render sponge iron less
practical than a conventional sulfur removal (Glaus) plant. Guard chambers
have been sized and estimated for two changes per year with 25 ppm of
H~S in the feed stream. Refer to paragraph 9.2.10.
61
-------
REMOVED BLOWER
«VI*ION
APPCD ABSORBER
ROBSTT BROWN ASSOQATES
CARSON, CALIFORNIA
CUSTOMER Environmental Protection Agency
PLANT
Contract //68-02-2838
LOCATION Case jy Flow
DRAWING NUMBER
REV.
153-1-9
KUINSLER VELLUM 1021-6
62
-------
ROBERT BROWN ASSOCIATES
CARSON.CALIFORNIA
CUSTOMER Environmental Protection Agency
PIANT Contract ff68-Q2-2??5"
LOCATION Case IVA Flow
DRAWING NUMBER
REV.
153-1-9A
3
VELLUM 1C2I-*
63
-------
5.7 CASE V
Case V differs from Case IV by being totally enclosed and by return-
ing vapors displaced from tankage by ship unloadings back to the ship.
Vapors displaced by barge leadings are similarly returned back to tankage.
In other words, this is a balanced displacement vapor control system between
marine vessels and shore tankage. Uncondensed vapors from refrigerated
ship ballast emissions and from gasoline barge loading emissions are
collectively transferred by blowers to floating roof blanket gas holders
on the basis that these marine vessels have arrived inerted with natural
gas. Although it would rarely occur, net excess blanket natural gas
could be returned to sales when all blanket tankage is full, or as dictated
by schedules. Blanket gas, however, is normally shipped away in tankers,
and treating units have much less importance in this case than they do
in Case IV. Refer to drawings 153-1-10 and 10A for this application to
both facilities, and to Tables 1 and 2 for related economic values.
Of importance here is that only breathing and working loss emissions
are received by flexible diaphram gas holders in Case V, and only standing
loss emissions from tankage with internal floating roofs in Case VA. These
emission losses in the latter case are very small, and the collection piping
and vapor storage from tankage is relatively very small. The only significant
amount of vapor expelled into floating roof gas holders is that from ship
ballasting operations at berth. Since the enclosed volume of this system
is reduced by ballast water displacements, these rates comprise the design
capacity of blanket gases treated and returned to sales. Treating units
are the same in principal as those in Case IV, including a standby flare
in the event that sales cannot accommodate operating schedules. Refer
to drawings 153-1-21A and -21B and paragraph 9.2.10.
64
-------
A restriction to this relatively effective, totally enclosed, vapor
control system is that ships and barges using either of these two terminal
facilities must be blanketed with natural gas, which is thereby removed
by ships unloading into the terminals. This gas should be utilised wherever
crude or gasoline is loaded onto the ships. If ballast is taken into
cargo compartments at sea, some compression, containment, and ultimate
burning of the gas displaced would be more desirable than exhausting it
to atmosphere. Associated retrofitting expenses have not been estimated.
In unit heating values, the import/export ratio of crude entering to
equal volumes of gas leaving at atmospheric pressure amounts to 1100 to 1.
Regarding tanker safety, it takes more air to explode a tanker of fuel
gas and hydrocarbon vapors than one inerted with flue gas. As a matter of
fact, the more flue gas there is with the hydrocarbon vapors (up to about
98 vol % ) the less air is required for combustion , as shown on Drawing No.
153-1-25. However, empty tankers with natural gas contain more combustible
volume than those with flue gas or air in their ullages, and for this reason
require special considerations which are outside the scope of this effort.
In order to provide more operating pressure drop flexibility in this
balanced displacement system, and margin of safety away from storage tank
relief valve pressure settings, the cost of 1/2" thick fixed tank roofs
has been added to this case, instead of the conventional 3/16" thick
roof, and duct sizes have been reduced accordingly. Thus, tank roofs are
4 8
more in line with ship cargo compartment thicknesses and pressure levels ' .
Refer to drawings 153-1-19A and -20A. A further safety measure to prevent
vacuum reliefs from opening on ship compartments is a 60 psig natural
gas repressuring header and hoses to ships that admits gas on pressure
65
-------
TOO LEAN TO BURN
FLAMMABILITY ZONES
FOR NATURAL GAS
100% AIR
21% OXYGEN
79% NITROGEN + INERTS
FOR TYPICAL H-C VAPOR MIX
(WITHOUT H2 OR PREDOMINANT
UNSATURATES)
100% NAT. GAS WITH
GASOLINE VAPORS
100% INERT GAS
%ARE BY VOLUME AT
APPROXIMATE STANDARD
CONDITIONS.
LEL = LOWER EXPLOSIVE LIMIT
UEL = UPPER EXPLOSIVE LIMIT
L«o I. O»T«.
ROBERT BROWN ASSOCIATES
CARSON,CALIFORNIA
CUSTOMER
PLANT
LOCATION
DRAWING NUMBER
153-1-25
REV.
o
66
-------
CUSTOMER Environmental Protection Agency
DRAWING NUMBER
ROBERT BROWN ASSOCIATES
CARSON. CALIFORNIA
PLANT Contract No. 68-02-2838
LOCATION par .;
yanr rnllprMnn Layout
MI.IMCLC*1 VC l-LUM ICBI -
-------
ROBERT BROWN ASSOCIATES
CARSON. CALIFORNIA
KLIN CLEM VELLUM 102 I -
CUSTOMER Environmental Protection Agency
PLANT
Contract No. 68-02-2838
LOCATION Farility #2 Vapor Collection Layout
DRAWING NUMBER
REV.
I53-1-1ZO
68
-------
control to compartments before vacuum relief valves open. Piping costs
have been added to Case V, and retrofitting costs have been estimated.
Since gas holder capacities have been predicated upon supplying
1^1 days of pipeline pumpout displacements in each terminal, gas holders
in this case are as large as those in Case IV. Low and high pressure
blanket gas distribution systems are also similar to Case IV. Refer
to drawing 153-1-23 for typical piping layouts and Table 3, paragraph
9.1.1, for blanket gas control pressure settings.
Controlled versus uncontrolled Base Case emissions in short tons/
year and installed vapor control costs arc:
EMISSIONS ST/Y
COSTS
Facility No. 1
Base Case
Case V
Case VA
Facility No. 2
Base Case
Case V
Case VA
H-C
2890
85
80
2578
67
63
NO
X
148
148
148
280
280
280
SO
X
111
111
111
20
20
20
CO
40
40
40
136
136
136
Part.
11
11
11
0
0
0
$M
-
40,160
56,140
-
14,750
19,960
The cost effectiveness of this control system is comparable to
other systems which have large blanket gas storages. However, it has
been assumed therein that the value of the commercial gas received balances
off that which is delivered to the terminal where the tankers receive
their crude oil or gasoline cargoes. Likewise, no credit has been
69
-------
provided for the gas received from barges. Although treating units
would seldomly be used, they have been estimated to cost as much as
those in Case IV, but direct operating costs and secondary emissions
have been reduced from a 90% loading factor in Case IV to a 10% factor
in Case V. Because of the close coordination needed between vessel
personnel, the wharfinger, and shore tankage operations with this vapor-
balancing system, two full-time operations personnel have been added to
the direct operating costs of this case for both facilities.
No sulfur emissions would be caused by adding sour crude vapors to
this totally enclosed vapor control system. Crudes with sour vapors,
however, would increase the frequency of spent sponge iron disposals
and fresh sponge iron replacements in treating unit guard chambers.
Except to the extent that sulfur is removed in the treating units of
Facilities No. 1 and No. 2, SO would be emitted from the combustion of
x
such exported sour gases elsewhere.
70
-------
o
i&
*2u
i-<*
S-J2
2fl5
k
N
L
<
1
5
y
z
j
lU
o- ^=4
ihfl^
?^S/
H^q^V
g
P?
^. fl Su
A ^ LJ 113
"jjj ^
"»< T
1
01
H
IU
10
4
0
2
>-
h
J
P
LL
<5!
I
I
a:
(j
i
. MO
2.
lt'^;TJ
I2-27-T7
REMOVED
RALLAiT DIJPL- TO TREAT
Jx
«
!/.'D3A%
ROBERT BROWN ASSOQATES
CARSON.CALIFORNIA
KLINCLEU VE ULUM 1O2I -
CUSTOMER Environmental Protection Agency
PLANT
Contract #68-02-^838
LOCATION Case ~SL Flow
153-HO
71
------- An error occurred while trying to OCR this image.
-------
5.8 CASE VI
This case compares with Case IV but utilizes nitrogen for gas
blanketing tankage instead of natural gas. Since this study basically
assumes relatively sweet crude vapors, no attempt is made in this case
to chemically remove sulfurous gases. The intent here is that the
containment, less normal blanket gas secondary emissions, will not
accummulate contaminating, corrosive, or other detrimental ingredients.
Excess blanket gas from cone roof tankage is refrigerated to recover
hydrocarbons and incinerated in this case. That from internal floating
roof tankage is directly incinerated, since about 99% of cone roof
emissions are restrained by internal floating roofs. Marine emissions
are handled as they are in Case I. Refer to drawings 153-1-11 and 11A
for this application to both facilities, and to Tables 1 and 2 for
relative economic values.
Nitrogen has been chosen for this case instead of carbon dioxide
because water vapor condensations are less corrosive and the costs
of the gases are about the same on a volume basis. Annual nitrogen
replenishing charges have arbitrarily been assumed to average that
required for 1 1/2 major storage tank fills per terminal. Nitrogen in
these volumes is priced at $60 per ton delivered as liquid in the Los
Angeles area. Leased equipment costs have been included.
Gas holder capacities are as large as those in Case IV and V. Only
low pressure blanket gas distribution is provided from these holdersf
however, because of the heat denand for nitrogen nake-up at punpout dis-
placement rates. Heat input, from leased ambient vaporizers, allows up to
100,000 SCFH of nitrogen from a leased liquid storage tank, This make-up
73
-------
rate normally supplies gas holders. If unsafe oxygen levels are detected
by storage tank oxygen analyzer-recorders, however, (see drawing 153-1-23,
para. 5.6) a small high-pressure nitrogen distribution line is used to
purge the tank to safe oxygen levels. Alarms are set at large
margins of safety below the upper explosive limit. Nitrogen purges
are automatic with alarms to notify personnel of maloperating conditions.
Valves do not automatically admit high pressure nitrogen blanket gas
on pressure control.
Controlled versus uncontrolled Base Case emissions in short
tons per year and installed vapor control casts are:
Facility No. 1
H-C
EMISSIONS ST/Y
NO
SO
CO
Part.
COSTS
$M
Base Case
Case VI
Case VIA
2890
323
320
148
148
148
111
111
111
40
40
40
11
11
11
-
42,190
50,090
Facility No. 2
Base Case
Case VI
Case VIA
2578
155
153
280
280
280
20
20
20
136
136
136
0
0 14,200
0 18,390
74
-------
.. . .J |
| I j *
DRAWING NUMBER
Environmental Protection Agency
ROBERT BROWN ASSOCIATES
CARSON.CALIFORNIA
Contract #68-02-2838
153-1-11
KI.INCLCM VILLUM IOJI-*
75
-------
ii-i-~7[^:/,Kc;/~. i^!-: ACSCKR [ I 1 I ;f 3 I. ^/-4A
. t.-i-i Vi-v.-» e« . 1 . »rj_. e« L -«"VD -J « I _>o l_:.Tf _I
:2-t- 77|Wr^.o-vct> Tt..^- -- '.....-,,^71 ] I | -\ A- WJJ!&]*
ROr£<7 E.?OWN ASSOCIATES
CARSON.CALIFORNIA
REMOVED HP CONTROL VAUVES I \_ _ j_,
CUSTOMER Eny j ronmpntal Protection Apency ] DRAWING NUM^R
:. /'68-02-2838
153-I-1IA
4
76
-------
5.9 CASE VII
This case, like Case V, is a balanced displacement system between
marine vessels and shore tankage, but it utilizes flue gas as a
blanket gas media instead of natural gas. Purchased nitrogen is not used
in this case, as it is in Case VI, because tankers would be
removing it from the terminals in volumes equal to daily terminal
tanker throughputs. Tankers utilizing terminals in this case are assumed
to already have stack inerting systems, and barges are assumed to be flue
gas blanketed. If such were not the case, however, special accommodations,
which have not been estimated, could be provided. Make-up flue gas is
taken from ship stack gases at berth following their on-board sea water
scrubbing operations. Dock-side turbo-blowers deliver this stack gas
at 1 psig to on-shore floating roof gas holders, the sizes of which
are the same as those for Cases IV, V and VI. Blanket gas from shore
side tankage is displaced into the tanker while it is unloading, so
that instead of the tanker normally filling its own cargo compartments
with flue gas, it is filling the gas holder on shore with displaced vapors.
Tanker inerting capacities are assumed to be at least 125% of their off-load-
o n
ing rate ' , although these capacities are more than ample for supplying
either terminal facility because each tanker leaves 20% ballasted. Also
barges in Facility No. 2 import blanketing flue gas. If flue gas make-up is
needed over that which is available from ships at berth, on-shore inert
gas generators are available. These generators exchange their hot flue
gas heat with sea water in coolers to about an 80 F dew point, which is
Q
approximately the dew point received from ships' scrubbers . Gas holder
storages provide a blanket gas reservoir at 7.0 osi pressure. Only a
77
-------
low pressure blanket gas distribution system is available because of the
low source pressure of flue gas. Refer to drawing 153-1-12 and 12A
for this application to both facilities and to Tables 1 and 2 for
related economic values.
Costs have been added to this case for a 1/8" corrosion allowance
on all interior surfaces exposed to the flue gas blanket, instead of
an application of corrosion resistant paint, in keeping with the
decisions at Valdez. Here, where scrubbed flue gas is also used as
a blanketing media, it was concluded after some study that the cost
of applying corrosion resistant print would be forstalled until such
time that corrosion probles indicated its necessity-^. Accordingly,
the fixed roof tankage for both terminals in Case VII are 3/8" thick,
1/2" being provided to allow more operating pressure drop flexibility
in this balanced displacement control system, as in Case V.
Controlled versus uncontrolled Base Case emissions in short tons
per year for Case VII are:
EMISSIONS ST/Y COSTS
H-C
Facility No. 1
Base Case 2890
Case VII 59
Case VI IA 59
Facility No. 2
Base Case 2578
Case VII 55
Case VIIA 55
NO
X
148
164
164
280
300
300
78
SO
X
111
123
123
20
21
21
CO
40
44
44
136
146
146
Part.
11
12
12
0
0
0
$M
47,130
64,140
18,100
23,800
-------
Here as in Case V, two full-time operations personnel have been
added to the direct operating costs at both terminals in order to
assure close coordination with vessel personnel, the wharfinger, and
shore tankage operations.
Stack emissions have increased due to the need to fire boilers for
Facility No. 1 and diesels for Facility No. 2 from 10% to 90% loads
during a few hours of non-pumping docking time daily, to supply on-shore
blanket flue gas storages. Additional fuel costs have been assumed an
operating contingency cost.
79
-------
ROBERT BROWN ASSOCIATES
CARSON,CALIFORNIA
KLIN CLIP VELLUM 1021 -
CUSTOMER Environmental Protection Agency
PLANT
Contract #68-1)2-2838
LOCATION
Case VII rlow
80
DRAWING NUMBER
REV.
153-1-12
-------
Environmental Protection Agency
DRAWING NUMBER
ROBOT BROWN ASSOCIATES
CARSON, CALIFORNIA
53-1-12
Case VIIA Flow
KLCNSL.ER VELLUM 1021 -8
81
-------
6.0 VAPOR CONTROL ECONOMICS
The cost of each vapor control system has been estimated as an incre-
mental cost to the total installed cost of new Base Case facilities. Vapor
control systems have not been estimated, therefore, as being retrofitted
to existing shore facilities. Engineering overheads and field indirect
costs have been pro-rated to major vapor control equipment as applicable
to the total terminal construction costs. All facility costs comply with
fourth quarter 1977 or first quarter 1978 prices in the Los Angeles, Calif.
area. Utility costs, however, apply to other areas as described below.
Gas and electric power availability in the quantities needed for various
vapor control systems is assumed to be reflected in their billing rates.
Marine vessel retrofitting costs per vessel have been estimated
separately.
82
-------
6.1 COST EFFECTIVENESS
Each vapor control system for each terminal facility, both with cone
roof tanks and with internal floating roof tanks has been process rated
and estimated for total installed cost and direct operating cost. Two
separate sets of utility rates have been used. The cost of each system
has been credited by those hydrocarbons the system returns to the terminal
thru-put. Net annualized costs are divided by those hydrocarbon emissions
recovered over base case emissions, less any secondary emissions. Secondary
emissions are defined, herein, as those resulting from the vapor recovery
system itself.
The cost effectiveness (CE) of each vapor collection system has been
expressed as:
ACC + DOC - CREDITS . . ,, .
in dollars/ton
CE =
U - C
where:
ACC (annualized capital charges ) = total installed cost (TIC)
difference times 0.1715 for capital recovery over a 15 year period, including
0.1315 for depreciation and interest and .04 for taxes, in $/year.
DOC (direct operating costs) = operating and maintenance labor dif-
ference and utility costs for the system, in $/year.
Credits = value of recovered hydrocarbons returned to the terminal
throughput, in $/year. Crude oil has been valued at $10.95 per barrel
with entitlements and gasoline has been valued at $16.40 per barrel.
U = total hydrocarbon emissions in the Base (uncontrolled) Case in
tons/year.
83
-------
C = Hydrocarbon emissions that cannot be controlled in each terminal
with a vapor collection (controlled) system, including secondary hydro-
carbon emissions, in tons/year.
Retrofitting estimates for tankers and barges are not included with
CE computations because the number of specific tanker and barge configura-
tions servicing these two terminals has not been defined. Tables 1 and 2
summarize the cost-effectiveness of each alternative vapor recovery system,
and their cost-effectiveness variables.
The use of segregated ballast in tankers would result in much less
total terminal cost effectiveness than this study has shown, especially
in the more costly cases. In cases IV, V, VI, and VII, for instance,
the majority of capital costs are for tankage emission control. The
rest, for ballasting or barge loading emissions, handles 75% of all
the hydrocarbon emissions from Facility No. 1 and 89% of all those
from Facility No. 2. Segregated ballast would eliminate the more
cost effective ballast emission requirements.
Cost effectiveness values for marine and tankage vapor control
systems vary from $1320 to $4010 per ton of hydrocarbon recovered in
Facility No. 1 and from $699 to $1670 per ton of hydrocarbon recovered
from Facility No. 2. The cause for this unapparent relationship
(the larger facility being less cost effective) lies in the capital
costs needed for handling massive gas quantities over relatively large
areas in the larger facility, and in the gasoline tanker and barge
operations which create large H-C emissions from the smaller facility.
While hydrocarbons recovered from Facility No. 1 are about 2560 T/Y,
those from Facility No. 2 are only 160 T/Y less, but costs vary from
$11 to $69 million for Facility No. 1, compared to only $2 to $29
million for the smaller facility.
84
-------
The cost effectiveness of large and small terminals having like
operations should favor the larger terminal, but perhaps not to the
extent that normal economics of scale prevail in processing plants.
Capital investments for these vapor control cases are more sensitive
to the Ibs./hr. of product handled, (low pressure hydrocarbon vapor)
than are processing plants of their liquid products.
85
-------
6.2 TOTAL INSTALLED COSTS
The total installed cost (TIC) of individual equipment item functions
has been modulated to include current delivered equipment costs to the
Los Angeles area plus all associated materials and labor costs for the
installation and functional operation of that piece (or pieces) of equip-
ment. Paragraph 9.0 presents design criteria affecting TIC estimates.
Equipment vendor prices have been solicited and received. Major vapor
collection and blanket gas distribution piping has been handled as "equip-
ment". Associated equipment costs include erection of the equipment, all
associated piping, electrical, instrumentation, structural, civil, paint-
ing, start-up, engineering pro-rates, construction indirects, spare parts,
and expendable costs necessary to fulfill its operating function. Assessing
these cost increments to each functional equipment item is intended to allow
the resulting module to be extracted from or added to a system in a fairly
realistic manner. Engineering pro-rates approximate 10% of most equipment
modules and construction indirects about 8%. Natural gas and nitrogen
blanket-gas initial terminal filling costs have been included as a vapor
control system TIC.
This study has not optimized alternative equipment details except in
major respects. Judgements have been made in these respects based upon
engineering and operating experience. The accuracy of TIC estimates is in
the order of + 20%. Total installed costs for ballast and barge emission
controls (Case I) equals $1.17 for Facility No. 1 and $1.99 for Facility
No. 2 per B/CD throughput. Total installed costs for marine vessel plus shore-
side tankage vapor control.systems equal from $1.21 to $28.51 per B/CD through-
put for Facility No. 1, and from $2.77 to $34.00 for Facility No. 2. (Cases
II and VILA). Cost intensities rise approximately 20-fold from Case I to
Case VII.
86
-------
6.3 DIRECT OPERATING COSTS
Functional equipment items have also been assessed with operating
and maintenance labor factors, maintenance material, utility rates and
other expendable costs. These operating costs have been reduced by load
factors that have been developed from the transfer schedules described
in section 4.0.
Utility rates have been selected from high and low cost areas for
gas and electric power to industrial users on the East, Gulf and West
-I O
Coast regions of the country:
high set (New Jersey) low set (Texas)
$/KWH* .050 .029
$/MMBTU/HR 3.225 1.275
* These are incremental power costs which include both demand and
usage rates for major industries. Costs have arbitrarily been
escalated 10% from the Federal Power Commission's last publication
of "Typical Electrical Bills" dated January 1, 1976.
Operating and maintenance labor has been considered as costing $25,000/
manyear, including normal benefits and payroll burdens. Terminal facility
staffs have been increased where believed necessary due to vapor control
system complexities. Alternatively, fractional operating and maintenance
labor and material costs have been assigned to equipment modules comprising
each vapor control system.
87
-------
6.4 UNCONTROLLED EMISSIONS
Base case emissions have been calculated for uncontrolled marine opera-
tions and for shore terminals with external floating roofs.
Marine operations expel the bulk of all terminal emissions by ballasting
to 20% of the cargo capacities in unsegregated compartments while at
berth. Ballasting rates will usually start by opening a sea water valve
to fill the compartment until the water level in the compartment approaches
the ship's water line. Ballasting water rate diminishes at this time and
13
if more water is desired in the compartment, ballast pumps are operated.
An overall average ballast rate of 50,000 BPH has been assumed for the
125M DWT tankers, while 20,000 BPH has been assumed for the 35M DWT tankers.
An emission factor of 0.6 Ibs. of hydrocarbon per 1000 gal. ballast water
has been used for crude oil tankers and 0.8 Ibs. of hydrocarbon per 1000
gal. for gasoline tankers. Ballast water has not been used with barging
operations, but gasoline loading into dedicated barges in Facility No. 2
has been given an emission factor of 4.0 Ibs. of hydrocarbon per 1000 gal.
loaded.
Base case external floating-roof tankage emissions have been calculated
from tr.e AP-42 standing loss formula:
, in Ibs/day
88
-------
For several years it has been apparent that the method of calculat-
ing hydrocarbon losses from floating roof tanks, as presented in API 2517
and adopted in AP-42, grossly overstates the losses to be expected
from well-designed and well-maintained floating roof tanks. In 1975
14
it was postulated that the major losses were wind-induced and that a
secondary seal of some sort could eliminate this effect. In 1976
2
experiments by CB&I supported this hypothesis and suggested that losses
much lower than those calculated by the API 2517 method would be more appropriate.
WOGA independently found at this time that emissions determined from this
formula were overstated by some 40 to 60%, according to small scale tests
on single seals. Since then a variety of multiple seals have been prepared
and some installed. These designs range from relatively simple close fitting
wiper blades designed to eliminate wind effects (and hence, the bulk of
the losses) up to full double seal designs that increase the effective
sealing surface as well.
The API has initiated a project to update its technical bulletins on
methods of estimating HC losses from external and internal floating roof tanks,
scheduled for completion in early 1979. A similar project is planned for
fixed-roof tanks.
This report is being prepared before a consensus has been reached,
therefore, regarding the proper emissions factor to be used for new
floating tanks equipped with, an advanced sealing system.. Accordingly, it is
necessary to fall back upon API 2517 as a base case, recognizing that the
cost effectiveness of various control alternatives will be less when this
consensus has been reached. Standing losses are shown
89
-------
below with a 0.1 seal factor to illustrate the overall terminal effect of
having double seals with that degree of sealing integrity. Other variables
in the above formula are quantified below for both terminal facilities:
Facility No. 1
M (vapor mol. wt.) crude = 52
P (true vapor pressure) crude = 4.6 psia
D (tank diameter)
V (wind velocity)
K (tank type)
K (paint factor)
K (commodity factor)
Number of tanks
340 feet
6 mph
0.045
0.9
0.84
24 crude oil
Facility No. 2
crude =52
gasoline =70
crude = 4.6 psia
gasoline 6.8 psia
262 feet
6 mph
0.045
0.9
0.84 crude
1.0 gasoline
6 crude oil
2 gasoline
Normal standing losses have also been calculated for slops tanks and an
API separator skim oil tank in each facility. Ballast water and oily
sewer water separator losses have been estimated, as well as fugitive
terminal emissions and transient (maintenance) emissions. Fugitive
emission factors for these terminals have been modified from oil refinery
emission factors in AP-42 by reason of the basis of refinery throughputs
versus terminal throughputs, and of the difference in the items of equip-
ment and their operating conditions in refineries and terminals. Fugitive
emissions have been thereby reduced 57% both for the base case conditions
and for secondary emission calculations.
Stack emissions from oil tankers and tug boat traffic within 5 miles
of each terminal have been calculated from a scenario of their operations.
90
-------
Large tankers and sea-going tugs for Facility No. 1 burn No. 6 fuel oil,
while smaller tankers and harbor tugs at Facility No. 2 use only diesel
oil.
Resulting uncontrolled base case pollutants have been calculated
in short tons/year as:
H-C
NO
SO CO Part.
x
Facility No. 1
Normal
Marine
Tankage
Fugitive
Transient
(No. 6 F,0)
Total
Facility No. 2
Normal
Marine
Tankage
Fugitive
Transient
(Diesel Oil)
Total
K = 1*0
s
2182
659
14
35
2890
2295
231
8
44
2578
*
K - 0.1
s
2182
66
14
35 148
2297 148
2295
23
8
44 280
2371 280
111 40 11
111 40 11
20 136 0
20 136 0
ideal double seal operation.
91
-------
6.5 CONTROLLED AND SECONDARY EMISSIONS
Controlled emissions are those atmospheric emissions which occur
after vapor recovery facilities are installed and in operation. They
include "secondary emissions", which are those resulting directly from
the vapor recovery equipment itself. Where incineration is used for
converting hydrocarbons to CO- and H~0, emission factors for combustion
pollutants have been used from EPA AP-42, second edition, for industrial
process boilers burning natural gas. These are considered normal emissions.
Only hydrocarbon emissions, however, apply to the economic cost effective-
ness concept. Fugitive and transient emissions other than hydrocarbons
have not been quantified. Total normal, fugitive, and transient secondary
hydrocarbon emissions are summarized below in short tons per year for each
vapor control system:
SECONDARY HYDROCARBON EMISSIONS ST/Y
Case No.
I
IA
II
HA
III
IIIA
IV
IVA
V
VA
VI
VIA
VII
VILA
Facility No. 1
263
263
29.8
28.8
293.3
291.1
295.9
291.5
31.0
30.5
274.7
272.5
11.1
11.1
Facility No. 2
98.6
98.6
8.4
8.5
107.4
106.7
112.0
108.0
9.3
8.8
102.7
101.2
3.3
3.3
-------
Nitrogen and flue gas blankets (Cases VI and VII) emit only saturation
quantities of hydrocarbon at most. Quantification of non-photochemically
reactive hydrocarbons is not within the scope of this work effort, but a
major portion of the secondary emissions from totally closed vapor control
systems with natural gas blankets (cases IV and V) would be methane.
Only pilot gas emissions have been assigned to emergency flares. Refrig-
eration compressor seal factors have been modified from refinery compressor
seal factors in AP-42 by quantifying the difference in the average number
of seals and temperatures involved. Fugitive terminal emission factor
adjustments to oil refinery capacity factors have utilized "controlled"
evaporative source emission factors from EPA-475/3-76-039, August 1976,
"Revision of Evaporative Hydrocarbon Emission Factors", Attachment C.
93
------- An error occurred while trying to OCR this image.
------- An error occurred while trying to OCR this image.
-------
6.6 MARINE RETROFITTING COSTS
Retrofitting costs have been estimated for typical barges and
U. S. Registered tankers with manifolded venting systems and stack risers.
Two hose connections are added to compartment vent manifolds so that
either the port or starboard side of the vessel can be connected to dock
hoses. A gate valve and flame arrestor is provided at each hose connection,
and a gate valve is provided at each stack riser so that normal compartment
venting can be restored at any time. Refer to drawings 153-1-13 and 15,
Tanker hose connections are 18" and 12" at Facilities No. 1 and
No. 2, respectively, and barge connections are 8" at Facility No. 2.
Retrofitting labor and material costs have been estimated per vessel
as follows:
Facility No. 1 Facility No. 2
Cases I, II, III, IV, VI
Cases V and VII
120M DWT Tanker
$58M
$74M
35M DWT Tanker
$23M
$32M
Barge
$4M
$4M
for tankers and barges which already have single vapor collection headers
and stacks, as illustrated on the referenced drawings.
Retrofitting costs have not been included in cost effectiveness
calculations because the number of various tanker and barge configurations
that would service these hypothetical terminals have not been defined.
96
-------
7.0 ENERGY RESOURCE CONSIDERATIONS
In the cost-effectiveness evaluation, cognizance has been taken
of the value of hydrocarbon vapors collected and recovered with the
monetary costs incurred. It is to be noted, however, that in the final
analysis, the total net loss or gain of energy may transcend cost
effectiveness. To this extent, consideration should be given to the
fact that passive vapor control systems offer better thermal effective-
ness than those which consume energy.
Cases VA and V offer the best thermal efficiencies in this study,
considering equivalent utility heat values and Cases II and IIA offer
the worst thermal efficiencies. The cost effectiveness of Case IIA,
however, is better than Case V. In general, the more capital intensive
vapor control systems evaluated are relatively more thermally efficient.
Such considerations give reason to weigh energy resource along with
monetary effectiveness.
97
-------
8.0 SAFETY AND RELIABILITY CONSIDERATIONS
The costs of providing best available technology for public, per-
sonal, and property safety have been estimated, but while it is possible
to quantify thermal and cost effectiveness, a comparative evaluation of
total terminal facility safety can hardly be quantified, or even ranked.
Historically, designers and operators in the petroleum industry have made
efforts to eliminate "naked oil", and avoid explosive or potentially
explosive vapor spaces. The blanketing of vapor spaces in tanks or
vessel compartments, and the elimination of vapor spaces altogether by
the use of floating roof tanks, are such examples. Designs have avoided
the need to depend upon seals, purges, or flame arresting devices to
prevent the ignition of large volumes of gas. The intrinsic nature of
vapor recovery and/or vapor incineration systems, however, intimately
involves these concerns. It must be recognized that these systems
necessarily represent an inescapable compromise with safety in the
interest of vapor recovery.
The distinction between potential emissions and controlled emissions
has been of increasing environmental concern. It is sometimes necessary
to evaluate a facility in terms of what would happen upon failure of the
emission control provisions. In such instances, external floating roof
tankage with small vapor volumes, and segregated ballast offer clear
safety and reliability advantages. These advantages are either contrary
98
-------
to or not discernible in cost or thermal effective analysis.
In addition to the concerns with controlling large low pressure gas
volumes are those relating to fires in large cone roofed crude oil tanks,
as described in paragraph 9.1.4 Fire Protection. In view of these con-
cerns, special design considerations and sizeable investments must be pro-
vided to vapor control systems to enhance safety and reliability:
Operating pressures well below relief valve settings,
Oxygen analyzes monitoring below each tank vacuum relief valve,
at marine vessel vapor collection headers, and at blanket gas
storages.
Automatic gas purges to reduce unacceptable oxygen concentration levels,
Flame arresters at major equipment,
Remote manually controlled block valves on major vapor collection
branches,
Circulating butane through marine vessel vapor collection systems,
Water seals before incinerators and refrigerators,
Automatic flare and blower shut down at loss of forward gas-flow to
flares,
Alarms to acknowledge automatic corrective actions taken for high
oxygen levels, no gas-flow to flares, etc.,
Enclosed drains to covered API separators,
Additional fire water capability for cone roof tankage,
Detonation barriers,
Strategic equipment and piping layouts.
Costs for the above provisions have been included in each alternative vapor
control case evaluated.
99
-------
9.0 VAPOR CONTROL DESIGN CRITERIA
Although basic design criteria is not expected to affect cost
effectiveness significantly, design concepts have assumed 3,000 psf
soil loadings (other than that for tankage), negligable frost lines,
no snow loads, ambient weather above freezing, 100 mph winds, 70 F
sea water temperatures, and essentially level terrain. Vapor temperatures
in marine vessel and shore tankage storages averages 75 F annually and
wind velocity averages 6 mph. Siesmic loads have not been calculated.
Some refrigeration power loads are high. Transformer capacity from
feeder voltage to 4160V has been assumed to be either available or supplied
by the power company and reflected in the power billings. All power
lines having 110V or less are run overhead, while the rest are run below
grade. NEMA Class I Group D Division 2 electric area classification has
been assumed throughout.
Pipe supports at grade are assumed to have adequate space available
for vapor control system piping. All large vapor collection and
distribution piping is elevated on shore, while that on piers has been
estimated with having special support brackets.
Land value has not been assessed to design arrangements, but
all vapor control system equipment has been layed out on terminal plot
plans for estimating pipe lines and electrical runs. Ample room is
available, including ground flares using 250' x 500' areas because of
low feed gas pressures. Refer to drawings 153-1-1 and -3, p. 21 and
26, for general plot plan arrangements.
100
-------
9.1 SYSTEM SAFETY FEATURES
Any vapor control system in a terminal must be as automatic as re-
liability and the best technology will allow, because the primary func-
tion and operating objective is to transfer bulk petroleum quantities.
Operating attention must be alerted, however, whenever a potentially
dangerous condition develops in the control system, and an automatic
sequence of operations should be actuated, wherever feasible, to mitigate
the dangers. This means a shut down of the vapor control system as a
last resort. At this time, fixed roof tanks breathe through their
PSVS (conservation vent valves), which is when oxygen analyzing control-
lers play an important part in preventing explosive gas mixtures.
Instrumentation has been estimated to accommodate the above phil-
osophy in each vapor control case. Alarms have been estimated both
locally and in a centralized terminal control room, and gas purges are
automatically actuated in all cases, for excessive oxygen levels, in
both terminal facilities at the following locations:
1. Each fixed roof storage tank, including slops and skim oil tanks
2. Each flexible diaphram gas holder
3. Each floating roof gas holder
4. Covered API separator
5. Marine vessel vapor collection headers, including Case VII.
Totally enclosed CasesV and VII have automatic emergency bleeds
to atmosphere at very high oxygen levels.
Case VII does not have a reliable automatic source of purge gas, except in
101
-------
item 5 above where a slave butane circulating system is available from
the marine emission refrigerator. When a high oxygen alarm sounds for
any item 1 through 4 above in Case VII, flue gas from ships' stacks or
from inert gas generators are manually attended. Liquid nitrogen
in pressure storage is available for reasonable emergency purges in
Case VI. When oxygen levels have been restored by these gas purges,
oxygen analyzing controllers automatically stop the alarms and purges.
Local overrides are provided in each case. Similar automatic oxygen
alarm and purge systems have been successfully operating in natural
gas blanketed oil refinery tank farms in recent years-'-" .
9.1.1 Control Instrumentation
Refrigerator units and blowers are activated by a flow switch in
collection headers for marine vessel emissions, and by a disc level
switch in flexible diaphram holders for tankage emissions. Blowers of
refrigeration vapor effluents are activated with the refrigerator unit and
blowers of flue gas from ships' inerting systems are manually activated.
Treating units are manually set to operate at a selected flow rate, as
are inert gas generators for Case VII. These operations are relatively
complex and need to be lined out during start up for fairly constant
production rates. Nitrogen flow for Case VI is activated from liquid
nitrogen supplies automatically for oxygen control, as mentioned above,
and manually for gas holder inventories. Conventional switch gear
technology for NEC Class I Group D Division 2 areas is applicable throughout.
Best available technology for oxygen analyzer-controllers is with
the use of sensor cells that produce a linear millivolt output potential
with concentrations of oxygen only. These controllers have been deve-
102
-------
loped to continuously monitor and record oxygen concentrations, except
during scheduled three-minute cycles once every hour while they automat-
ically calibrate their output with ambient air. Cell life lasts at least
9 months with that amount of oxygen exposure and these small cells are
very easy and inexpensive to replace. When output voltage drifts away
from that representing 20.9 vol. % during calibration cycles, either an
alarm can call an operater's attention to the need to manually recalibrate
the instrument, or recalibration can be automatic until such time that
the extent of deviation calls for cell replacements. In view of the
quantity of these controllers proposed for the two terminal facilities,
and to minimize maintenance staffing, automatic self-calibrating instru-
ments have been estimated in this study.
Tankage blanket gas control valves are self-contained pressure
regulators that are sensitively counter-weighted against atmospheric
pressures. Table 3 illustrates their settings for each vapor control
case, as shown on flow drawings 153-1-7 through 153-12A. Atmospheric
vacuum relief valves settings equate to a 25 psf vacuum on the basis that
fixed tank roof live loads are designed to support 50 psf, 25 of which
is for vacuum. Snow loads would add to this 50 psf live load. Pressure
relief valve settings in Table 3 show pressure ranges within which
pressure regulators operate.
103
-------
TABLE 3
BLANKET GAS PRESSURE CONTROL VALVE SETTINGS IN OSIG.
(Ounces Per Square Inch Guage)
Case
II
III
IV
V
VI
VII
Atmospheric
Pressure
Relief
Setting
+ 0.85
+ 0.85
+ 0.85
+ 2.27
+ 0.85
+ 2.27
BLANKET GAS SOURCE
Floating Roof Holder
start
to open
- 0.1
- 0.1
- 0.5
- 0.5
full
open
- 0.5
- 0.5
- 1.3
- 1.3
Purchased Gas
start
to open
- 0.5
- 0.5
- 0.9
- 0.9
full
open
- 1.3
- 1.3
- 1.3
- 1.3
Atmospheric
Vacuum
Relief
Setting
- 2.78
- 2.78
- 2.78
- 2.78
- 2.78
- 2.78
Flares and incinerators, both being multi-burner ground flares,
are protected from flash-backs by:
1. Enrichening lean or explosive gas mixtures with hydrocarbon
concentrations slightly too rich to burn.
2. Providing an enclosed 3" water seal, continuously maintained
automatically with conventional instrumentation.
3. Providing a continuous fuel gas bleed down stream of the water
seal and a flow switch so that if a positive gas flow towards
the burners is not detected, pilot gas is shut off, feed blowers
are deenergized, feed blower discharge valve is closed, local
alarms are sounded, and an alarm in the centralized control
is sounded.
Pilot flame detectors are also provided that activate the events listed
in Item 3 above if pilot burners are extinguished. A remote control
104
-------
panel provides a means of manually igniting pilots.
9.1.2 Operating Features
With daily functional operations of each vapor control system being
substantially automatic, manual operating burdens imposed upon either
terminal facility by vapor control systems is limited to handling
marine-to-dock hose connections, rectifying emergency upsets, and
managing blanket gas inventories.
Emergency upsets are manifested by the alarms mentioned above,
namely from oxygen analyzers or flaring operations. Excess oxygen con-
ditions are rectified by natural gas, nitrogen, or flue gas purges,
consistent with safe blanket gas inventories. If such purges de-
plete those inventories, then a shut-down decision needs to be made.
The major variables involved in this decision involve:
1) Scheduled tanker receipts and alternatives
2) Scheduled pipeline deliveries and alternatives
3) Availability of blanket gas from outside sources
4) Length of time required for maintenance to repair the situation
causing the high oxygen levels
5) Shutdown and startup time and manpower availability.
Storage tanks can be removed from service without interrupting operations
of the other tanks. In view of the large ducts and very low pressure
blanket gas conditions involved, aluminum slip blinds can be used between
flanges having jack screws, instead of investing in many large valves
which are rarely used. Valves must be used, of course, wherever a line
could reasonably need to be blocked out fast for safety reasons. Valves
have been estimated in vapor headers between groups of tanks, for in-
stance, as a fire safety precaution.
105
-------
Redundancy of equipment for operating flexibility is limited to
blowers and small refrigt.1 at ,s. condensate pumps. Other items of
equipment are too costly and massive for duplication in such an ancilliary
function as vapor control. Certain safety instrumentation is redundant.
Shutdown and start-up procedures must avoid large volumes of gas
passing through the explosive range, and the dryness of possible pyrophoric
iron sulfide surfaces of equipment which will be exposed to air.
Storage tanks to be taken out of service from a gas blanketed system
should be first filled with sea water (using the fire water pump) and
then blinded from other tankage. The small ullage remaining above the
water should then be swept gas free by purging with nitrogen from one
side of the roof to the other. Atmospheric emissions therefrom are
transient. With air vents open on the tank roof, sea water should drain
from the tank through the API separator until no visible oil appears, after
which it could run directly to sea until the tank is almost empty. Last
portions should go through the separator again. Tank shell manholes
are now ready to open. Interior surfaces are wet and iron sulfide
has little, if any, fuel to ignite. Limited preventative maintenance
may even be justified for checking fixed roof tank relief valves and
cleaning flame arresters.
Appropriate operating labor has been added to the equipment
modules that comprise each vapor control system.
9-1.3 Maintenance Features
Maintenance burdens imposed by vapor control systems are more
extensive than operating burdens because of the need to maintain reliab-
ility without redundancy in highly automated complexes. The handling
106
-------
of sour crudes greatly magnifies this difference, although operating
problems also are increased. A primary feature complicating maintenance
problems in vapor control systems handling sour crude vapors is the
potential danger of H.S to personnel, and of fires from pyrophoric
iron sulfide. Although the terminal models in this study does not define
the crude sulfur content, it is assumed in this section that some H-S
prevails. Maintenance problems with H9S are most imminent when install-
ing slip blinds in large vent lines to isolate storage tanks. Here
mobile cranes are needed to handle large aluminum blinds and impact
wrenches. Scaffolding may be needed. In any event, safety-tested
gas masks must be worn during the opening, spreading, blind-inserting,
and flange-tightening operations. Similar precautions and equipment is
needed for removing the blind. Other maintenance activities regarding
vapor control include that for:
- Inspecting and maintaining external floating roof seals
Rotating equipment
Blower packing glands and bearings
Cold refrigeration pumps
- Flexible diaphram inspections and replacements
- Floating roof gas holder fabric (dry seals)
- Fixed roof relief valve checks
Electronic control panels and switch gear
- Refrigerator compressors (reciprocating)
- Gas treating units
Compressor maintenance
107
-------
Sponge iron replacements*
Sea water shell and tube exchanger cleaning
- Flame arresters
- Oxygen analyzer cells
General instrumentation
* Sponge iron replacements vary in frequency with H~S concentra-
tions and the rate of natural gas circulations.
Maintenance labor and material costs have been added to the equipment
modules that comprise each vapor control system. Negative values have
been assessed to fixed roof tankage for the savings in not continuously
inspecting and repairing external floating roof seals.
9.1.4 Fire Protection
It should be recognized here that above ground fixed roof tankage
of the sizes in Facilities No. 1 and No. 2 is not normally used in the
petroleum industry for crude oil because of the potential hazards
caused by fires. Crude oil tankage usually contains a bottom layer
of water. Heat is eventually convected to this layer from surface
fires, resulting in boil-overs that can literally spill all of the
burning crude oil from the tank onto surrounding areas. Resulting
flame impingements, especially down wind, and radiation, can overheat
adjacent cone roof tankage. Atmospheric pressure reliefs will ignite.
If overpressures and/or heat destroys adjacent cone roofs and their
crude oil contents are ignited, a holocaust can result. Gas-blanketing
fixed roof tanks greatly reduces the possibility of ignition, but it
108
-------
does not eliminate it because of human errors, particularly by over-
filling, or not following established procedures while taking tanks
in and out of service.
National Fire Protection Assoc. requires 9080 GPM and 5390 GPM
of fire water for cooling cone roof tank surfaces in Facilities No. 1
and No. 2 respectively, compared to only 150 GPM and 112 GPM for
either external or internal floating roof tanks, because of the relatively
minor coverage between floating roof seals (or foam dams) and tank
shells . The difference in cost for additional coverage has been
charged to all cases (not alternative cases) except Case I, including
additional form concentrate inventories. Foam chambers, or subsurface
foam inlets, and fire water monitors, are considered to be consistant
in all cases with that in the Base Case for both terminal facilities.
9.1.4.1 Flame Arresting - Cost estimates have included circular
crimped ribbon type flame arresters-*-'»19 located on dockside piping
at each ship hose connection, at the inlet to each refrigerator and
incinerator (flare), at the inlet header from tankage to flexible dia-
phram receivers, and at the inlet header to floating roof gas holders,
even with non-combustible blanket gases. UL approved flame arresters
are available in smaller sizes (i.e.: 10"-12") and can be alternatively
used in a battery of parallel units between two headers. Flame
arresters are not located on storage tank vents, in keeping with the
20
API Committee on Safety and Fire Protection . Water manometer taps
should be provided to periodically test the pressure drop across each
flame arrestor, and maintenance should schedule fast clean-out procedures
with minimum shutdown time and expense for each arrestor. Sour crudes
109
-------
will increase the rate of arrestor plugging.
Bursting discs are located at the extremities of vapor collection
headers in a manner such that pressures resulting from normal burning
of combustible gas mixtures in the headers can be relieved without
serious equipment damage. Normal burning is meant to be that when
the gaseous products of combustion flow away from the flame front and
have a density and pressure less than that of the unburned gases.
Detonation occurs when the inertia of the products of combustion cause
substantially higher pressures than the fresh unburned gas before the
flame front. Here bursting discs and flame arresters have little, if
any, value unless they are arranged as detonation barriers.
9.1.4.2 Detonation Barriers - The range of gas concentrations in
volume percent with air that can result in detonations are-"-°:
Flammability Limits Detonation Limits
lean rich lean rich
Methane 5.3 13.9 8.5 11.0
Gasoline 1.3 7.3 2
The radial pressure exerted by detonation against the sides of a
collection header is directly related to the initial pressure of the
gases. At these low vapor control pressures, detonation pressure of
a methane-air mixture may only be from 500 to a peak of 1500 psia radially,
21
while axial pressures would be 1 1/2 to 3 times higher . Detonation wave
velocity is relatively low also because of the low initial pressure,
perhaps about 1000 feet per second. Some detonation velocities reach
1 O
6000 FPS . Consequently, detonation barriers may not be out of the
110
-------
question for vapor control systems. One such arrangement is shown by
21
drawing 153-1-24 . Time has not been available in this study, however,
for locating and designing such barriers, but estimated contingencies
have been provided for their application.
Ill
-------
DETONATION
SOURCE
TO
PLAMT ARE.A
RUPTURE DISC.
I"RASCHIQ RINGS OR
OTHER PACK IMG.
MATERIAL
STEEL GRID AMD
SUPPORTING SCREE-kl
_MO , » DAT!
JH
H
ROBERT BROWN ASSOCIATES
CARSON.CALIFORNIA
CUSTOMER ENVIRONMENTAL PROTECTION AGENCY
PLANT CONTRACT #68-02-2838
LOCATION DETONATION BARRIER
DRAWING NUMBER
153-1-24
REV.
O
KLINGLER VELLUM 102 1-8
112
-------
9.2 EQUIPMENT RELIABILITY
As mentioned previously, the primary function of oil terminals is to
transfer bulk petroleum liquids and not to operate vapor control systems.
Equipment reliability is obviously important, not only because it is
usually unattended, but also because large gas volumes can become poten-
tially dangerous by the inclusion of oxygen. The magnitude of these
equipment investments do not allow reliability in the form of redundancy
in most instances. Consequently, the equipment used should be intrinsi-
cally high in quality and commercially proven in the application employed.
9.2.1 External Floating Roofs
Tankage having these roofs apply to the Base Case and Case I vapor
control systems for both terminal facilities, including slops and skim oil
tanks as well as main tankage. Tank seals have been priced for secondary
seal arrangements, having a bottom mechanical shoe-type seal and a top
fabric wiper seal. Helper springs on top seals are not considered consis-
tant with tank seal factors (K ) of 1.0. Since alternative vapor control
systems compare only the difference in cost and emissions from the Base
Case, other storage tank features, such as the tank shell and floor con-
struction, foundations, and general appurtenance costs, cancel out. No
covers are assumed on the API oily water separators for the Base Case and
Case I facilities.
Tanks in Facility No. 1 are only 48 ft. high by 340 ft. in diameter,
presumably to accomodate a 3000 psf soil loading. It should be noted
that this height is not as economical in cost per unit capacity as is a
60- foot high tank, especially with floating roofs. This feature has
113
-------
appreciably lessened the cost effectiveness of vapor control systems with
internal floating roof tanks, in this study.
It is believed by RBA that if floating roofs were evaluated with
special double seals, instead of single seals, seal factors of as low as 0.1
could be obtained.
9.2.2 Internal Floating Roofs
Tankage with these roofs apply to all alternative vapor control
system cases. API separators in these cases are assumed to have conven-
tional coverings, but slops and skim oil tanks in these alternative cases
also have internal floating roofs.
Reliable designs for floating roofs in tanks 340 ft. or 262 ft. in
diameter would be difficult. However, they have been priced for having
single decks over a buoying means, and with a single fabric seal. Costs
were not obtainable for a lever shoe-type seal and upper wiper seal, comparable
in integrity to that of external floating roof seals, since these are not a
currently conventional design for such tankage^.
Fixed roofs have been priced for being 3/16" thick steel in all cases
except VA and VIIA. In these cases tank vapors are in balance with those
from ships at berth and 1/2" thick steel fixed roofs have been priced in
order to provide more operating pressures available below tank relief
valve settings. Case VIIA has an additional 1/8" corrosion allowance.
All fixed roofs are designed for a 50 psf live loading, 25 of which is for
vacuum. Snow loads would be added to the 50 psf live load. Pressure
relief is set by the weight of the roof, or 0.85 osig for 3/16" thick
roofs and 2.27 osig for 1/2" thick roofs. Vacuum relief is set at 25
114
-------
psf, or 2.78 osig vacuum. Since natural gas, nitrogen, or flue gas
blankets are employed in all alternative cases, except Case IA, only this
latter case has internal floating roof tankage with air vents in the upper
shell. Otherwise, blanket gas enters and leaves each tank through nozzles
as near to the center of the cone roof as structural roof supports will
allow and no vents, of course, exist. Refer to drawing 153-1-23 p. 59.
Only the price difference between these fixed and floating roofs
have been compared to the Base Case external floating roofs. The price
of painting all interior ullage surfaces to prevent carbonic acid corrosion
has not been added to the internal floating roof tankage costs for Case
VIIA, which utilizes flue gas for blanketing, in view of the decisions
made for Valdez. Here, where scrubbed flue gas is used as a blanketing
media, it was concluded after some study that the cost of applying errosion
resistant paint would be forestalled until such time that errosion probes
indicated its necessity . This approach has been applied to Case VII
as well.
9.2.3 Cone Roofs
Tankage with these roofs apply to all cases except Case I, where
external floating roof tanks remain as in the Base Case. All cone roofs,
including API separator covers, are of 3/16" thick steel, except those
in Cases V and VII where tankage vapors balance with those from tankers
at berth. Cone tank roofs in these cases are 1/2" thick steel to provide
for more operating pressures as mentioned above, and Case VII has an
additional 1/8" corrosion allowance. Pressure relief settings are as
described above. Blanket gas enters and leaves through nozzles as near to
the center of the cone roof as structural roof supports will allow.
115
-------
9.2.4 Flexible Diaphram Gas Holders
These gas holders are designed to collect net blanket gas expulsions
from cone or internal floating roof tankage at storage tank ullage pressures
of about 0.5 ounce per square inch (osi) for tanks with 3/16" thick fixed
steel roofs (that relieve to atmosphere at 0.85 osi) or at storage tank
ullage pressures of about 1.35 psi for tanks with 1/2" thick fixed steel
roofs (that relieve to atmosphere at 2.27 osi). Thus, the flexible diaphram
in these gas holders must be as light in weight as durability will allow.
Diaphrams are contained in steel tanks with self-supported roofs, and
are attached midway up the shell. Live roof loads of 25 psf have been
cost estimated. Snow loads would require additional loading. Full
bag volumes occupy to about 90% of the tank volume.
Vapor collection ducts have been sized and costs estimated for Buna-N
diaphrams with nylon inserts that weigh 0.2 osi. This material has been
used successfully in similar service at only 0,144 osi . A spare
blanket should be constantly in warehouse stock, however. A heavy 1/2"
thick steel disc in the center of the diaphram keeps it from becoming askew
in the tank as it rises and falls. A level switch outside of the enclosed
steel tank housing the diaphram, operated by the level of the center disc,
energizes tank effluent blowers or refrigeration units when the tank is
about 90% full, and deenergizes them when the center disc reaches the
bottom of the tank.
The size of most flexible diaphram tanks estimated for these models
73
are believed to be about maximum for those in commercial use . Larger
gas volumes, even at the same low pressures, are stored in counter-
weighted floating roof type gas holders, which are much more expensive
116
-------
per unit volume, as described below. Flexible diaphram tank groups have
been sized to hold at least one hour of maximum storage tank vapor expul-
sions at both terminal facilities.
9-2.5 Floating Roof Gas Holders
These gas holders are designed to recycle blanket gas for terminal
tankage in order to minimize the need for new gas supplies. Thus, an
operating pressure of 12" water pressure (6.91 osi), which can be supplied
by a blower, has been selected as a reservoir pressure.
Dry-seal type telescopic holders with tee fenders and flexible Buna-N
vinyl seals have been cost estimated as being subcontracted for construc-
tion in place. Group capacities have been based upon supplying terminal
throughput pump-out replacements for 1 1/2 days without a ship delivery.
Total volumes amount to 12.5% and 17.7% for cone roof and internal floating
roof tanks, respectively, at Facility No. 1. Those at Facility No. 2
collectively amount to 14.0% and 21.6% of the cone roof and internal
floating roof tankage there. The sizes of these individual holders has
been limited to large commercial applications, generally under 200 ft. in
diameter and height. The cost of these holders is the major reason for
the cost ineffectiveness of Cases IV, V, VI, and VII, and the reason for
not providing 100% terminal storage volumes for blanket gas.
9.2.6 Blowers
Blowers are motor driven, aluminum, centrifugal type, industrial
exhaust fans with axial blades. Seal leakage is restrained by graphite
base packing seals with stuffing box and lantern rings. Blowers have
been rated for a differential head of 20" water at actual inlet gas densities,
117
-------
except flue gas boosters at dockside and boosters to ground incinerators
or flares. All are aluminum construction, graphite seals, and explosion
proof motor drivers and have been selected for non-sparking features, in
Class I Group D Division 2 electrical areas. Similar blowers- have been
used successfully in explosive gas mixtures in other industries .
Blower services have been spared mostly by having 3-50% units. In
very small sizes they have been allowed a 100% spare. In tanker ballast
and barge loading emissions services, one unit has been provided for
each berthing location. Flue gas blowers that transfer stack gas to
floating roof holders in Case VII and those supplying waste gas to flares
and incinerators, are turbo-type, having a differential pressure of
1.5 psi.
9.2.7 Refrigerator Units
These have been priced as skid-mounted units with most of the
operating equipment in a pressurized housing for NEC Class I group D
Division 2 areas. Vapors are refrigerated and exhausted at atmospheric
pressures and liquid water condensate is pumped to the API separator.
Condensed hydrocarbons are pumped by the units to about 50 psig at
the coldest temperatures reached by refrigeration. When essentially all
hydrocarbons heavier than ethane are to be condensed, these temperatures
are about -170 F. Partial condensations, such as those required for re-
moving only tankage vapprizations from a saturated blanket gas media,
hydrocarbon liquids are pumped away at warmer temperatures, such as -80 F
or so. The more hydrocarbons are condensed and removed, the less
thermally efficient the unit becomes because these removals are not
118
-------
heat recoverable. Vapors leave the unit from 0° to 40°F, depending upon
how much hydrocarbon is condensed.
Continuous operations are realized by alternating coil sections for
refrigeration and defrosting. Liquid water is first removed at the
hydrate point, about 35°F. Additional water is removed as liquid by
defrosting. Where flue gas is used in Case VII, and essentially all
hydrocarbons are condensed, about two-thirds of the C0~ precipitates at
-152°F at the partial pressures involved, and then sublimates upon defrost-
ing. The remaining CO stays in the vapor state. Installed horsepower
has been estimated for start-up operations. Cost also include constant
monitoring capability. Marine emission refrigerators in all cases require
a slave n-C/ circulating system to increase hydrocarbon concentrations in
marine collection headers above explosive limits. Complete condensation
and vaporization mechanisms for this compound in the proper quantities
for all cases have been estimated.
Very little is known about hydrate formations at these low temp-
eratures and atmospheric pressures, but it is not expected to be a major
problem . Hydrates are an accumulation of water molecules that capture a
hydrocarbon molecule, and most frequently occur when high pressure, water
saturated, gas is depressured. About 17 mols of water are needed for
hydrating 1 mol of propane, and 7.9 to 8.5 mols are needed for 1 mol of
ethane . Hydrate formation would be collected in a solid state with ice,
and that portion of the propane and lighter hydrocarbons that do form
hydrates can be vaporized at defrosting temperatures. Even if vapori-
zations were collected and chilled again in the refrigeration (nonde-
fresting) coils, a build up of light ends may not liquify and consequently
escape to atmosphere as vapor. However, neither the amount of water vapor
119
-------
available at these cold temperatures, nor the low operating pressures,
are typical of the prerequisites for hydrate formation. Even though
traces of H-S may promote hydrations, and the propane constituent of crude
vapors is significant, no evidence is known that precludes the ability of
staged refrigeration to liquify propane and heavier hydrocarbons at -170°F
and 14.7 psia. It is known that gasoline vapors can be so condensed without
25
hydrate problems
Of more concern is the plugging up of refrigeration heat transfer
surfaces with solid COj in Case VII. None-the-less mechanical design
is expected to be able to provide enough heat transfer to accomplish the
required temperature levels.
9.2.8 Flares and Incinerators
Both incinerators and flares in this study are multi-tipped
vertical burners at grade having pilots and flames which are generally
concealed from ground-level views. Both are arrangements of vertical
pilots and burners having below grade header manifolds and a remote
ignition system. Adequate fuel gas is admitted into the upstream vapor
lines of so-called "incinerators" so that the mixture can sustain com-
bustion, i.e.: having at least 150 Btu/CF. Technically speaking, incin-
erators handling gases that are premixed for combustion become flares,
flares being defined as that which burns gases that sustain their own
combustion.
These banks of vertical burners require inlet waste gas pressures of
1.5 psig in order to burn crude vapors containing up to 15 vol. % propane
smokelessly Air fans, actuated by gas flow switches can provide
supplemental primary air to assure smokeless burning at these low gas pressures
if such is found to be necessary. An enclosed 3 inch water seal and flame
120
-------
arrester prevent back-flashes. Consequently, flare areas and burner
heads are large. Real estate requirements vary from 120 ft. by 100 ft.
to 150 ft. by 400 ft. Banks of burners are charged with gas by stages.
The number of pilot burners vary in number from four (4) to eight (8).
Each pilot burns about 200 SCFH of natural gas continuously from a
separate commercial natural gas source for safety reasons. Automatic
ignition is accomplished from a local panel just outside the fenced-in
burner areas. Flame failures automatically shut down feed sources to
these flares or incinerators, and sound alarms accordingly.
9.2.9 Inert Gas Generators
Inert gas generators are used in standby service for Case VII
only. These natural gas near-stoichiometric burners have been arbitrarily
sized for reducing an empty storage tank from 21 vol. % to 4 vol. % oxygen
within 48 hours while producing 1.0 vol. % oxygen. Scheduled tankage
turnarounds would be by tank flooding procedures, however, as described
under operating features, paragraph 9.1.2. Unit sizes produce 100,000
SCFH of a flue gas at about 10" H~0 with the following average compositions,
after being lined out:
0.5 vol. % oxygen
91.2 vol. % nitrogen
7.5 vol. % carbon dioxide
0.83 vol. % water vapor
neg. carbon monoxide
Water removal is accomplished physically by heat exchange with sea-water
cooling to a dew point of about 80 F. Carbon dioxide and/or further water
removals are considered too expensive for this standby facility, especially
121
-------
since tanker stack flue gas, the primary source of this blanket gas, is
not treated.
Inert gas generators are potentially dangerous in that more CO can
easily be generated than is shown above in attempting to minimize oxygen.
They are usually difficult to line out, and if their delicate flame
is extinguished, the generator produces a very explosive mixture of gases
into the blanketing system. For these reasons, Case VII preferentially
obtains its flue gas from ships' stacks at berth, assuming that the ships'
boilers operate at no more than 20% excess air ft reduced boiler loads.
Even so, this source is also ladened with CO, espe ially in Facility No. 2
where diesel driven tankers are serviced exclusively.
9.2.10 Treating Units
These units are needed to assure that excess blanket natural gas
volumes meet pipeline quality requirements n Cases IV and V. They remove
sulfur compounds, water, and vaporized hydrocarbons from blanket gas to be
either recycled or returned to sales. Because of the investment costs
for compressing large volumes of gas for sales at 350 psig, treating
units have been reduced in design throughput by the difference in load
factors between blanket gas demands and excesses, and recycled blanket
gas is passed through power recovery turbines back to low pressure storages.
Large feed gas holders allow these reduced treating unit rates to be fairly
constant and continuous. A flare is available for unit interruptions
or for disposals with filled blanket gas holders and no sales outlets.
Treating units consist first of a guard chamber of sponge iron to
remove stray sulfur compounds at 25 psig and fairly ambient temperatures.
122
-------
Any condensed hydrocarbons are also passed through the sponge bed for
sulfur removal. Water removal is then accomplished at a pipeline header
pressure of 350 psig. In cases having cone roof tankage, where significant
amounts of hydrocarbons are present, 10 F dew point temperature is reached
by indirect refrigeration. Defrost cycles gravitate water with any hydro-
carbons to a closed API separator in a closed drain system. In cases
where 99% of the vaporizations from cone roof tankage is restrained by
using internal floating roof tankage, water removal is obtained by glycol
absorption. Refer to drawings 153-1-21A and 21B. Atmospheric emissions
from the very small fired heater and regenerator vent are assessed to
Case IVA and VA, where these units are used. Transient emissions also
occur periodically from the opening of sponge iron chambers, the chang-
ing of their contents, and the cleaning of glycol filters. Refer to
paragraph 5.6, pages 58 and 60, for further details.
In some locations it may be feasible to return sour gas to the gas
utility for their processing, along with their other gas receipts. The
cost to clean up this gas has been estimated for Cases IV and V because
+
that cost would supposedly be reflected in the difference between sour
gas delivered and sweet gas received in any case. Local negotiations
could, perhaps, reduce the treating costs estimated.
A lead factor of 90% has been calculated for Case IV, but Case V
theoretically has no load factor, since blanket gases leave the terminals
in tankers. The same capital costs have been estimated for both cases,
but a 10% load factor has been assigned to Case V treating units.
123
-------
£
IU
(0
§
Ul
0
(0
OH
-L
u.
1
ES
'\
h "s "§
§
U UJ
5 m
3i
i-
fclU
111 Of
I--"*-
s
^-UJ
OS?
Ftf)
Js
ROBERT BROWN ASSOQATES
CARSON, CALIFORNIA
CUSTOMER ENVIRONMENTAL PROTECTION AGENCY
PLANT CONTRACT #68-02-2838
LOCATION TREATER UNIT WITH CR TANKAGE
DRAWING NUMBER
REV.
I53-I-2IA
O
KLINGLER VELLUM 1021-8
124
-------
6
UJ
cQ
0
J*
UJ
V?
o
Q.
0)
\J
$
£
t-
t/>
?
£
Ul
2
0}
o:
D
h-
O.
X
ill
/-I
_/ \
f?
b
N
V
I
PS
U.I
3U
J,
il?
ADDED eXR TURBINES
TS
»
I
-~J
.VISION.
ROBERT BROWN ASSOCIATES
CARSON.CALIFORNIA
CUSTOMER Environmental Protection Agency
PUNT
Contract #68-02-2838
LOCATION Treating Unit with IFR Tankage
DRAWING NUMBER
REV.
153-I-2IB
KLINCLER VELLUM IOZI-»
125
-------
9.3 PIPING ARRANGEMENTS
Vapor collection and blanket gas distribution piping has been sepa-
rately estimated and modulated for each vapor control system considered
in this study. Low operating pressures and high transfer rates have
created large, thin walled ducting up to 80" in diameter. Estimates
have been based upon 1/4" steel wall thicknesses for this low-pressure
ducting in sizes 6" and larger except in Case VII. Here 3/16" thick-
nesses have been estimated for additional flue gas errosion allowances,
as discussed under Case VII paragraph 5.9. Mitred elbows and plate flanges
are used in sizes 14" and larger. These lines are routed in pipeway
locations so that internal detonations will cause the least amount of
damage. Where longitudinal weld seams are used, a continuous weak (rip)
seam can be directed upwards, or tilted to the least vulnerable direction.
All vapor collection lines are sloped down to closed drain boots where
condensations can be manually drained upon high level alarm signals in a
centralized control room. Drains flow by gravity through a closed pipe
line to the API separator without intermediate emissions, or by pumps
from dock and beachhead locations.
Refer to drawings 153-1-13, -14, -15, -16, -19, -20, and -23 for
typical piping arrangements. The only piping that needs insulation is
that in which condensates from refrigerators are pumped. Here it is
important that the flashing of liquid does not occur before the cold
condensates are dispursed into main terminal transfer headers. All piping,
other than waste water, has been estimated as being above ground.
Water seal pots and drain boots are totally enclosed vessels which
automatically or manually drain through closed piping to the API separator in
126
-------
all except Cases I and IA. Three-inch seals are maintained by con-
ventional level controls and continuous water circulation in each
vessel.
127
-------
10.0 REFERENCES
1. U.S. Environmental Protection Agency. Supplement No. 7 for
Compilation of Air Pollutant Emission Factors, 2nd Ed., Research
Triangle Park, N.C., April 1977, Section 9.0.
2. Sohio/CB&I Floating Roof Emission Test Program, Final Report
November 18, 1976.
3. Discussions with Lt. Commander Cox, Marine Section, Engineering
Branch, U.S. Coast Guard G-MMT-2/82, 400 7th St. S.W., Washington, D.C.
4. Wilson, R. P. Jr. and Phani Raj, P.K. Vent System and Loading
Criteria for Avoiding Tank Overpressurization. U.S. Coast Guard
Office of Research and Development, Washington, D.C. Report No.
CG-D-59-77. September 1977. 62 p.
5. Discussions with Dr. John Erbar, Professor of Chemical Engineering,
Oklahoma State University, Stillwater, OK. January 1978.
6. Gammell, D. M. Inerting with Natural Gas Has Advantages. Oil and
Gas Jr. 75 (No. 6): 76-77 February 7, 1977.
7. American Gas Association. Purging Principles and Practice AGA
Catalog No. XK0775, 1975, p. 35 & 36.
8. Discussions with Anthony L. Rowek, Consultant, U.S. Coast Guard,
G-MHM-1/83, 400 7th St. S.W., Washington, D.C.
9. Discussions with George L. Steel III, Mgr. New Construction,
West Coast Shipping Co., 1052 W. 6th St., Los Angeles, CA.
10. Discussions with W. Lang&r, Technical Services Engineer, Carboline,
Inc., 350 Hanley Industrial Court, St. Louis, Mo. December 1977.
11. Guthrie, K.M. Process Plant Estimating Evaluation and Control.
Solana Beach, California, Craftsman Book Co., 1974. 600 p.
12. Discussions with Kenneth Cuccinelli, Associate Director of Energy
Systems. American Cas Association. 1515 Wilson Blvd., Arlington,
Virginia, January 1978.
13. Discussions with First Assistant Chief Engineer Thomas Lloyd aboard
70M dwt tanker "Chevron California" San Pedro, CA. November 1977.
>.
14. Paktank-Pacific Corp. application to EPA District IX for con-
struction permit for crude and product storage terminal.
15. Western Oil & Gas Association. Hydrocarbon Emissions from
Floating Roof Storage Tanks. Report dated January 1977.
128
-------
16. Gammell, D.M. Blanket Tanks for Gas Control. Hydrocarbon Process-
ing 12: 101-104, 1976.
17. National Fire Protection Assoc. Foam Extinguishing Systems 1976.
NFPA No. 11, Boston, Mass. 113 p.
18. Cubbage, P.A. Flame Traps for Use with Town Gas/Air Mixtures. Gas
Council Research Committee Report GC 63/1959, Midlands Research
Station, Grosvenor Place, London, S.W.I.
19. Wilson, R.P. Jr. and Attalah, S. Design Criteria for Flame
Control Devices for Cargo Venting Systems. U.S. Coast Guard
Office of Research and Development, Washington, D.C. Report
No. CG-D-157-75. August 1975. 43 p.
20. American Petroleum Institute. Flame Arresters for Tank Vents.
PSD 2210 1st Ed. 1801 K Street, N.W. Washington, D.C. May 1971.
21. Ghormley, E.L. Guard Against Detonation Hazards. Petroleum
Refiner. y}_ (No. 1): 185-190 January 1958.
22. U.S. Environmental Protection Agency. Evaluation of Methods for
Measuring and Controlling Hydrocarbon Emissions from Petroleum
Storage Tanks. EPA-450/3-76-036, Research Triangle Park, N.C.
November 1976 p. 6-28.
23. Discussions with Ken Sassen, Mgr. Vapor Recovery Systems, Trico-
Superior Tank and Construction Co. 6155 S. Easton Ave. Los Angeles,
November 1977.
24. Discussions with John Fehrenbacher, Buffalo Forge Co., 2500 W.
Sixth St., Los Angeles, CA. December 1977.
25. Discussions with Ray Edwards, Edwards Engineering Corp. 101 Alexander
Avenue, Pompton Plains, N.J., December 1977.
26. Discussions with M. Keller and R. Noble, Project Engineers, John
Zink Co., 4401 Peoria Street, Tulsa, OK. February 1978.
129
-------
11.0 ABBREVIATIONS AND CONVERSIONS
Abbreviations used in this report are defined below in English
units:
API = American Petroleum Institute (degree of density)
B/CD= barrels (42 gallon) per calendar day
BPH = barrels (42 gallon) per hour
Btu = British thermal unit
bbls= barrels (42 gallon)
DWT = dead weight tons
F.0.= fuel oil
fps = feet per second
ft. = feet
GPM = gallons per minute
KWH = kilowatt-hour
M = thousand
MM = million
osi = ounces per square inch
Part= particulates
PCV = pressure control valve
PSV = pressure safety valve
ppm = parts per million
psf = pounds per square foot
psia= pounds per square inch absolute
psig= pounds per square inch gage
RVP = Reid vapor pressure
SCFH= standard cubic feed per hour
ST/Y= short tons per year
Ounces, pounds, and short tons are Avoirdupois weights.
130
-------
Applicable English to Metric conversion factors are:
barrel = 0.159 cubic meters
Btu = 0.252 kilogram - calories
short tons = 0.907 metric tons
feet = 0.305 meters
miles = 1.609 kilometers
gallons - 3.785 liters
ounces per square inch = 4.394 grams per square centimeter
pounds = 453.6 grams
pounds per square foot = 488.3 grams per square meter
pounds per square inch = 70.30 grams per square centimeter
standard cubic feet (at 60°F) = .0268 standard cubic meters (at 0°C)
cubic feet = .0283 cubic meters
131
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
2.
3, RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Emission Control Technology for Marine Terminals
Handling Crude Oil and Gasoline
5. REPORT DATE
April. 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
Don M. Gammell
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Robert Brown Associates
500 East Carson Plaza Drive
Carson, California 90745
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2838
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report presents results of a study which developed basic background
information on emission control systems for a hypothetical deep water marine
terminal handling crude oil and an inland marine terminal handling crude oil
and gasoline. The study includes comparative cost analysis for alternative
emission control systems together with comparable safety and reliability analysis
for both marine terminal modules.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air Pollution
Control Methods
Hydrocarbons
Tankers and Barges
Storage Tanks
Air Pollution Control
Hydrocarbon Emission Cont
Organic Vapors
Mobile Sources
rol
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |