TO:	EPA Docket No. EPA-HQ-OAR-2003-0146

		

FROM:	Bob Lucas, EPA/SPPD 

DATE:	August 3, 2007

SUBJECT:	Storage Vessels:  Control Options and Impact Estimates 

I.	Purpose

This memorandum documents the methodology used to identify the control
options and estimate the costs and environmental impacts for storage
vessel control options. 

II.	Background

Section 112(f) of the Clean Air Act Amendments (CAA) directs the U.S.
Environmental Protection Agency (EPA) to assess source categories
regulated under Section 112(d) of the CAA and determine whether any
human health or environmental risks remain from the continued emissions
of hazardous air pollutants (HAPs) following implementation of maximum
achievable control technology (MACT) standards.  The CAA further states
that if the MACT standards do not reduce lifetime excess cancer risk to
the most exposed individual to less than one in one million, EPA must
set additional standards to protect human health and the environment, in
accordance with the interpretation set forth in the Benzene NESHAP. 
Additionally, the EPA is required to review these technology-based
standards and to revise them “as necessary (taking into account
developments in practices, processes and control technologies)” no
less frequently than every 8 years, under CAA section 112(d)(6).  The
Refinery MACT 1 (40 CFR Part 63 Subpart CC) was promulgated over 8 years
ago and is now being reviewed.   

III.	Summary of Existing Refinery MACT 1 Storage Vessel Control 
Requirements

Refinery MACT 1 requires storage vessels above a certain size and vapor
pressure to meet the control requirements specified in part 63 subpart G
(commonly referred to as the HON), except that they are specifically
excluded from the fitting requirements in §63.119(b)(5), (b)(6), (c)(2)
and (d)(2).  Refinery MACT 1 defines Group 1 storage vessels as: 

storage vessels at an existing source that have a design capacity
greater than or equal to 177 cubic meters and stored-liquid maximum true
vapor pressure greater than or equal to 10.4 kilopascals and
stored-liquid annual average true vapor pressure greater than or equal
to 8.3 kilopascals and annual average HAP liquid concentration greater
than 4 percent by weight total organic HAP; storage vessels at a new
source that have a design storage capacity greater than or equal to 151
cubic meters and stored-liquid maximum true vapor pressure greater than
or equal to 3.4 kilopascals and annual average HAP liquid concentration
greater than 2 percent by weight total organic HAP; or storage vessels
at a new source that have a design storage capacity greater than or
equal to 76 cubic meters and less than 151 cubic meters and
stored-liquid maximum true vapor pressure greater than or equal to 77
kilopascals and annual average HAP liquid concentration greater than 2
percent by weight total organic HAP.

Group 1 storage vessels must either use internal or external floating
roofs or must vent to a control device that reduces organic HAP by 95
percent or reduces organic HAP to an outlet concentration of 20 ppmv.

IV.	Identification of Control Options

EPA’s TANKS software program was used to identify storage vessel
control configurations that had the greatest HAP emissions.  External
floating roofs have significantly higher emissions than internal
floating roofs.  Additionally, a significant fraction of the emissions
from external floating roofs come from slotted guide poles, based on
preliminary TANKS modeling runs.  Based on TANKS modeling and literature
reviews, two storage vessel control options were identified.

  

Option 1:  install a gasketed sliding cover or a flexible fabric sleeve
and install a gasketed float or other device which closes off the liquid
surface from the atmosphere, i.e., the fitting controls required in the
HON in §63.119(c)(2)(ix) and (x) for slotted guide poles used on
external floating roofs.

Option 2:  install a geodesic dome or otherwise retrofit an existing
external floating roof tank to be an internal floating roof tank.

V.	Impact Estimates for Model Storage Vessels

Model Storage Vessels 

Three different model storage vessels were developed to simulate the
different types of material stored at refineries that would trigger the
Refinery MACT 1 Group 1 control requirements.   These were:

Crude oil storage tanks

Average throughput:  30,000 bbl/day

Diameter:  180 ft

Height:  40 ft

Calculated Turnovers:  190 per year

Gasoline storage tanks

Average throughput:  20,000 bbl/day

Diameter:  150 ft

Height:  40 ft

Calculated Turnovers:  182 per year

Other light distillate storage tanks (e.g., for wide cut jet fuels,
light naphtha, and light distillate intermediates)

Average throughput:  10,000 bbl/day

Diameter:  105 ft

Height:  40 ft

Calculated Turnovers:  186 per year

These model storage vessel parameters were developed using engineering
judgment after reviewing Louisiana State permit data and other sources. 
These data are provided in the petroleum refineries source
characterization report for refinery emission sources (RTI, 2002).

Model Unit Emission Reductions 

Emission reductions were estimated using TANKS version 4.09d.  The crude
oil tanks were modeled using the single preset crude oil properties in
TANKS, which has a Reid vapor pressure of 5 pounds per square inch
(psi).  The gasoline tanks were modeled using the preset properties of
gasoline with a Reid vapor pressure of 11 psi, and the other light
distillates were modeled using the preset properties of jet naphtha. 
The emissions from external floating roof tanks were modeled using the
preset meteorological data for Corpus Christi, Texas.  All tanks were
assumed to have pontoon-type, welded external floating roofs with a
primary liquid-mounted seal and no secondary seal.  The tanks were
assumed to be painted white and in good condition and the internal
shells were assumed to have light rust (default settings).  At baseline,
these tanks were assumed to have slotted guide poles with ungasketed
sliding covers and no floats.

For Option 1, gasketed sliding covers and a slotted guide pole sleeves
were assumed to be installed to control emissions from the slotted guide
pole.  The VOC emissions were estimated using TANKS 4.09d software for
the model tanks at baseline and with Option 1 controls; these results
are summarized in Table 1.  The HAP emission reductions were estimated
from the overall VOC emission reductions based on average refinery
process stream compositions developed in the original Refinery MACT 1. 
The vapor-phase organic HAP content for crude oil, gasoline, and naphtha
process streams are 8.9 wt%, 10.8 wt%, and 13.5 wt%, respectively.  The
HAP emission reductions for Option 1 for the three model tanks are also
presented in Table 1. 

Table 1.  Summary of External Floating Roof Tank (EFRT) VOC Emissions
– Option 1

Product	Tank Diameter (ft)	Tank Turnovers	Tank Throughput (MMgal/yr)
Total Emissions without guide pole controls

(lb VOC/yr)	Emissions with guide pole controls

(lb VOC/yr)	VOC Reductions 

(lb /yr)	Organic HAP Reductions

(lb/yr)

Crude oil (RVP 5.0)	180	190	459.9	14,500	8,000	6,500	580

Gasoline (RVP 8.3)	150	182	306.6	60,600	26,400	34,200	3,700

Jet Naphtha	105	186	153.5	17,700	6,700	11,000	1,480



For Option 2, each model external floating roof tank (with ungasketed
sliding cover without a float) was retrofitted with a domed external
floating roof.  Again, TANKS 4.09d software was used to estimate the VOC
emissions for the model tanks at baseline and with Option 2 controls. 
The HAP emission reductions were estimated from the VOC emission
reductions as in Option 1.  Table 2 presents a summary of the VOC and
HAP reductions under Option 2.  

Table 2.  Summary of Baseline and Domed External Floating Roof Tank
(DEFRT) VOC Emissions – Option 2

Product	Tank Diameter (ft)	Tank Turnovers	Tank Throughput (gal/yr)
Baseline VOC Emissions -without guide pole controls

(lb/yr)	VOC Emissions - with geodesic dome

(lb/yr)	VOC Reductions 

(lb /yr)	Organic HAP Reductions

(lb/yr)

Crude oil (RVP 5.0)	180	190	459.9	14,500	3,100	11,400	1,000

Gasoline (RVP 8.3)	150	182	306.6	60,600	3,700	56,000	6,200

Jet Naphtha	105	186	153.5	17,700	1,200	16,500	2,200



Model Unit Control Costs 

Control costs for Option 1 were taken from a cost analysis performed by
the Bay Area Air Quality Management District (BAAQMD) in 1999 (BAAQMD,
1999).  As these costs estimates were originally developed in 1999, the
costs were escalated to 2006 dollars using the Chemical Engineering
Plant Cost Index (CEPCI).  The average CEPCI in 1999 was 390.6; the
average CEPCI in 2006 was 499.6.  Therefore, the cost of the slotted
guide pole controls was estimated to be $2,600 ($2,000×499.6/390.6,
rounded to two significant figures).  These costs were considered
capital investments and were annualized over 10 years assuming a 7
percent interest rate.  There are no operating costs associated with the
slotted guide pole fitting controls; inspections were assumed to be
performed annually.  The inspections were assumed to take 5 technical
labor hours and cost $630/yr.  Table 3 provides the basis of the loaded
labor rates used for this analysis.  The unit control costs for Option 1
were considered to be average costs per tank and would not vary
significantly given the model tank sizes used in this analysis;
therefore, these costs were applied to all model tanks.

Table 3.  Summary of 2006 Labor Rates.

Labor Class	(A)

Labor Rate from the Bureau of Labor Statistics	(B)

Fringe Benefit Loading rate from BLS	(C)

Overhead and profit rate from MWC ICR	(D)

Loaded wage rate 

(A x B x C)

1. Professional specialty and Technical	48.27	1.43	1.67	115.27

2. Executive, administrative and managerial	59.15	1.40	1.67	138.29

3. Administrative support, including clerical	17.46	1.41	1.67	41.11

4. Installation, Maintenance, and Repair	24.28	1.40	1.67	56.77

5. Plant and System Operators, All	20.64	1.40	1.67	48.26

May 2006 Labor rates (accessed in Jun 2007).

http://stats.bls.gov/home.htm

http://stats.bls.gov/oes/current/oessrci.htm	total cost/tech hr*	126.30

	*Calculated as: D1+ 0.05D2 + 0.1D3

Where D1, D2, and D3 are cell references above.



Control costs for Option 2 were taken from a cost analysis conducted by
the South Coast Qir Quality Management District (SCAQMD) in 2001
(SCAQMD, 2001).  An estimate of $350,000 was made to construct a domed
external floating roof tank (DEFRT) from an external floating roof tank.
 These cost estimates were originally developed in 1999, so the costs
were escalated to 2006 dollars using the average CEPCI in 1999 of 390.6
and the average CEPCI in 2006 of 499.6.  Therefore, the cost of
constructing a DEFRT was estimated to be $450,000
($350,000×499.6/390.6, rounded to two significant figures).  These
costs were considered capital investments and were annualized over 20
years assuming a 7 percent interest rate.

These controls prevent the loss of products from the storage vessels. 
Therefore, the product not lost as a result can be sold; therefore, the
control costs are offset, to some extent, by the increased product
sales.  This VOC credit was calculated assuming the value of VOC to be
$1.50/gallon, based on average crude and gasoline spot prices in 2006
(based on September 8, 2006 OGJ crack spread spot prices of $65.34/bbl
($1.55/gal) for brent crude and $72.09/bbl ($1.72/gal) for product value
per barrel; reported in Oil and Gas Journal, September, 18, 2006, p.
70).  Assuming an average refinery process stream specific gravity of
0.8, the VOC credit is $480 per ton VOC reduced.

VI.	Nationwide Impact Estimates for Storage Vessels

Nationwide impacts were developed using a refinery-specific analysis. 
Data on process unit capacities for individual refineries were taken
from the Energy Information Agency’s (EIA’s) Refinery Capacity
Report 2006 (EIA, 2006b); the crude distillation capacity as directly
reported in barrels per calendar day (bbl/cd) was used for crude
throughput rates.  Production rates of gasoline and other light
distillates were estimated using a combination of the refinery-specific
production capacities from the Refinery Capacity Report 2006 (EIA,
2006b) and the net production quantities reported by Petroleum
Administration for Defense District (PADD) subregions (EIA, 2006a). 
Note that, except for the crude distillation capacity, all other process
unit capacities are reported only in terms of barrels per “stream
day” (bbl/sd).  A stream day capacity is the process rate capacity,
assuming continuous 24-hour operation for an operating day.  Due to
required maintenance and other events, most refinery units do not
operate 365 days each year. The subsequent downtimes limit the actual
annual production capacity of the refinery processes when evaluated on a
basis of 365 “calendar days.” 

The following methodology was used to estimate refinery-specific
distillate production:

Initial Distillate Estimates.  Preliminary refinery-specific distillate
fuel estimates were calculated by subtracting the heavy product
production capacities (e.g., lubes, asphalt, and petroleum coke) and the
aromatics production capacity (assumed to by tanks subject to the HON)
from the overall refinery crude capacity.  This value was then
multiplied by 0.88 to account for quantities of refinery fuel gas,
liquefied petroleum gases, and residual fuel oil, which were estimated
to be 12 percent of this pool.  This provided an estimate of the light
and middle distillates produced at the refinery in terms of bbl per
stream day.  There are a few refineries that only process intermediate
products; their distillate fuel estimate was calculated as 60 percent of
their catalytic cracking capacity plus their alkylate production
capacity.  

Stream Day to Calendar Day Conversion.  For refineries with crude
distillation units, the stream day capacities were converted to calendar
day capacities using the ratio of the two reported crude distillation
capacities (bbls/cd per bbl/sd) in the Refinery Capacity Report 2006
(EIA, 2006b).  There were some heavy crude or intermediate processing
refineries that do not have crude distillation units.  For these
refineries, a factor of 0.9 (bbls/cd per bbl/sd) was used to adjust the
distillate productions to calendar day productions. 

Sub-PADD Distillate Yields.  The product yields reported by EIA (EIA,
2006a, Table 21) were used to proportion the light and middle
distillate estimates between gasoline, light distillates, and middle
distillates.  Finished gasoline yields were reported directly.  Light
distillates were assumed to include finished aviation gasoline, half of
the kerosene yield (jet kerosene and other kerosene), and naphthas (for
petrochemical feed use and special naphthas).  Middle distillates were
assumed to include distillate fuel oil and half of the kerosene yield
(jet kerosene and other kerosene).  

Refinery-specific, Distillate-specific Production Estimates.  As the
production estimates in Step 2 only include light and middle
distillates, the sum of the gasoline, light distillate, and middle
distillate yields calculated in Step 3 were used to escalate the
individual yield estimates to calculate the relative proportion of the
total distillate production that was gasoline versus light distillates
versus middle distillates.  For example, the gasoline yield divided by
the sum of the distillate yields times the refinery-specific distillate
production generates a refinery-specific estimate of estimate of
gasoline production.  In this same manner, refinery-specific light and
middle distillate production estimates were developed.

These production estimates were then used to estimate the number of
model storage vessels that would be present at each refinery.  As
discussed previously, the model storage vessels were assumed to manage a
certain throughput.  These throughputs were used as follows to determine
the number of tanks at the refinery.

		Number of Tanksi = Xi + ROUNDUP(Quantityi / PerTankThroughputi , 0)

Where,

i	= index for product type

Xi 	= excess tank quantity; Xi =1 for crude, gasoline, and middle
distillate; Xi =2 for other light distillates and intermediates

ROUNDUP(number, 0) = Excel® function to round up to the next whole
integer.

Quantityi = refinery throughput estimate for product type i, bbl/cd 

PerTankThroughputi = model tank throughput for product type i,
bbl/cd/tank

This analysis resulted in the estimated total number of crude, gasoline,
other light distillate and middle distillate storage vessels per
refinery.  Refineries that specialized in asphalt products were assumed
to process crude oil that could be managed in Group 2 storage vessels;
consequently, the number of external floating roof crude tanks was set
to zero.  All middle distillate tanks were also assumed to be Group 2
storage vessels (based on a review of MSDS data for these products);
therefore, none of the middle distillate storage tanks were assumed to
have external floating roofs.  

Once the crude storage vessels at asphalt refineries and the middle
distillate storage vessels were removed from the analysis, the
proportion of the remaining tanks that would have external floating
roofs and the proportion of these tanks that would have slotted guide
poles were estimated.  According to the BAAQMD report, 60 percent of
floating roof tanks had external floating roofs and two-thirds of the
external floating roofs had slotted guide poles.  Assuming the
proportions are accurate for all refineries in the United States, the
number of external floating roofs and the number of external floating
roofs with slotted guide poles were estimated.  Again, the ROUNDUP
function was used so that the number of tanks subject to the control
options was always an integer. 

Using the approach described above, the number of external floating
roofs with slotted guide poles were estimated for each individual
refinery, and facility-specific impacts were estimated using the unit
costs and emission reductions described in Section V of this memorandum.
 Based on this analysis, there were 373 crude oil tanks, 386 gasoline
tanks, and 304 light naphtha and intermediate tanks nationwide that have
external floating roofs with slotted guide poles.   These tank totals
were used, along with the unit costs and emission reductions described
in Section V of this memorandum, to develop the nationwide impacts that
are summarized in Table 4.



Table 4.  Summary of Nationwide Impacts of Refinery Storage Vessel
Control Options

Control Option	Total Capital Investment  (million $)	Annual Operating
Costs  (million $/yr)	Total Annualized Costs  (million $/yr)	HAP
Emission Reduction 

(tpy)	Cost-Effective-ness

($/ton HAP)

1.  Guide pole sleeves	2.76	(3.88)	(3.49)	1,046	(3,340)

2.  Geodesic domes	478	(7.50)	37.7	1,715	22,000



VI.	References

BAAQMD.  1999.  Staff Report:  Amendments to Regulation 8, Rule 5,
“Storage of Organic Liquids”.  November 2.

EIA. 2006a. Petroleum Supply Annual 2005.  Prepared by the Energy
Information Administration, Washington, DC.  October 23.

EIA. 2006b. Refinery Capacity Report 2006. Prepared by the Energy
Information Administration, Washington, DC. June 15.

RTI.  2002.  Petroleum Refinery Source Characterization and Emission
Model for Residual Risk Assessment.  Prepared for U.S. Environmental
Protection Agency, Office of Air Quality Planning and Standards,
Research Triangle Park, NC.  EPA Contract No. 68-D6-0014.  July 2,
2002.

SCAQMD.  2001.  Final Socioeconomic Report For Proposed Rule
1178—Further Reduction of VOC Emissions from Storage Tanks at
Petroleum Facilities.  November.

TANKS 4.09d Software.    HYPERLINK
"http://www.epa.gov/ttn/chief/software/tanks/index.html" 
www.epa.gov/ttn/chief/software/tanks/index.html .

Technical Memorandum – Storage Vessels:  Control Options and Impact
Estimates

August 3, 2007

Page   PAGE  8 

Technical Memorandum

