-------
Table 7-35. RETURN ON INVESTMENT FOR PETROLEUM AND COAL PRODUCTS INDUSTRY
32
Number
Item
1
2
3
4
5
6
7
8
9
10
Current assets
Other assets
Net depreciable assets
Accumulated depreciation
Gross depreciable assets
(no. 3 + no. 4)
Mineral rights, etc.
Total assets
(sum of nos. 1,2,5,6)
Net profit after tax
Depreciation
Cash flow
(no. 8 + no. 9)
$40,625.50a
44,237.50a
59,701.00a
58,296.00a
117,997.00a
21,559.00a
224,419. 00a
12,767.00a
7,698.00a
20,465.00a
Capital recovery
coefficient
(no. 10 * no. 7)
Depreciation rate
(no. 9 * no. 5)
Asset life
(no. 5 * no. 9)xl.25
Return on investment
Ratio of net income to
revenue
0.09119
0.0652
19 years
6.2 percent
.071
Dollar figures are in millions of 1978 dollars, stocks are averages for the
year, flows are totals for the year. '.
The return on investment is the value for r which satisfies the following
formula:
Capital recovery coefficient = Life of asset
The formula was solved for the value 0.062 percent - r given the asset
life of 19 years and the capital recovery coefficient of 0.09119.
7-64
-------
method of analysis. The baseline parameters in Table 7-35 are assumed to
be representative of the current economic conditions facing benzene
producers. Table 7-36 shows a similar computation for the industrial
chemicals and synthetics industry, which is assumed to be representative
of benzene consumers. The procedure computes the net cash inflow to the
industry as a result of its operations, and divides this figure by an
appropriate measure of the stock of assets tied up in the industry. The
quotient of cash flow over assets is called the capital recovery coeffi-
cient. Given the capital recovery coefficient and the average lifetime
of the assets making up the capital stock, the return on investment can
be computed by the discounted cash flow method.
The following paragraphs of this section describe in some detail the
general procedure outlined above, and step through the application of the
procedure to the petroleum and coal products industry.
The cash flow into the petroleum and coal products industry is
computed from published income statement data by adding depreciation to
after tax profit. This measure of money inflow is more appropriate than
using net accounted profit because the depreciation flow is not actually
a cash outflow, but merely an accounting convention. Referring to
Table 7-35, the cash flow is $20,465 million ($12,767 million +
$7,698 million).
The amount of resources tied up in the industry must be evaluated at
original cost, not after subtracting the accumulated depreciation. This
is done to facilitate the comparison of the initial investment to a
typical year of cash inflow—the depreciated value of assets has no
meaning in discounted cash flow analysis. Referring to Table 7-35, the
total assets figure is $224,419 million.
Now the average lifetime of the assets, based on the depreciation
reported, is computed. The depreciation rate is the annual percentage of
decay of assets. The best available way to estimate this rate is to
divide the annual flow of accumulated depreciation by the gross depreciable
assets. In the petroleum and coal products industry, the result is
$7,698 million divided by $117,997 million, or about 6.52 percent. If
6.52 percent of an asset decays every year, then, assuming straight line
7-65
-------
Table 7-36. RETURN ON INVESTMENT FOR INDUSTRIAL CHEMICALS AND SYNTHETICS33
Number
1
2
3
4
5 '
Item
Current assets
Other assets
Net depreciable assets
Accumulated depreciation
Gross depreciable assets
$20,567.00a
8,830.25a
24,934.00a
25,210.25a
50,144.25a
(no. 3 + no. 4)
Mineral rights, etc.
Total assets
(sum of nos. 1,2,5,6)
Capital recovery
coefficient
(no. 10 * no. 7)
Depreciation rate
(no. 9* no. 5)
Asset life
(no. 5 * no. 9)xl.25
Return on investment
Ratio of net income to
revenue
1,124.75£
80,666.25C
8
9
10
Net profit after tax
. Depreciation
Cash flow
(no. 8 + no. 9)
4,036.00°
3,253.00a
7,289.00a
0.09036
0.0648
19 years
6.2 percent
0.068
Dollar figures are in millions of 1978 dollars, stocks are averages for .the
year, flows are totals for the year.
The return on investment is the value for r which satisfies the following
f ormula:
r
Capital recovery coefficient =
l--(Hr)"L1fe of asset
The formula was solved for the value 0.062 percent = r given the asset Itfe
of 19 years and the capital recovery coefficient of 0.09036.
7-66
-------
depreciation, the asset will be fully decayed after 1 divided by
0.0652 years, or 15.33 years. Thus, the average asset life is estimated
to be 15.33 years. However, the problem with this estimate is that
accounting depreciation will usually overestimate the true depreciation
rate of an asset because Federal tax laws allow corporations to use
accelerated depreciation to lower their taxes. The usual rule of thumb
in adjusting asset life to account for accelerated depreciation is to
inflate the asset life implied by the accounting depreciation rate by
25 percent. Thus, the asset life is estimated to be 19 years (15.34 x 1.25).
The capital recovery coefficient is the quotient of the cash flow
divided by the capital asset stock. (When the coefficient is calculated
using assumptions about asset lifetime and rate of return on investment
instead of cash flow and capital asset stock, it is called a capital
recovery factor.) The return on investment is the rate of interest one
would have to charge to equate the discounted values of the incoming
flows of cash to the original investment. In the present case, the
initial investment of $224,419 million earns money at the rate of
$20,465 million per year for a period of 19 years. Thus, given an asset
life of 19 years, it is necessary to find the value of r that satisfies
the following equation:
- 20,465 + 20,465 + 20,465 +
-
+ 20,465
(i+r)19
It can be shown that this formula is exactly equivalent to:
20,465 r
224,419 l-(l+r)-19
which can be solved to get r = 0.062, representing a 6.2 percent return
on investment.
Table 7-36 shows the results of a similar analysis on the industrial
chemicals and synthetics industry, with the resulting value of the return
on investment substantially the same as the 6.2 percent ratio calculated
for the petroleum and coal products industry. The asset life and deprecia-
tion rates are almost identical, as are the ratios of net profit to
revenue. In the following analysis, ROI is rounded to 10 percent and the
asset life to 20 years for both producers and consumers. This simplifying
7-67
-------
adjustment assures that estimates of the potential negative economic
impacts of the control options will be conservative (i.e., slightly
exaggerated.
7.4.3 Example Calculation of Economic Impacts
The throughput and cost characteristics of the model plants developed
in earlier sections of this report are summarized for ease of reference ---
in Table 7-37. These numbers form the basis for the computation of two
fundamental measures of impact on the model plants. The first measure is
the change in the rate of return on the overall plant and equipment as a
result of the expenditures stemming from compliance with an option, under
the assumption that price and quantity sold remain unchanged. This
measurement is intended to be instructive of the potential loss in earnings
the plant would suffer if it were totally unable to pass on any of the
increased costs. The opposite polar case, that of complete cost passthrough,
forms the basis of the second measure of impact to be computed here. The
second measure is the new price that the model plants would have to
charge in order to maintain a 10 percent return on investment, assuming
no change in quantity sold. In order to illustrate the method used in
computing these impact measures, this section shows two sample calculations,
one for a large benzene producer faced with Control Option III, and one
for a benzene consumer faced with Control Option III. Throughout this
section, the final results of the computations are rounded in order to
improve readability and not create an exaggerated sense of precision in
the results. Exceptions to this rule occur when greater precision is
required to show small differences between two numbers. Intermediate
calculations are carried out before rounding.
7.4.3.1 Baseline Characteristics of Large Benzene Producer in Example
Calculation. In order to evaluate the option-induced changes in asset
stocks and cash flows for the model plant, it is first necessary to
compute baseline stocks and flows. Given the throughput of the plant
from Table 7-37 and a price of benzene of $0.34 per liter, revenues are
computed for the model plant as $76.36 million ($0.34/liter x 224.6 million
liters). It is likely that the benzene operations of the large producer
are only a small part of a large petrochemical operation so that these
7-68
-------
Table 7-37. THROUGHPUT COST SUMMARY FOR MODEL PLANTS
Plant
Existing facilities
Large producer
Small producer
Consumer and
bulk storage
terminal
New facilities
Large producer
Smal 1 producer
Consumer and
bulk storage
terminal
Control
option
I
II
III
IV
V(A)
V(B)
I
II
III
IV
V(A)
V(B)
I
II
III
IV
V(.A)
V(B)
I
II
III
IVC.A)
IV(B)
I
II
III
IV(A).
IV(B)
I
II
III
IV (A)
IVCB)
Benzene
throughput
(10° liters
per year)
224.6
224.6
224.6
224.6
224.6
224.6
46.3
46.3
46.3
46.3
46.3
46.3
42.1
42.1
42.1
42.1
42.1
42.1
224.6
224.6
224.6
224.6
224.6
46.3
46.3
46.3
46.3
46.3
42.1
42.1
42.1
42.1
42.1
Capital cost
for capital
with 15-
year life
($)
0
0
25,400
25,400
284,100
242,000
0
0
0
0
250,300
208,200
0
0
0
0
230,500
188,500
0
25,400
25,400
266,100
244,000
0
0
0
241,9.00.
199,800
0
0
0
225,700
183,700
Capital cost
for capital
with 10-
year life
($)
0
10,400
'141,800
191,500
0
0
7,000
7,000
36,800
65,500
0
0
0
0
17,100
33,400
0
0
0
56,100
96,000
0
0
6,100
44,600
68,900
0
0
0
15,100
28,200
0
0
Total
annu-
al i zed
cost
($)
, ° a
(l,000)a
21,000
29,100
68,000
71,800
2,200
2,200
5,700
10,700
71,300
65,700
0
0
2,300
5,000
69,800
62,100
0
5,200
11,200
66,400
68,500
1,900
4,000
8,100
66,700
64,200
0
100
2,20.0
67,200
61,200
Credit is indicated by parenthesis.
7-69
-------
benzene revenues are only a small part of the total revenue of the entire
operation. The question arises as to the separability of the benzene
facilities from the rest of the operation, and the effect that the assump-
tion of separability has on the computed impacts. It can be argued that
the multiproduct firm has at least as much flexibility as a single-product
firm, so that by separating the benzene facilities from the rest of the
operation as if it were a separate plant, an upper bound is set on the
impacts—the multiproduct firm would be able to find more ways to minimize
impacts than the conceptual plant producing benzene only.
Given the revenue above and the ratio of net income to revenue of
0.071 as reported in Table 7-35, the net after-tax profit or income of
the model plant from benzene operations is estimated to be $5.42 million
(0.071 x $76.36 million). Given revenue and net profit, the assets of
the firm are computed as follows. Recall that cash flow divided by
assets gives the capital recovery coefficient; i.e.,
F = £ (7-1)
where F = capital recovery coefficient, CF = cash flow, and A = assets.
Further, cash flow equals net income plus depreciation, as shown in the
following formula:
CF = NI + DEPR (7-2)
where NI = net income, and DEPR = depreciation. Using an asset life of
20 years and applying the rule of thumb relating asset life to depreciation
life, a depreciation life of 16 years results, implying a depreciation
rate of 0.0625. Substituting this relation into Equations 7-1 and 7-2
produces the following relationship:
_ Ml + (0.0625 x A)
~ A
Solving for A gives:
A =
NI
(F-0.0625)
From the equations for calculating the return on investment, the capital
recovery coefficient for a 10 percent return on investment with a lifetime
of 20 years is 0.11746. Using the previously computed value of NI of
7-70
-------
$5.42 million, A is computed to be $98.62 million. Now rearranging
Equation 7-1, CF = F x A. The cash flow of the model plant is thus:
CF = 0.11746 x $98.62 million = $11.58 million
Similarly, solving for DEPR as the product of the depreciation rate and
the stock of assets:
DEPR = 0.0625 x $98.62 million = $6.16 million
The total revenue of a firm less the cost of goods sold gives the
before tax profit. At the corporate tax rate of 46 percent,3 the following
relation is applicable:
NI = (Revenue - Cost of goods sold) x (1-0.46)
Solving for the cost of goods sold:
Cost of goods sold = Revenue - (NI/0.54)
= $76.36
«^
ion
$5.42 million _ cc 00 .,,
--- n~~54 - 66.32 mill
ion
It will be convenient later on in the calculations to break the cost of
goods sold into two components, a depreciation component and a "miscellaneous"
component, which includes all expenses of the firm except depreciation
and income tax. Defining the "miscellaneous" component as S yields:
Cost of goods sold = S + DEPR.
Then, solving for S:
S = Cost of goods sold - DEPR
= $66.32 million - $6.16 million = 60.16 million.
Because the baseline estimates for cash flow, net income, depreciation,
miscellaneous annual costs, revenue, and assets for the large producer in
our example calculation are now available, the impacts can be computed.
Note that the marginal tax rate, MTR, is identical to the corporate tax
rate for corporations with very high taxable incomes.
7-71
-------
7.4.3.2 Impacts for Large Benzene Producer in Example Calculation.
Reference to Table 7-37 shows that the costs for the large producer under
Option III are broken down as $25,400 expended on a capital investment
for capital with a 15-year life, $141,800 on capital with a 10-year life,
and $21,000 total annualized cost. The total annualized cost includes
capital recovery allowances, capital overhead charges, and annual costs
for all other aspects of compliance with the option—maintenance, inspection,
recovery credits, and labor. It is convenient to break down the total
annualized cost reported in Table 7-37 as follows:
$ 3,340 Capital recovery for capital with a 15-year life
23,077 Capital recovery for capital with a 10-year life
-5,417 All other annual costs and credits
$ 21,000 Total annualized cost
The capital recovery charges are obtained by multiplying the capital
investment (e.g., C15 for capital with a 15-year life) by the capital
recovery factors for a 10 percent return on investment and the appropriate
asset life. For 10-year assets the factor is 0.1627, and for 15-year
assets the factor is 0.1315. The "all other" component includes the
capital overhead costs, and is calculated as the residual between the
total annualized cost and the combined capital costs. Because one component
of the "other" annual costs is a credit for benzene not lost to emissions,
the "all other" component costs may be either positive or negative.
Having separated out the costs in this way, the change in return on
investment that can be attributed to the option can now be computed. The
new cash flow after the imposition of the option is:
CF* = (PQ - S - M - DEPR*)(1 - 0.46) + DEPR*
where
PQ = price x quantity sold, which yields revenue
M, the miscellaneous annual costs and credits of the option
= total annualized costs (TAC) - (C15 x 0.1315 + CIO x 0.1627)
DEPR* = DEPR + (C15 + CIO) x 0.0625
7-72
-------
C15 = the capital outlay for the assets with a 15-year life required by
the option
CIO = the capital outlay for the assets with a 10-year life required
by the option.
Because by assumption, price and quantity remain unchanged, revenue also
remains unchanged, and the only differences in cash flow attributed to the
option are the increased annual costs of the option, M, which become a
direct charge against before-tax revenue, and the change in depreciation
charges from DEPR to DEPR*. Using the equations for DEPR* and CF*, DEPR*
is $6.17 million, and CF* is $11.59 million. (The baseline cash flow was
$11.58"mill ion.)
Recall that in the baseline case, the rate of return was the value
of r that satisfies the following equation:
CF r
Assets
Multiplying both sides by Assets:
- (1+r)-520
CF =
1 - (1+r)-20
x Assets
By identical reasoning it can be shown that the rate of return upon
implementation of an option is the value of r that satisfies the following
equation:
CF* =/
x Assets
VI - (1+r)
x CIO
( r \
\1 - (l+r)~ /
C15
This equation can be solved for r because the values for CF*, Assets,
CIO, and C15 are available. Unfortunately, the form of the equation
precludes solving it explicitly for r. Two practical alternatives in
this case are (1) to let a computer solve the equation using numerical
iteration, or (2) to use a Taylor series approximation. The latter
method was chosen because within the narrow range of values for the
impacts of this option, the Taylor series has neglegible approximation
error.
7-73
-------
Solving for r, the rate of return upon implementation of Option III is
estimated to be 0.099767. The second impact, the price change that is just
large enough to allow the firm to maintain its ROI, is now evaluated. The
new cash flow necessary to maintain ROI is computed using the equation:
CF1 = (A x F) + (CIO x F10) + (CIS x F15)
(7-3)
where
CF' = the new cash flow needed to maintain ROI,
F10 = the capital recovery factor for 10 percent ROI and a 10-year
asset life,
F15 = the capital recovery factor for a 10 percent ROI and a 15-year
asset life.
All other symbols have already been defined. Because CF' is related to
the new price as described in the following equations:
CF1 = NI' + DEPR*
Ml1 = 0.54(P'Q-S-M-DEPR*)
(7-4)
(7-5)
where NI' = new net income and P' = new price, Equations 7-3, 7-4, and
7-5 can be solved explicitly for P' as follows:
P' - lf(A x F)+(C10 x F1Q)+(C15 x F15)-(DEPR* x 0.46) . c . M~"l ,,',.
P _-j^ _ + S+MJ (7-6)
Substituting the numbers already obtained for the variables on the right
side of Equation 7-6, P1 = $0.340154 per liter for the example calculation.
This represents a price increase of about five hundredths of 1 percent.
This completes the example calculation for the large benzene producer
under Option III. The methodologies have been shown for computing the ROI
assuming no change in price, and the price assuming no change in ROI,
where all the information used as input to the calculation is found in
Tables 7-35 and 7-37. Similar methodologies are shown for estimating
the impacts on the consumers of benzene.
7.4.3.3 Baseline Characteristic of Benzene Consumers in Example
Calculation. The analysis of the benzene consumer differs from that for
a benzene producer in two fundamental ways. First, the costs of an option
7-74
-------
are not reflected in a change in price or ROI for benzene, but instead in
a change in price or ROI on a product containing benzene. Second, the
benzene consumer must pay more for benzene if the producers can pass on
their cost increases instead of absorbing them. That is, the consumer
has two potential sources of additional costs if the option is implemented:
(1) The firm's own capital and annual costs for bringing the plant into
compliance, and (2) the increased price the firm might have to pay for
benzene as a result of costs imposed on producers.
Table 7-38 shows the prices and throughputs for three different
types of benzene consumers. These three types of consumers utilize about
85 percent of all benzene consumed. The outputs of the plants are ;
based on the benzene input of the model plant and the fixed relation
between benzene input and final product output based on the chemical
reactions involved in manufacture. For example, in a stoichiometric
reaction involving the manufacture of styrene, 78.1 kilograms of benzene
would combine with 28.1 kilograms of ethylene to eventually produce
104.1 kilograms of styrene and 2.1 kilograms of hydrogen gas. However,
the typical yield attained in practice is about 84.percent of the theoretical
yield. Thus, 78.1 kilograms of benzene gives about 87.44 kilograms of
styrene. Based on the fact that benzene weighs 0.883 kg per liter, the
output of the model styrene plant is computed to be:
/,„ -, •-,-,• TJ. u 0.883 kg benzene .. 87.44 kg styrene
42.1 million liters benzene x a x a *
liter benzene 78.1 kg benzene
= 41.61 million kg of styrene
Given the price and quantity (throughput) data from Table 7-38 and
the baseline operating ratios from Table 7-36, the same methodology
that was used to compute the baseline operating characteristics of the
model producer plant is used to compute the operating characteristics of
the model consumer plant. The step-by-step calculations are shown here
for the styrene plant:
Revenue = Price x Quantity = PQ = $27.50 million
NI = 0.068 x Revenue = $1.87 million
A, Assets = NI/(F-0.0625) = $34.02 million
CF = F x A = $4.00 million
7-75
-------
Table 7-38. CHARACTERISTICS OF MODEL CONSUMER PLANTS
34, 35
Type of plant
Price
Quantity
Cumene
Styrene
Cyclohexane
$0.437kg.
$0.661/kg.
$0.396/liter
54.218 million kg.
41.61 million kg.
51.033 million liters
7-76
-------
DEPR = 0.0625 x Assets = $2.13 million
Cost of Goods Sold = Revenue - (NI/0.54) = $24.04 million
S = Cost of Goods Sold - DEPR = $21.91 million
7.4.3.4 Impacts for Benzene Consumer Example Calculation. Impacts
for the model consumer are computed in the same manner as for the producer
except that two impacts are calculated. The first impact is computed
based on the assumption that the benzene producers are unable to pass on
their increased costs as higher prices. The second impact is computed
based on the assumption that benzene producers are able to pass on all
of their increased costs as higher prices. The only difference in the
two calculations is that in the second calculation, the increase in the
cost of benzene to the consumer is added into the miscellaneous component
of total annual ized cost. Here are the step-by-step computations in the
no-change- in-benzene-price scenario:
DEPR* = DEPR + (CIO + CIS) x 0.0625 = $2.13 million
M = TAC-(C15 x 0.1315 + CIO x 0.1627) = -$482
CF* = (Revenue-S-M-DEPR*)(l-MTR) + DEPR* = $3.998 million
r = 0.099926
CF1 = (Assets x 0.1175) + (0.1627 x CIO) + (0.1315 x C15)
= $4.00 million
=1 (CF-
Q y
- (DEPR* x 0.46)
0.54
=0.661090
In the second calculation, for the full-producer-cost-passthrough scenario,
the formula for M changes to:
M = TAC-(C15 x 0.1315 + CIO x 0.1627) + (Pb - $0.34) x 42.1 million liters
where P. = the highest price that the benzene producers might charge for
benzene in order to maintain their ROI under the same control option for
which the consumer impacts are computed. For reasons that are discussed
in Sections 7.4.4.1 and 7.4.4.4, this price is always the price charged
by the small existing benzene producer. In the example calculation,
P. = $0.340191 per liter. Thus, because the model consumer uses
7-77
-------
42.1 million liters of benzene per year, the cost increase for benzene is
$0.000191 per liter times 42.1 million liters, which equals $8,041. In
the no-change-in-benzene-price computation, M was a $482 credit. In the
full-producer-cost-passthrough scenario, M equals $7,559.
Because everything else is the same under both options, only the
calculations that depend on M must be changed. Running through the
step-by-step calculations, CF* = $3,994 million instead of $3,998 million
as was just calculated for the no-change-in-benzene price scenario. This
reduction in cash flow implies a reduction in ROI. Solution of the
Taylor series gives r = 0.099767, as opposed to the previous value of
0.099926. Substituting the larger value of M into the P1 evaluation
gives a new price of $0.661283 instead of the $0.661090 value computed
earl.ier.
This concludes the example calculations for the benzene producer and
consumer. For the producers, the methodologies have been shown for
computing ROI assuming no change in price, and the change in price assuming
no change in ROI. In reality, the producers will be likely to pass on
some of the cost increase, and will have to absorb the rest. The estimation
of the exact mix of passthrough and absorption is unnecessary because
both impacts are inconsequential, so that any mix between the two would
also be inconsequential.
For the consumer, four cases of impacts have been computed, including
ROI and price changes based on no cost passthrough by producers, and ROI
and price changes based on full cost passthrough by producers. In each
case, the methodology is identical to that applied to producers. The ROI
change is based on the assumption of no change in the price of the product
produced, be it cumene, styrene, or cyclohexane, and the price change is
computed based on the maintenance of a 10 percent ROI. The impacts
calculated using the methodology discussed in the above examples are now
presented.
7.4.4 Economic Impacts for Model Plants
Table 7-39 shows the impacts for the existing and new producer model
plants, under the five control options for existing plants, and the four
control options for new plants. Tables 7-40 and 7-41 show the impacts
7-78
-------
Table 7-39. ECONOMIC IMPACTS FOR MODEL BENZENE PRODUCER PLANTS
Control
Plant option
Existing facility
Large producer I
II
III
IV
V(A)
V ;
V(B)
Small producer I
II
III
IV
V(A)
V(B)
New facility
Large producer I
II
III
IV(A)
IV(B)
Small producer I
II
III
IV(A)
ROI
under assump-
tion of
no change
in price
0.100000
0.100001
0.099764
0.099680
0.099423
0.099414
0.099907
0.099907
0.099707
0.099461
0.097178
0.097438
0.100000
0.099922
0.099857
0.099441
0.099443
0.099920
0.099741
0.099537
0.097344
0.097502
Price
under as'sump-
Percent
change
in ROI
0.000
0.001
-0.236
-0.320
-0.577
-0.586
-0.093
-0.093
-0.293
-0.539
-2.822
-2.562
0.000
-0.078
-0.143
-0.559
-0.557
-0.080
-0.259.
-0.463
-2.656
-2.498
tion of
no change
in ROr
0.340000
0.340000
0.340154
0.340209
0.340377
0.340383
0.340061
0.340061
0.340191
0.340353
0.341862
0.341687
0.340000
0.340051
0.34009.3
0.340365
0.340364
0.340052
0.340169
0.340303
0.341751
0.341644
Percent
change
in price
0.000
0.000
0.045
0.062
0.111
0.113
0.018 "
0.018
0.056
0.104
0.548
0.496
0.000
0.015
0.027
0.107
0.107
0.015
0.050
0.089
0.515
0.483
aUnits for price:
dollars per liter of benzene.
7-79
-------
TABLE 7-40. ECONOMIC IMPACT FOR MODEL BENZENE CONSUMER PLANTS
UNDER ASSUMPTION OF,FULL COST ABSORPTION BY PRODUCERS
Control
Plant option
Existing facilities
Cumene I
II
III
IV
V(A)
V(B)
Styrene I
II
III
IV
V(A)
V(B)
Cyclohexane I
II
III
IV
Y(A)
V(B)
New facilities
Cumene I
II
III
IV(.A)
IV (B)
Styrene I
II
III
IV(A)
IV(B)
Cycl ohexane I
II
III
IV(A)
IV(B)
ROI
under assump-
tion of
no change
in price
0.100000
0.100000
0.099912
0.099817
0.098068
0.098302
0.100000
0.100000
0.099926
0.099846
0.098374
0.. 098571
0.100000
0.100000
0.099899
0.099789
O.Q97774
0.098043
0.100000
0.099968
0.099892
0.098135
0.098329
0.100000
0.09.9973
0.099910
0.098430
Q. 098594
0.100000
0.099963
0. 099876
O.Q97851
0.098074
Percent
change
in ROI
0.000
0.000
-0.088
-0.183
-1.932
-1.698
0.000
0.000
-0.074
-0.154
-1.626
-1.429
0.000
0.000
-0.101
-0.211
-2.226
-1.957
O.QOO
-O.Q32
-0.108
-1.865
-1.671
0.00.0
-0.027
-0.09.0
-1.570
-1.406
-O.QOO
-0.037
-0.124
-2.149
-1.9.26
Price
under assump-
tion of
no change
in ROI*
0.430000
0.430000
0.430069
0.430145
0.431537
0.431350
0.661000
0.661000
0.661090
0.661187
0.662986
0.662744
0.396000
0.396000
0.396074
0.39.6154
0.397633
0.397434
0.430000
0. 430.026
0.430085
0.431484
0.431328
0.661000
0.661033
0.661110.
0.662918
0.662716
0.396000
0.39.6027
0.39.6090
0,. 397577
0.. 397711
Percent
change
in price
0.000
0.000
0.016
0.034
0.357
0.314
0.000
0.000
0.014
0.028
0.300
0.264
0.000
0.000
0.019
0.039
0.412
0.362
0.00.0
0,006
0.020
0.345
0.309
0.000
0.005
O.Q17
0.29Q
0.260
0.000
0.0.07
0.023
0,398
0.356
*Units for price: dollars per kilogram for cumene and styrene, dollars per
liter for cyclohexane.
7-80
-------
Table 7-41. ECONOMIC IMPACT FOR MODEL BENZENE CONSUMER PLANTS
UNDER ASSUMPTION OF FULL COST PASSTHROUGH BY PRODUCERS
Plant
Existing facilities
Cumene
Styrene
Cyclohexane
New facilities
Cumene
Styrene
Cyclohexane
Control
option
I
II
III
IV
V(A)
V(B)
I
II
III
IV
V(A)
V(B)
I
II
III
IV
V(A)
VCB)
I
II
III
IV(A)
IV(B)
I
II
III
IV(A)
IV(B)
I
II
III
IV(A)
IV(B)
ROI
under assump-
tion of
no change
in price
0.099940
0.099940
0.099723
0.099468
0.096239
0.096642
0.099950
0.099950
0.099767
0.099553
0.096835
0.097174
0.099931
0.099931
0.099680
0.099386
0.095666
0.096129
0.099940
0.099778
0.099544
0.096305
0.096669
0.099950
0.099814
0.099616
0.096891
0.097197
0.099931
0.099744
0.099474
0.095742
0.096160
Percent
change
in ROI
-0.060
-0.060
-0.277
-0.532
-3.761
-3.358
-0.050
-0.050
-0.233
-0.447
-3.165
-2.826
-0.069
-0.069
-0.320
-0.614
-4.334
-3.871
-0.060
-0.222
-0.456
-3.695
-3.331
-0.050
-0.186
-0.384
-3.109
-2.803
-0.069
-0.256
-0.526
-4.258
-3.840
Pri ce
under assump-
tion of
no change
in ROIa
0.430047
0.400047
0.430219
0.430421
0.432993
0.432669
0.661061
0.661061
0.661283
0.661543
0.664867
0.664448
0.396050
0.396050
0.396233
0.396447
0.399180
0.398835
0.430047
0.430175
0.430361
0.432940
0.432647
0.661061
0.661227
0.661466
0.664799
0.664420
0.396050
0.396186
0.39.6383
0.. 399123
0.398812
Percent
change
in price
0.011
0.011
0.051
0.098
0.696
0.621
0.009
0.009
0.043
0.082
0.585
0.522
0.013
0.013
0.059
0.113
0.803
0.716
0.011
0.041
0.084
0.684
0.616
0.009
0.034
0.071
0.575
0.517
0.013
0.047
0.097
0..789.
0.710
Hlnits for price: dollars per kilogram for cumene and styrene, dollars per
liter for cyclohexane.
7-81
-------
for three consumer plants under the same options. The reader is cautioned
that the data used in these calculations do not support the level of
accuracy expressed in the tables. The numbers do indicate that the
•economic impacts generally are very small; that price changes, if any,
are likely to be increases; and that changes in the return on investment,
if any, are likely to be decreases. The numbers also show the relative
magnitude of changes among model plants and control options. The following
sections present detailed discussions of these impacts.
7.4.4.1 Economic Impacts for Large Existing Benzene Producer. The
existing large producer has relatively little trouble meeting the require-
ments of any option. In Option I the tank configuration is such that the
large producer suffers no impacts whatsoever. In the other options, the
cost and asset bases are so large that the additions to the costs and
assets occasioned by the control options are inconsequential by comparison.
Even so, it is instructive to note that Options V(A) and V(B) have much
larger impacts than the other options. The changes in return on investment
or price are almost twice as large under Option V than under Option IV, the
next less expensive option.
7.4.4.2 Economic Impacts for Small Existing Benzene Producer. As
with the large existing producer, the impacts for the small existing
producer can be regarded as inconsequential with the possible exception
of the vapor control options, Options V(A) and V(B). Under these options
the price change is roughly one-half of 1 percent. The impacts of these
two options are much larger for the small producer than they are for the
large producer because both large and small producers must install roughly
the same amount of capital, but the large producer spreads the expenditure
over a much larger throughput, thereby achieving an economy of scale
relative to the small producer. The effect would be even more noticeable
for real-world producers that are smaller than the small producer model
plant.
7.4.4.3 Economic Impacts for Existing Benzene Consumers. The
impacts for the consumers are similar in pattern to those of the small
producer—that is, the impacts for the tank reconfiguration options,
Options I through IV, are quite small, but in comparison, the impacts for
vapor control options are quite large.
7-82
-------
The reason that the impacts for the consumers differ for the different
chemicals being produced is because the consumers vary as to the percent
of benzene embodied in their final output, and as to the total asset
base. The asset bases of the three model consumers are $28.85 million
for cumene, $34.02 million for styrene, and $25.00 million for cyclohexane.
Because all three plants have the same costs imposed, it is expected that
the ROI will be most greatly affected in the cyclohexane model plant
because it has the smallest asset base. Conversely, the smallest ROI
change under a given option should occur in the styrene plant. The same
reasoning holds for the price changes. Inspection of Table 7-40 shows
this reasoning to be borne out by the figures. :
The percentage of benzene embodied in the final output also affects
the impacts computed for the producer cost passthrough scenario in Table 7-41.
This is explained by considering the following figures. For the cumene
producer, the revenue is $23.31 million, and the amount paid for benzene
as an input is $14.322 million (0.340191/1iter x 42.1 million liters).
Thus, the value of benzene input is 61 percent of revenue. For styrene,
the figure is 52 percent, and for cyclohexane it is 71 percent. Thus, a
given increase in the price of benzene will have its greatest impact on
cyclohexane producers, arid its smallest impact on styrene producers.
This pattern of passthrough cost impact is exactly the same as the pattern
of impacts of their own costs, and because the effects are additive, the
same overall pattern of relative impacts is expected under the cost
passthrough scenario as existed under the no-change-in-benzene price
scenario, except that the impacts would be larger. Inspection of
Table 7-41 shows that this is true. Among all consumers, the largest
changes in price and ROI occur for the cyclohexane producer. Therefore,
this case is discussed in some detail. Note that the costs of Options V(A)
and V(B) are seven to nine times larger than the cost for Option IV, the
next less expensive option, but that no price increase exceeds 1 percent
even under the assumption of full cost passthrough on the part of the
producers. In judging the overall consequences of implementing the
option, it must be remembered that consideration of the price increases in
both benzene and products containing benzene is, in a sense, double
7-83
-------
counting. If one expects full cost passthrough by the producers, then
the full impact on the economy is eventually felt through the price
increases in products containing embodied benzene. The society loses
buying power due to a price rise in intermediate goods such as benzene
only when the price rise filters down to final goods and services. In
this sense, the passthrough impacts are probably the most meaningful
figures pertaining to benzene consumers. On the other hand, if the
impacts of the option on an individual industry are of interest, then the
passed-on increases from supplying industries are not directly applicable.
Thus, the no-change-in-benzene-price impacts are most appropriate in
assessing the consequences of the option on the benzene-consuming indus-
tries. In either interpretation, the price and ROI impacts listed in
Tables 7-40 and 7-41 are inconsequential, with the possible exception of
the vapor control options.
7.4.4.4 Economic Impacts for New Facilities. The economic impacts
for new facilities differ from those for existing facilities primarily
because the Petroleum Liquid Storage Tank NSPS requires that all new
external floating-roof tanks have primary and secondary seals, and only
the incremental costs of the options over the NSPS levels can logically
be attributed to the control options analyzed here. That is, for existing
facilities, the costs of going from the fixed-roof tank CTG to the control
option is attributable to the control option, whereas for a new facility,
only the costs of going from the NSPS to the control option is attributable
to the control option. Another difference between the costs for new and
existing facilities is that it is never more costly, and it is usually
much cheaper, to upgrade to a particular specification when the equipment
involved is still in design, rather than to retrofit existing equipment.
For the tank configuration options, Options I through III, the
difference attributable to the NSPS baseline can make a considerable
relative difference between comparable options applied to new and existing
facilities. However, in absolute terms, because the existing plant
impacts were already judged inconsequential, the new plant impacts (which
are smaller) are perforce also inconsequential. In the case of the vapor
control options, Options IV(A) and IV)B), the difference between new and
7-84
-------
existing plants is attributable to the savings of new over retrofit costs
for the piping and switches. These differences are rather small, averaging
4.2 percent for Option IV(A) and 2.8 percent for Option IV(B). Thus, the
new plant impacts repeat the pattern already observed for the existing
plant impacts—the tank reconfiguration options are uniformly inconsequential,
none amounting to more than a one-tenth of I percent rise in price,
whereas the vapor control options are many times more expensive.
7.4.4.5 Summary of Economic Impacts for Model Plants. Two general
observations serve to summarize the results of the economic analysis of
the model plants:
(1) Except for the vapor control options, the impacts can be considered
inconsequential. Among the tank configuration options, the
largest price impact is a rise of 0.104 percent—a price change
that would hardly be noticed in an industry subjected to the
large increases in price imposed by foreign oil suppliers;
(2) The impacts attributed to the vapor control options are also
quite small in comparison to the impacts imposed by foreign oil
prices, but in comparison to the tank configuration options,
the vapor control options are much more expensive. It is
worthwhile to note that the vapor control options could possibly
be the cheapest options for a very large producer. This would
likely be the case for a plant that has or expects to have
other vapor control requirements in addition to benzene tanks,
or for a plant that already has much of the vapor control
equipment in place. However, the general conclusion is that,
for the model plants specified here, the vapor control options
are decidedly less economical than the tank configuration
options, costing roughly three to ten times as much. In
individual cases, though, it might be economical for a specific
plant to utilize a vapor control system.
7.4.5 Economic Impacts for Bulk Storage Terminals
The economic analysis for bulk storage terminals differs from that
for producers and consumers in one main respect. The terminals are
7-85
-------
providing a service, whereas the producers and consumers are providing a
product. The reason this is important is that the costs of storage
services are only a small part of the overall cost of most products, but
they constitute the whole output of terminal facilities. Earlier sections
of this chapter have shown that the increases in costs resulting from
the proposed control options affect the producers and consumers very
little. Thus, the storage component of the $0.34 baseline price of
benzene must be quite small. However, a similar analysis for terminals
would reveal a different story because the storage cost component of
their output is 100 percent. Unfortunately, there are no data available
for computing price and ROI changes for terminals in as quantitative a
way as was done for the producers and consumers of benzene. Thus, none
are computed. Instead, the impacts are approached in a nonquantitative
way. Because the service being offered by terminals is an almost perfect
substitute for the terminal customer providing the same service, the
following two propositions are true:
(1) Any price increase in terminal storage services cannot possibly
raise the price of benzene or products derived from benzene
very far above the prices computed for those products earlier
in this chapter.
(2) As long as terminals pass on only the price increase necessary
to cover their increased costs, terminal customers will not
shift to other terminals or to self-provided services, because
the costs of these options will have risen by the same amount.
Put another way, these two propositions can be viewed as essentially
treating the terminal as an extension of the producer's or consumer's own
plant. The only difference is that, instead of bearing the impacts
directly, they are borne indirectly through the terminal owner.
When a benzene producer or consumer opts to use storage at a public
terminal instead of storage at a self-owned facility, it must be because
he or she views the terminal service as less costly than a self-owned
facility. Even when the terminal is remote from the main plant site,
so that the apparent reason for using the terminal is its location, the
benzene producer or consumer always has the option of building a storage
7-86
-------
facility near the remote site serviced by the terminal. The decision to
lease is evidence that the management of the benzene facility felt that
leasing was cheaper than self-provision at the time the decision was
made. The terminal must, therefore, be providing the storage services at
a competitive price that just covers costs and a fair ROI. If the terminal
were charging more than that price, the producers and consumers would
self-provide. If the terminal were charging less, it would go out of
business. The same argument holds after implementing an option, so
that the impacts for a benzene producer or consumer who leases storage
tanks would differ little from those of a producer or consumer who owns
storage tanks. The lease price should go up by almost exactly the same.
amount as the total annualized cost of the option (ignoring differential
tax treatments on the depreciation). If the terminal operator tried to
pass on more than the associated costs, it would become advantageous for
the benzene producer or consumer to self-provide. The terminal operator
would not pass on less, because of his or her wish to maintain ROI. The
terminal operator would sooner maintain ROI by using tanks to store other
substances with physical properties similar to benzene such as lube oils,
fuel oils, or glycols, rather than suffer a reduced ROI by providing
benzene storage below cost. Because terminals have very few of their
tanks in benzene service, this minor change in the mix of services provided
could be performed with a minimum of disruption in overall terminal
operations. Employment would be unlikely to change at all, and revenues
would change only in response to cost changes.
Consumers and producers of benzene will not necessarily shift away
from benzene storage at terminals as a result of increased prices for
these services, because the costs of the substitute, self storage, will
have gone up by roughly the same amount. Therefore, the cost considerations
which caused them to lease rather than buy in the first place will not
have changed as a result of implementing an option.
7.4.6 Analysis of Closure Option
The closure option is presumed to be the ultimate control for any
hazardous substance. However, such a drastic measure as closure is
rarely justified because of the excessive burden involved. The impacts
7-87
-------
of closure would be twofold. Direct impacts, such as a reduction in
employment, and writing off of capital stock in the benzene-producing and
benzene-consuming industries would affect the producers and consumers
themselves. Indirect impacts, such as price increases, quality reductions,
and reduced output would arise from the attempts of the downline users of
benzene-derived substances to buy the benzene-derived substances from
foreign sources or substitute other less suitable substances for them.
7.4.6.1 Direct Impacts on Benzene Producers and Consumers. A total
ban on benzene would cause the people employed in benzene production to
become unemployed or occupied by other jobs. The number of people so
affected can be roughly estimated as follows. The 1976 Annual Survey of
oc
Manufactures shows that for SIC 2911, Petroleum Refining, the value of
shipments was $77,507.3 million, with the industry employing 101,700
workers at that time. This implies a labor/output ratio of 1.31213 x
10 6 person-years per 1976-dollar of output. To restate the ratio in
first-quarter 1979 dollars, it must be multiplied by the Bureau of Labor
37
Statistics1 producer price index for refined petroleum products in 1976 ,
OQ
276.6, and divided by the same index for first-quarter 1979 , 350.6.
The result is 1.0352 x 10 6 person-years per 1979 dollar of output. The
total production of benzene in 1979 is estimated to be about 6,420 million
OQ
liters or about $2,183 million dollars, at the price of $0.34 per liter.
Multiplying the person-years per dollar coefficient by the dollar value
of total output gives the person-years embodied in the total output. The
result is 2,260 person-years, a rough estimate of the number of jobs
which would be displaced in the benzene-manufacturing sector in the event
of closure. A similar computation for benzene consumers suggests that
about 17,600 jobs would be displaced in the consuming industries, giving
a total of 19,860 affected jobs. It is important to note, however, that
the production of benzene is typically carried out in a facility that
produces a variety of chemicals, not just benzene. It is likely that the
individuals currently employed in benzene production would simply adjust
the mixture of time they spend producing various outputs, adjusting out
of benzene production and into other production; however, they would
likely be employed at essentially the same job at the same place of work.
7-88
-------
It is unlikely that any significant unemployment would result from
closure.
A similar analysis holds for the other major factor of production,
capital. The quantity of capital in benzene production can be computed in
much the same way as capital was computed for the model plants. Recalling
from Table 7-35 that the ratio of net income to revenue in the petroleum-
refining industry is 0.071, the net income to the petroleum-refining
industry that can be attributed to benzene operations is estimated as
$155 million (0.071 x $2,183 million). From Table 7-35 the capital
recovery coefficient for the petroleum-refining industry is 0.09119, and
the depreciation rate is 0.0652. Substituting these values into the
formula for assets derived from Equations 7-1 and 7-2, the capital stock
of the petroleum-refining industry that is dedicated to benzene production
is estimated at $6,156 million. A similar computation for the consumers
of benzene gives $10,563 million (These estimates are very sensitive to
small changes in the ratios used in the calculations and, therefore,
should be considered only ballpark estimates.) Just as with labor,
however, the capital involved in this production is likely to have alter-
native uses. Thus, if closure were imposed, it is unlikely that all or
even a substantial part of the total capital affected would be totally
written off. It would be placed into alternative service. To the extent
that certain items of equipment were useful solely in benzene-related
activities, these items would have to be scrapped, modified for non-benzene
use, or sold to benzene producers located outside the U.S. This forced
action would be the direct capital cost impact of the closure option.
7.4.6.2 Indirect Impacts on Users of Benzene-Derived Products.
There are two substitution alternatives available to the users of
benzene-derived products. The first is to purchase the same product from
a foreign producer, and the second is to substitute a non-benzene product
with similar characteristics. In the foreign-supply case, prices will be
likely to rise reflecting the increased transportation costs of the
foreign benzene, and the U.S. balance of payments will also suffer.
Furthermore, the banning of U.S. benzene production, about 40 percent of
world production, would create a worldwide upward shift in the supply
7-89
-------
curve for benzene, which could have the effect of increasing the world
price of benzene while simultaneously reducing the total quantity produced.
In the case of substitution of non-benzene-derived inputs for benzene-
derived products, the quality of the output produced from the substituted
inputs will be likely to have fallen from the pre-closure level, and the
price of the output may also rise. This is so in the absence of significant
technical advances creating substitutes that are not now available. If a
substitute input was available before the closure, the manufacturer would
already have been using it unless it was less suitable, more costly, or
both, when compared to the benzene-derived input being used at the time
of the closure. Thus, if closure is implemented, the substitute is most
likely to be less suitable, more costly, or both.
These indirect impacts, increases in prices, decreases in quality,
or combinations of both, are the most important impacts that could be
expected in the event of closure. Compared to the direct impacts, they
would be much further reaching and potentially more expensive. A quanti-
tative estimate of the cost of the indirect impacts is beyond the scope
of this study.
7.5 SOCIOECONOMIC AND INFLATIONARY IMPACTS
7.5.1 Inflationary Impact Statement Thresholds
Executive Order 12044 requires that the inflationary impacts of
major legislative proposals, regulations, and rules be evaluated. A
regulation is considered a major action requiring the preparation of
an inflationary impact statement if it exceeds either or both of the
following thresholds:
(1) Annualized costs of compliance, including capital charges,
equals $100 mi 11 ion.per year
(2) Total additional cost of production exceeds 5 percent of the
selling price of the product.
The following sections consider each of these thresholds in turn.
7.5.1.1 Annualized Cost Compared to $100 Million Threshold. In
order to compute the total annualized cost of an option affecting plants
in the United States in 1979, the number of such plants must be multiplied
by the corresponding cost per plant. Utilizing the estimate that there
7-90
-------
were 28 large producers, 34 small producers, 77 consumers, and 4 bulk
storage terminals existing in 1979, and using the cost data from Table 7-37,
we find that the total annualized cost of implementing Control Option III
in 1979 is $968,100.
In calculating the annualized costs for 1980, the additional total
annual costs for the new facilities constructed after 1979 must be added.
Assuming that there is no reduction in the number of existing facilities,
that there is a 5 percent annual growth rate in the number of facilities,
and that the new facilities pay a full year's worth of the costs, the
following expression can be used to compute the total annualized costs
(TAG) in-any year after 1979: :
TACt = TAC197g + [(l.OB)*"1979 - !][# plants in 1979][cost/piant]
Filling in the cost formula, assuming Option III for existing tanks and
Option II for new tanks apply, the total annualized cost in 1985 is
calculated to be $1,066,626 ($968,100 + $98,526). Annualized costs for
other combinations of options (i.e., other alternatives) and other years
are displayed in Table 7-42. No cost for any combination in any year
even approaches the $100 million threshold; therefore, based on this
criterion, none of the combinations represents a major impact.
In addition to the total annual cost of the combinations, the annual
expenditures on new capital should be of some interest. These expenditures
are calculated by multiplying the capital cost for each type of plant by
the number of plants of that type and summing the results. In 1979, the
total expenditures on new capital under Control Option III are calculated
to be $7,317,900. In subsequent years, the new capital expenditures
become much smaller because only the plants that come into existence in
that year need to buy the capital. In any year t, the number of plants
that come into existence may be computed using the formula:
New plants in year t = [(1.05)t"1979-(1.05)(t"1)"1979][# original plants]
Thus, the formula for calculating the new capital expenditures, K, in year t, is:
K. = (new plants in year t)(capital cost/plant)
U
The total cost in 1985 for Control Option II applied to the new plants is
thus:
7-91
-------
01
CO CO
Oro
U_ CO
°g
co _i
z: _i
o o
t-t a
«=t en
Z: t*~
»-H en
O UJ
oo:
-S
uj cy
_l I
CC CO
h-a:
=D i-l
CO Ll_
tS2
^5
P1-
CO CD
o z:
°p
Q CO
LU •—<
M X
l—i UJ
Si
ZeC
§3:
O O
CM
CJ
ra
•— O)
O 2=
ro
i— O)
i— OJ
-a 3:
(O
O)
O O LO
o o
LO CO
VO LO LO CO
ID
CT> CM CM LO
oocMr^ococococoooocQLo
CM CM vo 10
en co o oo
CM
CM
CM i — oo en LO CM en
cncnoo
i— 1 r— l
•to-
OTi
CO O O
CO
O CM
T-« r-« CM
CM
CD co en vo co ro CM co
CM en co
vo co en *"H co *5t* ^o co
cncncnooooo
LO
o «-i LO
CM en co
LO en LO
CM CM CM
o o
CO CD
en
vo
CO i-i -5l- CO CM
i-f CM CO
^J- «^- *»o
CM
r^.cooocooocococococococn
CO
O
O)
O)
(1)
o
fO
(U
r- en
10 C
to CO
-c: a)
01 c:
c E
•r- 3
lo
10 u
in
to 0)
cu
•4->-a
ro a>
O T-
i— O
ra a>
o o-
S- -r-
r- O
J3 l-
ro 4->
JZ O
o
LU O
ro
7-92
-------
[1.056-1.055][(81,500x28)+(44,600x34)+(15,100x77)+(15,100x4)] = $320,442
Capital costs for other combinations of control options and other years
are shown in Table 7-43.
7.5.1.2 Production Costs Compared to Five Percent Threshold. The
costs per liter calculated in the previous section translate to percentages
of price of 0.46 percent for the consumer and 0.42 percent for the producer,
when compared to the baseline price of $0.34 per liter3. These impacts
are considerably smaller than the 5 percent threshold. Therefore, none
of the options results in major impacts.
7.5.2 Foreign Trade Considerations
There are two aspects of foreign trade that must be considered in
evaluating the feasibility of an option: (1) Will implementation of the
option induce more importation of foreign goods, and (2) Will implementation
of the option reduce U.S. exports of the goods? An affirmative answer to
either question means that the option will worsen the U.S. balance of
payments and lead to reduced domestic industrial production. Both questions
could be answered in the affirmative if U.S. producers raised their
prices to pass on new costs, while producers in other countries, who do
not have to comply with U.S. laws, maintain their prices. In this case,
both U.S. and foreign consumers would tend to substitute other countries'
products for the U.S. output, reducing U.S. exports and increasing U.S.
imports.
In the case of benzene, only about 3.5 percent of the product consumed
in this country is imported, and only about 2.5 percent of the domestic
production is exported. Even if these small quantities underwent fairly
sizable proportional changes, their overall impact would not be large in
relation to the whole market. However, given the small ness of the price
changes expected to come about as a result of an option, it is unlikely
that the quantities imported or exported would be greatly affected through
price-induced substitution.
aThese cost-to-price ratios differ from the price change necessary to
maintain ROI in that the government absorbs some of the total annual!zed
costs through reduced taxes in the computation of price changes necessary
to maintian ROI, whereas these cost-to-price ratios compare per-liter costs
to baseline without considering where the costs are borne.
7-93
-------
u.
o
CO
^f
o
f«m ^J—^
•a: co
co _i
O CD
o a
o cn
I— t«-
cn
Ul rH
_i
ca a:
l__ l«
ra a:
CQ eC
c2 o1
|_« |
r— • H~
« t- ^rt
o:
O ''
o —
co co
i*Z
—i z:
I— \—
I-H
p cn
(O
•r—
.Q
E
O
O
•0 3
r— O)
O«— ^
z
•"•^w™1*
CO CQ
^^^^
I-H
T3 S
i— CO
CD Z!
^•^^-^
•a; «t
^"^
I-H
•O 3
o zi
^^ I-H
t— 1 I-H
I-H
•a 3
o z:
I-H I-H
I-H I-H
I-H
-0 X
o z:
HH I-H
I-H I-H
•O S
^™ Q)
1 — 1 I-H
S-
03
CO
CD
O
CO
CM
T— 1
cn
CSJ
o
o
in
co
i— i
in
co
•w-
O
o
vo
in
o
o
i— i
{_!
(y)
CD
CD
cn
JS,,
i-H
CO
A
r-.
o
o
CM
CTi
CM
in
^/)
o
o
0
CO
CO
CM
CT>
a*
i-H
in in
o in
CD ^^
in o
i— H ^^«
i-H i-H
in co
in «s»-
co r».
1^ CM
CT> CO
vo i*»
T-H I-H
o in
CD VO
co co
I— 1 I-H
CD CM
in cr>
r~* CM
o vo
r-H CO
in vo
CM CM
in cn
t^ CM
o vo
r-l CO
in vo
CM CM
CD CO
t~- CO
co co
o o
I— 1 I-H
O I-H
CO CO
cr> cr»
I— 1 i-H
co
CT>
CM
cn
CM
CM
in
CO
CO
r-H
CO
I-H
CO
CO
«d"
CM
^^
^-
O
CO
vo
CM
O
i-H
CO
vo
CM
CO
co
I— I
i-H
CM
CO
cn
r—4
CO
in
r^
CD
cn
CM
o
CO
in
vo
cn
i-H
in
in
in
,3.
vo
i-H
in
vo
O
CM
,_!
in
vo
o
cn
CM
in
o
o
CM.
T-H
OO
CO
cn
t-H
in o
cn vo
CM O
in co
in CM
co «*
CO r-H
in «si-
t~^ cn
co vo
vo vo
0 I-H
CM CM
CO CM
CO l~~-
r^. CM
CO i-H
«* in
CO CM
CO •*
in CD
O CM
CO CO
CO CM
CO <*
in CD
O CM
co co
in in
O CO
VO CM
CM CO
I-H t-H
CO CO
cn cn
I— 1 T-H
CO «*
i-H CM
CM cn
«* CO
cn vo
>* in
r^. co
co in
CM CD
in cn
I^, CO
CM CO
CM CM
o cn
CO VO
i^ vo
f>. «3-
CO VD
in in
«*• co
VO CO
«* CM
VO CO
co in
CO CO
«* CO
VO CO
«3- CM
vo co
co in
co co
r«> CM
cn cn
co m
CO «3-
i-H i-H
vo r^.
CO CO
cn cn
T-H i-H
O CO
^ CO
co r>
r>- en
«3- CM
vo r-^
"* 7
in o
CD CO
in cri
CO CO
o co
in vo
CM CM
co co
o «3-
cn in
CM CM
cn CM
in vo
CM a
in CD
cn in
o cn
1^ CO
co co
CM O
in CD
cn in
CD cn
r^ co
co co
i~H t^**
CM CO
CO CD
in vo
T— 1 I-H
oo cn
CO CO
en cn
i-H i-H
in
CM
CM
vo
i-H
CO
VD
CM
vo
in
vo
CM
in
fN^
VD
CO
in
vo
in
i^
cn
CO
CD
«=t-
in
cn
CO
o
CM
a^
CO
VD
T-H
o
cn
cn
i-H
4-^
03
^~
rVi
cn
C.
to
as
JO
"O •
to
4-> cn
fO~ C
3 TJ
O 03
r— • CO
(O *d
U
c
as i —
0
•o o
c
ta cu
1C
S-
03 C
r— "i —
r—
0 -0
•a a>
cn «t-
o? o
r-H Q)
0-
c: to
'""" ^
p— C
03 O
^j •r*B
•r- 4->
Q. O-
(C O
o
n~*
^ o
C +J
c:
$- 0
O 0
q-
t^.
l/l O
CO
s- to
4J O
•r- •!—
TJ 4->
c- n3
CU C
Q.-I-
X -Q
CO E
O
CO U
S-
ro CU
r~
CO 4J
o CO
10 in
4-> 3
c in
•t— 4->
c
to as
CO r—
S- Q.
cni—
•r- i—
U. as
ta
7-94-
-------
7.5.3 Industry Output, Employment, and Growth Impacts
One concern with regulation is that the price change brought about
by the implementation of an option might be sufficient to induce a change
in quantity sold, which would, in turn, induce a change in employment in
the industry. It is unlikely that the small price changes attributable
to any of the control options would produce a noticeable change in either
output or employment. Nonetheless, a simplified calculation of employment
impacts is presented below to suggest approximate upper bounds on the
changes in employment and output.
Because benzene is an intermediate good, its demand elasticity is
the same as the elasticity of substitution for benzene in the manufacturing
processes in which it is used. These elasticities can be estimated, but
such an estimation is beyond the scope of this work. However, it has
been shown that the price elasticity for inputs to production rarely
exceeds ,unity. Thus, as a rough estimate of the upper bounds of the
employment and output impacts of an option, changes in employment and
output have been estimated under the assumptions that benzene consumers
raise prices by 0.803 percent, the highest hike for any consumer in any
option, and that the price elasticity is unity. Under the unitary
elasticity assumption, a 0.803 percent rise in price implies a 0.803
percent fall in quantity sold, and under the assumption of fixed
coefficients of production, this implies a 0.803 percent fall in employ-
ment. Thus, given the initial level of employment of 19,860 person-years
computed in Section 7.4.6.1, a reduction in employment of 159 jobs is
estimated as the maximum employment reduction in both industries under
the most expensive option. It must be borne in mind, however, that this
impact does not lead directly to the firing of 159 employees; it is quite
likely that the displaced employees would simply become engaged in other
production activities at the same places of work. Even if all 159 individuals
became immediately unemployed, however, this event could hardly be
characterized as a "major" impact on the economy.
The economic conditions in the petroleum and petrochemical industries
have been a major concern because of the large and continuing rise in
world oil prices which began in 1973. Certainly the enacting of controls
7-95
-------
which serve to further raise costs in these industries should receive
double scrutiny because of this situation. For most industries, one of
the prime concerns in regulation is the effect of the regulation on
industry growth. The options evaluated here could possibly have some
minor effects on industry growth; however, any effects they could have in
this area would be quite inconsequential in comparison to the growth
effects brought about by the increases in world oil prices and the actions
of the government in the area of price regulations and supply allocations
intended to combat these price increases. Furthermore, in the petroleum
and petrochemical industries, unlike most industries in which growth is
one of the primary goals, the primary goals seem now to be efficiency and
conservation. Viewed in this light, those options in which benzene is
recovered or prevented from escaping have a positive benefit. In fact,
if the price of benzene continues to rise relative to the price of control
equipment, it is merely a matter of time until many of the options become
so attractive that firms will voluntarily employ them.
7.5.4 Impacts on Suppliers of Emissions Control Equipment
The primary concern in a regulatory action is usually confined to
the industry being regulated. It is usually assumed that the rest of the
economy is operating efficiently, and is sufficiently flexible to adjust
to the changes in the regulated industry as easily as it adjusts to
"normal" economic changes. However, when the relation of the regulated
industry to some other industry is too close to ignore the adjustment
effects in that other industry, it is necessary to expand the analysis.
In this regard, the assumption made is that the industries supplying and
being supplied by the producers and consumers of benzene are sufficiently
flexible to adjust to the small changes estimated here for the petroleum
and chemical industries, with one possible exception: the industry that
supplies floating roofs for tanks. A requirement for contact internal
floating roofs would increase demand faced by the firms which supply
contact roofs, and the firms that cannot now produce such roofs would
suffer a decrease in demand.
The decrease in demand faced by the sector of the industry that
makes only noncontact roofs would be likely to decrease profits that the
7-96
-------
companies in that sector are able to make. The extent of the reduction
in profit faced by the firm depends on two things: (1) the importance of
benzene tank roofs relative to the overall output mix, and (2) the flexi-
bility of the firm in rechanneling productive capacity intp new lines.
To the extent that benzene tank roofs represent a relatively small component
of total sales, and to the extent that the capital and labor used in
benzene tank roof manufacture can be used effectively in manufacturing
other items, this aspect of the option merely serves to rechannel productive
capacity, rather than cause economic hardship. Under certain circumstances,
this rechannelling is just as equitable as the rechannelling of economic
resources that comes about through the usual operation of the free market
forces: supply, demand, and competition. The main difference is that in
the free market case the economic forces are applied directly by the
market participants shifting their demand from the old product to the new
one with characteristics they prefer; however, because there is no market
for clean air, shifts in market demands cannot be depended on to bring
about economic changes. Instead, the individual demands of the people
are collectively channelled through the government where they reach the
producers through the enactment and enforcement of direct regulation.
The only conditions needed to ensure that the actions of the government
are equitable in the same sense that the operation of market forces is
equitable, is that the article in demand have bonafide benefits over the
existing product, and that the government is correctly interpreting the
demands of the people.
7-97
-------
7.6 REFERENCES FOR CHAPTER 7
1. Synthetic Organic Chemicals, U.S. Production and Sales, 1978.
U.S. International Trade Commission. Washington, D.C. Government
Printing Office, p. 15.
2. Synthetic Organic Chemicals, U.S. Production and Sales, respective
years. U.S. International Trade Commission. Washington, D.C.
Government Printing Office.
3. Chemical Engineering. January 30, 1978. p. 64.
4. Chemical and Engineering News. Vol. 57. June 11, 1979. p. 59.
5. Gunn, Thomas C., and Koon Ling Ring. CEH Marketing Report on Benzene.
Chemical Economics Handbook. Stanford Research Institute, Menlo
Park, California. May 1977. p. 618.5022N.
6. Gunn, Thomas C., and Koon Ling Ring. CEH Marketing Report on Benzene.
Chemical Economics Handbook. Stanford Research Institute, Menlo
Park, California. May 1977. p. 6185021C.
7. Gunn, Thomas, C., and Koon Ling Ring. CEH Marketing Report on
Benzene. Chemical Economics Handbook. Stanford Research Institute,
Menlo Park, California. May 1977. p. 618.5024B.
8. Industrial Chemicals Report. Radian Corporation, Austin, Texas.
1979.
9. Gunn, Thomas C., and Koon Ling Ring. CEH Marketing Report on Benzene.
Chemical Economics Handbook. Stanford Research Institute, Menlo Park,
California. May 1977. p. 618.5022W-Y.
10. Gunn, Thomas C., and Koon Ling Ring. CEH Marketing Report on Benzene.
Chemical Economics Handbook. Stanford Research Institute, Menlo Park,
California. May 1977. p. 618.5022F-^.
11. Average 1976 Unit Market Price: Synthetic Organic Chemicals,
U.S. Production and Sales, 1976. U.S. International Trade Commission.
1977. Total Sales, 1976: Company records.
12. Gunn, Thomas C., and Koon Ling Ring. CEH Marketing Report on Benzene.
Chemical Economics Handbook. Stanford Research Institute, Menlo Park,
California. May 1977. p. 618.5023B-C.
13. Gunn, Thomas C., and Koon Ling Ring. CEH Marketing Report on Benzene.
Chemical Ecnomics Handbook. Stanford Research Institute, Menlo Park,
California. May 1977. p. 618.5023F-G.
14. Standard Corporation Description. Standard and Poor's Corporation,
New York. Continuous update.
7-98
-------
15. Chemical Engineering. January 30, 1978. p. 33.
16. Chemical and Engineering News. April 4, 1977. p. 10.
17. Chemical Marketing Research Assn. Review. May 3, 1977. p. 82.
18. Chemical Age. October 13, 1978. p. 11.
19. Telecon. E.B. Dees, TRW, Inc., to Anthony J. Finizza of Atlantic
Richfield Company, Los Angeles, California. 12 December 1978.
20. Telecon. E.B. Dees, TRW to Paul Fritz, Corpus Christi Petrochemicals,
Houston, Texas. 14 December 1978.
21. Oil and Gas Journal. October 30, 1978. p. 36.
22. Chemical Marketing Reporter. September 18, 1978. p. 81.
23. Guthrie, K.M. Data and Techniques for Preliminary Capital Cost
Estimating. Chemical Engineering, p. 114-142. March 24, 1969.
24. U.S. Environmental Protection Agency. Control of Volatile Organic
Emissions from Petroleum Liquid Storage in External Floating Roof
Tanks. Report No. EPA-450/2-78-047. Research Triangle Park,
North Carolina. December 1978.
25. Telecon. Ailor, D.C., TRW with Larry Oxley, ALTEC. February 27, 1979.
Internal floating roof cost estimates. *'
26. Letter and attachments from Roney, E.W., PETREX, Inc., to D.C. Ailor,
TRW, Inc. February 28, 1979. Features of PETREX Internal Floating
Roofs.
27. Telecon. Houser, G.N., TRW, Inc. with Ken Wilson, Pittsburgh-Des Moines
Steel Company. January 25, 1979. Cost for installing aluminum dome
on external floating-roof tank.
28. Telecon. Ailor, D.C., TRW, Inc., with E. W. Roney, PETREX, Inc.
November 1979. Costs for removing noncontact internal floating
roofs.
29. Telecon. TRW Environmental Engineering Division. Survey conducted
of benzene tank users/owners to determine distances between benzene
storage tanks. July 1979.
30. U.S. Environmental Protection Agency. Guidelines Series, Control of
Volatile Organic Emissions from Storage of Petroleum Liquids in
Fixed-Roof Tanks. EPA-450/277-036 (OAQPS No. 1.2-089). Research
Triangle Park, North Carolina. December 1977.
7-99
-------
31. Telecon. TRW Environmental Engineering Division. Survey conducted
to determine inspection frequency and operating and maintenance
problems occuring with benzene floating-roof tanks.
September 13-18, 1979.
32. U.S. Federal Trade Commission. Quarterly Financial Report for
Manufacturing, Mining, and Trade Corporations, First Quarter 1979.
Washington, D.C., U.S. Government Printing Office, 1979. p. 12, 31,
33.
33. U.S. Federal Trade Commissions. Quarterly Financial Report for
Manufacturing, Mining, and Trade Corporations, First Quarter 1979.
Washington, D.C., U.S. Government Printing Office, 1979. p. 12, 25,
27.
34.
35.
Current Prices of Chemical and Related Materials.
Reporter. 215(26): 46-56.
Chemical Marketing
Gunn, Thomas C., and Koon Ling Ring. CEH Marketing Report on Benzene.
Chemical Economics Handbook. Stanford Research Institute, Menlo
Park, California. May 1977. p. 618.5022F-618.6022S.
36. U.S. Department of Commerce, Bureau of the Census, Annual Survey of
Manufactures, 1976. Industry Profiles (M76(AS)-7). Washington,
D.C., U.S. Government Printing Office, 1977, p. 130.
37. U.S. Department of Commerce, Bureau of Economic Analysis, Business
Statistics, 1977. Washington, D.C., U.S. Government Printing Office,
1977. p. 48.
38. U.S. Department of Labor, Bureau of Labor Statistics. Producer
Prices and Price Indices. U.S. Government Printing Office, January,
February, and March, 1979. Tables 4.
39. Gunn, Thomas, C., and King Ling Ring. CEH Marketing Report on
Benzene. Chemical Economics Handbook. Stanford Research Institute,
Menlo Park, California. May 1977. p. 618.5021C-618.5021D.
7-100
-------
APPENDIX A
EVOLUTION OF THE BACKGROUND INFORMATION DOCUMENT
-------
-------
1—
z
LU
=5
O
O
a
z
o
1— 1
^p
s:
o
LU
z
i— i
Q
Z
=3
O
C3
CJ
CO
LU
I—
LU
O
z
o
1— 1
I—
3
o
LU
.
rH
1
a>
XI
re
H-
c
o
•^'
4->
re
o
o
c
o
.p
u
re
4-
o
ai
i.
3
•P
re
- z
4_>
c
re
*~~ >s
3 (J
Ul C
c a)
O O!
u re
" S-
>> o
re
CL
E
O
O
3
•a:
CJ
z
*
E
re
x:
s-
3
a
c
0
'•P
re
E
£_
0
q-
c
•r-"
S_
O
•p
in
O>
3
D-
O>
S-
^4*
rH
rH
C
O
•r*
•P
U
CO
^^
r"
re
Q.
Q
O
r—
re
u
•r—
E
a>
CJ
c ^c
o •
x •
LU =3
p;
en
rH
„
CM
^_>
Ul
3
Dl
3
O
Z
«l
E
re
f~
t_
3
a
c
o
•p
re
E
S-
O
C
•^
s-
o
+J
Ul
(U
3
a-
at
s-
^-
rH
rH
C.
O
4J
(J
0)
CO
in
•p
U
3
73
o
s.
a.
E
3
^—
Q
5_
4_> >.
OJ C
o_ re
CL
c: E
3 0
CO O
p;
Q^
rH
„
CM
-P
Ul
3
Dl
3
•a:
o
z
*
E
re
^
s.
3
a
c
o
'.|~>
re
E
t.
O
t-
c
•r-
S-
o
-p
in
ai
3
cr
cu
s-
*4*
rH
rH
C
O
•^
^>
u
d)
CO
f
re
u
•i—
E
CD
_f™
U
O
s-
0)
Ul
0)
+J
re
4^
CO
r~" ^)
re c
•P re
in CL
re E
O 0
CJ 0
{^
en
rH
„
CM
+J
in
3
ai
3
ai
3 3
o- o-
re
OJ r—
x: -P
0 in
3 3
Dl D)
3 3
O
Z
M
E
re
x:
s-
3
a
c
o
'.p
re
JL.
0
C
•r—
S-
O
•P
Ul
cu
3
cr
(D
S-
^fr
rH
rH
C
O
•r^
4~>
O
O)
CO
.
^c
,
CO
.
=3
f.
>j
C
re
CL
E
O
o
r—
•C—
0
t^_
f—
3
0
£;
en
rH
„
CM
•P
in
3
ai
3
0
Z
«t
E
re
x:
s-
3
C
O
'•P
re
t.
o
4-
C
•^
S-
o
+3
Ul
ai
3
a-
0)
s-
*st*
rH
rH
C
O
•^
^_)
u
ai
CO
^.
c
re
CL
E
O
CJ
f».
•r—
O
73 re
S- C
re re
73 -f-
C 73
re c:
•p 1-1
to ^
^
cr*
rH
„
CM
•P
in
3
D>
3
4^
o>
^~
^j-
rH
rH
C
O
•1—
.4.3
u
J>»
C?/-x
re i —
o. re
E ••-
O -P
o c
CJ
M
o
u
X ui
H~ *'"'
u
c
c: re
0 S-
•P LU
in
3 C
o re
n: co
S- S-
CL) Ol
•P -P
•P -P
O> O)
^^ ^"
rH rH
rH r-t
C C
O O
•r- *f
•P -P
U U
(U (U
CO CO
o o
•p -p
O> O)
Ul Ul
c: c
o o
CL O-
in in
O) o>
Q£ Q£
•
^c
.
CO
. .
33 U
C
»• 1 — I
>}
C -
re
CL
en en
* •>
^^ 10
rH rH
i. t-
0) Q)
XI -Q
E E
Q) 0)
•P -P
CL Q.
rH
i-t
C
O
•^>
4^
CJ
O)
to
o
••p
O) /->
Ul 73
C (U
O 3
O- C
in -i—
Ci -P
CC. C
O
U
>^
f~
re
Q.
E
O
CJ
r—
•r-
O
r— *
r>»
O)
x:
CO
f^
en
rH
^
CO
CM
i.
QJ
X)
E
O)
4->
CL
O)
co
A-3
-------
TJ
O
3
C
'•P
O
O
r-i
m
a
•a:
Q.
M
(0
'.£
o.
CU
TJ
m
•p-
.c
D-
j_
(D
at
•
rH
C
o
43
u
OJ
CO
o
•p
Response
leum Products
o
•p >>
3
O
C
o
03
•p—
O)
•r-
rd
d.
m
t
o
o
"cO
u
at
o
o .
X CO
x •
„
oo
CM
1
CM
S-
o>
fZ
CU
-p r--
at en
CO rH
X
•p
U)
$_
CJ
to
3
a.
o
o
s-
ai
O)
rH
rH
C
•^
.p
U
at
CO
o
-p
Response
la
cj
€
tates Petrochei
CO
•P (O
U) Q.
CO E
O O
CJ O
f^.
en
rH
en"
CM
i-
0)
es
Q.
01
CO
X
'-p
to
t.
.c
CJ
U)
1-
o
0 ,
-p
•p—
to
-p
m
Q_
"cO
O
°e
tates Petrochei
CO
*&
•P CO
Ul Q.
CO E
O O
O CJ
p^
en
rH
A
CM
CM
i-
Ol
O)
O
1
I-H
^
O
D)
CO
O
JZ
CJ
S-
en
0) C
c°
o in
•P"
in -^
in c
•i- co
E -P
0)
•P
t- in
o >
O) c:
CO CO
^ o
CJ CJ
CO
r>>
en
1 rH
i.
cu s-
o oi
E -Q
Ol E
•P 01
Q. U
CU Ol
CO O
3
cu
D)
O
Q£
c
o
.p
CO
CO
•p
•p-
10
•p-
CO
p—
Q-
imical Company
CJ
£=
P
S-
o
CO
at
CJ?
s-
o
-p
U)
o
>>
s.
3
to
to
a> a>
C 'r-
O -P
CU CO
1— •*-
Q.
O
O
to
to
a*
Amerada H
CO
!*•••
en
rH
*>
I-H
Ol
t~>
O)
O
O)
a
eC
o
u_
c
o
-p
CO
OQ
at
O)
CO - .
S-
3
to
p
^>
J-
3
(O
to
C »P- ^**
0 -P TJ
x: -r- at
O.P— 3
CD-i- C
"cu
J13
O)
O
cu
a
A-4
-------
T3
a>
3
•P
C
0
1 1
S-
c
ro
a.
0
o
<1J
•p
(0
X
V,
O5
c
•r-
5-
Q.
D)
00
O)
ro
O
+j
U)
^_
O
>,
01
s-
in
(/>
(U 0)
C •!-
O -P
a.^-
0) •!—
r— O
ai ra
i — *+—
u
c:
i — i
n
ra
c
o
i ^
(P
c
ro
O
•r-
£_
u
0)
a
^>
^g
,_
ro
OJ
a)
in
O> a)
C •!-
0 -P
f- •!—
a) T-
r— O
a> ro
^^
t—
CO
a.
E
O
o
^_
•r-
0
•a
c
(O
j"l
iH
vi
rH
S-
s-
o
in
4-
O
5^
O
i-
3
in
o> a)
C •!-
0 +J
x: •!—
rSJ *O
Ol ra
i— <«-
^
c
ra
p;
o
0
o
r~"
ro
a)
c
c
o
o
00
r-~
CD
rH
^T
rH
S-
o
o
0
VI
o
«^
a>
Q.
O)
ro
S-
o
U)
,,_
o
-^
a>
S-
3
U)
(A
ai 01
c: -i-
o +->
x: T-
Q.I—
a» -i-
r— U
O) re
i — i—
c
0
s-
T3
C
ro
a>
3
u.
o
TO
ro
O
O
o
oo
CD
rH
^T
rH
S-
O)
E
a>
u
a>
x
I—
VI
c
o
in
3
O
-c
a)
D)
ro
o
U)
^_
o
>,
a)
s-
3
W)
a* a>
C v-
0 +J
x: T-
a> -r-
•— o
^^
r"
rO
Q
0
0
r—
•r—
O
S-
ai
ro
x:
o
oo
CD
rH
^r
rH
S-
E
a>
u
a>
a
3
VI
U)
a)
r—
i.
n)
_C
U
a>
ro
a*
D)
ro
o
-p
in
»
a>
s-
3
U)
in
a) ra
i— M-
^i.
e
ro
E
0
o
at
0
•r—
^
S-
a>
to
in
a>
-P
o
oo
CD
rH
^T
I-H
S-
a>
gs
a>
u
a>
a
i— i
VI
T3
C
ro
r—
U)
1— 1
a>
3
OO
0)
O)
rO
O
•P
U)
tj_
O
>,
a)
s-
3
m
ui
a> 0)
o -P
Q-r—
a> -i-
r— U
Ol ro
r- M-
D>
C
»^
f~
•r-
*t~
a)
T3
c
ro
o
s-
ro
o
oo
CD
rH
M
«*
rH
S-
a>
f*t
E
a>
u
ai
a
^c
'
VI
at
r— •
r-"
>
S-
ro
O
a>
CJ)
ro
O
-P
in
M-
0
>,
a) '.
S-
3
in
U)
a> ai
C •!-
0 -P
O-r-
a> T-
•— o
a> ro
t— <•-
d
c
HH
j^T
i
to
o
C_3
oo
CD
rH
^T
rH
S-
a>
XI
E
a>
0
a>
a
x
r-
V.
ra
c
01
~o
ro
in
ra
a.
0)
0)
ro
O
•p
in
M-
O
>)
a>
S-
3
U)
in
a) a)
C •!-
0 -P
X^ *P~
Q-r-
a> v-
r— U
a> ro
_
ra
s-
-P
C E
a> 3
§2
O -P
s- a>
0 0.
oo
CD
rH
,-jT
rH
i.
a)
XI
E
a>
u
a)
CJ
x
h-
*•
ro
in
a>
T3
O
ai
O)
ro
o
+J
in
H-
O
>,
aj
s_
3
U)
in
a> a>
C •!- /-S
O -P T3
x: T- a>
O-i— 3
ai -i- c
1— O -r-
oj ro -P
1— t- C
o
u
U)
ro
CD
i —
ro
3
•P
ro
O
0)
rO
O.
r—
LU
oo
CD
rH
^T
nH
J-
0^
.O
E
a>
u
ai
a
A-5
-------
\-l. Continued
"•fc
o>
.O
CO
C
o
'•%
LI
O
c
o
•p-
-p
u
CO
,,_
o
01
1
4J
c .
m
4->
r- >>
3 ai
o -P
.C v-
Q.P-
ai -i-
t— O
O) CO
r-t-
Q.
S.
0
o
•r-
a.
a.
in
in
in
in
•p-
s:
•p
U)
i-
LU
00
en
r-l
^T
1—4
s_
01
^
O)
o
0)
a
^
-
a.
a.
•p-
J-
at
OJ
(0
o
in
. O
1
3
in
in
at a»
C •!-
0 -P
.e-p-
Q.P-
O) -P-
p— O
0) CO
r-1-
1 c
o
•p-
cO
5-
0
a.
o
?
00
en
^T
i-i
O)
ps
0)
U
0)
a
X
i—
«
•p*
u
in
co
ai
en
co
S
in
o
1
3
in
in
at a>
C i-
0 -P
j= -p-
Ci-p—
0) •!-
p- O
0) (0
i— *•
•r~
0
Marathon
00
p^.
en
1-1
^T
pH
S-
01
-i
O)
u
0)
a
0)
in
^
in
e '
t
CO
1
O)
D)
2
o
in
O
!>
3
in
in
O) 0)
C f-
0 -P
Q.'r^
O) -P-
P— CJ
i25
p—
CO
U
•p-
01
u
-Q
O
00
en
rH
^r
pH
a>
Q
at
u
0)
a
x
1—
>;
0
in
ca
X
Ol
CO
o
in
«i
O
s-
3
in
m
Q) O)
C -P-
0 -P
Q.'p^
.— U
0) CO
1— 1-
^^
ta
i-
o
o
Monsanto
oo
en
1—4
«jT
rH
S-
0)
s
O)
u
o>
a
\—
-P"
O)
O)
m
01
a>
co
o
in
•I
O
Telephone survey
facilities
>>
co
Q.
E
0
o
Monsanto
00
en
rH
^T
rH
i.
O)
e
0>
u
01
a
§
U)
•p-
3
O
— J
CO
O)
2
o
in
t
o
1"
3
U)
in
O) ai
C -P-
0 -P
-C •!-
Q-t—
a> «P-
r— O
ai co
f— t-
^^
2l*
m
i-
o
(J
Monsanto
00
O>
rH
,*"
'•
i-
01
u
z
c
o
in
S-
01
•o
O)
z
Ol
0)
CO
o
in
..
O
Telephone survey
facilities
«'
Ol
o
Montrose
00
r«~
en
1-1
<-"
1 '
s-
Ol
E
Ol
(J
Ol
a
0)
r~-
'>
•a
t
to
0)
z
storage
q_
O
5
3
in
in
01 Ol
C •!-
0 -P
^= -p-
O.P—
01 •!-
i— O
Ol CO
\r- <*-
in
01
S-
•p
in
•o
t— 1
2
Q.
00
en
rH
P^
s_
Ol
1
Ol
u
01
a
P
, -r-
• in
si -
o
in
1
O
u
01
O)
CO
0
in
•i
o
Telephone survey
facilities
] continued)
•*t^
__
1
o
3:
Quintana-
00
on
s
s-
Ol
•i
01
u
O)
a
A-6
-------
-a
O)
c
•i—
c
o
o
1— 1
1
CO
XI
ro
«•-
O
•r"
p
rO
U
O
c
o
•r-
. *
U
ra
<4-
o
ai
s_
•p
ro
Z
*pl
ro
3' U
in c
c a>
0 O5
O ro
*> S-
>>0
1
0
o
CO
-p
(0
a
i
i — i
„
in
•r—
s_
i-
CU
O5
ro
O
-p
in
«i
O
>>
O)
S-
3
in
CO CO
C •!—
O -P
x: «i—
a.1—
CO *r~
i — U
cu ra
1 — f—
"re
u
CO
x:
o
-o
"o
x:
CO
00
CD
l-H
s:
S-
O)
XI
E
cu
u
at
1=1
<£
1
#1
S-
m
E
in
O)
en
CO
O)
ro
O
p
'1
O
>.
O)
>
S.
in
U)
C •!—
o P
OLI—
O> •!-
i — O
ai ro
I— t-
in
ra
o
'E
CO
x:
c
o
u
x>
3
ceL
00
l-H
j"
i-
a>
E
CO
o
CO
0
1
a
#*
>>
."Jl.
o
0)
s-
(0
ra
"co
a
a>
OJ
ra
S-
o
tn
l^_
o
^J
CO
S-
m
in
CO CU
c .««-
O P
x: T-
0.1—
CO .,—
^~ u
O) ro
h-
^>
CO
E •
'ra
x:
o
CO
O5
ro
S-
o
^>
U)
,
cu
i-
3
V)
in
CO CU
C •!-
0 -P
Q.Z-
Ol -r-
i — U
CO rO
r- <»-
(J
C
»— I
O
u
01
c
CO
00
CT>
rH
rH
S-
co
E
CO
u
CO
Q
.
OH
al
««
tn
ro
"co
3
C
CO
Q-
Ol
ra
s-
o
in
<4-
O
>^
CO
>
S-
Ul
in
CU CO
C T-
0 P
Q.I—
0) -i-
•— u
CO ro
r— "*-
au
•a
• r-
XI
(O
0
c
o
c:
=j
oo
r-H
S
!-
CO
E
CO
o
CO
Q.
„
•a
c
ro
in
i— i
CO
i—
•r-
>
r—
CU
z
CO
en
ro
S-
o
-p
in
tt_
o
>.
01
s-
3
in
in
CU CO
c= •.-
0 -P
x: •!-
Q.1—
Q) «r-
»~— o
ai ro
^~
cu
cu
to
to
=3
co
rH
3
S-
co
XI
E
CO
o
CO
Q
1—
«>
ra
>
cu
c
cu
C3
CO
D5
ra
o
p
in
tf_
o
>j
O)
>
^
in
in
CO CO
C *r—
0 -P
x: -r-
CXi—
O) -i—
i — O
CO ra
1— <*-
r—
CO
CO
-p
to
to
00
CD
rH
S
S-
XI
E
CO
u
cu
1=1
o
z
«>
E
ra
x:
s-
3
a
s-
01
^^
p
cu
^—
Q.
3
- i
0
I™1,
1"™
o
^-
^J.
rH
rH
C
O
•r™
•P
(J
CO
to .
^-
c
ra
i
o
o
'ro
(J
•i—
0)
x:
o
c <:
0 •
x to
X •
UJ =3
00
rH
CM
S_
co
XI
p!
cu
u
01
en
o
21
#>
E
ra
x:
s-
3
S-
0)
^J
p
cu
r~
Q.
3
^
O
r—
^—
O
1-
5;
rH
rH
C
0
•1—
-p
o
CO
to
in
-P
u
3
T3
0
Q-
E
3
CO
O
p >>
CO C
a. ro
a.
C E
3 O
to o
oo
CD
rH
CM
CU
£
01
u
cu
a
c^
z
•*
E
ra
x:
s_
3
0
t-
O)
p
^>
CO
r—
a.
3
5
o
r—
r^
O
*f—
«^.
rH '
rH
^^
C "O
O CO
•r- 3
•P C
(J *^*
to c
o
u
i —
ro
U
•r-
CO
x:
o
0
•p
CO
Q-
tn
cu
4^
rO
to
r- >>
ro c:
p ro
in Q.
ra E
0 0
o O
co
CD
rH
r,
t~~
CM
5-
0)
XI
E
CO
u
CO
a
A-7
-------
•o
Conti nue
r-J
1
O)
JQ
re
r™
c
o
•r—
re
u
o
C!
0
*r-
U
re
o
01
3
1
•P
re
10 C
c 01
o o>
u re
re
o
01
re
TO
a
o o
(O (0
^ ^*
Jw J—
S- S-
3 3
a o
s- s-
01 ai
•p -P
ai ai
• *
3 3
O O
i — r—
0 0
(|_ (fa
rH rH
C C
0 0
•r— •!—
-P -P
O U
a) o>
CO CO
M -o
>— i i —
ai
Chevron U.S.A.,
Atlantic Richfi
Company
CO CO
en en
rH i-H
CM CM
i. t-
Ol Ol
ja ja
§ i
S S
a 1=1
o
z
re
>-•
s-
3
a
1
ai
a.
o
r—
"o
"i™
rH
i-H
C
O
•i—
-P
U
ai
CO
u
0)
o
re
s-
3
a
i.
01
•p
4J
ai
a.
o
"o
*JJ
«*'
i-H
rH
C
o
'.p
u
0)'
CO
£
re
Standard Oil Co
(Indiana)
00
en
rH
CM
i.
Ol
_Q
Ol
o •
01
x:
K—
>i
•r-
O
in
re
X
01
letter
rH
rH
O
•P
O
O)
IA
O
•P •
Ol
in
O
0.
01
cc
i
o
o
•r-
O
re re
•a T-
c -a
re c
^j H~4
CO *-^
en
en
rH
O
i-H
2>)
re
c.
re
^T
^^
^-
o
p
ai
"3
LL.
4->
in
o
u
o
M-
>,
>
«.
ephone si
:ormation
Ol C
i — <-
(A
01
C
i
Pittsburgh- Des
Steel Company
en
en
i-H
r-T
i-H
^>
re
re
•~3
_
C
C
re
'C
o
43
re
s
t_
o
q-
(=
in
O
u
i.
o
1-
>,
t.
in
01
O
a.
ai
'oi
i—
re
T3
re
s-
01
•r-
O
oa
C IA
3 -^
0 S-
cn
en
i-H
«s
rH
i-H
^>
re
c
re
•-3
^»
O
A
in
ai
O *oi
D)
• C
re >
3 3
in in
„ a-
0 0
a. a.
ai a>
"a> lii
r— r~
i—
U
•r—
ai
o
^ 1
i_j fll
Tank Service, ]
Western Petrol i
Services
en en
en en
rH rH
rH~ P?
^^ ?*
re re
c c
re re
JK£
i—
c
0
p
o
letter
,3-
rH
C.
O
4J
(J
ai
CO
o
ai
in
o
a.
IA
a>
5^j
re
a.
1
u
Exxon Chemical
U.S.A.
en
S
r-H
^^
^*
re
c
re
0
•a:
o
o
u
in
•r-
(J
re
LL.
c
re
CO
i-
0)
Ol
^j-
rH
C
O
u
O)
CO
o
Ol
IA
c
o
Q.
in
01
u
I-H
CO
O
Ol
o
en
S
•t
en
rH
^
^*
re
c
re
•~3
-------
73
3
C
•r—
•P
O
CJ
rH
«=c
OJ
_a
03
r-
c.
o
4->
03
O
O
1
r"
' »f—
.p
U
03
n
o
OJ
3
.p
(0
.p
c
03
r— >\
3 (J
C
C OJ
O CJ)
U 03
*« S—
^» Q
f*J
<0
o.
e
o
o
OJ
-p
03
Q
X" X
1- r-
* *\
c . c
0 0
in in
3 3
O O
i- i.
OJ OJ
•p -p
OJ OJ
r— r— •
^. ^.
rH rH
rH rH
C C
O O
"-P '-P
(J «J
OJ OJ
to co
o o
-p -p
OJ 0)
in in
c c
o o
O. Q.
w in
oj oj
OS O£
^-
^1 c
c (0
03 a.
Q. E
E 0
f )
f—
i"™ *i~~
•r- O
0
i — OJ
3 .C
CD to
cn cn
r-- r-.
cn cn
rH rH
«• »»
cn cn
CM CJ
^^ >.
J_ J_
03 03
3 3
C C
(0 03
X
•r—
•P
in
•r—
S-
^*
O
in
3
Q.
S-
o
o
^_
0)
[ %
OJ
^.
rH
rH
C
o
u
OJ
to
o
•p
OJ
in
o
a.
in
OJ
ae.
E
OJ
o
s_
.p
OJ
a.
in
OJ
•P
[^
to
*~~ 5*>
03 £=
•P 03
in o.
03 E
O O
O CJ
f^.
cn
rH
A
0
CO
>>
t?
^3
C
03
X O
h"** ?^
£ i
OJ .C
OJ S-
3 3
to a
OJ
•f"
re
•p
c
OJ
in
s-
OJ
c
03
Q.
O
O
-p -p
in 3
> CJl
c
c: -P
03 OJ
i — OJ
a. s:
^
3
OJ
'o
•p
OJ (J
Q. C
I — I
in
Q. -
X
r- OJ
v^ -P
-C OJ
o_ a.
^^.
cn cn
i — i r^»
cn
rH
rH
^» OO
t- rH
03
S- U
OJ 03
LL. s:
_1
H- 1
•.
OJ
r^
r—
•r-
C
03
r~-
Q.
in
c
o -^
•r- C
in 03
in 1—
•r-
E -P
.. uj in
CL. OJ I—
UJ C
OJ <4-
0 N 0
•P C O
OJ rv'
73 CO
oj cn
+J M- C
•P O ••-
•I— +J
E -P 03
£1 C O
3 OJ i—
in E U-
OJ
•P S- 03
S- 3
O ,
0) C
03 03
u a.
•r- E
-C O
CJ CJ
cn
r--
cn
rH
OJ
C
3
"~3
O
^^
«l
(0
r"
i-
3
a
c
o
-p
(O
S-
o
<*-
c
•r-
j_
O
4—
-P
in
OJ
cr
OJ
3
rH
C
O
•f—
•p
u
a>
CO
1— >J
03 C.
u m
•r— O_
E E
OJ O
-C CJ
CJ
r~~
>5 re
F— C
S- •!-
aj E
> s-
oj ai
03 1—
cn
cn
rH
f.
00
i-H
>^
f_
3
•-3
O
3^
(0
s-
3
a
c
o
'-p
03
E
S-
O
t{_
r"
• r—
S-
0
4—
-p
in
OJ
3
CT
OJ
rH
rH
=
O
•r—
0
^
1—
3
•~3
CJ
z
03
-C
S- '
3
°
C
0
'J3
re
E
s- -
O
4_
C
•r—
S-
o
M—
4^
in
OJ
g-
OJ
^
rH
C
O
•1—
u
O)
CO
.
a.
S-
o
CJ
^~*
re
c:
•r—
E
S-
aj
r—
X
1—
CD
cn
cn
rH
•t
00
T-H
>J
f__
3
*^
o
z.
re
.c
s-
3
Q
C
O
-p
re
E
S-
o
4-
C
•F—
S-
o
4—
4J
in
OJ
3 '
O"
OJ
rH
rH
C
O
•r—
CJ
OJ
CO
•*-
in
re
c
•I—
E
S-
OJ
r—
73
OJ
•P
• f—
C
O
•P U
OJ C
Q- t-l
cn
cn
rH
*1
00
rH
>>j
i™~*
3
*~3
CJ
Z.
re
x:
S-
3
a
c
o
•p
re
E
t.
0
4-
c
•r—
S-
o
4-
•P
If)
OJ
3
cr
OJ
rH
rH
C
O
•r—
•P
U
OJ
CO
•
a.
s.
0
0
in
in
OJ
3:
re
73
re
OJ
E
cn
cn
r-H
*i
00
rH
>>
f—.
3
'TJ
r~3
Z
*•
OJ
C
C
o
>J
re
CO
S-
OJ
•p
•p
OJ
^~
^J-
rH
rH
C
o
•f—
•p
u
OJ
CO
o
4->
in 73
C OJ
O 3
a. c
in •r-
0) 4->
O£ C
o
o
OJ
o
•r*
S-
OJ
CO
r—
re
c
•p—
E
^
OJ
t—
>J
o re
73 a.
S- E
0 O
CD O
cn
t^»
cn
rH
•*
^t*
C\J
•P
-------
•1. Concluded
•c
0)
JO
CO
r-
C
o
p
u
o
Q
• |K
p
u
CO
0
0)
c
m
r- >»
u> c
C O>
O O)
U CO
- S-
(0
i-
o
u
0)
CO
Q
•
t-H
o
DJ
CO
u
•r—
o
s-
O)
^^
0)
r~
«*
Response to Section 11'
e-
o
u
u>
c
X
C3
S
en
rH
rC
S-
0)
•i
(U
o.
0)
to
x:
«»
j^
o
o
s-
j3
CO
0)
to
O)
* ^
0>
r~
*t-
Response to Section 11
.
(A
"cO
1
,0*
PetroUnited 1
Inc.
en
Cn
rH
O
CM
i.
0)
E
0)
Q.
0)
to
•"3
*
0)
O)
•a
•r"
S-
ja
-O
O
£'•
O)
'^
O)
'~
**•
Response to Section 11
o
I—
Amerada Hess
(Confident!' a
en
f*^
en
rH
r-T
rH
t-
O>
•g
U
o
o
_c
Ol
•f"
O)
i
c
o
»^
CO
c
0)
NAPCTAC Committee Pres
D)
C
'£
(0
0)
u
•I—
0.
Q.
LU
O
•S
^
r**«
rH
^~
'tl
O
•t
o
O)
0)
o
c
CO
to
c
CO
0)
•r™
E
01
•P
Telephone survey to de
inspection procedure
c
o
•r-
r—
"o -p
0- U
•r~
•r- P
«C 01
•r*
o a
O)
O) t—
•i- O
c c
CO O
to o
§
en
rH
«T
*^
i"
•«
u
*t
CO
C ~
O)
•a
CO
w
CO
a.
T3
O)
S-
CT
a)
2-
o
r- -P
>,
0)
c
*r"
1-
O)
r_
&.
c
a)
0
|
u
I
rH
io"
O)
C.
•-3
^
£
•r—
i-
P
CO
TJ
a>
s-
cr
at
t-
O)
E
P
O
3 C
(0
0)
It
O) 01
Q) O
r- P
.
O
u
"cO
u
•r-
E
O)
0
o
00
en
rH
us"
0)
C
3
x:
r-
«
3
S-
t
O
a.
TJ
O)
£-
•^
ey
O)
S-
a>
E
•r-
O
> u>
3 C
ui CO
-p
0)
§t
ft
U
t— <
o
u
CO
X
at
§
en
^
^
to
a>
c
3
•-3
U)
o
|
c o>
o c=
•^ «r*
P P
U' CO
0) o
Q-i —
U) *^
Survey to determine ir
procedure on internal
roof tanks
&
^
3 0
£- P
-------
APPENDIX B
INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS
-------
-------
CO
^
H-l
•j.
O£
LU
. .
cn
o
o
o
c ro CD
•I-.-P r—
S- C ^f
ro co i****
Q. E CO
co c
J- O Q£
o_ s- u_
S- > CD
o c co
4— ' ' ' *^*s
(A C (A
CO O -P
C -r- C
•r- -P OJ
r- U E
CO , ro
3 S- -P
cu o oo
.p
>> re -P
U r- U
c: 3 ro
CO OJ D.
OJ CU E
«=C Q£ 1-1
_
>
0 S-
_^_ o
O -P
3
CO -P
xj ro
.p
i — IA
^ CO
•^ ,J
in
"O in
S- *1—
"O IA
c ro
ro
•p "
j_ ^
u
•r- C
3 •«- •
•P (A
E U TD
O CO S-
s- to ro
4- "O
« C
w i — ro
CO -P
> s- in
•i- CO
•P -P OJ
ro a. e
c ro ••-
«_<*•!«
CO 0 0
-P O.
r— C O
ro •!- j-
0
S- CO S-
0 N 0
•P «i— 4—
ro s-
i — ro >>
3 E -P
OJ E •«-
CO 3 S-
s- in o
JZ
CO CO -P
x: s- 3
J— ftf *O
•o
ro 3
>>•*->
- s- ro
c o -P
o -p in
•i- ro
•P r— CO
Q. 3 x:
•r- OJ-P
i. CO
u t- -a
in c .
co co re >>
•ax: -P
•P m v-
co s-
-a 4- > o
c o -p- x:
3 -P -P
o co ro 3
s_ in c ro
OJ O S-
-^ Q. cu >>
o s- -P s-
ro 3 r- o
CD Q. ro -P
.
r—
1
OJ
CO T3
S- CO
"O
>
S- 3s T3
3
a. in in
LU C
O CO
s- -r-.x:
CO -P -P
x: ro
-p •— c
O 3 •!-
i a>
s- DJX:
CO -i- -P •
f~ t/) (A
+-> >i CO
o in XJ >
r~ *r~
o ro -a -P
•P m co ro
O -P C
a. a. o s-
•r~ O CO CO
x: s- 4- -P
tA O-4- i—
c ro ro
o -a
•r- C >> >,
•P ro r— s-
ro -P o
r— in C -P
ai c re ro
S- O O i—
•r- •!— 3
CO -P 4- OJ
x: o -r- co
f— ro c s-
^^
CO
+J
u
CO
ro
*>^
-p •
in co
•a s-
C CO
•i— -P
a.
co ro
x: x:
•P O
c c
•p -a
•r- CO
3t -P
tA
in T-
CO i—
o
S- CO
3 J-
o ro
CA
e emission
Iternatives
x: ro
c o
ro -p
ro
•P OJ
IA CO
3 S-
T3
C CO
'*" -P
CO
x: £
CO •
x: in
•P CO
>
>s»r~
XJ -P
ro
•o c
CO S-
•P CO
0 -P
CO i—
4- re
4-
ro >>
S-
>> o
S- -P
-P re
in i —
3 3
•O OJ
C CO
I-H J_
^^
^
o
ro
"3
OJ
CO
j_
CO > c
XJ 0
•r-
•o -p
CO U
-P CO
u oo
4- CO
ro
CA CO
CO -P
•i- O-
-p ro
•r- x:
•i—
0 C
ro •!-
-a
"O CO
C N
ro •!-
in ro
CO E
(A E
in 3
CO (A
o
O CO
S- S-
o. ro
o in
4- >
'o '-p
co ro
0- C
in s-
01
CO -P
x: i—
p~ ro
in
CO
•o «r-
O) -P
•p ro
o c
CO S-
4- CO
4- -P
ro i—
ro
in
CO >>
(A 5-
(A O
CO -P
u ro
O r—
S- 3
Q. OJ
CO
u s-
•I—
4- O)
•r- x:
0 -P
CO
Q. >>
CO
c x: -o
CO O CO
N -P- XI
c x: ••-
co 5 s-
x> u
>
o ro xj s-
r— S- CO
O O T3 -P
C -P CO Q.
x: IA > ro
u cu x:
CO E T- 0
-POX:
J- 0 C
r- 4_ « -i-
o
S- IA d ,
+j c co T-
c. o
x: E ro c
i— co x: -1-
IA
0)
*i— .
OJ
o
o
c
f~
u
cu
•p
r—
0
S-
•p
c
o
o
CO
r—
XJ
ro
o
•r—
r—
CL
O.
r—
^•)
ro cu
0 S-
ro
ro ~&
S- CO
cu -o
IA •!-
in
CO C
x: o
r- 0
•
c
o
•r-
•p
o
ro
CO
x:
•P
o
-P
IA
CO
>
•1—
-p
ro
c
S-
CO
-p
r~—
-p -o
•r- O)
•P 4- C
ro o -r-
C r—
J- CM >P
CO P- 3
•P P- O
r—
ro c co
0 S-
co -p- ro
x: -P
p- O UJ
CU CO
to OJ
ro
in S- -p
CU CO C
x: -a ro
u c >
ro 3 -a
o ro
5- (A IA
a. c T-
O- O T3
ro -P-
IA "O
£42 g.
0 E
4^ cu in
ro co
r- CO OJ
3 c ro
OJ CO -P
CU N C
s- c ro
cu >
CO XJ T3
> ro
•p- OJ
-P C S-
ro -P- -P-
C -P CO
s- re x:
O i — -P
•P 3
P- OJT3
CO C
co x: -p-
-a -P r—
-p
p~* "O 3
O C O
s- ro
•P CO
c in S-
o co ro
o >
•p- in
CO -P CO
> co >
•r- C -r-
•P S- -P
ro cu ro
C -P C
S- r— S-
01 ro co
•p -P
•- >>•—
ro s- ro
o
CO -P S-
x: ro ro
3 3
OJ O
. CU T-
IA S- -P
CU 5-
o cu ro
•p- x: a.
> -p
co a>
•a >> IA
XJ CO
*o ~a -P
S- CO
-P 5- OJ
C •!- C
0 3 >P-
o O--P
CO O
CU S- CO
> 1—
•r- CU CU
•P XJ IA
ro
C T3 S-
S- P- 0 .
a> 3 4- •*
•P 0
P- O (A 5-
cC C CO
x: o -P
O (A O.
•p- ro ro
. x: co x:
xj 3 s- o
•o
CO
3
C
•p"
•P
C
O
u
\*t*S
B-3
-------
•^J
at
3
•r*
•P
o
O
,
H
1
CD
at
,
XI
re
r-
-p
at
E
O
o
a
c
o
"43
re
E
O
4-
C
T3
3
O
s-
cn
o
re
CO
at
x:
•P
c
•r—
x:
•r—
C
o
•f—
-p
re
o
0
_j
O)r—
c re
•r- -P
£- C
re at
O. E
at <=
s- o
a. s-
•i—
s- >
o c
4- LU
U) C
at o
C -l-
•r- -P
r— U
at >
C3 O
>>1c
C 3
at en
en at
s-
4J -o -p • at u
at >•!—> O • •!-
o at i — •!- T- «i- u i — r — •>
ON ceaieato-pT-s-
•i— • c • at S— "o S- N •*— c s« at
•P OI CM EOI OJT--POQ.-P
re x» • c -P at 3 x:
re o e -I- • c >EO
at T- ex: o re i*- -P x c
W-P at-ps-re rere*r-
>x:at oi4--p cats- -o
•r-uto o)oo.4-re-pai'oe
4J re s- re o Q.-P c 3
tO O ** Q) W f"! r^ tO ^~ *0 O
s-s-tn >+Jowrex:re 4-
reo. T3O -P-PO W
a. Q. i- reatcu-r- atatat
E re at 4— *r- re o <— <— ^_ r%
O -P "OH- Q.re-r-4J3
«J>>a. CO)T3EO WO
t™ (0 fO ^— *^~ T3 H™ • C5 W
wox: 03 ateor— i—
-4^0 i— T- o u x: 3 ore
^s (0 fO ^~ ^~ *i~~ ~|~i* C? V)
O I— C -r-0 E 4-+J-PC
c3-i- ucatoo oere
at o T" o *"** c e at re re o
O)O)-a 4-U O'i-XS O-r—
4- xjc atuuureur^
+j «r- re i— x: c •!-•!- 3
+JO) O) O-P-i- -E+JT3T3
4-rex) x:«s- >>oeoc
oc +J r— -P 4- »s-cais-ro
t-c recot— +JO-PQ.
catre 4--I-O atwuo <*
O-PU OU(JC>30tO.S- •
•i-r- o oatTSi atr^.
wrew >>w4-T-i— COD3E
W C S- 0-P T-T-C3W
3WO re« O-r— O-r-WCZ
WOW E-Pr— S-S-X:W3OT-
•i-«i-W 3r— O1O-P+J r— U-P
•os-'i- wre>wc ato o
reE aiatatoox:ci=at
^C ^ 0) ^C *"" r— • "O CJ -P H™ *r™ *n" t/5
• •
re xi.
0
i -'e at
r— I— O r—
re i — re c X2
> re -P o re
at -I- e o e
u at at o
at -r- E w
> 4- c -a re
•r- at o c at
-P c s- re s-
re at T-
s- xi > -x:
re c r— u
o. at at re re
EX: -r- at •
o -P at o at
u w o 4- >
4- S- W O •!-
w o at -p
> « w re
>> c -a x: 4-> e
u o re -P u s.
c •!- i— at at
at -P T3 re 4- -P
o> re e at 4- r—
•a: 3 re x: at re
S-
at
4- -p
o a.
re
w x:
E o
at c
-p v-
c w
•r- O)
u
w s-
u o
re w
•r- O
£S
re -r-
E E
•i- at
^re
at 3
5.|
W •!-
C
O
o
x:
w
W
w
•p
u
at
c LU
re re
re w .
•P CM
c .
re
at -
i— IXt
r— 5-
o at
s- -P
•p o.
c re
o x:
o o
at e
•r™
•P TJ
re at
c w
s- w
l— W
*5
4-
O
W
-p
u
re i—
O. O. C
E o
•a
at
c
o
o
at -p at
S- T3 O 73
•r- o at •!-
01
c a*
UJ 5-
re o
j:: +j
•P T3
>» at
at i— u
W -P 3
o o ~o
x: at at
-P s- s-
u
re
a.
s-
re
at -o
re -a
at
w +J
+J 3
U X)
re *i—
o. s-
E -P
•O
s.
w re
+J -o
c c
re re
-P -P
3 W
i—
i— S
o at
a. c
u re
x:
-P
-p •!—
re »
x:
-P re
•!-•
W TJ
at at
01 E
c
re w
x: 3
u o
o
re •!-
s- at o
re xi re
•r- -I- C O)
s- >>r— s-
•i— xi re re •
u u > c
at 4-> T- o
a, 3 w at T-
w o >»x: -P
x> x: -P u
4- re a. 3
o c -a
+j at •!— at
w x: x: s-
r- D)-P S-
at 3 3 w
> o -a u -i-
a.
re
o_ O-PI— xi re o-p
B-4
-------
•o
OJ
T3
o
r—
o
o
r-i
QQ
0)
(0
l_^
(^
-p
c
OJ
3
Q
ttt
c.
^
, ^
rd
5-
O '
H- ,
.
•a
c;
3
o
o
_*:
u
rd
CQ
OJ
4->
c
•^™
.c
4^
5
0
•r-
rrt
*o
CJ
y Guidelines for Preparing
atory Action Environmental
t Statements (39 FR 37419)
O r— 0
C 3 (0
OJ O5 O.
O) CU E
(J CU CU
•i- -a > i—
r- QJ • OJ
qZ . ' O
in •*— co *~~ S-
r— «p O -P
O C "O S- C
s- aj e -P o
-p TJ <0 C U
C -r- O
o •> u cu
u cu oo >
5_ . CU •!— .
i — rd CO > -P
fd *f" rd
3 Oi •> -P C
-a cu co rd s-
•t- > . C . 0)
> T- CM S- OO -P •
•i— .p . OJ • r— LO
•Q fd CO 4-> CD rd •
C C r— CO
•r- S- in rd c OJ
-p o OJ -r- -p o
x: i— -i— x: -P *i—
.p rd -P .p U 4- -P
u aj o u
4- >> CU 4- CO CU
o s- co o -p co
o u
i/) .p ». w CD ro •>
•P ro CD -P Q.CO
Of— OS- E ,
rd 3 S- rO CU T- S-
O. O5 CU O.-P CU
E OJ -P E 0. C -P
v- s- o- -r- ro o a.
rd f™ •!-• rd
r— cu x: s- o -P x:
ro x: o cu 0.0
-p -P -PC E
c C ro •!- 3 c
cu -P T- 3 in •!-
E CU "O C
c cu >> >> cu o -a
O E i — S- M- O OJ
S- "O •!- ^l-
> .p -i- C -P CT-r-
C -P O C £- -P
cu "O ro o rd eu c
cu -P cu 3 c rd
S- in ••- U) CT CU 3
OJ 3 ' — CU".
+j OJ 3 x: S- x: in
O X> O" 1— rd h- T-
in
OJ
-P I -P
U 4- T> C rd T3
OJ 1 O ' — O XS C
s-x -p3OS-rd-p
•i- cu c x: o cu o
•a ocnu4-oErd
CS--I-3 OrOOJQ.
•r-0-POT3 Xr-E
LL.OS-S-COJX>'i—
O) 3 XJ rO O O
s_ -a -a ••- -P s- >,
ro .cuinc-prOD-S-
U) S- -P rO O-X: (0
w+j C-PO-PCT3
4J +Ju>>rOO)"a oc
CJ O rd S- -P rd >>•!— O
<0 rd Q. 0 3 3 O>-P 0
CL O.E-P"— OJCUO30J
e S-^rOi— cx:r— i— in
v- -r- T30 -P 0 r-
T3 C Q. rd C O rd
>, >, oj ro c x: a.
C C o e o >»-^- o cu
ro (D3 -i-Xl CUS-XJ
TD T3 TD "4- -P -P OJ
c ccajT'+ii — X:TJ
0 0-r-t— U33i— -Pt—
O O Q-CUOinOO3
CU CU S- E O-X> CU S- C O
CO coorOinrOS--Pr03:
«
XJ
Other considerations.
*
*!:*"
E
m cu o oj
+j o in x:
O C r— 4->
(0 (0 >> ro
Q. 0 T3 OJ
£ •!- rd OJ N
•r- 4- CU S- •!-
•r- S- CU rd E
x: c r- s- -I-
-P D5 ro ro T3 c
C •!- OJ •!-
ro in x: in "O E
OJ O -P -i-
x: oj •!- o o o
x: x: rd > -p
-O -P 3: D. rd
C EC
rd 4- in •>- OJ cu
OS- XJ J^
r— O CU rd
rd c -P in -p -P
+J 0 0 S- 0
c -r- rd cu c cu
CU O) Ll_ > C XI
E in TD rd
C 3 rd 0 C
O O • rd
s- in CD r- x: o
•i- •<- rd O
> -o s- •!- T- x:
c cu -P x: o
CU fO -P C 3 -i-
a. cu x:
OJ T3 ro -P in 3
in c x: o -P
S- ro 0 a. O in
OJ rO Q.
> m c at a. cu
-o oj •<- x: E +J
(0 > -P -r- in
r— -P C 4- OJ >>
ro ro 3 O in c
•r- C O S- rO
-P S- 4- OJ CU
c aj E > x:
cu -P aj o ~a -P
-P i— X) in ro ••-
0 rd 3
O. C OJ OJ
>, ro -P w Di
oj s- o ro o c
x: o c x: o
rd O E <0
4- i— ro •!-
0 3 Q-i— • ~
>, OJ T- OJ OJ
S? S- 0 -r- •!-
(0 x: -P 4- 4-
E CO O •«-•!-
EX: ro -p -P -P
3 -P cu in c c
0) -I- OJ OJ
4- 4- x -a -a
,
S-
Adverse impacts which cannot
be avoided should a regulate
alternative be implemented.
'
fd
*
-0 rH
0) •
> VD
r™™ •
o co
>
c c
•i— O
• ^
OJ -P
XJ 0
OJ
•a co
3 -
o co
x: aj
0 -P
•r- O.
x: ro
3 x:
o
in
OJ C
O •!-
3 T3
O OJ
in in
OJ in
S- 3
O
OJ in
XJ ^O
ro
> OJ
CU i-
•i- (0
S-
-P C
OJ O
S- •!-
S- -P
•r- U
ro
•a
C T3
(0 OJ
in
cu o
i— O-
XI 0
•r- S-
w o.
S-
OJ OJ
> -^
aj -P
s-
S- C
1— 1 -r-
aj x: ~a
i — O r—
Irreversible and irretrievab
commitments of resources whi
would be involved with the
regulatory alternatives shoi
one be implemented.
*
.Q
B-5
-------
-------
APPENDIX C
EMISSION SOURCE TEST DATA
-BENZENE STORAGE TANKS-
-------
-------
APPENDIX C - EMISSION SOURCE TEST DATA
C.I INTRODUCTION
This appendix describes the emissions source test data obtained
prior to.and during the development of the Background Information Docu-
ment (BID). The facilities tested are described, the test methods used
are identified, and the data obtained presented.
C.2 ESTIMATING EMISSIONS FROM FLOATING-ROOF TANKS
The emissions from external and internal floating-roof tanks storing
benzene were estimated in the BID using equations developed for EPA by the
Chicago Bridge and Iron Company (CBI). This section summarizes the test
methods, test results, and conclusions from this study.
C.2.1 Description of Test Facility
The benzene emissions test program was performed in a test tank at
CBI's research facility in Plainfield, Illinois. The test tank was
20 feet in diameter and had a 9-foot shell height (see Figure C-l). The
lower 5'-3" of the tank shell was provided with a heating/cooling jacket
through which a heated or cooled water/ethylene glycol mixture was
continuously circulated to control the product temperature.
The effect of wind blowing across the open top of a floating-roof
tank was simulated by means of a blower connected to the tank by either a
30-inch or 12-inch diameter duct. An inlet plenum with rectangular
openings was used to distribute the air entering the test tank shell.
This air exited from the tank through a similar plenum into a 30-inch
diameter exit duct. The 12-inch diameter air inlet duct was used for the
internal floating roof simulation tests, and the 30-inch diameter inlet
duct was used for the external floating roof simulation tests (which
required larger air flow rates). While one size of inlet duct was in
use, the other size was always closed.
C-3
-------
o
(O
I
o
4J
(O
CD
in
-------
C.Z.I.I Principal Instrumentation. The principal instrumentation
consisted of the following:
1. The air speed in the inlet duct was measured with a Flow Technology,
Inc., air velometer, Model No. FTP-16H2000-GJS-12.
2. The total hydrocarbon concentrations were measured with Beckman
Instruments, Inc., Model 400, total hydrocarbon analyzers. Two
instruments were used, one for the inlet and one for the outlet.
3. The airborne benzene concentration at the test facility was
measured with an HNU Systems, Inc., portable analyzer, Model
PI 101.
4. The local barometric pressure was measured with a Fortin,
Model 453, mercury barometer.
5. During unmanned periods (nights and weekends) the barometric
pressure was measured with a Taylor Instruments, aneroid baro-
meter, Weather-Hawk Stormoscope Barometex No. 6450.
6. The temperatures were measured with copper/constantan thermo-
couples and recorded with a multipoint potentiometer, Doric
Scientific Corp., Digitrend, Model 210.
C.2.1.1.1 Analyzer calibration. Calibration gas mixtures were
provided by Matheson Gas Products Company for the purpose of calibrating
both the total hydrocarbon analyzers and the portable analyzer. Gas
mixtures of three different benzene concentrations in ultra zero air were
used:
0.894 ppmv
8.98 ppmv
88.6 ppmv
The inlet air analyzer and the portable analyzer were routinely
calibrated with the 0.894 ppmv benzene calibration gas. The outlet air
analyzer was calibrated with the gas mixture closest to the concentration
currently being measured by the analyzer. Both total hydrocarbon analyzers
were calibrated at the beginning of each 8-hour shift, and the portable
analyzer was calibrated at least twice a week.
C.2.1.2 Product Description. The benzene used during the testing
program was Nitration Grade Benzene as defined in ASTM-D-835-77.
C-5
-------
C.2.2 Test Method
The testing was done in three phases, each using a different type of
floating roof. Phase I used a contact-type internal floating roof.
Phase II used a noncontact-type internal floating roof. Phase III used
a double deck external floating roof.
A total of 29 tests were conducted during the three phases. Conditions
were varied in order to determine the:
o Emissions from a tight primary seal.
o Emissions from a tight primary seal and secondary seal.
o Effect of gaps in the primary and/or the secondary seal.
o Contribution of deck fittings (penetrations) to emissions.
o Effect of vapor pressure (temperature) on emissions.
C.2.2.1 Description of Floating Roof and Seals.
C.2.2.1.1 Phase I, contact-type internal floating roof. A
cross-sectional view of the position of the floating roof within the test
tank is shown in Figure C-2.
A flapper secondary seal was used during some of the tests. This
seal was 15 inches wide, with internal stainless steel reinforcing fingers.
A sketch of its installation on the rim of the contact-type internal
floating roof is shown in Figure C-3.
Description of test conditions—The test conditions for Phase I are
summarized in Table C-l. This table presents a brief overview of the
various temperatures, seal configurations, and deck fitting sealing
conditions for the Phase I emissions tests.
C.2.2.1.2 Phase II, noncontact-type internal floating roof. The
internal floating roof for the Phase II tests was fitted with shingled,
flapper type primary and secondary seals. A plan view sketch of a portion
of a shingle-type seal is shown in Figure C-4. Also, the dimensions of a
single piece, or shingle, of the seal is shown. Figures C-5 and C-6
describe the details of the shingled, flapper type seal that was installed
in lieu of the single continuous flapper seal used during the propane/octane
tests. Figure C-5 shows a cross-sectional view of the position of the
noncontact-type internal roof within the emissions test tank.
Description of test conditions—The description of test conditions for
Phase II are summarized in Table C-2. This table presents a brief overview
C-6
-------
REMOVABLE EXTERNAL,
CONE ROOF
AIR PLENUM
3O"0 AIR DUCT
AIR OPENING
CONTACT-TYPE INTERNAL
FLOATING ROOF
SR-B RESILIENT
FOAM SEAL
RIM SPACE HEATING
«, COOLING COILS STf"1
UCT LEVEL
Figure C-2. Position of the contact-type internal floating
y roof within the emissions test tank.
C-7
-------
BOTTOM OF AIR OPENING
FLAPPER-TYPE
SECONDARY SEAL
CLIPS ON 3" CENTERS FASTENING SECONDARY
SEAL TO RIM OF ROOF
•PRIMARY SEAL IMMERSED IN BENZENE
:ONTACT-TYPE
INTERNAL FLOATING ROOF
Figure C-3. Rim mounting of the flapper secondary seal.
C-8
-------
L±J
D; o
o o
u- oi
oo cs
•z. -z.
O r-l
r-l O
O —I
Z U.
o
O ,-J
UJ UJ
O
UJ
>- a.
ai >-
«C I—
CO
O
O)
res
o
O
O
01
0)
o
z
M
M G
O~r4
•^
O -P
a -u
fa
o u
AS id
13 K
• rH W
O «3 DJ
Won)
CO W O
• rH
U (d
01 V
CO CO
^t
^4
«d »H 01
E id &
•H 0) Id
M CO U
04
•0 ft ~
o B sr
V4 0) o
-P .
0) O
0> Z
EH
4J
(0
O
rH
id
tH
4J
jrl
P<
•8
id
01
n
C
O
•o
0)
(!)
01
C
D
0)
o
z
o
c
o
0)
c
o
o
CO
rH
1
rt*
£4
M
-P
M
0)
rH
id
-H
JJ
^1
id
d.
•o
01
tH
0)
(0
G
D
•O
0
m
a
CO
o
0)
c
o
z
at
c
o
0
CO
CJ
1
o«.
W
n
a>
6-1
rH
id
•H
4J
^
m
(Xi
•o
0)
(0
0)
U!
C
•d
0)
G
o
z
o
c
o
z
0
o
rH
CO
1
P4
43
a]
a
EH
Id
•H
4J
M
M
04
•O
0)
id
0)
CO
o>:
id
o
CO
o
o
G
O
o
o
rH
^t
|
<
A.
W
•a
a
ns
a*
CO
•a
0)
ns
0)
CO
o
•a
o
c
0
z
in
r*^
0
i-i
flj
IX
•o
o
id
CO
•o
o>
id ~
01
CO
a
G
O
Z
0)
G
O
Z
fc
CN
f^»
_x
•JT*
•H
1
O
CO
^
rH
I
Pi
w
•a
0)
id
o>
»
•a
"«j
0)
(0
o>
o
z
n
u
0)
c
o
z
in
r-
id
0)
CO
0)
I
01
0)
>4
_
C4
f-
£
•Jf
rH
1
in
r*.
n
•H
1
p.
W
•a
V
id
V
CO
o
r-t
IS
=
*^
X
^yv
rH
1
m
r-
"3-
rH
I
p t
W
•a
(U .
aS
CO
•o
0)
rH
ns
0)
CO
0)
c
g
0)
c
o
z
Q)
C
o
z
in
r-
m
rH
1
Oi
a
•o
0)
as
0)
CO
•O
0)
rH
c
o
s
Ml
rs
e
Jf .
m
VO
1
^
Ai
C-9
-------
-4-
12
INDIVIDUAL PIECE OF SHINGLE-TYPE SEAL
TANK SHELL-
RIM PLATE-
STEEL CLAMP BAR
PLAN VIEW
(the same detail was used for both
primary and secondary seals)
Figure C-4. Installed shingle-type seal.
C-10
-------
REMOVABLE EXTERNAL
CONE ROOF
AIR PLENUM
3OV AIR DUCT
Figure C-5. Position of the noncontact-type internal floating
roof within the emissions test tank.
C-ll
-------
SECONDARY SEAL
STEEL CLAMP BAR
BOLTED JOINT
PRIMARY SEAL
FOAM TAPE
FABRIC SEAL FOR MOUNTING BRACKET
MOUNTING BRACKET FOR SECONDARY
SEAL
STEEL CLAMP BAR
7
FOAM TAPE
RIM PLATE ' DECK SKIN-
DECK SKIN CLAMP BEAM ASSEMBLY
DECK SKIN
Figure C-6. Cross-sectional view of the shingle-type
seal installation.
C-12
-------
«C
2F t1-
D- o
«s
U- o
CO £
o
o
oo
UJ
ffll
5
S
«o
0)
f^
10
0)
CO
K,
•-J
w •
« E
0-H
CU«H
0) O
•P -H
4J
•O
«.*
i— 1 U
*0 0)
0) -O
to
•o
f-l (0
•H
n » •
Mi-Id)
O -rl iH
~ cum (0
g di
0) U (0
•P-H
43 O
jn »A m
x
•P O
9 «
J3-O
<3<0
CM
1 (n
< G
Pi-H
10
10 0)
•a n
CO f-l
0 C
•H O
n-l —4
,.
4) it
0 O>
JQ n C
•H-H
» « *»
fH fl) 4J
01 10-H
1 *W
*•€ ^1
CU i4 ^4
Vl O
to o«o
to
U 10
(0 •-!
o. a
O 10 f-l
tO rH 0)
D-D-
CO n
W
C-O
O"O
o a •
O.C E
E 4J ra
«d-H 0)
w ? m
e
^"
•H
10
0)
CO
0) O
(0 CO
w
0)
o
2
•O
0)
r-t
10
O
CO
•o
0>
0)
(A
I
n
o>
10 r-4
6 *
«
CO
CO
O.
<0
a>
I
X
n
0
§
C
O
O * «
"
•g
f-l
At
O
VI
m
*
I-a f fa
O O o
M fr« —
w o
0) %
H
H
o
I
«n
r-
CM
04
H
n
o
s
in
r-
m
IN
«S
k
H
C-13
-------
of the various temperatures, seal configurations, and deck fitting sealing
condition for the Phase II emissions tests.
C.2.2.1.3 Phase III, external double deck floating roof. A cross-
sectional view of the position of the double deck roof within the test
tank is shown in Figure C-7. The figure also illustrates the metallic
shoe seal mounted on the double deck external floating roof. When a
secondary seal was required, the flapper type secondary seal from Phase I
was reused. However, in order to fit it to the double deck roof, the
length of the secondary seal had to be shortened because of the slightly
smaller diameter of the double deck roof.
Description of test conditions—The test conditions for Phase III
are summarized in Table C-3. This table presents a brief overview of the
various temperatures, seal configurations, and deck fitting sealing
condition for the Phase III emissions tests.
C.2.3 Emissions Test Data
C.2.3.1 The Effect of Vapor Pressure on Emissions. Several emissions
tests (EPA-5, EPA-9, and EPA-15) were initially conducted to determine
the effect of the product vapor pressure, P, on the emissions rate. This
relationship was evaluated during these tests by varying the product
temperature in the pilot test tank which had been fitted with a contact-type
internal floating roof and a liquid-mounted primary seal. The product
temperatures maintained during the three respective tests were 100 F
(EPA-5), 60°F (EPA-9), and 75°F (EPA-15). Based on these tests, the
emissions are directly related to the vapor pressure function, f(P):
f(P) =
14.7
1 -
14.7
V. >
o.3"\
C.2.3.2 The Effect of Seal Gap Area on Emissions. Several tests
were performed to determine the rates of emission as a function of seal
gap area.
Table C-4 presents the seal gap areas tested and the measured emissions
for the Phase I testing of a contact-type internal floating roof. Several
conclusions are apparent from these tests:
1. A comparison of the emissions measured during tests EPA-5,
EPA-9, and EPA-15 with the emissions measured during tests
C-14
-------
30"* AIR DUCT
V
REMOVABLE EXTERNAL
CONE ROOF
AIR PLENUM
fi
B-
RIM SPACE HEATING
& COOLING COILS "^
8
SHELL HEATING
& COOLING
JACKET -" .
DOUBLE DECK EXTERNAL
FLOATING ROOF
z
Figure C-7. Position of the double deck external
floating roof within the emissions test tank.
C-15
-------
UJ
a
re u.
o« ^^
o; oi
o
u. en
in »-<
§ ^
(-4 O
o «a:
° g
t; P
LU X
H- UJ
U. ^
O CJ
UJ
& °
S UJ
£ CO
S =5
g
°?
O
(U
s
(O
a
4)
4*
o
55 ,
• <-4 n
o a a
O 0) o
M £4 *-
Pi
44 .
n o
Q) £3
*
tH
(0
O
«M
•O
0)
A
41
CO
a
o>
e
4J
a
HJI 1 1
o n
0 0)
O -u
4)
O
O
a
o
o
o
S3
in
|N»
f>
CM
1
Pu
W
§
o
o
S5
*
W
X
B
t-l
1
in
t-
^»
CJ
I
Pi
H
_X _
^p r*"
i-i ro
rH
i
•-)
(0
X
' e
C4
X
l-l
1
in
r^
P.
in
CM
I
Pi
H
4)
n
o
n
o)
E
€V
X
s
r-l
I
JS
m
t*»
in
cs
I
04
w
^.
JN
_x
>!.
f-1
1
C4
u
0)
X
E
C*l
X
s
^
' 1
cs
m
f^.
vo
c^
1
04
w
4)
O
m
4)
4)
A
O
Si
in
f>«
(•••
c^
I
cu
W
£
C4
X
N
*v^
t-l
1
W
to
41
X
w
_x
C*4
v.
H
1
*3"
in
p^
CO
w
1
I^J
pi
W
'
o
c
53
V
a
CM
X
f>|
•*v
iH
I
rf
in
f«*
ot
ts
1
Pi
C-16
-
-------
*t
fn
uj
i— *
tf\
V i
UJ
i— i
UJ
CO
3C U_
n. o
o
eco:
0.
UJ CD
•z.
§1—1
^—
a: =t
u_ o
CO U_
§ ,
t— < ^C
CO Z
co a:
i— i UJ
m g»»
i— i
UJ
Z UJ
111 Q_
g£
UJ 1
CQI—
^J
0 «=C
p^ jy
=3 O
CO 0
o
I—
S-
cu
-Q
•z.
JC
a.
E
in
0.
^
o
r-l
a.
E
in
r—*
*v^
q-
^*^^
CM
C
•r-
4/1
O.
re
p
^-^
•f*^
I-
^^^
M
c
•p.
C^
C/)
Q-
(T3
CT1
M—
o
^^
10
ai
I—
^** CD CM CO CO
• • • • •
CO *3- !"•» CM CM
t-l
co i-i oo r». r-.
CO O IT) rM r-l
f-t
1^. 00 OO O O
• • • * •
CM IT) «3- t-( T-I
1 1 1 O O
III
.a .a .a o o
CO
I r-l T-) 1 »H
1 CM 1 CM
O «* CM O ^~
•a
c
re
re
*• «in r-i vo CM co
in CT» T— 1 r—l « — I rH T— 1
III 1 1 1 1
^t ^c ^c ^c ^c *^ ^c
a. a. a. a. a. Q- o.
UJ UJ UJ UJ UJ i.i in
^
a
43
•i •'
^W
-i->
' S-
€U
CU
E
re
•r-
TD
•4*J
o
M-
O
C\J
cu
JC
i
«r-
cx.
?
re
•r"
V)
a.
in
•
t— i
•!->
re
cu
£=
N •
c- r—
cu re
i-l Q)
10
re >>
•o re
CU "O
+-> £=
re o
i— O
3 CU
O J/>
r-»
re o
o •z.
re .a
C-17
-------
EPA-11 and EPA-16 clearly demonstrates that increasing gap
areas in the primary seal increases emissions.
2. A comparison of the emissions measured during tests EPA-5,
EPA-9, and EPA-15 with the emissions measured during test
EPA-12, in addition to a comparison of the emissions measured
during tests EPA-11 and EPA-13, demonstrates that the addition
of a secondary seal reduces emissions.
3. A comparison of the emissions measured during tests EPA-12 and
EPA-13 shows that, as long as the secondary seal has no gaps,
the emissions rate is generally independent of the amount of
gap in the primary seal.
No relationship between seal gap area and emissions could be
established from the Phase II testing of a noncontact-type internal
floating roof. This was probably a result of the type of primary and
secondary seals used during the tests.
Table C-5 presents the seal gap areas and the measured emissions for
the Phase III testing of a double deck external floating roof. Several
conclusions are apparent from these tests:
1. A comparison of the emissions measured during tests EPA-23 and
EPA-24 demonstrates that small gap areas in the primary seal do
not increase emissions.
2. A comparison of the emissions measured during tests EPA-23 and
EPA-27, in addition to a comparison of the emissions measured
during tests EPA-24 and EPA-25, demonstrates that the addition
of a secondary seal reduces emissions,
3. A comparison between similar cases in Tables C-4 and C-5 demon-
strates that the emissions from an external floating-roof tank
are much higher than the emissions from a contact-type internal
floating-roof tank similarly equipped.
C.2.3.3 The Development of Seal Factors (K ) and Wind Speed
Exponents (n). The emission factors (K and n) for internal and external
9
floating roofs with primary seals and primary and secondary seals were
developed from the emissions tests data previously discussed. The emis-
sions factors for contact internal floating roofs and external floating
roofs having primary seals and primary and secondary seals are average
C-18
-------
CD
00
UJ
UJ
CO
«=c o
0-0
COO
CO -J
«C
CO
UJ X
2: o
UJ UJ
CQ O
O UJ
UJ —I
C£ CO
=> ZD
CO O
2°
o
cu
S
re
c >
O «
•i- t:
CO -•.
to c/
•r- JC
IxlC
re
O)
VI
^
s.
•o
E
0
o
cu
co
*re
cu
10
>.
S-
n
Q.
0.
E
LO
JC
0.
o
r—
O.
E
Q) — »
,__ N 4-»
^^ ,_. ii
^J VI ^**.
O CNJ
' « -i-
c-.
C/)
S- 0.
ja en
2: o
»" N +>
*2'i— M—
o "evT^
t- ex c
re -r-
c-,. *
S- Q.
cu re
JQ O>
2: o
"tn
CU
1 —
t
CO O CO O l"%
«* i— «=}• r— CM
0 0
CM O CM O CO
CO r— CO i— CM
CO CO •*
o en CD en r*».
CM CM r—
1 1 1 1 CO
1 1 1 1
r—
re o re o CM
1 1
co co • co
O O CM . CM CM
co r^ «*• ir> - *o
CM CM CM CM CM
1 1 1 1 1
D- Q- O- O_ O-
1 1 1 1 1 1 1 1 i UJ UJ
re
-------
seal factors developed from these test data and field tank gap measurement
data. Using a methodology similar to one discussed ir American Petroleum
Institute (API) Publication 2517,1 the test data from selected EPA Phase I
and Phase III tests were weighted to represent gap measurement data
collected by the California Air Resources Board (CARB) during seal gap
area surveys on external floating roof tanks. Based on engineering
. judgment, it is reasonable to assume that they are also representative of
the seal gaps on internal floating-roof tanks.
Consequently, the emission factors for a contact-type internal
floating roof with a primary seal (cIFRps) were estimated based on the
weighted average of tests EPA-15 and EPA-16, which have no measurable
seal gap and 1.3 square inches of seal gap per foot of tank diameter,
respectively. Because 65 percent of the tanks surveyed by CARB had no
measurable gaps, the emissions measured during test EPA-15, the test with
no measurable gap, were weighted at 65 percent. The remaining 35 percent
was assigned to the emissions measured during test EPA-16.
Similarly, the emission factors for a contact-type internal floating
roof with primary and secondary seals (cIFRss), an external floating roof
with a primary seal (EFRps), and an external floating roof with primary
and secondary seals (EFRss) were estimated by applying appropriate weighting
factors to the EPA test data to represent the CARB tank survey data.
Table C-6 summarizes the emission factors for internal and external
floating roofs.
Some of the data collected during the Phase I and Phase III tests
were not used to develop emission factors. Data collected during Phase I
tests EPA-1 through EPA-4 were not used because these tests were performed
primarly to evaluate the performance of the test facility. Data collected
during test EPA-10 were voided because the secondary seal was not compatible
with benzene. Data collected during test EPA-11 were not used because
the seal gap area was unrealistically large.
Data collected during Phase III test EPA-25P were not used because
of a failure of the secondary seal. Data collected during test EPA-28
were not used because the seal gap area was unrealistically large.
Additionally, while the testing did not specifically address the
control effectiveness of placing a fixed roof over an external floating
roof, it is reasonable to assume that the emissions from a tank so modified
C-20
-------
Table C-6. EMISSION FACTORS AND THE BASIS OF ESTIMATION'
Roof and
seal3
cIFRps
cIFRss
ncIFRss
EFRps
EFRss
Primary Secondary
seal gap seal gap
EPA (inVft tank (inVft tank
test(s) diameter) diameter)
EPA- 15
EPA- 16
EPA- 13
EPA-14
EPA- 17, 18
EPA- 23
EPA-24
EPA-29
EPA-25
EPA-26
0 no seal
1.3 no seal
21 0
21 21
0, 1.3 0, 1.3
0 no seal
.3.4 no seal
14.4 no seal
3.4 0
3.4 1.3
Weighting
factors ,
(%)
65
35
90
10
NA
10
85
5
75
25
Emission
j Factors
Ks
12.7 0.4
3,6 0.7
10.3 1.0
48.6 0.7
57.7 0.2
cIFRps - contact internal floating roof with a primary seal, cIFRss -
contact internal floating roof with primary and secondary seals,
ncIFRss - noncontact internal floating roof with primary and secondary
seals, EFRps - external floating roof with a primary seal, EFRss -
external floating roof with primary and secondary seals.
C-21
-------
would be equivalent to the emissions from a contact internal floating-roof
tank similarly equipped.
C.3 ESTIMATING EMISSIONS FROM FIXED-ROOF TANKS :
As discussed in Chapter 3, the working and breathing loss equations
from AP-42 were used to estimate benzene emissions from fixed-roof tanks
storing benzene. However, breathing losses estimated using these equations
were discounted by a factor of 4, based on recent fixed-roof tank tests
conducted for the Western Oil and Gas Association (WOGA), EPA, and the
German Society for Petroleum Science and Carbon Chemistry (DGMK).
C.3.1 WOGA and EPA Studies
During 1977 and 1978, 56 fixed-roof tanks were tested for WOGA and
EPA. Fifty of these tanks, which were tested for WOGA, were located in
Southern California and contained mostly California crudes, fuel oils,
diesel and jet fuel. These tanks were in typical refinery, pipeline, and
production service. The remaining six tanks, which were tested for EPA,
contained isopropanol, ethanol, acetic acid, ethyl benzene, cyclohexane,
and formaldehyde, respectively.
C.3.1.1 Test Methods for the WOGA and EPA Studies. The test methods
for the WOGA and EPA studies followed the methods described in the American
Petroleum Institute (API) Bulletin 2512, "Tentative Methods of Measuring
Evaporative Loss from Petroleum Tanks and Transportation Equipment,"
Part II, Sections E and F. This document recommends that the emissions
from a fixed-roof tank be estimated by measuring the hydrocarbon
concentrations and flow rates leaving the tank.
In the WOGA study, the volume of vapors expelled from a tank was
measured using a large and a small positive displacement diaphragm meter
and a turbine meter connected in parallel. Three meters were used so
that the potential range of flow rates could be covered. These meters
were connected to the tank with flexible tubing. Vapor samples, which
were taken from the tank using a heated sample line, were analyzed con-
tinuously with a total hydrocarbon analyzer. With continuous monitoring,
fluctuations in the hydrocarbon concentration could be noted. Periodically,
grab samples were taken and analyzed with a gas chromatograph, providing
details on hydrocarbon speciation.
In the EPA study, the volume of vapor emitted from a tank was measured
by positive displacement meters of either the bellows or rotary-type,
C-22
-------
depending on flow rate. Both meters were mounted so they could be manually
switched for positive and negative flow through a one-way valve which was
weighted, when applicable, to simulate the action of a pressure-vacuum
valve. Vapors from the tank were sampled using a heated sample line (to
reduce condensation in these lines), and then monitored with a total
hydrocarbon analyzer calibrated specifically for the chemical in the
tank. For the formaldehyde tank, a thermal conductivity gas chromatograph
was used instead of a flame ionization detection gas chromatograph.
C.3.1.2 Test Data and Conclusions from the WOGA and EPA Studies.
In these studies, 33 tank tests were available for correlation with the
API 2518 breathing loss equation which is the basis for the breathing :
loss equation in AP-42. Table C-7 lists the emissions measured during
each of these tests and the emissions calculated using the API equation.
Measured versus calculated emissions for each of these tanks are also
presented in Table C-7. Of the 33 tanks tested, only two had measured
emissions larger than those calculated using the API breathing loss
equation. In general, the API equation overestimated breathing losses by
approximately a factor of 4.
An additional 13 tank tests from the WOGA study were available for
evaluating the emissions from a fixed-roof tank in continuous working
operation. However, because of limited and scattered data and the fact
that breathing losses could not be separated out of the emissions, no
suggestions were made for developing a new correlation for working losses
from fixed-roof tanks.
C.3.2 DGMK Study
During 1974 and 1975, emissions tests were conducted by the German
Society for Petroleum Science and Carbon Chemistry (DGMK) on a 3,000
cubic meter fixed-roof tank storing gasoline. The tests were designed to
evaluate the effects of climate and method of operation on the emissions
from the tank over a long period of time.
C.3.2.1 Test Methods for the DGMK Study. A large number of parameters
were measured and recorded during the tests, including volume of vapor
leaving the tank, concentration of hydrocarbons in the emitted vapor, gas
pressure and temperature in the tank, liquid temperature, liquid level,
ambient temperature, air pressure, and solar radiation. In addition,
C-23
-------
Table C-7. MEASURED AND ESTIMATED BREATHING
LOSSES FROM FIXED-ROOF TANKS
Test no.
USEPA # 1
2
3
4
5
6
7
' 8
9
10
11
12
WOGA # 1
2
. 3
4
5
6
7
8
9
10
Type of Measured breathing API calculat«dg
product Toss breathing loss
(bbl/yr) (bbl/yr)
Isopropanol
Isopropanol
Ethanol
Ethanol
Ethanol
Acetic acid,
glacial
Acetic acid,
glacial
Ethyl benzene
Ethyl benzene
•Cycl ohexane
Cyclohexane
Cycl ohexane
Crude
Crude
Fuel oil
Crude
Fuel oil
Diesel
Crude
Crude
Crude
Jet component
20
22
8
4
8
24
45
15
19
27
23
19
0
o
0
0
1
0
224
164
222
0
(continued)
C-24
59
59
49
54
46
75
93
39
44
172
141
167
17
51
91
10
101
21
607
257
856
44
Measured/cal cul ated
0.34
0 .37
0.16
0.07
0.17
0.32
0.48
0.38
0.43
0.16
0.16
.0.11
0.00
0.00
0.00
0.00
0.01
0.00
0.37
0.64
0.26
0.00
-------
Table C-7. Concluded
Test no.
WOGA # 11
12
13
14
15
16
17
18
19
20
21
Type of
product
Crude
Crude
Crude
Fuel oil
Crude
Crude
Crude
Crude
Crude
Crude
Diesel
aAPI Bulletin 2518, "
Measured breathing
loss
(bbl/yr)
1
6
240
3
84
339
1,086
0
9
0 *
20
API calculated Measured/ calculated
breathing loss3
(bbl/yr }
26
74
167
17
138
490
783
61
298
2
38
Average
0.04
0.08
1.44
0.18
0.61
0.69
1.39
0.00
0.03
0.00
0.53
0.29
Evaporation Loss from Fixed-Roof Tanks."
C-25
-------
using discontinuous measurements, vapor samples were analyzed in a laboratory
for speciation and total hydrocarbons.
The flow rates from the tank were measured using three bellows gas
counters connected to the breathing valves on the tank. Three gas counters
were used so that extremely high and extremely low volume flows could be
determined. The three bellows gas counters were installed on the roof of
the tank. The pressure drop across the counters was 20 mm water column
at full load. The additional pressure drop caused by the counters was
compensated for by installing a new set of breathing valves.
An electrically-heated sampling line was connected from the outlet
of each of the bellows gas counters to the measurement room. The vapors
were analyzed with a flame ionization detector (FID) for total hydrocarbon
content. Grab samples were also analyzed using two different gas
chromatographic techniques to determine total hydrocarbons and individual
components.
C.3.2.2 Test Data and Conclusions from the DGMK Study. Table C-8
presents the measured breathing and working Tosses and the losses calcu-
lated using the API 2518 breathing and working loss equations. A comparison
of the measured and calculated losses indicates the measured breathing
losses are only 24 percent of the estimated breathing losses. In addition,
measured working losses are approximately 96 percent of the working
losses estimated using API 2518.
C-26
-------
Table C-8. COMPARISON OF MEASURED LOSSES WITH
THOSE CALCULATED USING API 2518
Loss and
time period
Breathing loss9
69 days
46 days
45 days
160 days (total)
Working loss
69 days
46 days
45 days
160 days (total )
alncludes withdrawal
Measured
(Mg)
2.0
0.6
0.7
3.3
12.2
11.3
5.9
29.4
loss.
3
Calculated
(Mg)
6.6
3.9
3.2
13.2
12.2
12.9
5.6
30.7
Measured/cal cul ated
(Mg)
0.30
0.15
0.22
0.24
1.0
0.88
1.05
0.96
C-27
-------
C.4 REFERENCES FOR APPENDIX C
1. American Petroleum Institute. Evaporation Loss, from External
Floating-Roof Tanks. API Publication 2517. February 1980.
2. Letter and attachments from Moody, W. T., TRW, Incorporated, to
Richard Burr, U.S. Environmental Protection Agency. April 24, 1980.
Emission Factors for VOL and Benzene.
3. American Petroleum Institute. Evaporation Loss from Fixed-Roof
Tanks. API Bulletin 2518. June 1962.
C-28
-------
APPENDIX D
METHODOLOGY FOR ESTIMATING LEUKEMIA MORTALITY AND
MAXIMUM LIFETIME RISK FROM EXPOSURE TO
BENZENE EMISSIONS FROM BENZENE STORAGE TANKS
-------
L
-------
APPENDIX D - METHODOLOGY FOR ESTIMATING LEUKEMIA MORTALITY
AND MAXIMUM LIFETIME RISK FROM EXPOSURE TO
BENZENE EMISSIONS FROM BENZENE STORAGE TANKS
D.I INTRODUCTION
The purpose of this appendix is to describe the methodology used in
estimating leukemia mortality and maximum lifetime risk attributable to
population exposure to benzene emissions from benzene storage tanks. The
appendix is presented in three parts:
• Section D.2, Summary and Overview of Health Effects, summarizes
and references reported health effects from benzene exposure.
The major reported health effect is leukemia. Mortalities
cited in the BID include only the estimated leukemia deaths
attributable to exposure to benzene emissions from benzene
storage tanks at existing petroleum refineries, chemical
plants, and bulk storage terminals although other, sometimes
fatal, effects are known to result from benzene exposure.
• Section D.3, Population Density Around Petroleum Refineries,
Chemical Plants and Bulk Storage Terminals, describes the
method used to estimate the population at risk; i.e., persons
residing within 20 km of existing facilities having benzene
storage tanks.
« Section D.4, Population Exposures, Mortalities, and Risks,
describes the methodology for estimating benzene emissions
from model plants, calculating expected population exposures,
and estimating the number of leukemia deaths and maximum
risk of leukemia attributable to benzene emissions from
benzene storage tanks at 143 existing petroleum refineries,
chemical plants, and bulk storage terminals.
0.2 SUMMARY AND OVERVIEW OF HEALTH EFFECTS
D.2.1 Health Effects Associated with Benzene Exposure
A large number of occupational studies over the past 50 years have
provided evidence of severe health effects in humans from prolonged
D-3
-------
inhalation exposure to benzene. Some 300 studies of the health effects
of benzene have recently been reviewed and analyzed -'n terms of application
to low-level ambient benzene exposures that might occur in a population
residing near a source of benzene emissions.
The reviewers concluded that benzene exposure by inhalation is
strongly implicated in three pathological conditions.that may be of
public health concern at environmental exposure levels:
• Leukemia (a cancer of the blood-forming system),
• Cytopenia (decreased levels of one or more of the formed
elements in the circulating blood), and
• Chromosomal aberrations.
Leukemia is a neoplastic proliferation and accumulation of white
blood cells in blood and bone marrow. The four main types are acute and
chronic myelogenous leukemia and acute and chronic lymphocytic leukemia.
The causal relationship between benzene exposure and acute myelogenous
leukemia and its variants in humans appears established beyond reasonable
doubt.1
The term "pancytopenia" refers to diminution of all formed elements
of the blood and includes the individual cytopenias: anemia, leukopenia,
thrombocytopenia, and aplastic anemia. In mild cases, symptoms of pancyto-
penia are such nonspecific complaints as lassitude, dizziness, malaise,
and shortness of breath. In severe cases, hemorrhage may be observed,
and death may occasionally occur because of hemorrhage or massive infection.
Patients with pancytopenia may subsequently develop fatal, acute leukemia.
Chromosomal aberrations include chromosome breakage and rearrangement
and the presence of abnormal cells. These aberrations may continue for
long periods in hematopoietic and lymphoid cells. Ample evidence exists
that benzene causes chromosomal aberrations in somatic cells of animals
o
and humans exposed to benzene. The health significance of these aberra-
tions is not fully understood. However, aberrant cells have been observed
in individuals exposed to benzene who have later developed leukemia.
Some types of chromosomal aberrations may be heritable. Quantitative
estimates of heritable genetic damage due to benzene cannot be made from
D-4
-------
data on the frequency of somatic mutations, although this damage may be
occurring at concentrations as low as 1 ppm in air.
The review concluded that man may be the only species yet observed
to be susceptible to benzene-induced leukemia. Evidence for production
of leukemia in animals by benzene injection was considered nonconclusive.
Moreover, benzene exposure by oral dosing, skin painting, or inhalation
has not been shown to produce leukemia or any other type of neoplastic
diseases in test animals, although other effects, including pancytopenia,
have been widely observed.
D.2.2 Benzene Exposure Limits
It should be noted that where the health effects described above
have been associated with benzene exposure, the exposure has been at
occupational levels. That is, the benzene exposure levels associated
with the effects have been high (10 ppm up to hundreds of parts per
million of benzene, except in a few cases of exposure to 2 to 3 ppm
benzene) or they have been unknown.
Benzene exposure was first associated with health effects in occupa-
tional settings, so initial attempts to limit benzene exposures were
aimed at occupational exposures. With recognition of the toxic effects
of benzene and its greatly expanded use after 1920, several occupational
exposure limits were established in the United States.3 These limits,
originally in the range of 75 to 100 ppm, were successively lowered as
more information on benzene toxicity became known.
For example, the American Conference of Governmental Industrial
Hygienists (ACGIH) recommended a benzene threshold limit value of 100 ppm
in 1946, 50 ppra in 1947, 35 ppm in 1948, 25 ppm in 1949, and 10 ppm in
1977. ' The National Institute for Occupational Safety and Health
(NIOSH) recommended an exposure limit of 10 ppm in 1974 and revised it
downward to 1 ppm in 1976.5 The current Occupational Safety and Health
Administration (OSHA) permissible exposure limit is 10 ppm6.
Occupational exposure limits were initially established to protect
workers from adverse changes in the blood and blood-forming tissues. The
most recently recommended limit of 10 ppm is based on the conclusion that
benzene is leukemogenic in man (NIOSH and OSHA7) or a suspected carcinogen
in man (ACGIH4).
D-5
-------
The EPA Administrator announced in the June 8, 1977, Federal Register
his decision to list benzene as a hazardous air pollutant under Section 112
of the Clean Air Act. A "hazardous air pollutant" is defined as an "air
pollutant to which no ambient air quality standard is applicable and
which . . . may reasonably be anticipated to result in an increase in
mortality or an increase in serious irreversible, or incapacitating
reversible illness."
D.2.3 Health Effects at Environmental Exposure Levels
Little information is available on the health effects of nonoccupa-
tional exposures of the general populace to benzene. Virtually all of
1 7
the studies cited ' were on the working population (mostly males) exposed
to higher than ambient benzene levels on a work cycle. Applying these
studies to chronic (24 hours per day) low-level exposure to the general
population (including infants, the ill, and the elderly) requires
extrapolation.
The recent analysis of benzene health effects concluded that the
evidence of increased risk of leukemia in humans on exposure to benzene
for various time periods and concentrations was overwhelming but that the
data were not adequate for deriving a dose-response curve.
However, EPA's Carcinogen Assessment Group (CAG), acknowledging the
absence of a clear dose-response relationship, has estimated the risk of
2
leukemia in the general population from low-level benzene exposure.
Data from three epidemiological studies of leukemia in workers (mostly
adult white males) were used to estimate the risk of developing leukemia.
The annual risk factor derived for benzene-induced leukemia was 0.34
deaths per year per 106 ppb-person years of exposure.
A nonthreshold linear model was used to extrapolate this estimated
risk to the low levels (below 5 to 10 ppb) to which some populations may
be exposed. For example, if 3 million persons are chronically exposed to
1 ppb benzene, the model predicts there will be 1.02 leukemia deaths (3
x 0.34) per year in that population. Use of a "linear" model means that
the model would predict the same number of leukemia deaths among 3 million
people exposed to 1 ppb benzene as among 1 million people exposed to
3 ppb.
D-6
-------
The risk factor (0.34 deaths per year per 106 ppb-person years) was
used in estimating the number of leukemia deaths attributable to benzene
emissions from benzene storage tanks at petroleum refineries, chemical
plants, and bulk storage terminals. Other effects of benzene exposure
(including deaths from causes other than leukemia) were not included in
the estimated number of deaths. The risk factor equated one leukemia
case to one death (that is, each case was presumed fatal).
Several sources of uncertainty exist in applying the risk factor.
First, the retrospective occupational exposure estimates may be inaccurate.
CAG calculated the 95-percent confidence intervals for this risk factor
to be 0.-17 to 0.66 deaths per 106 ppb-person years if exposure estimates
in the three studies extrapolated are precisely correct, and 0.13 to 0.90
if exposure estimates are within a factor of 2. Second, the composition
of the exposed populations around petroleum refineries, chemical plants,
and bulk storage terminals may vary from that of the populations used as
a basis for the CAG estimate; the risk factor assumes that the suscepti-
bility to leukemia associated with a cohort of white male workers is the
same as that associated with the general population, which includes
women, children, the aged, nonwhites, and the ill. Third, the true
dose-response relationship for benzene exposure may not be a linear
nonthreshold relationship at the low concentrations to which the general
population may be exposed. Fourth, the risk factor includes only leukemia
deaths and not other health risks. No quantitative estimate of the
uncertainty in the risk factor due to the latter three factors has been
attempted.
D.3 POPULATION DENSITY AROUND PETROLEUM REFINERIES, CHEMICAL PLANTS,
AND BULK STORAGE TERMINALS
The population "at risk" to benzene exposure was considered to be
persons residing within 20 km of facilities having benzene storage tanks.
There are 143 such facilities in the United States: 28 large benzene
producers, 34 small benzene producers, 77 benzene-consuming plants, and 4
bulk storage terminals. These facilities are referred to as "plants" in
the ensuing discussion. Populations residing within radial distances of
1, 5, 10, and 20 km from each plant were estimated from an existing
D-7
-------
8 ?
population file. This file consists of a grid of I-km cells covering
the continental United States, each with an assigned population. The
population assigned to each cell was the 1975 estimated population,
extrapolated from the 1960 and 1970 populations of the census enumeration
district in which each cell occurs, assuming that the population is
uniformly distributed within each of the 256,000 census enumeration dis-
tricts. The population around each plant was determined by summing the
populations of all cells occurring in annular areas at radial distances
from the plant center of 0.5 to 1 km, 1 to 5 km, 5 to 10 km, and 10 to
20 km. The estimated total populations exposed as a function of distance
from the plant site are reported in Reference 8, Table A-4.
There are some uncertainties in the above method. First, the
assumption of uniform population distribution, both within enumeration
districts and annular areas, may not be precisely correct. For urban
areas the assumption is probably reasonably valid, but there is some
uncertainty for rural areas 10 to 20 km from the site. Another area of
uncertainty is the use of 1960 and 1970 population data. However, these
are the latest available in the form required. No attempt was made to
quantify the range of variability in the population figures.
D.4 POPULATION EXPOSURES, MORTALITIES, AND RISKS
D.4.1 Summary of Methodology for Calculating Deaths
The locations, descriptions, and capacities of all 143 U.S. plants
known to have benzene storage tanks were compiled. Using these data as
discussed in Chapter 3, four basic "model" plants were then developed to
characterize the benzene storage facilities of large benzene producers,
small benzene producers, benzene consumers, and bulk storage terminals.
The model plants contain seven, four, two, and two benzene storage tanks,
respectively, of various sizes. The benzene emissions rates from the
various storage tanks were then estimated for the baseline using the
available data. Each of the 143 existing plants was assigned the model
most resembling it. All model plants were assumed to be located along
the Texas-Louisiana Gu=lf Coast.
The omnidirectional annual average benzene concentrations (i.e., the
concentrations estimated assuming that the wind blows equally from all
D-8
-------
directions) in ambient air resulting from benzene storage tank emissions
were determined to a distance of 20 km from each model plant, according
9
to the Industrial Source Complex (ISC) dispersion model, rural mode 1.
Lake Charles, Louisiana, meteorological data for the 1973-1976 period
were used in the dispersion model. This period was considered
representative of dispersion conditions in the areas where the majority
of benzene plants are located.
The population around each actual plant location was then correlated
with its modeled benzene concentrations to yield a benzene dose to that
population in ppb-person years. The methods for determining populations
are described in Section D.3 of this appendix.
2
From health effects data, the EPA Carcinogen Assessment Group
derived a leukemia risk estimate of 0.34 deaths per year per 10 ppb-
person years from exposure to benzene. The methodology for estimating
the leukemia risk factor is described in Section D.2.3 of this appendix.
The leukemia deaths per year attributable to exposure to benzene
emissions from benzene storage tanks were estimated by multiplying
0.34 x 10~6 deaths per year per ppb-person year exposure times the expo-
sure in "ppb-person years," as described in Section D.4.2. The leukemia
deaths so calculated are summarized in Table D-l for each plant, with a
total for all plants of 0.31 deaths per year.
D.4.2 Estimates of Leukemia Deaths
The general equation for estimating the number of leukemia deaths
attributable to benzene emissions from a particular plant (e.g., Plant X)
8
ISIi
Dx =
where,
10-20 h,~ h+?
I (R)(l/3.2) (2)(7t)(p1)(a)(D1D** - D^ ^)/(b+2) , (1)
i = 0.5-1.0 2 1
D = estimated number of leukemia deaths per year from benzene
emissions from the plant (e.g., Plant X).
R = the risk factor (0.34 deaths per year per 106 ppb-person
years).
p. = density of population at risk, in area (i) around Plant
X.
D-9
-------
CO
2£
£C
UJ
cs
eC
O£
o
co
UJ
ar
UJ
M
z:
UJ
CQ
•«—
§
U_
co
o
H-4
c/to
CO CO
*— i z:
UJ 1— <
1—
UJ I— I
2= 0
III 2:
tvl O
UJ
CO UJ
z:
ae* i— i
o; uj
U_ CO
CO CQ
eC UJ
UJ a
o z:
1—4
UJ
UJ
_J
o
UJ
I—
> co
ca
CO
"O
s
>~
[ *
E
3
o
E
o
•r-
1 *
CO
o
o
1
•r—
0
CO
4->
+•>
CO
E
O
•r™
cn
CO
s-
^.^
f~)
^_^»
^.^
^
CO
a.
O
i \
E CO •
CO T3 O
r— O E
D- O
l£> CM CO CO CO «*
tf» j*.^ CD *5j" *-d~ CD
CO CM O •>!- «=J- CM
O CM O O O O
O 0 O O O O
0 O O 0 O 0
• CO • CO CO CO
E en E en en en
CO —1 CO _J —1 —1
s- s- s- s- s- s-
O_ Q. D_ Q_ Q- Q_
s-
co
+J
in
CO CO
.G S-
E 0 CO
o 3 en
•r- O CO
=3 O z!
CO
•a
E
CO CO X
f— 3 m -r-
i — CO CO CO O
E «i- E t— E S-
CO > O CO CO O
-O •»-> 1— 3 >>
E in E CO •
•i— CO • CO 3 4->
_i SB z: D_ CD co
•-3 <~s >- a£ ce. i— i
z: z: z: o- D- >
r—
•i—
O
.E in
i — -4-> in
•r-i— CO
O CO 'ZL
co in
•a S a. co
O E E -r- "O
E O CO O i— CO
O CO r— E r- S-
CO X X -E E T- CO
CC x CO in o -E E
LU UJ 1— «=C 0 Q- eC
c_3 i— i
Q I— 1 1— i
O
D- O O
•r- r-l CM CO •* If) ID T-
cn cn
• CO CO
i— i OL C£
^o co CP> v
•»-> CO
•f™ *n~ -N<^ ^^
CJ> J= O O
a. o o
CO •— 3= 3=
S- CO
co ~a in in
S CO 3 3
CO r— O O
t— •!- S- S-
CO .E CO CO
a D_ ss:
1 1 i ^^ »g^ g^^
Q Q- D_ Q-
l-l»
•p—
o
^ _-.!! ^"t
•i- S- i—
CO CO -r-
>, -ao
••-> If- E
+J •— CO E
CO 3 •»-> 3
C3 CD CO CO
I-H
E
O
Ps^ CO O"* O •! —
r— 1 OI
CO
01
CM LO «5f
in 01 oo
tn CM in
i-H i— 1 O
o o o
o o o
O O 0
• CO •
E en E
co —i co
si si C
Q_ Q- Q. '
E
CO 0
i— in
•i--O 3£
JO >> 0
O O CO
s: CQ --o
en
^.
3 CO
J3 i—
in 3
-l-> O
QJ +J CD
r— CO CO
•r- r— O
XJ 4-» in
O CO CO
S cj a.
—1 >- CO
«=C i«i S!
in
CO
•i—
S-
O)
E
•r—
t|
CO •—
o: •!-
o
CO
r— -O E
cn E o
E CO S-
CO r" >
•r- JZ CO
S- cn .G
r- «C 0
>
E
0
r-l CM CO •!-
t— 1 1— i i— i cn
CO
CtC
t--. CO CM CM
O «^ ^J* i — 1
O O O 0
0 0 O O
o o o o
CO • CO CO
cn E cn en
_i co — J — i
tl si tl s-
Q- O- Q- O.
E
O
in in
•r- _i^ co
-o o >> o
CO O CO 3
s: o CQ _i
S-
co .
^ ^)
•r— -1-3
o: +j T- o
E 0 TD
•o o co
0 E >>r-
O CO CO O
3 — 1 CQ H-
_i _i i— i :n
i— i i— i 2; O
1 •
CO
o
i— r—- *r-
•r- -r- E
O O CO r-
f"'**-"
r— E O O
r— O
CO T- 3 E
JZ E O 3
co r3 a co
•3- LO us r-.
1— 1 t-H I— 1 i— 1
-o
CO
3
E
!p
E
O
O
~ — *
D-10
-------
w
CU
3
C
"f"
O
O
t— 1
o
CU
s
1C
1—
S-
10 CU E -i
•i- Q. O CU
E s- c
cu 10 f- cu
.i£.E N
^3 4 * C f"*
CU (C (C CU
_! d) O> _Q
-o :>>
t
'
E
o
•^
+J
(C
o
o
V
4J
E CU
fC "O
i — O
0- 0
3
cu
s-
3
O-
X
O)
0
^™
O)
T3
O
c
>,
C
o
0
•r"
0
CU
res
4-*
CO
«•••*•»
C:
0
01
cu
S-
>}
_o
^— •*
>^
c
fO
Q.
>—
0
o
c
i~~3 cy\ c\i *j' o^ co [ 3 c 3 f"^ 01 f™~* (~ *) r*«^ f***. t,o tjj LO ^o uo c\j co cxj **j CD o^ *>j' 0*1
oo co ft *ij' i"1") c^ co vo co f**^ i~~^ ^o oo oo "^j o^ ^H f1** co ^*^ cvj cxi r^» v~^ ixs f^ r^^
co CD o*> t^ co co ^o cr* *^ LO CD CD co ^~ CD 0*1 ^* T— ^ r^* co *~H r*^ CD ^* co LO r— i
OOI^ O CO CMOi— iOCO*?J-C\Ji-HLOT-*OCOCOOCMCMOJOT-HO^O
OOOCD OOOOCDOOi— iOOOOOOOOOOOOOOO
OCDO O O OOOOOCDCDOCDCDOOOOOOCDOOOOCD ,
oooo ooooooooooooooooooooooo
*
•OCUd) (I) •&) • *OCU f-- en ^ ^ CD o> CD E E CD CD £ E CD CD CD E E E E CD CD
CO * i | i CO L J CO CO i F j-.i * t J CO CO f | CO CO — -* i--T , 1 CO CO CO CO i -- J
S-S-S-S- S-J-S-S-i-S^J-S-S-S-S-S-S-i_S-S-S-S-S-S-S-S_l-
O.O.Q.O. O.Q.Q.O.O.Q.O.O.O.O.O.Q.O.O.Q.O.O.Q.Q.O.O.O.O.
cu
o (/> -a 10
Oi CU S- CU
3 CU C (C •— C C C CE
CU i— E T- ES-OO (COlOOOrt)
•I— r— OE S-CCW-IJ «r- W) S- +J l/> «r—
OS-CQCT T3 lOM->S-S-OS-OrtSS-S-|s4S-«4-rCEO>M-IM
r— CU (O ~a • • i— M-r-S-S-CUS-CUSS-S-fOS-M-XreCUr— «»-reS
rt3 Q • ^^ (/) -— »• -P 3>jCU S- S-O> 3 -i--S->>
S-E3CCCUS-CU JZ-M'i- ^ ^=E J=E J=+J
c_5 O-i- (O4J +->O O'r-+J
J=EOiOEO-+J S- Or— E C' &. C C S- C &. 4-> C_)E
CJCU (CCUCU Q-OCUO5 O
Gi ^3™ ^3 i i QJ i ^3 (/J ^^ |Q c; [ ^ r~ ft (Q .^ >r.
•^^ (C *t~^ ^^ i ^> cC M ^^ ^— 5^ cC H3 t« *^ ^» ^7) Z3 OO CU ^*j t. y" ^^ c ^^ ^j ^^
_J 0. CO DO OO Ol— 1— O.I— C_)3TCJ>rcC_)OQ3IO-Li-£QO.OO3OI— Cd5
•=C«=t'=i: 01
c cu c
res -a -a o o- T-
C i— r— -r- C
•r- ' — CU CU +J C -i-
01 "of- ~a -i— -i— to (c i—
CU CU O S- *»— t- C to O CU
o 4-> cu.s-«^:^= s-cu-r-r— i— Q;
•r-i — T-T3+->XJOO CU+JS-tCnSi — i —
>(O C -r- CUC'i--r- 4-> (C 0) S- S- 1C -M -r-
s-o za ja o-rtsoio; E4-5E-*->-i->o ccuo
CU'i— S- -t-5 i— i CO •=£ C C v- CUCUi —
COE r— i— ro ccoooc — cua>E i— -OOET-O
CU -i- -i— O O i — (G - — -i— -r-'i-S-! — OC-3CU -r- EOOO4J
oi-c o oo -^o -M -M i— cu (C c j= Or— cus:^: c
CUCJC lxJO>COT-OCEQ-4-'4->CUCCCJC i — Q- 1 -4-> i — (8
•i- _ OM- CCO S-On3«CEt.tO'OSS OH-CUCUS-tO-i-oo
J^j r.-» r^~ CO (O fl3 t/1 CD CD ^£ ^^ [i TI ^£ T3 c , c _|^ «•**
E4-J-PJ=jrOOS-S-OX3OECU
-------
•o
cu
E
t %
O
CJ)
*
1
a
cu
S
CO
t-
•
S-
co CU E °
•i- a. o cu cu
E S- E S-
CU t/l t- O) 3
3 4-> S- E 0
CU CO (O CU CL
_l CU CU -Q X
•o >> cu
CO
1—
cu
"O
o
£
E
3
Q
E
O
•r—
4->
CO
O
O
"
CJ
•
CU
to
•4-9
co
E
o
en
cu
s_
17
^
a.
E
o
o
. ^
E CU •
CO T3 O
r— O E
D- 0
O CO O CM i— 1 OO 1 — •>!- «d- 1 — . <-H CM
f^ O^ l*1^ LO O^ CO CO ***>J CO CO CO i p
CM O O O OO O i— 1 t-H OO r-i O O
ooooooooooo o
ooooooooooo o
ooooooooooo o
_i oo oo oo _i oo oo — i — i oo oo oo
O_ Q_ Q- Q- D- D- Q_ Q-"b_ D- Q- D-
E
E O E E
O 1/5 CO O O
+j E -I- co in
t/1 -r- S- CO 01 tO S- 00 S- S-
> O Nl CJ S— O CJ CJ '1 CJ f| r—
t— 4->cCCUS-4->CUCU<4-CUM- 4J
cnflcazrnuj^izz'oz:'^ CD
•c— 'I— *r- M—
i \ t ^ t \ I *
t/) I/) l/l t/>
•I- 'I- -1- S- •!-
>> £^, .EJC-E-E-O 0
•i- OS- C_> O 4-> O E T3
CJ) CO S- CO (O
(/> CU E 3 CO33 3S- O
ca en cu Q. s_ co a. a.+-> a. cu o
XS-CUS-CUCUS-S-S-S-TD
CUOSOCUT3OOOOCU t —
1— caoooooooQ-cjz: uj
XXXXXXXXXXX 00
H-t-l-l-l-l-l-h-l-l-t- ^
^^
l_»
CU E*— • CJ
3 S- E •—
^-» O CU 3 M-
-O 3T 4-> OO -i-
CU r— i — t/)» O
3 O CO t/5 tO 'r— -i— CU COVO
E 4JQ-Q.EOO3CU Q-r-^
•i— E •!- -I— CO T3 O
•»-> cOi— i— -»->r— r— -E-i-CJEE >i
E W)r— r— Ei— i— •*-»•»-> CO OO 4J
O E -r- T- T- O) CU 3 E X •<- -i- •»-»
O O^TJ=3^E^:O3-CUEE CU
— s: Q- Q- o-co oo co oo i— =3 =0 o
1 — 1 I — 1
-1" ^"
E E
0 0
•r— LO ^O f^* OO C3*» O i — 1 CM CO "d" LO "r— CD
CD ^* «^~ ^" ^ -^J" LO LO LO LO LO LO Ol LO
cu cu
D: D:
CTV LO 1 — <—l CO VD
oo oo r*^ CT» i — 1 1 — i
CO CO O1 •* O O
O CO LO CM l£> CD
o o o o o o
o o o o o o
O 0 0 0 0 O
E E E E E E
OO OO 00 OO CO OO
S- S- S- S- S- S-
D. a. a. o_ D_ CL.
cO
O) CU CU 00
t— i— i— O
CU CU CU CJ> 3 3
CD C7) CD r^ ^™
E E E CO 3 3
cC eC -=C S- r— r—
4-> 0 0
CO (/) CO E E E
o o o o o o
E
o
J_fr
D^
E
•r-
E
^ J= O
3 o -a
— - CO E T3 3 3
O) 3 E i— i—
E CO CD O 3 3
O CU E r— r-
en cnco -E o o
S- E CJ E E
CO O r— «r- O O
c_> _i uj a; a: a:
«« «=C eC «C t-i I-H
o o o o :c 3=
E
•O ^-~^-N CU
i— -o -o -a
CU S- S- E
•r- CO CO CU
If. -o "O Q.
^= E E CU
O CO CO "O i —
•r- •(->•»-> E -r-
a: oo oo i— < o
O E T3
•r- E E CO S-
4-> O O -I- CO
E S- S- ••- -O
CO O.> > CO E OO
^~ CJ CU CU 3 CO Cc
4J s- -c: -G ca 4-> ui
eg: c=c o <_> 3: oo s:
X OO
1-1 o
E 0
o
•i- r-» oo cr> o •— i CM
CD LOLOLOIOIQCO •
CU •— '
OL t— i
•* oo
^v LO
i—t r-i
T— 1 CM
0 0
O 0
0 C>
E E
0 0
o o
S-
cu
•)-9
4-> CO
cu cu
CO _E
S- 0
CU 3
E 0
O i—
CO CIS
J^
o
Si
ca o •
4-}
•a io
§5
o ••-
03 CD
z: -z.
-a
•1—
E
CO
E
CO
0*
(0
0 4->
•r- E
s- o
Jo.
3
o
HH
H— 1
E
0
•r" CO ^"
en CD CD
cu
a:
* — ^
•o
cu
3
E
•1— •
E
0
o
D-12
-------
S-
"O
cu
3
E
*|"^
4-*
E
O
4-3
^
t—4
1
O
cu
T^
(0
i—
«3 CU E -Q
••- Q- O CU CU
E S- E S-
CU 10 <+- CU 3
.*: .E N to
3 4^ t. E O
CU to to CU Q.
_1 (U O> -Q X
•O >i CU
03
CU
-o
O
E
O
j ^
tO
o
o
-~
>^
[ ^
E
3
O
0
>^
1 ^
•1—
o
cu
03
CO
*c
o
•r-
cu
>^
.0
>^
E
a.
E
0
o
] *
E CU •
03 -a o
r— O E
0- O
^— ^
-a
cu
E
•i—
4->
E
O
O
1—4
t—t
E
O
•r—
01
CU
a;
vo C5 o** T— i ^^ ^j~ T— i cxi r*^ r*-» r*^ r***
*?}* CO ^D t~^ V.O O^ OO tO CZ^ t_ fi t^j ^^
^j1 !-•) t— i *5j* c 3 CSJ 00 C7^ t-Q iiO to tO
^t* C\J LO CO CM CO r— H C^ ^3 CO CD CO
oooooooooooo
oooooooooooo
oooooooooooo
lOlOlOlOl/)V5lOlOl/>J/)lOt/>
EEEEEEEEEEEE
OOOOOOOOOOOO
OOOOOOOOOOOO
S-
O) (O
X X 4-> i—
CU CU to 03 i—
(/) tO CU O> CO «r-
CU E CU -E 03 S- E
Ei— EOr— O-O03 03
O "O O (/) "O 3 E' O} ^>
•i--0-«--O-OOOo3 03
E i- E 3 -i- i— E i- 3
zs s: ^> z s: CD o :z CD
03
cu
-a w 3
i— r— E
CU i— CU
•r- to O.
H-.E CU U_
E +J i— CU to 03 V)
•r- CU >s i-~ *S) tO tO O CO 03
E03-QCU T-3i-.— -QEr—
"O *^ Isl <-- TD | ^ (O O^ 3 ^— >} 3
E *r~ O3 5- )'r- CU 03 3 O)
r— ,1— ii— E T-
CU C_5 CU 4-» t— -r- J3
J=.~O .E to reS M- S-
O r— -a CJ3CUCU1003
0 S- 0 T3 S Ct: 0.0
i— JE03OOTDEE -i-
E CU f~ T3 cU O CU *"~* O O r~~ E
OEOEE03-1- E O i— O
X E "^ O3 E X ^- f ^ E &— **— 'r""
X 3 CU4-»CU CU t— O O O-E E
i — i
t—t
H— 1
E
O
to *»o ^o **o v^ r*** r*^> r^ r**> r*^ r**« r^** 01
cu
ce:
c\j ^t- ^- r-* LO o ^o f**» to o^ CM
O^ ^T^ f^^ O^ CT^ LO C^? CO CO f1*^ C^
CO ^D LO ^O C^l C\J t" 1 <™ H ti 3 c,3p vj"
OCNJO-vi-CJOOOOOO
ooooooooooo
ooooooooooo
ooooooooooo
«/ltO«/)t/)tOtOt/5tOtO«/)«/)
EEEEEEEEEEE
OOOOOOOOOOO
OOOOOOOOOOO
03
cu ••-
i— cu JC >5 to
4-> S- 0. E •— 4->
t/) O i — CU i — E O3
O3ES-CU-EO3O3 i — i — .E
O T- CU -O OI-SI to CU CU CU 3
4->>i— CU1003ENN03
Sr— O3'i- •— S- CU >>4->4-> E
CU03CU-Ei— 03r— 03CUCU03
*zca ccio.«a:so.333^
cu
r—
>, >, -0 T-
4^ CU 03 E CU >
•i- r- -r- . (U • i— -I-T- t-i >4J4-5
S-O CU>> tOS-tO
03ES-"OCUl03 E03CU
3 T- CU 03 C31TO 0 -r- S i—
TC> 4-> > i — TDEi — i — 4J S-
i — i — 03 1-1- 3i — 03 S- 3 03
O>(Oa>.ES-OT-CU03CU-E
OCQCQO-CQSS^SZO
LUO.eCI:>^'>>>
0 5: Q- 0. 0. 3 3 33 3 3
TD
CU i— tJ
E a> i-
il ^ M^ "JO 03 10
o o ^: o E cu cu
T— O *t— O3 °t~~ "O
.E i r- E >>r— i- -i-
<_3o3Qi CUO-r- 4J-Q
4J j= o in s-
-OEOr— C_3E 3O3
S-CU-r-T-tO 03T3 T3O
03E4->OS--OUE E
•O-i-E CUCU-r-O3i — l— IE
E +-> 03 H- Q.-I- S- • r- O
03 E r— r— Q.1— CU -E JO CD •!-
4-d O 4-* 3 O r™" E t/1 O *"* E
ooo«cc3^:<:«5 co »— f csj co ^* LO tc* r^«» •!"•
r>. r». r^ co co oo co co oo co co en
cu
o;
^- •
00
1— 1
o
o
o
o
to
E
O
o
03
tf)
O
O
r—
03
O
t/)
3
1—
4_>
,_
o
i
-------
•o
0)
3
E
•r—
'•4-*
E
0
°
>
1—4
0
CU
1 •"
JQ
ftf
*"
s-
re cu E -Q
•r- O. O CU CU
E S- E S-
CU to t|_ CU 3
-*: .E N to
^*j .1% f ^— £^
co re re co o-
— 1 CU CO JQ X
-o >> cu
re
"cu
T3
i
>>^
-P
E
3
O
E
O
•r—
40
re
o
o
*
>^
4-J
CJ
CU
re
4J
oo
. — ..
E
0
•r—
en
cu
S-
>)
.Q
^—^
>5
E
re
o.
E
0
4-3
E CO •
re ~o o
r— 0 E
0- O
CM OO CM OO «*
to ^~ ^~ LO to
CM O O CD CM
CD CD O O O
CD CD CD O O
O CD CD O O
O CD O O CD
to to to to to
E E E E E
0 0 O O O
CJ CJ O CJ CJ
^— ^—
l— r— E
re re o
.E JE cu to
•o co to -ci .**:
>> s- s- re o
o re re cu re
ca s: s: s: '•D
co ^> ^>
S- 4-> 4-> CO
3 •!— -i— s- re
JD. 0 CJ 3 i—
CO -Q 3
4-> 4-> 4-> E O
4-> S- S- CO O)
cu cu cu -a re
r— > > E O
4-> i— i — re to
re re re s— re
cj cj cj ca a.
>->->->- oo
•r—
E O.
o a.
.E "i— "I—
O 4-> tO
•r- « CO
"•—• i— S- S- •!-
•a -i- T3 o to
CU O O Q. CO
3 O S- T-
E T3 CS OS
•r^ E • CJ
4J re • la_ 4->
E r- U_ • E CO
O -E eC -t- S-
O CO • • i — -i—
x ^ c^ ro C3 CD ' '
•—< >
E E
O 0
•i- Cn O t-H CM OO -i-
OI OO O"» CT» CTi O^ OI
CU CU
fv" rv
«* CM LO «* 00
o «— < cr> ch cn
r»« oo CD LO oo
O CM O O 0
CD CD CD O O
CD O O O CD
O O O O O
t/) CO CO CO tO
E E E E E
O 0 0 0 0
cj cj cj cj o
1_
•r—
re -o
•— >> E
cj -a re
O -3 >»TO
o 4-> s- re ••-
CJ OO CD CQ IS;
-o to
E •!-
re s-
r— S- >>
to o 4-> x»
i— i 4-» s: -i- E
a> o re '
CU O)4-> i—
3 3 10 >>-O
r-^ re re re *r~
CQ CO UJ CQ SZ
_i _J _j t-i i— i
i— i i — i i— i y ^y
r—
re
0
•r—
^
d)
f* r^ ^~
o re re
O 0
t— -a •!-•!-
•r- 0 r- E E
O 4^ O Q) CU
E -E -E ^:
-^ re .E o o
S- CO O
re E «r- s s
•— o a; o o
CJ S Qi Q Q
i — i
E
O
«d- LO to r-. co -i-
cu
a:
Oi— )LOO^CMOtO«^«^OCOCT>OO
f«_l ,_^ ^yj f"^ (^Q cj^ tO CD t~H ^" t-H ^" «d" CM
OOOOr- IOOO1LO«*OOC\Jr- li— (
Or-lOOOOOOOOOOOi-H
OOOOOOOOOOCDOOO
OOOOOOOOOOOOOCD
oooooooooooooo
totocoiotoiocotocototototoio
EEEEEEEEEEEEEE
OOOOOOOOOO O-O O O
CJOOOOCJCJCJCJCJOOOCJ
CO
o
CD E CO E E E
E r— o a> o o o
•r- O i — T- E T- +J to
J^4->-r-COretO COS-CO CO-OIO
(^ ^3 ^* £ '"3 £^
CU r— to
r— CU *r- &L *r-
^— CO > >> 3 CO i- CO
> O CL) O 'r-4->re CJ-i—
EDii — tOOIS- CJS- 1 S-E
o i — tJ.Ere reotocore4->rea-a. to
O4-> S- Ei r- r— X S- O) E S-C03
rerereocucu3cuOi— reo-i-o
'-3CQCJQSCDI— t— o_cjQ-cjca:E
ci;cCe:Ci:a-1.:]-ca-^,><><...><><><><><
•=C_ I_I_I_I_JOI— 1— 1— 1— 1— h-H-
re
E
•r-
O
S-
i ^
CU
T3 Q.
tO r—
CO re TD -r- re X
-E O S- *t- to O CO
o T- re JE: cu -r- 4->
CO E T3 O 4-> S- O
o CO>>ET- recus-
3: -EEreo; 4->E-*->
r— cj re 4-> oo -5 cu
re E r— i— Q-OO O d) O) — 'CL.
re re ct: -r- -I— E E — •<- co co i — —
> o CUeC
re cuooi — i — ^Q E Oi — i — i — re tos:
S-EO3333J=4->CUCUOOI-LJ
cyj CD T~H CM oo ^" LO co r*^ co o~i CD i • CM
O1OOOOOOOOOOr-l r-l'r-i
t— (i— li— It— (t— 1.— 1<— Ir— li— li— 4i— (i—<.— 1
"
^^^
•^J
CO
3
E
E
O
O
D--14
-------
TD
cu
O
CU
s
S-
ta cu E -Q
•r- 0. O CU CU
E S- E S-
53 M- CU 3
.ii .E N IA
3 +•> S- E O
CU > cu
to
CU
•^j
O
J^
o
.!_>
(0
o
o
$
3
O
O
.^^
+j
0
cu
E CU •
to TD O
i— 0 E
O- O
g5^2ocMcocMcoilOlOt/>t/>lOt/)lOl/>«/ltOtO
OOOOOOOOOOOOOOOOOOO
ooooooooooooooooooo
E EEEE EE E
1C O O O O tO O as O O O
$_ S_ tO ^> t/) 5— ^- t/)(/)(/)^C/)(/)^> 3 I/) i-
OCUS-T-CUCUCUOCU'i— -r-OCUCUCUCnOCUCU
i\it»-os-t->M-Ni>s-s-Ni>ot- cn^: o t-
S« tj— r— (4*. 03 r— S- t. tO ^— CU H— CU ^" CU M—
SH CU O (O CU tO CU ^ tO to to i~ n3 3 CU >- to 3 CU
rn *~3 1 1 1 ~T" »~3 o rO OQ CD ~r" T" CQ O 21 ^O O" O 21 f~3
3
O -r- -r-
to in in
S— CQ 3 'r- i- -I—
.C +J CU 4-> •>— -4-* -E .E -E *O
4J4J +J.r-+J+J-i-> -r-O-MS-POE
S-E ES-OEtOOr— E O S-CUM- tO
OOtOS^C Oi— CUO>» toeC-r-'r-l/lr—
ni ^^ t/i J-J* i .** *rt ^5 U tO ^" V) QJ (O ^> -|—3 C7i"C3 C?^ QJ
CUtOCU>>S-XIOOXtO3CUXS-S-EfCS--O
u_ CQ o CQ n_ i— CQ c_j|— on:coi— OQ — IGOOZ:
>^ >^ >< >^ >^ ?^ ?^ ^^ ^^ ^^ ^^ ?^ ^^ p^ ^^ ^^ p^» p^1 p^
1— 1— 1— 1— 1— H-f— 1— 1— 1— H-I- 1— t— H-l— H-h-h-
to
to
CJ3 •—
to
£ . 'E ^-» E cu o
,— 3 i— CU i— E • «O -O -i-
^-, tO +-> -r- -E -r- 3 E -r- 'f-
•oo to o o °^,to"£'o
-3= i — E-r-OO Xt/)T3 fOfOtOl£>
E 5 O T-OO+J-MCUCUQ-S-CU LUOO-I-..
•r- ^I4JU1 O^: EEE+J-t-(OlOOOO
O 3D- Xi — S- ^3 E E -r- 4-> •!— fO E X X -r- -r- -r-
O O3i — XStOOOOXCU-E-MSCUCUEEE
i — i
E
o
'5l ^-li— Ir-li— Ir-HrHi-HCMCMCMCMCMCMCMCMCMCMCOCO
CU i— 1 r— lt~lrHr— It— It— It— fi— 1«— It— <<— '«— 1 i— '*— 1 «~ 1 i— « t— 1 i— I
O£
co in
CM LT>
r-H CO
o «=»-
o.o
O O
o o
«-l «-l CO LO
r> «* o ID
CT1 i-H O LO
CM r-* O CM
o o o o
o o o o
• • • •
o o o o
(/) tO
c: c
o o
O O
s- o
a> —i
HP •
3 -1-5
CO CO
to t/> to to
O O O O
o o <_> o
re
in 4-J tO
cu to a)
•— O O i—
CD O O> Cn
C »0 n3 C
«=c s- s: ec
4->
to c c: to
O O 03 O
_j o co _i
o to
•a -r-
to 3
s- o
O —1
UJ co
GO O
cu
•a o
E -a i—
3 E -a ID
cr> o o TD
CU E O E
tn JT s T—
O T3 3
f— •!- CU i-
CU
o
o
CJ) C_3 O O
en
cu
+-> t/)
-»-> E
CU O
CJ3 S
E
O
CM CO •!-
co co en
1/1
o
J«£ •!-
O E
O (O
s- cn
(O O
E E -a i—
O O E fO
S- S- O T-
> > E O
CU O) (O d)
^= ^: -r- Q.
O CJ Q CO
«d- LO 10 r--.
co co co co
D-15
-------
•o
cu
•o
3
i—
O
E
O
O
.
t— 1
1
Q
CU
_C\
to
J—
t* • iffr f^^
§o
cu cu
E S- E S-
O) 01 «»- CU 3
J*: .E N to
3 -P i. E 0
CU fO (O CU Q.
_J (U CU -O X
-o >> cu
(O
•o
0
s:
>fc
•p
E
o
0
E
o
•p
(O
o
o
'
>}
-p
•r-
O
CU
•p
to
40
co
*— -*
c
o
•i—
O1
CU
S-
>}
i O
*«_rf-
>^
c
fO
Q_
CJ>
•p
E CU •
(0 -a o
•— 0 E
D- O
O i-l
CD CO
o o
O 0
o o
o o
E E
O O
C_3 C_3
cu
QJ
CT^
E
S-
o •—
i (j
Ol
CU
r— • E
CU-— O
O) E 10
E 0 S-
«=C crt eu
s--o
W tO E
O 0 CU
^C ""3
C3 ?^
GO
I— -J
CO «C
o ^
^— »r- l-H
fO E ?*^
o cu a:
— - -r- .E LU
-0 E 01—
cu cu
3 -E CU LlJ
E O 01 CD
•i- O «C
-P o s- a:
E O -P O
O -P E 1—
o -r- o co
* — "^ ^" vx
X _J
t— 1 ~~^
CQ
E
0
5 00 CT> H-I
cu ro oo i— <
D: «-H t-H »— 1
fv^.
o
00
o
o
o
h-
CO
^
E
cu
SI
O)
cu
1 —
1-^
"*
^
0
o
Qi
tn
cu
cu
O '
<=c
Q-
r^i
IO •
E 0
•t- E
E "— '
«^_
CU •*
t- W
CU
E 0
O -i-
T3 >
S- S-
O 0)
CD CO
t— i i — i
i — t >
E E
0 0
•r— O *r—
03 *5j~ 01
CU «— 1 CU
o: o:
o
1^ O <—l
CO CM f~-
O CD OJ
CM »-< O
O O 0
O O O
O CD C3
h- 1— H-
CO CO CO
CQ CO CO
ol 01-01
•r* *r~ 'r*
i- S- S-
s- s- s-
(O fO
-------
I
Q
CD
U_
co
O
£
• .
to
o
J^ s_
o cu
fj cr
3
to to
CO E
O
*o cj
cu
•P CU
CO E
E CU
OV N
«r— E
CO CU
cu JQ
T3
II
CU
si .
(O CO
E
CO O
cu
O S-
£ cu
0
CU 3
-ET3
1— O
S-
Q_
E E
CO CU
i — N
O- E
"O t""1
3 r™
Or—
S- fO
to E
CO
E II
0
'•P E
CO CO
•p •
E S-
CU Q.
O
E «
0 S-
o cu
o
CU 3
ET3
CU O •
N S- r-
E Q- CO
CU E
.Q CU *r-
E E
CU CU S-
•P IM CU
fO E -P
•r- .O CU
•P CT>
CO CU tO
cu cn &-
S- 0
O fO -P
"O II ^*
CU r—
to • 3
3 O) -O
C7>
1 — 1 II
cu
T3 • I—
o s- co
s: a. DQ
42
H-
04 -C
>k
-0 O
E
O E
JC O
CO •!-
to to •
CU S- OJ
S- -P
3 Et-
a> cu o
•i- O
t- E 5-
o o
O5 CJ -P
c: o
•r- -a to
TJ CU M-
> fO CO
•r- E
T> -r— E
•P -r~
•O CO JC.
E CU -P
fO •!—
CU 3
OViE
E -P CU
>> CO CO
• — CU
•r- 3 «•
+-> l/l +->
i — t/l S-
3 CO O
£ Q-.
-P CU
>» CO S-
JD iE
-P CD
CU >— CO
E tO
•r- > CU
CO S- i£Z
4-> QJ 40
IQ' -P
O E E
CU
s- cu -o
to o cu
E to
•fJ CU
E *a 3
CU T- O
O 1— «/l
S- E •!-
CU O "O
Q-O
CO
ID -P fO
cn E
CU «
H- o -a
O S- CU
cu co
CO Q. O
-P Q-
•I- LO X
E cn cu
i — CO CU
CU E CU
O *r- 3
E
cu co to
-O-P s-
•r- i — CU
<4- 3 -1^
E CO S-
o cu o
0 S- 3
_a
•
o>
E
-a
E
3
o
-p
cu
*o
CO
cu
o
CO
1 E -O CJ -P
0 CU E -r- E
co E CO-PCX
CO +J CO i —
c* \^* *\ ty^ .{_} CU *^ O ^~
-PCOECUEEfO-PlO
•i— CU CO O O
cOS-EEi — cUCO-i-
•i— T- a. >> to 4-> E
E to co i— 3 -i- cu
CU T— CU tOi — E fl3 M— -E
JiiE'1^ CUfOOOCUO
3 CU 3 •!- O E
CUJ^i — -P-r-M— OCU "
r— 3OEEO-PiCltO
CU E -r- CU CU
c^— r— • VP— co if^ E ''O CU *r~
O -P O O CU -E S-
— o to co cj •> T—
CO: E ^= E CU S- -i— E M—
tO O 3 3 'r^ CU *O O CU
O vr- S- 'O E -r- S-
+o •>• CU CU «i— T- -P
CU tO E S- E to -r- E
iC i— 0 to -i- E ET3 3
+JO-1— H-OCUT3CU
Q--P CU CU O CU to r—
iE tO tO S- S- -Q O
-(-> S- IT- CU CO E S-
•r-4-> 3iE E CU Ot-H-P
3 X Q-.-I-3 3 T3 to CU
cu o cu 3 r— a.
*O *^> •* r— r~~" fO •
CU E TJ O CJ CO £
+JtOr— ES-ECOEO
fO fOO-P-i— cCOS-
•i— • E S- O CU -E •{- <4-
CJ O CU CU Q- CO 4->
O E CO CU CU to CO
tona cu q- to j_ s_ E
tOCUOV OfO3S-O
> >y O ^£5 10
CO' *"*> ^E JE -1-^ ^— O_ fO^ CO
.r* +J 4-> •!— O X 'r-
•O r— E >CUOtOT-S- (OCU
•P -P -P CU O CU CU £
sr to jc: '•— ja c: o c
•i— i — C>-£CUtOO
tO3S-3 3NOJ3
•POCU CU. E E E S-
S-i — ^^CUJE CUOfO
CU tO S— •^* 4J CU t^^ &- CJ •
O O O -P iE JE O "O
E 3 E +J 0 S- CU
3 CU TO •!-• • T3 T-
$_ H- E •> to >M-
r- 3 r— CO E -P
iQ 4-> " Q_r— CU -P S- E
cO co ^. co O ^C. -^ E CU co
S-.CUOCUCU 3O>:E3
CUf)-E+JQ_ CUE-PO"
~T3 CUEE
o id — o^: o cu >-i—
CJ ^-3 cO E 3 t^ • CU f~~ CU
£ X -r- -0 r- S-
^) « « O CU O co
fO -P CU *O -P •« CU S-
-c co -M cu CUCU;E-PCO
^_i t— Vp— U> t/5 S- i- -P E l~~
•r-^ifOi— tB3 OtO
T3U-3 CU COTDCJE
CU CU > C/l O E T—
N >>;E CU i — O- tO t- E
•r- • -C 4-> i — COX Oi-
E CU -P E CU •» CU
D5i — i — « CU T- CO E -P
O •*""* to CO E E CU to O
OtOCU-PCUS-E-r-'r-CU
CU *P ^^ E N CU CU E •$-* CT>
S- cOE-PNCUfOfO
CU fO 4— CU E Q-i — S-
CU ^" E ^ CU CU O 3 O
.a -P ^E T- cn^a -P o.-i->
4-> CU fO >> O CO
'{J g~* •[— «\ C*" ^-. t^— ^) j^
i— -i-SE-POO -^
3 CU -P "i— i—
OS-T3 S- cnc04-> co CO3
^* (^ QJ "i^j y ^j -r~ t- ^T^
to CU-Pi — T- -^ CU E CU
Q- CO -i- -P r— <*- CU E~O
-P Q.-I- -E CO 3 M- E CU E
i— icOOOE-QCUtOCncO
-a
D-17
-------
(1/3.2)
a and b
i =
D. and D. = distances from plant to outer edge (D. ) and inner
edge (D. ) of area i (e.g., for the area 5-10 km from
the plant, D. = lo'km and D^ = 5 km).
factor converting ug/m3 to ppb, the units in which R
is expressed.
values describing the dispersion pattern of benzene in.
air around Plant X, according to the equation B. - a D.,
in which B. is the benzene concentration at distance D.
from the plant. Values of a and b are unique to each
annular area i around each model plant.
the particular area in which p. occurs (i progresses
from the area 0.5 to 1.0 km from the plant to the area
10 to 20 km from the plant).
I = summation of deaths per year from all areas (i).
This equation is a mathematically rigorous method for estimating the
exposure to the population within any area between i, and i"2 km from
the plant, taking into account that with constant population density (p.)
more people reside near the outer edge of the area than near the inner
edge, and that the benzene concentration (B.) decreases with distance
from the plant. The equation is derived in Reference 8.
Values of a and b were calculated for each annular area associated
with each model plant as follows:
ln(B1 /B. )
b = ln(D, /D. )
(2)
a = B, /(D, )
b
(3)
B
in which i-. is the benzene concentration at the inner edge of area i
D B
(i.e., at distance i-.), and i'2 is the omnidirectional annual average
benzene concentration at the outer edge of area i (i.e., at distance
i'2). B. values for each distance (D.) from each model plant are listed
in Reference 9.
D-18
-------
Population density (p..) for a particular annular area around a
particular plant is obtained by dividing the total population in that
area (P.) by the area in square kilometers; i.e.:
Pi - P,/[n(D,2 - D.2)]
i i i2 !-,_
(4)
P. values for each plant and annular area are listed in Reference 8.
In summary, for each annular space around a particular plant, the P.,
D
'
D
values are taken from Reference 8. B. values at all distances
(D.) are taken from Reference 9. Values of b, a, and' p.' are calculated
from Equations 2, 3, and 4, respectively, for each annular area. Then,
using Equation 1, exposures in ug/m3-person years are calculated for each
annular area, divided by 3.2 to convert ug/m3 to ppb, the units in which
R is expressed, and multiplied by R to yield the number of deaths in each
annular area. These deaths are summed to give D , the annual leukemia
deaths attributable to benzene emissions from Plant X.
The total estimated number of leukemia deaths per year attributable
to benzene emissions from all plants was determined by the equation:
Total estimated number _ _
of leukemia deaths/yr (D.) = D., +
D
143
(5)
The total numbers of estimated leukemia deaths attributable to
benzene emissions from existing benzene storage tanks are given in the
last column of Table D-l on a pi ant-by-pi ant basis, in deaths per year,
assuming the baseline is effective as discussed in Section 3.3.1.1. The
number of deaths expected under each of the control options can also be
derived using the same methodology.
D.4.3 Example of Leukemia Death Calculation
Plant no. 4 from Table D-l was chosen for an example calculation of
the number of leukemia deaths attributable to benzene emissions from
benzene storage tanks. For a determination of the number of deaths
according to Equation 1, numerical values are needed for R, a, p., i,,,
D
i,, and b. In turn, for a determination of p. from Equation 4, the
numerical value of P. for each annular area must be known. For a
D-19
-------
determination of b and a from Equation 2 and Equation 3, respectively,
B B
numerical values of i., and i'2 must be known for eac!i distance.
Calculations are shown in Table D-2. The valuer; in the first three
lines of Table D-2 are common to all plants. They show the distances at
which concentrations and populations were measured and the risk factor
(R). Line 4 shows the population (P.) in each annular ring, obtained
from Table A-4, Reference 8, for Plant 4.
Lines 5 and 6 show the benzene concentrations at various distances
from the plant for the applicable model. These are found as follows:
Table A-l, Reference 8, indicates that the "large producer" model plant
applies to Plant no. 4. Table B-l, Reference 9, indicates the omnidirec-
tional annual average benzene concentrations by distance for this model.
Note that concentrations for 1, 5, and 10 km from the plant apply to the
B B
outer edge ( i^) of one ring and the inner edge ( i-,) of the adjacent
ring.
Lines 7 through 11 show the calculations. These are shown below for
the outer ring. From Equation 4:
!* - Di)] = 287,456/D*(202 - 102)] = 305.0
From Equation 2:
b = ln(Bi2/Bi1)/ln(Di2/Di1) = ln(0. 0342/0. 0903)/ln(20/10) = -1.401
From Equation 3:
= Bi2/(Di2) * 0.0342/C20)"1'
-1.401 _
All the values needed for using Equation 1 are now available, so:
Deaths for 10- to 20-km ring 01Q_20 = (R/3.2)2np.a(Dl£*2 - Di|j+2)/(b+2) ,
D10_20 = (0.34 x 10~6/3.2)2n(305.0)(2.27)(200-599 - 10°'599)/0.599, and
D1Q_20 = 0.001580
D-20
-------
^J"
I—
z
«=C
_J
CL,
•*
CO
pi
UJ
o
•a:
ul
^
UJ
1 1
o
h—l
^~
"^
— i
0
S
UJ
1
a.
X
UJ
CM
1
Q
CO
!Q
*""
CO
•$•)
CO
73
If.
o
CO
u
S-
o
GO
•^"•^
E
^1
4J
E
CO
r—
Q.
73
E
3
O
s-
CO
CD
•^w
s-
S-
CO
E
E
CD
CM
1
O
1—4
O
I— 4
1
LO
LO
1—4
r—4
1
r—t
CD
S-
o
u
CO
u_
CO •
E O
•r- E
~"
1 1 1
h-f— r- r—
* * E CD CO
CO CO * CO ""O "73
CM 00 r— CD O CD O
* * ^ SjCO S^cO
(/)(/)•• • •• •
CP- M- «4_ Cj_ S- «4- j_ Q) Ci_S-O)
CO CD CO COO COOCDCOOCD
cr o: a: o; q- o:cr._j a: M- _j
ID
1
CD O O VO CM CO
CM i-4 r-t LO «* O
^J" CO CD
X « CD O
^^ * *
«* CO O O
CO CM
•
O
«0
1
O
O LO 1-4 «* CO
i— < IO
CD ^d" CO
LO t-4 X CD CM IO
•» • •
«* CM CD CM
CO CM
O
10 •
O
r- 4
O
X CM 5 co
— — E O « •>
CM i—4 O CO CO CO
•r— -p- CO S— CD CD
Q C3 S- CO -O T3
CO O. CO CO
. •» « Q. ^_^
CO CO 1 CD CD
CD Dl J3 CD E E
TD TD 0. E -p- *r-
CO CO CL -p- S- S- OO
eC CD CD CO S- S- O
1— EEJ=S- CO CO i— i
«=t -p- -p- 4-> CO 4-* E 1—
^ t ^ -i "i ° ^ 3
^ s- S- 73 E ZD
UJ CO CO - — E 4-5 4-> O
^" 4~* E CO CO CO — -1
>— I 3 E S- T- 0 0
O O O E E
4-> 4-> CO E O •— ^ O
*t- O O CO CJ CO
CD CD -r- E E
EECOcO ECD ECD
cOcO-p-p- C03- C03-
4-> 4J S- 3 N N
CO CO Q. E CM E t-4
•P- «p- •» O CD "P" CO *P™
OOQ£a_ CQ CQ CQ CQ CQ
i-4 CM CO «5f LO 4_> +j J3
CO CO CO CO co
3 333 3
cr cr cr cr cr
UJ UJ UJ UJ li I
0
CO
LO
t— 1 1—4
O l>» O
CD <3- CM O
LO i-< CM O
0 1
CO
CO
LO
O
•Sf CM O
O «3- LO O
LO i— 1 CM O
O 1
CO
LO
vo
CO
CD t-1
r^ 10 o
CD •* CO O
LO t-4 CM CD
co
jQ
O
CM
CM O
J~t CD CO O
> 1 i-l JD E .O CM r—
4-* . «p- •- — «p- »r— CO
•P- CM CM CQ CM Q . — »
CO T- -^. -p- S- ~— E S-
EQ CMQ COCO -P->>
CO T- — ' CO -I- ^
"O S= CQ — » >> Q- CO CO
1 — I — - CM t= J= J=
E' E-r- S-CM 4->4->
O T- p- CQ CD •— •• CO CO
•P- O. Q- CM CO CO
4-> II II • T3 T3
CO II CO CO '
P— -Q CO JZ ~~. p—
3 M- 4J CU CO X
0. Q- CO ^-^ 4-> Q
O «» .. <|J Q
CX J3 CO C3 I—
r-^ oo CD o 1-4
t— 4 T— 1
X
•r—
E
CO
r>
O.
CO
CO
JZ
tj
O
t-4
t
Q
CO
|i ••
(O
I —
c^
•P>
•a
c
to
ft
CO
CD
O •
E CO
CD t-
S- +3
CD CO
CH co
CO t3
D£
* E
l~4 CO •
"^C •!— •
CL. "5
CD E
P- CD CO
-O E CL
CO •!— CL
r— 4J CO
CO
E P- CO
•i— ^5 •p.
E P^ 5
s ^
O c_> if-
•^ E °
•p— [
CO 1
S- LO Q
CO •
4-> i —
E u _a
CO ^cp
"el -r-
o •»
co >K co
CD P-/ S-
E CO
S- o CO
O Q.
S- CD CL
E CO
CO T-
CD S_ p-
0. Kl
>> CO 4->
4-> -r- 0
*CO CO
73 S_ .p-
O^* ^~
-i- \J Jfc-
s: u- H-
CO JD O
D-21
-------
The same calculations are made for ranges 0.1 to 1 km, 1 to 5 km,
and 5 to 10 km, but there is one modification. In the calculation of
population density (pQ ^_^ and number of deaths (DQ j^), Di;L is 0.5 km
(not 0.1 km) because the population is assumed to reside in the area 0.5
to 1 km from the plant. In the calculation of b and a, i, = 0.1 km.
The total annual leukemia deaths for the plant (D )_are the sum of
J\
the deaths for each ring; i.e.:
Deaths for Plant 4 (D"4) = I(DQ l_I +
D
5_1Q + D10_2Q)
D~4 = 0.000240 + 0.001365 + 0.001058 + 0.001580
D^ = 0.004243 deaths/yr
Deaths attributable to benzene emissions from any of the 143 plants
may be calculated in the same manner.
D.4.4 .Estimate of Leukemia Risk
The estimated leukemia deaths shown in Table D-l are. based on estimates
of omnidirectional annual average benzene concentrations around benzene
storage tanks. Because atmospheric dispersion patterns are not uniform,
some population groups will receive above-average benzene exposures and
will, therefore, incur a higher risk (or probability) of developing
leukemia.
Maximum annual benzene concentrations were estimated as follows:
For each model plant, the benzene storage tanks were assumed to be in a
straight line parallel to the most prevalent wind direction, in order to
maximize calculated annual average concentrations. The most prevalent
wind direction in Lake Charles is south and north; thus, the storage
tanks in each model plant were placed in a straight north-south line to
maximize the combined effect of tank emissions on ambient air benzene
concentrations.
Maximum annual risk is the estimated probability that a person who is
constantly exposed to the highest maximum annual average benzene concentra-
tion in the ambient air around a particular source for 1 year will develop
leukemia because of exposure to benzene emissions from that source.
Maximum lifetime risk is estimated by multiplying the maximum annual risk
by 70 years.
D-22
-------
D.4.4. 1 Example of Leukemia Risk Calculation. The maximum lifetime
risk of leukemia was calculated for a person, who was assumed to reside
at the point of highest maximum annual average benzene concentration
outside the model plant with the greatest benzene emissions from benzene
storage tanks. The maximum risk of leukemia associated with these emissions
is calculated as follows:
First, the highest maximum annual average (MAA) benzene concentration
associated with benzene storage tank emissions from any of the model
plants is selected from Table 8 in Reference 9. This concentration, 16.8
|jg/m3, occurs 0.1 km from the plant boundary of the "large benzene producer"
model plant. This model plant has a benzene production capacity of 224.6
x 106 liters/yr (Table 3-2).
Second, the benzene producer with the largest existing capacity is
selected from Table 7-7. This producer, which is listed as Plant no. 4
in Table D-l, has a capacity of 700 x 106 liters/yr (185 x 106 gal/yr).
The benzene concentration (16.8 ng/m3) based on the model plant capacity
of 224.6 x 106 liters/yr was scaled up proportionately to the existing
plant capacity of 700 x 106 liter/yr; i.e.:
Actual MAA benzene cone. = (model benzene cone.) .<«%;] plant cgality') •
Actual MAA benzene cone. = (16.8)(700 x 106/224.6 x 106), or
Actual MAA benzene cone. =52.4 ug/ms.
This figure is converted from pg/m3 to ppb by dividing by 3.2:
Maximum annual average benzene concentration = 52.4/3.2 = 16.4 ppb.
The result, 16.4 ppb, indicates that the person most exposed to
benzene from any of the 143 plants, assuming he or she resides 0.1 km
from the boundary of Model plant 4, receives an exposure of 16.4 ppb
2
continuously, or for 1 person-year annually. By applying the risk factor
of 0.34 x 10 deaths per year per ppb-person year to this exposure, the
annual risk can be calculated, viz:
Hsk Semia = (0'34 X 10"6 deaths Per year/ppb-person year) x
risk of leukemia (15.4 ppb-person years), or
D-23
-------
Maximum annual -6
risk of leukemia = 5.58 x 10 .
Because lifetime risk is expressed as a probability to one person of
dying of leukemia, the units have been deleted for convenience. Techni-
cally, the number represents deaths per year for one person. The lifetime
risk of leukemia, assuming a 70-year lifespan, is simply 70 times the
annual risk, or:
Maximum lifetime = (5 5g x 10-6)(70) = 3 9 x 1Q-4
risk of leukemia v yv
The risk associated with the emissions from any specific plant or model
plant may be calculated in the same manner.
D.4.5 Validity of Estimates
Several uncertainties exist in the estimated number of leukemia
deaths and the maximum leukemia risk. Primary sources of uncertainty are
listed below:
• Risk factor (R),
• Populations at risk,
• Estimated benzene concentrations around plants, and
o Benzene exposure calculations.
Uncertainties in the risk factor (R) are discussed in Section D.2.3,
and uncertainties in populations "at risk" (P.) are discussed in Section D.3.
The other factors are discussed below.
D.4.5.1 Estimated Benzene Concentrations. The estimated benzene
concentrations are derived from several factors which follow:
• Configuration of the model plant,
• Emission rates from the model plant, and
• Dispersion patterns of the emissions.
Uncertainties associated with these factors could not be quantified,
but their qualitative effects on the estimated number of leukemia deaths
are discussed below.
The configuration of the model plants assume from two to seven
benzene storage tanks in a north-south line, with a center-to-center
spacing of 183 meters. Current benzene emission rates from various type
D-24
-------
and size storage tanks were estimated and uniform emission rates assumed.
Four emission models were used, and each model was matched with an actual
plant. No corrections were made for differences between actual and model
plant capacities (except in calculating maximum leukemia risk).
Several sources of uncertainty occur in the dispersion model.
First, it is unlikely that any plant duplicates its corresponding model
plant precisely, so uncertainties due both to differences in actual and
model plant capacities and to assumed locations of tanks within plant
boundaries may be expected. Second, the model used 1973-1976 weather
data from Lake Charles (a near coastal city) to project dispersion patterns
for all existing plants, and did not incorporate the effects of terrain.
Thus, when.applied to hilly, inland areas, the model may introduce
inaccuracies. Third, the model assumes there is no loss of benzene from
atmospheric reactions or ground level absorption. If such losses occur,
the actual concentration of benzene will be less than the estimated
values.
A final source of uncertainty is that the model measures benzene
2
dispersion only to 20 km. If the linear risk model is accurate, expo-
sures at distances greater than 20 km, however small, may be important.
If such exposures occur, the estimated number of deaths would be higher
than estimated here.
It is estimated that benzene concentrations predicted by the disper-
Q
sion model are accurate to within a factor of 2, barring large inaccur-
acies in estimated benzene emission rates.
0.4.5.2 Benzene Exposure Calculations. Benzene exposure calculations
assume that persons at specific locations are exposed 100 percent of the
time to the benzene concentrations estimated to occur at each location.
The assumption of continuous exposure to residents introduces some uncer-
tainty, both in estimated number of leukemia deaths and in maximum leukemia
risk. No numerical estimates of potential variation are available.
Furthermore, the maximum lifetime risk assumes that a particular plant
operates at full capacity for 70 years. There is necessarily a discrepancy
between the methods used to measure distance from the plant for benzene
concentrations and for populations. Benzene concentrations at the 0.1 km
D-25
-------
distance are measured from the plant boundary. This discrepancy introduces
some imprecision (<2 percent) in the "ppb-person years" benzene exposure
calculations used to estimate the number of leukemia deaths. The maximum
lifetime risk estimate is not affected.
D-26
-------
D.5 REFERENCES FOR APPENDIX D
1. U.S. Environmental Protection Agency. Assessment of Health Effects of
Benzene Germane to Low Level Exposure. EPA-600/1-78-061. September
1978.
2. U.S. Environmental Protection Agency. Carcinogen Assessment Group
(R. Albert, Chairman). Population Risk to Ambient Benzene Exposures.
January 1980.
3. National Institute for Occupational Safety and Health. Criteria for a
Recommended Standard—Occupational Exposure to Benzene. HEW Publica-
tion Number (NIOSH)74-137. 1974.
4. American Conference of Governmental Industrial Hygienists. Threshold
Limit Values for Chemical Substances and Physical Agents in the Work-
room Environment with Intended Changes for 1977. 1977.
5. National Institute for Occupational Safety and Health. Revised Recom-
mendation for an Occupational Exposure Standard for Benzene. August
1976.
6. Occupational Safety and Health Administration. Occupational Safety
and Health Standards, 29 CFR 1910.1000, Table Z-2. Publication 2206.
1976.
7. 42 FR 27452. May 27, 1977.
8. Suta, B. E. Assessment of Human Exposures to Atmospheric Benzene from
Benzene Storage Tanks. SRI International, Center for Resource and
Environmental Systems Studies Report No. 119. August 12, 1980.
9. H. E. Cramer Co., Inc. Calculated Air Quality Impact of Emissions
from Benzene Storage Facilities. Prepared for the U.S. Environmental
Protection Agency. Report No. TR-80-141-04. Salt Lake City, Utah.
July 1980.
D-27
-------
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-78-034a
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Benzene Emissions from Benzene Storage Tanks •
Background Information for Proposed Standards
5. REPORT DATE
December 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality Planning and Standards
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
1O. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3063
12. SPONSORING_ AGENCY NAME AND ADDRESS
DAA for Air Quality Planning and Standards
Office of Air, Noise, and Radiation
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Draft
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Standards of Performance for the control of emissions from benzene storage tanks
are being proposed under the authority of Section 112 of the Clean Air Act. These
standards would apply to all new and existing storage tanks having a capacity of
4 cubic meters or larger, which are to be used for the storage of pure benzene.
Existing sources will have to comply with the standard within 90 days of its
effective date, unless a waiver of compliance is secured from the Administrator.
This document contains background information and environmental and economic
impact assessments of the regulatory alternatives considered in developing the
proposed standards.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air pollution
Pollution control
Carcinogenic
Benzene storage tanks
Contact floating roofs
Petroleum refineries
Chemical manuf. plants
Equipment standard
Volatile Organic
Compounds
National Emissions
Standards for
Hazardous Air
Pollutants
Air Pollution Control
13 B
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
277
2O. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------
-------