EPA-4 50/3-74-021
MARCH 1974
DEVELOPMENT OF A
METHODOLOGY TO ALLOCATE
LIQUID FOSSIL FUEL
CONSUMPTION BY COUNTY
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
-------
EPA-450/3-74-021
DEVELOPMENT OF A
METHODOLOGY TO ALLOCATE
LIQUID FOSSIL FUEL
CONSUMPTION BY COUNTY
by
Josettc C. Goldish, Franklin D. Trowl, John R. Ehrenfeld,
Khee M. Chng , and Richard Stockdale
Walden Research Corporation
359 AJlston Street
Cambridge, Massachusetts 02139
Contract No. 68-02-1067
Project No . 1
EPA Project Officer: Charles O. Mann
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Air Quality Planning and Standards
Research Triangle Park, N. C. 27711
March 1974
If
-------
This report is issued by the Environmental Protection Agency to report technical
data of interest to a limited number of readers. Copies are available free of charge
to Federal employees, current contractors and grantees, and nonprofit organizations
as supplies permit - from the Air Pollution Technical Information Center, Environ-
mental Protection Agency, Research Triangle Park, North Carolina 27711, or from
the National Technical Information Service 5285 Port Royal Road, Springfield,
Virginia 22151.
This report was furnished to the Environmental Protection Agency by Walden
Research Corporation, Cambridge, Massachusetts, in fulfillment of Contract
No. 68-02-1067. The contents of this report are reproduced herein as received
from Walden Research Corporation. The opinions, findings, and conclusions
expressed are those of the author and not necessarily those of the Environmental
Protection Agency. Mention of company or product names is not to be considered
as an endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-74-021
11
-------
TABLE: or CONTENTS
Section Title Page
I INTRODUCTION 1-1
A. Background 1-1
B. Study Objectives and Limitations 1-1
11 SUMMARY 2-1
A. Data Base 2-1
B. Method 2-11
1. Completion of State Data 2-12
a. Residential 2-12
b. Commercial 2-13
c. Industrial 2-14
d. Heavy-Duty Vehicles (HDV) 2-14
e. Light-Duty Vehicles (LDV) 2-15
2. County Allocation of Fuel Oil 2-15
a. Residential 2-15
b. Commercial 2-16
c. Industrial 2-16
d. Heavy-Duty Vehicle (HDV) 2-17
e. Light-Duty Vehicle (LDV) 2-18
C. Results for 13 Selected Counties - 1970 and 1971. 2-18
References 2-26
III DISCUSSION OF METHODOLOGY 3-1
A. Residential 3-1
1. State Fuel Oil 3-1
2. County Fuel Oil 3-3
3. Additional Data 3-4
B. Commercial 3-4
1. State Fuel Oil 3-4
2. County Fuel Oil 3-5
111
-------
TABLC OF CONTENTS (Cont.)
Section Ci t. h- Pane
C. Industrial 3-10
1. State Fuel Oil 3-10
2. County Fuel Oil 3-11
D. Heavy-Duty Vehicles 3-11
1. State Gasoline and Diesel - HDV 3-11
2. County Gasoline and Diesel - HDV 3-14
E. Light-Duty Vehicles 3-17
1. State Gasoline - LDV 3-17
2. County Gasoline - LDV 3-17
References 3-20
IV SULFUR CONTENTS AND SEASONAL FLUCTUATIONS 4-1
A. Sul fur Contents 4-1
1. Sulfur Content Reported for NEDS Point
Sources Using Oil : 4-1
2. Burner Fuel Oils, MIS, Bureau of Mines 4-4
3. Fuel Oils by Sulfur Content, MIS,
Bureau of Mines 4-4
4. Oil Availability by Sulfur Levels, Bureau
of Mines, 1971 4-9
5. Import Supplement to Oil Availability by
Sulfur Levels, Bureau of Mines, 1972 4-9
B. Seasonal Fluctuations 4-11
References 4-17
V COMPUTER PROCESSING 5-1
A. Introduction 5-1
B. Program Descriptions 5-4
1. The WALDEN PREPROCESSOR Program 5-4
2. The WALDEN COUNTY FUEL OIL ALLOCATION
Program 5-4
3. The'WALDEN RESIDENTIAL/COMMERCIAL SEASONAL
SUMMARY Program 5-7
4. The WALDEN INDUSTRIAL SEASONAL SUMMARY
Program 5-7
IV
-------
TABLE OF CONTENTS (Cont.)
Section
Title
Paae
VI RECOMMENDATIONS 6-1
A. Improvement of Data Base 6-1
1. Bureau of Mines Fuel Oil Sales Data
(Annual) 6-1
2. Bureau of Mines, Burner Fuel Oils Data
(Annual) 6-4
3. Census of Manufactures, Fuel and Electric
Energy Consumed 6-5
4. Highway Administration Data 6-5
5. R.L. Polk Data 6-5
6. Point Source Data 6-6
B. Improvement of Present Study 6-6
1. Choice of Fuels 6-6
2. Linear Correlation Between Commercial Fuel
Oil Use and Socio-Econonric Data 6-6
3. Rural/Urban Driving Patterns 6-7
4. Excise Tax Data 6-7
5. Truck Vehicle-mile Data 6-8
6. Re-examination of Assumptions 6-8
7. Re-examination of Regression Coefficients ... 6-8
References 6-10
APPENDIX A RESIDENTIAL FUEL USE A-l
APPENDIX B REGRESSION ANALYSIS OF COMMERCIAL USE OF OIL FOR
VARIOUS SUBCATEGORIES B-l
APPENDIX C REGRESSION ANALYSIS OF URBAN VS RURAL DRIVING
PATTERNS ON A COUMTY-BY-COUNTY BASIS C-l
APPENDIX D ANALYSIS OF AN ALTERNATE METHOD TO ALLOCATE MOTOR
FUEL USED BY HEAVY-DUTY VEHICLES ON A COUNTY-BY-
COUNTY BASIS D-l
-------
LIST OF FIGURES
No. Description Page
3-1 Comparison of Various Fuel Consumption Rates for Vehicles
of Various Gross Weiqhts 3-15
4-1 Summary of Availability of Sulfur Content Data from Local
and State Air Pollution Agencies 4-10
5-1 Computer Processing Phase for Fuel Oil Allocation
Program 5-2
5-2 Area Source Coding Form 5-6
5-3 Flow Chart of Countywide Fuel Oil Allocation Program 5-8
5-4 NEDS Point Source Coding Form 5-9
LIST OF TABLES
No. Description Page
2-1 Sources Required for Input Preparation County-Wide Fuel
Oil Allocation Programs 2-2
2-2 Sources Contacted for Vehicle Mile Data ..: 2-5
2-3 Comparison of Reported Fuel Oil User Categories and
Required User Categories 2-11
2-4 1970 Results for Selected Counties 2-19
2-5 1971 Results for Selected Counties 2-20
2-6 Seasonal Fluctuations of Fuel Use for Residential and
Commercial Spaceheating - 1970 2-22
2-7 Seasonal Fluctuations of Fuel Use for Residential and
Commercial Spaceheating - 1971 2-23
2-8 Summary of Seasonal Fluctuations of Residual Oil Used by
Industrial Sources 2-24
2-9 Sulfur Content Averages for Distillate and Residual Oil
Burned by Area Sources 2-25
VI
-------
LIST OF TABLES (Cont.)
Np._i_ Description Page
3-1 Summary of Estimation Methods for Residential Fuel Oil
Used for Spacehea ti ng by State 3-2
3-2 Regression Results for Fuel Oil Used by Comrrercial
Subcategories 3-6
3-3 1970 Commercial Coal, Oil, Gas Use by State 3-8
3-4 1971 Commercial Coal, Oil, Gas Use by State 3-9
3-5 Nationwide Fuel Intensity Ratios 3-12
3-6 Average Miles Per Vehicle by % Rural Categories 3-19
4-1 Residual Oil Consumed by Point Sources by Sulfur
Content 4-2
4-2 Percentage of Residual Oil Consumed by Point Sources
by Sul fur Content 4-3
4-3 Sulfur Content Analysis Based on Bureau of Mines Data 4-5
(1970) 4-5
4-4 Sulfur Content Analysis Based on Bureau of Mines Data
(1971) 4-6
4-5 Imports of #4 Fuel Oil by Percent Sulfur Content by
States: Jan.-Dec. 1971 4-7
4-6 Imports of Residual Fuel Oil by Percent Sulfur Content
by States: Jan.-Dec. 1971 4-8
4-7 Agencies Compiling Sulfur Content Data 4-12
4-8 Seasonal Fluctuations of Fuel Use for Residential and
Commercial Spaceheating - 1970 (Based on Degree-Days) 4-13
4-9 Seasonal Fluctuations of Fuel Use for Residential and
Commercial Spaceheating - 1971 (Based on Degree-Days) 4-14
4-10 Summary of Seasonal Fluctuations of Residual Oil Used
by Point Sources 4-15
A-l Estimation of Distillate Oil Consumed for Residential
Use - 1970 A-3
VI1
-------
LIST OF TABLES (Cont. )
No. Description Page
A-2 Estimation of Residential Consumption of Distillate
Oil for Spaceheating - 1970 ' A-5
A-3 Correlation Matrix A-8
A-4 Residential Distillate Oil Consumption Based on Multiole
Regression Equation A-8
A-5 Estimates of Distillate Oil Consumed for Home Heating
Eased on a Per Housing Unit Requirement According to
Average FT? '. ". A-10
A-6 Summary of Estimation Methods for Residential Fuel
Oil Used for Spaceheating A-ll
A-7 Gas Companies Which Provided Residential Gas Data A-12
A-8 Residential Gas Use Results A-14
A-9 Correlation Matrix for Six Groups Combined A-16
A-10 Residential Distillate Oil by County - Maine A-17
B-l Room/Employee Ratios for Hotels by State B-4
B-2 1970 Commercial Use of Fuel Oil in Massachusetts B-5
C-l Results of Regression Analysis to Determine Differentia-
tion in Rural Vs Urban Driving Patterns C-l
C-2 Driving Patterns for Selected 100% Rural Counties in
Georqi a C-3
C-3 Average Miles Per Vehicle by % Rural Categories C-4
D-l County Fuel Consumption Results for 1971 Using Interstate
Vehicle Mile Methods D-5
D-2 County Consumption of Motor Fuels by Heavy and Light-
Duty Vehicles Using Existing Methods - 1971 D-6
Vlll
-------
I. INTRODUCTION
A. BACKGROUND
The Environmental Protection Agency has developed an extensive
nationwide data base providing air pollutant emission estimates for area
and point sources. The data in this National Emission Data System (NEDS)
need to be updated on a regular basis to provide significant statistics to
the EPA and other branches of government. The data on point sources will
continue being updated by means of legal reporting requirements for point
sources, using questionnaires and/or personal contacts with the establish-
ments concerned. It was found necessary, however, to obtain a more routine
method of updating the area source data.
This project includes the development of a methodology, whereby
annual fuel oil consumption by stationary sources and motor vehicles can
be collected by the EPA on a continuing basis, and allocated to individual
counties. The resulting county-wide fuel oil figures will serve to update
similar figures of previous years, presently available in NEDS format. It
is expected that up-to-date air pollution emissions from area sources burning
fuel oil can be derived from these consumption figures.
E. STUDY OBJECTIVES AND LIMITATIONS
The purpose of this contract effort was to develop a method whereby
annual fuel oil consumption data could be routinely collected by the EPA on
a continuing basis and allocated to individual counties with a probable error
of 10 percent or less. The methods which have been developed for each
county determine distillate and residual oil consumed by industry,
1-1
-------
commerce and residential units, as well as gasoline and diesel consumed by
light and heavy-duty vehicles. For the purpose of this study, residential
units in structures of 20 units or more were considered under the commercial
category. Furthermore, it was assumed that all on-highway use of diesel is
consumed by heavy-duty vehicles.
In addition, seasonal fluctuations of fuel oil use were studied by
user category and geographical region, and a collection of references on
sulfur content in fuel oil were analyzed.
The first phase of the study developed the methodology to be used,
and applied the methods to selected counties. These counties were selected
to include examples of urban as well as rural areas, high as well as low
fuel oil use areas, a variety of climates and regional economic structures.
The 13 selected counties were:
Bel knap, New Hampshire
Franklin, Massachusetts
Worcester, Massachusetts
Baltimore, Maryland
Palm Beach, Florida
St. Louis, Missouri
Minnehaha, South Dakota
King, Washington
Galveston, Texas
Jefferson, Alabama
Boulder, Colorado
Los Angeles, California
San Diego, California
The second phase included the collection and processing of 1972 data
for all counties. This report describes the results of the first phase of
the project. The results of the second phase are available in the form of a
computer listing and magnetic tapes.
1-2
-------
The basis of the approach used by Wai den to arrive at the county
allocation methods was the development of linear correlations between fuel
use and other demographic and economic factors. Statewide fuel oil data
were distributed to the counties within each state, based on the developed
correlations.
The limitations of the resulting methods are summarized below:
(1) Statewide fuel oil data were found to be incomplete and not
available by the categories required for this study. Methods had to be
developed, therefore, to reduce the available information to the required
categories. This meant that, aside from the inaccuracy of some of the
published state data, an additional error factor was introduced in the
figures which were used as the basis for the county-wide allocation methods.
(2) It was hoped that independent checks would be possible for
several of the 13 selected counties. Vial den was only able to perform rough
checks on the resulting figures, since the data which would permit a more
accurate validation process were not available or could not be made avail-
able.
(3) In the majority of the states, it is estimated that the developed
county-wide figures have an accuracy of 10 to 15 percent. In some states,
however, the statewide figures were quite inaccurate, making it practically
impossible to determine the accuracy of the developed county figures.
1-3
-------
(4) The fuel crisis confronted by the United States during 1973-
1974 has resulted in significant changes in the fuel consumption patterns
across the nation. Lower thermostat settings in homes and businesses,
shorter working hours, and the unavailability of motor fuels as well as
lower driving speeds will have altered some of the numerical values of the
historical correlative relationships developed here.
1-4
-------
11. SUMMARY
A. DATA BASE
Whenever possible, Walden has attempted to use data which appear
annually or more frequently. Table 2-1 summarizes the major sources used.
Table 2-2 specifies the contacts made with the various state highway admin-
istration offices throughout the country.
Other data sources are referenced throughout this report, but
the sources in Tables 2-1 and 2-2 are essential to prepare the input to
the computer programs containing the methods developed to allocate fuel
oil to individual counties.
The data base is divided into two major categories:
(1) state data
(2) county data
The state data include fuel oil figures, as well as socio-economic and
demographic variables. The county data include solely the latter. The
basic state fuel oil figures used were taken from the Bureau of Mines
Annual Fuel Oil Sales publication, for stationary sources, and from
Highway Statistics, published yearly by the Federal Highway Administration,
for mobile sources.
The seasonal fluctuations data were taken from climatological
reports (1) and summaries made from NEDS industrial point source data.
2-1
-------
in
Z. 5j
4 C?
h-0
«< o
ix
e£o
\
!5 o
1 Q.O
CM :=: -J
y «**
>^
fj
_ri
4J
o
f i
1.
£
XI
u
Vh
CI
2S
19 «
r- »-
X rt
0 >
ex
O. 01
** <3
0
V
u
V.
o
o
u
1
c~- .
Gl «
t> f
"- r
r «
Vt
VI 1
o -
t*- *
-1
r
_C
01
.r
ca
- ro
: -j- 01
/ vj <-
o f > »-
< i -
i^ -
- lf> 0
/j i.' *
i zn
: ^ c c
- f> -r- <
". C> 4» «
l~- C C
^- ^_^ * 1
J i-
o_
T1 CI 4 -> <
-i v: c:
1- 11 O <
1 ^ K
VI
:>
i
\i
_>
->
-^ cu fu c: *
VI CI O
Ql
I-
IU i
~}
i
_J-
-~
no '>
a \ o
r* CD
Irt
i~ -
> o in
a o ri:
i-*
O)
=
yj
o
«"
f>
Ci T
1"
f J
- 4
10
-
Ci
11 t
0
.t C
_ i
- *-» *-
Lt C~» «J C
O CJ " r
.. c»» irj *
( > i c>
r ' ' , > C1^ < *
LJ 1"^. ») V- *
- O ^xj C>
1> t" I
*~ » J--
> O t J
V^J- *J
> * o
f.
f
r.n
0
* » i
ti *0
»
O.J
~j <"5 i
CJ "-) O
i- to t -
Cvl t "
- O
I- - C>
OJ (_>
1-* » V-
r: i - o
O C.1 CJ O
n o 4-» o +J
fX
tri
1/1
J-
o
* V- k" fX
!-. (^ O O
'J XT O
M i. (X
fJ tj CJ
c: xt * to
j o a> . »
t) f-~ CX
r- r O
M *r- (_J
rt ^»
* -CI to
!/) »
(k
,
X3
DO
1_
X3
at (
U_ (
i
t-
eu
c
0>
!.>
i
ft»
r-1
r
..
O
x>
0
"M
.
-J
^f
CJ
r
;B
LIT
«/>
^ 0 .._ <^> H- ^D 0 .
«9
C
O
r-
9
4J
P
CI O
C >, JC
a
Vt f
« 3 a
0) 4-* O -
»-r- > C
g^S -g
t» *J «X -r-
>J-r- V- O
M-0 O i.
^t
fO
a> w
I- >, 4J
f
r"
O
o>
c
^«
3
UI
4J
*
C
3
oc
">fc>_J
r-. r
*/> ^
rj
L.
e
« r-
a
tn 4J
*-» c
K QJ
«3 x'
n. ~-
01
M V-
^
«»
»
o
CI
4-*
u
r-«
&
U
ID
ttl
19
01
O
u
1
o
C7
c
O *vi
Cft O
c 2;
r*
£ V-
w» O
:< vi
0 V«
w c
01 CI
c t_>
£ o
*- (II
o »
csl
o>
o
«
o
**
a.
CJ
o
I/I
r>
»
If)
U
^*
tA
1.
a
*j
u
^»
c
CI . 1
JC M
Vt L. >»
»- » 4->
3 C
CLO 0
*-*r- fH
t.
<_> CI
o
°
*> 4J
C IO
O CI
*J =:
en
c j
c o
I/I
^t c
«
Ql f
O *-
I. 0
t^ r
O 3
LJ tj U-
^J 0
-^ CJ
\J
in
/> t
0 1
V) f
I- «J
** -r-
r- 4-1 CJ
O v> S-
O.D «*
r- O
1- 4-> r
>> w
n. 4->
O CJ
u
L.
C> 0
n- CJ
1
r-
>, ea
IO r-
O 4J
C
o> ci
Ol XI
^ T~
O> v>
CI CI
o i.
r
>>
cx
0
o
L.
c>
ex
VI
In
XI
i
0>
CI
o>
01
Ql C C C
S-5 55- S
» i-
01 c-
l/l o
i.
_o o
n
X E
* ** V~*
O
U r~*
"." iy
O O
V. O
4J -J
^
n5 *
vtz: o
i»
*»** ** *y
4-> 01 >
JC
u o ei
_c «
x: <«- vi a
** 10
s " o
>> 0. 01 i
*- Ol ** nj
t.t CI 'r~
t; t-J V
3 to o
3.1- "o
4-> *»
(^ IO
XI
c
13
Q
f^
10
4J
l/>
-c>
Gl
M
t_
1)
r
^
.
O.
O
o
*
l/>
1
*_>
ra
*j
,_-.
«o
c
o
1 '
IO
r-»
r
»
»
c
u o
*C 4J
01 U
Cl 4-1
° 42
r- C
«- VI Ol
ci q^-
E" ^> *r*
6«j; >
u> 01
XJ .C
*- C VI
O i9 *4^
2-2
-------
UJ
z>
^"^
-^
S
%-
i
CM
UJ
C^
«t
H-
.^:
rj
.0
O
O
^
rtJ
4J
t/)
O
O
t.
0
u.
-^
..Q
fO f&
fc: *""""
x ra
e<
CL
rx ej
tC 4->
CJ
rj
CJ
CJ
t_
3
O
to
CJ
',£
O
cn
c; c-
t C
4J -
C <-
r- C
rx
J
-
2
g
o
i^!
U
O
cn
c c-
-r- C
C C
C
(X
O
^~
J
\l
+J CJ 4J (J
c:
£C
c:
S- (
> 4
O
cr>
^ OJ Q
"5
_>
DI
r:
s-
GJ
b
a
o
u
cn
- L9 cr
0
O
w"
CJ
Cl
c: r
.- c
c; r
-- c
V.
n~
0
U
«4-
'*
O
m
j t: r
_J -r- C
r 4-> -^
3 c <
si < r
Q.
o
I)
V-
CT>
cn
M tr c
-> ;- C
i cr. c
\J -r- c
tx
CI
o
14-
M-
O
D>
.1 C C
r *-> «.
J C C
vj -- C
J-
tx
^J
3
i-
3
vl
4JCJ 4JO 4JCJ 4->CJ
n
c:
CJO OQ .
O
cn
- c:
- i~
J (D
r-> >
^T> O
» cr
c J-
O OJ
M > -1
Cn O
- c
c i-
O CJ
J > -1
CJ1 O
~
Ct
JJ
Cn
r <-^ c o c CD c:
.i: -XT .x: x: xr .XT
to
/>
to
in
nj ro
CO
L^ (/)
. _rcl
m to
ro .
3 => :
in tO
in
O ro
to
r
ID
r^.
o
O
l->
>^
in
l"O
to
I
i.
CJ
CJ X> £X
S-
3
O
in
CJ
x>
o
SI
^>
_o
o
c:
O
jj
rO
^
CL
O
O.
eu cvj
4-j r~^
OJ
q.
o
«>
^
o
4->
f0
ri
o.
o
o
t/>
v>
c
CJ
o
0
crt
cn
^x,
.
.
S)
,
IZD
l»
c.
fU
-M
JD
Cl
C
*
O
1_
OJ
_O
V_:
i1
rO
>f
1 <
C
4-> X>
t: -i-
o o
o t-
>^
t.
CJ
CJ
XI
-cr in CJ
in S- £~.
i ra in
r OJ -r-
T XI
O-O -J
^-'t D-
c
o
cn
c
r- 1
-cr
in cr
rj j;
in
" 3
QJ O
u u:
CJ H-
I- 0
t;
t.j in
in
E
O
0
cn
ro
-^
cr>
o
XI
H
CJ
u
OJ
0
CO
' »
4- . C\
O CJt
O r\3
4' . O
0-CJ r-
ci
;;
01 Oi '.1
00 r-0
O
^J-
Cr.
CO
X>
C
4-> ro
C. r~
3 ra
cn o
a v-
- CJ
O
L»")
CM
«»
in
CJ
1 CJ
4-» 3
ro c: O
i- OJ in
t- '~
4-> >, CJ
10 O »
in F 3 i -.-
3 E; x> o. xi
0 0
^^
i*.
CJ >,
> ^
xr m
in s-
r- ro 4->
o in
X) >> 3
3 cn
0-0 3
r- <
*
C CO
o
4-> ^>
cn .
C
r to
XT C
in t-
13 O
^5 4J
ro
QJ O.
O
S- in
C != 0
- o E
1
t-
CJ
JO
5
o
z.
c. +J
o o.
4J O
cn Q
c
x: to
in
ro 3
3
in
CJ U
O -r-
i_ *->
OJ in 31 in
=? c-lfc
K C
3 '§
4- C3
O
-f-t
4~> . r:
o. 1 3 M
c> o
CJ Q O
c^ ^^
O -,
rj
4-> C"
O-CJ C7
Cl CJ ".
in
c
CJ
CJ
to
.
to
c:
o
CJ v-
ET +->
<- 13 f
r t-
O 4-> 10
in ci in o
ra o o
d CTlr t-
O -r- 3
0 t- E 0
x> in
3i l~~ 1 CJ
OJ CJ O f
r3 CJ XT XI
4-J -r- O O
in x> > >~
S-
ro
CJ >,
>> »-
>> 01
^ E >
5- O CD
ro !-
O V- -a
>1 O
in s- x:
CJ rg i/i
JT CJ -r-
U w ^ f
S- ra XJ
r3 > O 3
to
U. -r- t.
- ro D-
C S- 0
O 4J . J.
r- in CJ ft.
rO C CJ T3
4j -f cr
i- r - ro
o x> r:
tl *-» CJ
cr >> cn 10
ftj ro C ~~J
i- :: i
t -*^ ^ ~\j
crn in iy
o r*r I^r d-
in
CM
o x:
cn
^- -r-
rj CJ
t- >>
OJ X>
O.
in
CJ CJ
I* 3
r- >-
E 4->
>^
<_
OJ
in OJ
J-
ra "o
QJ CJ
in
Cl i
_c~ ^3
4-5 fA^
1
O
in o_
OJ LO ** "
CO C
o ra >-, o
.^- s_ ra 3 v-
O XT GJ rO
r=- *- cn i- 4-»
O-r- 3 »-
L. s: ca o
o . c
*-> 4-> i » in
o a. ra c cr
?: Cl U O ro
0 CJ -^ in S-
XJ X) 4-> TT> H-
CJ CJ ro rj
4-» l/> U. t_ O tf-
O 4J CE: O
E xi cr
cr cn eC O- O
CL.
UJ
to
0
c
0
cn
Cl
4->
CJ
tn
i.
Cl
r
Cl J-
> a
r- CJ
*"
o
+-> t
cn O
J^ *O
(/>
3 X>
in' I
3 <->
10 C
c r»
OJ O
tj tj
0 f
o
3 Q.
ro
CJ
t- O _J
£l C3 O£
M
O
ra
S-
u
o
u
in
c
o
+J
ro
i-
«
.^^
cn
CJ
t.
^,
o
p
4_>
x>
c
-------
LU
O
^-^*
1
C\J
LLl
m t
CO
1
t:
«ti
4-J
,O
O
0
>^J
,;
0
V-
o
tv.
"U
QJ
l/l
OJ
CJ .O
4 » rj
rj »
X >
O *-.;
i.
O
i.J nV
i t *
1 1 -i
« i t >
* > t >
"' *'' V IM
<|l 1 1 ' *
i (vi v : (
' 01 ' " i\i
\) ~ CX.
r . « -
ll (.XJ 1 * C )
n - r -- t~
C >- CO ai
cj - ;. c.t
- , c:
--z i. *
a« o c
r - CJ ' > O
* -^ C O 4 '
.1") S_ O O U.
t) o c: c.
t: > n - -,-
0 O CO -CI
I > ^; r - 4/1
z-i a> o «3
i
1 O
1 0
J_ O
O r-
** ~ W Qj rj
CJ i/i O
m -r- rs o
c: s.,
O O r 1/»
r- Cii nj ;D
r - n c3 T-
*"X .0 t) ro
C71
>> 5..
J- 3 (U
f.i c: >»
- - O QJ
u_ ui .at:
.v o> r^ «u >'
(,' c. a>
s- 1 3 .v o
f- - O "-. r
ii! O rc :
ci n ' c o
4 * 'J *. '
,c.' ,'"T' '' ;(" -1
D
t o '
t.j
-» >
-.:
0 - C
Ol 'tJ
«^_ , £._
Cii i
i ) c: "
O 0 "
C r - O
f\5 -i - -1 >
U ~: en
- c:
S I O <
Ci I » - i -
L" in s-
C^
0
o
OJ
I/)
f
r- rj
0 D.
flj "O
tl CJ
t: fX
O *-
CJ D.
*ll
C!
»jj^
CJ>
0
VI
M O
C3 4->
C')
c: c
fO «r-
O -C
f
t > O
[ i t_j
n
I" V
c" i;
C t
t»
'I
*1
t >
1 »
I"
LJ tv^
t" .
'' (.'>
LX <_J
CJ
L' C^
^ -
cu c*
> 0
O 4->
CD O)
C
' !
U~l JL-
^
CNJ
O
CJ
in
1 .
, c: re>
in JD
-,- p S-
>, trt t/> ftl rj
QJ ra o C >}
> CJ O- t:
CL> >y ' f3 C\J
O !>
r- C
r- S- (U
a -t-> ll
i- O 4J O
O CJ S-
o i fa
O- t^J o o
11 -I-)
£- (/) CTl
3^ c
IJ CJ ZJ -
on ,cr
M- X) r3
11 Q) S~
cr u> E"
i : OJ 11
:: -r- M o
1- C 0
*< CJ O t-
O '-'"> t_J CJ
'-'"' >.' ir" i)
»>-. O 5« C_>
i i : a
C.I tw UJ C'
2-4
-------
a:
r
o
UJ,
_J
K4
s:
LU
I
^,.4
1C
CM UJ
CM
UJ O
_J U.
CO
1 U
I
O
I
O
to
UJ
or
O
to
CJ
XJ
,
rO
-£
CU
4->
CJ
.
O
^
.
GJ
r-
O
rO
cr
o
CJ
n-
o
(./^
CJ
_ri
ej
O-
i.
o
in
rrt
en
73
-C
l/l [;
'3 -r
O
i^- CU
CJ -O
O-i
S_ ro
O >
4- ro
r) S-
4 rO
rO <1J
0 >>
S-
01
1- i;
O 0
n rx
CJ *
1- S-
CO r
-ro
(/>.
cr o ro
0 O 1
-.- T- CT»
*-> > CO
ro I- CM
4-> OJ
O r
O- l/> O
10 cr CM
ro -r-
l ro
4-> C)
1- S- C
O CJ O
n f~
i/l CX
4-> tr 01
CX rj ,
CJ S_ OJ
f~i H 1
ra
5_
O
>*-
in
r
ro
0 >>
oj rj
- u
E
oj ,n
o in
i ~o
_cr ro
CJ O
:=- i-
CJ
tr
i
s:
_c
u
GJ
~J
S_
c s;
o
4^ cr r~-*
ro O lO
4-> -r- CM
X- i/l I
O -r- C3
cn. > cvj
> O
4-J ro -Z".
t. ^ n.
ro -iT O
O- CTl^-
CJ -r- CJ
f~\ T" )
^t
1
< 0
o
o rm
V~ f~~
in i-
i O
ro ^-
o >,
O) r3
r O
r- (J
E
CJ -0
U l/l
r- X)
^r ro
GJ O
> i-
4J
a
o
f^
f- _
GJ
QJ
OJ
1
.
s:
c
o
in
in T
-* > ^J-
S- -r- =}"
O O ro
-^j ro
1-0 1
0 i CJ
i^ rx r
5^
o c
cj r3
4-> O
ro O
CJ ro
C
i- O
o
OJ
ri r-
ro *o
~o
CJ ro
r >
,- ro
K
C
OJ O
0 4->
r- ro
s^ B
CJ i- O
> 0-.-
>«- in
o cr i3
CJ
.^
r-
rx
f:
rj
~T~_
^
=
O
c: -=J-
c\i
C r~"
0,1 LO
O I
to !"»
Jsi CSJ
i- s.. r-
o :^
» r~-«
O C' *
-- O iO
X) **~
rs *-
C_ O '*
G)
**- CT. c
o t; o
i ^^
4-> e o
CL rj r
CJ CJ
o n_ i
>>
-o "c^
o r3
4-> O
r3 U
CJ <^
S-
cr
^ o
o
CD
r3 i
4-1 ^O
CJ O
-o t
O -3
r >
E
0 0
O 4->
r- ftJ
x: =i
CJ i- fi
> O -r-
<*- en
0 C r3
^ -r- JD
tn
4->
O
CO
^3
JH
U
rO
i/l
\fi
rj
s:
^rj-
cn
vo
CM
i I
10 o r^.
_v o r^.
i- c: CM
O T-
C p~~
O LtJ O
-i * J"
JD -r-
CL «^- *
ro CU
<*- i- c;
o ( o
» OJ CX
4-> -t-> CJ
O- O r -
CJ -M CU
0 UO h-
r~~
o
CJ
1
^
-o c
o r»
*-> O
rO O
o rt
c:
s- o
o
o
^j _O
rO r3
X> r- *
QJ r3
» r3
E
C
CJ O
O 4-*
-r- rd
-E: G
0 S-
> o
V
o c:
T3
c
f3
vi
» «
- U>
KJ CJ1 tO
4->
^ - * >.
O V- CO
o ^: o
*/) OJ
ro ,rr
S- U
I i . *
*\j ai
**- o r:
O i/l O
C> -C"
- a: rx
4-> O
f-^-^3 t .
CJ ci O
Q rO h~
r
QJ
G>
r-~
>^
o c
O Z3
4-> O
rj O
i
CJ r&
S-
S- O
o
CJ
*_* _Q
n3 r3
-0 r
QJ *5
r >
r- 1X3
n
c
CJ O
i *r
U 4-J
r- rd
SE
> O
tf_.
0 C
4->
ra
o
4J
O
OJ
c
c
o
0
t31
cr
cr
c:
ro
CX
'I
o
CJ
o
n- 10
<*- -r-
OT30
CJ vl-
- CJ LO
cr "^ in
O h- t
r r^
rO S~ *3-
4-* ^Z
o «co
rx 4-* ^~
i/j cr to
ro I-
5- IX
I O
CJ
*t- cj L:
0 > O
OJ -<^
en ix
4J CJ
cx-o >
Cl C CJ
O ro 1
t
OJ
OJ
^
"O i~
OJ -3
4-> 0
ro 0
GJ rO
S- 6
O
OJ
ra r
4~* _O
ro ro
-a
OJ r3-
E
OJ O
o
o cr
-2£
i
O
^
CJ
tr
2-5
-------
C3
111
g
1
21
O
<->
CM
1
CM
J
?
fll
, '
ftt
r
O»
«
ai
4~*
rj
r i
o
'^
CJ
r-
i
0
ra
i »
c
o
0
I*
o
«^
10
lO
en
n-
o
4_>
C
4^
X
Lul
ai
nl
l/l
t_
1. n
rl 1
ftl t
.* i
i/l
in
O IO
in
c
ro
i.
l
V-
o
3
n
Cl
>_ >,
3 a
*~ CO
c i- in
O 1C 1
4J -CO
«u in r^
4-1 0
o 4-» r-~
cx in t
in f- r^
f- 4J *-*
IU id
U 4J
H- «/>
^
ft
n
QJ
t.
Ol
CJ r
r C
.- 0
to
CJ "O
r 13
O O
>- V.
J3
CJ CJ
> 4J
13
f 4-*
0 VI
*°
»
r;
c;
QJ
t.
V. fit
«« I".
in
in
CJ r .
-. t.
ci n
S.. 11
a.
CJ
i- «- -
o -O
*t~ ry
r .
«) T-
4-* IO
10 r>
O IO
cn co
c *f
/- ro
c «r
c: i
ID CO
CX, l£>
?"» *"~~~
/O CM
t O
.£- co
cn »
f
3;
*- OJ
O 41 c
o o
3 -C -C
O.
O CJ
t. f
»x"
>^
J^
J~
U)
f
1*-
0 10
Ol
c
r
c:
c.
a
rx.
O
ro
4J in
C CO
CJ 1
£3 CM
-> Cf\
i- C CM
CX CJ.
CJ v. cn
000
^l *~~*
ID t.
_f- . .
Cl " O
cr c
rr: o o
r- .C
o m cx
4-J .r- 0
V^ Ql-
>^
ft
s- c
3 O
Ol
c- VI
u- -a
CJ 0
V"
E o
o /o
r +*
U V>
JH L*
^ *i
'5 -S*
13 C
c o
in * »
. V i L
Cl (11
1^ l/>
n "v.
>-. uj
Cl.
CJ
O .Cj
y_
IO pr~
4-k 4U
O >
O IO
O)
o
f
n-
o
c
o
1 * (*)
10 c «a-
V- IO *T
4-> cr i
in o ro
»- o co
c *~- ro
B .
> "~^
M »
i: c
CH'r- CJ
- c c
3; c o
10 -C
oj ^- cx
4J Q. Cl
IO
4JV- O 1
VI
o
13 O
O U
W O -i
o >*- s-
4-> t- ro
IO IO t.
4-> 1- l-
TO
IO
*JL
»,
?c
.cr
o
t.
i >
t
1*1 >,
IM 1
"i ' 1
in l.
-i 11
(1 IU
'*."
«i i"
i- -.-
CJ
t. , -
o .r»
M- ro
IXI *r
4-> <'J
0 10
^ I/I
o cr
cr ^
O i
Cl *~*
5» "~"
r-
C5 J- f~
~ IO
in »c\i
10 c; cn
:: -r- in
£71 'S
il 1. ^?
CJlr
"*- Ol£>
0 1-^--
rx.
4J cO * '
cx a>
o cn c:
Ci cr o
CJ c; 75.
4J C CJ
4J i C*
10 o- 1
^
§>,
O *O
o ^
>,cn
VI
01 O
3 J""
CnCJ
Q> r4
C
i- o
E
in =
CJ X> 4->
r^ IO f
u o o
f- 1- O.
JC CJ
CJ Cl Oi
10 o
1*3 VI 13
3 01
C I. i '
C 0 -r;
O
r-
.C
0
c
o
j^
a
4J
O
CX
11
C
IO
l-
vo
* IO
« CM
>> 1
O CO
rf fo
01
re o r^
CJ VI 10
3 >
l/> O
Ci
H~ cn ti
0 C 0
f f*
. c cx
4J tr o
CJ r O
O Cl. 1
O-
_%^
C
M O
L 4~>>
3 13
V- O
S-
o c:
*r-
O
CJ 4->
*0r-
-i Cl
JC i-
CJ
0
c
4->
a cr
10 o
o o
0 >,
^ -a
c
id
cn
^£
u
sc
o
aj
o
c
13
o
>> c
ro rj
5 c
-c: r:
cn o ro
^- S_ r
rr co r
0 S- CO
CO
4J «
cx c:^-.
CJ O I;?"
o - o
fO *^* *^^
'r- ^
- a
Cn O
». 01 C
r- r: q
c cx
4-» C CJ
in rj r
Cl r O
3: o-»
i-*C
0 O
V-
r-*
f » fj\
C r
3
0 >.
o -o
3
JD VI
in
Cl <3
1- -r-
3 0
o> at
T- CX
t- 13
O
CJ V.
u o
r ^J
JC IO
OJ 4->
> VI
IO
c
o>
J-
r=-
4J
VI
_o>
u
H-
3
i-
V-
O
>
0 10
r-
co
VI Csl
>> 1
r? o
.i r^»
Ol
i ro
O
it- t-~
O '
cn
c
cx et o
o cr cr
f ~| i-j o
» -CT
ci a. cx
4-> Cl
4J c; o
^>
4J
CL
O
in r3
CJ "O
r^
»- T3
4^
r~" r-^
0 0
r- l~
CJ >,
> C
c
0 t-
o
o '
10 4J
^ f*
3
0 0
s: o
0
r-
C
"cn
r-
r^
^
i * >
r)
C.) <--
o
Oi
c:
c;
c:
fO CJ
r i
ex. q
I. NJ
o o
ia
c: r
o an o
i- O
in in
«- V; CM
* »^-
cr LO
o
>i4« CM
000
-: ci in
j~ i/j '
C"
ac to
o CJ
M- -r- C
O 4' O
in JC
»- cx
4J 4J CJ
O. « t
CJ *> O
Q,-!- "0
c c's
3 3 O >>
O O 4J 13
O O 13 3:
>,r in cn
"°'io >*- DC
*J li-
ra O 4J
r- r> CJ -r- -
.c o CJ cn n
CJ U I- 0
> CX VI »-
I CX IO 4->
UU f !-
3 . lO 4J
C t- » 4J IO
c: o «> o 4J
«£ **- V- 4^ l/>
^k^
^
0
4~*
r:
OJ
2-6
-------
s ,
O
UJ
» 4
H-
o
CJ
^~^
CM
1
CM
UJ
ra
«=c
l-
CJ
^
rj
ff-
rd
5
QJ
ra
CD
o
OJ
r--
CJ
ra
4J
t-
O
0
OJ
3
*
i *
c:
E
i-
ra
CJ
I i
a)
JQ
ra
>r
ra
«3;
ra
4J<
ra
Q
Cf
O
4->
C
Ol
+J
UJ
OJ
4-1
+J
CO
CO
r
C CO
O UO
r- 1
CO CO
CO CO
r- CO
P 0--.
O O P*-
<_) .- r
> CO
>, S-
(0 CJ
? t°
"oir fl)
f ro C
-TOO
r- .C
O ft CX
4_> JC CJ
ro O
4-> IU OJ
t/> 1 h-
^*
ex
OJ
-^ rd
to ro
O T3
r- -0
E GJ
4Jt
O ro
f t
0 CJ
JC
CJ >,
> c.
a
c.
o S-
o
n)
+J ^1
ra +-*
0 E
3
0 0
:z u
t$
a
«3
XI
c:
*-*
$
6 r~~
en
t- f-
CL o
it-
QJ GJ
JQ
"§
rO i
ro ra
r 0
cn
i -»->
X
C. 0
0 c
ra (^
4-J r^
ro en
JC
CJ
V.
OJ
CO
O
c:
c>_
O
3 C
ra o
CJ 4J» CO
i- C -=r
CQ O r-N
1 1
CO . C\J
>* V- LO
ra y
r* fL f^
cn »-> r
r- C CM
E '
<»- CL
O O "
r CJ
C CJ tZ
o > o
i- C) JC
to o ex
r- CJ
r- c: cj
O ra |
^
+J
c
3
O CO
o s-
>> CJ
JQ >,
f> CO
OJ
t- >>
3 C.
cn OJ
r- >
V- QJ
CO CO
OJ -O
r ra
12
CJ r-
l~ tt
U K>
JC >-
OJ 0
> *-
r
O
c
f
r
j^
*-4
V
0
c
o
CO
>
>r_
CD CM
C CM
0 C 1
- -r- to
4J S- 10
ra QJ CM
4J» ex
s- E
CLCL, r
C
1- E
« Q)
1- cn c
0 C 0
C CL
4-> E CJ
cx ro r
O . O
CJ
CJ ra
!-- -C
& -
CJ 4->
r C
CJ 3 4->
r- O GJ
JC 0 >,
CJ
.£) QJ
>
ro co .,
3 X> GJ
c r) o
c o a
> CLLD
rp J-r-
^ "=:'*'
c5»""*
"*~ t!
2: o
*fr- C
0*0
nj n.
*-> CJ CJ
CL J-«~
W 3 CJ
Q CO I
^Q*
4J
O.
j*i rd
W r?
ey -o
r "O
f= QJ
4->
(U n9
r r~*
0 QJ
r- J-
CJ >1
> c
o i-
0
ra
+-» >>
o c
o o
QJ
0)
CO
CO
CJ
c
c
OJ
I
<*-
O c.
Q>
C. O
OJ ll.
0
* i- ^?
c: n: i
o ro
r- 1
co -c cr\
CO CJ CV
r- i_ CO
F S^-
c3 co c^
CJ QJ i
ra T3
S c
cn o
r- C5> C
re c 9
O £~ d.
4J K C!
ra ra i
4-> C
to CL 1
^*
CL
OJ
co ra
Q) -u
"O
E QJ
o ra
U QJ
_c~^
CJ >,
> C
ra
c
O S-
o
*- >^
ra 4J
T3 C
3
O 0
'Z. U
ra
c
«^-
"o
t.
r3
O
JC
t_
o
c
c.
c
0
CL.
Q
c
0
*- CD
«*- r
ra co
S- ro
1 i
o co
rj f^.
ex ^ *
o ra ro
t~t -T* eD
>> -S-
ra S-
s -s.
C71 " CJ
r c C
_r o o
r- .C
O CO CX
*J -r- O
ra > i
U 'r CJ
OO CD 1
^*
CL
CJ
j»f ra
co ra
O "O
r- -O
S CJ
CJ ro
CJ OJ
r- i~
Q) >,
> t?
ra
0 S.
o
ra
4^ ^^
ra *J
o c
3
O O
z. ci
r]
c
o
i.
ra
o
JC
4J)
3
0
1/V
rj
cn
>r_
T-
S--
c>
o
CO O
r- CO
r- LO
O 1
CO
« c 10
-r- CO
4-> ^t
a. c-
QJ CJ *3"
CD r O
ra i_
^ s:
C7) - CJ
^- cn c
-ECO
r- JC
CJ c a.
4J C OJ
ra ra r
» 0)
CO CL. V
CJ o
c
f
>^
QJ +?
c
CJ 3
r- O CO
.E O -i-
QJ CO
> >> ra
1-^ CO i
cn -o m
ra 3
O C
C S- c
O 13
4-* r~* C
i >
l/>
0
'>
CJ >,
5.- i
rx :.T
t-
C, v-
«*- a
ra QJ
4-> C
0-J
c: -c
r- Cn
c: -
c: c.
ra ~"-^;
ex.
a ac
c
(
JC >
O -r- CO
J- O «*
ro co
CJ >> to
o: s_ r^.
3 "3-
co * ^
>> cnCM
ra C *
£ c
en c
i- 0
CL. O
«1- C
o o
. ra CL
4J» O CJ
ex s- ,
0) 3 CJ
Q CJ3 1
>^«^
+J rO
§ ^
0 >,
^- ^ >
JQ (A
QJ
f3 t-J
4-> cr
*O CJ
-o s-
01 -M
1 VI
r- rs
E E
.a» -*
» M
O "O
J^ O
O S-
f
r~ r
^ CJ * *
ra E
5 CL
Cni QJ
GJ C
a: > o
OJ .C
OJ O CX
4-» GJ
ro -0 r-
4-> C GJ
c/o ra c
4J>
CO
JQ 3
cr
4-> CJ
CX i-
CJ
CO
CO 3
oj E
r *
E v>
-0
QJ ra
O I
_t_ ^^
CJ r
> ra
c: i-
o o
V-
ra i
4-> >,1-
ra *J ro
o c E
0 0 >,
z: o .a
ra
E
ra
r9
5
2-7
-------
^ ^
Q
UJ
P
i «
|^^
0
CJ
"""'
CM
CM
UJ
1
CO
r^
XJ
ni
.,
ro
^C
CJ
r'
CJ
,
O
f
OJ
1
n
4->
U
r3
£-
6
CJ
n-
o
CJ
E
-^
\ \
c.
E
S-
r3
O-
CJ
Q
QJ
X)
j -
r
;>
e£
ra
ro
O
t-
o
4_>
C
CJ
X
UJ
4_>
<3
l/l
O
S_
CJ
f:
3
in
QJ
rj
C
r
rj
_O QJ
i
r- Cn
ra cr
ra "-T
b
A3 «
4 > r
rl o
CJ <4-
cn
c
c:
ra
i
CL.
-o
C
o
>4- CM
i» r~-»
ra cr,
!_ o r-~
r 4-> 1
in «cj-
i trnLO
c; co
| * .r
O_ > ^
CJ -r- I
d _1 O
IO
ra s-
? =-
cn - o>
a; o o
oj in ex
| » .r CJ
4-J -r- QJ
UO O I
>,
C
13
O
^>-
X)
ra
ro
XJ
1>
C
r- O
CJ -O
, . fQ
O O
.c
CJ QJ
i 4-*
rO in
3
C !~
cr o
<£ M
o.
CX
r*
in
r>
in
a:
l^-
o
0)
XI
f-
ci
4->
n.
CJ
c:
r-
C) U
i rj
_O Qt
«--
-r- Ol
rO C
> T-
<3 "5
CJ
ra .
4-» r -
ra O
C3 V-
cn
c
c
c
ra
XJ
c
ro
U r-
r- «3-
<»- ro
v- ir>
i- cn
r- in CO
CJ CO
«k ^>
in CJ.
ro C^ O
5 IT)
r: . ^
en s-
i-
"^ * 01
M- c C
O O O
t- _cr
in ex
4-> -r- CJ
CX > r
CJ -r- OJ
o a i
s-
o
S- CX
O CJ
<*- a:
V) r
0 «
t- 3
en c "in
14 n ra
^
OJ c: .c
r -r- tn
E^Z
a c: o
r O 4-J
O rcj
r- in 4->
-C X> 1X1
QJ ra
> O ra
S- C
ro QJ T-
3 4-> in
c KJ n
C 4J O
rj
c:
>r
^3
O
_J
C
r-
OJ
XI
ro
'J5
ro
J_
r)
CJ
in
^
O
~>
Cl
o.
1- >>
0 C
*+ rj
rj V.
O C'
<«- t-
O r3
en
O OJ
i- S-
!
> .
OS r-
* 4H , .
fZ CT>^3"
o c i
r- -r- CO
40 n CO
O f~ «^"
4J ra
I- r *">
o a. «rr
ex o
in c o*
PO -
1_ 4J
4J CJ
*- s- t:
o o o
C-i. C"
in CX
4-> C CJ
CX 0 r
QJ S- OJ
Q t- r-
0
XI
in in
QJ -H
!_ CX
3 -r- CJ
en o c
O 3
*f- CJ «-^
in c
OJ X -1-
r3 4J cn
in c;
oj d x)
c c
i- X) QJ
i CJ
O in l_
O X) CJ
en >,
r i'S^""
f^j 4-J (Q
n c u
C 3 tn
C 0 -r-
*i U 4-
o
»T_
O
tu
r
Gt
j-j
rj
»~
a
ra
£_
(-3
CJ
IA
*~j
c>
;t
CJ
J.-
cx
o
*4-
rj >^
O - j
Cl 0
C
10
rr
cn 3
c x:
a
c; a
C C) CO
ra co ro
r r~-~
a. . i
S_ CO
-Z CO
cn
4-J »
CX ^
cj 4-> un
Q 0-0
cj tn
ro
:-- en
'uw- CJ
r; t" o
OJ -C
cj v- a.
4_J (_,"> CJ
r3 o ^^
+-> i- O
to a. i
^
c
o
U
o
>> <->
X
P3 _0
XJ S
O
OJ 2:
»~
E c
f" ~
OJ >,
i in QJ
O XI >
r- ro i-
0 S- CO
r O
I - i -
QJ
IX
x_
o
H3 "o
+-* l-
(T3 »O
o ic
-*J>
CJ
t-
v>
X)
c
tn
cr.
c
c: CM
ra LO
CM
d, r^.
i
. c: VD
M CJ CM
O. QJ
O V- CM
O CD O
>> -
ra s-
S i.
Cn QJ
r- C C
X o q
QJ I/I CX
4-J ' QJ
ro > r
4-J 'r- OJ
to o i
+->
c
0
O
^h-
XI
r3
4-'
X)
GJ r
, c;
r- O
I/I
OJ XJ
I ' C3
O O
-r- S-
QJ QJ
> 4-J
1 4->
ro in
^3
C i-
C 0
C
O
N
«=C
q'
'3
CJ
XI
ra
5
ra
t
rj
CJ
M
^3
O
^
C)
CX
o
Ij
ra
n
a
>s,
GJ
S-
3
cn
C 0
r- CO
C 0
c: r-
13 1
r CM
CX CO
CO
in
in in ^-^
>> O CM
ro c: o
Cn S-
2Z ;"--
"^ . QJ
it- r: t:
o o o
r- JH
in cx
4J -r- CJ
CX >
Cl QJ
0 0 t
in
3
X< C3
4->
in U5
o> :
3 C
tj . «»
in
OJ X<
r 13
r- O
E !- =
in
OJ r n
r i OJ
O <3 4->
£*-£
QJ O to
5^ 'I
XI
r >, r3
ro 4-> 0
3 C C£
C 3
C 0 'I-
cj^ CJ O
r3
XJ
ra
OJ
2:
c
*'~
QJ
-O
ra
'(3
13
j_
t1
OJ
in
o
^
QJ
I.
CX
O
ij-
ra
rd
O
t-
ra
OJ
in
OJ
XJ
d
<8
1O
cn CM
cr «3-
r- CM
C 1
c: ^_ cn
, ra ir>
o_ cn
cr. ^"^
o ca O
5 LO
CJ> V-
r- r^
- o
«4- c cr
O 0 0
in CX
4J '- 0
CX > r
QJ -r- OJ
I \ r~> f
->v
c~
o
o
Jw-
o
«
4_>
x>
CJ
r .
-r-
E
OJ
r in
O X!
x: o
o s_
>
r" r *
(3 (3
3
C 5-
C 0
<^ M
V>
t/>
p;
fD
^i
V-
_
4-*
CJ
.s~
$_
3
CJ
* to
4-> t~^»
c: tr>
O CvJ
t- i
JV- r
13 CXI
cj m
t/i rr -
>> o to
rl - O
3r */v "C*
cn >
CJ
*t- CT1 Cl
O C 0
C C_
*-> Cl O
C_ rj r
0 r GJ
CD 0- H-
^"»
4_>
O.
O
_ii r3
to fO
O T3
-r "O
jr o
4_J
O P3
O O
^- 4-
j^3
o >>
> c
rj
c
0 S-
o
ro
4-> >.
ro 4->
x> c:
3
O O
z: o
ra
o
_r~
ro
O
C
'
Cl
Q
ry
.
'5
ra
s_
OJ
in
O
--^
CJ
CX
i- 1-
O 01
t- -Q
A3 CJ
4J tl
ra CJ
cn O
i*-
o
c
o
tn
>
>,
CD
* CJ
C IO
. o
in S- VJTJ
in o cn
-;; X> CM
t " 3 »~~«
C_J tO r
10
rj 1_
jv iC
cn » oj
-r- en cr
or ci o
r- .C
cj tr ex
4-> ir cj
4-> 1 CJ
to rx. r
^
c:
3
o
U
^J
XI
13
I *
13
X)
OJ
r~*
E
CJ
t m.
O XJ
x: o
OJ S~
>
r
r *
O f3
^3
C $-
c: o
^
o
>-»
-------
o>
n)
r
£j}
d
CJ
i-
4_»
O
f3
C*
o
<_>
*4 *
^-^ o
o
UJ C)
r-} I
;=: o
1
y *>
c5 4~»
C-> f.
**-* O
C-I
CM 4-»
i i-
CXJ CJ
tx
UJ 0
_l O
CO
o
f .
J3
/H
r
>.
*o
n3
O
V-
o
4_»
C
O
>^
UJ
o
rtf
00
TO
C
cn
cr
E
fU
I CXI
cn r
o o
1- *£>
CX. 1
* f""1-
*J"
4-J
cx C:-
(U O OJ
l/l «J-
>)( -, ,-
»o >
-r: O
Cn o
*r- Cft C
1C c; q
CJ C O-
4-> c cj
4-> r- 'o
CO (X h~
>^
4_>
c:
rs
o
O rQ
>» r3
-Q XJ
W "O
O QJ
i 4-J
r cO
CJ
CJ 1-
'o >,
*> cr
-c J5
CJ
> c:
o
cr
O OJ
^-~
fO J3
+J rO
X? *f*
nj
O >
2: nj
rS
t/)
01
J~
OJ
rs:
r:
r"
CJ
t
«o
f--
fO
rj
J-
rtj
0
;)
O
>
CJ
CX
£_
O
»*-
f3 >>
n3 rs
o
c
CO
c:
c o
C 4-»
D. to
» . ^r
c i- co
o 1C ro
r- 1
l/l » kO
»/» c cn
r- O CsJ
f- .,-
O *" <^O
tJ > 1
*^- o->
>>C3 "*-*
o
i: c -
C7> O *~*
4_> CJ CJ
tO O H-
>-
t->
C
r»
o
0
s.^
Jd
fd
/3
-a
CJ
r
E
0)
r CO
0 X)
.cr o
CJ )V-
J>
1 *-*
ra ra
3
C S-
C 0
«* *-
(/^
A3
nj
^
rj 4->
c c c " -O
r- ^ r -,- 0
o ci aj D r^
- D 1 r - _U (T\
.O r- - ,O .rt JJ r
O j.j ->
-. - t ~ ~r- -.- 0> **
r1 -. - *,1 r j . , -
> i > -~ (.1 i -
rj ^ t:\ cj -, -
f i « > f )
*- V t. .1 . .0
nj O «il f-l »
> 1 .f J ( ) r P »-r -
-^ >» >> ^- M
r aj >
n *A i/i .n <\J
Zl "t- T " 1 (
c^ ;-r rj> ( ? n i t ~.
> cv; > > O
C> r~- cJ o ;,-->"»
S- cn ro S- s_ nj
CX .- r^ tx tx ±-r~
c-» o aj
I- r- S-- r V- 5- -0
Or O O 0 t-
U~ 0 l+~ 4J V- *+- CJ rj
CT> 4~» >j +!j 4-> <1>
rJ J^I n,'. u r3 rj r3 O O»
C54-* O<: O^ CJO.O
cr .c
* .- o
c: c: s-
0 U cr nj
r V) *r rtJ CJ
4J ^- 14- ^- i^
own- xi o. ,) 0 C OC
to V- «3 "o
r3 1 c: *o
Cn c cn CO ~-; CO « CM 0) Cxi I- CT*
d_ i rj i/j i >-, t ex. * i t/) fi ci *
» t- T- yD O O i.~> t- *J- CL> 1 nJ f*^-
* LT> V C CT) T- r «tj- - *-7 CM CXI ^T X3 l-O
r^ c.) cxi r- -r - *^r cx) ex; cu u r--
4 J X) -T » ' J 4 J - * CXJ ^
iv * -> «^ i ** : i rn ' c\ c^ ^ - to - * **-,-»
cj <-r >i-« CM a- to cj o *^-- ^-> in .-s ° ^^
ci f - o o . . , r 3 ' - o ti c_>
C--> _-; vo ^- ~ CTt ist P>V, > »j.> >; iO
^y ^_x .u v. ' o i-_T * - ^fc>'1 " ^^ * -C." « ^ ,d * *
rj cr> t- cj r> cn cn s~
it .. DI - 4^ ±> i: cj " -c :n ~ «
Cnf-O ^CJpTLCO CilCJ O *C»
't- - c: M~ c ci o '.; cr *- .c: cz v- cr u *4- n tr
moo ooo 0.20 rz:uo ooo ooo
, j^ «r- ,c~ 4J _'-~- j_ _c; -r __^ -r _d
CJ^TCX -vicx ai-cx cjrticx .tocx w cx
4-> CJ 4-> -^ QJ 4-> O O 4-» C? O 4-> -r- CJ -M -r- GJ
f3 . r CX>i rXJCJ.* rot-l. CX>r O_>r
4-> (- CJ CJ -i CJ 4-> O* CJ 4-J O CJ O -i CJ CJ »- CJ
tOSCh- OOI uoOl tnr^F C3QH- C3OH-
>
r CJ O 4->
JT» r- -r- >* CT
o *a xii--*j>> >v rj
4-> * r3 O -* 4-> +J O
<_^ c:c_> c o
l/j > (-g .j ~T-J Q t^_ Q >^
n3 >ncr ov- o -o
O rJ " >> s- >, ro
n "o d o ri 03 "d
4-> X5 i/> CX 4-1 r3 4->
n3 CJ -o «J r3 « CJ
*o * ao>>-oc: x> »
QJ E^a -^s-c o«=C cj Er3
r 4-* FrS=« = r 4-»
r 1> CJ O XS
. .CCJ -r- S- >-, VI r W r l/> JCCJ
O O4-» J^OO-r- U"O O~O CJ4->
f- > *O O < I-i *-> *- rd -r- n3 > rtf
JC>, r >_CrO^!0 X:O r
o ' QJ >^cr)4-» oi- cjl- CJ
>C fl3t- f~4~»-r-tO> > «3i-
O 13 fOcrrC i r: r 13
r C>, Z>34->r r 4->I 1 £=>,
roin cc COOCJ rarj^ rani dc:
3-0 r3c; ra o 3 RJnj
CC3 coOi-CZS^-CXCJ-
co OS- .,->,4-J4_> coo cro OJ-
*
o cj v- J= -e: v-
10 C -r- -M 4-> O
V> C r 1-31
r- -r- 0 OOO
z; s: o ^ i^ o
2-9
-------
£J
~^
s
fes
tJ
CM
CM
CI
J3
rJ
r
r-
;.
01
u
o
o
z:
^
01
t-
u
4^
o
o
4fr,
O
Cf
I;
f>
g
fe
&
0
o
JO
r-
rj
;
id
O
It.
O
*J
g
X
Ul
o
3
«/»
cn
C
r-
C
C
«&
»
cv.
M
I
>k
V) 1^-
o
r-.
VI IO
^* t
id co
j5 Jjp}
m4J>
r- VI *-*
a: og
o
i.
Q. » ii
CI C C
o o o
CI VI f)
*»-r- CI
«9 > r-
t/IOr-
^3
T3
r^Ef
O E
fv^ 4r*
Ol
r VI
0 0
%a fc.
** r*"
19 19
o
r-«»- 3
r- 4->
^ »
CI C O
^L> O ^
SxC
-z»ja *-»
.C
3
r>
c
CJ
JTJ
f
r~
4
id
i.
01
10
£>
»
^
CI
J- CI
CX C
it" ^3
0
V- i.
o
id
id *a
I"} *^~
o
c
cn
£'5
c 3 tn
id R) in
«X 1
* *r *"
*- "~
O. -.
o !-»
o > o
J^tCD *«*
S o*
J= C
cn-J: e»
i v. ex
ft 0«
*-> V- OJ
|5
i
_g>
*j
*o
» ^
m *r
r C
E
Vf
«>-O
r Id
U O
T- L.
"S 01
<9
id i/i
cn
c
1?
o
;<
cn
c.
c.
id
n.
«^
c in
O CJ CM
r- C 1
VI > «»
E O
O «>
«J t. O
>v^
n>
O>*r- dl
S°§
>.J=
CJ
(Q l« r
u n «a r
0 J- 1-
r- U
fll r O_
> O W>
id
t:
id
c
o
s:
-o
c
in
c
en id
c >
C i r
c cn
13 3 LO
CO CM
Q ci
01 *- o.
*J
id -a
"O O
ai a
^*«J
B J^
r*** C
U 19
01 0
*
i 01
Idt
C ftf
o >
3S id
o
JC
id
o
r:
r'
CI
.r>
,-
*3
n>
t-
CX C
3
o
<*- i.
o
id
. i
jt O IO
jc t- r»
C> Id
rr; u> 10
CJ 0
«r-CS(M
o - -
o
4J AJ ..
CX CJ
ei cnc
a c o
ei'c ex
*J C 0
J Or-
«Jr~ O
^>
§
8
_^
id
ft}
"°
4i)
^«
E
Ctl
» VI
w id
-S£
>
!-
^5 **o
ll
I
cn
c
3
r<
i:
<-..
rf>
r .
*.«
VI
j..
at
VI
b
^
£
ex
t.
o
r:
u
^
t- VI
1- c r^
CJ t^
S CM
c: o 10
O 1
*r- CO
0 E«0
L o**"^
O CO
ex > o
IA -r- in
t- cn
. c ex
*» T- CJ
f* O^r1
OUJr-
|J
i
u
^
id
id
t»
U *"*
.r C
B
Vt
ttl TJ
r- «O
"£ CU
^ rfc^
r-5
a «i
5<£
S
JJ
o
"*'
ri
r i
1 1
a
ij)
t.
O
v>
o
^
O)
S-
ex
i_
o
id
4-*
C-l
o>
c:
id
r
OL.
j*^
id
Cn
^* l/l
\o
c~ r**
O 1
T- OJ
*» o
S C>
Or
>
§r^
O*0
JC
XO
4-» 13 >»
Id 3 4-1
"° C §
ci id o
i/
E c >»
ra jj
o
o-o c
«r- O O
JC O-r-
0 t 4J
r t.
a ft u>
c i. "Si
coo
«x «- u
V
id
S
n:
c: *
" *0
nn..
'r> >
- c»
K« in
j» 1-
4-» O
ci 10 jc:
)- T)
CX
o ci
t- c. n
0 -^
V- -<>
Rl 'r~ *-C 4-*
C
o
-r~
l^
r~
5»
r*
^%
Ol
C
-r"
C
n
Ov
(M
VI CM
>< *
J CO
Ol
r- -r^fc
50 o
- cn
O 4J>^-»
o
l-
4J id
O CI O
_f> JS
O *'i ^^
«- 0
Id r
_X
o
r
Id
M
S
r-
f*
VI «l
«3»
cn^3
§2
id
M >
4J Id
Sx
^ rj
1- vt
CI t.
c a)
c:
c o
Rl «J
4J O
rv( 4-1
4-> Ol
^>'r*
r "O
01 <~
^*: o
»- O
r O
O
r- *
*~* 3
^j
^t"
C*
r C»
*O 4->
V>
- C
> *r-
*O i^t
« c
O
1u >,
o Jo
4J M
C 01
Or
r?^
U
3 4*
*O 4JI
o
>**
r 4-1 M
4J V.
SO CI
4-> 0
VI C
CI VI >r«
IVO O»
4J4J **
Id Olr
JC U *
4-» C
4J 0
iA rx-r-
oi o cn
4-> *O Ol
0 « t-
"*r^3
_*^ ( ~TT
4J S C
2 VI
-------
The sulfur content data were taken from Bureau of Mines publica-
tions [2,3,4], summaries made from NEDS point source data, and a survey
of state and regional sulfur content data collected by the various local
air pollution agencies.
B. METHOD
The Bureau of Mines fuel oil data are available by state by the
categories listed in the left-hand column of Table 2-3. The categories
required for the study are listed in the right-hand column of Table 2-3
TABLE 2-3
COMPARISON OF REPORTED FUEL OIL USER CATEGORIES AND
REQUIRED USER CATEGORIES
Categories Used by Categories Required
the Bureau of Mines for this Program
1. Heating Oils #1, 2, 4, 5 and 6 1. Residential
2. Distillate Oil and Residual Oil 2. Commercial
Shipments for:
a. industrial use 3. Industrial
b. oil company use 4. Light Duty Vehicle
c. railroad use 5. Heavy Duty Vehicle
d. vessel-bunkering use
e. military use
f. electric utility company use
g. miscellaneous (including on-
highway and off-highway uses
for diesel)
2-11
-------
It was therefore necessary to convert the available state fuel oil data
to the desired categories, before attempting to allocate the oil use to
the various counties. Consequently, the developed methods are discussed
under two separate headings:
(1) Completion of State Data
(2) County Allocation of Fuel Oil
1. Completion of State Data
a. Residential
Various methods were examined to estimate fuel oil consumed
by residential units on a statewide basis. The more sophisticated methods
included variables for which the available data would have to be averaged
by state. It was felt that the accuracy of such methods would be reduced
by the use of the averaged data, and it was therefore decided to use the
EPA method [5] to calculate statewide residential distillate oil consump-
tion, whenever this would prove necessary,* i.e.,
Residential distillate oil (gallons) = [number of housing
units using oil for spaceheating x .18 (gallons/degree-
day/unit) x degree days**] + [number of housing units using
oil for hot water x 250 gallons]
*0nly necessary when not all counties within a state are being processed.
** Degree Days: A measure of the departure of the main daily temperature
from G5°F: one degree day for each degree of departure below the standard
of 65°F during one day.
2-12
-------
The residential category includes only housing units in
structures of less than 20 units [6], and it is therefore accurate to assume
that all residential fuel oil burned is distillate oil.
b. Comercial
For the years for which the Census of Manufactures Special
Report Series on Fuel and Electric Energy Consumed [8] is published the
commercial oil is calculated as follows:
Commercial distillate oil = all distillate oil categories,
except power plants - residential distillate oil - distillate
oil used for manufacturing
Commercial residual oil = all residual oil categories, except
power plants - residual oil used for manufacturing
Otherwise, the calculations are based on Bureau of Mines
data [7] and data derived by Wai den as summarized below:
Commercial distillate oil = [a* x (distillate oil used
for heating - residential distillate oil)] + distillate
oil for military use
Commercial residual oil = [a* x residual oil used for
heating] + residual oil for military use
*a is a commercial fraction based on commercial employment
b is an industrial fraction based on manufacturing employment
a + b = 1
2-13
-------
c. Industrial
For the years for which data are available from the Census
of Manufactures report mentioned above, distillate and residual oil state
totals are taken directly from that publication. Otherwise, the following
calculations are performed:
Industrial distillate oil = [b* x (distillate oil used
for heating - residential distillate oil)] + distillate
oil for industrial use + distillate oil for oil company
use
Industrial residual oil = [b* x residual oil used for
heating] + residual oil for Industrial use + residual
oil for oil company use
d. Heavy Duty Vehicles (HOV)
Heavy duty vehicles are defined as all vehicles weighing
more than 6.000 Ibs gross weight. Four subcategoHes are considered in
these calculations:
HOV,: 6.001 - 10,000 Ibs
HDV2: 10,001 - 20,000 Ibs
HDV3: 20.001 - 26,000 Ibs
HDV4: Greater than 26,000 Ibs
*a is a commercial fraction based on commercial employment
b is an industrial fraction based on manufacturing employment
a + b = 1
2-14
-------
Statewide HDV gasoline use is calculated as follows:
i=4
HDV
gasoline use = [ £ (HDV- x average miles./miles per
1
gallon-j) + (commercial buses x average gallons/bus) +
(school buses x average gallons/bus)] - [diesel and
butane use]
HDV diesel use by state is reported by the Federal Highway Administration [9],
e. Light-Duty Vehicles (LDV)
It is assumed that only a negligible number of light-duty
vehicles use diesel.
LDV gasoline use = total gasoline sales [9] - HDV gasoline
2. County Allocation of Fuel Oil
a. Residential
Countywide use of residential distillate oil is calculated
by means of the formula:
Residential distillate oil use (gallons) = (.01288 x
degree-days + 30.41 x average rooms per housing unit
- 79.54)/.14
This relationship was developed by Maiden for this project by means of a
stepwise regression analysis based on residential gas data obtained from a
number of gas companies throughout the nation (see Appendix A).
2-15
-------
b. Commercial
County-wide distribution of commercial distillate and resi-
dual oil is based on the-following steps:
(1) Calculation of fuel used in each county by hospitals,
hotels, schools, colleges, and laundries, based on correlative relationships
between fuel use and employment developed by Walden (see Appendix B).
(2) Determination of the fractions of this fuel use attri-
butable to distillate and residual oil based on the fuel use patterns in
each state.
(3) Calculation of distillate and residual oil consumed by
housing units using oil in structures of 20 units or more in each county.
It was assumed that units in structures of more than 50 units consumed
residual oil and that all other residential units included under the com-
mercial category consumed distillate oil.
(4) Subtraction of the oil totals for the above six sub-
categories from the state totals for commercial distillate and residual oil
use.
(5) Subtraction of the employment totals for the categories
mentioned under (1) from the respective county and state employment totals.
(6) Allocation of the remaining state fuel oil figures by
means of the adjusted county and state employment figures.
(7) Addition of the oil used by the six subcategories to
the county oil figures obtained in Step 6.
c. Industrial
County-wide distribution of industrial distillate and resi-
dual oil is performed by means of industrial employment figures which have
been adjusted by means of a fuel intensity factor which is industry dependent
(two-digit SIC).
2-16
-------
adjusted industrial employment.
Industrial fuel oil . = -
county i adjusted industrial employmentj
x industrial fuel oil . . T
state I
d. Heavy-Duty Vehicle (HDV)
County use of diesel by HDV's is obtained by using county
truck registrations [10] as the apportioning factors to be applied to the
state totals.
trucks > 6000 lb-
Diesel HDV = - 1 x diesel HDVcta.Q T
county i trucks > 6QOO 1bi state I
In order to. determine county use of gasoline for HDV's
the following calculation is performed:
(gasoline HDV)county . = J HDV.. x avg. miles j(I)/MPG. (I)
J '
population (census yr)^
+ - ; - x (commercial busesT x 7276 gallons/bus)
population (census yr) l
population (census yr)^
+ - : - x (institutional buses x 1058 gallons/bus)
population (census yr)j I
trucksi
- x (diesel and butane)
trucks i
2-17
-------
e. Light-Duty Vehicle (LDV)
The method developed for county allocation of gasoline
used by LDV's is as follows:
(1) If vehicle miles are available by county they are to
be used as the distributive factor in the form:
vehicle miles-
gasoline LDV . . = - x total gasoline,..,...,- T
county i vehicle miles state l
- gasoline
(2) If vehicle mile data by county are not available, the
sum of the registered automobiles and trucks under 6,000 Ib in each county
[10] is adjusted by means of a rural /urban miles per vehicle index and
used as the distributive factor in the form:
LDV. x index.
gasoline LDVcounty . = _J - ! - x total
(LDV, x Index,)
1=1 1 1
- gasoline
where state i consists of n counties.
C. RESULTS FOR 13 SELECTED COUNTIES - 1970 and 1971
Tables 2-4 and 2-5 show the results of the methods summarized above
for 13 selected counties for 1970 and 1971 respectively. In order to obtain
these results, it was necessary to subtract point source fuel oil use from
2-18
-------
CO
UJ
t_4
r
^"*
)
CD
( 5
o
UJ
1
p^
=3- uj
CM UJ
co
UJ
CO CD
>
C3 CU
(/I
>*-^
rO O
CD
Q ^^
CO
>1 C
> O
(O ,
QJ '
GJ T~ fr<
e CD
'o =1
to >, O
CD 13 4->
1 »
cn
f^
a
-o
i
i i/>
as a)
'r **"*
s_
4->
(/>
-a
c a>
fO
,
,
i
-*->
01
a
,_
f^
=5
-0
i
t"3 QJ \f)
o 'a!
S- S-
CJ S-
E:
Q
r Ol
i~~ fO
-U i
c: i
r- £) to CD C\J «^J" to CD to O CTt
cOLnr^Lnr^-i r^-^j-, i i r~-ro
air ^j-ioroc?iCMOi ioior-^i
CMtoro ror i i co
,
OCDOOOOOOOCDCDOO
CO CD CD i LD to CTi CO r^. CT> r i LO
i cO'^t-^i-rOi i r i cvjcxiioin
i «^- LT> vl" fO I **
CM
^t* ^t* CD "^J- CD O> «^?" CD CD CM *^j" C\J CD
COCMO<^> criuiSo i cncn
CM <^- CO CD r
*> *>
to ro
( CD CD r ro f^^ *^" CO CO CM r^ CO CD
r~- CM ^T . CM 1
LO r I r
^-LococnLncr,Or-cr>cMcr.LOLn
r CTt CD CO C\J CD CD LO r OO ^" CO
r LOCO UO O . CO O
* * "
r-OOCD<-OsjOr^sI^OOJi r CO
«^j" co co i r^. vo ^** t*o UD r CD T^*
C\jCn^l£>CMi i LO
CO CO CO
Ci o cut- i/) rjoc
OOJO> J-OJC-i c" "* o
i- C *r- (U E t/) r O tS -C CQ
OJ *^ C13 ~O *r QJ -^ 1 C QJ QJ
ij j ^> jj ^- _\^ c £ > 01
*4 C/l C !3 r S^- <"3 r C i" r C
r~^ I to CO CO 3» 1-L- (/) CO ?-~ d- CI3 N^
2-19
-------
oo
UJ
t~
r~~>
O
c_>
Q
UJ
m |
1 C_>
CM uj
. j
UJ UJ
_J CO
<^ ClC
r 0
j^t
OO
r
1
CO
t-Lj
r
r»^.
CT>
i
a "0]
t/>
>> O)
ro Q
CD
1C
-t-J
^5
Q .
CO
> O
ra ,
CD r
~ i _' rti
CD
C 3
r- >, 0
i <-> -cr
0 zi -i-J
to Q '
rp
.C
CD
_J
rd
TD
r
( CO
fO fl)
i rv
S_
J
"O cu
c »->
i i ra
, .
,
1
4-J
CO
o
ro
X>
.,
i VI
ro CU ^~~-
r- a: in
O i
5- CD
CD S-
E i.
E ro
O CD CO
C_J> -4-i
ro
r- O
i r~
-I-1 r
CO
.,
Q
i CD
CO +>
i- CO
C i
CD T-
-a 4->
i CO
CO -i
GJ Q
a:
;>j
4_>
CI
3
O
' O
0
cn
r--
in
i
CD
CO
cn
cn
CM
CD
CD
vf
(
CO
CO
1
cn
co
ro
CM
ro
cn
CM
i
«^j-
c:
o
^_
cu
14-
M-
O)
0
0
ro
,
r^
i
0
CD
CO
^j-
in
1 '
CD
O
O
[
ro
CXI
co
CO
in
r~
t~^
O
CO
r^
ro
r>
ro
r^.
r-~
r^
r>
CM
CD
CO
ai
i
cu
CJl
C3
j^
u c:
S- ro
3 L"-
cn
r
in
CD
CD
=3-
O
,-_
O
o
,^
CO
CO
r
co
o
ro
,
in
i
,_
CO
OO
CO
oo
CO
CO
1
3
o
1
4_)
OO
i cn
00 CO
cr* r^
CM
oo co
OO CO
CD *^J~
r CM
CD 0
CD OO
in CM
cn in
, CM
o cn
i CD
,
ro ,
^J"
CD CD
p.
co cn
CO ^^
CXI
CO
e~
Q. CO
co _e
C Q)
-^ £Z
i C
O) -r-
CQ 21
r^. CD
co in
CO
O CM
in CM
CM r^
1--, LO
r
0 CD
O CO
cn o
00 .
CT> r
cn r^^
^J- CD
<^t* '
CM
CM CO
CM
f
OO CM
cn in
\
00 .
o
,
r~
O C
fti C3
O) *->
CQ CO
CU
E >
CO ro
D_ CD
CD
CD
«^T
CM
O
CO
CO
CM
OO
CD
O
i
cn
CO
CD
CD
,
, .
in
*
r
CO
CO
00
cn
o
01
^
oo
cn
c
r
^
2-2Q
-------
the statewide commercial and industrial fuel oil totals, and also to
subtract I he corresponding point source employment figures from the
employment data for the 13 counties. Thus, the figures in Tables 2-4
and 2-5 are for area sources only.
Tables 2-6 and 2-7 show the seasonal fluctuation patterns for
residential and commercial fuel oil use for 1970 and 1971. These figures
are based solely on degree-day fluctuations. Seasonal fluctuations in
industrial residual oil use are summarized in Table 2-8 by two digit SIC,
and are based primarily on summaries made by Wai den from NEDS point source
data.
Table 2-9 shows average sulfur content for distillate and residual
oil used in the selected counties in 1970 and 1971. These figures were
taken from a variety of Bureau of Mines publications, as well as summaries
made by Wai den from NEDS point source data.
These county-wide figures are available for all counties of the
United States and Puerto Rico for 1972 in the form of a computer printout
and NEDS area source punched cards provided to EPA-NADB, Durham, North
Carolina.
2-21
-------
TABLE 2-6
SEASONAL FLUCTUATIONS OF FUEL USE FOR RESIDENTIAL AND COMMERCIAL
SPACEHEATING - 1970 (BASED ON DEGREE-DAYS)
County
Belknap, N.H.
Franklin, Ma.
Worcester, Ma.
Baltimore, Md.
Palm Beach, Fl .
St. Louis, Mo.
Minnehaha, S.D.
King, Wa.
Galveston, Tx.
Jefferson, Al .
Boulder, Co.
Los Angeles, Ca.
San Diego, Ca.
Jan. -March
51%
51%
51%
59%
76%
57%
51%
37%
72%
59%
45%
48%
45%
April -June
12%
13%
13%
9%
0%
7%
10%
20%
2%
4%
15%
17%
19%
July-Sept.
3%
3%
3%
0% .
0%
0%
3%
6%
0%
0%
3%
0%
0%
Oct. -Dec.
34%
33%
33%
32%
24%
36%
36%
37%
26%
37%
37%
34%
36%
2-22
-------
TABLE 2-7
SEASONAL FLUCTUATIONS OF FUEL USE FOR RESIDENTIAL AND COMMERCIAL
SPACEHEATING - 1971 (BASED ON DEGREE-DAYS)
County
Bel knap, N.H.
Franklin, Ma.
Worcester, Ma.
Baltimore, Md.
Palm Beach, Fl .
St. Louis, Mo.
Minnehaha, S.D.
King, Wa.
Galveston, Tx.
Jefferson, Al .
Boulder, Co.
Los Angeles, Ca.
San Diego, Ca.
Jan. -March
50%
52%
52%
58%
94%
59%
51%
40%
71%
63%
41%
42%
49%
April-June
14%
15%
15%
11%
6%
9%
10%
20%
3%
9%
16%
15%
17%
July-Sept.
3%
2%
2%
1%
0%
1%
3%
6%
0%
0%
5%
0%
0%
Oct. -Dec.
33%
31%
31%
30%
0%
30%
36%
34%
26%
28%
37%
43%
34%
2-23
-------
TABLE 2-8
SUMMARY OF SEASONAL FLUCTUATIONS OF RESIDUAL OIL
USED BY INDUSTRIAL SOURCES
SIC
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Industry Group
Ordnances
Food
Tobacco
Textile
Apparel
Lumber & Wood
Furniture
Paper
Printing
Chemicals
Petroleum
Rubber & Plastics
Leather
Stone, Clay, Glass
Primary Metals
Fab. Metals
Machinery, exc. electr.
Electr. Equipment
Transportation
Instruments
Miscellaneous
Winter
36%
29%
25%
28%
37%
25%
31%
26%
38%
27%
25%
30%
32%
26%
26%
28%
29%
35%
40%
27%
37%
Spring
22%
29%
25%
25%
25%
25%
25%
25%
23%
25%
25%
25%
25%
25%
28%
25%
25%
24%
16%
25%
25%
Summer
14%
21%
25%
22%
13%
25%
19%
24%
13%
23%
25%
20%
18%
. 24%
24%
22%
20%
17%
11%
23%
14%
Fall
28%
21%
25%
25%
25%
25%
25%
25%
26%
25%
25%
25%
25%
25%
22%
25%
25%
24%
33%
25%
24%
2-24
-------
TABLE 2-9
SULFUR CONTENT AVERAGES FOR DISTILLATE AMD
RESIDUAL OIL BURNED BY AREA SOURCES
(percentages)
County
Belknap, N.H.
Franklin, Ma.
Worcester, Ma.
Baltimore, Md.
Palm Beach, Fl .
St. Louis, Mo.
Minnehaha, S.D.
King, Wa.
Galveston, Tx.
Jefferson, Al .
Boulder, Co.
Los Angeles, Ca.
San Diego, Ca.
1970
Distillate
0.21
0.21
0.21
0.21
0.17
0.28
0.28
0.21
0.17
0.17
0.32
0.21
0.21
Residual
1.90
2.00
2.00
1.47
2.00
1.67
1.67
1.64
1.73
1.73
2.15
1.64
1.64
1971
Distillate
0.21
0.21
0.21
0.21
0.19
0.27
0.27
0.21
0.19
0.19
0.32
0.21
0.21
Residual
1.68
1.85
1.85
1.00
1.30
1.64
1.64
1.56
1.73
1.73
2.15
1.56
1.56
2-25
-------
REFERENCES - SECTION II
1. C1imatologi cal Data, monthly summarized station and divisional data,
U.S. Dept. of Commerce, National Oceanographic and Atmospheric
Administration, Ashville, N.C.
2. Burner Fuel Oils, Mineral Industry Surveys, Bureau of Mines, Bartles-
ville, Okla.
3. Fuel Oils by Sulfur Content, Mineral Industry Surveys, Bureau of Mines,
Washington, D.C.
4. Oil Availability by Sulfur Levels, Bureau of Mines, Washington, D.C,,
1971.
5. Guide for Compiling a Comprehensive Emission Inventory, Environmental
Protection Agency, Research Triangle Park, N.C., March 1973.
6. 1970 Census of Housing - Detailed Characteristics, U.S. Dept. of
Commerce, Washington, D.C.
7. Mineral Industry Surveys, Annual Fuel Oil Sales, U.S. Bureau of
Mines, Washington, D.C.
8. 1972 Census of Manufactures, Special Report Series: Fuels and Electric
Energy Consumed, IT.S. Dept. of Commerce, Washington, D.C.
9. Highway Statistics, U.S. Dept. of Transportation, Federal Highway
Administration, Washington, D.C.
10. Registration data available from R.L. Polk Co., Detroit, Mich.
2-26
-------
III. DISCUSSION OF METHODOLOGY
A. RESIDENTIAL
1. State Fuel Oil
Four different methods were attempted to arrive at the
residential use of distillate oil by state:
(1) Using EPA's suggested estimate of 0.18 gallons/unit/
degree-day [1] to determine fuel oil consumed for spaceheating require-
ments.
(2) Using the formula:
# of oil burners x avg size (Btu/hr) x 8760 (hr/yr) x load
140,000 (Btu/gallon)
to determine fuel oil consumed for spaceheating requirements [2].
(3) Using a stepwise regression analysis which included
the independent variables: degree-days, price of fuel, per capita in-
come, and average number of rooms per housing unit, to determine fuel
oil consumed for spaceheating requirements.
(4) Using the formula:
# of oil burners x heat loss x degree-days x use factor
140,000 Btu/gallon x design range
to determine fuel oil consumed for spaceheating requirements, where the
heat loss was dependent on the average square feet per housing unit [3],
These calculations are discussed in detail in Appendix A.
Table 3-1 shows estimates for selected states of residential
fuel oil used for spaceheating as calculated by each of these methods. In
general, it was felt that the need to average and estimate so many of the
3-1
-------
TABLE 3-1
SUMMARY OF ESTIMATION METHODS FOR RESIDENTIAL FUEL
OIL USED FOR SPACEHEATING BY SFATE
Based on:
Massachusetts
Maryland
Missouri
Washington
Maine
Connecticut
Florida
Alabama
EPA<]>
25,678
9,853
2,494
9,533
8,700
15,984
1,927
128
Equipment^ '
42,005
10,307
3,466
16,109
8,326
19,665
1,389
245
Multiple, x
Regression^ '
14,896
7,048
1,234
4,786
3,951
10,475
n.a.
n.a.
Square Feet^ '
n.a.
n.a.
3,236
11,062
n.a.
n.a.
2,466
143
Sources:
(1) EPA Guide for Compiling a Comprehensive Emission Inventory
(2) FuelOil and Oil Heat; Intermediate Boiler Study, Walden
(3) Vlalden, see Appendix A
(4) Independent Gas Association of America; Yankee Oilman
3-2
-------
variables for regions as large as states, reduced the significance of
the more sophisticated methods. It was, therefore, decided to use the
[ I'A factor ho arrive at statewide residential distillate oil used for
spaceheating. Distillate oil used for nonspaceheating was calculated
by multiplying the number of housing units using oil for hot water
heating [4] by 250 gallons per year [1].
.
This calculation is used only when selected counties are
being processed. When all counties within a state are processed, the more
refined residential distillate oil figures for each county are summarized
to provide the state figure.
2. County Fuel Oil
Using residential gas data by municipalities provided by
various gas companies throughout the country, Walden performed a more
detailed stepwise regression analysis, resulting in the relationships
shown below.
gas use (Mcfs) = 0.01288 x degree-days + 30.41 x average
rooms per housing unit - 79.54
The multiple correlation coefficient (R^) was 0.67 and the
standard error was 20% of the mean. For more details on this regression
analysis, the reader is referred to Appendix A.
Assuming that gas and fuel oil consumed for residential use
are utilized in approximately the same fashion, Walden converted this for-
mula to produce results in gallon of oil consumed and applied it to the
housing units using oil in each county [4] to estimate the distillate oil
used by these units.
3-3
-------
3. Additional Data
For census years, the number of housing units using oil for
spacehe.iting and nonspaceheatimj purposes are easily available [4]. In
order to estimate the number of housing units using oil for spaceheating
in each state in the intermediate years, the following method is used:
(1) The number of oil burners installed in new homes [5]
during the time-span between the year of concern and the census year is
expressed as a percentage of the number of oil burners in use at the end
of the census year (factor 1).
(2) The number of conversions to oil burners [5] minus the
oil burners lost to other fuels [5] during the time-span between the year
of concern and the census year is expressed as a percentage of the number
of oil burners in use at the end of the census year (factor 2).
(3) Housing units using oil in year t = housing units using
oil in the census year prior to t + [(factor 1 + factor 2) x housing units
using oil in the census year prior to t].
Other possible sources to update the census data on housing
units were considered,but both the Construction Reports [6] and data pub-
lished by the F. W. Dodge Co. [7] were found to be incomplete for residen-
tial units.
The same state percentage change is assumed for housing units
using oil in all counties within that state. This is only an estimate, and
the error factor is estimated to be within 2 or 3 percent.
B. COMMERCIAL
1. State Fuel Oil
The commercial category is an extremely complex one, and in
the county-wide allocation methodology, Wai den has attempted to stratify
3-4
-------
the commercial users into a few subcategories which show distinct dif-
ferences in fuel use patterns.
On a statewide level, the fuel oil figures include all oil
burned in stationary sources which is not included under the residential,
industrial, or power plant categories. This means:
Commercial distillate oil = all distillate oil categories,
except power plants - residential distillate oil (see page
3-3) - industrial distillate oil (see page 3-10).
Commercial residual oil = all residual oil categories, ex-
cept power plants - industrial residual oil (see page 3-11).
2. County Fuel Oil
The logic of the methodology to determine commercial fuel
oil consumed on a county-wide basis is that of separating out the major
categories which consume fuel oil in a special way, and distributing the
remaining fuel oil by means of adjusted county commercial employment
figures [9].
For this purpose, Wai den performed several linear regres-
sions to determine the correlation between employment and fuel used by
the following subcategories:
(1) Hospitals
(2) Schools
(3) Colleges
(4) Laundries
(5) Hotels
3-5
-------
The results of these analyses are discussed in detail in Appendix B,
and are summarized below in Table 3-2.
TABLE 3-2
REGRESSION RESULTS FOR FUEL OIL '
USED BY COMMERCIAL SUBCATEGORIES
Category
Hospitals
Schools
Colleges
Laundries
Hotels
Dependent
Variable
Oil
Oil
Oil
Oil
Oil
Independent
Variable
Employment
Employment
Employment
Employment
Rooms*
R2
0.81
0.58
0.67
0.72
0.96
Slope
0.715
2.97
0.546
0.355
1.09
Intercept
+208.5
+ 76.2
- 40.9
+ 8.4
+ 41.5
Rooms/Employment ratios have been calculated by state (see Appendix B).
Using these relationships it is possible to separate out
the fuel oil used by categories in each state based on employment
figures [9], once it is known what percentage of the commercial establish-
ments in each state use oil and what grade of oil (distillate or residual)
is likely to be used.
In order to determine the fuel choice for the subcategories,
it was assumed that the ratio of coal, distillate'oil, residual oil and gas
for the subcategories in a state would not differ significantly from the
fuel use pattern in that state for the commercial category as a whole.
3-6
-------
Therefore, Walden collected data by state on commercial coal use [10] and
commercial gas use [11]. The results for the selected states are shown in
Tables 3-3 and 3-4 for 1970 and 1971, respectively.
The above methods enable us to calculate the distillate and
residual oil use for the five categories in each county.
In addition, Walden estimated the distillate and residual oil
consumed by housing units in structures of 20 units or more in each county.
This was done by assuming that of the units using oil, those in structures
of 50 units or more used mostly residual oil [12,13] and all others used
distillate oil. Since the 1970 Census of Housing did not provide suffi-
cient data to separate out the units :n structures of more than 50 units
on a county basis, it was decided to aoply the state percentages repre-
sented by these units to the county housing unit figures.
The final method for county allocation of commercial fuel oil
can thus be reduced to the following steps:
(1) Calculate distillate and residual oil consumed in each
county by hospitals, schools, colleges, laundries, hotels and residential
units in structures of 20 units or more.
(2) Adjust the statewide distillate and residual oil figures
to exclude the oil consumed in all counties by the above categories.
(3) Adjust county and state commercial employment to exclude
employment for the first five subcategories.
(4) Distribute the remaining state fuel oil figures by means
of the following method:
adjusted commercial employment in county i
fuel oil consumed in county i = adjusted commercial employment in state I
x fuel oil consumed in state I
3-7
-------
TABLE 3-3
1970 COMMERCIAL COAL, OIL, GAS USE BY STATE
State
New Hampshire
Massachusetts
Maryland
Florida
Alabama
Missouri
Texas
South Dakota
Colorado
California
Washington
Total Btu
(Billions)
7,569
235,601
50,386
40,641
33,888
96,039
140,667
12,031
68,584
307,991
54,987
Coal
1.3%
0.6%
3.5%
6.6%
6.4%
3.2%
0.1%
24.5%
11.9%
0.8%
3.6%
% of
Distillate
31.3%
35.2%
16.5%
26.4%
7.3%
13.6%
16.4%
9.4%
6.5%
6.7%
20.3%
Total
Residual
50.8%
52.0%
46.1%
18.2%
3.3%
9.2%
3.3%
0.6%
2.3%
24.2%
39.7%
Gas
16.5%
12.2%
33.9%
48.8%
83.0%
74.0%
80.3%
65.5%
79.2%
68.3%
36.3%
3-8
-------
TABLE 3-4
1971 COMMERCIAL COAL, OIL, GAS USE BY STATE
State
New Hampshire
Massachusetts
Maryland
Florida
Alabama
Missouri
Texas
South Dakota
Colorado
California
Washington
Total Btu
(Billions)
10,215
249,841
53,178
43,577
33,249
90,647
131,627
10,899
70,874
318,268
54,198
Coal
0.4%
0.1%
1.7%
3.0%
5.0%
2.0%
0.1%
17.3%
7.8%
0.0%
2.5%
% of Total
Distillate
31.6%
37.0%
14.4%
23.2%
7.8%
14.0%
18.9%
0.0%
7.2%
11.5%
21.2%
Residual
53.2%
50.0%
48.7%
24.2%
2.0%
5.8%
2.4%
0.1%
2.0%
15.5%
35.0%
Gas
14.7%
12.8%
35.2%
49.6%
85.1%
78.1%
78.6%
82.6%
83.0%
73.0%
41.3%
3-9
-------
(5) Total commercial fuel oil in each county would be the sum
of Steps 1 and 4.
Methods suggested by the EPA to allocate commercial fuel oil
were previously based on a straightforward distribution by means of employ-
ment ratios [1]. This assumed the same ratio of fuel use per employee for
all the various types of establishments included in the commercial category.
The results of the regression analyses performed for the various subcate-
gories by Walden show that this assumption was incorrect. The methodology
used in this project to allocate commercial fuel use to counties is, there-
fore, considered to be a significant improvement over previous methods.
C. INDUSTRIAL
1. State Fuel Oil
For the years for which industrial fuel oil use is available
by state and SIC group from the Census of Manufactures [14], the distillate
and residual oil consumed by manufacturers in each state are taken straight
out of that publication. Otherwise, the Bureau of Mines data [8] are used
to arrive at total industrial fuel oil use by state as shown below:
Industrial distillate oil = [b* x (distillate oil used
for heating [8] - residential distillate oil (see page
3-3 )] + distillate oil for industrial use [8] + distil-
late oil for oil company use [8]
tnanufactyri ng empl oyment
b= commercial + manufacturing employment ' where the commercial category
includes wholesale and retail trade, finance, insurance and real
estate, and services [10].
3-10
-------
Industrial residual oil - [b* x residual oil used for
heating [8]] + residual oil for industrial use [8] +
residual oil for oil company use [8]
2. County Fuel Oil
The logic of the methodology to determine industrial fuel oil
consumed on a county-wide basis is that of adjusting the county employment
figures [9] for two-digit SIC groups by a fuel intensity factor, and using
these adjusted employment figures to allocate the state fuel oil figures to
the appropriate counties.
For those years for which distillate and residual oil data
by state and two-digit SIC are available [14], the fuel intensity factors
are represented by Barrels/Employment factors, which are developed by state
and SIC groups. For the years in which such state data are not published,
nationwide fuel intensity factors by two-digit SIC are used to adjust the
employment figures. Table 3-5 shows the nationwide fuel intensity ratios
which were applied to the 1970, 1971 and 1972 industrial employment figures,
D. HEAVY DUTY VEHICLES
1- State Gasoline and Diesel - HDV
Four heavy-duty vehicle (HOY) truck categories are used:
manufacturing employment
b= commercial + manufacturing employment* where the commercial category
includes wholesale and retail trade, finance, insurance and real
estate, and services [10].
3-11
-------
TABLE 3-5
NATIONWIDE FUEL INTENSITY RATIOS
SIC
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
19
Industry
Food
Tobacco
Textile
Apparel
Lumber & Wood
Furniture
Paper
Printing
Chemicals
Petroleum
Rubber & Plastics
Leather
Stone, Clay, Glass
Primary Metals
Fab. Metals
Machinery, exc. electr.
Electr. Equipment
Transportation
Instruments
Miscellaneous
Ordnance
All Manufacturing
Fuels Purchased
Million $^
423.0
8.5
123.3
34.6
147.1
27.2
462.7
46.6
772.2
384.4
74.0
14.8
511.9
1,103.9
174.8
166.5
107.3
160.1
28.3
27.1
21.8
4,829.1
Employment
Thou/2)
1,595
71
948
1,376
555
446
668
1,082
881
136
558
304
592
1,268
1,354
1,996
1,881
1,817
405
422
343
19,762
Ratio
0.27
0.12
0.13
0.03
0.27
0.06
0.69
0.04
0.88
2.83
0.13
0.05
0.86
0.87
0.13
0.08
0.06
0.09
0.07
0.06
0.06
0.24
Sources: (1)
1970 Annual Survey of Manufacturers, Fuels and Electric
Energy Used by Industry Groups, U.S. Dept. of Commerce,
Washington, D.C.
(2) 1970 County Business Patterns, U.S. Summary, U.S. Dept.
of Commerce, Washington, D.C.
3-12
-------
HDV-] are trucks with gross weights between 6,001 and ] 0,000 Ib.
HDV2 are trucks with gross weights between 10,001 and 20,000 Ib.
HDV3 are trucks with gross weights between 20,001 and 26,000 Ib.
are trucks with gross weights greater than 26,000 Ib.
In order to obtain the total motor fuel consumed by all HDV
categories in each state, the following calculation is performed:
1=4
Motor fuel for HDV = £ (HDV.J x average miles -j /miles per
i=l
gallon-,-) + commercial buses x average gallons per bus +
school buses x average gallons per bus
The breakdown of trucks into the four weight categories is
available from the Census of Transportation (published every 5 years), or
from R. L. Polk in Detroit, Michigan. For 1971, such a stratification
was available for 15 states [15]. Wai den estimated the correct break-
downs for the remaining states for 1971, based on geographical location.
By means of interpolation between the 1967 data [16] and the 1971 data [15],
the 1970 breakdown of trucks by weight categories was obtained. For years
for which the Census of Transportation data on the R. L. Polk figures are
available, such manipulation of the data will be unnecessary.
It is expected that county-wide truck registration data by
weight categories will be supplied on a regular basis to the EPA by R. L.
Polk of Detroit, Michigan.
3-13
-------
Average miles per gallon by weight category are taken from
the Census of Transportation and are held constant over the five years
following that Census. Upon publication of the new Census of Transporta-
tion, these averages must be updated.
Miles per gallon figures by weight category were derived from
Road User and Property Taxes, a tri-annual publication of the Federal High-
way Administration. The curve from which this information was derived is
shown in Figure 3-1.
Data on buses were obtained from both the Federal Highway Ad-
ministration [15] and the Automobile Manufacturers Association [17].
Once total motor fuel consumed by HDV's has been calculated
for each state, gasoline usage by HDV's is obtained by subtracting total
diesel and butane use [15] from the motor fuel figures for each state.
Diesel totals for HDV's are estimated annually by the Federal Highway Ad-
ministration [18].
2. County Gasoline and Diesel - HDV
County use of diesel by HDV's is obtained by using county
truck registrations [19] as the apportioning factors to be applied to
total diesel used by HDV's in the respective state.
Diesel HDVrnim. . = trucks > 6000 Ib d1 , HDV
county n trucks > 6000 Ibj State l
In order to calculate the county use of gasoline for HDV's,
the county-by-county R.L. Polk truck registrations by weight categories
are used. If these data are not available, county registrations of trucks
3-14
-------
CD
o
o
o
o
2
z
<
or
cT>O
zp
O
o
o
tl_
o
<5
\»
CD ,_
J<
O
c:
o
0
-
(\i
k
O
o
Q
en
i
0)
in
CO
o
0
OJ
1
_ _ CD r- ID in
N011V9 b3cl 93~IIW
OJ
3-15
-------
by weight categories are estimated by assuming that the proportion of
registrations in each category is the same for all counties within a
state. The remaining calculation is similar to that used to obtain
the state totals for this category:
(gasoline HDV)CQunty . = [^ HDVj x avg milesj(I)/MPG.(I)
population (census yr)-
*- 7 - : ; - : - r*- x commercial buses x 7276 gallons/bus
population (census yr)j
population (census
+ - 7-~r~. - ; - * - r~ x institutional buses x 1058 gallons/bus
population (census yr), 3
trucks.
- . x (diesel & butane),
trucks I
An alternate method to allocate gasoline and diesel use by
heavy-duty vehicles to counties was analyzed. This method was based on a
one-time study on truck vehicle miles performed by the Federal Highway
Administration. The method consisted of the following steps:
(1) Calculate the vehicle miles traveled on a statewide
basis by respectively diesel and gasoline trucks on interstate highways.
(2) Estimate the corresponding county vehicle miles by
means of county level interstate highway mileage totals.
(3) Divide the county vehicle miles by the average miles
traveled per gallon of fuel to obtain the gasoline and diesel consumed
by heavy-duty vehicles on interstate highways in each county.
(4) Estimate the gasoline and diesel used in each county
by heavy-duty vehicles on other than Interstate highways.
(5) The sum of steps (3) and (4) for respectively gasoline
and diesel consumption would be the county-wide figures for the consumption
of motor fuels by heavy-duty vehicles.
3-16
-------
The method was not used because the procedure was considered too time-con-
suming and costly, and also because the truck vehicle mile data required
would not be available on a yearly basis. For a more detailed description
of the analysis of this method, the reader is referred to Appendix D.
E. LIGHT DUTY VEHICLES
1. State Gasoline - LDV
It was assumed that the diesel consumed by LDV's is negli-
gible.* This assumption should probably be re-examined in five years.
Statewide use of gasoline is reported both by the American
Petroleum Institute [20] and the Federal Highway Administration [15]. It
was found that the FHWA state data were more up-to-date and statistically
adjusted, and Maiden decided to use these figures instead of those of the
API. Statewide use of gasoline by light-duty vehicles is obtained by sub-
tracting the gasoline used by HDV's (see page 3-13) from the total gasoline
reported for each state.
2. County Gasoline - LDV
County-wide gasoline figures for LDV's are obtained preferably
by using vehicle miles as a distributive factor on the statewide total gaso-
line figures [21] and subtracting the county use of gasoline by HDV's from
the result. For those states for which no county-wide vehicle miles are
available, registrations of light-duty vehicles by county [20], adjusted by
a rural/urban factor, are to be used.
*
All on-highway consumption of diesel was assumed to be used by HDV's.
3-17
-------
The rural/urban factors were derived from a statistical analy-
sis performed by Walden (see Appendix C). In short, Walden attempted to
find significant relationships between miles traveled by cars annually and
the degree of ruralness of each county, using data for seven states. The
multiple correlation coefficients were found to be too low and the standard
errors too high to permit the use of regression curves for this estimation
process. The resulting factors are summarized in Table 3-6.
For those states for which no vehicle miles were available,
county-wide registrations of automobiles and trucks of less than 6,000 Ib
weighted by the indexes shown in Table 3-6, were used as the distributive
factor to be applied to the statewide gasoline totals. This resulted in
total gasoline use by county. By subtracting the previously calculated
gasoline used by HDV's, the county-wide gasoline use by LDV's is obtained.
3-18
-------
TABLE 3-6
AVERAGE MILES PER VEHICLE BY % RURAL CATEGORIES
State
California
Washington
Kansas
Iowa
Georgia
Maine
Arkansas
Seven State Avg.
Indexes
Entire
Sample
15,152
18,722
16,155
15,643
23,140
15,261
18,652
Modified
Sample
13,047
17,092
15,179
14,604
20,262
15,261
.17,932
16,709
100
_< 25%
10,729
11,940
10,706
11,468
13,185
11,210
11,039
11,217
67
26-50%
12,202
15,793
13,622
12,141
17,680
15,396
15,908
14,572
87
51-75%
14,803
15,778
15,306
15,501
19,687
14,876
18,144
17,238
103
76-100%
16,920
21,855
17,225
15,862
22,461
15,995
19,458
19,261
115
3-19
-------
REFERENCES - SECTION III
1. Guide for Compiling a Comprehensive Emission Inventory, Environmental
Protection Agency, Research Triangle Park, N.C., March 1973.
2. Systematic Study of Air Pollution From Intermediate Size Fossil-Fuel
Combustion Equipment. Maiden Research Corporation. Cambridge, Mass.,
March 1971.
3. Fuel Trades Fact Book, New England Fuel Institute, Boston, Mass.,
March 1973.
4. 1970 Census of Housing - Detailed Characteristics, U.S. Dept. of
Commerce, Washihgton, D. C.
5. Fuel Oil and Oil Heat, Cedar Grove, N.J., October 1972.
6. Construction Reports, 1971, Housing Authorized by Building Permits
and Public Contracts, U.S. Dept. of Commerce, Washington, D.C.
7. Reports available from the F.W. Dodge Division of McGraw Hill.
8. Mineral Industry Surveys, Annual Fuel Oil Sales, U.S. Bureau of
Mines, Washington, D.C.
9. County Business Patterns, U.S. Dept. of Commerce, Washington, D.C.
10. Minerals Yearbook, U.S. Bureau of Mines, Washington, D.C.
^- Gas Facts, American Gas Association, Arlington, Va.
12. Personal communication with Mr. Nespeco of the National Oil and
Fuel Institute, New York, N.Y.
13. An Analysis of the Economic Impact of the Massachusetts Air Pollution
Control Regulations, Walden Research Corporation, Cambridge, Mass.,
December 1972.
14. 1972 Census of Manufactures, Special Report Series, Fuel and Electric
Energy Consumed, U.S. Dept. of Commerce, Washington, D.C.
15. 1971 Highway Statistics, U.S. Dept. of Transportation, Federal .Highway
Administration, Washington, D.C.
16. Census of Transportation, 1967, U.S. Dept. of Commerce, Bureau of
Census, Washington, D.C.
17. 1971 Motor Truck Facts, Automobile Manufacturers Association, New
York, N.Y.
3-20
-------
18. Special diesel estimates by state provided to Walden by Mr. L.L.
Liston of the Federal Highway Administration, Highway Statistics
Division, Washington, D.C.
19. Registration data available from R.L. Polk Co., Detroit, Mich.
20. Total Gasoline Consumption in the United States, American Petroleum
Institute, Washington, D.C.
21. Obtained by contacting all state highway departments in the U.S.
3-21
-------
IV. SULFUR CONTENTS AND SEASONAL FLUCTUATIONS
A. SULFUR CONTENTS
The following sources contain information which was used in this
project to determine the sulfur content of fuel oils used in the United
States:
(1) Actual data from NEDS point source files.
(2) Burner Fuel Oils, Mineral Industry Surveys, Bureau of
Mines, Bartlesville, Oklahoma (annual).
(3) Fuel Oils by Sulfur Content, Mineral Industry Surveys,
Bureau of Mines, Washington, D.C. (monthly).
(4) Oil Availability by Sulfur Levels, Bureau of Mines,
Washington, D.C., August 1971.
(5) Import Supplement to Oil Availability by Sulfur Levels,
Bureau of Mines, Washington, D.C., June 1972.
1. Sulfur Content Reported for NEDS Point Sources Using Oil
The NEDS data were summarized by Wai den and the results are
shown in Tables 4-1 and 4-2 for selected counties. In general, it may be
said that the average sulfur contents which could be derived from these
data are not applicable to area sources. The distillate oil used by point
sources is biased towards grade #4 distillate oil, and the sulfur contents
of the residual oil used are also slightly upward biased. It is therefore
not recommended to use the sulfur contents derived from these summaries for
area sources.
4-1
-------
CQ
CO
LU
O
fv*
i
O
co
1
1 I
Oh-
r\ 2t ., ^
._ H CO
^"" ' c
f 1
o o in
I 1 1
y rv iH.
"~i *~-^
co u. .
O^o
o co £-
1 >-N
t i CQ
o
1
e^
"~>
0
i i
00
UJ
o
CO
A
O
CO
1
LO
CM
LO
CM
1
CM
^^
CM
1
LO
i
fe°^
LO
a
r
1
^
r
1
U~>
o
8
CD
1 1 1 1
1 1 1 1
CD O O
o cn r~.
o en CM
I ** r»
1 CO CO
1
O 0 0
LO O CD
CO CD CD
* * n 1
CM <* r 1
IO UD
I l
0 CD O
O CD O
«* CO O
1 r, « «
1 r VQ LO
=! LO O
*
CD O
CD 0
cn o
i *i
1 1 r^ tD
1 1 CM U3
0 O O
O CM O
^}- r^ o
| r, « ^
i LO r^s co
i p«^
co
o
r*^
1 1 r 1
1 1 * 1
r
. .
. fO TD
1C rtJ SI 2:
^>^ * **
^ S- OJ
C O) t-
O. -r- -fJ O
rO i co E
c -^ cu -i-
-i£ E O +->
p f0 S i~~
CU 5- O rtS
CQ LL. 3: CQ
1 1 1 1 1 1 1 I 1
1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1
CO CO
CO "=f
1 III III
1 III III
O 0 0
O UD O
co o r-~
1 » * 1 1 1 1 1
1 r 1 1 1 1 1 1 O
CM
O CD 00
O CD CO 0
cn CD «* o
ft W^ ft ft
1 I LO O 1 1 CM 1 i
i i cvj r^ i i i CM
cn r
r-^
0 CD iO
CO O CO 0
CO O CO r
1 * 1 « 1 1 1 «
i r- i cn i i i r--
i CO
^J- O
CM «*
1 1 1 1 CO 1 1 1
1 1 1 1 - 1 1 1
LO
fQ fXS
, . O . (O O
u_ o x i as
2: -n |_ < . « c_>
> O CO
O CO rO c C r O
HJ «- -C fOOO "OJCD
QJ 3 (TJ *^g [ ^ to i- D^ CU
coO-c: coS-Qjc-r-
j Cl) » CD CU "O <«- r
r .C Ci *- 3 «/) C
fS-M't -r (dCUOOnJ
Q_coS!I i^CDrDCQ-_ICO
4-2
-------
1
s:
» *
o
IX
ca
a
LU 1
ZD LU
oo f
0 0
o o
C\J | p^
' * 1 ' )
* CD U.
LU _, =^
1 J }
ca < ^
£ a>-
^ »-H CQ
LU CO
o; i>i
0
u- cr:
o rD
CD
LJ OO
CD
»
LU
a:
LU
Q-
o
CO
A
CD
CO
1
LO
CNJ
LO
CM
1
CM
*S
CM
1
"1
,_!
LO
,_!
i
c?^
f
1
LO
CD
^^
tn
o
V
1 1 1 1 1 1 1 1 1 1 I 1
1 1 1 I 1 1 1 1 1 1 1 1
.
1 LD CD Cn 1 I 1 I I 1 1 I
iroojoiiiiiiii
c: c.
OOQLOI 1 r 1 1 1 r 1 1
CD «=r co i i iii it
r_
i CM co cr> i «* «* i i i i i
1 r CM «S- 1 1 1 1 1 1
i i cn co i iiacni ivoi
ii r i i en cn i i co i
i in n cn i LO i r i 1*3-1
1 CO 1 CO 1 1 1 r 1
r~ ^**
i i CD i icni i icni i
i i o i i S. QJ d
C -^ O) > _J S~ O r3 »d -M »r »r fO CJ O O
CQ Li- ^3 CO d. CO ST ^ C3 *^> CQ «J
1
1
1
1
1
1
cn
CO
LO
CO
oo
i
i
fd
O
o
CO
a
i
ca
c
cd
oo
4-3
-------
2. Burner Fuel Oils, MIS, Bureau of Mines
In this publication, which appears yearly, samples of fuels,
selected by their manufacturer as typical of that year's production of
that specific grade and brand [1] are taken. Walden applied the results
shown in these reports to selected counties (see Tables 4-3 and 4-4).
Although the Bureau of Mines sulfur figures are for large
regions, lacking the detail required for this project, it seems more
accurate to use the average sulfur contents reported by the Bureau of
Mines for area source calculations, instead of using the average sulfur
content of fuel oil used by the point sources in each county. The
Bureau of Mines sample is quite small for #4 and #5 grade fuel oil, but
comprises 134, 149 and 109 sample points for respectively #1, #2 and #6
fuel oil in 1970.
It would be extremely useful if the sample size for this
survey were significantly increased, enabling summaries of sulfur contents
of burner fuel oils by smaller regions. This would provide the EPA with
better sulfur content data to be used in area source emission calculations,
3. Fuel Oils by Sulfur Content, MIS, Bureau of Mines
In this monthly publication, sulfur content data are shown
for #4 fuel oil and residual oil imported into the United States. The
December issue usually contains a summary of the type shown in Tables 4-5
and 4-6. The problem with these data is that the sulfur content applies
to the fuel as imported into the state shown. It has been found that the
4-4
-------
s- %
t .
v^_X
<^
h-
^""^
1
LT)
r -.-,-.
CD
"cu
li. "+J*
,^^
>> cn
f
in '
-M
C LO
OJ =fc=
-t->
0
CJ
c
"% l^J-
(^ r^-
^
CO
a
*T>
^J)
CXI
r-C ~^
r^
>j
\ i
f~
^~l
O
CJ
0
(
CD
CJ
p^
» -1
O-
«cC
to ro
r^j r--.
i i
to
cr> O CJ -J-> to
ra 10 £ ca to s-
c: ^i o -r- OJ cj
^i c o 4-» e > **-
i f3 t- r r~" r *i
OJ >- O f3 r3 ra GJ
C3 U- ~3i CO d- CO "^
, .
CO
CXI
CD
CXI
co
CD
0
^
a
o -
s: co
* «
c
VO CO
i r
ro cx» r
r
O CD C
r
r-
^ cn c
i CD
o o
*
ft3
0
ra C
-CJ
o to
CJ CJ 1
_«
n
»
^»
r>
/)
r
D
1)
3
E_
3J
-
rs
Q
J7
t/> rd ' O >>
i _c: " ra CJ cn CJ
n ra i- 3: CTJ CJ >
o -c
_1 QJ
OJ C-r- J-
TD -«=C O
3
C r cn CO
sr
rs c: 10 c:
4-> »- O -r- o ra >>
CO 21
C3 i£ _J CO S- CJ
»-> <4- r-
<
/) O JQ
-«r ^-t
.» >^^
T3 V) ,
C
C $-
S- CJ
01 jc.
-M 4-1
to 13
rd O
LJLj CO
i
r3
^
4-J
C
cu
CJ
i
cr t-
-r-
+-> *i
;
HJ i
O !
2! C -.
t «
>^ QJ
-^ 4-*
CJ to ^
O O r-
ttT _£ x-
^ 4-> -r-
* to f^3
-- >
a a
^- O -t->
y u o
c c
- GJ
~ O-
B
-» r3 ra
- CO
-- * c:
4-5
-------
*
~
^
h^
^«
CJ
CO
1 , I
l-U
» i
c
u_
CD
ZD
_J
^c
«c
1
2*:
LJ
f
~-*
0
CJ
Ci
__D
U-
1
^D
CO
».o
^te
^
rrt
OJ CU
o. .c
i!? ^
(
LO
'r- *~
CD
ni
LL 4->
-C
>> CD
.0 -r-
CO
.[ ^
c: LO
01 =iS
4->
0
0
5_
r3
it- **
.->
CO
CI
cr>
ra
S-
£L)
> CM
-=C =fc
i
=^
^.^
-o
4->
c:
~5
O
CJ
c:
o
*
01
OJ
ct:
*
CL
cv cxi o-)
tj"> cxj vf i£) «^-
. . .
CJ 00 r r- i
t
c
-)
CM 00 *
co ra r vo re*
0 C r^ CO SZ
-
1
* *
co cn «rf- . in «4
viP co <£> fd r^
CD CD CD C CD
c
CD r~- CD CO CM
1 CO 1^. r I
CV f CM OO CM t-
. 1-
CD O CD O CD C
r~
CM CM CD CTi r
r-- ^*- co 01 r-- c
CD ' O CJ O CD
CD CD CJ O O
L
cu
r CD CD
« -o U- x i o. ra c
-craiz.^^ l -3; :£. co "CJ
_^ -^ " O 00
"~-S-o"oCC COftJ i CD
"CCJi- r3OO c rOCUtn
ci.'r- 4-> o GJ -i-1 in ^> cs i- "=» cn cy
-Ji tnj£tJWS_ Oi; CJ C-r-
c-i>:c)-^- oajtojx) Xca
-Vt:£><4- Cr- C7) c
i r3S_i i , t)- .c 13 ccoc
CJS_or3 f3OO-»J-^- O'r-Or3
cnu i:ca d-tD-D cos: C3 i^_ico
-
-
C i-
1
<3 r
M i
C . !
n
C. C1
C 5- ^- I-' C
i- CI ra L-
GJ JZ S- >, CJ
-M +j -M _~^ ^_>
"i 3 c: o f "
r3 O QJ O CJ i-
1 i ) CO CJ ex "S *.
J
11
^*
c
:>
-j
'ji
tj
-
3
""
»*
yi
c
'Z
*~
o
a
^j
n
K.
3
a
»
-^
T>
^
D
U
_J
_
^
2J
d
L-
^i
O
j-j
>>
i)
>
!_
3
O
=» t
^- CJ
-> lf~ f
/I O _O
ZJ tJ
3 <^ r
' | l ,f~.
-* lo r3
fl f3
3 £;
>- 0 -t->
U U O
c t;
C)
r; , i
ex
E
-^ (J rj
- co
-^ +: cr
4-6
-------
LU
o; au
ZD S-
LJ_ S~
_J (O
=D JD
to
rs
I O
2: ^:
LU 4->
O '
a;
LU r
ex r-^
CTl
>- i
ca
O
_J LU
1-1 Q
O I
-J<
LU -)
U-
CO
«3- LU
=tfc I
_
O CO
CO >-
I ca
C£
O
Q.
<
__
2^
_U
z:
p
g
_J
z:
LU
CJ
c£
1 i 1
LLJ
o.
CJ
) I
C£
I
to
f"H
C3
_J
=c
o
o
o
CM
LU
O
o
o
CM
1
O
O
o
^^
1
«
o
LO
0
LU
^
f-
co
CO LO LO LO
r-. LO CM LO
r *3~ r
r r
1 t 1 1
LO 1 1 1
CO
r-.
«d- i vo LO
i «3- LO
^- r~_ r
r
i LO cr> i
LO f-v
CM
co
h- Q
ZD LU ct
o to _i
i i ZD tO
Q i :r HH
2T C-> CJ
e£ LU LU ^ LU
_i s: 2: co o
>-> Uj z: >-ns) o
2: o < < 3;
LU O 21 SI Qi
0 3
LU
n_
=d-
D3
CM
t
r>
v^
_
*x.
no
no
_J
i
1 1 1 I~~. LO
CD «a-
CM
1 1 1 CM CM
CO
«^~
'
CO CO CO OO CTl
CM CO LO CM CO
i ->T fx. LD LO
LO
i co o c> r*^
LO CO *d" CO
CO «d" CTl
r- CO
0 - 2:
"Z. LU
_j a: z: a: a: _i
f >-
<: 3 i rs >~ to
^c ^ ^^
_i _i a: 3 3 z:
uo
0
0
CM
CM
CO
xf
CO
^f-
T~
0
^J*
^J*
f*-s.
CM
CM
CO
o
_J
^^
\
0
»-
CO CM O O1
r~- o c^- c3
CM CO «*
1 CTl O 1
CO CM
i co i en
r O
^j-
CO 1 1 1
fv.
CNJ
1 1 1 1
1 1
1
0 O
*-H Di
h~" ^C ^C
^j* ^f ^^ ^_J H-H
«C Q t-i 2:
1 i O ZC i i
I a: a: i o
< o o D; ct;
_J LU O HH
a: LU o :s: >
LU
0
_j
3O
O
LO
01
o
CO
CM
CM
«f
CO
M
1
_J
c^
h-
O
(
co
CO
«vt*
CM
CM
O
px.
co
CM
co
CO
«^f-
en
LO
1
^2
_j
c^
0
n
CO
^3
Q
2:
=)
O
O
ZL
LU
O_
LU
co
LU
CQ
o
CO
CO
co
-------
TABLE 4-6
IMPORTS OF RESIDUAL FUEL OIL BY PERCENT SULFUR CONTENT
BY STATES: JAN-DEC 1971 (thou barrels)
DISTRICT
STATE
P.A.D. I
NEW ENGLAND
CONNECTICUT
MAINE
MASSACHUSETTS
NEW HAMPSHIRE
RHODE ISLAND
TOTAL
CENTRAL ATLANTIC
DELAWARE
MARYLAND
NEW JERSEY
NEW YORK
PENNSYLVANIA
TOTAL
LOWER ATLANTIC
FLORIDA
GEORGIA
NORTH CAROLINA
SOUTH CAROLINA
VIRGINIA
TOTAL
P.A.D, II
ILLINOIS
MICHIGAN
MINNESOTA
TOTAL
P.A.D. V
CALIFORNIA
HAWAII
WASHINGTON
TOTAL
U.S., TOTAL
PERCENT SULFUR CONTENT
0-.50
872
3293
50
4214
29
354
27241
60528
16689
104840
274
274
-
-
-
-
109328
.51-1.00
10885
18C81
no
1768
30844
27547
6023
57485
27307
118361
14241
15
14256
3145
3145
-
-
166606
1.01-2.00
3865
2700
11759
3493
3568
25385
315
6682
1939
12862
4627
26426
17296
397
2350
2179
2797
25019
62
62
66
246
21
334
77225'
OVER 2.00
17297
14269
19657
2757
3833
57813
346
3565
1419
35967
7007
48305
28528
5204
5078
7335
29192
75337
733
12
745
416
416
182615
TOTAL
32917
16970
52790
6360
9219
118256
690
38149
36622
166842
1 55630
297932
60339
5601
7428
9529
31989
1 14885
' 3145
795
12
\ 3953
66
663
21
750
53S7?6
SOURCE: OFFICE OF OIL AND GAS
NOTE.. DATA MAY NOT ADD TO TOTALS SHOWN BECAUSE OF INDEPENDENT ROUNDING
1 INCLUDES 271,000 BARRELS OF CRUDE OIL FOR DIRECT BURNING AS FUEL.
4-8
-------
various fuel oils are frequently blended, sometimes by the dealers and
sometimes by the industrial establishment consuming the fuel, in order
to comply with the sulfur in fuel regulations for the particular region.
Thus, a fuel oil dealer in Massachusetts may sell residual oil with a
sulfur content of 1% as imported to customers outside the 13 cities
and towns around Boston, but may blend this residual oil with fuel oil
of a lower sulfur content for his Boston customers in order to comply
with the sulfur restrictions for that area.
4. Oil Availability by Sulfur Levels, Bureau of Mines, 1971
This report essentially surveys the data discussed under 2.
and 3. in great detail for earlier years.
5. Import Supplement to Oil Availability by Sulfur Levels^
Bureau of Mines, 1972
This report supplements the discussion in the 1971 report.
In addition data from state and local air pollution agencies
indicate that most do not collect sulfur content data. As estimated in
Figure 4-1, fewer than half of the agencies collects sulfur content infor-
mation and of those that do, only about 20% compile the data for reference
purposes. Most sulfur content information was obtained from samples taken
at industrial plants and bulk oil terminals or from fuel oil dealers. The
forms used to record such data varied widely, from permit application forms
to forms designed by the air pollution agencies. The majority of agencies
4-9
-------
NC SJLFUR CCMTEMT
INFCRMATIOH COLLECTED
Figure 4-1. Sunmary of Availability of Sulfur Content Data
from Local and State Air Pollution Agencies
4-10
-------
collecting data on sulfur contents of distillate and residual oils take
fewer than 5 samples per year.
A listing of agencies v/hich may have summaries on file of
sulfur content data is shown in Table 4-7.
In general, it is recommended that the data of the Bureau of
Mines publications [1,2] be used to arrive at average sulfur contents for
area source emission calculations. These average sulfur contents should
then be checked against the sulfur in fuel regulations in the various areas.
Where the data are available (see Table 4-7), it is recommended that sulfur
content information available from local agencies be considered in addition
to the above sources.
B. SEASONAL FLUCTUATIONS
Residential and commercial use of fuels is highly dependent on
degree-days, since most of the fuel is consumed for spaceheating purposes.
Degree-days have been used here to estimate the percentages of fuel oil
burned by residential and commercial users in each quarter [3]. Estimates
of the seasonal fluctuations in residential and commercial fuel use are
shown in Tables 4-8 and 4-9 for selected counties for 1970 and 1971.
To get a clear picture of the seasonal fluctuations of industrial
fuel oil use, a summary was made of seasonal information available for
point sources on the NEDS, files. The results of this summary are shown in
Table 4-10,by industry groups. It is suggested that such a summary be
made after each annual updating of the NEDS point source file.
4-11
-------
TABLE 4-7
AGENCIES COMPILING SULFUR CONTENT DATA
1. Dept. of Health, Town Hall Annex, Greenwich, Ct. 06830
2. D.A.P.C., Prince George's County Health Dept., Cheverly, Md. 20785
3. Dept. of Air Quality Control, 4525 Indianapolis Blvd., E. Chicago, In.
4. State of Maryland Dept. of Health & Mental Hygiene, Environmental
Health Administration, 610 N. Howard St., Baltimore, Md. 21201
5. County of Sacramento Health Agency, Environmental Health Services,
6730 Folsom Blvd., Sacramento, Ca. 95819
6. Anne Arundel County Dept. of Health, Air Quality Control Sect.,
3 Broad Creek Pkwy., Annapolis, Md. 21401
7. Manatee County Health Dept., 202 Sixth Ave., East, Bradenton, Fl. 33506
8. Puget Sound Air Pollution Control Agency, 410 W. Harrison St., Seattle,
Wa. 98119
9. Environmental Improvement Agency, Pera Building College & W. Manhattan
St., Santa Fe, N.M. 87501
10. Utah State Division of Health, 44 Medical Drive, Salt Lake City, Ut. 84113
4-12
-------
TABLE 4-8
SEASONAL FLUCTUATIONS OF FUEL USE FOR RESIDEflTIAL AfiO CO.'uXERCIAL
SPACEhEATIfiG - 1970 (ERASED Oil DEGREE-DAYS)
County
P-lkr-p, fi.H.
rr~r,.-J "ir,, ?-'.a.
l.'orc^tor, f-ia.
Eal ti.-.oro, Md.
Pel- Beach, FT.
c-t- I r ., -,* r Mn
O '^ . t_ ^ i^ i -> 3 1 I W
Mirnehaha , S.D.
King, l,V..
Calves ton, Tx.
Jefferson, AT .
Boulder, Co.
Los Angel 8$, Ca .
Ss.n Diego, Ca.
Jan. -March
51%
51?
5155
59?;
7b%
57%
51%
37%
72%
59%
45;:
432
45%
April -June
12%
13%
13%
9%
0%
7%
]0.i5
2C;^
2.1
4%
15^
17%
19%
July-Sept.
3%
3%
q<>'
V//C
0%
0%
c\c'
U/3
3%
6%
0%
0%
3^
OJi
OK
Oct. -Dec.
34%
33%
33%
32%
24%
36%
36%
37%
26%
37%
37%
34%
35%
4-13
-------
*TABLE 4-9
SEASONAL FLUCTUATIONS OF FUEL USE FOR RESIDENTIAL AND COMMERCIAL
SPACEHEATIKG - 1971 (BASED ON DECREE-DAYS)
County
Bel knap, N.H.
Frankl in, Ma.
Worcester, Ma.
Baltimore, Md.
Palm Beach, Fl .
St. Louis, Mo.
Minnehaha, S.D.
King, Wa.
Galveston, Tx.
Jefferson, Al .
Boulder, Co.
Los Angeles, Ca.
San Diego, Ca.
Jan. -March
50%
52%
52%
58%
94%
59%
51%
40%
71%
63%
41%
42%
49%
April -June
14%
15%
15%
11%
6%
9%
10%
20%
3%
9%
16%
15%
17%
July-Sept.
Oo/
O/9
2%
2%
1%
0%
1%
3%
6%
0%
0%
5%
0%
0%
Oct. -Dec.
33%
31%
31%
30%
0%
30%
36%
34%
26%
28%
37%
43%
34%
4-14
-------
TABLE 4-10
SUMMARY OF SEASONAL FLUCTUATIONS OF RESIDUAL OIL
USED BY POINT SOURCES
(Percentages)
SIC
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Industry Group
Ordnance
Food
Tobacco
Textiles
Apparel
Lumber & Wood
Furniture
Paper
Printing
Chemicals
Petroleum
Rubber & Plastics
Leather
Stone, Clay, Glass
Primary Metals
Fab. Metals
Machinery exc. electr.
Electr. Equip.
Transportation
Instruments
Miscellaneous
Winter
36
29
25
28
37
25
31
26
38
27
25
30
32
25
26
28
29
35
40
27
27
Spring
22
29
25
25
25
25
25
25
23
25
25
25
25
25
28
25
25
24
16
25
25
Summer
14
21
25
22
13
25
19
24
13
23
25
20
18
24
24
22
20
17
11
23
14
Fall
28
21
25
25
25
25
25
25
26
25
25
25
25
25
22
25
25
24
33
25
24
4-15
-------
The seasonal fluctuations of gasoline and diesel consumption are
very slight when lumped together by quarters for the various states.
Whether the quarters are taken to be January through March, April through
June, etc., or December through February, March through May, etc., seems
to make little difference. Month by month fluctuations are slightly more
significant and are available from the Federal Highway Administration by
state [4]. July and August usually show the highest motor fuel usage in
most states.
4-16
-------
REFERENCES - SECTION IV
1. Burner Fuel Oils, Mineral Industry Surveys, Bureau of Mines, Bartles-
ville, Oklahoma.
2. Fuel Oils by Sulfur Content, Mineral Industry Surveys, Bureau of
Mines, Washington, D.C.
3. Climatological Data, U.S. Dept. of Commerce, National Oceanographic
and Atmospheric Administration, Asheville, North Carolina.
4. Highway Statistics, Federal Highway Administration, Washington, D.C.
4-17
-------
-------
V. COMPUTER PROCESSING
A. INTRODUCTION
In order to facilitate the annual calculations required to obtain
fuel oil consumption and seasonal fluctuations figures for the more than
3,000 counties, the methods described in the previous chapters were pro-
grammed in Fortran IV for use on an IBM System 360.
Card input forms for three separate card files were designed to
be used in these programs. These files contain:
- data required to allocate fuel oil consumption by county;
- data required to determine seasonal flucuations;
- point source employment data.
The computer processing flow is indicated in Figure 5-1 and sum-
marized below.
1. The input card-files are transferred onto tape by means
means of an IBM program which generates card-image tapes
for each file.
2. Each tape goes through a sort/merge procedure which outputs
a magnetic tape with card-image data sorted in the right
sequence.
3. A tape containing all NEDS point source data for point sources
using oil and the point source employment data file are used
5-1
-------
c o
.- 1/1
O
» rj -» * /
'<- '»- i'» v> c= I
! o o r > -.3 o /
-O «/i
-F-
~&
L: o
o ra.
Q- ii
5-2
-------
as input to the WALDEN PREPROCESSOR PROGRAM. This program
prepares an output tape with all the point source data which
are relevant for the following County Allocation Program.
4. The sorted fuel oil allocation data file and the pre-processed
tape from the previous step are input to the WALDEN COUNTY FUEL
OIL ALLOCATION PROGRAM. This program allocates fuel oil,
gasoline and diesel by user categories to the counties of the
United States. The results are printed and also punched out
in the NEDS area source format. In addition, a diagnostic
listing is provided, indicating areas that should be checked,
e.g., if the industrial use of distillate oil by point sources
is larger than the total industrial distillate oil use, the
distillate oil used by industrial area sources for that region
will automatically be set to zero and reported. This diagnostic
listing should be analyzed to ensure that the problem did not
arise due to coding or keypunching errors.
5. The seasonal fluctuations data file is input to the WALDEN
RESIDENTIAL/COMMERCIAL SEASONAL SUMMARY PROGRAM, which produces
a county by county breakdown of the percentage of residential
and commercial fuel oil consumed in each quarter.
6. A special card-image tape can be produced from the NEDS disk
files containing all point sources using fuel oil. This tape
is input to the WALDEN INDUSTRIAL SEASONAL SUMMARY PROGRAM,
which produces a statewide breakdown of the percentages of
fuel oil consumed in each quarter, by 2-digit SIC. A national
summary by 2-digit SIC is also provided.
5-3
-------
B. PROGRAM DESCRIPTIONS
Below a short description is given of each of the four programs
containing the methodologies described in the previous chapters.
1. The HALDEN PREPROCESSOR program
Input: - point source fuel oil tape, containing data on all
point sources which consume fuel oil
- point source employment tape, containing employment
data for all point sources which consume fuel oil.
This tape is in variable NEDS card-image layout (see
page 5-5).
Output: - preprocessed point source tape containing fuel oil and
employment data summarized by county and various com-
mercial and industrial subcategories.
Program function: - The WALDEN PREPROCESSOR Combines all the point source
information required for the main fuel oil allocation
program by county on a single tape. This tape contains
summarized county figures for manufacturing fuel oil
use and employment by 2-digit SIC group, and commercial
oil use and employment by the various subcategories,
i.e., hospitals, hotels, schools, colleges, and
laundries, as well as the total categories wholesale
trade; retail trade; finance, insurance, and real estate;
and services.
2. The WALDEN COUNTY FUEL OIL ALLOCATION program
Input: - preprocessed point source tape
- fuel oil allocation data tape: this is a sorted card-
image tape (see pages 5-10 to 5-14 for card layouts)
Output: - printed listing of fuel oil used by county
- diagnostic listing
- NEDS area source cards by county (see Figure 5-2 for
card layouts*)
*0nly the darkened fields in Figure 5-2 contain data in the punched out cards.
5-4
-------
I I
- r
5-5
-------
uoipy
ys
13
a 3 -
3 a"&
S
s
T 8
Ql"
* 1 s
;
£
S
«
II
I2
_i
Ji
1 £
*&
gi
c °
g
S *
** 3 5.
i
*
» j;
g3
5S
o -
K£
Sife
:»
0 JS
*?
Ms
« =
3~
5.1
|o
is
i a
NS
a"g
ss
m ^
M-
r "* "
K
^
J3
?
P
P
^
S
L3
3
10
S
W
3
S
s
§
s
a
u
to
-3-
ifl
I?
^
S
a
g
s
i
s
3
^
y.
s
&
s
s
s
s
s
3
<
R
n
S
R
S
M
K
fi
J5
U
ft
r*
^
2
t
S
*
«
^
5»
^
2
* " *
4/1
r «wnv 3
s
o *S;
1 ?s
5 i;
2 - -
*
r
1^5
°" * !
s ;
5 c «
OE ft £
e r*
S '
;
s
1 Is
ft vf '^ ;
!
J« j
gfe
I -5
ils
°p
_j
fuS
«5
* ^3j
*
<
t
1
«
»
1
t
!
i
*
i
i
r
>
'
;
»
r
3
it
'"
5
«
5
V
2
X
2
>
o
U-
-o
o
-------
Program function: - The WALDEN COUNTY FUEL OIL ALLOCATION program contains
all the methods described in Chapter III to allocate
by county the fuel oil consumed by residential, com-
mercial and industrial area sources, and the gasoline
and diesel consumed by light-duty and heavy-duty
vehicles. Since the program is intended to apply to
area sources only, point source fuel oil and employ-
ment data are subtracted from respectively the state
fuel oil totals and the state and county employment
totals. Figure 5-3 shows the approximate program
flow. A more detailed description of the program, with
clear indications concerning the various program steps
is available in the program documentation provided to
EPA-NADB.
3. The HALDEN RESIDENTIAL/COMMERCIAL SEASONAL SUMMARY program
Input: - seasonal fluctuations data tape: this is a sorted
card-image tape (see page 5-15 for card layout)
Output: - printed listing of the percentages of annual residential
and commercial fuel oil consumed during each quarter of
the year by county
Program function: - This program merely summarizes the monthly degree-day
data on the input by quarters for each county
4. The HALDEN INDUSTRIAL SEASONAL SUMMARY program
Input: - card-image NEDS point source tape (see Figure 5-4 for
card layout)
Output: - printed listing of the fuel oil consumed in each season
of the year by industrial point sources by county and
2-digit SIC.
Program function: - The program summarizes the point source fuel oil consump-
tion data by county and by season for each industry group
and provides a national summary by industry group at the
end.
5-7
-------
Read County
& State Cards
Calculate State *
Totals for
Fuel 911 Used }
By Stationary and j
Mobile Sources I
Read
Point
Source Tape
-1
Subtract Point Source!
Fuel Oil Use From ;
State Totals j
j
Read Point
Source Employment
Data
Subtract Point
Source Employment
Figures from County
and State Totals
Apply Methodology for
Countywide Fuel Oil
Allocation
Print and
Punch Results
Figure 5-3. Flow Chart of Counts-wide Fuel Oil Allocation Program
5-8
-------
'>'.-'s, [R!J g^>l fcs ^?
1". J ?! 5 »-! J5 ] 5L_^« -r.-
'* a £ is?" . "---: " - A
S_
O
a
o
u
QJ
u
3
O
in
o
d.
O
uj
S-
cn
iZ
5-9
-------
I I
§
«c
o
i !
PJ033J i
I j ' I I
' a
i
!
PJ003J
. !
t
I <
tl"
iizrzl
5-10
-------
o
to
pa
JO
-I
I V
0) t
in a
Pop
(.Ce
t: en
c: o .
-------
D
ex.
o
i
aieis -
ATA
KA
0
in
O
\n
.(0
I I
; r
UFACT
-\
5-12
-------
en o i- -r-
> o OJ c_:
«^ oc o- ID
>-
(
o
Q
s
v. v>
P-IOD3U
JO JBa^
5-13
-------
a
C£
S
3
a.
5
or
u
s
R
o
IX)
CM
0
to
CO
CM
O
CO
CM
0
to
CM
O
to
CM
O
to
CM
O
CO
CM
O
1
c
b
CJ
£2
',
ro"
ro '
i
i
i
i
-
--
~
|
o
£
g
i-
i;
ui
o
_t
£
i
*-»
iS
i o
00
0
1 ii
k
ri
LO
r*>
CJ
to
ro
o
Jo
o
o*»
_
O
«/>
DUg23J
* - A
a;
o
^
*-»
c
3
O
CJ
94*US
-
r
-
*
.
-
-
r
,.
'-
~
-.
,
.
-
--- _|
I
<
-
0
4O
Ul
_LJ I
5-14
-------
£2
o
5-15
-------
-------
VI. RECOMMENDATIONS
A. IMPROVEMENT OF DATA BASE
1. Bureau of Mines Fuel Oil Sales Data (Annual)
The basic source for fuel oil used by stationary sources is
the Bureau of Mines annual report showing fuel oil sales by state [1], This
report is published annually in October, with data of the preceeding year.
Conversation with Mr. James M. Diehl of the Division of Fossil Fuels, which
prepares the report, revealed that the figures are based on the return of
a questionnaire (see pages 6-2 and 6-3) which is sent out to oil refiners
and fuel oil dealers throughout the country. Approximately 5,000 question-
naires are completed and returned to the Division of Fossil Fuels for further
processing. It is estimated that these returns represent about 75% of all
distillate oil shipments, and 65% of all residual oil shipments. It is noted
that dealers processing less than 10,000 barrels during the year are not
required to fill out the questionnaire.
Mr. Diehl informed us that the returns of the questionnaires
were high, although response was not obligatory. Questionnaires are not
sent out to firms which, in the past, have refused to complete them.
Statistical completion of the sample consists mostly of a simole prorating
procedure. Based on the above, it is felt:
(1) that the sample is unnecessarily biased towards larger
firms,
(2) that the sample may exclude valuable information due
to the fact that questionnaires are not mailed out to all refiners and
dealers,
6-1
-------
G-1337-A
UK'ITrD STATCS
DEPARTMENT Or THE INTERIOR
BURrAU OF WINES
WASHING1ON, D.C. 20240
FUEL OIL AND KEROSINE SHIPMENTS
AND 5NVENTOPJES
Bmtjret Itumtti No 42 - R(y;.lo.
Approval iiplre.t M'srrh 1OT2.
INMIVinUAI. COMPANY
DATA-CONFJUKNTIAL
Th« daUt furn'fhfl In thli report *"!U
rm treat**) In CoiitMente hy the !><-j>i<(U
m*nl o( tiie Intih-ior, eicipt tfiHt they
n>»y be d!»cV>»e«l to detenu* agtncleo-
f F!ra«? correct if name or address has changed.)
SEE INSTRUCTIONS AND DEFINITIONS ON REVERSE SIDE
1. SHIPMENTS OF FUKL OIL AND KEROSINE DURING THE YEAR BY STATES OF DESTINATION AND BY USES.
(Insert names of States In column headings)
Type find usa
Code
2911-
Barrels
Barrels
Barrels
A. Distillate-type fuel oils:
1. Direct shipments to consumers by your company for
a. Heating:
(1) Grade fl for automatic burners
(2) Grade,*! for oil other heating ...41(L
(3) Grade #2 42°
(4) *Grade#4( % heavy % light)..
b. Industri.il use 40l
402
c. Oil-company use
d. Railroad use ..40?..
404
e. Vessel bunkering.
f. Military use
E. All other uses
409
2. Shipments to dealers and resellers
3. Total (Sum of Al and A2) 4U
B. Residual-type fur:! oils: ,
1. Direct shipments to consumers by your company for
a. Heating. ,
(1) "Grade *5 ( L% heavy, light)
(2) GraJe *G
501
b. Industrial usi'
502
c. Oil-company use .
KOI
d. Railroad use
e. Vessel bunkering
505
f. Military une
1OH
g. All other uees
2. Shipments to dealers and resellers
3 Tot«J(Sumof 1)1 and FJ2) ' '
C. Keroslnw
1 IMrect shipment* to consumers by your company for-
| 300
b. All other usf;fi (excluding Jet fuel). -2PJL
2. Shipment* to dealorn and resellers '. -
3. Total (Sum of Cl and C2) 3H
6-2
-------
II. I\VK.\TOIll I-:S OF FUKI, OILS AND KKHOSINK
,.,..,_....>
Inventories by type
A. lU'^lMti-nj; tif year:
1. DictlllnU^tvpe fu^l oil (including dlcsel)
'2. |{c»UImil-t\pf fuel oil (including heavy diesel)
3 KeroMne (excluding jet fuel)
li. Knit of \o:ir
1 Pistil!dte-ty$H' fue) oil (including diese!)
^ Kc^idu.il type fuel oil (including heavy diesel)
Code
2911
477
577
377
483
58H
388
HiirruN
B.irrtln
RurrcU
if this company changed ownership during the year, ple:-.~e report name- and address of present owner, and date sold:
(Maine)
(AiUrfss)
(Date sold)
S.ijnnture
O.T.c, .' r-j*i--.
Dale of report
INSTRUCTIONS FOR COMPLETING THIS FORM
Report all Ggures in barrels of 42 gallons.
Companies which shipped less than 10,000 bar-
rels (420,000 gallons) of fuel oil and kerosine during
(he year arc not requested to complete this form.
Form should be so noted and icturned.
Tliis report of the distribution of your shipments
should cover all light and heavy fuel oils and kero-
sine shipped or sold during the year. Include all
domestic shipments of these products, whether pur-
chased or of your own manufacture. Report ship-
ments by state of destination and inventories by
state of location. Do not include shipments to
Puerto Rico, as such information is compiled by the
U. S. Department of Commerce. A consolidated re-
port covering all your company's fuel-oil and kero-
sine shipments and inventories will be acceptable.
Please return one copy of the completed form by
April 15. Before submitting the report, please com-
pare it with the previous year's report and verify
any inconsistencies such as the addition or elimina
tion of states or products and any sizable differ-
ences in quantity data between the two years.
DEFINITIONS
DISTILLATE-TYPE FUEL OIL. Include ASTM
-^ .ides 1, '2, and 4 and distillate-type diesel fuol oil.
RESIDUAL-TYPE FUEL OIL.-Include ASTM
grades 5 and 6, heavy diesel, Navy special and
Bunker C oils used for 'generation of heat and/or
power. Include acid sludge and pitch used for re-
fin nry fuel.
KECiOSfNTC. Include petroleum distillates suit-
able for use as an illuminant when burned in a wick
lamp and kerosine sold for range oil.
HEATING. Report all fuel oils used for heating
purposes. (Indicate the percentage of heavy and
light oils contained in grades 4 and 5 fuel oils.)
INDUSTRIAL USE (excluding heating and oil-
c.-mpany uses). Report under this item all fuel
oil shipped to mines, smelters, and plants engaged
in producing manufactured products.
OIL-COMPANY USE. Report all fuol oil, crude
oil, or acid sludge used aa fuel nt your refineries, by
your pJp«linuH, or in your field operations. Ship-
ments to other oil companion for field use should be
included but exclude shipments'for use vs refinery
chaiging stocks. Oil used to heat buildings, and
operation of marine equipment should be reported
under their proper categories.
RAILROAD USE. Include all fuel oil shipped to
railroads, except that used for heating buildings
operated by railroads which should be reported as
shipments for "Heating".
VESSEL BUNKERING. Report all fuel oil and
diesel oil shipped for ships bunkers and other ma-
rine purpose's including own-company use. Exclude
shipments to the Armed Forces.
MILITARY USE. Include all fuel oil shipped to
the Armed Forces, regardless of use.
ALL OTHER USE. Include on and off highway,
agriculture, utilities and any other use category
not covered above.
INVENTORIES. Report all beginning and end-
:r.g inventories held by your company for the year
ir.dicuted on form.
6-3
-------
(3) that the statistical extrapolation methods should be
refined,
(4) that it would be of interest to other government agencies
using these Bureau of Mines reports if the user category "residential"
were included in the report.
It is of extreme importance to this study that the actual accuracy of the
Bureau of Mines statewide data be determined and, if possible, that the
report reflect some of EPA's needs. If this seems to be too difficult a
task, in view of the interdepartmental cooperation required, it is recom-
mended that EPA study the possibility of developing a methodology to collect
statewide fuel oil consumption data by grade of fuel and user category on
an annual basis for its own use.
2. Bureau of Mines, Burner Fuel Oils Data (Annual)
This report is an annual mineral industry survey of the
Bureau of Mines, which shows sulfur content data for samples of fuels.
The sampled fuel oils are selected by their manufacturer as tyoical of
that year's production of that specific grade and brand.
At present, the sample size is large enough to allow the data
to be summarized by five geographic regions. It is recommended that the
sample size be significantly enlarged to permit summaries by smaller regions,
thus providing more accurate average sulfur contents to be used in emission
calculations for area sources.
6-4
-------
3. Census of Manufactures, Fuel and Electric Energy Consumed
The Census of Manufactures has published a special report
on fuel consumed by industry, showing fuel use by type of fuel by state
and two digit SIC. It may become an annual publication, in which case
it is likely to be an important source of data for this project. The
EPA should indicate its interest in seeing this information produced
annually.
4. Highway Administrati on Data
On a local level, it is recommended that the state Highway
Administration agencies be encouraged to collect vehicle mile data by
county.
On a national and local level, it is recommended that the
Federal Highway Administration attempt to publish yearly truck registra-
tions by weight categories common to all states. It would seem logical
to use the categories used by the Census of Transportation.
5. R.L. Polk Data
In the absence of truck registration data by consistent weight
categories published by the Federal Highway Administration, it is recommended
that the EPA subscribe to annual publications of R.L. Polk of Detroit,
Michigan, showing automotive registrations and truck registrations by weight
categories on a county by county basis.
6-5
-------
6. Point Source Data
It is recommended that point source employment figures be
routinely recorded on NEDS variable data forms, to facilitate the prepara-
tions required for this project.
B. IMPROVEMENT OF PRESENT STUDY
1. Choice of Fuels
In many cases, especially for commercial users, linear cor-
relations were developed in this study between employment and fuel con-
sumption. The determination of these relationships makes it possible to
derive fuel use from county-wide employment data, which are easily avail-
able [2]. It is not clear, however, how to determine what percentage of
the employees in each county are employed in establishments using fuel
oil.
For the purposes of this study, rough estimates were made to
arrive at the appropriate fuel oil figures. It is recommended that a study
be undertaken to make an inventory of actual regional fuel choices taken
by various types of establishments.
2. Linear Correlation Between Commercial Fuel Oil Use and
Socio-Economic Data
A preliminary attempt was made during this phase of the study
to arrive at linear relationships between fuel oil consumed and employment
for various commercial categories. It is felt, however, that it would be
6-6
-------
useful to study the commercial category in depth in a separate study,
which could result in fuel consumption figures by two digit STC for the
commercial categories in each state.
3. Rural/Urban Driving Patterns
A considerable amount of data was collected on vehicle miles
traveled per car in the counties of seven states. These data were correlated
with four different variables in a stepwise regression analysis. The results
were not significant enough to be finalized and included in this study.
Other independent variables should be collected for these counties in order
to find the proper relationships which will explain the variability on a
county level of average miles traveled per car.
It is recommended that this approach be studied further, in
order to arrive at better gasoline consumption data on a county level.
4. Excise Tax Data
It is recommended that the Internal Revenue Service be ap-
proached by the EPA to determine if the IRS might be able to cooperate
with the EPA in an effort to arrive &t gasoline sales data on as detailed
a regional level as possible, based on excise tax data. Some such data
are available at present, but it seems that the excise tax is not reported
by retail outlets.
6-7
-------
5 Truck Vehicle-mi 'le Data
It is recommended that the EPA study the results of the
alternate method analyzed in Appendix D to allocate motor fuel consumption
by heavy-duty vehicles to the various counties. If the time and cost re-
quired to implement this method can be justified it is suggested that the
EPA approach the Federal Highway Administration with the request to publish
the required truck vehicle-mile data on a regular basis.
6. Re-examination of Assumptions
Two basic assumptions should be re-examined within the next
five years.
(a) It was assumed here that the amount of diesel consumed
by light duty vehicles could be ignored. There may be a change in this
pattern in the future and the assumption should be re-examined.
(b) It was assumed here that residential units using oil in
structures of 50 units or more would use mostly residual oil as opposed
to distillate oil. Because of recent air pollution regulations, larger
size buildings are being converted to distillate oil. It is therefore
necessary to re-examine this assumption in the near future.
7. Re-examination of Regression Coefficients
Linear relationships were developed in this study based on
historical data provided by various gas companies', state highway agencies
and the NEDS point source file. In view of the change in consumption
6-8
-------
patterns brought about by the energy crisis confronting the country at
the time of this writing, it is recoiraended that the EPA re-examine the
coefficients of these linear correlations using 1973 data.
6-9
-------
REFERENCES - SECTION VI
1. Fuel Oil Sales, Annual, Bureau of Mines, Washington, D.C.
2. County Business Patterns, U.S. Dept. of Commerce, Washington, D.C.
6-10
-------
APPENDIX A 9
RESIDENTIAL FUEL USE
1. State Fuel Oil
Four different methods were attempted to arrive at the resi-
dential use of distillate oil for spaceheatimj by state:
a. Using EPA's suggested estimate of .18 gallons/unit/degree-
day (1) to determine fuel oil consumed for spaceheating requirements.
b. Using the formula:
# of oil burners x avg size (Btu/hr) x 8760 (hr/yr) x load
140,000 (Btu/gallon)
to determine fuel oil consumed for spaceheating requirements (2).
c. Using a stepwise regression analysis which included the
independent variables: degree-days, price of fuel, per capita income, and
average number of rooms per housing unit, to determine fuel oil consumed for
spaceheating requirements.
d. Using the formula:
# of oil burners x heat loss x degree-days x use factor
140,000 Btu/gallon x Design Range
to determine fuel oil consumed for spaceheating requirements, where the heat
loss was dependent on the average square feet per housing unit (3).
a. EPA METHOD
In order to get an estimate of the residential spaceheating
requirements for distillate oil, the following calculations were
performed. The number of housing units using oil for spaceheating (1) were
recorded for each state. To eliminate units located in large apartment
buildings, which may not use distillate oil, the percentage of units in
A-l
-------
structures containing 20 units or more was subtracted (1). Using average
degree-days for each state (2) and the factor of .18 gallons/unit/degree-
day (3), Wai den estimated the distillate oil consumed by these residential
units for spaceheating. Then we proceeded to calculate the distillate oil
consumed for non-spaceheating purposes by state. The number of housing
units using oil for hot water heating (1), which is the major use of non-
spaceheating fuel oil consumption, was multiplied by 250 gallons (3) for
each state. The sum of these estimated spaceheating and non-spaceheating
fuel oil consumption figures was then compared to the total #1 and #2
heating oil reported for each state by the Bureau of Mines. The percentages
attributable to residential use for each state are shown in Table A-l.
b. OIL BURNER METHOD
Using the formula,
# of oil burners x avg. size Btu/hr x 8760 (hr/yr) x load /*\
140,000 (Btu/gallon)
we obtained the results shown in Table A-2.
c. STATEWIDE REGRESSION ANALYSIS
It was decided to perform a stepwise regression analysis in
order to arrive at the relationship between fuel consumed for homeheating
and degree-days, per capita income, fuel price, and the average size of
a housing unit by state. Due to the fact that considerable data are avail-
able on gas usage in publications of the American Gas Association, the
analysis was performed for gas usage.
Walden used average therms for homeheating per homeheating
customer as the dependent variable (y). These were obtained by state
by multiplying the total residential gas use, by the percentage of gas
used for homeheating and dividing this by the total number of homeheating
customers. The independent variables for each state were obtained from
the following sources:
A-2
-------
i
c£
UJ
I
CO
c^
t
o
cr>
i
UJ
CO
)
)
« '
UJ
H-
1
1
* i
t
CO
I-H
Q
Lt_
o
,
o
£
s:
i_i
i
CO
UJ
CM
t->
. - O "d
O 1 C
N X 03
O r
i
Ol -
-l-% CM
rU ~-tt- CJ>
i C
. . , -O !-
C.J T- E 4->
CO CL)
r- i 1C
cT3 ^fc
i-
0
4-
0)
C
r
-P
CL
rtJ CO
i 1
r- O
i ^ -^
CO
1
Q
-o
CU CD
4-J t~
(O !-
£ 4-9
r- (C
CO
i LOCOCMCOLOOO'-OOi VO
CMt-^CMi ncMCDr^-Lnun^l-
CT> CM CT1 C^O UO CTl *^~ CM t LO CO CT»
i CO F CT* f~^ CM CTi CT^ "^ CX) CM r^
l£)CMi ^COCMi CMOOCM' >
^j ^_ , CO CO CM «3- CM
i
«* CM rO C\J CM CT> i i Gnu"). i
i LT> i OOlOi i
i CO CM CM CM
«*
CM
oooococnoO'^t-r^-.cor^Ln
CM ^O CO LO CO O^ CO r~~ O^ CM LO 1
i LO <*CMCTli COCTiOOl
. LO CM ' i
4->
rO 3
r- O
CO C O -r- (1)
(Q ftS fO i- "O 4-5 i- fO 13
ErtSCCOOrtSOfO "O '!-!-
TO Vj/ O C M i- CD 3 "r" CTl*'"
JD CO N fti -r~ O C^ fO i- t. (T3
/y r£ .^- ^^ , i - £3 ^_ (^_j O O ^
i r s-t-rcioocu ! ^t-^COLOi VO^t-LOOOOOLO
CNJi ^O CM O*»CNJi -^O^d"
OJ i i LD CM r
co LO r LO ro ro co i t^ vo ^* r--
^t" CTi *^" r~~ O"» 1^*- ^~ CXJ LO
^j~ r*^. o CNJ c\J r~
*» M
^ ^J"
-
o<=^"O'*oor>" roof^)COi -oococoLO
CM CD CD LO r~ CO CJ1 LO ^^ LO
i i CM i
\
CO
1 t
T '
4->
OJ
HS CO r3
CO >, E 13 3 C +J
r fO ^ fO C" r" fO O
OC COO-'- (OOOJf
Ocro (arscocjr rc-rciJ
-CT-T- to co-p-i c r>>cojz: c:
H3 r~* "O S C CT Z3 *r- i- CO CJ CI
"Oi CO(0 LO
CO CTl ^ CO
CTl O i O
CO CO O CM
LO r-~ C\J
CXJ CT> ^" CO
CO i
co =* o
(V^ £J"j QQ Q^
<=} CTl i
CNJ i
r
CL
Q.
r- -r- UJ
CO t- fO ^
CO 3 C CO
r- O (X3 fO
CO CO 4-> S^
CO CO CT o
1- !- O <1J
A-3
-------
. ".
~o
cu
3
£
y~
o
u
,__
i
CD i»o co r** r1** co o** **o ro co co
ro . cNjLOOrcnuncsjuor coojLnooc\jCOr^r-
LO OO CD O"* P*1^ i~~~ CO ^t* {& ^J" r LO LO OO ^)~ ^~ CO *:d~ CD
r^cricr>cvjcococQLnioc\]oorocor^.co lO CO CD CD OO t^-* *~ Lf>
C\l CT> ^O I*1** ^O *st" ^^** CT* O*» **O U") c\J ' r~" * ^1"
ir>r-^-CSjr CNJCO OJCO
* +
st- in n
ioocM«a-vor^» cnocT>Lncricr>cOLOCMr-=t-Ln
oo co in i^» t^> co *^ CM co r*«« co ro co ^o ^o r
«3ioo> «*t^ooco cy»Ocn«*^-i i cor^co
in o in CM co o ^* o\ ^ co CM ' co o
CM «d- i i CM r
(d ra
tt) c c
^ «r (y io "O *^~ tj
<- r 4J - C i +->
JT>>O oo cmoo
t^ Q^ ^) t NX nJ i * ^, \x QJ
f^- c/) *r~ _^f (Q rf) fQ ^> t/} rO fO CJ fO
E t- X t- C-J Q E « 'f
> to cc
"O r* »3 5^ >- .C _C f~ o (/> O c" _C G) i/l O *r-
3e3:35i-i-'r-i ojco^^cxcoils.
zzsszisss^oooci-QitoooH-l £D>>
CM o r- ool CM
co o oo I r~-
in r-~ . i--.
co ro en vr>
co
i i
*X) r CD CO
IT) i CM
r CM
CO LO «* i
co in r*^ r~
LO Crt CO r~
Ol LO
r«>
A3
r~
C
c: -r-
o en c
1 1 £_ ,^_
O>*»~ en
c: > c c
- o >-
>-C 4-> O E
I/) Ml/) O
fd ^
S 13 3 3
»
(^j
,
f~%
C
o
cr>
c
[^
10
ra
*^;
ft
IO
OJ
C
sE
Ij
*^
o
3
(O
(1)
4-
CO
-
'ro
^
c:
cz
to
(U
i/>
r
r-
O
r
0)
U-
^M^.
^_
'
s_
ai
c.
in
0
fO
r~
r-~
cn
S-
i
>
_QJ
L.
O
0-
>
in
JC
»->
XJ
o
4_>
ft}
U
r-
-o
C
r
>
JC
a>
^
-,-T
0
2
A
CD
cu
Q)
_c^
31
«
0
CD
r
X
col
4r
ii
a
*
*
,
4J
r9
o
T-
^1
c~
CT
r
to
J
A-4
-------
-a cu
o s-
.c- s-
|J ro
0> -Q
< .
_-j
0- O
UJ -c:
-a 4-
01 0
in TD
ZD CT) O'
-r- -M O) CD
o ra s: s-
UJ O> i-
f OJ .c +-> ro
e=C 4-> CD C -Q
_1 rO 0 CU
1 ^T< f~
1 i fO fc: *
i i r OL O. 3
| T LO T- O
00 4J =3 _E
> i IO S- CT-P
Q -r- O UJ
o 4-
O 0
2: cri
O
1 1
Q-
^. CO
*~~) ^?r
CM OO >-< S-
i z: h- o
<: o «=£ 4J
O LU O
LU re res
i 1 i , 1 1 1
J I LU u
-> 01
C> O ui r-~ O') co kD m tM i in 01
r^ ixi r--
co LO oo in >* m in CD cri cvi en
CsJ * * vj" c\J C\J
1^3 CO CSJ LO ^J" LO ^.O * ' LO CO r*"** CO
C\J C3 ^D tZ^ P^1* ^X5 ^«O C\J *^3~ CO CD LO
CO ^O LO CJ f -^ ^£5 C3 LO LO C\J CO **£)
OO VO C\l C\J WD CT> r CO LO CNJ O C\J
^d" r r^ co co p
I** p^x Q^ LO LO LO LO C\J CVi C^) CO O"l
oooooooooooo
LocMcnoorOi r o <£> ro en o
i r i i LOCT>OO OO
r r i r
CU l/l
S_ 4J X> ro
r- +J C 4-> -r-
JC 0) rO 3 >> c
E-MJCI-I4J i-S-r- S-E
a)in:orocucu>-ow5:i
illllcSlllfild
in in r-~ 01 in r-- j en CT -^J «;! '-t 10 <-j- oo CD
r - LO >.o r-- ur> cxi 01 iii cu r-«- i"-- in
O CM OO i «J- O O CU I.O LO 1,0 CM r
S^^oScoSc^co^SScSSScS
O"^ r**^ r | « f*^ CO C5^ ^D C\J CxJ lO CO r~"
cvi r***» r*"^ c\i co LO LO oo CNJ co LO r*^« co o^ c~3 LO
OOOCDOOOOOOOOC5OC5O
CM O ^" 'd'OOi LOi CMO^OO1^! CO
g-g££ggg§g!ggg°£
ra ra
ro E E
r- "o 'o
ro S ro fO E t/) *!-~ -p ' rO
r-T-OOrOrora rO-i-(/)O S- -^
r- .E _E OVi >i rOEOOJ Or3rb
Cn P -P +-> ^_ S f O*f*r O C rO o*) en i-
i- _CU O O 0) r T- J^ E r !- -r; O -r; rO
-------
- GO < LOCOCOliirOOCMr Oi CO f-- <*£> "zf < "> f^~> CT> CT>
-- co r-r ro CM n LT> n oo M co cr> i 01 vo oo r--. co n * "*o
o-^j-r-i i i en i i CM r""> UD i.n cr> «=r
v * *» **
CO CM r P
CM
en
o
CM o CM > i
CM
cooooi r-^r^or-^i
CO [-^ O 00 i CO i *£>!
r, n rt
i 10 r^. <
i LO to cr> vo r- un «3- UD r cM^cocor^oouD
CMCMCMi i CM' O < i CM CM CM CM CM CM i i CM CM '
ooooooooooooooooooooo
i oo co en
i
co i r--i CMOOCM i
- CM
CM
r^
en
JD
o
o
o
CU
re
I
o
-a
c
ra
o o
-^ -i^
rO ra
a o.
4J 4->
S- 3
o o
"
cu
CD
CO rO
co £
CU ra
a.
a_
r- I/)
to ra
i/> co
M E J3 CO ra
ra
E
o
-E
ra
O)
CU r
I <
O
"3 DTO
c c: ra
s: < _j o H-
E ra O r ra > __ .
o-a>,o4->cuiDi-
s:i-isorjs2rcc
O E
o o
I- +J
X ra CD
« OJ E E E
as: o -r- o
"3 N X: C7)
to CU
ra S_
3 O
ra
rO
o
tt)
o
i-
3
o
oo
A-6
-------
Average degree-days from the 1970 Statistical
Abstract, of the U.S.
Per capita income from the 1971 Statistical
Abstract of the U.S.
Price of residential gas from the 1971 Gas
Facts, American Gas Association
Average rooms/housing unit from the 1970 Census
of Housing, General Characteristics
In the stepwise regression, the variables were selected in
the following sequence: degree-days (x,), price (x,-,), per capita income
(x.J, average rooms/housing unit (x»). The final equation was:
y = 0.00011 x] - 0.34450 x£ + 0.00021 x3 + 0.22904 X4 - 0.93373
where y is expressed in thousand therms/homeheating customer
X-, is expressed in degree-days
x~ is expressed in dollars per thousand cubic feet (or million Btu)
x3 is expressed in dollars
x. is expressed in average number of rooms
2
The adjusted multiple correlation coefficient R equalled 0.672.
The F-statistic, F(4.45), was 25.377, which indicates a 0.995 confidence
level. The adjusted standard error of estimate was 0.205 and the range
of residuals was 0.893. The t-statistics for the coefficients showed a
0.9995 confidence level for the coefficients of x,, x2, and x3 and a 0.975
confidence level for the coefficient of x..
Leaving out the variable indicating average number of rooms per
2
housing unit we found the R reduced to 0.646 and the adjusted standard
error slightly higher, 0.211. The confidence level remained more or less
the same. Interestingly, the range of residuals was reduced to 0.860,
although the standard error increased. This reduction is not significant
enough to warrant deletion of this fourth variable.
The correlation matrix shown in Table A-3 showed no significant
intercorrelation of the independent variables.
A-7
-------
TABLE A-3
CORRELATION MATRIX
y
xi
X2
X3
X4
y
1.000
0.672
-0.193
0.362
0.214
xl
0.672
1.000
0.196
0.217
0.286
X,
-0.193
0.196
1.000
0.281
0.276
X3
0.362
0.217
0.281
1.000
-0.087
X4
0.214
0.286
0.276
-0.087
1.000
TABLE A-4
RESIDENTIAL DISTILLATE OIL CONSUMPTION BASED
ON MULTIPLE REGRESSION EQUATION
State Thou Barrels Consumed (1970)
Massachusetts 14,896
Maryland 7,048
Missouri 1,234
Washington 4,^786
Maine 3,951
Connecticut 10,475
A-8
-------
Using the resulting equation and tank wagon prices for #2 fuel
oil shown in the 1970 November issue of Fuel Oi 1 and Oil Heat for selected
cities and assuming that these prices are representative for the entire
state in which the cities are located, we obtain the fuel oil figures
shown in Table A-4.
d. METHOD BASED ON HEAT LOSS AND DESIGN RANGE
The Yankee Oilman, March 1973 Fuel Trades Fact Book shows the
following formula:
Fuel Units Consumed Annually = Heat Loss x Annual Degree Days, x Use Factor
Btu per Fuel Unit x Design Range
where the heat loss is the hourly amount of Btu's lost per square feet of
area multiplied by the total square feet; the use factors are fuel and use
dependent and listed in a table in the publication; the design range is de-
fined as the difference between inside temperature (70°F) and the design
outside temperature for the area.
Wai den obtained the state-by-state data needed to use this
formula from the following sources:
(1) Average square feet per housing unit
for selected states (needed for heat
loss calculations) - Independent Gas
Association of America, Comparison
of Seasonal Househeating^ Costs,
December 1972.
(2) Design range - "Systematic Study of
Air Pollution from Intermediate Size
Fossil Fuel Combustion Equipment,"
Walden Research Corp., Cambridge,
Mass.
(3) Degree days - 1970 Statistical Ab-
stract of the United States, U.S.
Dept. of Commerce, Washington, D.C.
Other factors were taken from the Yankee Oilman publication in which the
»
formula was shown. The estimated distillate oil used for spaceheating is
listed in Table A-5 for selected states.
A-9
-------
TABLE A-5
ESTIMATES OF DISTILLATE OIL CONSUMED FOR HOME HEATING
BASED ON A PER HOUSING UNIT REQUIREMENT
ACCORDING TO AVERAGE FT2
State
Pennsylvania
Indiana
Illinois
Michigan
Iowa
Missouri
Nebraska
Georgia
Florida
Kentucky
Tennessee
Alabama
Oklahoma
Idaho
Arizona
Washington
Distillate Oil
(thou barrels)
21,487
7,055
7,165
11,962
2,839
3,236
833
1,123
2,466
1,683
1,143
143
21
1,808
33
11,062
A-10
-------
Table A-6 shows a summary of the four methods.
TABLE A-6
SUMMARY OF ESTIMATION METHODS FOR RESIDENTIAL FUEL
OIL USED FOR SPACEHEATING
Based on:
Massachusetts
Maryland
Missouri
Washington
Maine
Connecticut
Florida
Alabama
EPA*"
25,678
9,853
2,494
9,533
8,700
15,984
1,927
128
to)
Equipment^ '
42,005
10,307
3,466
16,109
8,326
19,665
1,389
245
Multiple/o}
Regression^ '
14,896
7,048
1,234
4,786
3,951
10,475
n.a.
n.a.
(4\
Square Feetv '
n.a.
n.a.
3,236
11,062
n.a.
n.a.
2,466
143
(1) EPA Guide for Compiling a Comprehensive Emission Inventory
(2) Fuel Oil and Oil Heat; Intermediate Boiler Study, Wai den
(3) Walden Research Corporation
(4) Independent Gas Association of America; Yankee Oilman
A-11
-------
2. County Fuel Oil
Walden collected sufficient community-by-community data to perform
stepwise regression analyses of the type shown below:
Residential gas use = a degree days + b median income
X A
+ c average rooms per housing unit + d
X
The regressions were performed on a community-by-community basis for two
reasons: (1) the majority of gas companies do not individually service
enough counties to provide a significant sample for a regression analysis,
(2) it is quite common for a utility to service only a portion of a given
county, in which case that utility's sales in the county would not reflect
total consumption.
Walden contacted various gas companies across the country to obtain
community residential gas sales figures for the year 1970. Table A-7 shows
the companies which provided data for the regressions. Detailed degree-day
data were obtained from the Climatological Data Center in Asheville, North
Carolina (6).
TABLE A-7
GAS COMPANIES WHICH PROVIDED RESIDENTIAL GAS DATA
Utility
Area Serviced
Wisconsin Gas Co.
Northern Illinois Gas Co.
Boston Gas Co.
Pacific Gas and Electric Co,
Southern Union Gas Co.
East Ohio Gas Co.
Southern Wisconsin
Northern Illinois
Metropolitan Boston
Central California
Texas, New Mexico, Arizona
Metropolitan Cleveland
Median income figures for all communities with populations over 2500 were
obtained from the Census of Population (7). Average number of rooms per
dwelling unit was obtained in the 1970 Census of Housing (8) for all com-
munities over 2500.
A-12
-------
The regression results are summarized in Table A-8 and discussed
in detail below:
In Wisconsin, where 21 communities were analyzed, the first vari-
able selected in the stepwise regression analysis v/as average number of rooms
2
per dwelling unit (R - .581). The second was annual degree-days. When this
was entered into the regression, the multiple correlation coefficient in-
creased to .687. There was a wider range in the size of dwelling units than
there was in the number of degree-days for the municipalities considered.
In using degree-days as a variable, it is important to note that the degree-
day figures are taken from the nearest weather station which may be as much
as thirty miles away. Degree-days also do not account for a wind chill fac-
tor, which can be very significant in many cases.
When gas use for residential space heating only was used as the de-
pendent variable in these same 12 communities, the order of significance for
the independent variables was not changed, but the correlation coefficients
in each case were increased slightly.
Southern Union Gas Co. provided Walden with data for 22 communities
in Arizona, New Mexico, and Texas. The most significant variable in this
2
case was degree-days (R = .755). For this area, the degree-day variation
was very large, ranging from 1558 to 7076 degree-days. The second most
significant variable, average number of rooms, varied by only .7 rooms,
from smallest to largest.
The third analysis was performed for 23 communities in the Cleve-
land, Ohio area. Because these communities were all within thirty miles of
one another, the range of degree-days was almost negligible. There were
only three stations in the area from which degree-day readings could be
used, further reducing the significance of the degree-day data. Median in-
come was selected as the most significant independent variable.
California data for 87 communities were obtained. Median income
explained a major proportion of the variability in fuel use for this analy-
sis. The most striking characteristic about these communities was that the
degree-day figures did not exceed 3200, and, in most cases, they ranged
A-13
-------
co
i
UJ
CO
U~l
*J
cc
ro
^tfc
O)
[Q
i
s~
ra
^"
CM
-It:
OJ
J2
ro
*f
S-
ro
^>
=*fe
OJ
o
ra
1
i-
ro
-!->
a.
OJ
o
a)
"j2
i t
01
4_J
rc
^y
c^-
tn
t^.
.
*^j-
CM
CM
LO
CM
to
in
CM
CO
in
f^.
ro
CM
CO
to
E
O
0
r--
ro
CM
CO
1
1
1 o
to
3
ro
to
cn
*
C\J
CM
vj-
to
_
^^
CM
CM
O
^
CO
>^
rO
Q
.
Q
to
sl-
i
in
to'
E
0
O
r*"
CM
to
ro
CM
1
CJ
to
2
in
r^
r--
^j-
CM
^
CO
to
Ln
CM
1
CD
^'
Oi
^
o
0
c
t 1
^
cn
cn
ro
O
!,
1.1
>^
ra
Q
.
Q
^-^
CM
CM
CO
1-^
to
E
0
o
cn
to
cn
cn
CM
i
o
to
3.
LT)
0
CM
«;}-
CM
cn
r
VO
cn
in
«^j-
^j-
to
6
o
1
ro
CM
to
1
*
0
co
^
ro
Ln
ro
,
CM
ro
f^
ro
j~>.
CM
O
I.
to
>^
ra
Q
.
a
cn
i
CO
in
in
CO
E
O
O
in
^j-
r
.
CD
CM
1
*
0
to
3
CO
1
cn
,
CM
0
in
f~-
CM
CM
cn
CD
0
~-^
ai
E
o
o
cr
CM
O
O
=!
vj.
CO
>^
>
ro
Q
CM
CO
ro
,_!
to
£^
O
1
c
~D
CO
cn
-d-
*=f
r_
^
CT>
cn
to
ro
cn
r-^
CM
v__>
to
E
O
o
a:
,
r
O
>>
rO
a
r
o
CM
f^
'
C
o
1
'~>
o
CO
CD
CO
.
CO
1
CM
CO
CM
CO
"~
to
to
CO
cn
CD
CD
CLI
E
0
CJ
r^^
f-^
cn
cn
ro
cn
co
o
c~
0
o
cn
.
Ln
o
^
f^^
CM
°?
cn
to
to
in
Ln
o
CD
a>
o
u
to
CM
in
to
in
,
"*
n
'i-o
"
A-l
CM
0
"^
cn
^
,
r~-
«d-
cb
o
o
I.
I/)
>^
ra
Q
Q
t »
co
in
0
o
CJ
o
CJ
c
CM
,
LO
«^J-
r~~-
r-^
^j
~
o
4
to
CD
"~|
cn
oo
ro
to
CO
ro
i
r-
'
LI
g=
O
o
^
cn
r---
O
O
's
0")
^^
ra
c ^
.
Q
co
«^-
o
0
OJ
o
c
r^^
ro
ro
s^-
to
CO
1
^J
'o
CJ
^J-
CM
ro
.
cn
^
to
en
CO
CM
r
CD
to
S-
in
o
o
cu
o
o
o
to
ro
r\
to
«3-
^
t H
CM
CM
cn
txi
1
CM
to
CM
to
to
r^
co
«=}-
i
x^_^
CO
E
0
o
Qi
CM
to
o
o
ai
Q
O
CM
in
«^-
o
1
cn
co
~
tn
cn
0
.
co
t~~
CM
cn
O
r--
O
"*
O
CM
CO
co
cn
ro
r~
CM
I/)
E
0
o
to
i^O
r-.
i
^j.
"*"
to
rO
an
r^J-
o
.
LO
^i"
"
r
CM
CM
to
ro
o
in
fvt
r^
i
o
CO
ro
a
^
CM
LT>
co
,
to
CO
a.
^
0
S-
1
5
*^J-
^J~
Ln
cn
CM
«^-
r^
to
*~~*
to
CM
CO
CO
CO
^ . f'
CO
E
o
o
a:
ro
i
O
CO
ro
Q
^
CM
r-~
tn
cn
i
a.
~^
o
S-
CI3
1
^
CM
O
in
CO
ro
r
CO cn
«tf- CO
CM O
CM r-~
ro CM
^
CM CO
in to
^
cn
r~-
o
o
ro
^- -^
I/)
E
O
O
a;
P**- CO
iri oo
, , .
0 0
to to
ra ra
a a
Q a
i^^
<* CO
CO LD
CM
in r>«
to i
CO
-------
from 1500 to 2500 degree-days. This meant that a much higher proportion of
total gas use in these communities was devoted to non-space-heating applica-
tions than in any of the previous areas analyzed. Degree-days, in general,
do not explain the variability in residential gas use in warmer climates.
2
Mean income was the most significant variable in this case (R = .676).
The correlation coefficient between degree-days and gas use was only .022.
Data for 68 communities in Northern Illinois were obtained from the
Northern Illinois Gas Co. More than three quarters of these ccrrmunities
were in the immediate Chicago area. This limited the degree-day figures to
only five weather stations in the area with a total range of difference of
less than 500 degree-days. In compiling the degree-day data in this case,
it was exceedingly difficult to assign a certain weather station's degree-
day figure to a given community with any assurance of accuracy. Secondly,
the Chicago area is generally subjected to strong winds coming off Lake
Michigan during the normal heating season. This is no doubt a significant
factor in gas consumption for space heating purposes, which our degree-day
figures did not reflect. Median income was the most significant variable
in this case (R2 = .610).
The final group studied included 19 communities in the Boston area.
The most significant variable in this case was average rooms per dwelling
unit with a correlation coefficient of .320. In the Boston area, unlike
the other areas studied, gas is not the primary heating fuel. Furthermore,
the proportion of older dwelling units in the Boston area is probably much
higher than in the other areas studied. These older units probably have
spaceheating characteristics which differ significantly from those in
newer housing units, due to boiler conversions and consequent diminished
combustion efficiency.
A regression was performed on all six groups combined. This gave
a total of 231 communities. As might be expected, the degree-days repre-
sented the most significant variable because of the wide range of readings
available on a nationwide basis from below 1500 to over 8000 degree-days.
The multiple correlation coefficient produced at this first step was .510.
With the addition of the average rooms per dwelling unit, the correlation
A-15
-------
coefficient was markedly improved to .674. Median income was not introduced
into the regression because as the correlation matrix (Table A-9) shows,
there is a very high correlation between average rooms per dwelling unit and
median income (r = .806) indicating that median income is not an independent
variable.
TABLE A-9
CORRELATION MATRIX FOR SIX GROUPS COMBINED
Gas Use
Degree-Days
Median Income
Rooms Per D.U.
Gas Use
1.000
.717
.572
.637
Degree-Days
.717
1.000
.329
.361
Median Income
5.72
.329
1.000
.806
Rooms Per D.U.
.637
.361
.806
1.000
A regression was also performed on all the groups excluding the
Metropolitan Boston communities. This left a total of 212 communities in
the regression. The results, however, were not as good as those for all
six groups. The first-chosen variable, which again was degree-days, gave
a correlation coefficient of .527. When the next variable, average roon,'s
per dwelling unit, was entered into the regression, the multiple correla-
tion coefficient increased to only .666, lower than for all 231 communi-
ties, with both variables entered into the regression.
Using the formula:
residential Dist. Oil (gallons) = [(.01288 x degree-days +
30.41 x avg. rooms per housing unit - 79.54)7
.14] x # of housing units
Walden calculated residential distillate oil use for all counties in Maine
(see Table A-10) and compared the sum of these figures with the Maine state
total obtained by using the EPA method (see Table A-l, column C). The figures
were within 12% of each other. Similar calculations were performed for
Massachusetts, where the two state totals differed by less than 1%. These cal-
culations increased our confidence in the selected method to be used to cal-
culate residential distillate oil by county.
A-l 6
-------
TABLE A-10
RESIDENTIAL DISTILLATE OIL BY COUNTY - MAINE
(thou gallons)
County
Androscoggin
Aroostook
Cumberland
Franklin
Hancock
Kennebec
Knox
Lincoln
Oxford
Penobscot
Piscataquis
Sagadahoc
Somerset
Waldo
Washington
York
Total Maine
Total Maine (EPA Method)
Distillate Oil
29,203
30,851
64,903
9,365
13,853
34,706
12,255
8,285
15,946
44,396
6,788
9,189
16,591
8,565
11,851
38,910
355,657
398,622
A-17
-------
REFERENCES - APPENDIX A
1. 1970 Census of Housing - Detailed Characteristics, U.S. Dept of Commerce,
Washington, D.C.
2. Statistical Abstract of the United States, 1970, U.S. Dept. of Commerce,
Washington, D.C.
3. Guide for Compiling a Comprehensive Emission Inventory, Environmental
Protection Agency (June 1972).
4. Systematic Study of Air Pollution from Intermediate Size Fossil^ Fug!_
Combustion Equipment, Wai den Research Corp., Cambridge, Mass. (1970
5. Communication with Mr. Griffith, Department of Statistics, American Gas
Association, Arlington, Virginia.
6. Climatological Data, Monthly Summarized Station and Divisional Data,
U.S. Dept. of Commerce, National Oceanographic and Atmospheric
Administration, Ashville, North Carolina.
7. 1970 Census of Population - General Social and Economic Characteristics,
U.S. Dept. of Commerce, Washington, D.C.
8. 1970 Census of Housing - General Characteristics, U.S. Dept of Commerce,
Washington, D.C.
A-18
-------
APPENDIX B
REGRESSION ANALYSIS OF COMMERCIAL USE
OF OIL FOR VARIOUS SUBCATEGORIES
B-'l GENERAL
It was found that the commercial category consisted of such a great
variety of business enterprises and institutions, that it would be in-
accurate to use employment as the distributive factor to allocate total
commercial fuel oil by county. It was therefore decided to analyze the
relationship between fuel oil use and employment for several subcategories
in order to determine the fuel oil use for these categories in a direct
way, based on the number of employees in each subcategory in each county
(1).
Fuel oil consumption of individual companies and institutions was
extracted from the NEDS files and basically compared to employment data.
This initial analysis has indicated fairly good correlations, and has
reconfirmed the fact that the various subcategories use fuel oil in
very different ways. The independent variable in these single variable
regressions was employment, but the slope and intercept of the regres-
sion lines differed considerably.
It is recommended that this analysis be refined and pursued for
other categories as well. The initial results are discussed below.
B-2 HOSPITALS
The NEDS files provided Walden with a significant sample of hospitals
using fuel oil in Baltimore, New Hampshire and Massachusetts. Staff data
on hospitals were obtained from the American Hospital Association (2).
For the 14 hospitals in Baltimore, fuel oil use was correlated with
an R^ of .91 to employment. For the 15 hospitals in New Hampshire, the
O
R was .86, and for the 40 Massachusetts hospitals (excluding state hos-
f\
pitals), the R was .68. Data for all three states were combined, resulting
in a correlation coefficient of .81 and the regression line:
fuel oil use (thou. gallons) = .715 x employees (thou.) + 208.5
B-l
-------
B-3 SCHOOLS.
Fuel oil use was available from NL'DS for schools in Baltimore and
Massachusetts and was compared to employment data for the specific schools
(3, 4). The correlation coefficient, R2, was .44 for Baltimore and .58
for Massachusetts. Due to the fact that many of the Baltimore schools
used fuel oil as a secondary fuel, it was decided to use the Massachusetts
results based on a sample size of 17. The regression line was:
fuel oil use (thou. gallons) = 2.97 x employees (thou.) + 76.2
B-4 COLLEGES
Fuel oil use for colleges in Massachusetts, New Hampshire and Maine
was available from NEDS. Staff data for these colleges were obtained
from an HEW survey of employees in institutions of higher learning (5).
2
The Massachusetts data showed an R of .76 for 15 sample points.
The New Hampshire and Maine data were combined and showed an R2 of .65
for a total sample of 10 colleges. Finally, the data for all three states
were combined, producing an R2 of .67 and the regression line:
fuel oil use (thou. gallons) = .546 x employees (thou.) - 40.9
B-5 LAUNDRIE'S
Fuel oil use for laundries available for Baltimore and Boston was
correlated with employment (6). Originally, very poor correlations were
obtained. Walden analyzed the Baltimore data, and found that many
laundry store fronts were included in the computer listing of sources in
Baltimore. Walden proceeded to separate out all the dry cleaning plants,
SIC 7211, 7216 and 7217, and re-ran the regression analysis on dry cleaning
plants only. An R2 of .72 was obtained for a total of 8 sample points.
It is obvious that more data need to be collected for this subcategory.
In spite of this, it was decided to use the resulting regression line:
fuel oil (thou. gallons) = .355 x employment (thou.) + 8.4
B-2
-------
Walden was not able to obtain employment data for the various hotels
reported as point sources. Therefore, a roundabout method was developed
to derive fuel oil consumed by hotels from employment data.
Fuel oil use by hotels obtained from NEDS and from the public relations
offices of the Hilton Hotels was compared to the number of rooms for those
hotels (7), resulting in an R^ of .96 for a total of 18 sample points.
Then, the number of hotel rooms in each state was correlated with the
number of hotel employees in each state (8) resulting in an Fe of .98.
State by state rooms/employees ratios were developed (see Table B-l) and
the resulting model consists of the regression line:
fuel oil (thou. gallons) = 1.09 x rooms +41.5
where rooms are determined by dividing the employment in hotels in each
county by the appropriate state ratios.
Example: To demonstrate the proportion of fuel oil used by these sub-
categories and residential apartment buildings, Walden calculated the
fuel oil used in Massachusetts by the enterprises and institutions dis-
cussed above, using the statewide coal, distillate oil, residual oil and
gas distribution shown in Section III of this report. The results are
shown in Table B-2.
B-3
-------
TABLE B-l
ROOM/EMPLOYEE RATIOS FOR HOTELS BY STATE
State
AL
AK
AZ
AR
CA
CO
CN
DE
DC
FL
GA
HI
ID
IL
IN
10
KA
KY
LA
ME
MD
MA
Ratio
.28
.21
.29
.39
.27
.29
.37
.13
.58
.34
.33
.56
.21
.32
.30
.28
.23
.30
.38
.16
.29
.34
State
MI
MN
MS
MO
MT
NB
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
RI
SC
SD
TN
TX
Ratio State
.27 UT
.35 VT
.28 VI
.30 WA
.20 WV
.26 WI
.80 WY
.23
.27
.23
.31
.22
.21
.31
.22
.29
.36
.36
.28
.16
.28
.32
Ratio
.22
.26
.29
.26
.29
.31
.20
B-4
-------
TABLE B-2
1970 COMMERCIAL USE OF FUEL OIL IN MASSACHUSETTS
(Thou. Barrels)
Hospitals
Schools
Colleges
Laundries
Hotels
Apartments
Z 6 categories
Total Commercial
Distillate
449
299
221
35
274
1>802
3,080
14,219
Residual
663
441
326
51
404
809
2,694
19,482
B-5
-------
REFERENCES - APPENDIX B
1. County Business Patterns, U.S. Dept. of Commerce, Washington, D.C.
2. American Hospital Association Guide to the Health Care Field, Ameri-
can Hospital Association, Chicago, 111., 1970.
3. Annual Report of Faculty Racial Composition, Baltimore City Public
Schools, Office of Research Reports and Statistical Records, Center
for Planning, Research and Evaluation, Sept., 1972.
4. Obtained from files of the Department of Education of the Common-
wealth of Massachusetts.
5. Number and Characteristics of Employees in Institutions of Higher
Learning, 1966-1967, Dept. of Health, Education and Welfare, Wash-
ington, D.C.
6. Dun & Bradstreet computer print-out provided to Maiden by the EPA.
7. Hotel and Motel Red Book, American Hotel and Motel Association,
New York, N.Y.
8. Hotels, Motor Hotels and Motels, Census of Selected Service Industries,
1967 Census of Business, U.S. Dept. of Commerce, Washington, D.C.
B-6
-------
APPENDIX C
REGRESSION ANALYSIS OF URBAN VS RURAL DRIVING
PATTERNS ON A COUNTY-BY-COUNTY BASIS
Stepwise regression analyses were performed on county-by~county data
to find a relationship between miles traveled in each county per registered
car and the socio-economic factors and highway patterns for each county.
The independent variables collected for this purpose were: percentage of
population considered rural, miles of interstate highways, per capita in-
come, miles of state highways, and population density.
The percent of county population considered rural was obtained from
the 1970 Census of Population (1). The number of miles of interstate high-
ways in given counties was obtained from the Federal Highway Administration
(2). Per capita income by county was obtained from the Department of
Commerce (3). State highway mileage by county was obtained from the
California Highway Department for California only (4). Population density
by county was obtained from the 1970 Census of Population (1). Miles
traveled per registered car for each county was derived from total county
vehicle mile figures, provided by the highway departments of the states in-
volved and county car registration totals supplied by R. L. Polk of Detroit,
Michigan. The results of the regression analyses are summarized in Table C-l
below.
The results were very discouraging. Not only were the multiple corre-
lation coefficients low, but the intercepts of the regression lines and the
standard errors were extremely large. In all states, there are some counties
with very extreme patterns. When we eliminated the data for the six counties
in California, which showed over 30,000 miles traveled a year per vehicle, a
o
greatly increased R (from .297 to .440) and a sharply reduced standard
error (from 9423 to 2377) were observed. Similar changes are observed when
some of the extreme data are eliminated in other states.
C-l
-------
"~J~
o
LU
_J
CO
=c
r
z.
o
1 1
h-
V 4
1
UJ
o:
LU
[-1-
22:
o 2
LU
LU I
ti C£
S ex
OC
LU O
1 'Z.
LU v- <
O >
O 5
h- Q
CO ^
t 1 «c£
CO CO
>- C£
t \
<£
z: co
^C ^"
z. _J
» * rv*
co rs
co o:
LU
o: z:
CD ii
ill
DC:
u.
o
CO
i
ra
CO
LU
1
S- +->
CU Q.
4-3 QJ
C 0
I i
(O 4->
> rO
4->
"O to
C 1
CM *->
~a
i. CU
rO 4->
> O
QJ
-a i
E QJ
CM CO
TD
S- QJ
ro -t->
> O
QJ
-a i
C QJ
CM CO
»
^-
ro 4-3
:=> ro
4J
»-> in
to I
r 4-^
S- .
ro 14-
> M-
QJ
4-> O
O
at
4-* *~~
CO
CM
^
fx»
CTl
CM
^-^
00
in
«N '
1)
t,
f
rO
O
10
«^J"
LO
in
UD
r*x.
en
CM
VO
«^-
^«
10
OJ
E
.
4J
to
QJ
^_
QJ
-l->
C
CO
r^
CO
^}"
<>o
CO
r
r
ro
S-
3
^_
S-5
o^
IO
r
A
r*^
CTl
en
r-^
CO
CO
f, ^
en
CO
^_^«
.
^:
in
rO
3:
CM
CM
i£>
f^
O
O
3-
CO
oo
to
QJ
'g
.
4_j
to
QJ
S-
cu
+J
c:
o
CM
^^
CM
r
r
t
ro
S-
3
i-
6«
in
CD
n
«5^-
Ln
o
CO
en
r-.
CO
^-^.
o
o
r
* *
to
rO
to
c~
ro
i^
O
cr>
t-.
I-,.
f^.
in
CO
UD
oo
f^N.
fM-
to
QJ
E
.
^_>
to
cu
£_
cu
4^
C
en
CO
.
UD
00
O
r
rd
S-
3
L
&«
O
^J-
oo
*
CO
00
«^
CO
m
r
3-
*~^*
en
en
ro
» i
,
CM
cr>
in
CM
,
m
oo
CO
en
in
cu
E
.
4J
to
CU
S-
QJ
-M
C
O
<^J-
un
en
CM
CM
f
rO
i.
3
j^
S^
CM
VO
r^*.
*i
en
^t
CM
O
o
* "
ro
i
a>
o
cu
CD
en
*
j
~
en
o
.
f~~
.
t-^.
CM
in
cu
E
.
4_>
to
QJ
t.
QJ
4J
C
CTl
en
en
.
r-
to
in
r
ro
^_
3
i_
&-S
p>^
in
CM
CM
0
CM
r
00
^
s~^
IO
r
% '
QJ
r
ro
2:
C3
CD
cn
r^^
) .
CD
,
^j-
cn
i^
ifi
QJ
E
4_>
to
CU
t-
QJ
4_>
C
CM
r
in
CO
CM
^~
r^
(0
^,
3
(?9
VO
co
in
M
^*
CM
«3
^-»
CD
O
CO
^^
m
i^.
^^^*
to
rd
1/1
C.
rO
^^
S-
«=*
^j-
«=)
,
1
1
1
1
-o
cu
4->
ra
f
E
f
cu
a.
cu
«-»
en
r^
CM
1
i
oo
i
>>
4«>
f
to
r~
O)
-a
a.
o
Q.
cn
*^-
10
VO
J^I
\
^-
0
*
^ ^
f*^
(£>
ro
-a
r
s-
o
r
U.
"O
QJ
S-
C1J
4J
in
r
cn
QJ
$-
^,
fC5
U
S
OJ
O-
l/>
(U
r"^
t
s
4-
O
TD
rO
CU
(/I
C
r
-o
i_
cu
-(->
t/1
cn
QJ
S-
j_
rO
CJ
S-
QJ
a.
c:
O
^
cn
w
r-
QJ
n
2
rO
>«
4->
C
QJ
c~
QJ
Q.
QJ
Q
*
C-2
-------
When a large percentage of the sample points have to be eliminated
in a regression analysis, the significance of the regression is sharply
reduced. In Georgia, for example, the miles traveled per car range from
3167 for Chattahoochee County (27% rural) to 56,976 miles for Charlton
County (100% rural). Furthermore, for counties classified by the census
as 100% rural we find very dissimilar miles travelled per car (see Table
C-2 below). This variability is probably due to factors relating to the
geographical location of the counties with respect to industrial regions;
factors which are not easily quantified.
TABLE C-2
DRIVING PATTERNS FOR SELECTED 100% RURAL COUNTIES IN GEORGIA
County
Atkinson
Baker
Brant! ey
Calhoun
Charlton
Clay
Miles/Car
37,229
27,921
48,170
23,579
56,976
20,701
Interstate Mi les
0
0
0
0
0
0
Per Cap.
Income
1,828
1,696
1,820
1,875
1,988
1,745
In short, the above analyses showed too vague a correlation between
driving patterns and rural/urban factors to be implemented in this pro-
ject. It did show that the vehicle miles per registered car fluctuate
tremendously from county to county, giving us a rough idea of the extent
of our estimation errors in countywide gasoline consumption when car
registration figures are used for distributive purposes instead of the
preferred vehicle miles figures. It is reassuring to note that many states
have now started collecting vehicle-miles data on a county level. It would
C-3
-------
be of interest to the EPA to indicate the usefulness of these data to
the proper state highway department officials, in order to avoid having
to use car registrations to distribute statewide gasoline use over the
various counties.
To compensate for these errors to some extent, Walden summarized
the average miles driven per vehicle by four categories: 25% rural or
less, 26 to 50% rural, 51 to 75% rural and more than 75% rural, leaving
out the extreme values that varied by more than 50% from the mean of the
total sample in each state. The results are shown below in Table C-3.
TABLE C-3
AVERAGE MILES PER VEHICLE BY % RURAL CATEGORIES
State
California
Washington
Kansas
Iowa
Georgia
Maine
Arkansas
Seven State
Indexes
Entire
Sample
15,152
18,722
16,155
15,643
23,140
15,261
18,652
Avg.
Modified
Sample
13,047
17,092
15,179
14,604
20,262
15,261
17,932
16,709
100
_<25%
10,729
11,940
10,706
11,468
13,185
11,210
11,039
11,217
67
26-50%
12,202
15,793
13,622
12,141
17,680
15,396
15,908
14,572
87
51-75%
14,803
15,778
15,306
15,501
19,687
14,876
18,144
17,238
103
76-100%
16,920
21 ,855
17,225
15,862
22,461
15,995
19,458
19,261
115
The method for countywide allocation will be as follows:
(1) If vehicle miles are available by county, use vehicle miles for
distributive purposes on statewide gasoline for LDV figures.
C-4
-------
(2) Otherwise multiply the LDV registrations in each county by the
appropriate miles/vehicle index according to the percentage of the popu-
lation considered rural in that county.
(3) Summarize these results for all counties to obtain a state total
(4) Use the modified county registrations, thus derived, to distri-
bute the statewide gasoline for LDV figures.
C-5
-------
REFERENCES - APPENDIX C
1970 Census of Population, Characteristics of the Population, Number
of Inhabitants, U.S. Dept. of Commerce, Washington, D.C.
Computer listing obtained from the Interstate Reports Branch of the
Federal Highway Administration, Washington, D.C.
Computer listing obtained from the Regional Economics Division of the
Department of Commerce, Washington, D.C.
Historical State Highway, County Road and City Street Statistics,
1957-1970, State of California, Department of Public Works, Division
of Highways, December 1971.
C-6
-------
APPENDIX D
ANALYSIS OF AN ALTERNATE METHOD TO ALLOCATE MOTOR FUEL USED BY
HEAVY-DUTY VEHICLES ON A COUNTY-BY-COUNTY BASIS
An alternate method to allocate gasoline and diesel used by heavy-
duty vehicles to counties was analyzed. It must be emphasized that this
method was not incorporated into the computer program and that data were
not collected for this method for 1972.
The Federal Highway Administration provided state-by-state totals
of vehicle miles traveled by trucks on interstate urban and rural highway
systems. It also provided estimates of the percentage of all trucks on
interstate highways which are diesel, distinguishing between urban and
rural interstate systems. By applying these respective percentages to
the urban and rural interstate highway truck vehicle mile totals for each
state, we derived estimates for total vehicle miles driven by diesel trucks
on both urban and rural interstate highways in each state.
Using county level interstate urban and rural mileage totals, also
provided by the F.H.W.A., the state vehicle-mile totals can be allocated
to counties. Dividing the county diesel interstate vehicle mile totals by
5.1 MPG (1), we arrive at an estimate of total gallons of diesel fuel con-
sumed by trucks on interstate highways on a county basis. This would
be in the form of a rural and an urban subtotal.
A similar procedure can be employed to estimate gasoline consumed
by heavy-duty vehicles on interstate highways. Using this method, it is
first assumed that all heavy-duty vehicles not using diesel, use gasoline.
D-l
-------
This is justified, since less than .1% of the truck traffic on interstates
is propane or LPG. It is therefore assumed that total interstate highway
truck vehicle miles minus the diesel vehicle mile total will be the gasoline
truck vehicle mile total. This statewide gasoline vehicle mile figure is
then allocated to the various truck weight classes by using truck registra-
tions by weight class (2), adjusted to reflect the differing average miles
driven per year by trucks of different weight categories (3). At this
point, the vehicle mile total for trucks weighing less than 6,000 Ib (LDV)
is eliminated from further calculations. The weight class vehicle mile
subtotals can now be divided by their respective mile per gallon estimates
which have been derived by Walden previously (see page 3-15). The resulting
t
weight subtotals of gasoline consumption can now be combined and allocated
to counties according to interstate mileage in each county, as in the
diesel method.
The above methods provide gasoline and diesel totals by county con-
sumed by heavy-duty vehicles on interstate highways. It is necessary to
allocate the remaining motor fuel, which is not consumed on interstate
highways, to the various counties as well.
1. Diesel
a. Subtract the interstate diesel consumed in the U.S. from
the total diesel consumed on highways in the nation.
b. Allocate the resulting diesel figure to each state, by
using the state percentage of the original national total.
c. Allocate the state diesel figure obtained in Step b to
the various counties by means of total truck registrations.
D-2
-------
The total diesel consumed by heavy-duty vehicles in that county is the sum
of the diesel consumed on interstate highways and the results of Step c
above.
2. Gasoline
a. Subtract the heavy-duty gasoline use on interstate highways
from the total use of heavy-duty motor fuel in that state.
b. Subtract the state diesel consumption total (both inter-
state and non-interstate) from the state total resulting
from Step a. This gives the consumption of gasoline by HDV
on non-interstate roads in each state.
c. Allocate the state gasoline total obtained in Step b to
the various counties by means of total truck registrations.
The total gasoline consumed by heavy-duty vehicles in that county is the
sum of the gasoline consumed on interstate highways and the results of
Step c above. These modifications should eliminate the occurrence of
negative totals for county heavy-duty vehicle gasoline and diesel use,
which could occur in counties with high interstate mileage and small
truck registration totals. The results of these new methods applied to
the thirteen test counties for the year 1971 are shown in Table D-l,
and can be compared with the results using the existing methods for the
same year in Table D-2.
This method represents a significant improvement over the method
used so far by Walden to allocate gasoline and diesel use by county. It
is found to have the following drawbacks, however:
D-3
-------
a. It is a very time-consuming method with regard to the
required input preparation.
b. The required data are not presently available on a yearly
basis.
D-4
-------
1
Q
LjJ
, |
CO
«t
(
U:
Q
O
:r
r
LU
s:
LU
1
t i
5:
i ' t
o
fc-H
LU
>
LU
1
to
LU
) » *
"Z. CO
O
CD r
K-< ro
to CD
f
i O
cr> to
i "O
C.
Qi ro
O to
1 1 ^
O
1 t
ZD
to
LU
C£
2^
O
i i
1
to
p**
8
i
LU
U_
f
; ^
o
o
>>
i n c
4-> 1 i
o -i-1 o
I -c co
en ro
r- CD
_J
ro O>
4-> CO
o cu
5
OJ
r fd
O) 1 4->
CO C CO
QJ O *-
r- Z 0>
a 4->
C
t 1
4-»
i (O
aj +->
CO CO
ni SM
Q -S
C
1 1
4-> a>
3 C
r- Q-r-
rO 1 r
I \ ^^^ .p^
*-* ^) O
O > co
OJ CD
"T~
>) a; -t->
^ c fd
Q <- 4-»
1 i CO
*~ .-. *^
>> o >-
> co O)
rO ra 4->
OJ CD C
3= -
1
if
Jv O fl^
_l ji ^/f i -.*
9 (O
Q 41 4J
1 C! (f>
^
4J
C
3
O
O
CT» tO i «3" OJ
^f 00 ^ ««- LO
OJ OJ 00 ID OJ
00 r^. CTt to tO
OJ O LO r
»
oo
LO gj ^ g j-
o> o oj «3- sf
r OJ CT\ f *£>
r 00 OJ
vo «* r-~ <^- to
LO r^ oo c> CD
V£) OJ OJ «3" r
LO ^ *3* i OO
tO r
O"l tj t^1^ t!I^> U^
^^ ^^ O^ C^ to
OJ OO tT> CO
vo o «* co
OJ r
r^ o co o r
LO OO CO CO r~~
tn to to LO LO
OJ OO LO i
r
OO r"" LO CJ f^
CO ' r"^ LO r^ CO
f~~> tO CO r"~*
OO LO i i
00 OJ
«3- CT> 00 O r
r*^ oJ oj LO oj
CO t^. O CO LO
tO O LO LO ^*
OJ LO OO i
to
0} f"*
C r- 0 0
co cr> fu i- tl)
LO i LOOl
I^-LO^OI^-OOLOO
oooj«^-r tor cnr--
O CSJ OJ <^~ r LO to OJ
LO OJ i *
C\J *^" I**1* CO ^f ^*^ C\J VO
o^ csj ro LO o^ f~^ i^^ r^**
O^ LT> vO O^ *«1" f^"* CO
r CO OJ
CD ^J" VO CD ^" r LO Ovf
co r*^ ^j* ^o LO o^ r^* LO
o>v0ojcomcr»coo
^£> ^- i F cr»
LOCXJ^O^LOOOU}
^t' r*^ oj oj r^» co LO P
cr> LO ^^ r*^ LO o co r*^
oj ^3* oj r^. co LO en
i OO
OOOljOLOr 0>
i i ojr~-otOLocT)
OJOJi OO'eJ-COLOr
OO ^J- QO OO r*~ *^* LO LO
i 1 ( ^3*
oo f^^ t7^ oj co oo r^^ to
COtOi r OCOLOLO
r OlOCT>^CJ>«3-tO
i OJ «3" OJ ' LO
r
COOOr COCOOJ«J-OO
r^> <^* co LO CTI r^ cr> ^j*
Oi LOO-ltDOOOLO
r^ oj oo LO co «s- cri
i r OO
\
>> c
-M 3
r- O
0 CJ
-------
TABLL U-2
COUNTY CONSUMPTION OF MOTOR FUELS BY HEAVY AND LIGHT-DUTY VEHICLES
USING EXISTING METHODS* - 1971
County
Jefferson
Los Angeles
San Diego
Boulder
Palm Beach
Baltimore
Worcester
Franklin
St. Louis
Bel knap
Minnehaha
Galveston
King
Heavy-Duty Vehi
Gasoline Consumption
29,960
154,800
34,860
6,035
17,250
34,190
18,580
2,750
10,400
1 ,030
2,466
5,722
32,880
cle
Diesel
Consumption
15,790
171,300
39,800
4,240
8,674
19,490
310,800
26,470
5,194
987
2,765
503
24,040
Light-Duty Vehicle
Gasoline Consumption
181,400
2,731,000
499,000
55,120
143,900
467,900
11,860
1,883
87,070
19,500
25,230
71 ,080
469,100
*See page
D-6
-------
rrCHNICAL HI;KM I DATA
' rcc.d iHiiruamri'S on r/i.p revrrr:C /"/"'<' (
I (' i t'O.I i NO
. _£ PA^i507_3z7 4^0.2.1
-; 1 I TLE AND SULJTITLfc
Development of a Methodology to Allocate Liquid
Fossil Fuel Consumption by County
5. PERFORMING ORGANIZATION CODE
HhCII-'ll Nt f. ACOuSSIOONO.
DAI F
March 1974
7 AUTHORIS)
a. PERFORMING'ORGANIZATION REPORT NO.
Josette C. Goldish, Franklin D. Trowt, John R.
Fhrpnfpld, Khee M. Chna. Richard Stockdale
-"> PcRFORMING~OR'~.ANIZATION NAME AND ADDRESS
Wai den Research Division of Abcor Inc.
359 Allston St.
Cambridge, Massachusetts 02139
1O. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-1067
2. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, N. C. '27711
13. TYPE OF REPORT AND PERIOD COVERED
1974 .
14. SPONSORING AGENCY CODE
5. SUPPLEMENTARY NOTES
6. ABSTRACT
Methods were developed for the routine determination of distillate and
residual oil consumption by industrial, commercial, and residential consumers,
as well as for gasoline and diesel fuel consumed by light and heavy duty motor
vehicles. The resulting data are allocated to counties for input and storage
in the National Emissions Data System (NEDS) area source format. In addition,
seasonal fluctuations of fuel oil use by consumer category and geographic region,
and references for determining sulfur content of fuel oils on a county basis,
were analyzed. The report summarizes the methodologies that were developed
and describes the computer processing techniques for reporting the data.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS ' |c. COSATI Field/Croup
Fuel consumption
Fuel oil
Gasoline
Diesel fuel
NEDS
Area source
Seasonal Fluctuations
Sulfur content
Counties
g. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport)
Unci_a_s_s i f i e d
21. NO^ OF PAGES
140
20 SECURITY CLASS (This pa?
Unclassified
22. PRICE
EPA Form 2220-1 (3-73)
-------
------- |