EPA-420-S-75-900
"STUDY OF POTENTIAL FOR MOTOR VEHI CLE
FUEL ECONOMY IMPROVEMENT"
SAFETY IMPLICATIONS PANEL REPORT
REPORT NO. 2 OF SEVEN PANEL REPORTS
January 10, 1975
PREPARED BY
THE
U. S, DEPARTMENT OF TRANSPORTATION
AND THE
m i ENVIRONMENTAL PROTECTION AGENCY
-------
EPA-420-S-75-900
PREFACE
This Safety Implications Panel Report is Number Two (2)
of a group of seven (7) prepared by special panels of the
Task Force established under the joint chairmanship of DOT
and EPA to conduct a study of the practicability of a fuel
economy improvement standard of 20% for new motor vehicles
produced in the 1980 time frame. Each panel addressed a
major impact area and drew on a variety of sources in pre-
paring its report, including previous DOT and EPA research,
and industry and public comments.
Materials developed by the various study panels were
used in preparing the Report to Congress entitled "Poten-
tial for Motor Vehicle Fuel Economy Improvement", dated 24
October 1974 (second printing, 18 November 1974i Assumptions
and conclusions expressed in the panel reports, however, are
those of the respective panels and do not necessarily reflect
official positions or policies of the U.S. Department of
Transportation, the U.S. Environmental Protection Agency,
or the study Task Force.
The complete Panel Reports set consists of the following:
Report
No.
1
Policy Assessment Panel Report
Report
No.
2
Safety Implications Panel Report
Report
No.
3
Air Quality and Emissions Panel Report
Report
No.
4
Technology Panel Report
Report
No.
5
Economics Panel Report
Report
No.
6
Fuel Economy Test Procedures Panel Report
Report
No.
7
Truck and Bus Panel Report
i
-------
TABLE OF CONTENTS
Section Page
1.0 FUEL ECONOMY AS A FUNCTION OF WEIGHT,
PERFORMANCE, AND DRIVING SCHEDULE 1-1
2.0 TRAFFIC CONTROL FOR SAFETY AND FUEL ECONOMY.. 2-1
3.0 WEIGHT VERSUS SAFETY 3-1
3.1 WEIGHT IMPACT OF CURRENT AND ANTICIPATED
FUTURE SAFETY STANDARDS 3-2
3.2 ACCIDENT STUDIES 3-3
3.3 MATERIALS SUBSTITUTION 3-9
3.4 SUMMARY..., 3-10
4.0 EFFECT OF SPEED LIMITS ON FUEL ECONOMY AND
SAFETY 4-1
4.1 FUEL ECONOMY . 4-1
4.2 SAFETY 4-1
iii
-------
LIST OF FIGURES
Figure Page
1. Bounds of Fuel Economy Improvement From
Automobile Weight Reduction With and Without
Engine Resizing 1-8
2. Bounds of Fuel Economy Improvement From
Reduction of Engine Size 1-9
3. Relationship Between Vehicle Weight and
Chance of Serious Injury to Unbelted Driver
in Two-Car Crash 3-6
4. Percentage Fatal/Serious Injury Versus
Vehicle Weight. 3-8
5. Fatality Statistics in Urban Areas 4-4
6. Fatality Statistics in Rural Areas 4-5
7. Fatality Statistics in All Areas 4-6
v
-------
LIST OF TABLES
Table Page
1. Fuel Economy (EPA Certification Data) and
Other Statistics by Automobile Market Class
and Manufacturer (Model Year 1974) 1-2
2. Fuel Economy (EPA Certification Data) and
Other Statistics by Automobile Market Class
and Manufacturer (Model Year 1973) 1-3
3. Fuel Economy (EPA Certification Data) and
Other Statistics by Automobile Market Class
and Manufacturer (Model Year 1970) 1-4
4. Simplified Baseline Data for 1973 and 1974
U.S. Auto Fleets 1-5
5. Summary of Results Regarding the Sensitivity
Coefficient of Fuel Economy to Various
Changes and Driving Conditions 1-7
6. Major Passenger Car Safety Related Weight
Changes 3-3
7. Highway Accident Statistics for Seven States
(Delaware, Georgia, Idaho, Kentucky, No.
Carolina, So. Carolina, and Texas) 4-9
vii
-------
1.0 FUEL ECONOMY AS A FUNCTION OF WEIGHT,
PERFORMANCE, AND DRIVING SCHEDULE
The sensitivity of fuel economy to automobile weight
depends strongly on the population or subpopulation of auto-
mobiles chosen for the determination of the sensitivity, and
on the driving schedule under which the fuel economy is deter-
mined. The subject has been examined in several contexts and
the main findings are summarized below:
In the context of the total U.S. auto fleet, information
has been examined and assembled from generally available auto-
mobile statistics* and from EPA certification data based on
simulated urban driving. The results are summarized in Tables**
1, 2, and 3 for model years 1974, 1973, and 1970 respectively.
These findings are presented as a function of automobile mar-
ket class and of manufacturer.
Further compression of this information may be obtained
by a reasonable simplification. Specifically, the automobiles
in the market classes of Standard, Intermediate and Specialty
can be grouped together in a simple class under the label of
"Large". Automobiles within this class of course differ ap-
preciably in fuel economy, weight and other characteristics,
but these differences are small compared to the differences
observed between Large and Compact or, Subcompact automobiles.
Such a compression scheme may be applied to any major entry
in Tables 1, 2, and 3. As an example, it is offered here for
the total U.S. fleet, in years 1973 and 1974. The results, ap-
pear in Table 4 which provides simplified baseline data.
The baseline data of Table 4 may be used for the evalu-
ation of various scenarios regarding changes in the mix of
*Wards Automotive Yearbook and Automotive News Almanac.
**Note that all data appearing in these tables are sales
weighted averages, except the fuel economy data, which
are sales weighted harmonic averages.
1-1
-------
TABLE 1. FUEL ECONOMY (EPA CERTIFICATION DATA) AND OTHER
STATISTICS BY AUTOMOBILE MARKET CLASS AND MANU-
FACTURER (MODEL YEAR 1974).
"r-TCT
MPR
1974 *
STANDARD
INTERNED
COMPACT
SUBCOMPACT
SPECIALTY
ALL
GM
% SALES
FUEL ECON
INERTIA WT
C.1«D.
WT/C.I.D.
AXLE RATIO
16.0*
9.1
4850
405
12.0
2.84
8.8%
9.1
4307
332
13.0
2.68
4.9%
14.3
3666
303
12.1
3.04
4.6%
20.4
2789
140
19.9
2.53
7.2%
8.6
4568
387
11.8
3.05
42.3%
10.0
4327
346
13.0
2.87
FORD
% SALES
FU£L ECON
INERTIA WT
C.I.D.
WT/C.I.E.
AXLE RATIO
6.9%
9.4
5031
401
12.5
2.81
4.8%
10.5
4500
329
13.7
2.86
3.4%
15.2
3500
258
13.6
2.91
4.5%
19.2
3000
122
24.6
3.47
4.7%
14.4
3938
276
14.3
3.21
24,3%
12.3
4124
291
15.5
3.03
CHRYSLER
% SALES
FUEL ECON
INERTIA WT
C.I.D.
WT/ C.I.D.
AXLF RATIO
3.8%
8.9
4828
409
11.8
2.74
2.9%
10.1
4000
309
12.9
3.06
6.0 %
14.6
3500
229
15.3
3.15
/
0.3%
3753
380
9.9
13.0%
11.3
4002
303
13.6
3.00
AMC
% SALES
FUEL ECON
INERTIA WT
C.I.D.
WT/C.I.D.
AXLE RATIO
0.3*
10.4
4160
332
12.5
3.15 -
0.7%
11.3
4022
298
13.5
3.19
1.5%
13.9
3230
2S0
12.9
3.20
1.3%
14.9
3039
250
12.2
3,54
0.3%
11.8
3487
304
11.5
3,54
4.1%
13.2
3394
268
12.6
3.30
VW
% SALES
FUEL ECON
INERTIA WT
C 4 I.
WT/C,I.D,
AXLE RATIO
/
/
/
4.2*
21.6 **
2350 **
96 **
24.5**
3.90 **
/
4.2%
21.6**
2350 **
96 **
24.5**
3.90**
TOYOTA
% SALES
FUEL ECON
INERTIA WT
C.I.D.
WT/C.I.D.
AXLE RATIO
/
/
/
/
/
/ J
/
*.«%
19.8 **
2375 **
106 **
22.4**
1.90 *•
/
/
/
/
2.6%
19.8**
2375 **
106 **
22.4**
3.90 **
OTHER
IMPORTS
% SALES
FUEL ECON
INERTIA WT
C.I.D.
WT/C.I.D,
AXLE RATIO
/
/
1-6%
16.1*#
3050**
121**
25.2**
4.10**
7.9 %
17.4 **
2525 **
104 **
24. 3**
3.68 **
/
/
2613**
107**
24.4**
3.75**
TOTAL
DOMESTIC
% SALES
FUEL ECON
INERTIA WT
C,I,D„
WT/C.I.D.
AXLE RATIO
in.m.
9.2
4884
403
12.1
17.2%
9.7
4292
325
13.2
2.91
15.0%
14.6
35 22
260
13.5
3.18
10.4%
19.0
2908
153
19.0
3.06
12.5%
10.3
4283
342
12.5
3.10
83.7%
10.9
4169
320
13.5
3.07
TOTAL
IMPORTS
% SALES
FUEL ECON
INERTIA.WT
C.I.D.
WT/C.I,D,
AXLE RATIO
/
/
/
/
1.6%
16.1
30S0
121
25.2
4.10
14.7%
18.9
2450
102
24.0
3.79
/
16,3%
18*6
2509
104
24.1
3.82
ALL
% SALES
FUEL ECON
INERTIA WT
C.I.D.
WT/C.t.D.
AXLE RATIO
27.8*
9.2
4884
403
12.1
17.2%
9.7
4292
325
13.2
2.91
17.4%
14.7
3506
260
13.5
3.18
25.1%
18.9
2642
123
21.5
3.48
12.9%
10.3
4283
342
12.5
3.10
100%
11.8
3904
287
14.9
3.17
Source* Sales & specs from Automotive News *7 months ** Estimate
1-2
-------
TABLE 2. FUEL ECONOMY (EPA CERTIFICATION DATA) AND OTHER
STATISTICS BY AUTOMOBILE MARKET CLASS AND MANU-
FACTURER (MODEL YEAR 19 73).
\£LASJ
1973
STANDARD
INTERMED
COMPACT
SUBC0MPAC1
SPECIALTY
ALL
GM
% SALES
FUEL ECON
inertia wt
C.I.D.
WT/C.I.D.
axle ratio
19.096
10.0
4848
391
It.4
2.91
11.4*
10.7
4357
350
12.4
3.13
4.6*
12.7
3546
315
11.3
3.07
4.0*
19.4
2500
140
17.9
2.63
6.7*
9.5
4299
392
11.0
3.05
45.7*
10,8
4308
351
12.6
2.98
FORD
% SALES
FUEL ECON
INERTIA WT
C.I.D.
WT/C.I.D,
AXLE RATIO
8.0)6
8,9
4635
396
11.7
2.75
5.1*
8.7
4000
336
11.9
3.IS
3.2*
12.9
3000
271
11.1
2.98
4.3*
19.2
2750
119
23.1
3.49
3.1*
8.3
4282
388
U.O
2.86
23.7*
10.2
3889
315
13.6
3.02
CHRYSLER
% SALES
fuel ecoij
INERTIA WT
C.I.D.
WT/ C.I.D.
AXLE RATIO
4.5S
10,0
4671
375
12.5
3.12
3.0*
9.9
4000
323
12.4
3.23
5.2*
16.2
3376
246
13.7
3.23
0.4*
10.0
3500
360
9.7
3.23
13.1*
11.8
3968
311
12.9
3.19
AMC
% SALES
FUEL ECON
INERTIA.WT
C.I.D.
¦WT/C.I.D.
AXLE RATIO
0.3*
11.2
4379
332
13.2
3.15
0.5*
12.2
4214
306
13.0
3.15
l.»
18 J 9
3427
251
13.7
2.73
1.2*
18.0
3259
251
13,0
2.73
0.2*
12.9
3434
304
11.3
3.54
3.4*
15.9
3568
269
13.3
2.88
W?
% SALES
FUEL ECON
INERTIA WT
C.I.D.
WT/C.I.D.
AXLE RATIO
/
/
/
3.9K
22.0
2316
96
24.1
3.83
3,9*
22.0
2316
96
24.1
3.83
TOYOTA
% SALES
FUEL ECON
INERTIA WT
C.Z.D.
WT/C.X.D.
AXLE RATIO
/
/
/
2.0*
20.4
2398
106
22.6
3.8
0.5*
19.0
2600
120
21.7
3.7
2.5*
20.1
2438
109
22.4
3.78
OTHER
IMPORTS
% SALES
FUEL ECON
INERTIA WT
C.I.D.
WT/C.I.D.
AXLE RATIO
/
0.4%
13,3
4000
168
23.B
3.92
0.6*
16.2
6.W
19.9
24170
100
24.7
3.63
0.5*
18.6
2500
143
17.5
4.43
7.7*
19.0
2558
107
24.1
3.70
TOTAL
DOMESTIC
% SALES
FUEL ECON
INERTIA WT
C.I.D.
WT/C.X.D.
AXLE RATIO
31.8*
9.7
4925
389
12.9
Z.90
20.OK
10.0
4456
341
13.1
3.15
14.2*
14.3
3399
274
12.4
3.08
9.5*
19.1
2738
124
22.1
3.02
10.4*
9.2
4090
387
10.6
3.01
85,9*
10.9
4219
329
13.5
3.01
TOTAL
IMPORTS
% SALES
FUEL ECON
INERTIA.WT
C.I.D,
WT/C.I.D.
AXLE RATIO
/
0,4*
13.3 '
4000
168
23.8
3.92
0.6*
16.2
12.W
20.6
2415
99
24.4
3.36
1.0*
18.8
2500
130
19.2
4.01
14.1*
19.9
2468
103
24.0
3.42
ALL
% SALES
FUEL ECON
INERTIA WI
C.I.D.
WT/C.I.D.
AXLE RATIO
31.W
9.7
4929
389
12.7
2.90
20.4*
10.0
4447
337
13.2
3.17
14.8*
14.4
3389
274
12.4
3.08
21.8*
19.9
2557
110
23.2
3.15
11.4K
9.6
3963
366
io.a
3.09
100.0*
11.6
3979
296
14.8
3.06
Sourcai Sal.# « cpaea from Automotive Newi,w»sdt Automotive
1-3
-------
TABLE 3. FUEL ECONOMY (EPA CERTIFICATION DATA) AND OTHER
STATISTICS BY AUTOMOBILE MARKET CLASS AND MANU-
FACTURER (MODEL YEAR 19 70).
1970
COMPACT
SUBCOMPACT
ALL
% SALES
FUEL E CON-
INERT IA WT
C.I.D.
WT/C.I.D.
AXLE RATIO
20.0*
10,9
4720
389
12,2
5.0*3
12,0%
13.1
4041
334
12.1
3.10
2.9%
13.7
3500
264
13.3
3, 09
0.3*
2500
140
17.9
2.53
5.5*
11.7
4005
356
11.2
3.52
40.7*
11.6
4320
356
12.2
3.12
* SALES-
FUEL ECON
INERTIA WT
C.I.D.
WT/C.r.D.
AXLE RATIO
12.5%
11.1
4526
359
12.6
3.13
5.4*
11.7
4000
293
13.7
2.81
4,8*
19.2
2814
177
15.9
2.84
0.9*
2250
110
20.5
3.55
3.6*
LI.5
3938
323
U.2
2.96
27 .2*
12.2
3967
323
13.7
3.01
* SALES
FUEL ECON
ItfERTIA ¥T
C.I.D,
WT/ C.I.D.
AXLE RATIO
6.4*
12.3
4587
346
13.3
3.08
3.1*
13.3
3755
302
12.4
3 .07
5.3*
15.3
3500
244
14.3
3.22
/
1.6*
12.0
3850
321
12.0
3.23
16.4*
13,3
4007
302
13.3
3.14
* SALES
PUEL ECON
INBBTIft.WT
C.I.D,
HO/C.I.D.
AXLE RATIO
1.2*
13.1
3913
284
13.8
2.94
1.0*
16.5
3000
207
14.5
2. 96
0.5*
17.9
3000
199
15.1
2.73
0.4*
15.2
3500
296
11.7
3.15
3.1*
15.0
3419
247
14.0
2 .95
* SALES
PUEL BOON
INERTIA WE
C.I.D.
WT/C.I.D
AXLE RATIO
7.0*
21.4 *
2379
97
24.5
4.13
7.0*
21.4 *
2379
97
24.5
4.13
* SALES
PUBL ECON
INERTIA VT
C.I.D.
WT/C.I.D,
AXLE RATIO
0.1*
14.7
3500
138
25.4
4,11
2.2*
23.0 *
2412
100
24.1
3.90
2.3*
23.0
2459
102
24.2
3.91
OTHER
IMPOSTS
* SALES
FXJZh ECCK
INERTIA WT
C.I.D.
WT/C.I.D,
AXLE RATIO
0.5*
3072
130
23.6
4.11
2.9*
20.0
2343
83
28.2
4.01
3.4*
20.0
2450
90
27.6
4.03
TOTAL
DOMESTIC
* SALES
FUEL ECOH
INERTIA WT
C.I.D.
WT/C.I.D.
AXLE RATIO
38.9*
11,2
4636
372
12.5
3.08
21.7%
12.7
3979
317
12.6
3.01
14.0*
16.1
3230
222
14.5
3.08
1.7*
2500
140
17.9
3.16
11.1*
11.6
3941
339
11.6
3.28
37.4*
12.2
411S
326
12.9
3.09
TOTAL
IMPORTS
* SAL£S
FUEL ECOH
ikwtcr.wt
C.I.D.
WT/C.I.D
AXLE RATIO
/
o.e*
3L2B
131
23.9
4.11
12.1*
21.3
2 364
94
25.1
4.03
12.7*
21.3
2400
96
25.0
4.03
* SALES
FUEL SCON
ISEOTXA WT
C.I.D,
WT/C.I.D.
AXLE RATIO
2-0.9*
11.2
4636
372
12.5
3.08
21.'?*
12.7
3979
317
12.6
3.01
14.«*
16.1
3225
216
14.8
3.12
13. •*
21.0
2381
100
23.6
3.91
11.1*
U.8
3941
339
11.6
3.28
100*
12.9
3902
297
14.3
3.20
Sourc*j Salop 6 apses from Automotive (few*
* Estimate
1-4
-------
TABLE 4. SIMPLIFIED BASELINE DATA FOR 1973 AND 1974 U.S. AUTO FLEETS
1 1973
1974
I LARGE
COMPACT
SUBCOMp.
ALL
LARGE
COMPACT
SUBCOMP.
ALL
% SALES
63.6
14.8
21.6
100
57.5
17.5
25.0
100
FUEL ECONOMY
MPS .
9.8
14.4
19.9
11.6
9.6
14.7
18.9
11.8
INERTIA WT
LBS
4600
3390
2560
3980
4580
3510
2640
3910
C.1.0.
368
274
110
298
366
260
123
287
WT/C.I.D. j
12.5
12.4
23.3
14.8
12.5
13. 5
21. 5
14.9
SOURCE: TABLE 1 AND 2.
-------
the U.S. fleet. The three appearing classes, namely Large,
Compact and Subcompact may be thought of as six, five and
four passenger cars respectively. A further entry, if de-
sired, may be made regarding a safety figure of merit as a func-
tion of the automobile weight. Thus, simple trade-offs can
be exercised.
In the context of individual automobiles, no significant
test data are available regarding the effects of automobile
weight, performance and driving schedule on fuel economy.
Thus, the approach followed here consists of utilizing exten-
sively test data available for engines, drive trains and other
automobile components of the 19 73 model year. Such data have
been used in conjunction with simulated automobile travel un-
der various driving schedules of current interest.
Table 5 gives the values of the sensitivity coefficient
of fuel economy to:
(1) a weight change (without other changes)
(2) a weight change accompanied by a proportional change
in engine displacement, (weight/CID ratio unchanged)
(3) an engine displacement change (without other changes)
The sensitivity coefficient in each case gives the per-
cent change in fuel economy per one percent change of weight
or other parameter as shown. Alternatively the sensitivity
coefficient may be visualized as the slope of the percent
variation of fuel economy plotted against the percent vari-
ation of automobile weight, or engine size. The three cases
are shown in Figures 1 and 2. These illustrations correspond to
the entries made under EPA Composite Driving Schedule, in Table
5.
Note that various driving schedules of current interest
have been considered. The results appearing under composite
schedules are suggested as fair indicators of fuel economy
under conditions of nation-wide mileage accumulation. Speci-
fically, the EPA Composite is a harmonic average of EPA Urban
1-6
-------
TABLE 5. SUMMARY OF RESULTS REGARDING THE SENSITIVITY COEFFICIENT OF FUEL ECONOMY
TO VARIOUS CHANGES AND DRIVING CONDITIONS
DRIVING
type schbd
OF CHANGE
EPA
URBAN
EPA
HIGHWAY
SAE
URBAN
SAE
SUBURBAN
SAE(55 MPH)
INTERSTATE
SAE(70 MPH)
INTERSTATE
EPA
COMPOSITE
SAE(55 MPH)
COMPOSITE
SAE(70 MPH)
COMPOSITE
WT. CHANGE
.26
.21
.31
.31
.15
.20
.25
.26
.27
ONLY*
+ .04
+ .04
+ .06
+ .08
+ .04
+ .05
+ .04
+ .04
+ .03
WT, CHANGE;
(WT/C.I.D.)
.70
.6S
.75
.74
.54
.57
.68
.68
.68
CONSTANT*
+ .09
+ .13
+ .11
+ .10
+ .08
+ .15
+ .10
1 +
o
CD
+ .10
C.I.D.
CHANGE
.43
.44
.41
.44
.37
.38
.44
.41
.41
ONLY*
+ .09
+ .13
+ .11
+ .12
+ .08
+ .15
+ .09
+ .08
+ .10
•CHANGE IN FUEL ECONOMY PER ONE PERCENT PARAMETER CHANGE.
-------
and EPA Highway, weighted respectively 55% and 45%, to reflect
FHWA statistics of nation-wide driving. Similarly the SAE
composites include 27.5% SAE Urban, 27.51 SAE Suburban and
45% of SAE Interstate at either 55 MPH or 70 MPH. It is
evident that when composites are used, the differences between
driving schedules are minimized.
| 20-
O
z
o
U
D
u.
&
UJ
s
I
J—
z
UJ
U
.Or
fif*'
-r
10
20
1
30
40
PERCENT REDUCTION OF AUTOMOBILE WEIGHT
Source: Table 5, EPA Composite
Figure 1. Bounds of Fuel Economy Improvement From Automobile
Weight Reduction With and Without Engine Resizing
1-8
-------
Figure 2.
Source: Table 5, EPA Composite
PERCENT REDUCTION OF ENGINE SIZE
Bounds of Fuel Economy Improvement
From Reduction of Engine Size
-------
2.0 TRAFFIC CONTROL FOR SAFETY AND FUEL ECONOMY
The primary emphasis of the U.S. Department of Trans-
portation is to promote the safe and efficient movement of
both people and goods throughout our country. Safety is us-
ually measured in the number of accidents, injuries and/or
fatalities along with their associated social and monetary
costs. Efficiency can be indicated by travel times, fuel
consumption, comfort, convenience and other user costs associ-
ated with the operation of the facilities involved. Because
of the recent energy crisis, the measure of fuel consumption
has become one of the more prominent measures of efficiency,
Therefore it behooves the Department of Transportation to
embark on a major effort to improve the safety and fuel con-
servation of our transportation systems.
One of the major potential methods to improve the opera-
tional safety and efficiency of our street and highway system
is through the improvement of the operational controls for the
traffic using these facilities. Studies have shown that motor
vehicles traveling at a uniform speed of about 35 miles per
hour provide the highest mileage rates per unit of fuel con-
sumed. ^ At approximately 35 mph our controlled-access road-
ways operate at their peak potential capacity and we obtain
(2)
maximum throughput. J Uniform speed without vehicle or
pedestrian conflicts also provides the safest traffic flows.^
Therefore, the application of appropriate traffic surveillance,
communication and controls, can substantially improve the move-
ment of people and goods over our roadways network.
The traffic signal is one of the basic traffic control
devices used on our urban streets. Its function is to assign
the right-of-way to conflicting streams of traffic at roadway
intersections. In cities of 1/2 to 1 million people, 60 per-
cent of the signalized intersections are periodically congested,^
2-1
-------
With modern technology, the timing of these traffic sig-
nals can be adjusted and coordinated to promote continuous
movements for major platoons through the street network thus
providing travel time saving and fuel conservation. Through
the proper timing of traffic signals, the number of stops can
be minimized thus saving fuel and improving safety by reducing
the chance for rear-end accidents. Through vehicle detectors
located on the approaches to signalized intersections, traffic
flow information can be collected and traffic signal timing
patterns can be determined by electronic computer traffic sig-
nal controllers to assign green times to keep most of the traffic
moving at all times. The fluctuation of traffic demands requires
a flexible traffic control system to best serve these demands.
The electronic digital computer can best supply this con-
trol flexibility. It is estimated that the implementation of
modern urban traffic control signal systems can generally pro-
vide a 20 percent reduction in travel time, about a 15 percent
savings in fuel and a 10 percent improvement in accidents.^
Our freeway system was initially conceived to promote the
rapid safe movement of motor vehicles by minimizing traffic
conflicts through the control of access to the freeway. The
access control was obtained by limiting the number and the
spacing of entrance and exit ramps to the freeway and spacially
separating the crossing roadways. In most of our urban areas,
however, the traffic demand for use of freeways periodically
exceeds the traffic capacity of the facilities and the smooth
flow of traffic breaks down into a stop-and-go operation. Such
traffic congestion creates a hazardous situation as well as
wastes vehicle fuel. Through the electronic traffic surveillance
of freeways and the adjacent roadway facilities, freeway en-
trance ramps can be controlled more positively to limit the
traffic using the freeway in order to assist in maintaining
smooth flow operations. The metering of entrance ramp vehicles,
2-2
-------
one-at-a-time, has also improved the safety of the vehicle
merging operation into the freeway traffic stream by eliminating
the rear-end collision hazard between ramp vehicles. Ramp
metering in Atlanta decreased ramp rear-end accidents by 85 per-
cent. ^ Freeway controls in general have increased average
peak period speeds by about 15 percent and reduced accidents
by some 40 percent.^ Fuel savings from freeway control pro-
jects may range from 10 to 20 percent.^
Because of the traffic surveillance capabilities of freeway
control projects it is possible to rapidly detect capacity-
reducing incidents and therefore quickly dispatch emergency aid.
Twenty-one million emergency stops per year occur on urban free-
ways of which 10 percent occur on the traveled way and require
removal.^ The delay due to incidents totals to 750 million
vehicle hours per year on U.S. urban freeways.Thus it can
easily be shown that incident detection and rapid response can
more quickly aid the needy motorist, clear away the capacity
reducing obstacle and return the freeway to normal operations
in a shorter time than if a freeway surveillance and control
system were not in operation. The quick response to injury
producing accidents will save lives.In addition, when a
freeway stoppage is detected, drivers upstream can be warned
of the congestion ahead and advised to use an alternate route.
It is estimated that 400 million gallons of fuel are wasted per
year because of delays caused by urban freeway incidents.
Through the selective use of controls on our streets and
highways, certain types of vehicles can be given preference
over the movements of the rest of the traffic. The use of ex-
clusive bus lanes on city streets and freeways and the granting
of priority for providing green lights to buses at traffic signal
controlled intersections can improve the operation of mass tran-
sit which may in turn promote more ridership on transit vehicles.
2-3
-------
In areas that have installed exclusive bus lanes as part of
a comprehensive traffic management program that includes improved
signalization, vehicle accidents have been reduced by 15 to 20
percent, and pedestrian accidents have been reduced by 35 percent
and bus passenger injuries have been reduced by more than 40
percent.The limiting of some freeway entrance ramps for
the use of bus and carpools and the reservation of a high speed
freeway lane for bus and carpool use can encourage drivers to
divert to multi-occupancy vehicles. Parking controls favoring
carpool vehicles can also provide carpool incentives. The more
drivers who can be enticed to join a carpool or to ride a bus
decreases the number of vehicles on the road during peak hours
which results in improved traffic operations, saves fuel and
reduces chances for accidents. If the average automobile oc-
cupancy were increased from 1.4 to 1.6 persons per vehicle
through carpooling, there would be a savings of 3.2 billion
f 141
gallons of fuel a year.
Traffic safety and fuel savings can be obtained through
the prudent application of traffic control techniques which
have been proved to be successful in pilot research studies
that have been conducted in various parts of the country.
Although many cities are installing or are planning to install
computer controlled traffic signal systems, many of the systems
proposed are not capable of taking advantage of advanced traffic
signal control strategies. The computers selected for control
purposes should be able to utilize programs written in a common
language such as FORTRAN. More engineers should be trained in
the design, operation and maintenance of advanced electronic
traffic control systems so that city and state agencies can
adequately implement and operate the new sophisticated systems.
The DOT should promote and assist State and local governments
in implementing the needed traffic control improvements to
save both lives and fuel.
2-4
-------
FOOTNOTES (SECTION 2.0)
[1) Fuel consumption tables in Paul J. Claffey's NCHRP
Report 111 on "Running Costs of Motor Vehicles as Affected
by Roadway Design and Traffic/' Highway Research Board, 1971
show that passenger vehicles use less fuel at speeds between
30 and 40 miles per hour (See Table 6 on page 17). Other
vehicles have minimum fuel consumption around the same speed
range or a little below (See Table 10, page 21 and the figures
in appendix A).
(2) Many studies have confirmed the fact that the highest
volumes can be obtained if traffic is moving at about 35 mph.
A typical example is shown in figure 3.44 on page 66 of the
1 - 1965, Special Report 87 of the
(3) Many studies have shown that controlled access highways
are safer than surface roadways, that "T" type intersections
are safer than four legged junctions, that minimum accident
involvement rate occurs for those traveling from the average
speed to 10 mph above it etc. which all come down to the less the
number of conflicts the less chance of accident occurrence.
The series of chapters on "Traffic Control § Roadway Elements
Their Relationship to Highway Safety/Revised" published by the
Highway Users Federation for Safety and Mobility from 1968 to
19 71 summarizes the literature very well to point out this
finding.
(4) The fact that up to 60 percent of the signalized inter-
sections are periodically congested is pointed out in the final
report by the Polytechnic Institute of Brooklyn in their final
report for NCHRP Project 3-18(2) on "Traffic Control in Over-
saturated Street Networks." This finding is a result of a
nationwide survey they conducted which is shown in figure 2-2
on page 45 of their report.
(5) An analysis o£ cities which have installed computer traffic
control signal systems was presented by Charles R. Stockfisch
in his report on "Selecting Digital Computer Signal Systems"
published by FHWA in 19 72. These reported findings presented in
Table 3 on page 60 show that for 307 intersections studied,
delay was reduced 20 percent, stops reduced 27 percent, accidents
reduced by 13 percent and travel time was reduced 28 percent.
Considering the reduction of the number of stops and the travel
time, it is safe to assume fuel saving will be about 15 percent
considering just one stop-start cycle requires 19 percent more
fuel per mile than a steady state driving speed of 30 mph.
2-5
-------
(6) The results of the Atlanta ramp metering experience are
reported in Paul E. Everall's report on "Urban Freeway Sur-
veillance and Control-The State of the Art" published by FHWA
in 19 72. The Atlanta results can be found on page 119 of the
report.
(7) Paul Everall's report on "Urban Freeway Surveillance
and Control" also presents the accident results for a number
of freeway control projects in table 34 on page 144. The in-
crease of freeway speeds reported for various freeway control
projects includes an increase of 34 percent in Detroit, 50 per-
cent in Los Angeles, 60 percent in Houston and 70 percent in
Chicago. Dallas has experienced about 15 percent increase in
speeds. It is felt that the more conservative 15 percent
might be expected for general application since the pilot re-
search studies were installed at problem areas whereas the
more general application would cover larger systems such as
is installed in Dallas.
(8) The fuel saving is based on an estimate that the level
of congestion would be reduced through freeway controls and
therefore the stop and start situation would be reduced. As
noted before, the stop and start movement of vehicle consumes
a significant amount of fuel.
(9) The number of emergency stops per year on U.S. roads was
estimated in a report on "Post-Crash Communications" prepared
by AIL, a Division of Culter-Hammer for the National Highway
Safety Bureau in 19 70. The figures quoted was obtained from
Tables 4-26 and 4-27 on page 4-34 of the report.
(10) The delay due to incidents was obtained from the same
report by AIL from the Table 4-29 on page 4-36.
Cll) It has been pointed out by several experts that quick
response to injury accidents can save lives. Dr. Haddon
quoted a report by the California State Health Department in
a Spring 1966 Symposium on "Traffic Safety a National Problem"
(See Eno publication of 1967 page 156) which "shows that
patients with exactly the same injuries have four times the
probability of dying if those injuries occur in automobile
accidents in rural areas instead of an urban area." The AIL
report mentioned above in (9 § 10) points out in pages 6-4 and
6-5 that the number of fatalities could be reduced if delays
in receiving proper medical care were reduced.
(12) The gallons of fuel used because of delays was deter-
mined from the 750 million vehicle hours of delay and using
an idling fuel consumption rate of 0.54 gallons per hour; the
2-6
-------
rate is about 400 million gallons of fuel. Claffey in NCHRP
Report 111 uses 0.63 as the composite automobile gallons per
hour fuel consumption rate.
(13) The figure used for the effectiveness of bus lanes
was obtained from a report entitled "Bus Priority" published
by the Transport and Road Research Laboratory of England,
TRRL Report LR 5 70. It is a summary of several studies
presented in table 2 on page 15.
(14) The average automobile occupancy rate of 1.4 was ob-
tained from several publications. This rate is quoted any-
where from 1.2 to 1.6 in different reports. The fuel savings
was determined from the following analysis
5.55 billion barrels of petroleum in 1971
46 percent is gasoline
71 percent of the gasoline is used for autos
34 percent of auto usage is for the work trips
5.55 (.46)(.71)(.34) = 0.615 billion barrels for auto
work trips
with 42 gal per barrel, 42(.615) = 25.8 billion gallons
Using a 12.5 percent reduction due to increased passenger/
veh 1.4-1.6
25.8 x .125 « 3.2 billion gallons per year
Or as reported 106 billion gallons of gasoline used in 1973
106 x .71 x .34 = 25.6
25.6 x .125 = 3.2 billion gallons of gas saving if
average auto occupancy increased
from 1.4 to 1.6
2-7
-------
3.0 WEIGHT VERSUS SAFETY
The relationship between automobile weight and safety is
difficult to quantify, but qualitative conclusions can be
drawn. Automobile size classes provide a convenient scheme
for general conclusions; however, these classes are not well
defined and the weight variation within a class can be quite
large. One of the most universal definitions for class dis-
tinctions is that of the Motor Vehicle Manufacturers Associa-
tion which separates the classes by wheelbase, (Ref. 1, p. vii).
The following table shows the classes, wheelbase range and a
typical, but not totally inclusive, curb weight range for
each size class:
It is common practice to equate car size with car
weight. While this relationship is generally true, the
above table shows that weight may overlap, though wheelbase
does not. Smaller cars commonly have somewhat lighter con-
struction than large cars; the weight overlap usually occurs
when one compares the largest and/or most expensive cars in
a given class with the smaller less expensive cars in the next
bigger class.
Car weight varies for a given make and model year
depending upon optional equipment. This tends to increase the
amount of overlap between classes. Foreign cars and full
size cars tend not to vary so much, on the order of 6 percent
or less for 1974 models, because there is less relatively
heavy optional equipment available. Domestic subcompacts can
Size Class
Subcompact
Compact
Wheelbase
Range (in.)
Curb Weight
Range (lb.)
1,800-2,750
2,875-3,575
3,550-4,100
3,950-5,100
Intermediate
Full Size
94-101
102-111
112-118
119-up
3-1
-------
easily vary 8 perpent, compacts 12 percent and intermediates
10 percent.
Weight changes can be particularly critical with the
smaller cars because of the smaller crush distances and in-
terior volume. The basic car structure which absorbs crash
energy is usually the same regardless of options, and the
absorption capability is probably related to the curb weight.
The weight that determines the energy which must be absorbed
is the weight at the time of the accident. This weight is
affected by payload, passengers and baggage, as well as
equipment; the effect can be greater in small cars than in
large cars. Three passengers increase the weight of a 2,000
pound car by 20 percent over the weight of the car and a
driver; the corresponding factor for a 4,500 pound car is
about 10 percent. Fortunately, this effect is mitigated
somewhat by low passenger density, however, a subcompact car
can easily weigh 25 to 30 percent more than curb weight at
the time of the crash.
3.1 WEIGHT IMPACT OF CURRENT AND ANTICIPATED FUTURE
SAFETY STANDARDS
1. Current standards have added approximately 260
pounds to automobiles as shown in Table 6.
2. Future bumper standards may add 45 pounds (sub-
compact) to 100 pounds (full size) if steel designs
are used. "Soft-nose" designs may decrease pre-
sent system weights by 50-100 pounds.
3. New brake system requirements may add 5-25 pounds
depending upon whether antilock is required on a
particular car.
4. The 30 mph passive restraints may add 55 to 80 pounds
to a car with no weight change considered for mounting
structure or increased length to provide deceleration
space. The weight change may be greater for small
cars. The 45 to 50 mph passive restraints may add
an additional 150 to 270 pounds to car weight.
3-2
-------
5. Probable weight increases for future standards
are:
Subcompact 2SO-375 pounds
Compact 325-450 pounds
Intermediate 525-450 pounds
Full Size 325-450 pounds
The major reason for the large range is the rela-
tively undefined effect of "soft-nose" bumper
designs. The effect; i.e., weight reduction, will
probably be greater on larger cars.
TABLE 6. MAJOR PASSENGER CAR SAFETY RELATED WEIGHT CHANGES
Standards in Effect Weight Increase (lbs)
100 Series 5
201 - 204, 207, 210 32
208 (Belts) 35
214 (Side Door Strength) 50
215 (Bumper) 141
263
Issued Standards Not Yet In Effect
215 (Bumper Corner Requirements)
9/1/75 9
105-75(Hydraulic Brakes) 9/1/75 5 -25
Possible Future Standards
Before 1980 FMVSS 208 (30 mph)
Part 581 Mo Damage Bumper
After 1980 FMVSS 208 (45-50 mph)
Total
3.2 ACCIDENT STUDIES
There have been a limited number of studies using acci-
dent data to investigate the relationship between car weight
and safety. Unfortunately each investigator established his
3-3
~ 55 -
80
lbs.
~ 45 -
100
lbs.
-150 -
270
lbs.
-250 -
450
lbs.
-------
own weight classes so results are not directly comparable.
Also, each investigator assigned the makes and models to his
classes by using some average weight. The effect is not too
serious for generalized results, but absolute numbers cannot
be obtained.
Another factor clouding accident data analysis is that
accident data are primarily collected for purposes other than
assessing crashworthiness, e.g., police reports are primarily
for roadway traffic control and law enforcement and insurance
reports are primarily for damage claims.
The question of involvement rate also clouds the issue;
do large and small cars have difference accident involvement
rates and, if so, why? Some evidence indicates that small
cars do have a higher involvement rate than large cars. The
Highway Loss Data Institute (HLDI) studies first party claims
and derives claim frequency per 100 insured vehicle years;
they attempt to standardize the data by adjusting for driver
age, under or over 25, and amount of deductible, $50 or $100.
Their results for 1972 and 1973 models as reported by the
Insurance Institute for Highway Safety (IIHS) show that com-
pacts and intermediates have a claim frequency approximately
20 percent greater than full size cars and subcompacts have
a claim frequency 33 percent greater than full size cars
CRef. 2; p. 3).
The Highway Safety Research Institute, University of
Michigan, has published data indicating that small cars are
over-involved in "out-of-control" single car accidents
(Ref. 3, p. 11^ and in collisions with other small cars
(Ref. 4, p. 5). In part, this over-involvement has been
attributed to average driver age with small cars having
younger drivers. The evidence is not yet conclusive so it
is not unreasonable to assume a constant involvement rate
regardless of size in any study to determine the effect of a
reduction in average vehicle size.
3-4
-------
The relationship between car weight and safety has been
studied intensively on various samples, but such studies have
not been extensive. As mentioned above, each investigator
devises his own weight classes. O'Day, et/al, studied "small
cars" with a licensing weight of 3,100 pounds or less and
"large cars" with a licensing weight of 3,300 pounds or greater;
the 200 pound gap was intentional to provide separation. They
found (Ref. 5, p. 5) that the chance of a car being involved
in an accident was not highly related to its weight, but once
involved the chance of injury in the car increases 2.5 percent
for each 100 pound decrease in weight.
Mela studied a set of 32,980 two-car crashes which
occurred in New York State during calendar years 1969-71.
He divided the cars into five weight classes (Ref. 6, p. 48-7).
He found that the driver's chance of serious injury increased >
or decreased by about 5 percent for each 100 pound increase or
decrease in his car weight; his chance of serious injury in-
creased or decreased approximately 2 percent for each 100
pound increase or decrease in the weight of the other car in
the collision (Ref. 6 p. 48-8). The data also implied that if
all cars involved in two-car crashes weighed 3,170 pounds, the
overall injury rate would be the same as it was for the popula-
tion studied which had an average weight of 3,360 pounds. The
effect of a given weight change on safety may vary with car
population. Figure 3 shows one relationship between changes
in the average weight of the vehicle population and the chance
of serious injury to unbelted drivers in two-car crashes.
Mela concludes that a shift in the U.S. passenger car popula-
tion from its present weight distribution to one composed
largely, i.e. 50 to 60 percent, of subcompact and compact cars
with no cars exceeding 3,250 pounds could produce up to 25
percent more serious and fatal injuries. Doubling of safety
belt usage would enable smaller cars to be used with about the
same injury rate as now exists (Ref. 6, p. 48-11, 17, 18).
Scott also investigated a New York State accident file
and found that when the primary vehicle shipping weight is
3-5
-------
04
I
PERCENT
SERIOUS
INJURY
UNBELTED
DRIVERS
9
8
7
6
5
4
-
50
-
40
-
30
20
BASE CASE, N.Y.
10
-
^e^STATE 1969-71
-10
—
-20
1 1
1 J i
2000 2500 3000 3500
AVERAGE VEHICLE WEIGH (LBS)
4000
PERCENT
CHANGE
FROM
"BASE
CASE"
Figure 3. Relationship Between Vehicle Weight and Chance of
Serious Injury to Unbelted Driver in Two-Car Crash
-------
under 2,500 pounds the severity is essentially the same
regardless of the weight of the secondary vehicle; this
severity is 2 to 3 times that of a larger car, above 3,250
pounds, striking another large car (Ref. 7, p. 13-15 and
13-22). He also found that (Figure 4) belted drivers of cars
in the 1,000 to 1,999 pound class have about the same percent of
serious and fatal injuries as unbelted drivers of cars in
the 3,250 to 3,999 pound class (Ref. 7, p. 13-16). These
findings correspond well with those of Mela.
Campbell and Reinfurt studied accidents occurring in
North Carolina in 1966 and 1968-71. They established an
injury index with 100 as the average value. An index of
120 means 20 percent more frequent serious injury to a
driver than average and, an index of 80 means 20 percent
less frequent serious injury than average. In car vs. car
crashes the index chart crosses 100 at approximately 3,000
pounds for the two cases: (a) all injuries and (b) serious
and fatal injuries (Ref. 8, p. 21 and 22).
In later work Campbell, O'Neill and Tingley studied
crashes of 1970-73 model cars dividing them into five cate-
gories: full size, intermediate, compacts, subcompacts and
import cars. Their data show that the smaller cars, including
imports, tend to have twice as many single car, predominately
ran-off-road, accidents as the combination of full size and
intermediate cars. This holds true for both belted and
unbelted drivers (Ref. 9, p. 12-3). This supports the findings
of the Highway Safety Research Institute mentioned above.
Using an index system similar to that described above the
data show that belted drivers in subcompact and compact cars
have approximately the same chance of injury as unbelted
drivers of intermediate and full size cars (Ref. 9, p. 12-6).
This also implies that greatly increased usage of seat belts
would enable smaller cars to be used with about the same
injury rate as now exists. This supports the Mela findings.
Current national policy is that the present rate is
3-7
-------
0)
• H
I-
3
u
in
DO
"O c
C
rH
t/l O
o >
•H C4J
• pi 00
rH tfl-H
CO 4-i 0)
~J C 3e
W 4)
tu "0 -O
¦r-t Q)
+-> U -P
c u a
D < u
u -H
K (-1 T3
DOG
&. i-f
12
10
Driver Unbelted
Driver Belted
_L
_L
X
J
1000-1999 2000-2499 2500-3249 3250-3999 4000-5499
Vehicle Weight (lbs)
Source: "A Safety Comparison of Compact and Full-Size Auto-
mobiles," by Basil Y. Scott, N.Y. State Department
of Motor Vehicles, presented at 3rd International
Congress on Auto Safety, Vol. I, 15 July 1974 in
San Francisco.
Figure 4. Percentage Fatal/Serious
Injury vs. Vehicle Weight
3-8
-------
unacceptable and must be reduced. On that basis any change
to a higher percentage of smaller cars would be self-defeating.
Efforts to increase belt usage have met with well pub-
licized resistance. Interlocks have increased usage some-
what, but laws may be necessary to achieve the usage level
required to accommodate a switch to smaller cars. Passive
restraints may be the solution; however, matching passive
restraints to smaller cars is at present an unresolved prob-
lem. Even with universal passive restraints equivalent to
present belts, the injury rate for a population of smaller
cars would remain approximately the same as the present level
for all cars.
The above findings indicate that injury levels become
a problem if the curb weight of an automobile drops below
5,000 to 3,200 pounds. This conclusion is based upon acci-
dent data and is, therefore, a measure of the capability of
cars using present designs and manufacturing techniques.
The National Highway Traffic Safety Administration is
sponsoring the Research Safety Vehicle (RSV) Program which
has as a goal a 3,000 pound car with crashworthiness at
speeds up to 50 mph. This program is currently in Phase I;
the results of Phase III are to be applicable to cars of the
mid-1980's. Such cars would not represent a significant
portion of the automobile population until 1990.
3.3 MATERIALS SUBSTITUTION
More significant from a near term point of view would be
material substitutions enabling production of relatively
large, light weight cars. Aluminum and plastic are commonly
mentioned substitutes. Aluminum prices are increasing; alu-
minum production requires large amounts of energy; and there
are valid questions as to the availability of aluminum in
sufficient quantities. Recent unpublished findings under a
DOT contract indicate that increasing aluminum usage from the
3-9
-------
present 81 pounds per car to 200 pounds per car would require
increasing the industry capacity by 15 percent. Such an
increase would reduce car weight about 120 pounds. Plastics
used in automobiles are largely petroleum products so a major
limitation on their use is self-evident.
Likely near term substitutes are the various high
strength low alloy (HSLA) steels. These steels have signi-
ficant strength advantages, but require larger forming equip-
ment as the punch forces must be higher. Another problem is
variation in weldability from run-to-run and batch-to-batch.
An in-service problem exists because the thinner sections
will be less tolerant of corrosion (Ref. 10, p. 30-32).
Recent unpublished findings under a DOT contract indicate
that extensive use of HSLA steels could reduce the weight of
a standard size passenger car by 300 to 400 pounds.
Increasing the use of aluminum and incorporating HSLA
steels could conceivably reduce the weight of a standard size
car by 10 percent by 1979 or 1980. A similar reduction
appears possible for intermediates; the reduction for the
smaller size classes would be somewhat less. Such cars
could represent a significant proportion of the automobile
population by 1985. The effect of materials changes on injury
rate may not be commensurate with the magnitude of the reduc-
tion in curb weight because the weight of the payload is un-
changed. Cars in the various size classes would have the
same relative strength that they now have, but crash energy
levels based upon weight at impact would be reduced by a
smaller percentage than the reduction in curb weight.
3.4 SUMMARY
1. The relationship between automobile weight and safety
is difficult to quantify, but qualitative conclu-
sions can be drawn. As it is common practice to
equate car size with car weight, automobile size
3-10
-------
classes provide a convenient scheme for general
conclusions; however, the weight/size relationship
is not exact. For any given make/model the basic
car structure which absorbs crash energy is usually
the same regardless of options. The weight that
determines the energy that must be absorbed is the
weight at the time of the accident and is affected
by payload and equipment. The effect of loading
can be greater in small cars than in large cars.
Consequently it is mandatory that any legislation
on "size" or "weight" of cars define these terms
rigorously.
2. Various investigators have studied accident data to
determine the relationship between weight and safety.
Their findings are not directly comparable, but it
appears that the chance of serious to fatal injury
increases considerably at curb weights less the
3,000-3,200 pounds. Doubling seat belt usage would
enable these smaller cars to be used with about the
same injury rate as now exists, recognizing that
present national policy is that this rate is unaccept-
able.
3, The above conclusion is a measure of the capability
of cars using present designs and manufacturing
techniques. Material substitutions are the most
likely means of maintaining car size while reducing
car weight. Such substitutions must preserve struc-
tural strength if injury rates are to be maintained.
High strength low alloy steels appear to be the most
promising materials ; extensive use of these steels
could reduce the weight of a standard size passenger
car by 300 to 400 pounds. Combined with some in-
creased use of aluminum the weight reduction should
approach 10 percent for a standard size car by 1979
3-11
-------
or 1980. The weight reduction will be somewhat
less for the smaller size classes.
References (Section 3.0)
1. 1974 Model Year Passenger Car and Truck Accident Inves-
tigators Manual, Motor Vehicle Manufacturers Association
of the United States, Inc., Detroit, Michigan
2. "Status Report," Vol. 8, No. 23, December 20, 1973,
Insurance Institute for Highway Safety, Washington, DC
3. "Small-Car Accident Involvement," Brown, HIT Lab Reports,
November 1970, Highway Safety Research Institute,
University of Michigan, Ann Arbor, Michigan
4. "Crashes Between Large and Small Cars: The Realworld
Perspective," Epstein and O'Day, HIT Lab Reports,
April 1972, Highway Safety Research Institute, Univer-
sity of Michigan, Ann Arbor, Michigan
5. "A Statistical Description of Large and Small Car Involve-
ment in Accidents," O'Day, Golomb and Cooley, HIT Lab
Reports, May 1973, Highway Safety Research Institute,
University of Michigan, Ann Arbor, Michigan
6. "How Safe Can We Be In Small Cars", Mela, Proceedings of
the Third International Congress on Automotive Safety,
National Motor Vehicle Safety Advisory Council, DOT,
Washington, DC
7. "A Safety-Comparison of Compact and Full Size Automobiles,"
Scott, Proceedings of the Third International Congress on
Automotive Safety, National Motor Vehicle Safety
Advisory Council, DOT, Washington, DC
8. "Relationship Between Driver Crash Injury and Passenger
Car Weight", Campbell and Reinfurt, November 19 73, Highway
Safety Research Center, University of North Carolina,
Chapel Hill, N.C.
9. "Comparative Injuries to Belted and Unbelted Drivers of
Subcompact, Compact, Intermediate and Standard Cars,"
Campbell, O'Neill, and Tingley, Proceedings of the Third
International Congress on Automotive Safety, National
Motor Vehicle Safety Advisory Council, DOT, Washington, DC
10. "High Strength Low Alloy Steels for Auto Parts, " Maloney,
Heimbuch and Rose, Automotive Engineering, July 1974,
Society of Automotive Engineers, New York, N.Y.
3-12
-------
4.0 EFFECT OF SPEED LIMITS ON FUEL ECONOMY AND SAFETY
4.1 FUEL ECONOMY
During the recent oil crisis a national speed limit of
55 mph was established in order to decrease fuel consumption.
It was indeed expected that by lowering the speed limit to
5 5 mph there would be appreciable savings in fuel. Limited
data available indicate that nationwide gasoline consumption
for the period November 1973 through March 1974, when com-
pared to the same period for the year before, is down about
3.8 percent; at the same time the data show that motor
vehicle travel has also been reduced by the amount of 3.7 per-
cent which is close to the reduction in fuel consumption.
A similar estimate of reduction in fuel consumption is also
obtainable from the well publicized figure of 200,000 barrels
of fuel being saved every day. These estimates are based
on very gross data which may be obscuring the small effects
of lower speed, but until more accurate information becomes
available we must conclude that the lower speed limits have
had a small relative effect on fuel consumption.
4.2 SAFETY
The lower speed limit went into effect for the primary
purpose of saving fuel and little or no major consideration
was given to the substantial safety results which could be
achieved by implementing a lower speed limit. The effect
that the travel speed has on both the frequency and the
severity of highway accidents has been the subject of many
research studies. The general conclusions of these studies
have been that the probability of a collision is greater for
any vehicle which travels at speeds increasingly different
from the average speed of all traffic and that the severity
of accidents increases as speed increases.
4-1
-------
More recently, many European countries have adopted
some type of speed limits outside urban areas and reports
are becoming available on the positive effects that this
reduction in speed is having on safety. In Germany the
results show that in the month of January 1974, after
adjustment for traffic volumes, that country experienced
a 61 percent reduction in both fatalities and injuries.
Denmark experienced a SO percent reduction in fatalities
and a 40 percent reduction in both injuries and accidents
for the month of November. The month of December showed
a preliminary reduction of 30 percent for all three. France
shows a 23 percent reduction in fatalities and a 12 percent
reduction in injuries. During the 4 month period of
December 1973 through March 1974, Luxemburg experienced a
slight increase in number of accidents but a 40 percent
reduction in fatalities. Only one fatal accident was
recorded during the first 2 months, but there was no notice-
able reduction in the remaining two. These and similar
results have induced the European Council of Ministers of
Transport to conclude that the new speed limits have led to
a reduction in the number and seriousness of road accidents,
and to consider that it is unwarranted, from a road safety
standpoint, to reintroduce freedom of speed on the road.
Some of these reductions in fatalities are much larger
than the ones found in the United States but we must be aware
that in most cases these European countries went from unlimited
speed to a reasonable speed limit, while the United States
has lowered the speed limit by approximately 10 mph on a small
portion of the highway system which already was controlled by
reasonable speed limits.
Nevertheless, the results experienced in this country
during the energy crisis period are just as significant as
the ones reported in Europe. Since the adoption of the 55 mph
speed limit, various analytical efforts have been initiated
4-2
-------
to assess the contribution of speed to the improved safety
level. It is still too early to arrive at definite conclu-
sions as to the true benefits of lower speed limit, however,
it is undeniable that, whatever changes have occurred, the
results are that approximately 6000 lives have been saved,
representing a 25 percent reduction in the number of high-
way fatalities over the first 6 months of 1974 when compared
to the same period in 1973. This represents a potential
saving of 12,000 lives per year (if the reduction is main-
tained.
Figures 5-7, which cover the period from July 1973
through March 19 74, provide a summary of the level of reduc-
tion experienced in travel, fatality and fatality rate.
These figures show that:
1. There has been a reduction in travel,
2. The percent reduction in travel is about the same
in both rural and urban areas,
3. The percent reduction in fatalities is much greater
than the reduction in travel,
4. The fatality rate (number of fatalities per 100
million vehicle miles of travel) has shown a per-
cent reduction similar to the percent reduction in
fatalities.
It becomes aparent from Figures 5-7 that reduction in
travel accounts for less than half of the reduction in
fatalities; the remainder must be associated with other
changes in type and condition of travel. Some of these
changes are the reduction in speed limit, more uniform flow
of high speed travel, reduction in the number of high risk
recreational trips, vehicle occupancy, changes in driver
attitude, etc. To isolate the effect that lower speed
limits alone have had on road safety may not be possible,
but a conservative estimate can be obtained by assuming
4-3
-------
JULY AUG SEPT OCT NOV DEC JAN FEB
1973 1974
Figure 5. Fatality Statistics in Urban Areas
4-4
-------
Figure
6. Fatality Statistics in
4-5
Rural Areas
-------
+ 10 -
+ 5 -
O
z
<
o
<
u
oe
0 -
-5 -
-10 -
-15 -
-20 -
-25
TRAVEL
FATALITIES
FATALITY RATE
I
NOV
~T~
DEC
JULY
1973
AUG
SEPT
OCT
JAN
1974
"T-
FEB
Figure 7. Fatality Statistics in All Areas
4-6
-------
(1) that lower speeds have had no effect on the frequency of
road accidents and (2) that this is the only change which has
affected the severity of accidents that do occur. With this
background, an analysis was undertaken of the accident data
available from seven states covering a period of 3 months
after the adoption of the lower speed limit. The seven
states which published monthly accident summary reports for
the minimum period of 3 months are Delaware, Georgia, Idaho,
Kentucky, North Carolina, South Carolina and Texas. The
month during which a lower speed limit was implemented
differed among these states but the analysis for each state
began with the month and included the next two months and
compared them to the corresponding months from the previous
year.
The results show that these states experienced a 13.6
percent reduction in the total number of road accidents.
Based on assumption (1), the reduction can be entirely
attributed to changes in both quantity and type of exposure.
All other things being equal, it would be expected that this
13.6 percent reduction in accidents would result in a similar
reduction (13.6 percent) in all types of injury.
Assumption (2) implies that the effect of lower speeds
is to lower the severity level of the remaining accidents;
it would be simple, but incorrect, to take the difference
between the actual and the expected (13.6 percent) reduc-
tion for each injury class and attribute it to the effect of
lower speeds on that injury class. A more plausible result
is obtained if we hypothesize that lower speeds do not pre-
vent injuries but reduce their severity. That is, some
accidents that might have produced fatalities have now caused
serious injuries, some serious injuries (A-Inj.) have become
moderate, some moderate injuries (B-Inj.) have become minor,
and some minor injuries (C-Inj.) have been prevented.
4-7
-------
Using this hypothesis to estimate the reduction in each
of the injury levels due to lower speeds, an analysis of the
accident data from the seven states produced the following
results:
• Reduction in quantity and type of exposure
accounted for 13.6 percent reduction in accidents,
fatalities, and injuries at each severity level.
• Lower speed may account for an additional reduction
of 12 percent in fatalities, 20 percent in serious
injuries, 5 percent in moderate injury, and less
than 1 percent in minor injuries.
The same type of analysis was conducted on the set of
multiple vehicle, single vehicle, and pedestrian accidents
to investigate whether the energy crisis is having a different
effect on these different types of accidents. Results for
each case are presented in a tabular form in Table 7.
The three major findings of this analysis were:
• The reduction in number of accidents was essentially
the same for all three types (13.8 percent for multiple
vehicle; 14.1 percent for single vehicle; and 12.9
percent for pedestrian).
• The severity of single vehicle accidents was not
affected by lower speeds.
• The severity of multi-vehicle accidents was reduced
by more than 20 percent due to lower speeds.
There is no available explanation for the last finding
and additional efforts will be required to both verify and
interpret the results.
In summary, the tentative conclusions which could be
drawn from the analysis are that changes in travel exposure
have uniformly reduced accidents and associated injuries by
approximately 14 percent; that lower speeds are having a
4-8
-------
large effect in the additional reduction of 12 percent in
fatalities and 20 percent in serious injuries; and that
the reduced injury severity occurred primarily in multiple
vehicle crashes.
TABLE 7. HIGHWAY ACCIDENT STATISTICS FOR SEVEN STATES
(Delaware, Georgia, Idaho, Kentucky, No.
Carolina, So. Carolina, and Texas)
MULTIPLE VEHICLE ACCIDENTS
Expected
(E)
(A)
(E - A)
(R)
fR/ED
Prior
(13.8%)
Expected
Actual
Reduction
Year
Reduct i on
Totals
Totals
Di f fe rence
Reduction
(Percent]
Fatalities
967
133
834
643
191
191
22.9
A-Injuries
10 ,405
1 ,43S
8,970
6,761
2 ,209
2 ,400
26.8
B-Injuries
14,219
1,961
12,258
13,106
-848
1,552
12.7
C-Injuries
17.386
2.39 8
14.988
15 .406
-418
1 .134
7.6
Accidents
147 759
127! 3-B5
SINGLE VEHICLE ACCIDENTS
Prior
Year
Expected
(14.lt)
Reduct ion
(E)
Expected
Totals
(A)
Actual
Totals
(E - A)
Difference
(R)
Reduct ion
(R/E)
Reduction
(Percent)
Fatalities
687
97
590
621
-31
-31
-5.3
A- Injuries
6,225
879
5,346
4,754
592
592
11.1
B- Injuries
6,858
968
5,890
6,885
-995
-40 3
-6.9
C-Injuries
3 .910
552
3,358
3,668
-310
-310
-9.2
Accidents
37 719
32,192
PEDESTRIAN ACCIDENTS
Prior
Year
Expected
(12.9%)
Reduction
(E)
Expected
Totals
(A)
Actual
Totals
(E - A1
Difference
(R)
Reduction
(R/E)
Reduction
(Percent)
Fatalities
370
48
322
261
61
61
19.6
A- Injuries
951
123
828
723
105
166
20.0
B- Injuries
957
124
833
981
-148
18
2.2
C-Injuries
563
73
490
494
-4
14
2.8
Accidents
" 2,(573
2,327
ALL ACCIDENTS
Prior
Year
Expected
(13.64)
Reduct ion
(E)
Expected
Totals
(A)
Actual
Totals
(E - A)
Difference
(R)
Reduction
(R/E)
Reduction
(Percent)
Fatalities
A- Inj uries
B- Injuries
C- Injuries
2,223
19,060
24,417
23 .752
302
2,580
3,310
3,220
1,921
16,480
21,107
20,532
1,69 7
13,438
23,399
21,370
224
3,042
-2,29 2
-838
224
3,266
1 ,974
136
11.6
19 .8
4.6
0.7
Acc idents
214 467
18 5 ,638
Note: See the following page (4-10) for an explanation
of the column headings.
4-9
-------
LEGEND FOR TABLE 7
Prior year - This first column contains the totals
during the based period for the specified type of
accident.
Expected reduction - The second column contains the
stated percentage of the previous column. The percent
figure represents the actual percent reduction in
total accidents and is obtained from the given totals
for all accidents.
Expected totals - This column contains the difference
between the first two columns and represents the
totals which would be found if the reduction in injuries
of various types were proportional to the reduction
in accidents.
Actual Totals - In this column the totals for the
current three months period are listed.
Difference - This column contains the values of the
difference between expected and actual totals.
Reduction - This column contains what is believed to
be the true value of the reduction experienced in
each injury class. The value for each class is
obtained by adding to the computed difference (Col. 5)
the reduction found for the next higheT injury class.
The value represents the number of injuries of a given
class, the severity of which has been lowered by one
level.
Reduction (Percent) - This last column expresses the
reduction in each class as a percent of the expected
totals. Negative values in this column reflect the
fact that the reduction in injuries was lower than
the reduction in accidents, while positive values
reflect the additional reduction which is tentatively
associated with lower speeds.
4-10
------- |