Federal Motor Vehicle Safety Standards; Fuel System Integrity

DEPARTMENT OF TRANSPORTATION

National Highway Traffic Safety Administration

49 CFR Part 571

[Docket No. 92-66; Notice 3]

RIN 2127-AF36

Federal Motor Vehicle Safety Standards; Fuel System Integrity

AGENCY: National Highway Traffic Safety Administration (NHTSA),

Department of Transportation (DOT).

ACTION: Advance notice of proposed rulemaking (ANPRM).

-----------------------------------------------------------------------

SUMMARY: This notice announces the agency's plans to consider upgrading

Federal Motor Vehicle Safety Standard (FMVSS) No. 301, Fuel System

Integrity, by making the current crash requirements more stringent and

by broadening the standard's focus to include mitigation concepts

related to fuel system components and environmental and aging tests

related to components. This notice requests comments on the agency's

plans to explore a three-phase approach to upgrading the standard. The

notice also requests data, methods, and strategies, which may assist in

the agency's regulatory decisions in defining specific requirements and

test procedures for upgrading the standard.

DATES: Comments must be received on or before June 12, 1995.

ADDRESSES: Comments should refer to the docket and notice numbers above

and be submitted to: Docket Section, National Highway Traffic Safety

Administration, 400 Seventh Street SW., Washington, D.C. 20590. Docket

hours are 9:30 a.m. to 4 p.m., Monday through Friday.

FOR FURTHER INFORMATION CONTACT: Dr. William J.J. Liu, Office of

Vehicle Safety Standards, National Highway Traffic Safety

Administration, 400 Seventh Street SW., Washington, D.C. 20590.

Telephone: (202) 366-2264. [[Page 18567]]

SUPPLEMENTARY INFORMATION:

Introduction

The National Highway Traffic Safety Administration (NHTSA) is

announcing its plans to consider upgrading Federal Motor Vehicle Safety

Standard (FMVSS) No. 301, Fuel System Integrity. The purpose of this

rulemaking is to further reduce fatalities and injuries from fires

resulting from motor vehicle crashes. Specifically, the agency is

considering whether to make more stringent the current crash

requirements applicable to vehicles with a gross vehicle weight rating

(GVWR) of 10,000 pounds (4,536 kg) or less. It is considering also

whether to broaden the standard's focus to include ways to prevent or

decrease the severity of vehicle fires by exploring regulations related

to fuel system components and tests of the resistance of components to

environmental and aging factors.

Today's notice outlines NHTSA's plans to explore a three-phase

approach to upgrading the standard. In Phase One, the agency would

evaluate performance criteria for components to ensure that the flow of

fuel from the tank is stopped in a crash. Phase Two would involve

defining upgraded crash test performance for frontal, side, and rear

impacts (e.g., higher test speeds, additional impact barriers, etc.).

During Phase Three, NHTSA would address the effect of environmental and

aging factors such as corrosion and vibration on components in the fuel

system.

Today's notice also summarizes issues related to vehicle fires and

discusses the agency's recent work in this area. The agency is seeking

public comment on the merits of the agency's rulemaking efforts to

explore alternative ways to upgrade the present standard. Today's

notice also supplements a previous notice published on December 14,

1992, in which the agency requested comments about making FMVSS No. 301

more stringent (57 FR 59041, Docket 92-66, Notice 1).

On December 2, 1994, Secretary of Transportation Federico Pena

announced a settlement of an investigation by NHTSA of an alleged

safety defect in certain General Motors (GM) pickup trucks with fuel

tanks mounted outside the frame rails. Under that settlement, GM will

contribute over $51.3 million for a variety of safety initiatives.

Among other things, the settlement will fund research on ways to reduce

the occurrence and effects of post-crash fires. All relevant results of

this research will be placed in the public docket for this rulemaking.

The Fire Problem

While vehicle fires are relatively rare events (occurring in only

one percent of towed vehicles in crashes), they tend to be severe in

terms of casualties. The agency's General Estimates System (GES)

reports that, in 1992, approximately 21,000 passenger cars, light

trucks, and multipurpose vehicles had a fire related to a crash. Based

on an analysis of the agency's Fatal Accident Reporting System (FARS),

four to five percent of occupant fatalities occur in crashes involving

fire (the fatality being due to burns and/or impact injuries). Overall,

the fire itself is deemed to be the most harmful event in the vehicle

for about one-third of these fatalities.

An analysis of 1979-1986 National Accident Sampling System (NASS)

data (Reference: ``Fires and Burns in Towed Light Passenger Vehicles,''

Docket No. 92-66-N01-001) shows that about 29,000 occupants per year

were exposed to fire in towed light passenger vehicles (cars, light

trucks, and multipurpose vehicles), of whom three percent received

second or third degree burns over at least six percent of the body. The

Abbreviated Injury Scale (AIS) defines these burns as moderate and more

severe (AIS 2 and greater). Half of those with moderate and more severe

burns had second or third degree burns over more than ninety percent of

the body; these maximum-severity (AIS 6) burns are always fatal. These

estimates are based on all 47 occupants with moderate and more severe

burns received in vehicle fires that were investigated as part of the

NASS during the eight years from 1979 to 1986.

NASS investigated vehicle fires that involved another 44 occupants

with moderate and more severe burns between 1988 and 1990. The eleven

years of NASS data suggest that each year 280 surviving occupants and

725 fatally-injured occupants received moderate or more severe burns

(AIS 2 or greater). These injuries and fatalities may have been caused

by burns or impacts.

NASS 1988 to 1990 data also indicate that potential escape from the

fire was made more difficult for most occupants (87 percent) with

moderate or more serious burns because they (1) were sitting next to a

door that was jammed shut by crash forces, (2) did not have a door at

their position, or (3) had a part of their body physically restrained

by deformed vehicle structure.

Federal Motor Vehicle Safety Standard No. 301

FMVSS No. 301, Fuel System Integrity, first became effective for

passenger cars in 1968. The requirements in the current standard apply

to all vehicles with a Gross Vehicle Weight Rating (GVWR) of 10,000

pounds (4,536 kg) or less since September 1, 1977, and to school buses

that have a GVWR greater than 10,000 pounds (4,536 kg) GVWR since April

1, 1977. FMVSS No. 301 only applies to vehicles that use fuel with a

boiling point above 32 degrees Fahrenheit (0 degree Celsius).

FMVSS No. 301 limits the amount of fuel spillage from fuel systems

of vehicles tested under the procedures specified in the standard

during and after specified front, rear, and lateral barrier impact

tests. The standard limits fuel spillage due to these required impact

tests to 1 ounce (28.4 grams) by weight during the time from the start

of the impact until motion of the vehicle has stopped and to a total of

5 ounces (142 grams) by weight in the 5-minute period after the stop.

For the subsequent 25-minute period, fuel spillage during any 1-minute

interval is limited to 1 ounce (28.4 grams) by weight. Similar fuel

spillage limits are required for the standard's static rollover test

procedure, which is conducted after the front, rear and lateral impact

tests.

The required impact tests for all vehicles that have a GVWR of

10,000 pounds (4,536 kg) or less are: a 30 mph (48.3 kmph) frontal

fixed rigid barrier impact with the barrier face perpendicular to the

line of travel of the vehicle or at any angle up to 30 degrees from the

perpendicular; a 30 mph (48.3 kmph) rear moving flat rigid barrier

impact with the barrier face perpendicular to the longitudinal axis of

the vehicle; and a 20 mph (32.2 kmph) lateral moving flat rigid barrier

impact in a direction perpendicular to the longitudinal axis of the

vehicle (i.e., with the barrier face parallel to the longitudinal axis

of the vehicle). The weight of the moving barrier is 4,000 pounds

(1,814 kg). A rollover test is conducted following the barrier impacts.

The required impact test for large school buses that have a GVWR

greater than 10,000 pounds (4,536 kg) is a 30 mph (48.3 kmph) moving

contoured rigid barrier impact at any point and angle. The weight of

the barrier is 4,000 pounds (1,814 kg). The static rollover test is not

required for large school buses.

The standard does not apply to large non-school buses or other

vehicles that [[Page 18568]] have a GVWR greater than 10,000 pounds

(4,536 kg).

December 14, 1992 Notice

On December 14, 1992, NHTSA published a Request for Comments notice

in the Federal Register (57 FR 59041, Docket No. 92-66, Notice 1)

stating that the agency ``is considering initiating rulemaking to

upgrade the protection currently provided by'' FMVSS No. 301. The

notice also requested answers to specific questions related to test

impact speeds, impact barriers, effect of vehicle aging on the

likelihood of fire, contribution of occupant entrapment to the

likelihood of fire-related injuries, etc.

NHTSA received 35 public comments by October 1994 including

comments from most of the major vehicle manufacturers, the American

Automobile Manufacturers Association (AAMA), Advocates for Highway and

Auto Safety (Advocates), the Center for Auto Safety (CAS), and the

Insurance Institute for Highway Safety (IIHS). Commenters raised issues

regarding the safety need, the adequacy of the current test procedures,

the availability and necessity of developing new test procedures, and

the existence and feasibility of countermeasures. Many commenters

stressed the need for further detailed investigation of real-world

crash data to determine the causes of vehicle fires and fire-related

occupant fatalities and injuries. In addition to support for the test

procedures currently used in FMVSS No. 301, commenters suggested

several alternatives, including substituting the dynamic side-impact

test procedures of FMVSS No. 214 for those currently specified in FMVSS

No. 301, adding frontal offset crash conditions, and developing new

barriers that might be more representative of real-world crash

conditions.

The agency has initiated work related to several fire safety issues

that need to be considered to define mitigation concepts to reduce

fatalities and injuries. Due to resource considerations, not all the

safety issues discussed in the previous notice are included in this

notice. The issues discussed in this ANPRM include crash conditions,

origin of fires, and vehicle age.

Agency Efforts Related to Fuel System Integrity

NHTSA has undertaken the following activities to more-fully

understand motor vehicle fires. These include comparing fuel system

safety requirements in this country with those in other countries,

conducting extensive test crashes related to fuel system integrity, and

analyzing data of real-world crashes.

Comparison of U.S. and Foreign Fuel System Safety Requirements

FMVSS No. 301's requirements have been compared to the following

foreign fuel system integrity standards: (1) The Canadian CMVSS No.

301, Fuel System Integrity (Gasoline, Diesel); (2) the Economic

Commission for Europe (ECE) Regulation No. 34, Uniform Provisions

Concerning the Approval of Vehicles with Regard to the Prevention of

Fire Risks (01 Series, Amendment 1, January 29, 1979) (Thirteen

European countries have agreed to adopt ECE Reg. No. 34, including

Germany, France, Italy, Netherlands, Sweden, Belgium, Czechoslovakia,

United Kingdom, Luxembourg, Norway, Finland, Denmark, and Romania); and

(3) the Japanese Standard, Technical Standard for Fuel Leakage in

Collision etc. (Amended on August 1, 1989).

The Canadian CMVSS No. 301 has requirements identical to those of

the U.S. FMVSS No. 301.

In terms of application to vehicles: FMVSS No. 301 applies to all

vehicles 10,000 pounds (4,536 kg) or less GVWR and school buses over

10,000 pounds (4,536 kg) GVWR. ECE Reg. No. 34 only applies to

passenger cars, and the Japanese standard applies to passenger cars and

multipurpose passenger vehicles 5,600 pounds (2,540 kg) or less.

In terms of required impact tests: As described above, FMVSS No.

301 requires frontal, rear and side impact tests at 30, 30, and 20 mph

(48.3, 48.3 and 32.2 kmph), respectively, plus a static rollover test,

for vehicles 10,000 pounds (4,536 kg) or less GVWR. FMVSS No. 301 also

requires a 30 mph (48.3 kmph) impact test for school buses over 10,000

pounds (4,536 kg) GVWR.

The ECE Reg. No. 34 requires a 48.3 to 53.1 kmph frontal fixed

barrier impact test and a 35 to 38 kmph rear moving flat barrier impact

test. The flat rigid barrier weighs 1,100+20 kg. A pendulum can be used

as the impactor. ECE Reg. No. 34 does not require a rollover test. The

standard requires a hydraulic internal-pressure test for all fuel tanks

and special tests (impact resistance, mechanical strength, and fire

resistance) for plastic fuel tanks.

The Japanese standard requires a 50+2 kmph frontal fixed barrier

impact test and a 35 to 38 kmph rear moving flat barrier impact test.

The flat rigid barrier weighs 1,100+20 kg. A pendulum can be used as

the impactor.

In terms of test performance requirements: all three standards

limit fuel spillage. As in FMVSS No. 301, the ECE Reg. No. 34 and the

Japanese standard, in general, also limit fuel spillage to about 1

ounce/min (28.4 grams/min). The Japanese standard lists the ECE Reg.

No. 34 and FMVSS No. 301 as examples of equivalent standards.

In summary, FMVSS No. 301 applies to more vehicle classes and to

higher vehicle weights than the ECE Reg. No. 34 or the Japanese

standard. FMVSS No. 301 requires testing in all crash modes (frontal,

side, rear, and rollover). ECE Reg. No. 34 and the Japanese standard

require only frontal and rear impact tests. FMVSS No. 301 uses a much

heavier moving barrier for impact tests than the ECE and Japanese

standards (1,814 kg vs. 1,100 kg). However, FMVSS No. 301 does not

require a hydraulic pressure test for fuel tanks, a battery retention

requirement, or additional tests for plastic fuel tanks; ECE Reg. No.

34 does. In addition, the ECE Reg. No. 34 requires that ``no fire

maintained by the fuel shall occur'' and no failure of the battery

securing device due to the impact. Since ECE Reg. No. 34 also requires

filling the impacted vehicle's fuel tank ``either with fuel or with a

non-inflammable liquid,'' the no-fire requirement is actually

interpreted from the observed fuel leakage. It is the agency's

understanding that in practice, when the ECE Reg. No. 34 tests are

conducted, the fuel tank is filled with non-inflammable liquid.

Safety Issues Related to Vehicle Fires

A. Crash Conditions

The crash conditions discussed in this section refer to real-world

crash conditions that result in vehicle fires and their implications

for compliance test conditions and performance requirements for the

current FMVSS No. 301. To further refine the relationship between real-

world and laboratory crash conditions, this notice has examined certain

engineering parameters such as impact speeds, impact locations, objects

struck, and damage patterns.

Laboratory Crash Test Results

Between 1968 and 1994, the agency has conducted 563 FMVSS No. 301

compliance tests in the frontal impact mode: 14 failures resulted (3%),

the last occurring in 1992. Effective September 1, 1976, the standard

was amended by requiring rear impact tests for all vehicles and side-

impact tests for passenger cars only. Side-impact testing was extended

to all vehicles and became effective on September 1, 1977. For model

years 1977 through 1994, 331 rear impact and 25 side-impact compliance

tests have been conducted; 26 rear impact failures (8%) and 1 side

[[Page 18569]] impact failure (4%) resulted. In computing these failure

rates, the rollover test is considered a part of the frontal, rear, or

side impact test.

The agency conducted a research test program on FMVSS No. 214, Side

Impact Protection, for light trucks. Since December 1988, 24 crash

tests have been conducted, 2 tests produced fuel leakage at a rate

higher than FMVSS No. 301 requirements. Both tests used the FMVSS No.

214 test protocol.

Between 1979 and 1986, 12 out of 201 (6%) frontal New Car

Assessment Program (NCAP) tests indicated leakage at a rate above the

fuel spillage requirements of FMVSS No. 301 at 35 mph (56.3 kmph). In

addition, during the same period, NCAP conducted 53 FMVSS No. 301 rear

impact tests at 35 mph (56.3 kmph), and 6 (11%) leaked at a rate above

the fuel spillage requirements of the standard. Rollover tests were not

conducted following any of the frontal or rear impact NCAP tests. Some

of these vehicles were retested at 30 mph (48.3 kmph), but none failed.

In 1993, NCAP resumed examining FMVSS No. 301 fuel spillage

requirements, and added a rollover test following the frontal impact

tests. To date, only one of the approximately 80 vehicles tested leaked

at a rate above the requirements of the standard at the higher speed.

Between April and June 1993, the agency conducted six baseline

vehicle crash tests (all 1993 models) as part of its initial research

effort for exploring potential upgrades to FMVSS No. 301. In addition,

the Federal Highway Administration (FHWA) conducted a seventh crash

test for the agency. Information on the seven tests has been entered

into the docket.

The test conditions for the seven crash tests represent a baseline

of delta-v (change of velocities), impact barrier, and impact location.

The tested cars were chosen based on their high sales volume as well as

agency experience with the cars in other test programs.

The six NHTSA tests include two in each of the crash modes:

frontal, side, and rear. Three tests used a 4,000-pound (1,814-kg)

moving contoured barrier--a frontal impact into a Chevrolet Corsica at

65 kmph (40.5 mph), a side impact into a Toyota Corolla at 49.4 kmph

(30.7 mph), and a rear impact into a Ford Escort at 56.6 kmph (35.2

mph). None of these three tests resulted in a loss of fuel system

integrity.

The other three tests were: a frontal impact of a Chevrolet Corsica

into a 305-mm (12-inch) diameter stationary pole at 56.3 kmph (35 mph),

a side impact into a Toyota Corolla with a 1,361-kg (3,000-pound)

deformable moving barrier (FMVSS No. 214 side impact barrier) at 87.1

kmph (54.1 mph), and an offset rear impact into a Ford Mustang with the

same type of FMVSS No. 214 moving barrier at 84 kmph (52.2 mph).

The only fuel system failure was a ruptured fuel tank from the rear

impact to the Ford Mustang by the FMVSS No. 214 deformable moving

barrier, resulting in a delta-v of about 39 kmph (24 mph). The head and

chest injury measurements on the instrumented driver and passenger

dummies exceeded the criteria specified in FMVSS No. 208, Occupant

Crash Protection. Thus, the survivability of this crash in the absence

of a fire is questionable. However, the agency would like to point out

that FMVSS No. 208 is for frontal tests and the test dummies used for

the tests were not specifically designed to collect impact data for

rear impact tests.

The crash test conducted by FHWA was on a Toyota Corolla, which was

crashed into a 203-mm (8-inch) diameter stationary pole directed at the

fuel tank location, in a side impact orientation at 32.2 kmph (20 mph).

There was no fuel system integrity failure. No dummy instrumentation

was used in this test.

The agency also conducted other frontal impact tests. These tests

primarily consisted of high speed, vehicle-to-vehicle offset crashes.

In addition, several side impact tests were conducted using the FMVSS

No. 214 test procedure. Since December 1990, a total of 25 crash tests

have been conducted. One test, involving a Chevrolet Corsica, resulted

in a small fuel leak from the fuel return line (within FMVSS No. 301's

limit). This test was conducted in an oblique configuration with a

Honda Accord striking the left front corner of the Corsica.

At the request of NHTSA's Office of Defects Investigation (ODI),

the Vehicle Research Test Center (VRTC) conducted 24 side-impact crash

tests (including one test with no instrumentation to determine

appropriate test speed) of the 1973-1987 General Motors full-size

pickup trucks and peer pickup trucks of the same vintage. These tests

were conducted as a part of a safety defect investigation, EA 92-041.

Seven of these tests were FMVSS No. 301 type side impact tests, three

were FMVSS No. 214 moving deformable barrier tests, three were vehicle-

to-pole side impact tests, and eleven were various vehicle-to-pickup

side impact tests. Reports of these tests are included in the public

file for EA92-041.

The summary report for this test program notes that the FMVSS No.

301 type tests produced no leaks in a test of a new replacement fuel

tank; however, one of the four GM trucks tested with ``as received'' GM

tanks leaked an amount in excess of the FMVSS No. 301 requirements in a

rusty area. Non-tank components of one Ford and one GM truck did leak

during the static rollover test.

In the three GM truck tests using the FMVSS No. 214 barrier, one at

53.1 kmph (33 mph) and two at 72.4 kmph (45 mph), one caused a leak in

the seam of the tank which resulted in a damp area, while the other two

did not leak.

In the vehicle-to-vehicle tests, the ride height of the striking

vehicle was adjusted to simulate heavy braking. At 72.4 kmph (45 mph)

with a Taurus striking car, the GM fuel tank significantly leaked at

the sending unit, filler nose, and a rusty area and small cut in the

tank. Although no leakage was noted from the fuel tank during a similar

test of a Ford F-150, significant fuel leakage was noted from the fuel

reservoir mounted on the inside of the left rail.

For the 80.5 kmph (50 mph) tests, significant leaks were noted from

the GM vehicles (in ``as received'' and new condition), but no leaks

were noted during a similar test on an F-150.

In the 96.6 kmph (60 mph) tests, both the GM and Ford F-150

vehicles leaked significant amounts, with the GM truck rupturing and

the Ford F-150 trucks being punctured, forming small holes.

One pole test was conducted at 48.3 kmph (30 mph) on a GM pickup

truck with significant vehicle damage and significant fuel leakage. In

the pole tests, at 32.2 kmph (20 mph) the GM tank leaked significantly,

but in a similar test of a Ford F-150, no leakage was observed.

Data Analysis of Real-World Crashes

Accurate data on vehicle fires are scarce, which makes it difficult

to define cause/effect relationships under all circumstances. Unlike

many other crashes, investigations of crashes involving fire are

hampered by the destruction of evidence needed for crash reconstruction

and analysis. The origin of fire in vehicle crashes needs to be

understood better to help define possible countermeasures and

performance requirements.

NHTSA has reviewed real-world crashes involving fuel system

integrity at great length. This analysis includes a review of the

National Accident Sampling System (NASS) file, a recent analysis by the

agency of the Fatal Accident Reporting System (FARS) data, a detailed

hard copy study of accident cases involving fire from NASS and

[[Page 18570]] FARS, and an analysis of State accident files.

The NASS review referenced in the December 14, 1992, Request for

Comments notice, ``Fires and Burns in Towed Light Passenger Vehicles''

(Docket No. 92-66-N01-001), noted that most fires occurred in crashes

with a delta-v of less than 32.2 kmph (20 mph). This figure is from all

fires, regardless of injury level.

When the same NASS files were analyzed for occupant burn injuries

at AIS 2 or greater, the sample size was very small, even after the

1991 data were added. The delta-v for frontal impacts resulting in fire

was estimated to be from 33.8 to 106.2 kmph (21 to 66 mph), with a 66

kmph (41 mph) median, based on 14 cases. The delta-v for side impacts

was estimated to be from 16.1 to 66 kmph (10 to 41 mph), with a 43.4

kmph (27 mph) median, based on seven cases. The delta-v for rear

impacts was to be estimated from 12.9 to 96.5 kmph (8 to 60 mph), with

a 41.8 kmph (26 mph) median, based on 11 cases.

The following are estimates of the delta-v's. For vehicle- to-

vehicle crashes, a 32.2 to 64.4 kmph (20 to 40 mph) delta-v range could

result from impact speeds in the 64.4 to 128.8 kmph (40 to 80 mph)

range for equal mass vehicles. Similarly, the same delta-v range could

be the result of other high impact speeds for crashes involving unequal

mass vehicles.

The FARS study analyzed real-world crash data related to vehicle

fires to establish which barrier design most closely replicates the

damage seen in real-world fatal crashes involving fire. Preliminary

results of the agency's FARS study indicate that the combined 1979-1992

data from FARS for light vehicles of model years 1978 and later include

9,440 vehicles with a post- crash fire, of which 2,840 were crashes

where fire was classified as the most harmful event. Of the latter

vehicles, approximately half were involved in single-vehicle crashes,

and half were in multi-vehicle crashes.

For frontal and side fatal crashes involving a fire, approximately

60 percent involved multiple vehicles, while for rear-impact crashes

involving in a fire, approximately 90 percent of the crashes involved

multiple vehicles. Narrow objects, including trees and poles, account

for approximately 40 percent of the objects struck in single vehicle

crashes resulting in a fire.

The agency recently completed a detailed hard copy study of a

sample of accident cases involving fire from NASS and FARS. The

detailed case study report has been entered into the docket of this

notice. The title of the report is: ``Fuel System Integrity Upgrade--

NASS & FARS Case Study,'' a NHTSA sponsored research study, by GESAC,

Inc., DOT Contract No. DTNH-22-92-D- 07064, March 1994.

The GESAC study selected 150 NASS cases for detailed analysis,

which were selected from recent years and involved fire with any

occupant injury of AIS 2 or greater. One of the objectives of the

analysis was to suggest a laboratory simulation for accidents that led

to vehicle fires. The suggested crash simulations include impact mode,

speed, barrier, location, and orientation.

The report presents information on a possible barrier test that

most accurately ``simulates'' crashes that resulted in ``moderate'',

``severe'', and ``very severe'' fires. A ``moderate'' fire is defined

as fire damage to between 25% and 50% of the vehicle surface, a

``severe'' fire has fire damage to between 50% and 75% of the vehicle

surface, and a ``very severe'' fire has fire damage to more than 75% of

the vehicle surface.

For this analysis, only the cases for which a simulation was

defined were included. Simulations were not defined, for example, for

cases where the fire originated outside the vehicle or where the crash

conditions were too complicated--these events included multiple

impacts, undercarriage impacts, or rollover events, etc. Based on these

criteria, there were 64 vehicles selected for simulations.

For vehicles receiving frontal damage, the report indicates that a

pole would be the most common simulation barrier type. For rear damage,

a moving deformable barrier with a partial overlap (a partial width of

the vehicle involved in the crash) was cited most often as a simulation

procedure. For side impacts, a pole impact was the most common

simulation procedure. The GESAC report also presents information on

impact speed for these simulations.

For frontal impacts, the delta-v ranged from 23 kmph to 105 kmph

(14 to 65 mph) with a 55 kmph (34 mph) medium delta-v. For rear

impacts, the delta-v ranged from 11 kmph to 73 kmph (7 to 45 mph) with

a 42 kmph (26 mph) medium delta-v. Overlap, which is defined as the

percentage of the frontal or rear width engaged in a crash, ranged from

40% to 100% for frontal crashes, with an average level of 72% overlap.

For rear crashes, the overlap ranged from 30% to 95% with an average

level of 71%. This real-world crash is similar to the Ford Mustang

test, discussed in the previous section, that resulted in a ruptured

fuel tank.

Based on these analyses, NHTSA tentatively concludes that in

developing any new performance requirements, it should consider

alternatives to the FMVSS No. 301 barriers in addition to possible

changes in impact speeds. Possible alternatives to be considered are

changes to simulate single vehicle crashes, pole tests, and offset

tests.

NHTSA also needs to consider the likelihood of an occupant

surviving the crash forces in high severity crashes that are associated

with many fire fatalities. To address this issue, the agency may have

to develop new test dummies that are capable of collecting meaningful

data at higher impact speeds and in rear impacts.

To further define crash conditions that lead to fires, NHTSA

anticipates conducting additional analysis of the FARS and NASS files,

the GESAC study, and experimental crash testing. Additional full-scale

crashes are being considered to help identify possible upgraded

performance requirements.

Response to the Request for Comments Notice

Impact Speeds

FMVSS No. 301 specifies that the frontal and rear crash tests be

conducted at 30 mph (48.3 kmph) and the lateral crash test be conducted

at 20 mph (32.2 kmph). The December 1992 notice asked about appropriate

test speeds.

In response to that notice, Advocates and CAS supported testing

with increased impact speed. Specifically, Advocates stated that impact

testing for all crash modes should be conducted at least at 56.3 kmph

(35 mph). It also stated that the current side impact 32.2 kmph (20

mph) test speed of existing FMVSS No. 301 is especially inappropriate

in light of the agency's current consideration of dynamic lateral test

regimens for light trucks. CAS stated that based on crash protection

technology in new vehicles, the standard should be amended to provide

for no fuel leakage in a 72.4 kmph (45 mph) frontal fixed barrier

crash, a 72.4 kmph side moving barrier, and a 72.4 kmph fixed rear

barrier.

In contrast, Mazda, Mitsubishi, Volkswagen (VW), Toyota, GM,

Chrysler, Mercedes-Benz, BMW, Ford Motor Company and the American

Automobile Manufacturers Association (AAMA) questioned the need for

testing at higher impact speeds or stated that more data are needed

before considering such an increase. For instance, Toyota stated that

the data and analyses on injuries and deaths from vehicle fires are

insufficient to support a compliance test requirement for higher impact

speeds. Similarly, Mercedes stated that increased impact speed as part

of a compliance test does not appear to have [[Page 18571]] great

potential for increasing real-world fire safety. AAMA stated that the

difference in impact speeds for side versus front and rear tests is

representative and reasonable.

Impact Barrier, Location, and Orientation

FMVSS No. 301 requires either fixed or moving rigid impact barriers

for the crash tests as described previously in this notice. In the

December 1992 notice, NHTSA posed several questions about the

appropriate barrier, including whether the current impact barriers

should be replaced by the moving contoured rigid barrier for testing

large school buses.

National Truck Equipment Association (NTEA), Mazda, Advocates, VW,

Toyota, AAMA, BMW, and Ford said no; and no commenter favored this

approach. NTEA objected to extending the existing contoured barrier to

other vehicles because of economic considerations. Mazda stated that

the FMVSS No. 214 barrier represents real-world crashes better than the

contoured barrier.

In the December 1992 notice, NHTSA also asked whether the current

barriers are representative of typical real-world crash situations.

While GM and BMW stated ``yes,'' Advocates, Ford, and Volvo said

``no.'' GM stated that the FMVSS No. 301 moving barrier side impact

test is an appropriate surrogate for real-world side impact

circumstances because it properly measures the fuel system performance

regardless of component location. Advocates stated that the current

perpendicular barrier crash test conditions for frontal and rear impact

tests should be replaced by offset and angle impacts. Advocates also

suggested that the current side impact test should be replaced by a

pole impact test, claiming that such a test is more representative of

real-world situations.

The December 1992 notice also asked whether all vehicles with GVWR

of 10,000 pounds (4,536 kg) or less should be subjected to the impact

test requirements for large school buses. Advocates, VW, Toyota, AAMA,

Mercedes, BMW, and Ford all opposed this approach, while no commenter

favored it. These commenters stated that the contoured barrier does not

simulate vehicles in use now.

Another question was whether the FMVSS No. 214 dynamic side impact

test should be incorporated into FMVSS No. 301, thereby replacing FMVSS

No. 301's current lateral requirements. Of the twelve commenters

responding to the question 11 answered ``yes'' (Mazda, Advocates,

Mitsubishi, VW, GM, Chrysler, AAMA, Mercedes, BMW, Ford, and Volvo).

Only Toyota said ``no.'' In general, the commenters stated that the

FMVSS No. 214 side impact test conditions are more representative of

real-world accidents than the current FMVSS No. 301 side impact test

requirements. GM and AAMA also suggested allowing the FMVSS No. 214

test as an optional test to the FMVSS No. 301 side impact test. In

contrast, Toyota stated that available accident data do not demonstrate

the need to replace the FMVSS No. 301 test with the FMVSS No. 214 test.

B. Origin of Fires

The origin of fire in vehicle crashes needs to be understood better

to help define possible countermeasures and performance requirements.

The agency's NASS collects information on the origin of fires in

towed light vehicles. NASS classifies fires as either minor or major.

Fires were classified as major if they involved the whole passenger

compartment or several different compartments such as the engine

compartment, trunk compartment, undercarriage, etc. Approximately 65

percent of crash-induced light vehicle ``major'' fires began in the

engine compartment, 28 percent began in the fuel tank or another part

of the fuel system, which includes the fuel supply lines, vent lines,

and tank filler neck, and seven percent others.

A recently published British article also concluded that the engine

compartment was the most common source of fires. This was attributed to

the varied electrical and mechanical systems. The article stated that:

``Investigators found that a disproportionately high number of crash/

collision fires start in cars built after 1985--especially where the

vehicles are fitted with a fuel-injection system. The investigations

also showed that fuel line integrity was more at risk from heat and

fire than from impact damage.'' (Ref: ``CACFOA Urges Action by Car

Manufacturers on Fire Risks,'' Fire Prevention, October 1992.)

C. Vehicle Age and Fires

Both the FMVSS No. 301 evaluation report referenced in the December

14, 1992, Request for Comments notice and more recent analysis of real-

world crash results indicate that older vehicles involved in crashes

represent a disproportionate number of cases in which there was a fire

compared to newer crash vehicles. The agency's FARS analysis showed

that vehicle age has a statistically significant relationship to fire

in fatal crashes. The agency is conducting an extensive statistical

analysis of fire occurrence in fatal and other crashes, as a function

of the factors that may influence the likelihood of post-collision

vehicle fires. Fire occurrence in FARS was examined in fatal crashes

with any occurrence of a fire and in those crashes for which the fire

was the ``Most Harmful Event.'' Preliminary results indicate that as

vehicles (especially passenger cars) age, the likelihood of a fatal

fire increases. The preliminary findings also indicate that while

trucks involved in fatal crashes have a somewhat higher rate of fire

occurrence than cars, there is not an increase in the likelihood of

fire as light trucks age.

Preliminary findings indicate that for cars, light trucks, and vans

as a group and with all other factors held constant, a vehicle that is

ten years older than another is on average, 29.3 percent more likely to

be involved in a fatal fire. Most of this increase is found in cars.

Although there is an indication that as light trucks and vans age the

probability of a fire increases in fatal crashes, the estimated

increase is less than the increase for cars only. However, the number

of cases in the current data base is insufficient to produce

statistically significant results using vehicle age as a variable.

The combined data for cars, light trucks, and vans do not suggest

any relationship between vehicle age and likelihood of involvement in a

fatal crash where the most harmful event is fire. Nevertheless, post-

crash fires should be avoided to the extent practicable. The possible

effect of vehicle aging, therefore may need to be addressed in an

upgrade of FMVSS No. 301.

To address the problems associated with older vehicles,

requirements may need to address such factors as corrosion, stress

cracking, fatigue, and mechanical damage. Various aging tests are

available, such as the Salt Spray (Fog) Test (ASTM B117), Humidity

Test, Laboratory Cyclic Testing and Electrochemical Testing to simulate

corrosive environments. However, if the problem of aging in relation to

fuel system leakage and fires were attributed to cracking of fuel

hoses, etc. then there are other options. Standards with performance

requirements for aging of fuel lines and tanks may be one approach to

mitigating this problem.

A question related to this subject was posed in the December 1992

notice. Eight commenters did not support setting up an aging test

standard within FMVSS No. 301 (Mazda, Mitsubishi, Toyota, GM, AAMA,

Mercedes, BMW, and Ford). Advocates and Volvo [[Page 18572]] supported

a component test procedure for aging. VW opposed aging tests on a total

vehicle basis but not for components.

Mitsubishi indicated that the design of various replacement parts,

their materials and conditions of use and exposure will all vary, and

it is not practical to set up a standard specifying time or mileage

limits for each part. BMW stated that age-related degradation can occur

not only in fuel system components, but also in other parts,

components, and structures and could be a significant factor related to

degradation, along with differences in vehicle use, operational and

environmental conditions and maintenance.

Mazda, VW, and Volvo recommended periodic inspection or replacement

of certain fuel system components. Mazda recommended it be performed by

the vehicle owner and VW suggested upgraded periodic inspections for

vehicle condition be performed under local or state government

programs. Mazda also stated that, in the long term, durability testing

of critical fuel system components may be advisable.

Advocates strongly supported simulation of fuel system component

deterioration and overall system performance loss due to aging effects.

Advocates suggested utilizing test standards to detect the deleterious

effects of aging and/or exposure to operating or environmental

conditions that degrade fuel system integrity.

The agency requests specific comments on the wisdom and

practicability of adopting existing test procedures or developing new

component test procedures related to aging effects. Individual fuel

system components could be evaluated using accelerated aging or

corrosion treatment tests.

Phased Rulemaking Approach

Based on the above discussions and preliminary analyses, the agency

is considering research and rulemaking activities to amend FMVSS No.

301 to address the following areas:

1. The definition of performance criteria for fuel system

components directed at mitigating the cause and spread of vehicle

fires.

2. The modification of the existing FMVSS No. 301 crash test

procedures and performance criteria to better simulate the events that

lead to serious injury and fatalities in fires.

3. The definition of the role of environmental and aging factors

such as corrosion and vibration as it affects fuel system integrity,

and, if appropriate, the specification of performance criteria related

to this area.

The agency is considering whether to initiate rulemaking using a

phased approach. The basis of this approach lies in the varying

complexity of addressing the different issues listed above. The initial

phase would focus on requirements for component performance, the second

phase would address system performance, and the third phase would deal

with issues related to environmental and aging effects.

Phase 1: Component Level Performance

A. Objectives of Component Approach

The first phase would focus on the specification of performance

criteria, at a component level, to attempt to ensure that the flow of

fuel from the fuel tank or fuel lines will stop in a crash. It would

also focus on minimizing the possibility of an electrical spark of

sufficient intensity to act as an ignition source. These specifications

would primarily affect fires that originate in the engine compartment.

However, they would also help to shut off the fuel flow for all crash

modes, including a rollover crash.

Shutting off the fuel flow quickly during or immediately after a

crash will eliminate a major fire and fuel source and therefore should

both reduce fire incidents and limit the spread of fire, if one were to

start. It also appears that many new vehicles incorporate different

techniques for addressing this problem. An electric current shut-off

device would minimize the possibility of a spark. The performance

associated with the fuel shut-off and the electric current shut-off

devices can be incorporated into the present crash tests in FMVSS No.

301 or other compliance tests such as those conducted as part of FMVSS

No. 214.

As discussed below, the agency is also seeking comment about

component test requirements for fuel tanks, fuel pumps, the vehicle's

electrical system, and engine fire extinguishes.

The agency requests information on the performance, cost, and

practicability aspects of various systems in shutting off the fuel flow

and the electric power. The agency also requests comments on ways to

develop a practicable test procedure and to define specific criteria

with sufficient objectivity that test variability is reduced to a

minimum. In the event that other, more appropriate, component tests

would satisfy the objectives of the Phase 1 effort, interested parties

are requested to provide this information to the agency.

B. Components Now in Use

The agency believes that technology already exists for detecting

and identifying conditions when the fuel flow should be shut off. Most

new vehicles sold in the United States are already equipped with

devices that shut off the fuel pump in any collision that causes the

engine to stop.

In some vehicles, sensors detect the consequence of severe engine

damage (rotation stops for camshaft, crankshaft or alternator) and

immediately shut off the fuel pump. Often, signals from more than one

sensor are used to determine if the engine has stopped running and the

decision for fuel pump shut-off is left up to the vehicle's onboard

computer (such as the Engine Control Unit or Electronic Control

Module). Manufacturers also use a ``central'' for collecting and

routing crash signals through a central collision detection bus.

Other vehicles are equipped with an inertia switch. Inertia

switches can be used to shut off the fuel flow as well as the electric

current. Inertia switches operate on sudden impact to open the

electrical circuit to the fuel pump or the battery during the crash. An

inertia switch can be designed to operate at various levels of impact

intensity and direction, and thus could be effective in all crash

modes.

The agency requests information on the different components used in

vehicles for shutting off the fuel flow or electric current.

C. Component Test Procedures

Fuel system components must operate in a real-world environment

surrounded by extreme conditions imposed by modern engine technology.

The materials and parts used to assemble fuel system components are

already subject to manufacturers' specifications, often derived from or

directly related to other engineering standards such as the

publications of the American Society for Testing and Materials (ASTM).

Some of the test requirements are generic to many of the ASTM

standards, for example: vibration, shock, endurance testing,

temperature cycling, temperature extremes, compatibility with other

materials, etc.

Comments are requested regarding the extent and scope of component

test requirements that should be developed as part of the FMVSS No.

301.

The agency has identified the following fuel system and vehicle

components as potential candidates for this approach:

a. Fuel tank, including filler pipe

b. Fuel pump(s) [[Page 18573]]

c. Vehicle's electrical system

d. Engine fire retardant/extinguisher

The agency has not included fuel lines in this proposed list

because the potential to shut down the entire fuel delivery system when

the fuel pump shuts down already exists. Comments are requested about

this decision.

a. Fuel tank, including filler pipe. During a vehicle crash, the

fuel tank may receive crash forces great enough to move or dislodge the

tank from its mountings and/or to rupture the tank. If the tank moves

significantly, the filler pipe, which is attached to the vehicle body

to provide access during refueling, may rupture or break away. If the

filler pipe ruptures, fuel could spill. Fuel spillage can be expected

under some crash conditions even if the fuel pump is shut off.

One concept would include a check valve located in the filler pipe

that is normally closed to prevent fuel flow but that would open

automatically during refueling. For example, inserting of the pump

filler nozzle could cause the closed check valve to open to permit fuel

flow; withdrawing the nozzle would cause the valve to close.

Another concept would use a check valve similar in function to the

valves used on heavy truck crossover fuel lines. Applied to the filler

neck, this concept would require a large valve, normally open, that

would close automatically upon detachment of the filler neck due to a

crash.

Comments are requested on how filler check valves should be

evaluated during safety compliance tests. For example:

1. Should the filler valve pass a simple go no-go test or should

the valve be subjected to many cycles of operation?

2. What test condition would be appropriate for filler check

valves: dynamic pendulum or other impact tests?

3. What are the critical engineering parameters that would

characterize the proper operation of a filler pipe check valve?

4. Are there alternative ways to control spillage from broken

filler pipes?

b. Fuel pump(s). Today's passenger cars, light trucks, and vans use

electrically operated fuel delivery pumps almost exclusively. Some

electric fuel pumps shut down if certain engine operating parameters,

such as crankshaft rotation, indicate that the engine has stopped. The

agency is interested in how manufacturers use engine sensing to control

fuel pump operation and under what conditions the fuel pump is shut

off. Specifically:

1. Is current sensing time response adequate to prevent fuel

spillage? If not, what would improve response time?

2. How does cessation of engine rotation typically relate to the

frontal crash pulse; i.e., after engine disintegration begins, how long

does it take for the rotating parts to stop?

3. During this time interval, how much fuel spillage could occur,

assuming that the crash has damaged the fuel lines, making fuel

spillage imminent?

4. How would sensing engine rotation provide benefit to vehicles

involved in a rear impact? rollover? side impact? in any crash where

engine damage may be slight?

5. With regard to vehicle rollover, would a separate rollover

switch prevent fuel spillage? Could this function be practicably

combined in a single switch that would respond to all crash modes?

6. Does fuel pump shut-off prevent gravity-induced fuel flow

through the pump?

7. Should a single fuel pump cutoff switch be used to replace the

functions currently performed by sensing engine rotational parameters?

8. What advantages/disadvantages would such an installation incur?

Some manufacturers currently use inertia switches to interrupt the flow

of electricity to the fuel pump when a crash is sensed, thereby causing

the fuel pump to shut down.

1. Could an inertial switch be substituted for the systems that

sense engine shut down to disable fuel pumping?

2. Under what conditions would such a substitution be impracticable

or too costly?

3. What sensitivity of operation should an effective inertia switch

have?

4. Can inertia switches be manufactured with sufficient durability

and reliability to function for long periods of time unattended in a

relatively harsh automotive environment?

5. Are there any other features of an inertia switch that would be

detrimental to occupant safety; e.g., what measures must an occupant

take to restart the vehicle after an inertia switch has stopped fuel

flow?

The agency is also interested if manufacturers or others have any

alternative techniques for accomplishing fuel shut-off during a crash.

c. Vehicle's electrical system. Other means exist to cause the fuel

pump to shut down in a crash. For example, a battery shut-off device

could remove all electrical power from the vehicle at the onset of a

crash. However, battery shut-off may have unintended consequences if

electrically operated door locks or windows are rendered inoperative

during a crash. Comments are requested regarding the relative costs and

practicability of battery shut-off devices.

d. Engine fire retardant/extinguisher. After ignition takes place,

vehicle fires could be controlled or extinguished if the proper

equipment were available and functioning. Examples of equipment that

could help control or extinguish a fire include an onboard fire

extinguisher mounted in the engine compartment and fire retardant

blankets. A fire extinguisher using carbon dioxide or other gaseous

mixtures could be operated by means of existing vehicle sensors (such

as the inertia switch) or by other signals. Fire retardant blankets

attached underneath the vehicle's hood could drop down onto the engine

to smother a fire in the event of a crash. Comments are requested on

the costs and practicability of these concepts.

Phase 2: System Level Performance

The second phase would focus on the process of defining upgraded

crash test performance for frontal, side, and rear impacts. The present

crash tests specified in FMVSS No. 301 require a frontal fixed barrier

impact at 30 mph (48.3 kmph), a moving barrier impact of 20 mph (32.2

kmph) into the side of a stationary vehicle, and a moving barrier

impact of 30 mph (48.3 kmph) into the rear of a stationary vehicle.

From the information discussed in this notice, it appears that the

present tests in FMVSS No. 301 may not be representative of the

severity of the crash conditions associated with fatal and severe

injury-causing fires. However, it is difficult at this time to define

specific upgrades to these crash conditions without further tests. Some

potential tests that appear promising for upgrading FMVSS No. 301 test

procedures are the offset/oblique tests in the frontal mode, the FMVSS

No. 214 offset barrier in the rear test mode and a pole impact or FMVSS

No. 214 barrier for the side impact.

As identified in the GESAC study, a key objective for such tests

may be to limit the engagement to a narrower area than engaged with

current barriers. The specific crash conditions that cause fuel system

loss of integrity must be defined, along with an understanding of which

crashes would be survivable if fire was avoided. Accident data analyses

and crash testing are being considered to further explore these issues,

which is expected to be the second phase of [[Page 18574]] rulemaking,

which may be conducted concurrently with the first phase.

The agency requests comments on the performance aspects and

practicability of this approach.

Phase 3: Environmental and Aging Effects

The third phase would explore the issue of environmental and aging

effects on vehicle condition and the possible relationship to fire

occurrence. The agency's preliminary analyses of FARS and State

accident files indicate that the likelihood of fire increases with the

age of the vehicle. The analysis also attempted to determine the

possible differences, if any, in the occurrence of fire in fatal

crashes in states that typically experience more inclement weather

(i.e., snow and ice) and as a result, use more salt and other corrosive

substances on public roadways, when compared to other states.

Passenger cars registered in the ``salt belt'' states and involved

in fatal crashes were found to have an approximately 25 percent greater

rate of fire occurrence in fatal crashes, compared with passenger cars

in fatal crashes in the ``sun belt'' states. (It should be noted that

when the fire itself was deemed to be the most harmful event in the

vehicle, the ``salt belt'' states had a lower rate compared to the

``sun belt'' states.) It is not clear at this time whether this

possible relationship between vehicle aging, weather and use of salt

and similar substances and fire occurrence may be due to environmental

characteristics, to changes in vehicle design, to differences in

operator characteristics, or a combination of these factors. If this

disparity can be attributed to environmental factors, it may be

possible to add environmental tests, such as corrosion, to FMVSS No.

301.

Further work is needed to associate vehicle fires with

environmental and aging factors and to define possible performance

tests. Because of this, the agency is considering addressing this

problem in a third phase of rulemaking.

The agency requests comments on this phased approach. This approach

may be implemented either sequentially or concurrently, depending on

the timing of the research.

Rulemaking Analyses

NHTSA has considered the impact of this rulemaking action under

Executive Order 12866 and the Department of Transportation's regulatory

policies and procedures. The agency has determined that this notice is

significant under Department's policies and procedures. The agency

notes that the increase in vehicle production costs and corresponding

increases in consumer costs that would result from upgrading the

requirements of FMVSS No. 301 would depend on the stringency and nature

of the new requirements and the extent to which present and planned new

production vehicles would already meet them, i.e., the type and extent

of vehicle changes that would be necessary. Since the agency is still

in the research and analysis phase of the rulemaking, including

assessing new vehicle hardware and fuel system crash integrity, it

cannot provide a cost estimate at this time. Nevertheless, a more

comprehensive discussion of this notice's cost impacts is discussed in

the Preliminary Regulatory Evaluation, which has been placed in the

public docket.

Submission of Comments

Interested persons are invited to submit comments on the proposal.

It is requested but not required that 10 copies be submitted. All

comments must not exceed 15 pages in length (49 CFR 553.21). Necessary

attachments may be appended to these submissions without regard to the

15-page limit. This limitation is intended to encourage commenters to

detail their primary arguments in a concise fashion.

If a commenter wishes to submit certain information under a claim

of confidentiality, three copies of the complete submission, including

purportedly confidential business information, should be submitted to

the Chief Counsel, NHTSA, at the street address given above, and seven

copies from which the purportedly confidential information has been

deleted should be submitted to the Docket Section. A request for

confidentiality should be accompanied by a cover letter setting forth

the information specified in the agency's confidential business

information regulation 49 CFR Part 512.

All comments received before the close of business on the comment

closing date indicated above will be considered, and will be available

for examination in the docket at the above address both before and

after that date. To the extent possible, comments filed after the

closing date will also be considered. Comments received too late for

consideration in regard to the ANPRM will be considered as suggestions

for further rulemaking action. Since NHTSA will continue to file

relevant information as it becomes available in the docket after the

closing date, it is recommended that interested persons continue to

examine the docket for new material.

Those persons desiring to be notified upon receipt of their

comments in the rules docket should enclose a self-addressed, stamped

postcard in the envelope with their comments. Upon receiving the

comments, the docket supervisor will return the postcard by mail.

Issued on April 6, 1995.

Barry Felrice,

Associate Administrator for Safety Performance Standard.

[FR Doc. 95-9025 Filed 4-11-95; 8:45 am]

BILLING CODE 4910-59-P