[Federal Register Volume 69, Number 72 (Wednesday, April 14, 2004)]
[Notices]
[Pages 19897-19901]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 04-8418]


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

DEPARTMENT OF TRANSPORTATION

National Highway Traffic Safety Administration

[Docket No. NHTSA 2002-12366 Notice 2]


General Motors Corporation; Ruling on Petition for Determination 
of Inconsequential Noncompliance

    General Motors Corporation (GM) has determined that the seat belt 
assemblies in approximately 1,870,000 of the company's model year (MY) 
2001-2002 vehicles fail to comply with the requirements of Federal 
Motor Vehicle Safety Standard (FMVSS) No. 209, Seat Belt Assemblies (49 
CFR 571.209). The identified noncompliance involves the emergency-
locking retractors (ELR) \1\ in the seat belt assemblies for the 
vehicles' front outboard seats. Some of the ELRs in these assemblies do 
not lock before the belt webbing extends 25 mm (1 inch) when they are 
subjected to an acceleration of 7 m/s2 (0.7 g), as required under 
paragraph S4.3(j)(1) of the standard. Pursuant to 49 CFR Part 573, GM 
filed a Noncompliance Information Report with the National Highway 
Traffic Safety Administration (NHTSA) on April 19, 2002.\2\
---------------------------------------------------------------------------

    \1\ FMVSS No. 209 defines an ``emergency-locking retractor'' as 
``a retractor incorporating adjustment hardware by means of a 
locking mechanism that is activated by vehicle acceleration, webbing 
movement relative to the vehicle, or other automatic action during 
an emergency and is capable, when locked, of withstanding restraint 
forces.'' 49 CFR 571.209, S.3.
    \2\ Although not referred to in GM's Noncompliance Information 
Report, the failure of the ELRs also constitutes a noncompliance 
with FMVSS No. 208, Occupant crash protection.
---------------------------------------------------------------------------

    In general, manufacturers of motor vehicles and replacement 
equipment are required to notify owners of, and provide a remedy for, 
noncompliances with FMVSSs. 49 U.S.C. 30118-30120. However, 49 U.S.C. 
30118(d) and 30120(h) authorize manufacturers to file petitions for an 
exemption from these notification and remedy requirements on the basis 
that the noncompliance is inconsequential to motor vehicle safety.
    GM submitted such a petition on May 3, 2002.\3\ The petition stated 
that the noncompliance occurs because the vehicle-sensitive ELR 
mechanism in a small number of seat belt assemblies can be disabled by 
atypical handling during transit from GM's safety belt supplier, TK 
Holdings, Inc. (TKH), to the seat suppliers or during installation in 
vehicle seats.
---------------------------------------------------------------------------

    \3\ GM submitted a revised petition on July 30, 2002 (Docket No. 
NHTSA-2002-12366-4), which replaced the May 3, 2002 petition (Docket 
No. NHTSA-2002-12366-3) in full. However, GM stated that the 
subsequent petition did not change the substance, rationale, basis, 
or conclusion of the original petition.
---------------------------------------------------------------------------

    The ELR in the vehicles in question incorporates two different 
types of locking mechanisms. The first is a vehicle-sensitive mechanism 
that was used to certify compliance with FMVSS No. 209, and which, when 
functioning, meets the requirements of the standard. The second locking 
mechanism is a voluntarily supplied, webbing-sensitive one that does 
not meet the requirements of the standard (although webbing-sensitive 
ELRs can be designed to comply with FMVSS No. 209). GM asserted that 
the failure of the vehicle-sensitive mechanism was inconsequential to 
safety because the webbing-sensitive system offers a level of 
protection nearly equivalent to that provided by a compliant ELR under 
the conditions that it and TKH evaluated. GM also submitted a 
calculation, based on a number of assumptions, which it asserts shows 
that less than one person would be likely to sustain a moderate to 
severe injury as a result of the noncompliance.
    The vehicles covered by the petition are all MY 2001 and most MY 
2002 C series and K series (C/K) vehicles (such as the GMC C/K pickups, 
GMC Yukon, Chevrolet C/K pickups (e.g., the Silverado), Chevrolet 
Tahoe, Chevrolet Suburban, Chevrolet Avalanche, and Cadillac Escalade), 
and GM's S series and T series (S/T) vehicles (such as the GMC Envoy, 
Chevrolet Trailblazer, and Oldsmobile Bravada). As described below, the 
webbing-sensitive mechanism in the ELRs in the C/K vehicles will lock 
up the retractor when the webbing is exposed to 2.0 g (the force of 
gravity), while the webbing-sensitive mechanism in the ELRs in the S/T 
vehicles does not lock up the retractor until the webbing experiences 
3.0 g. The C/K vehicles constitute approximately 80 percent of the 
vehicles covered by the noncompliance determination.

[[Page 19898]]

    On July 8, 2002, NHTSA published a Federal Register notice 
announcing the availability of GM's petition and affording the public a 
30-day comment period (67 FR 45179). No comments were received. GM and 
TKH met with agency staff on four separate occasions to provide and 
discuss the results of various tests they had conducted to assess the 
risk of increased injury due to the noncompliance.\4\
---------------------------------------------------------------------------

    \4\ GM requested, and was granted, confidentiality for the 
presentations made to NHTSA during these meetings. This document 
will include some general information about the test results shared 
with the agency, but will not reveal detailed information about the 
confidential materials. All non-confidential documents related to 
the inconsequentiality petition are posted in the DOT Docket 
Management System Web site at http://dms.dot.gov in Docket No. 2002-
12366.
---------------------------------------------------------------------------

    For the reasons discussed below, NHTSA has concluded that the 
noncompliance with FMVSS No. 209 in the C/K vehicles is inconsequential 
to motor vehicle safety, while the noncompliance in the S/T vehicles 
(equipped with the less sensitive webbing-based mechanism) is not. 
Accordingly, GM's petition is granted in part and denied in part.

I. ELR Requirements of FMVSS No. 209

    FMVSS No. 209 specifies certain requirements for all seat belt 
assemblies manufactured for use in passenger cars, multipurpose 
passenger vehicles, trucks and buses. Among these requirements is one 
requiring each belt assembly to have either an automatic-locking 
retractor, an ELR, or an adjusting device that is within the reach of 
the occupant (S4.1(g)(1)). However, all passenger cars and light trucks 
are equipped with ELRs pursuant to S7.1.1.3 of FMVSS No. 208, Occupant 
Crash Protection, which requires that the safety belt assemblies in all 
forward-facing, outboard designated seating positions in vehicles with 
a gross vehicle weight rating of 10,000 pounds or less be equipped with 
ELRs meeting the requirements of FMVSS No. 209.
    ELRs are designed to provide maximum freedom of movement unless the 
belted occupant is subjected to a rapid acceleration or deceleration. 
In a vehicle-sensitive ELR, the locking mechanism is activated in 
response to a rapid deceleration of the vehicle, such as results from a 
collision or sudden braking. In a webbing-sensitive ELR, the locking 
mechanism is activated based on the rate at which the occupant extracts 
webbing from the retractor housing. In many cases, vehicle 
manufacturers voluntarily equip their vehicles with ELRs that have both 
a vehicle-sensitive mechanism and a webbing-sensitive mechanism. The 
two types of mechanisms do not behave identically, with each offering 
some advantages over the other.
    FMVSS No. 209 permits both webbing-sensitive and vehicle-sensitive 
ELRs, and either type may be used for certification as long as it meets 
the conditions set forth in S4.3(j) of the standard. S4.3(j)(1) 
requires an ELR to lock before the webbing extends 25 mm (one inch) 
when the retractor is subjected to an acceleration of 0.7 g (7 meters/
second2). S4.3(j)(2) prohibits the locking of a webbing-
sensitive retractor at 0.3 g or less; and S4.3(j)(3) prohibits the 
locking of a vehicle-sensitive retractor when the retractor is rotated 
15 degrees or less from its orientation in the vehicle.
    The test procedure under which the compliance of ELRs is assessed 
is found at S5.2(j) of FMVSS No. 209. The ELR is subject to an 
acceleration of 0.7 g within a period of 50 ms while the attached belt 
webbing is extended to 75 percent of its total length. The test is 
conducted differently depending on whether the ELR is webbing-sensitive 
or vehicle-sensitive, but both types of ELRs must lock in response to 
the specified acceleration at different angles to account for various 
possible crash scenarios.
    When FMVSS No. 209 was first adopted, the standard required ELRs to 
lock when subjected to an acceleration of 0.5 g, and not to lock in an 
acceleration of 0.2 g. In 1970, the agency proposed to increase this 
0.5 g level to 2.0 g and to increase the no-lock level to 1.0 g because 
it was concerned the then-existing requirements resulted in safety 
belts that cinched up on the user to a degree that was uncomfortable, 
possibly inhibiting belt use. 35 FR 4641 (March 17, 1970). In response 
to comments that the proposed acceleration levels were too high, the 
agency decided to set the acceleration level at which locking is 
required at 0.7 g and to set the no-lock level at 0.3 g. 36 FR 4607 
(March 10, 1971). Those were the levels that GM had suggested in its 
comments on the Notice of Proposed Rulemaking.

II. The Noncompliance

    From May 2000 to May 2002, GM installed front seat belt assemblies 
in almost two million MY 2001 and 2002 C/K series and S/T series 
vehicles with ELRs manufactured by TKH that were equipped with a 
vehicle-sensitive mechanism that, when functioning, met the 
requirements of S4.3(j) of FMVSS No. 209. However, as GM subsequently 
discovered, these ELRs could be damaged during handling and 
installation. In assemblies with damaged retractors, the plastic cross 
bar at the top of the weight pendulum interferes with the ELR actuator 
and renders it inoperative. As a result, these vehicle-sensitive 
mechanisms do not function at all, and they would not lock the safety 
belt in the event of sudden vehicle deceleration or rollover. However, 
the ELRs in these vehicles were also equipped with a second, webbing-
sensitive locking mechanism, which functions as designed, 
notwithstanding the breakage of the vehicle-sensitive mechanism. This 
mechanism will limit the webbing payout of the safety belt, although 
not in precisely the same manner or under the same conditions as the 
vehicle-sensitive locking mechanism.
    The webbing-sensitive mechanisms in the ELRs installed in the C/K 
series vehicles were designed to lock up the retractor at 2.0 g, with 
the objective of meeting the requirements of the Economic Commission 
for Europe's Regulation No. 16, Uniform Provisions Concerning the 
Approval of: Safety-belts and Restraint Systems for Occupants of Power-
driven Vehicles; Vehicles Equipped with Safety-belts (ECE R16) and be 
sold in Europe.\5\ The webbing-sensitive mechanism in the ELRs 
installed in the S/T series vehicles were designed to lock up the 
retractor at 3.0 g, since these vehicles were only produced for the 
U.S. market and were not designed to meet ECE requirements. GM has not 
claimed that the webbing-sensitive mechanism would allow the ELRs on 
any of the noncompliant vehicles to meet the 0.7 g acceleration 
requirements of FMVSS No. 209.
---------------------------------------------------------------------------

    \5\ We offer no opinion as to whether these C/K vehicles would 
satisfy all of the European requirements.
---------------------------------------------------------------------------

    According to GM, the noncompliance was initially discovered by TKH 
in January 2002. Seat belt assemblies that had been shipped to a 
Belgian test facility for type approval under ECE R16 were returned to 
the manufacturer because the vehicle-sensitive locking mechanisms of 
the ELRs were broken. During inspections of completed seating units at 
seat assembly plants, TKH discovered that the vehicle-sensitive ELR 
mechanism was not functioning in a small number of seat belt 
assemblies. TKH concluded that atypical handling during transit likely 
damaged the vehicle-sensitive mechanism so that it would not function. 
To address this, on January 15, 2002, TKH initiated a 100 percent 
inspection of the seat belt assemblies upon their arrival at the seat-
manufacturing facilities.
    This inspection practice was intended to, and apparently did, 
identify failures

[[Page 19899]]

that arose during transit. However, based on its inspection of some 
seat belt assemblies after their installation in seats, TKH discovered 
that handling of the assemblies at the seat-manufacturing facilities 
during installation also could disable the vehicle-sensitive ELR 
mechanism. Consequently, during March and April of 2002, TKH initiated 
a 100 percent inspection of the safety belts in assembled seats. 
Beginning in April 2002, TKH also implemented a design change to the 
vehicle-sensitive mechanism to improve its robustness, in order to 
prevent breakage during shipping or installation. GM and TKH are 
confident that all vehicles produced after April 30, 2002 are equipped 
with belt assemblies that comply with the emergency locking 
requirements of FMVSS No. 209.
    On the basis of its inspections, TKH has estimated that the 
mishandling during transit could cause the failure of the vehicle-
sensitive mechanism in 58 out of every one million retractors and that 
mishandling during seat assembly could lead to the failure of this 
mechanism in an additional 32 out of every one million retractors.

III. GM's Petition for an Inconsequentiality Determination

    GM's petition for a determination that the noncompliance is 
inconsequential to motor vehicle safety took two separate approaches.
    First, GM submitted a ``risk analysis'' in which it estimated that 
of the approximately 3,740,000 seat belt assemblies in 1,870,000 
vehicles produced between May 2000 (i.e., the earliest vehicle 
production start date among the affected vehicles) and April 29, 2002 
(i.e., the date after which it has confidence that the noncompliance 
was eliminated), there were approximately 271 noncomplying assemblies. 
It then contended that very few occupants would actually be exposed to 
any possible increased risk due to the absence of a vehicle-sensitive 
ELR.
    Second, GM submitted the results of a series of frontal sled tests 
comparing the performance of C/K and S/T vehicles with compliant ELR 
systems to those vehicles equipped with ELRs with only a webbing-
sensitive mechanism. GM asserted that this data demonstrated that the 
webbing-sensitive locking mechanisms performed nearly identically to a 
properly functioning vehicle-sensitive ELR mechanism.
    In GM's opinion, the existence of the webbing-sensitive locking 
mechanism, combined with the very low frequency of potentially 
noncomplying retractors, renders this noncompliance inconsequential 
with respect to vehicle safety.

IV. NHTSA's Consideration of the GM Inconsequentiality Petition

A. General Principles

    Federal motor vehicle safety standards are adopted only after the 
agency has determined, following notice and comment, that the 
performance requirements are objective and practicable and ``meet the 
need for motor vehicle safety.'' See 49 U.S.C. 30111(a). Thus, there is 
a general presumption that the failure of a motor vehicle or item of 
motor vehicle equipment to comply with a FMVSS increases the risk to 
motor vehicle safety beyond the level deemed appropriate by NHTSA 
through the rulemaking process. To protect the public from such risks, 
manufacturers whose products fail to comply with a FMVSS are normally 
required to conduct a safety recall under which they must notify 
owners, purchasers, and dealers of the noncompliance and provide a 
remedy without charge. 49 U.S.C. 30118-30120.
    However, Congress has recognized that, under some limited 
circumstances, a noncompliance could be ``inconsequential'' to motor 
vehicle safety. It therefore established a procedure under which NHTSA 
may consider whether it is appropriate to exempt the manufacturer from 
the duty to conduct a notification and remedy (i.e., recall) campaign. 
49 U.S.C. 30118(d) and 30120(h). The agency's regulations governing the 
filing and consideration of petitions for inconsequentiality exemptions 
are set out at 49 CFR part 556.
    Under the statute and regulations, inconsequentiality exemptions 
may be granted only in response to the petition of a manufacturer, and 
then only after publication of a notice in the Federal Register and an 
opportunity for interested members of the public to present 
information, views, and arguments on the petition. When NHTSA does not 
receive any public comments, as in the present case, the agency will 
draw upon its own understanding of safety-related systems and its 
experience in deciding the merits of a petition. An absence of opposing 
argument and data does not require us to grant a manufacturer's 
petition.
    ``Inconsequential'' is not defined either in the statute or in 
NHTSA's regulations. Rather, the agency determines whether a particular 
non-compliance is inconsequential to motor vehicle safety based on the 
specific facts before it.
    There have been instances in the past in which NHTSA has determined 
that a manufacturer has met its burden of demonstrating that a 
noncompliance is inconsequential to safety. For example, a label 
intended to provide safety advice to an owner or occupant may have a 
misspelled word, or it may be printed in the wrong format or the wrong 
type size. If the manufacturer shows that the discrepancy with the 
safety requirement is unlikely to lead to any misunderstanding, we have 
granted an inconsequentiality exemption, especially where other sources 
of correct information are available (e.g., in the vehicle owner's 
manual). See IMPCO Technologies; Grant of Application for Decision of 
Inconsequential Noncompliance, 65 FR 14009 (March 15, 2000) (NHTSA-99-
6269-2); TRW, Inc.; Grant of Petition for Determination of 
Inconsequential Noncompliance, 58 FR 7171 (February 4, 1993).
    The burden of establishing the inconsequentiality of a failure to 
comply with a performance requirement in a standard is more substantial 
and difficult to meet, and the agency has not found many such 
noncompliances to be inconsequential. One area in which the agency has 
granted such petitions has been where the noncompliance is expected to 
be imperceptible, or nearly so, to vehicle occupants or approaching 
drivers. For example, NHTSA has determined that the following three 
noncompliances with FMVSS No. 108, Lamps, Reflective Devices, and 
Associated Equipment, were inconsequential: where certain headlamps had 
a slight decrease in long-range visibility and a slight broadening of 
beam patterns, where the photometric output of certain center high-
mounted stop lamps (CHMSL) was blocked by blackout paint, and where a 
CHMSL illuminated briefly absent braking when the hazard button was 
fully depressed. In these cases, there was deviation from the 
performance requirements of the standard, but in each case, the 
noncompliance was determined to be so minor as to be inconsequential. 
See General Motors Corporation; Grant of Application for Decision of 
Inconsequential Noncompliance, 63 FR 70179 (December 18, 1998) (NHTSA-
98-3813-2); Subaru of America, Inc.; Grant of Application for Decision 
of Inconsequential Noncompliance, 66 FR 18354 (April 6, 2001) (NHTSA-
2000-8201-2); General Motors Corporation; Grant of Application for 
Decision of Inconsequential Noncompliance, 66 FR 32871 (June 18, 2001) 
(NHTSA-2000-7312-2).

[[Page 19900]]

    Arguments that only a small number of vehicles or items of motor 
vehicle equipment are affected by a noncompliance will not justify 
granting an inconsequentiality petition. Likewise, we have rejected 
petitions based on the assertion that only a small percentage of the 
vehicles or items of equipment covered by a noncompliance determination 
are likely to actually exhibit the noncompliance. In many cases, it may 
not be readily apparent which vehicles or items of equipment are 
actually noncompliant. More importantly, the key issue in determining 
inconsequentiality is not the aggregate safety consequences of the 
noncompliance as a percentage of all drivers, but instead, whether the 
noncompliance in question is likely to increase the safety risk to 
individual occupants who experience the type of injurious event against 
which the standard was designed to protect. See Cosco, Inc.; Denial of 
Application for Decision of Inconsequential Noncompliance, 64 FR 29408 
(June 1, 1999) (NHTSA-98-4033-2).

B. Assessment of GM's Arguments in Support of Its Petition

    GM's attempt, through its ``risk analysis,'' to minimize the 
potential safety impact of the noncompliance by arguing that there is a 
very low likelihood of any particular individual being exposed to an 
increased risk is not compelling, and we reject it for the reasons 
discussed above and in previous agency denials of inconsequentiality 
petitions (e.g., Cosco, Inc., ibid). The percentage of potential 
occupants that could be adversely affected by a noncompliance is 
irrelevant to the consequentiality of the noncompliance. Rather, the 
question is whether an occupant who is affected by the noncompliance is 
likely to be exposed to a significantly greater risk than an occupant 
in a compliant vehicle.
    However, on the basis of the sled test and simulation data provided 
by GM, the agency has concluded that GM has adequately demonstrated 
that the potential safety consequences of the failure of the vehicle-
sensitive locking mechanisms in the ELRs in the C/K vehicles to 
function properly are inconsequential. While the webbing-sensitive 
systems in these vehicles do allow slightly increased belt payout 
compared to a functional vehicle-sensitive system, and lock slightly 
later in a crash event, these differences do not appear to expose a 
vehicle occupant to a significantly greater risk of injury. Conversely, 
the absence of a properly functioning vehicle-sensitive retractor in 
the seat belt assemblies installed in the S/T vehicles results in a 
significant derogation of their performance compared to their 
performance with a complying assembly, which precludes a determination 
that the noncompliance is inconsequential in those vehicles.
    For both the C/K and S/T vehicles, GM estimated the performance 
differences between a vehicle with a fully functional, compliant ELR 
and a vehicle with an ELR that has a broken, non-functioning vehicle-
sensitive mechanism and a functioning webbing-sensitive locking 
mechanism. This analysis was based upon a series of tests conducted by 
TKH. GM analyzed three scenarios in which there could conceivably be an 
increased risk: (1) Injuries due to an occupant moving closer to the 
front of the vehicle following pre-crash braking; (2) injuries in 
frontal crashes; and (3) injuries in rollover crashes.
    With respect to the first scenario, GM presented confidential test 
data from in-vehicle panic braking tests conducted by TKH in an S/T 
vehicle at three different speeds with test dummies and human 
volunteers, as well as simulated sled tests for both C/K and S/T 
vehicles. Although there is a very slight increase in the amount of 
belt payout when the vehicle-sensitive mechanism is disabled, we have 
concluded that it is unlikely to significantly increase the risk of 
injury during pre-crash braking events in any of these vehicles.
    To assess the potential increase in risk in a frontal crash, GM 
analyzed the results of 30 frontal sled tests at differing test 
speeds.\6\ The tests were conducted with both a 50th percentile adult 
male test dummy and a 6-year-old child test dummy. The webbing payout, 
estimated lock time, and dummy head excursion were recorded in each 
test. In eight of the tests, maximum chest acceleration readings were 
recorded in accordance with the procedures in FMVSS No. 208. The 
noncompliance consistently had a greater effect in the S/T vehicles 
than in the C/K vehicles, although with both types of vehicles, the 
dummy injury measurements did not increase significantly and were well 
below the maximum values permitted under FMVSS No. 208.
---------------------------------------------------------------------------

    \6\ The highest test speed was 32 km/h (20 mph). TKH explained 
that ``higher severity pulses will increase [the] web extraction 
rate, causing [the] web sensor to lock faster, and [be] more similar 
to [the] vehicle sensor.'' To support this statement, TKH's May 30, 
2002 presentation stated that the differences in webbing payout and 
in the estimated lock times recorded in the 16 km/h (10 mph), 24 km/
h (15 mph), and 32 km/h (20 mph) sled tests generally decreased as 
the test speed increased. NHTSA agrees that it is likely that a 
webbing-sensitive ELR mechanism will lock up more quickly in a 
severe frontal crash than in a low-to-moderate severity frontal 
crash.
---------------------------------------------------------------------------

    For example, in two 16 km/h (10 mph) frontal sled tests of the C/K 
vehicles with a 50th percentile male dummy, the webbing payout of the 
2.0 g webbing-sensitive locking mechanism averaged only 2.5 mm (0.1 
inch) more than that allowed by the compliant vehicle-sensitive ELRs, 
the average estimated lock time increased by 5 ms, and there was no 
difference in forward head excursion. In two 16 km/h (10 mph) frontal 
sled tests of the S/T vehicles with a 50th percentile male dummy, the 
webbing payout of the 3.0 g webbing-sensitive locking mechanism 
averaged 25.0 mm (1.0 inch) more than that allowed by the compliant 
vehicle-sensitive ELRs, the average estimated lock time increased by 
approximately 6 ms, and there was an average increase in forward head 
excursion of 47 mm (1.85 inches).
    Similarly, in a 32 km/h (20 mph) frontal sled test of a C/K vehicle 
with a 50th percentile male dummy, the webbing payout was only 5.0 mm 
(0.2 inches) more than that allowed by the compliant ELR, there was no 
increase in the lock time, and there was no difference in forward head 
excursion. But in two 32 km/h (20 mph) frontal sled tests of the S/T 
vehicles with a 50th percentile male dummy, the webbing payout averaged 
39.4 mm (1.55 inches) more than that allowed by the compliant ELRs, the 
average estimated lock time increased by 6 ms, and there was an average 
increase in forward head excursion of 21 mm (0.8 inches).
    Sled tests using the 6-year-old child test dummy were conducted at 
16 km/h (10 mph). In the C/K vehicles, the webbing payout of the 2.0 g 
webbing sensitive locking mechanism was, on average, 3.8 mm (0.15 
inches) more than that allowed by the compliant ELRs, the average 
estimated lock time increased by 3.5 ms, and the head excursion 
increased by an average of 3.8 mm (0.15 inches). In the S/T vehicles, 
the webbing payout of the 3.0 g webbing sensitive locking mechanism 
was, on average, 25 mm (1.0 inch) more than that allowed by the 
compliant ELRs, the average lock time increased by 13 ms, and the head 
excursion increased by an average of 65 mm (0.65 inches).
    NHTSA has concluded that the extremely small increases in webbing 
payout and lock time, with little or no increased head excursion, 
reflected in the tests of the ELRs installed in the C/K vehicles do not 
demonstrate a significant likelihood of increased injury due to the 
absence of a complying ELR in these vehicles. Accordingly, the agency 
has determined that the risk of

[[Page 19901]]

injury posed by the noncompliant systems in these vehicles in a frontal 
crash is not significantly greater than if they had a compliant ELR. 
However, the differences in the amount of webbing payout, lock time, 
and head excursion between compliant and noncompliant ELRs in the S/T 
vehicles were significantly greater than the differences experienced in 
the C/K vehicles.
    With respect to the performance of the noncompliant vehicles in a 
rollover crash, in its July 30, 2002 submission, GM acknowledged that, 
in a rollover, ``We would expect that the noncomplying belt would not 
lock up as early as the complying belt, but we have no way to be sure 
how great a difference there would be.'' However, during a November 19, 
2002 meeting at the agency, TKH presented confidential test data from a 
rollover simulation that it performed. TKH asserts that this simulation 
represents the worst-case scenario relative to the ability of these 
vehicles' webbing-sensitive systems to adequately restrain an occupant 
in the event of a rollover.\7\ These tests yielded data with respect to 
webbing payout, final belt position, and head and chest displacement.
---------------------------------------------------------------------------

    \7\ Although the simulated rollover utilized by TKH is 
relatively benign in terms of crash severity, we agree with GM and 
TKH that it presents a ``worst case'' scenario for the purpose of 
assessing the likelihood that an occupant of a vehicle with only a 
webbing-sensitive ELR would be adversely affected by additional 
webbing payout in a rollover, since a more violent crash would 
likely cause the webbing-sensitive system to lock more quickly than 
in the simulation.
---------------------------------------------------------------------------

    The data indicates that, in both cases, ELRs with only a webbing-
sensitive locking mechanism allowed somewhat more head and chest 
displacement than the compliant vehicle-sensitive ELRs. However, the 
increases in the S/T vehicles (with a 3.0 g webbing-sensitive 
mechanism), was significantly greater than the increases experienced in 
the C/K vehicles (with a 2.0 g webbing-sensitive mechanism); e.g., the 
increase in head displacement was approximately twice as large in the 
S/T vehicles as in the C/K vehicles. This data leads us to conclude 
that the absence of a vehicle-sensitive locking mechanism in the ELRs 
installed in the S/T vehicles will significantly increase the safety 
risk to occupants in a rollover crash, while the increased risk 
associated with the noncompliance in the C/K vehicles is not likely to 
be significant.
    On the basis of the foregoing, NHTSA has determined that GM has 
adequately demonstrated that, under the specific facts and 
circumstances presented here, the noncompliance with FMVSS No. 209 in 
the C/K vehicles is inconsequential to motor vehicle safety. 
Conversely, the noncompliance in the S/T vehicles is not 
inconsequential. Accordingly, GM's petition for an exemption from the 
duty to recall these noncompliant vehicles is granted in part and 
denied in part.

    Authority: 49 U.S.C. 30118, 30120; delegations of authority at 
49 CFR 1.50 and 501.8.

Kenneth N. Weinstein,
Associate Administrator for Enforcement.
[FR Doc. 04-8418 Filed 4-13-04; 8:45 am]
BILLING CODE 4910-59-P