A. Summary of the Final Rule
The amendments in this final rule are essentially the same as those proposed in the NPRM, but with compliance requirements for the rear impact upgrade to be phased-in. Instead of providing that all vehicles must comply at the end of a several year period, as proposed in the NPRM, the agency is providing that compliance with the rear impact upgrade will be phased-in over an additional three-year period, without credits for early compliance. The lead time for the side impact upgrade is the same as proposed.
As proposed in the NPRM, the final rule establishes a rear impact test procedure that specifies striking the rear of the test vehicle at 80 ± 1 km/h (50 mph) with a 1,368 kilogram (3,015 pound) MDB at a 70 percent overlap with the test vehicle. The MDB is located 50 millimeters (2 inches) lower than the face of the Standard No. 214 barrier to simulate pre-crash braking. This replaces the 48 km/h rear moving barrier crash test previously required under S6.2 of Standard No. 301.
Also as proposed, the final rule eliminates Standard No. 301's side crash test and replaces it with the side impact crash test currently specified in Standard No. 214. S6.3 of Standard No. 301 had required a vehicle's side to be impacted by a barrier moving at 48 km/h. This final rule incorporates into S6.3 the side impact crash test in Standard No. 214, which is also amended by the final rule to specify that a stationary vehicle be struck on either side by a 1,368 kg (3,015 pound) MDB moving at a speed of 53 ± 1km/h.
NHTSA notes that while it has conducted research to explore the desirability of revising the Standard No. 214 barrier, additional research would have to be conducted before the agency could decide whether proposing a revision might be worthwhile. Thus, even if a revision were ultimately adopted, it could not be implemented until well beyond the implementation of this upgrade to Standard No. 301.
Instead of providing that all vehicles comply at the end of a several year period, as proposed in the NPRM, the agency is providing that compliance with the rear impact upgrade will be phased in by increasing percentages of production over an additional three-year period, without credits for early compliance or compliance above the required percentages.
At least 40 percent of vehicles manufactured on or after September 1, 2006, but before September 1, 2007, must comply with the new rear impact requirements. At least 70 percent of the vehicles manufactured on or after September 1, 2007, but before September 1, 2008, will have to comply, and starting September 1, 2008, all vehicles manufactured will have to comply with the upgraded rear impact requirements. The final rule amends 49 CFR Part 586, establishing reporting and record keeping requirements concerning the phase-in. However, vehicles manufactured in two or more stages will not be required to meet the rear impact upgrade requirements until September 1, 2008, when all vehicles must be certified as complying.
All vehicles manufactured on or after September 1, 2004, must comply with the upgraded side impact requirements.
B. Rear Impact Test Procedure
After reviewing real-world fire-related crash data, various vehicle offset crash tests, and the comments, NHTSA is adopting the rear impact test procedure as proposed. The final rule replaces Standard No. 301's current rear impact test procedure with one that specifies striking the rear of the test vehicle at 80 ± 1 km/h (50 mph) with a 1,368 kilogram (3,015 pound) MDB at a 70 percent overlap with either side of the test vehicle. The MDB face is located 50 millimeters (2 inches) lower than the face of the Standard No. 214 barrier to simulate pre-crash braking.
The agency is not adopting DC's recommended impact speed of 56-64 km/h for the reasons stated in the NPRM. Namely, the agency believes that the upgraded test procedure will simulate a type of rear vehicle-to-vehicle collision that can result in post-crash fire in an otherwise survivable crash: a high speed offset rear strike to the vehicle that results in fuel leakage from a breach in the fuel system and, potentially, a rapidly spreading fire that results in fatalities and injuries. As NHTSA noted in the NPRM, NASS estimates show that the majority of fatal and nonfatal occupant burn injuries in rear impact crashes were in the 34 to 48 km/h (21 to 30 mph) delta-v range. The agency believes that the offset rear impact test procedure specified in this final rule will simulate vehicle-to-vehicle crashes with a delta-v range of 32 to 48 km/h (20 to 30 mph).
The agency is also not adopting the recommendations of DC, GM, and Ford to specify design changes to the MDB and cart system. DC, GM, and Ford commented that, in their rear impact testing under the proposed test procedure, the MDB honeycomb appeared to bottom out. NHTSA, in the rear impact testing that it conducted in support of this rulemaking, took detailed post-crash test honeycomb crush measurements. None of the measurements indicated complete bottoming out of the honeycomb, and only a few of the measurements indicated about 85 percent compression.
Honeycomb, by design, is limited to approximately 85 percent compression, at which point it begins to stiffen considerably until it becomes infinitely stiff at 90-95 percent compression. During the 85-90 percent compression phase, it is similar to vehicle structures that become progressively stiffer as the crush in a crash increases. The measurements that were at 85 percent compression were observed near the edges of the MDB face, and since the total area was miniscule compared to the overall block of honeycomb, there will be little or no effect on the total or local forces exerted. Therefore, the energy absorbing honeycomb element has fully served its function of spreading the loads to the soft and hard structures of the vehicle and dissipating its share of the crash energy by the time that nearly full compression occurs. Accordingly, the agency believes that bottoming out of the honeycomb is not a concern.
DC, GM, and Ford also commented that, in their rear impact testing, the uprights supporting the cart face inadvertently contacted the struck vehicle. None of the commenters provided details on specific tests. However, the agency observed such contact in several of its rear impact tests. The NHTSA tests were conducted with the Standard No. 214 barrier, which has uprights that extend approximately 100 mm (4 inches) above the backing plate. The agency conducted film analysis of its tests and found there was no contact between the uprights and any significant vehicle structural components. The only vehicle components that contacted the uprights were trunks or tailgates that were already deformed. These contacts came late in the crash event and did not influence the outcomes.
The agency also notes that GM conducted several rear impact crash tests with a non-Standard No. 214 barrier. The cart used by GM had uprights extending approximately 600 mm (24 inches) above the backing plate, or 20 inches taller than the uprights on the barrier used by NHTSA. Despite this significant difference, there was no difference in the GM and NHTSA test results. Accordingly, the agency believes that contact between the cart uprights and the struck vehicle is not a concern.
GM noted that in some of its tests the barrier face underrode the struck vehicle and, upon rebound, the upright that contacted the vehicle became "caught" on vehicle structure (e.g., bumper, frame cross member, etc.), with undetermined effects on the struck vehicle. Ford stated that, in some of its tests, right angle corners of the barrier face hung up on vehicle trim, potentially affecting test repeatability.
NHTSA did not observe either of these phenomena in any of its testing. The agency notes that the top edge of the deformable element of the barrier is 31 inches above the ground. Consequently, for the barrier face to underride the struck vehicle, the rear end of the vehicle would have to be lifted approximately 20 to 24 inches off the ground. The agency believes that such lifting is highly unlikely. Moreover, the agency believes that any possible effects from the phenomena observed by GM and Ford would be secondary since they occur after the maximum crush damage of the test. The agency notes that damage to the fuel systems tested by NHTSA and GM appeared to result from crush damage rather than from any secondary damage. Accordingly, the agency believes that the points raised by GM and Ford are not of concern.
NHTSA is not adopting VW's suggestion to not lower the barrier face by 50 mm (2 inches) to simulate pre-crash braking. NHTSA has determined that this change will have no measurable effect on the performance of the barrier.  The agency's tests indicate that the center of gravity of the barrier will drop about 7.4 mm (0.29 inches), which is well within the 25.4 mm (1 inch) allowed tolerance for center of gravity locations. The moment of inertia of the barrier about the longitudinal (roll) and transverse (pitch) axes will be reduced 0.1 percent and 0.02 percent, respectively, and there will be no change in the vertical (yaw) axis. The agency notes that the device currently used to measure these parameters is not capable of measuring such small changes in the moment of inertia. Moreover, these small changes will not produce any measurable effect on the test results.
The agency is not adopting the recommendation of Advocates, IIHS, and AAD to specify the use of a heavier (4,000 pound) barrier because, as the agency noted in the NPRM, in an 80 km/h (50 mph) rear impact offset crash test, a 3,015-pound barrier effectively reproduces the damage profile seen in real world crashes that most often lead to fires. If a heavier barrier were used, the proposed rear impact crash test would no longer reproduce that profile. In addition, the agency has conducted its crash tests in support of this rulemaking with a 3,015-pound barrier. The agency would have to conduct further research and development before a heavier barrier could be proposed for use in any test procedure. 
Honda claimed that the MDB overrode the rear of the test vehicle but was not specific about the vehicle test in which the override occurred. We presume that Honda may have examined the test film of a Honda Accord test conducted by GM under the GM C/K settlement agreement with NHTSA. While NHTSA personnel made suggestions and witnessed the testing, we did not have direct control over the conduct of the tests. The Standard No. 301 upgrade test protocol was not precisely followed, but was instead modified according to what GM believed to be a worst case. Two major exceptions to the Standard No. 301 upgrade protocol were the test speed and the barrier height. GM believed that testing near the Standard No. 214 height and at 85 km/h would provide a worst case scenario. In GM's test with the Honda Accord, there was indeed severe override, which we believe was due to the additional 2.5 inches in the height of the MDB face, and to a lesser degree the additional speed of the test.
VW also recommended that the proposed rear impact test be on the side of the vehicle where the filler pipe is located. Our test results indicated that fuel leakage is not dependent on the location of the filler pipe; rather, it is dependent on how the overall fuel system is protected against the impact. Therefore, we are not incorporating VW's suggestion and vehicles must comply when the rear impact test is conducted on either side of the vehicle.
Further, our cost study indicated that compliance with this final rule will not require major structural redesign (as stated by Alliance and DC), or necessitate an increase in vehicle stiffness (as stated by VW). Because there is no need to increase the vehicle stiffness, this final rule does not increase the potential for whiplash injuries in lower crashes as suggested by VW.
C. Side Impact Test Procedure
The agency is replacing Standard No. 301's current lateral crash test with the side impact crash test specified in Standard No. 214. The Standard No. 214 side impact crash test specifies that a stationary vehicle be struck on either side by a 1,368 kg (3,015 pound) MDB. As noted above, Standard No. 214 currently specifies an impact speed of 54 km/h. In order to provide an appropriate tolerance without affecting the stringency of the test, the agency proposed to change the test speed to 53 ± 1 km/h and adopt it for Standard No. 301. No comments addressed this issue, and we are adopting that proposal.
The agency is specifying that the MDB be lowered 50 mm (2 inches) for the rear impact test to simulate pre-braking. However, the agency is not specifying that the MDB be lowered for the side impact test. The test conditions of Standard No. 214 were intended to reflect a "worst case" scenario for occupants riding in the struck vehicle in that the striking vehicle was assumed not to be braking prior to impact. A braking vehicle would strike lower than a non-striking vehicle, potentially engaging more of the side sill of the struck vehicle. Full engagement of the sill dissipates the crush energy more effectively than engagement of the door structures located above the sill, resulting in les intrusion and deformation along the struck side where fuel system components (e.g., fuller filler neck and tube) are located. NHTSA believes the integrity of those fuel system components would be tested in a more severe environment if the barrier were not lowered. For these reasons, NHTSA has decided not to lower the MDB for the side impact test.
D. Door System Integrity
NHTSA believes that a post-crash door operability requirement could be a practicable, reasonable safety enhancement. However, the agency has decided not to add a post-crash door operability requirement to Standard No. 301 or Standard No. 206 in this rulemaking. The agency has not developed a practical, objective, and repeatable test procedure for testing door operability. The agency is specifically concerned with developing specifications for the type and magnitude of force needed to test door operability. The agency notes that none of the commenters who supported a door operability requirement suggested a test procedure.
Accordingly, NHTSA will need to conduct research before proposing any post-crash door operability requirement and will consider adding a post-crash door operability requirement to Standard No. 206 or Standard No. 301 in a separate rulemaking.
E. Lead Time
In the Preliminary Regulatory Evaluation for the NPRM, the agency said that most vehicles needing modification to meet the upgrade rear impact test would need only minor modifications and estimated that those modifications could be completed in 33 months. Based on that estimate, we provided a lead time of three years.
The Alliance said "significant vehicle redesign and retooling for production will likely be required in a number of vehicles." It suggested "(t)he small number of tests conducted by the NHTSA, often with a sample size of one vehicle, simply is inadequate to identify whether vehicle changes are required or for any manufacturer to assure compliance for all its vehicles." The Alliance suggested that the agency phase in the requirements, beginning not earlier than three years after the issuance of the final rule, according to the following annually increasing percentages of production: 25%, 40%, 70% and 100%.
In their comments on the NPRM, DC and Honda argued that some vehicles would need more than just minor modifications and that therefore additional lead time should be provided. Both suggested that the entire rear ends of some vehicles would have to be redesigned, although neither identified any specific models in need of such changes. Honda generally cited crash testing in support of its argument, but gave no details about that testing. DC suggested the same phase-in recommended by the Alliance. Honda suggested that the agency phase in the requirements, beginning not earlier than three years after the issuance of the final rule, according to the following percentages of production: 10%, 30%, 70% and 100%.
While the agency continues to believe that a three-year lead time is sufficient for most vehicles in need of modification, it agrees that it is desirable to provide additional lead time to accommodate any new models that were designed and developed based upon the current requirements. The agency recognizes that vehicle platforms, once developed, are typically used for a number of years without major structural modification. We also recognize that in order to meet the requirements of FMVSS No. 301, a vehicle, without modification, must meet the static roll over requirements following an impact in a barrier test. Consequently, we have decided that the upgraded rear impact test will be phased-in, beginning on September 1, 2006, according to the following percentages of production: 40%, 70% and 100%. We believe that this combination of lead time and phase-in will allow sufficient time for existing platforms to be redesigned to comply with all of the requirements of FMVSS No. 301, and that the four year phase-in proposed by the Alliance is not necessary.
We also believe that it is not necessary to allow an optional "0%, 0%, 100%" three-year phase-in for limited-line manufacturers as proposed by Porsche. A similar phase-in exception for limited line manufacturers is present for the advanced air bag requirements of Standard No. 208. However, the advanced technology requirements that compounded the disparity between the phase-in requirements of the Advanced Air Bag Rule for limited line manufacturers and more diverse manufacturers is not present here.
In the Preliminary Regulatory Evaluation for the NPRM, we said
Since almost all vehicles pass the Standard No. 214 test without fuel leakage and all manufacturers have done these tests on their passenger cars and light trucks and vans up to 6,000 pounds GVWR, the agency is proposing a one year leadtime after the final rule for implementing the Standard No. 214 test requirement for the lateral test.
Some vehicle manufacturers supported the one-year lead time proposed in the NPRM for the side impact test upgrade, but recommended that the agency add a phase-in after this lead time. The Alliance asked for additional lead time for vehicles not previously subject to Standard No. 214, saying
Until a complete and thorough evaluation program is completed for each model, the actual and exact extent of changes to each vehicle cannot be ascertained. For this reason, we recommend that vehicles not previously subject to FMVSS 214 requirements also have the same three-year lead time and four year phase-in schedule as proposed above for the rear impact requirement, with the allowance for early compliance. Starting the lead time and the phase-in for both the rear and side impact requirements at the same time would also be logical and provide clarity since both are contained in the single notice.
As we note below in the section on costs, only one out of more than 100 vehicles tested failed Standard No. 301's fuel leakage requirements using Standard No. 214's side impact test. Based on those test results, the agency believes that few vehicles, approximately 1%, will have to be modified to meet Standard No. 301's leakage requirements using Standard No. 214's side impact test. Therefore, the one-year lead time without a phase-in is adopted, as was proposed.
The target population of crashes includes multi-vehicle crashes in which a passenger vehicle is struck in the rear by another passenger vehicle and the fire starts in the struck vehicle. There are an estimated 58 burn-related fatalities and 119 non-fatal burn-related injuries annually in the target population. The non-fatal burn injuries in that population of crashes were mostly minor and were typically not the most severe injury to the occupant. The agency estimates that approximately 8 to 21 fatalities will be prevented once all vehicles on the road comply with the upgraded rear impact test. The cost per life saved is estimated to be $1.96 million to $5.13 million ($41 million/21 lives to $41 million/8 lives). The agency is not estimating the number of reduced non-fatal burn-related injuries because there are only a few cases each year in which the injured person's most serious injury was a burn injury.
There are fewer than 100 fatalities annually in multi-vehicle side impacts that result in fire. The agency believes that the Standard No. 214 side impact test is somewhat stricter than the existing lateral impact test in Standard No. 301. However, the agency was unable to quantify any benefits from switching to the Standard No. 214 side impact test.
NHTSA disagrees with the VW comment that the benefits of this rulemaking are too low compared to its costs. VW did not provide any data to support their comment. However, the agency believes that VW's cost estimates may be based on costs of issues, such as post-crash door operability, seat back failures, and dummy responses, which were discussed in the NPRM but not adopted in this final rule.
The agency estimates that the average cost for vehicles that will need to be modified to comply with the upgraded rear impact test is $5.31 per vehicle.  Based on its belief that the test failures in the agency's testing were more the result of design differences than vehicle weight differences, the agency estimates that 46 percent of the vehicle fleet does not currently meet the upgraded rear impact test. It further estimates that approximately 16.7 million vehicles are sold each year in the U.S. Together, this information indicates that the total cost for the fleet will be approximately $41 million per year.
Using the Standard No. 214 side impact test as the Standard No. 301 side impact test will eliminate the cost of conducting a unique Standard No. 301 test as well as the cost of an extra test vehicle. Since the average current Standard No. 301 side impact test is roughly $4,300 and the average test vehicle costs about $21,000, the total savings would be about $25,200 per vehicle model. 
Only one out of more than 100 vehicles tested failed Standard No. 301's fuel leakage requirements using Standard No. 214's side impact test. Based on those test results, the agency believes that few vehicles will have to be modified to meet Standard No. 301's side impact leakage requirements using Standard No. 214's side impact test.
NHTSA disagrees with the Honda comment that the agency's cost estimates are too low. The agency's cost estimates are based on the changes that will be needed to remedy those noncompliant vehicles needing only minor modifications. Since most vehicles readily pass the fuel leakage requirements using the Standard No. 214 side impact test, we do not believe modifications will be required which are not minor. Neither Honda nor any of the other vehicle manufacturers provided data indicating that the costs of modifying vehicles to comply will be greater than the agency's estimates.  Furthermore, 54 percent of the vehicles tested were able to pass at the higher test speed and the measures required to address the failing vehicles do not involve structural changes.
H. Additional Issues
None of the commenters provided any real-world data on the relationship between crash severity and the risk of occupant injury due to fire. Thus, as discussed above in the section on Costs and Benefits, it appears that the data files NHTSA used in developing the NPRM are the best available data sources.
The agency agrees with the Alliance's comment that the issue of occupant protection would best be discussed in the context of a rulemaking focused on that issue. The agency notes that it is in the process of developing a final rule for Standard No. 202 and a proposed rule for Standard No. 207, and will address rear impact protection in those rulemakings.
NHTSA agrees with the DC comment that the issue of seat back failure best be discussed in the context of a rulemaking focused on that issue. The agency notes that it is considering an upgrade of Standard No. 207, where this issue will be addressed.
NHTSA agrees with the GM comment that using various dummy sizes to study the safety consequences to vehicle occupants due to the rear impact test upgrade should be a subject for future research. The agency will address the use of various dummy sizes in rear impact tests in separate rulemakings.
The agency is adopting GM's suggested revision of S7.1.6(b) of Standard No. 301. That paragraph is revised to read as follows:
S7.1.6(b) Except as specified in S7.1.1, a multipurpose passenger vehicle, truck, or bus with a GVWR of 4,536 kg or less is loaded to its unloaded vehicle weight, plus the necessary test dummies as specified in S6, plus 136 kg or its rated cargo and luggage capacity weight, whichever is less, secured in the load carrying area and distributed as nearly as possible in proportion to its GAWR. For the purpose of this standard, unloaded vehicle weight does not include the weight of work-performing accessories. Each dummy is restrained only by means that are installed in the vehicle for protection at its seating position.
This provision is revised to simplify the language and provide clearer instruction as to how weight is to be distributed on a vehicle under test conditions. NHTSA is not adopting Alliance's recommendation to retain the language that addressed distribution when the axle's proportional share is exceeded. While the revised provision eliminates that language, it does provide that weight is to be distributed as nearly as possible in proportion to the GAWR.
NHTSA is not adopting GM's recommendation that the agency replace the language in S8.1.1 of Standard No. 208, S6.1 of Standard No. 212, S7.7 of Standard No. 219, S7.1.6 of Standard No. 301, S7.1.6 of Standard No. 303, and S7.2.3 of Standard No. 305 with identical language. The agency believes that such revisions are outside the scope of this rulemaking. However, the agency will treat GM's recommendation as a petition for rulemaking and will make a decision on whether or not to grant the petition within the next 120 days.
NHTSA is also not adopting GM's and VW's suggested revision of S7.1.6(a). The agency believes that the language in that paragraph is sufficiently clear.
For the reasons discussed above in the section on Rear Impact Test Procedure, the agency is specifying that the barrier face in the rear impact test procedure is lowered 50 mm (2 inches).
NHTSA agrees with the comments from the Alliance, GM, DC, Porsche, and Ford that the agency must assess the safety basis for a pole side impact test, as well as develop an objective, repeatable test procedure, including pole contact locations, closing velocities, pole sizes, and modes of testing, before the agency proposes such a test. Any future proposals to utilize a pole side impact test will address these issues.
NHTSA agrees with the comments from the Alliance and GM that the agency should consider a future revision limiting fuel system leakage in any Standard No. 208 crash test to the levels currently specified in Standard No. 301. In frontal impacts at 35 mph (56 km/hr) into the barrier, there have been 10 failures out of 406 NCAP (New Car Assessment Program) tests, since 1979. While rare, the agency will continue to monitor fuel system leakage in the NCAP tests to determine whether future upgrades would be appropriate.
NHTSA believes that there will be no change to the certification responsibilities of second-stage and final-stage manufacturers as a result of this rulemaking. The agency has reviewed the conformity statements from first-stage manufacturers submitted by the NTEA and believes them to be reasonable. Under these conformity statements, the first-stage manufacturer installs the entire fuel system, and the second-stage or final-stage manufacturer can make minor alterations without violating the pass-through certification from the first-stage manufacturer.  Further, vehicles manufactured in two or more stages will not be required to comply with the rear impact upgrade until the final stage of the phase-in.
In cases in which the second-stage or final-stage manufacturers make significant changes to the fuel system, they may not be able to use the pass-through certification, and may have to certify that the vehicle complies with Standard No. 301. If it is not economically feasible for these manufacturers to perform the compliance testing or engineering analysis, the manufacturers may apply for a temporary exemption under 49 CFR Part 555.
The agency also notes that it is currently involved in a negotiated rulemaking process with the NTEA, first-stage manufacturers, and other stakeholders regarding the certification process for vehicles manufactured in two or more stages. The agency intends to develop changes to the regulations governing the certification of such vehicles through this process.
The issues raised by Dynamic Safety concerning fuel siphoning and by IIHS and Advocates concerning fuel cutoff devices all pertain to frontal fire protection. NHTSA research identified rear impacts as the most common type of crashes that result in fires. The agency would need to conduct research to determine the extent of the problem related to fuel siphoning as described by Dynamic Safety. If future NHTSA analysis of real-world crash data indicates that there is a safety problem that warrants further regulatory action, the agency will consider additional changes to Standard No. 301. The agency notes that preliminary research of frontal impacts under the GM settlement agreement indicated that the vehicles equipped with fuel cutoff devices performed similarly to vehicles without such devices. Consequently, the agency does not anticipate that any real world benefits would result if the agency required fuel cutoff devices.
 The agency has docketed its findings on which this determination is based. See Docket No. NHTSA-00-8248, entry 3.
 The agency notes that the impact with the 3,015-pound barrier at 50 mph produces an impact energy 2.09 times greater than the impact energy produced by the current Standard No. 301 test. If the agency specified the use of a 4,000-pound barrier at 50 mph, the impact energy would be 2.78 times greater than the current Standard No. 301 test.
 This is a $0.31 increase from the figure used in the NPRM.
 The figures in the NPRM were $4,000 for the average lateral test for Standard No. 301 costs, $20,000 for average test vehicle cost, and a total of about $24,000 per vehicle model.
 The agency believes that Honda's cost estimates are overstated since they include vehicle models that complied with the amendments in this final rule when the agency tested them.
 The first-stage manufacturer typically certifies that the incomplete vehicle meets the requirements of Standard No. 301. This relieves second-stage or final-stage manufacturers of the burden of conducting Standard No. 301 compliance tests, unless those manufacturers make substantial changes to the fuel system.