There are many factors that influence crashes caused by flat tires/blowouts, including speed, tire pressure, and the load on the vehicle. Blowouts to the front tire can cause roadway departure, or can cause a lane change resulting in a head-on crash. Blowouts in a rear tire can cause spinning out and loss of control. As discussed in the target population section, a target population can be estimated for tire problems, but the agency doesn't know how many of these crashes are influenced by tire design or under-inflation. The agency's best estimates of these effects are discussed below.

    The target population is 414 fatalities and 10,275 non-fatal injuries that occur annually in light vehicles in which the cause of the crash is a flat tire/blowout. Puncture is the most common reason for a blowout. However, there are also many cases where a tire is punctured, loses air, and then fails later after being driven a distance under-inflated. There are no data on whether the tire failed because of a nail puncture, hitting a curb, de-beading, low tire pressure with or without overloading, or normal wear out. Thus, it is difficult to estimate what percent of the tire problem crashes are the result of tire failure modes that might be affected by this proposal.

    In the Tire Pressure Monitoring System (TPMS) analysis, the agency assumed that under-inflation is involved in 20 percent of the cases that caused the crash. The agency assumed that the influence that under-inflation has on the chances of a blowout are influenced by both tire pressure and the properties of the tire. Thus, we assumed that better inflation would take care of 50 percent of these cases and we assumed that better tires could take care of 50 percent of this problem. Thus, 41 fatalities (414 x .2 x .5) and 1,028 injuries were assigned to the TPMS rule. This leaves the target population for this rule at 373 fatalities and 9,247 injuries.

    The impact of the proposed rule will be to increase the strength, endurance, and heat resistance of tires by strengthening the standards on road hazard, bead unseating, endurance, high speed tests, and by adding an aging requirement. The impact of strengthening the standards is that certain tires would be eliminated from the U.S. marketplace. Although there is not a direct one-for-one correlation, the agency believes that P-metric tires that could not pass the proposal would currently be rated either B or C for temperature resistance under the Uniform Tire Quality Grading System (UTQGS). The recalled Firestone tires that were used on the Ford Explorer are C-rated for temperature resistance and did not pass the test protocols proposed for the agency's test program.

    Table IV-1 shows our estimate of how many tires would have failed the combination of high speed and endurance tests under the different alternatives. Tires had to pass both tests to be considered as passing the proposed tests. There were 15 P-metric tires tested and 7 LT tires tested for both high speed and endurance.

    Table 1
    Estimated Percentage of Tires Failing the Alternative Tests

    P-Metric Tires Alternative 1 Alternative 2 Alternative 3
    High Speed Test 0 13% (2 of 15) 33% (5 of 15)
    Endurance Test 0 20% (3 of 15) 60% (9 of 15)
    Both Tests 0 33% (5 of 15) 67% (10 of 15)
    High Speed Test 0 0 71% (5 of 7)
    Endurance Test 0 29% (2 of 7) 86% (6 of 7)
    Both Tests 0 29% (2 of 7) 86% (6 of 7)

    For the proposed Alternative 2 endurance test, out of the 15 P-metric tires tested, we estimate that three would have failed, two B-rated tires and a C-rated tire (1). For the high-speed test, 2 different P-metric tires probably would have failed, a B-rated tire and a C-rated tire (2). Out of the 7 LT-tires tested, two would have probably failed the endurance test (LT tires are not rated for UTQGS) and none would have probably failed the high-speed test.

    Approximately 1,700 different P-metric tire models were rated in the UTQGS for 1999. Of those tire models, 328 (19 percent) have a C-rating for temperature resistance, about 50 percent have a B-rating for temperature resistance, and about 31 percent have an A-rating for temperature resistance. The agency has no sales data for these tire models, so it does not know what percent of tires sold are C-rated, B-rated, or A-rated tires. Each tire model is required to pass all of the tests. Given the test data we have on endurance and high speed (where both C-rated tires tested probably would have failed and where three of seven B-rated tire models might have failed the proposed test), the agency assumes that 33 percent of the P-metric tire models (not sales) would fail the proposed test (5 of 15 total tires tested failed) = 33 percent. Weighting all C-rated tires (17 percent) and 3 of 7 B-rated tires (43 percent times 50 percent) = 38.5 percent, not very different from the straight percentage of tires tested of 33 percent and the weighted estimate assumes that all C-rated tires would fail the proposed test based on testing only 2 of the 328 C-rated models).

    There are approximately 5,000 LT tire models are in the marketplace. The test data we have on endurance and high speed indicate that two of the seven (29 percent) of the LT tires tested would probably not have passed the proposed Alternative 2 test. Weighting P-metric and LT tires results in the estimate that 32.8 percent (33% * .95 + 29% * .05) of the light vehicle tires would not have passed the proposed Alternative 2 test.

    Obviously, the agency needs much more test data on P-metric and LT tires to better understand what percent of the tire models or what percent of the tire sales would not pass the proposed set of tests. Comments providing test data and sales data are requested.

    While it is intuitively correct to upgrade the tire standards (i.e., stronger tires will lead to less blowouts, tire failures, and de-beading problems), the agency cannot make a direct link between the present standard and the proposed upgrade of the standard, in terms of tire failures. Obviously, Alternative 1 would result in no benefits, since we estimate that all of the tires tested would have passed the test criteria for this alternative.

    Alternative 3 would intuitively result in more benefits than Alternative 2, since the test criteria are more stringent and more tires fail Alternative 3 testing. The only tires that passed the Alternative 3 criteria are A-rated tires for temperature (and one A-rated tire failed the Alternative 3 criteria). Thus, if the agency chose Alternative 3, it appears that the only tires that could be sold would be A-rated tires for temperature. Again, the agency cannot quantify how much improvement in benefits there would be for Alternative 3 compared to Alternative 2.

    One comparison that the agency can make is to consider the results of the high-speed tests that were run to failure in the Phase 1 testing and compare the times it took for the tires to fail compared to the proposal. The proposal (Alternative 2) is for a 90 minute high speed test. For P-metric tires, the average times to failure (see Table II-4) for the six test series where the tires were run to failure were 93 minutes for A-rated (for temperature resistance) tires, 100 minutes for B-rated tires, 72 minutes for C-rated tires, and 108 minutes for LT tires. If the two C-rated tires that were tested were representative of all C-rated tires, and the proposed level was compared to the average of the six test series, then the average C-rated tire would need to be improved by 25 percent (90/72 - 1) to meet the proposed high-speed test. However, the proposal is not an average of the six tests run (proposed at 85% load and 220 kPa), but is closer to an average of the four tests (80% load at 210 and 240 kPa, and 90% load at 210 and 240 kPa). Since the tests at 210 kPA were not run to failure, but were stopped after 90 minutes if the tire did not fail sooner, a fair comparison could not be made which would allow a better estimate of how much improvement will be needed for C-rated tires. However, ignoring the bias that results from stopping the 210 kPa test at 90 minutes, the average of these four tests (87, 88, 100, and 60) for C-rated tires is about 84 minutes, which means that the average C-rated tire would need to improve by about 7 percent (90/84 - 1) to meet the proposal.

    It is hard to use the endurance tests to determine how much improvement is needed to pass, since the Phase 2 tests were run at higher loads than proposed. Many of the failures occurred very close to what the agency considered a passing grade. However, there was considerable variability in the tires tested and reducing that variability will be a considerable benefit in reducing tire failure.

    For LT tires, there were no failures in the high-speed test, the only failures were in the endurance test. Two Goodyear Wrangler tires failed very quickly in different endurance tests. The other failure occurred very close to what the agency considered a passing grade.

    Based on the tires tested, and comparing how well the tires did in the tests compared to what the agency estimates they need to do to pass the proposed tests (Alternative 2 high-speed test and Alternative 2 endurance test), the agency estimates that the high speed test will improve tire safety by about 7 percent and the endurance test will improve tire safety by about 15 percent. The agency considers these results additive, such that the total benefit from these two tests will be 22 percent for those tires that currently don't pass the test.

    There are four other tests that the agency either believes the benefits will be low or that we cannot quantify currently. Based on preliminary testing of a small sample of tires, all of the tires tested met the proposed upgraded road hazard impact test and de-beading test. Thus, we anticipate zero to minimum benefit from upgrading these two tests. The agency has not done enough testing for the aging test to form an opinion of its potential benefits. The low pressure - endurance test would have no benefit, since all of the tires tested passed. The failure margin for the low pressure - high-speed test appears to be very high. Benefits for this test cannot be easily characterized. There is some overlap of the benefits for this test and the benefits of the high speed and endurance test and the tire pressure monitoring system.

    It appears that there is significant variability in tires and if this variability can be reduced, many of the failed tires could pass the proposed test.

    The problem the agency has in estimating benefits is that while the agency knows intuitively that any improvement in how tires do in these tests will improve safety, it does not know how to translate the test improvement into real world benefits. Furthermore, it is hard to estimate what improvement might occur if variability in tires were reduced in the real world. Will the impact be large or small? Comments are requested to help answer this question.

    At this time, the agency knows that improving tires will be beneficial in reducing tire failures and crashes resulting from tire failures. The question is, do these upgraded requirements result in tires avoiding a heat-related or structure related problem long enough that the tire is discarded because of a worn tread or some other reason before it fails.

    We have made an estimate of the target population. There are an estimated 373 fatalities and 9,247 injuries in the target population. However, we do not have a good estimate of effectiveness. Assuming that the improvement needed to pass the high-speed and endurance tests (estimated to be 22 percent) related to a reduction in flat tires/blowouts, the total potential improvement would be 82 lives saved (373 * .22) and 2,034 injuries avoided if only those tires in the target population were those that needed improvements. If the tires having flats and blowouts were a random selection of all tires and only benefits accrued to those tires currently not passing Alternative 2 (weighted to be 32.8 percent), then the benefits would be 27 lives saved (373 * 0.22 * 0.328) and 667 injuries reduced.

    The 27 lives saved and 667 injuries reduced are estimated to be the benefits of the high-speed test and the endurance test.

    The agency examined the UTQGS ratings for temperature resistance, traction, and tread-wear to determine if there is a correlation between these factors. The question was whether there would be other safety benefits if many of the tires rated low for temperature resistance were taken off the market. However, no correlation was found between temperature resistance rating and tread wear, nor between temperature resistance rating and traction.

    1 For Alternative 2, the bases for assuming a tire would fail the endurance test were tests where the speed and kPa were as proposed, and there was a failure during the time or within one hour of the time that the tire was tested at up to 110% of load. For example, when the load was 100% and 110% for 8 and 10 hours, respectively, and the failure occurred in less than 19 hours, we assumed this tire would fail the proposed test.

    2 For Alternative 2, the bases for assuming a tire would fail the high speed test were when the test conditions were 85% load at 220 kPa, and the failure occurred during the 160 km/h portion of the test, or some equivalent values.