V. BENEFITS

Human Factors Issues

There are two human factors issues involved with Tire Pressure Monitoring Systems (TMPS). The first is what information is presented to the driver and how it is presented, and the second is whether the warning makes the driver pull into the next service area to check the pressure.

Regarding the information that the driver sees, the agency is proposing alternative display icons for comment. Some testing has been done on the understandability of these icons. The indirect measurement systems can only provide a warning light that tire pressure is low. The direct measurement systems could display individual tire pressures and tell the driver which tire(s) are low. Although individual tire pressures are not proposed to be required, this analysis assumes that manufacturers of direct measurement systems will display individual tire pressures because it will be helpful to drivers in terms of fuel economy, tread wear and safety.

We anticipate that drivers will react differently to the different amounts of information. Some drivers will keep track of the individual tire pressures and will add pressure to their tires whenever necessary, say at 10 percent below placard, even before the warning is given. These drivers will accrue more safety benefits and more benefits in terms of fuel economy and tread life than drivers that wait longer for a warning. On the other hand, some drivers who currently check their own tires frequently enough to avoid significant underinflation may start to rely on the TPMS to indicate underinflation, rather than checking their tires frequently and filling them up whenever they were below the placard level. We believe this would happen more often for an indirect system, where only a warning light comes on when tire pressure goes below a specified threshold, rather than a direct system where individual tire pressures could be monitored continuously. These drivers would actually accrue fewer safety, tread wear and fuel economy benefits than they did without the TPMS. The agency has no information that would help it estimate what percent of drivers would put to use the information on individual tire pressures.

The second question is whether drivers, given a warning, will stop and inflate their tires back to the placard pressure. We do not expect driver compliance with the TMPS telltale, which is amber or yellow, to be 100 percent. We have found no data with which we can predict compliance levels. We assume more than 50 percent of drivers will want to make sure they don't get a flat tire and be stranded somewhere, so they will fill the low tire(s). Given just a telltale, some drivers will try to just fill one low tire. Given a reading of tire pressure on all four tires with a direct measurement system, the driver will know which tires are low and need to be filled.

For this analysis, we will assume that the equivalent of 80 percent of the drivers will react to a direct measurement system that gives them a continuous readout of tire pressure and to a continuous warning light when their tires get 20 percent below the placard and will inflate their tires the next time they refuel, given the gas station has the equipment. This takes into account the group that will fill their tires more frequently because they have continuous information, than those who would just fill their tires when given a warning. We assume that with an indirect measurement system 60 percent of the drivers will inflate their tires back up to the placard level when given a warning. Thus, for Alternative 1, we will be using 80 percent, and for Alternative 2, we will be using the weighted average of 66.6 percent (80% * 0.33 + 60% * .67).

Stopping Distance

Tires are designed to maximize their performance capabilities at a specific inflation pressure. When tires are under-inflated, the shape of the tire's footprint and the pressure it exerts on the road surface are both altered. This degrades the tire's ability to transmit braking force to the road surface. There are a number of potential benefits from maintaining the proper tire inflation level including reduced stopping distances, better handling of the vehicle in a curve or in a lane change maneuver, and less chance of hydroplaning on a wet surface, which can affect both stopping distance and skidding and/or loss of control. An estimate will be made of the impact of TPMS on stopping distance, but other benefits from improved maneuverability cannot yet be quantified.

The relationship of tire inflation to stopping distance is influenced by the road conditions (wet versus dry), as well as by the road surface composition. Decreasing stopping distance is beneficial in several ways. First, some crashes can be completely avoided by stopping quicker. Second, some crashes will still occur, but they occur at a lower impact speed because the vehicle is able to decelerate quicker during braking.

In Chapter III, a variety of stopping distance test results are discussed. In tests conducted by Goodyear Tire and Rubber Company, significant increases were found in the stopping distance of tires that were under-inflated. By contrast, tests conducted by NHTSA at their VRTC testing ground found only minor differences in stopping distance, and in some cases these distances actually decreased with lower inflation pressure. The NHTSA tests also found only minor differences between wet and dry surface stopping distance. It is likely that some of these differences are due to test track surface characteristics. The NHTSA track surface is considered to be extremely aggressive in that it allows for maximum friction with tire surfaces. It is more representative of a new road surface than the worn surfaces experienced by the vast majority of road traffic. The Goodyear tests may also be biased in other ways. Their basic wet surface tests were conducted on surfaces with .05" of standing water. This is more than would typically be encountered under normal wet road driving conditions and may thus exaggerate the stopping distances experienced under most circumstances. On the other hand, crashes are more likely to occur under more hazardous conditions, which may mean the Goodyear data are less biased when applied to the actual crash involved population. Generally speaking, the Goodyear test results imply a significant impact on stopping distance from proper tire pressure, while the NHTSA tests imply these impacts would be minor or nonexistent at lesser water depths. This analysis will estimate stopping distance impacts using the Goodyear data to establish an upper range of potential benefits. A lower range of no benefit is implied by the current NHTSA test results.

Impact Speed/Injury Probability Model

In order to estimate the impact of improved stopping distance on vehicle safety, NASS-CDS data were examined to derive a relationship between vehicle impact speed (delta-V) and the probability of injury. Following is a description of the derivation of this model.

Data: From 1995-1999 CDS, all passenger vehicle occupants involved in crashes where at least one passenger vehicle used brakes.

Methodology: (1) The percent probability risk of MAIS 0, MAIS 1+, MAIS 2+, MAIS3+, MAIS4+, MAIS 5+, and fatal injuries was calculated for each delta-V between 0 and 77 mph. The percent probability risk of each MAIS j+ injury level at each delta-V i mph is defined as the number of MAIS j+ injury divided by the total number of occupants involved at i mph delta-V. If j=0 represents MAIS 0 injuries and j=6 represents fatalities, the probability of injury risk can be represented by the following formula:

    i = 0 to 77, j = 0 to 6

Where :

p⁺_i,j = percent probability risk of MAIS j+ injuries at i mph delta-V,

I _i,j = the number of j+ injuries (i.e., MAIS 0, MAIS 1+, MAIS 2+, …, fatal) at i mph delta-V

T_i = total number of occupants at i mph delta-V

Note that p⁺_i,0 = percent probability risk of MAIS 0 injuries at i mph delta-V and p⁺_i,6 = percent probability risk of fatalities at i mph delta-V. I_i,0 = the number of MAIS 0 injuries and I_i,6 the number of fatalities at i mph delta-V.

(2) The risk-prediction curve for each j injury level was derived using a mathematical modeling process. The process used delta-V as the independent variable (i.e., predictor) and p⁺_i,j as the dependent variable and modeled all the data points (delta-V, percentage risk) for each j injury level. For example, for MAIS 1+ injuries, the process used the data points: (0, p⁺_0,1), (1, p⁺_1,1), (2, p⁺_2,1), …, (75, p⁺_75,1), (76, p⁺_76,1), (77, p⁺_77,1) to derive the MAIS 1+ risk curve. Table V-1 shows all the risk-prediction formula. These formulas were developed under two assumptions: a) no one was injured at 0 mph, i.e., P⁺_0,0 = 100 percent, and P⁺_0,j = 0 percent for j=1…6, and b) everyone was assumed to have at least MAIS 1 injuries for 36 mph and higher delta-V, i.e., p⁺_i,0 = 0 , for i >=36 mph. This assumption was based on the injury distribution derived from 1995-1999 CDS.

Table V-1
Injury Probability Risk Curve Formula

Injury Level	Risk-Prediction Formula
MAIS 0
MAIS 1+
MAIS 2+
MAIS 3+
MAIS 4+
MAIS 5+
Fatal (j=6)

(3) The percent probability risk p_i,j was calculated for individual MAIS level. For MAIS 0 (j=0) and fatal injuries (j=6), p_i,0 = p⁺_i,0 and p_i,6 = p⁺_i,6 . The percentage risk for each MAIS 1 to MAIS 5 injury level is the difference between the two predicted risks. Thus, p_i,1 (risk of MAIS 1 at i mph delta-V) = p⁺_i,1 - p⁺_i,2, p_i,2 = p⁺_i,2 - p⁺_i,3, p_i,3 = p⁺_i,3 - p⁺_i,4, p_i,4 = p⁺_i,4 - p⁺_i,5, and p_i,5 = p⁺_i,5 - p⁺_i,6.

(4) Adjusted total row percent risk to 100 percent. Because of statistical measurement variation and predicting errors, the row risk percentages at some delta-Vs do not add to 100 percent. To adjust to a total of 100 percent for these delta-Vs, an adjustment factor (f_i) is applied to every risk probability. The adjustment factor is 100/(actual total percentage), i.e.,

   where j = 0…6.

The adjusted risk probabilities for i mph delta-V would be f_i * p_i,j. For example, at 10 mph delta-V, f₁₀ = 100/85 = 1.1765. The risk probability for MAIS 0 becomes 52.5 (= 44.6*1.1765) and MAIS 1 becomes 43.5 (= 37.0*1.1765). These adjusted risk probabilities are higher than those predicted by the original curves listed in Table V-1. However, the general shape of each curve does not alter significantly. Table V-2 shows the adjusted percent probabilities of risk. Note that cell probabilities were rounded to the nearest tenth. Therefore the sum of the individual cells may not total exactly 100 percent.

Once this relationship was established, crash data from 1999 CDS and FARS were distributed across this matrix to establish a "base case" injury distribution. This was done separately for 3 different groups of crashes stratified according to the speed limits on the roadways where crashes occurred. The roadway stratification was selected because stopping distances are largely dependent on initial pre-braking travel speed, and speed limits were assumed to provide a reasonable stratification for this variable. However, actual travel speeds differ from speed limits. For this analysis, it was assumed that actual travel speeds were 5 mph higher than the mean speed limit in each category. The 3 speed limit categories were 0-35mph, 36-50mph, and 51 mph and over. The mean speed limits for each category were 30, 44, and 57. There were only minor differences between speed limits for wet and dry surfaces, or for passenger cars and LTVs. Therefore, the same average speed limit is used regardless of road surface or vehicle type. Allowing for a 5 mph difference for travel speed, the three assumed average speeds that represent the speed limit categories are 35, 49, and 62 mph.

Table V-2
Adjusted Percent Probabilities of Injury Risk

Table V-2
Adjusted Percent Probabilities of Injury Risk
Continue

Separate target populations were also derived for passenger cars and LTVs, and for crashes that occur on wet and dry pavement. These distinctions were necessary because stopping distance is strongly influenced by pavement conditions and vehicle characteristics. In addition, LTVs have significantly different levels of under-inflation than passenger cars and this impacts calculations of delta-V reductions. Note that the presence or absence of anti-lock brakes also has a significant influence on stopping distance. However, because reliable data on the presence of these systems is not included in crash databases, these differences will be accounted for at a different stage of the analysis. A total of 12 separate target population cells were thus produced. The fatalities and injuries for each cell are summarized in Table V- 3 for passenger cars and Table V-4 for LTVs. Table V-5 summarizes the target populations across all passenger vehicles.

Table V-3
Passenger Vehicle Occupants in Crashes Where
at Least One Passenger Car Used Brakes
1995-1999 CDS, Annual Average

Table V-4
Passenger Vehicle Occupants in Crashes Where
at Least One LTV Used Brakes
1995-1999 CDS, Annual Average

Table V-5
Passenger Vehicle Occupants in Crashes Where
at Least One Vehicle Used Brakes
1995-1999 CDS, Annual Average

Preventable Crashes

The impact of small reductions in stopping distance will, in most cases, result in a reduction in the impact velocity, and hence the severity, of the crash. However, in some cases, reduced stopping distance will actually prevent the crash from occurring. This would result, for example, if the braking vehicle were able to stop just short of impacting another vehicle instead of sliding several more feet into the area it occupied.

The benefits that would accrue from preventable crashes would only impact that portion of the fleet that:

a)   Has low tire pressure, and
b)   Would be notified by the TPMS
c)   Is driven by drivers who will respond to the warning

Data from NHTSA's tire pressure survey (discussed in Chapter III) indicate that 74 percent of the on-road fleet has at least one tire that is under-inflated. For these vehicles, notification of this under-inflation would not be given until the system is triggered. For example, under Alternative 1, it is estimated that direct TPMS will trigger at roughly 20% below placard pressure, or roughly 6 psi for passenger cars and 7 psi for trucks. The portion of the vehicle fleet that is below these levels will potentially experience some reduction in crash incidence due to improved stopping distance. Data from NHTSA's tire pressure survey indicate that 36 percent of passenger cars and 40 percent of LTVs have at least one tire that is 20 percent or more below recommended placard pressure. However, in order to experience this reduction, the driver must respond to the warning. NHTSA has no data to indicate what portion of drivers will take action in response to this warning. For this analysis, it will be assumed that 80 percent would respond to direct systems. Eighty percent is chosen to represent a level that reflects the heightened consumer awareness that would come with systems that constantly monitor and display tire pressure levels. A lower response rate of 60 percent is assumed for indirect systems, which only provide information when the systems reach the warning level.

The portion of crashes that would actually be preventable is unknown. However, an estimate can be derived from relative stopping distance calculations for vehicles that were involved in crashes. The average stopping distance was calculated for the existing crash-involved vehicle fleet, and for that fleet if they had correct tire inflation pressure. The method used to calculate these stopping distances is described in a later section of this analysis. The results indicate that the existing passenger car fleet would, on average, experience a stopping distance of 137 feet, while the crash-involved LTV fleet experienced an average stopping distance of 131.5 feet. These differences between passenger car and LTV stopping distances reflect the distribution of injuries by speed and road conditions for each vehicle type. By contrast, the average stopping distance for passenger cars with correctly inflated tires would be 132.1 feet, while for LTVs it would be 127.3 feet.

In theory, current crashes occur under a variety of stopping distances but if these distances were shortened due to improved inflation pressure then a portion of these crashes would be prevented. Crashes could be prevented over a variety of travel speeds and braking distances. For example, a vehicle might be able to avoid an intersection crash by slowing quickly enough to miss a speeding vehicle running a red light. In an angular head-on crash, better braking could reduce the chance of two vehicles striking their corners, given that crash avoidance maneuvers are also taking place. An example for rear impacts could involve sudden braking to avoid a vehicle swerving to cross lanes on an interstate highway. We anticipate that a large portion of the fatality and serious injury benefits for crash avoidance would occur in intersection crashes, since both vehicles are moving at high speeds, and a small change in braking efficiency could result in the avoidance of a high-impact crash.

NHTSA does not have data that indicate average stopping distance in crashes. Under these circumstances, it is not unreasonable to assume that crashes are equally spread over the full range of stopping distances. Under this assumption, the change in stopping distance under proper inflation conditions can be used as a proxy for the portion of crashes that are preventable. With equal distribution of crashes across all stopping distances, the portion of crashes that occur within the existing stopping distance that exceeds the stopping distance with correct pressure represents the portion of crashes that are preventable. For passenger cars, this portion is (137-132.1)/137 or 3.6 percent of all current crashes. For LTVs, this portion is (131.5-127.3)/131.5 or 3.2 percent.

Benefits from preventable crashes were thus calculated as follows:

Ip_(s)=Pp*I_(s)*Pu*Pr

Where,

Ip_(s) = Preventable injuries of severity (s)

Pp = portion of crashes that are preventable

I_(s) = Existing injuries of severity (s)

Pu = portion of vehicles with under-inflated tires that will receive notification from TPMS

Pr = portion of drivers who will respond to the TPMS notification

The results of this analysis are shown for passenger cars under Alternative 1 in Table V-6 . The combined results for all vehicles under Alternative 1 are shown in Table V-7, and for Alternative 2 in Table V-8. Note that these results have been adjusted to reflect a small amount of overlap that occurred in the separate examination of passenger car and LTV crashes. An adjustment factor of .968 was applied to account for this overlap. This factor was derived by comparing the sum of the two separate crash counts to a total count based on all passenger vehicles.

The benefits from preventable crashes, shown in Tables V-6, 7, and 8, were assumed to occur over all crash types and severities. This assumption recognizes that there are a variety of crash circumstances for which marginal reductions in stopping distance may prevent the crash from occurring. Crash prevention may be more likely under some circumstances than others. For example, it is possible that a larger portion of side impacts might be prevented than head-on collisions. In side impacts where vehicles are moving perpendicular to each other, improved braking by one vehicle reduces the speed at which it enters the crash zone and potentially allows the second vehicle to move through the crash zone, thus avoiding the impact. In a head-on collision, both vehicles are moving toward the crash and a reduction in stopping distance for one vehicle may be less likely to avoid a high-speed crash than in the case discussed above for side impacts. Further, if a separate analysis were conducted for different crash types and severities, the portion of crashes prevented would be greater for crashes at higher speeds. However, NHTSA does not have sufficient information to conduct a separate analysis of each crash circumstance and has used an overall estimate across all crash types instead. Comments are requested on this assumption.

Table V-6
Potential Benefits from Preventable Crashes,
Passenger Cars Adjusted for Properly Inflated Vehicles,
20% Notification Level, 80% Response Rate, and Overlap

NOTE: Negative signs indicate reductions in injury levels.

Table V-7
Potential Benefits from Preventable Crashes, All Passenger Vehicles
Adjusted for Properly Inflated Vehicles,
Delta-V Distribution, 80% Response Rate, and Overlap

Alternative 1

Non-Preventable Crashes

In the vast majority of crashes, small changes in stopping distance will not prevent the crash, but will reduce the speed at impact and thus the severity of the crash. As noted above, 3.6 percent of braking passenger cars and 3.2 percent of braking trucks could have avoided crashes with proper tire inflation. The remaining 96.4 percent of passenger car crashes and 96.8 percent of LTV crashes would still occur, but at a reduced impact speed. To estimate the impact of reduced crash speeds, changes in stopping distance will be estimated and used as inputs to recalculate impact speeds for the population of non-preventable crashes These changes in impact speeds will then be used to redefine the injury profile of this crash population shown in Table V-2, and safety benefits will be calculated as the difference between the existing and the revised injury profiles.

Stopping Distance

Stopping distance can be computed as a function of initial velocity and tire friction. The formula for computing stopping distance is as follows:

SD = Vi²/(2*g*Mu*E)

Where:

SD =Stopping Distance (in feet)

Vi = initial velocity (mean speed limit for specific data group + 5 mph)

g = gravity constant (32.2 ft/second squared)

Mu = tire friction constant (ratio of friction force/vertical load )

E = ABS braking efficiency (estimated @ 0.8)

About a third of all passenger vehicles sold in the U.S. do not have anti-lock brakes, although the portion is higher in the on-road fleet. For these regular braking systems, the term for anti-lock brake efficiency (E) would not be used.

Calculating Mu

The value of Mu is dependent on surface material (concrete, asphalt, etc.), surface condition (wet vs. dry), inflation pressure, and initial velocity. The Goodyear Tire and Rubber Company submitted a model they developed by testing tires under various circumstances that predicts Mu based on Vi and inflation pressure. Separate models were developed for Mu at both peak (the maximum level of Mu achieved while the tire still rotates under braking conditions) and slide (the level of Mu achieved when tires cease to rotate while braking (i.e., skid)). The models are as follows:

Ms = 0.2339537+(0.0034537*ip)+(0.0003625*Vi)-(0.000049*Vi²)

Mp = 0.4374907+(0.0024907*ip)+(0.003075*Vi)-(0.000095*Vi²)




Where:

Ms = Mu slide value

Mp = Mu peak value

ip = inflation pressure (psi)

V_i = initial vehicle speed (mph)

Mu Surface Adjustments

The above formulae were derived from tests conducted on a Traction Truck surface (this is a specific surface calibrated to specifications of the companies OEM customers). In order to relate them to real world surfaces, predicted values from the formulas were compared to actual test results on 2 surface types (asphalt and concrete). From this, a surface adjustment factor was obtained for each surface. For asphalt, the factor was 1.22. For Concrete, it was 2.00. Although most road surfaces are asphalt, the test surfaces tend to be slicker than roads that have experienced wear. NHTSA and Goodyear engineers both felt that the frictional qualities of the concrete test surface are most like those encountered on actual roads. Therefore, calculations of stopping distance will be based on the Concrete surface adjustment factor. The formulae thus become:

Ms = (0.2339537+(0.0034537*ip)+(0.0003625*Vi)-(0.000049*Vi²))/2

Mp = (0.4374907+(0.0024907*ip)+(0.003075*Vi)-(0.000095*Vi²))/2

The models provided by Goodyear were developed using wet traction test data, and are thus appropriate for wet surfaces only. Goodyear tested the tires with .05" of water on the track surface. This is more than would typically be encountered under normal wet road driving conditions and may thus exaggerate the stopping distances experienced under most circumstances. On the other hand, crashes are more likely to occur under more hazardous conditions, which may mean the Goodyear data are less biased when applied to the actual crash involved population With these caveats, this analysis assumes the data to be representative of the crash involved population on wet surfaces. To adjust for dry surfaces, NHTSA used data provided by Goodyear to develop models that predict adjustment ratios for dry surface conditions. The data on which these models are based is listed in Table V-9. The models take the following form:

DFs = -0.022778*ip+.0485*Vi+1.437222

DFp = -0.0075*ip+0.03225*Vi+1.0575

Where:

DFs = slide dry surface adjustment factor

DFp = peak dry surface adjustment factor

The formula for Mu peak and slide on dry surfaces thus become:

Ms = ((0.2339537+(0.0034537*ip)+(0.0003625*Vi)-(0.000049*Vi²))/2)*DFs

Mp = ((0.4374907+(0.0024907*ip)+(0.003075*Vi)-(0.000095*Vi²))/2)*DFp

Table V-9
Measured Mu Values by Surface Condition,
Speed and Inflation Pressure

Source: Goodyear Tire and Rubber Co.

Anti-lock and Normal Braking Systems

Roughly 2/3 of all passenger vehicles sold in the U.S. have anti-lock brakes, but the portion is smaller in the on-road fleet. For vehicles with anti-lock brake systems, Mp is used to calculate stopping distance because it represents the peak controlled braking force that anti-lock brakes attempt to maintain. For vehicles with regular brake systems, Ms is used because it represents the level of friction encountered under normal braking by most drivers without assistance from anti-lock brakes. Also, for these regular braking systems, the term for anti-lock brake efficiency (E) would not be used.

Delta-V

Changes in stopping distances were then used to calculate the decrease in crash forces (measured by delta-V) that would occur due to the decrease in striking velocity of the vehicle. The formula used to calculate striking velocity is:

V(d) =

Where:

V(d) = velocity of vehicle at distance d after braking

a = deceleration

d = distance traveled during braking of vehicle

In this case, V_(d) is a measure of the speed at which the vehicle with under-inflated tires would be traveling when it reaches the distance at which it would have stopped had its tires been correctly inflated (d). Deceleration (a) is calculated for the vehicle with under-inflated tires. The derived formula for deceleration is:

a = (V(d)²-Vi²)/(2*d)

Since V = 0 at d, the formula becomes:

a = (Vi²)/(2*d) (the negative sign that would precede the formula indicates deceleration and will be ignored from this point on)

The distance over which a is calculated is the stopping distance for the vehicle with under-inflated tires. This will be designated as SDu. The formula thus becomes:

a = (Vi²)/(2*SDu)

Where:

SDu = stopping distance with under-inflated tires

The striking velocity is then expressed in mph by multiplying by 1/ 5280 ft.*3600 sec. hour. The delta-V experienced by each vehicle would be dependent on vehicle mass. For this analysis, the mass of each vehicle was assumed to be equal, giving a delta-V of 1/2 V(d) for each vehicle or:

DELTA-V = (V(d)*3600/5280)/2

Where:

DELTA-V = the change in velocity resulting from increased tire pressure.

The base case target population represents the injury profile that results from the fleet of passenger vehicles that were on the road at that time. In order to determine the inflation pressure that exists in that fleet, NHTSA conducted a survey of both recommended and actual inflation pressures on vehicles. Details of that survey are discussed elsewhere in this analysis. The results of the survey indicate that 74% of all passenger vehicles are driven with under-inflated tires. However, because TPMS would not notify drivers of low pressure until it dropped 20% or 25% below placard, no stopping distance benefits would accrue to vehicles with smaller tire pressure deficits. Weighting factors were derived from the tire pressure survey to represent the affected population under each alternative. For Alternative 1, these weights were drawn from the population that had at least one tire 20% or more under-inflated. For Alternative 2, these weights were drawn from the population that had at least one tire 25% or more under-inflated. In the case of Anti-lock Brake systems under Alternative 2, the population was also restricted to cases where the maximum inflation pressure of any tire exceeded the minimum pressure by at least 25%. The distribution of each level of under-inflation is shown in Table V-10 for both Alternatives.

Table V-10
Percent of Vehicles Under-inflated Within Notification Levels

As noted previously, the value of Mu in the formula for stopping distance is dependent on inflation levels. For each speed limit category, a set of delta-Vs corresponding to each under-inflation level was calculated. In each case, an average placard pressure of 30 psi was assumed for passenger cars. For LTVs, an average pressure of 35 was assumed. The rates of under-inflation in Table V-10 were used to weight the change in delta-V that results from each corresponding psi under-inflation level to an overall weighted average change across all levels. The resulting changes in delta-V are summarized in Table V-11 for each passenger car and LTV target population category for ABS systems, non-ABS systems and combined systems, based on weighting factors representing the relative portion of the vehicle fleet that has Anti-lock brakes. Similar results are summarized for Alternative 2 in Table V-12. Note that these estimates do not reflect any impact for vehicles with inflation levels that are less than the assumed set point for the TPMS system. For Alternative 1, this analysis assumes a set point of 20 percent below the placard pressure, or 6 psi based on the assumption of a 30 psi recommended pressure. Benefits would only accrue to those tires that are more than 6 psi beneath their recommended pressure. For LTVs, benefits would accrue for those tires that are more than 7 psi beneath their recommended pressure. Alternative 2 assumes a set point of 25% below placard for non-anti-lock brake systems and this results in higher average delta-V changes for these systems under Alternative 2 than Alternative 1, due to the higher level of potential improvement within this more limited population of vehicles. However, for vehicles with anti-lock brakes, the systems would only operate in cases where the highest tire pressure exceeded the lowest tire pressure by 25% or more, and this less rigorous level of notification results in a lower average delta-V change under Alternative 2 than under Alternative 1.

Table V-11
Weighted Average Reductions In Delta-V
from Improved Tire Inflation Pressure

Alternative 1

Table V-12
Weighted Average Reductions In Delta-V
from Improved Tire Inflation Pressure

Alternative 2

Calculation of Safety Benefits

Safety benefits were calculated by reducing the delta-V for each injury by the appropriate level for each specific target population category shown in Tables V-11 and V-12. Functionally, the injury totals for each delta-V category were redistributed according to the injury probabilities of the reduced delta-V level. This resulted in a new injury profile. Totals for each injury severity category were then compared to the original injury totals to produce the net benefits from reducing delta-Vs. An example of the original target population distribution and the revised distribution is shown in Tables V-13 and V-14. Note that the revised distribution shown in Table V-14 represents a whole number delta-V change (in this case, 8 delta-V). Since actual average reductions were fractional, interpolation was used to calculate the results of the fractional reductions. These interpolated results are reflected in Table V-15. Table V-15 summarizes the results for all scenarios for passenger cars under Alternative 1.

Adjustments to Non-Preventable Crash Safety Benefits

A number of adjustments must be made to the benefit estimates in Table V-15. These include:

1)   Adjustment for crash braking distance distribution

2)   Adjustment for portion of vehicle fleet with no under-inflation or under-inflation less than notification level

3)   Adjustment for driver response

4)   Adjustment for target population overlap travel speeds would be about 11 percent of those based on maximum impact for passenger cars, and 10 percent for LTVs.

Table V-13
Passenger Cars, Original Injury Distribution
>=51 MPH Speed Limit, Wet Pavement

Table V-14
Passenger Cars, Modified Injury Distribution
>=51 MPH Speed Limit, Wet Pavement

Table V-15
Estimated Passenger Car Stopping Distance Impacts
Alternative 1, Unadjusted

Braking Distance Distribution

Table V-15 represents safety impacts that would occur from the reduced stopping distance of a tire at the point where it would stop if pressure were corrected. It represents the maximum change in delta-V that would occur in cases where the actual braking distance in the crash just equals the correct stopping distance. In reality, crashes occur over a variety of braking distances, and the change in delta-V is a direct function of this distance. This relationship is illustrated in Figure V-1 below. The change in delta-V is virtually non-existent in crashes where braking distance is minimal, but becomes significant as the distance traveled during braking increases.

Figure V-1
Generalized Relationship Between Change in
Delta-V and Traveling Distance

To account for the variety of possible outcomes, a factor was calculated based on the relationship between calculated delta-V changes and travel distance. The techniques used to calculate this factor are fully described in Appendix A. The results indicate that the impacts over the variety of travel speeds would be about 11 percent of those based on maximum impact for passenger cars and 10 percent for LTV's.

Properly Inflated Vehicles

As previously mentioned, 26 percent of all vehicles have no tires under-inflated. In addition, many vehicles have a level of under-inflation that would not trigger a warning from the TPMS. The target population used in the above calculations assumes a full fleet of under-inflated vehicles and must be adjusted for the portion of the fleet that is not under-inflated, and that will be notified of the problem. The portions differ by Alternative and vehicle type. Based on NHTSA's tire pressure survey under Alternative 1, only 36 percent of passenger cars and 40 percent of light trucks would potentially benefit from a TPMS. Under Alternative 2, 27 percent of passenger cars with anti-lock brakes, 26 percent of passenger cars without anti-lock brakes, 21 percent of light trucks with anti-lock brakes, and 29 percent of LTVs without anti-lock brakes would potentially benefit from a TPMS.

Driver Response

Table V-15 also represents the benefits that would accrue if all drivers responded immediately to the TPMS and inflated their tires to the proper level. Since this is unlikely to occur, an adjustment was made to represent the driver response rate. These rates vary for each alternative. For direct systems, a response rate of 80 percent is assumed. Eighty percent is chosen to represent a level that reflects the heightened consumer awareness that would come with systems that constantly monitor and display tire pressure levels. A lower response rate of 60 percent is assumed for indirect systems, which only provide information when the systems reach the set point. Since Alternative 1 involves only direct systems, the factor for that alternative is 80 percent. Alternative 2 involves both direct systems on vehicles with conventional brakes, and indirect systems on vehicles with anti-lock brakes. A weighted average of the two systems, 66.6%, was used for Alternative 2.

Overlapping Target Populations

As previously noted separate target populations were derived for passenger cars and light trucks because the under-inflation profile is different for these vehicle types. These populations were stratified based on the vehicle braking. However, a comparison of the two separate injury counts to a single count done for any passenger vehicle indicated that a small amount of double counting resulted from a simple addition of the two separate braking vehicle populations. Based on this comparison, an adjustment factor of .9685 was applied to the benefit estimates to eliminate the overlap.

The above 4 adjustments were accomplished by multiplying the results in Table 15 by factors of .11, .36, .80, and .9685. Similar adjustments were made for each vehicle type and Alternative. Table V-16 summarizes the total adjusted non-preventable crash benefits for passenger cars under Alternative 1. Table V-17 summarizes the benefits from non-preventable crashes under Alternative 1 for both passenger cars and LTVs. Table V-18 summarizes total benefits for all crashes and vehicle types under Alternative 1. Table V-19 summarizes total safety benefits for all crashes and vehicle types under Alternative 2. The results indicate a potential safety impact under Alternative 1 of 158 fatalities eliminated and roughly 21,000 nonfatal injuries prevented or reduced in severity from improved stopping distance. Under Alternative 2, an estimated 97 fatalities and 13,000 nonfatal injuries would be prevented or reduced in severity. Alternative 1 thus offers benefits that are potentially 60% higher than Alternative 2.

These estimates represent the upper bound of results based on the variety of test results currently available. As previously mentioned, other test data from NHTSA's VRTC indicate that stopping distance impacts may be insignificant. A lower range estimate of no impact is implied by the VRTC test results. Neither of these estimates can be considered to be a likely result because both are derived from test data that may be inadequate to represent real world crash situations. In Chapter III, the results from both Goodyear and VRTC tests are discussed. In tests conducted by Goodyear, significant increases were found in the stopping distance of tires that were under-inflated. By contrast, tests conducted by NHTSA at their VRTC testing ground found only minor differences in stopping distance, and in some cases these distances actually decreased with lower inflation pressure. The NHTSA tests also found only minor differences between wet and dry surface stopping distance. It is likely that some of these differences are due to test track surface characteristics. Moreover, the wet surface tests were conducted on a surface that was only sprayed with water. Given the unworn condition of the track, these tests may not have properly represented the slick conditions that result when road surfaces become wet. The Goodyear tests may also be biased. Their basic wet surface tests were conducted on surfaces with .05" of standing water. This is more than would typically be encountered under normal wet road driving conditions and may thus exaggerate the stopping distances experienced under most circumstances. On the other hand, crashes are more likely to occur under more hazardous conditions, which may mean the Goodyear data are less biased when applied to the actual crash involved population. Still, it is likely that the Goodyear tests represent a more extreme condition than would be expected under most wet driving circumstances. Thus, it is likely that the Goodyear tests produce estimates that overstate the impact of proper tire inflation pressure, while the VRTC tests produce estimates that understate these impacts. Although NHTSA is confident that the impacts lie within this range, there is no data to determine exactly where within this range the most likely impacts are. Therefore, the "best estimate" of impacts is assumed to be an average of the upper and lower estimate. These results are summarized in Tables V-20 and 21 below.

Table V-16
Estimated Passenger Car Stopping Distance Impacts
Adjusted for Properly Inflated Vehicles,
Delta-V Distribution, 80% Response Rate, and Overlap

Table V-17
Estimated Non-Preventable Crash Stopping Distance Impacts,
All Passenger Vehicles Adjusted for Properly Inflated Vehicles,
Delta-V Distribution, 80% Response Rate, and Overlap

Alternative 1

Table V-18
Total Estimated Stopping Distance Impacts, All Passenger Vehicles
Adjusted for Properly Inflated Vehicles,
Delta-V Distribution, 80% Response Rate, and Overlap

Alternative 1

Table V-19
Total Estimated Stopping Distance Impacts,
All Passenger Vehicles Adjusted for Properly Inflated Vehicles,
Delta-V Distribution, 67% Response Rate, and Overlap

Alternative 2

Table V-20
Mid-Point Estimate Total Stopping Distance Impacts,
All Passenger Vehicles Adjusted for Properly Inflated Vehicles,
Delta-V Distribution, 80% Response Rate, and Overlap

Alternative 1

Table V-21
Mid-Point Estimate Total Stopping Distance Impacts,
All Passenger Vehicles Adjusted for Properly Inflated Vehicles,
Delta-V Distribution, 67% Response Rate, and Overlap

Alternative 2

Fuel Economy Benefits

Correct tire pressure will improve a vehicles' fuel economy. Current radial tires are a vast improvement over the old-fashioned bias-ply tires, yet they still use more fuel when they are run under-inflated, although not as much as bias-ply tires. According to a 1978 report ⁽¹⁾, fuel efficiency is reduced by one percent (1%) for every 3.3 pounds per square inch (psi). More recent data provided by Goodyear indicates that fuel efficiency is reduced by one percent for every 2.96 psi, fairly close to the 1978 estimate.

For this analysis, we assumed that there was no effect of tire over-inflation, and that savings only started once the warning went on. In other words, if the placard pressure were 30 psi, and a warning were given under Alternative 1 at 24 psi (20 percent below placard), no benefits are assumed for those vehicles that have tires with lowest pressure above 24 psi. For Alternative 1 and 2, data from the tire pressure survey was used to estimate the average under-inflation of all 4 tires for those vehicles for which a warning would be given. Table V-22 provides the average under-inflation and the percentage of the fleet that would get a warning by the TPMS by alternative.

Table V-22
Analysis of Fleet Tire Pressure Survey

Tables V-23 and V-24 show the weighted vehicle miles traveled by age of vehicle for passenger cars and light trucks. They also show the 7 percent discount rate and the assumed price of gasoline. The projected price of gasoline was taken from a DOE projection from January 2001 ⁽²⁾. It excludes fuel taxes, at $0.38 per gallon, since these are a transfer payment and not a cost to society. Year 1 for these gasoline prices is estimated to be 2004, when the TPMS requirements will be in place. Obviously, these gasoline prices are much lower than the current prices at the pump ($1.70 in May 2001, or $1.32 excluding taxes). However, the projections are for gasoline prices to steadily decline from 2001 through about 2005 when they will level off.

Table V-23
Passenger Cars Vehicle Miles Traveled, Discount Factor, and
Assumed Price of Gasoline in (2001 Dollars)

Passenger Cars

Table V-24
Light Trucks Vehicle Miles Traveled, Discount Factor, and
Assumed Price of Gasoline in (2001 Dollars)

Light Trucks

The baseline miles-per-gallon figure for cars was 27.5 mpg at perfect inflation, and for light trucks was 20.7 mpg at perfect inflation. A sample calculation for passenger cars for Alternative 1 is:

The average of all four tires on a passenger car that would be warned based on our survey would be 6.1 psi lower than placard. Since 1 percent fuel efficiency is equivalent to 2.96 psi lower, the average passenger car with a warning would get 2.060811 percent higher fuel economy. With a baseline of 27.5 mpg, the average fuel economy of those vehicles warned that increased their tire pressure up to placard would be 27.5 * 1.02060811 = 28.0667 mpg. Based on our estimated vehicle miles traveled by age, scrappage by age, a 7 percent present value discount rate and estimated fuel costs per year, the baseline passenger car (at 27.5 mpg discounted by 15 percent to account for real on-road mileage) would spend $3,631.32 present value for fuel over its lifetime. Those drivers warned who filled up to placard pressure and achieved 28.0667 mpg (discounted by 15 percent to account for real on-road mileage) would spend $3,558.00 for fuel over their lifetime. The difference is $73.32. Since 36 percent of the fleet get a warning, and it is assumed that 80 percent of the drivers would fill their tires to placard, the average benefit is $21.12 ($73.32*0.36*0.80). The estimated benefit for each subgroup under the different alternatives is shown in Table V-25.

Table V-25
Fuel Economy Benefits Compared to the Baseline Fleet
Present Discounted Value over Lifetime
(2001 Dollars)

Weighting the Alternative 2 fuel economy benefit by the percent of the fleet with ABS-based systems (67 percent) and direct measurement systems (33 percent) results in an estimated $12.02 for passenger cars and $20.77 for light trucks. Weighting light trucks (50 percent) and passenger cars (50 percent) results in the following overall benefit in fuel economy shown in Table V-26.

Table V-26
Fuel Economy Benefits Compared to the Baseline Fleet
Present Discounted Value over Lifetime
(2001 Dollars)

Tread Life

Driving at lower inflation pressure impacts the rate of tread wear on tires. This will cause tires to wear out earlier than necessary and decrease tire life. When a tire is under-inflated, it puts more pressure on the shoulders of the tire and does not wear correctly. This analysis will attempt to quantify the impact of increased tread wear on consumer costs.

Based on data provided by Goodyear (see Docket No. NHTSA-2000-8572-26), the average tread life of tires is 45,000 miles and the average costs is $61 per tire (in 2001 dollars).

For Alternative 1

Assuming a direct measurement system, the TPMS warns the driver anytime a tire is 20 percent or more below the placard and the driver inflates all of the tires back to the placard levels, then we can estimate the impact on tread life using the following calculations.

Goodyear provided data estimating that the average tread wear dropped to 68 percent of the original tread wear if tire pressure dropped from 35 psi to 17 psi. Goodyear also assumed that this relationship was linear. Thus, for every 1 psi drop in inflation pressure, tread wear would decrease by 1.78 percent [(100-68%)/(35-17psi)]. These effects would take place over the lifetime of the tire. In other words, if the tire remained under-inflated by 1 psi over its lifetime, the tread wear would decrease by 1.78 percent or about 800 miles (45,000*0.178).

Data from our tire pressure survey indicated that 2,136 out of 5,967 passenger car tires (36 percent) had at least one tire under-inflated by 20 percent or more below the placard level. The average under-inflation of the 4 tires for these vehicles was 6.1 psi. Thus, on average, passenger cars lose an estimated 4,880 miles (6.1 * 800 miles) of tread life for each tire due to the way they are currently under-inflated that could be remedied under Alternative 1 if everyone filled all their tires back up to the placard pressure when they were notified by a TPMS. If we assume that 80 percent of the people actually inflate their tires properly, then on average about 3,900 miles of tread life would be saved per tire.

If the average current lifetime of tires is 45,000 miles at current inflation levels, the average lifetime could be 48,900 miles with a TPMS. The agency estimates that the average lifetime per passenger car is 126,678 miles. Thus, currently the average car would have 3 sets of tires on their car over its lifetime (new, at 45,000 miles, and at 90,000 miles) and with TPMS the average car would have 3 sets of tires purchased (new, at 48,900 miles, and at 97,800 miles). The benefit to consumers is the delay in purchasing those tires and getting interest on that money at an assumed 7 percent rate of return. Using a mid-year 7 percent discount rate, the discounted present value of these delayed tire purchases is estimated to be $14.62 for those passenger cars that would be notified by a TPMS that they are under-inflated. Since 36 percent would be notified, the present discounted benefits are $5.26 ($14.62 * 0.36) and 1,404 miles (3,900 * 0.36) of tread life.

For light trucks, data from our tire pressure survey indicated that 1,564 of 3,950 light truck tires (40 percent) had at least one tire under-inflated by 20 percent or more compared to the placard. The average under-inflation of the 4 tires for these vehicles was 7.7 psi. Thus, on average, light trucks lose an estimated 6,160 miles (7.7*800) of tread life for each tire due to the way they are currently under-inflated that could be remedied if everyone filled all their tires back up to the placard pressure when they were notified by a TPMS. If we assume that 80 percent of the people actually inflate their tires properly, then on average 4,930 miles of tread life would be saved per tire.

If the average current lifetime of tires is 45,000 miles at current inflation levels, the average lifetime could be 49,930 miles with a TPMS. The agency estimates that the average lifetime per light truck is 153,706 miles. Thus, the average light truck would have 4 sets of tires on their truck over its lifetime (new, at 45,000 miles, at 90,000 miles, and at 135,000 miles) and with a TPMS the average light truck would have four sets purchased (new, at 49,930 miles, at 99,860, and at 149,790 miles). Using the same methodology as for passenger car tires, the benefit in delaying purchasing tires is estimated to be a present discounted benefit of $42.00. Since in 40 percent of the vehicles at least one tire is under-inflated by 20 percent or more, the average benefit for light trucks is estimated to be $16.80 ($42.00 * 0.40) and 1,972 miles (4,930 * 0.40) of tread life.

For Alternative 2

We have to consider both ABS-based vehicles and non-ABS-based vehicles since they are represented by a different group of vehicles in the tire pressure survey. For Alternative 2, we assume that two-thirds (67%) of the vehicles would have ABS-based indirect measurement systems and one-third of the vehicles (33%) would have a direct measurement system. For the ABS-based vehicles we assume the TPMS warns the driver anytime there is a 25 percent or more psi differential between tires. For the non-ABS-based vehicles, we assume a direct measurement system will provide a driver warning anytime one or more tires is 25 percent or more below placard. If we assume the driver inflates all of the tires back to the placard levels, then we can estimate the impact on tread life using the following calculations.

For direct measurement systems

Data from our tire pressure survey indicated that 1,575 out of 5,967 passenger car tires (26 percent) had at least one tire under-inflated by 25 percent or more below the placard level. The average under-inflation of the 4 tires for these vehicles was 6.8 psi. Thus, on average, passenger cars lose an estimated 5,440 miles (6.8 * 800 miles) of tread life for each tire due to the way they are currently under-inflated that could be remedied if everyone filled all their tires back up to the placard pressure when they were notified by a TPMS. If we assume that 80 percent of the people actually inflate their tires properly, then on average 4,350 miles of tread life would be saved per tire.

If the average current lifetime of tires is 45,000 miles at current inflation levels, the average lifetime could be 49,350 miles with a TPMS. The agency estimates that the average lifetime per passenger car is 126,678 miles. Thus, currently the average car would have 3 sets of tires on their car over its lifetime (new, at 45,000 miles, and at 90,000 miles) and with TPMS the average car would have 3 sets of tires purchased (new, at 49,350 miles, and at 98,700 miles). The benefit to consumers is the delay in purchasing those tires and getting interest on that money at an assumed 7 percent rate of return. Using a mid-year 7 percent discount rate, the discounted present value of these delayed tire purchases is estimated to be $16.30 for those passenger cars that would be notified by a TPMS that they are under-inflated. Since 26 percent would be notified, the present discounted benefits are $4.24 ($16.30 * .26) and 1,131 miles (4,350 * 0.26) of tread life.

For light trucks, data from our tire pressure survey indicated that 1,148 of 3,950 light truck tires (29 percent) had at least one tire under-inflated by 25 percent or more compared to the placard. The average under-inflation of the 4 tires for these vehicles was 8.7 psi. Thus, on average, light trucks lose an estimated 6,960 miles (8.7*800) of tread life for each tire due to the way they are currently under-inflated that could be remedied if everyone filled all their tires back up to the placard pressure when they were notified by a TPMS. If we assume that 80 percent of the people actually inflate their tires properly, then on average 5,570 miles of tread life would be saved per tire.

If the average current lifetime of tires is 45,000 miles at current inflation levels, the average lifetime could be 50,570 miles with a TPMS. The agency estimates that the average lifetime per light truck is 153,706 miles. Thus, the average light truck would have 4 sets of tires on their truck over its lifetime (new, at 45,000 miles, at 90,000 miles, and at 135,000 miles) and with a TPMS the average light truck would have four sets purchased (new, at 50,570 miles, at 101,140, and at 150,710 miles). Using the same methodology as for passenger car tires, the benefit in delaying purchasing tires is estimated to be a present discounted benefit of $47.71. Since in 29 percent of the vehicles at least one tire is under-inflated by 25 percent or more, the average benefit for light trucks is estimated to be $13.84 ($47.71 * 0.29) and 1,615 miles (5,570 * 0.29) of tread life.

For ABS-based systems

Data from our tire pressure survey indicated that 1,622 out of 5,967 passenger car tires (27 percent) had a 25 percent or more tire pressure differential. The average under-inflation of the 4 tires for these vehicles was 4.9 psi. Thus, on average, passenger cars lose an estimated 3,920 miles (4.9 * 800 miles) of tread life for each tire due to the way they are currently under-inflated that could be remedied if everyone filled all their tires back up to the placard pressure when they were notified by a TPMS. If we assume that 60 percent of the people actually inflate their tires properly, then on average 2,350 miles of tread life would be saved per tire.

If the average current lifetime of tires is 45,000 miles at current inflation levels, the average lifetime could be 47,350 miles with a TPMS. The agency estimates that the average lifetime per passenger car is 126,678 miles. Thus, currently the average car would have 3 sets of tires on their car over its lifetime (new, at 45,000 miles, and at 90,000 miles) and with TPMS the average car would have 3 sets of tires purchased (new, at 47,350 miles, and at 94,700 miles). The benefit to consumers is the delay in purchasing those tires and getting interest on that money at an assumed 7 percent rate of return. Using a mid-year 7 percent discount rate, the discounted present value of these delayed tire purchases is estimated to be $8.84 for those passenger cars that would be notified by a TPMS that they are under-inflated. Since 27 percent would be notified, the present discounted benefits are $2.39 ($8.84 * 0.27) and 635 miles (2,350 * 0.27) of tread life.

For light trucks, data from our tire pressure survey indicated that 831 of 3,950 light truck tires (21 percent) had a 25 percent or more tire pressure differential. The average under-inflation of the 4 tires for these vehicles was 6.1 psi. Thus, on average, light trucks lose an estimated 4,880 miles (6.1*800) of tread life for each tire due to the way they are currently under-inflated that could be remedied if everyone filled all their tires back up to the placard pressure when they were notified by a TPMS. If we assume that 60 percent of the people actually inflate their tires properly (some of them might only fill some tires and not all of their tires), then on average 2,930 miles of tread life would be saved per tire.

If the average current lifetime of tires is 45,000 miles at current inflation levels, the average lifetime could be 47,930 miles with a TPMS. The agency estimates that the average lifetime per light truck is 153,706 miles. Thus, the average light truck would have 4 sets of tires on their truck over its lifetime (new, at 45,000 miles, at 90,000 miles, and at 135,000 miles) and with a TPMS the average light truck would have four sets purchased (new, at 47,930 miles, at 95,860, and at 143,790 miles). Using the same methodology as for passenger car tires, the benefit in delaying purchasing tires is estimated to be a present discounted benefit of $24.63. Since in 21 percent of the vehicles there is a tire pressure differential of 25 percent or more, the average benefit for light trucks is estimated to be $5.17 ($24.63 * 0.21) and 615 miles (2,930 * 0.21) of tread life.

In summary, assuming that half of the vehicle sales in the future are passenger cars and half of the sales are light trucks, the average present discounted value benefit for tread wear savings for Alternative 1 is $11.03 ([$5.26 + $16.80]/2) and 1,688 miles ([1,404 + 1,972]/2) of tread life. For Alternative 2, the average benefit for tread wear savings for direct measurement systems is $9.04 ([$4.24 + $13.84]/2) and 1,373 miles ([1,131 + 1,615]/2) of tread life. The average benefit for tread wear savings for the ABS-based indirect measurement system is $3.78 ([$2.39 + $5.17]/2) and 625 miles ([635 + 615]/2) of tread life. Assuming that 33 percent of the fleet uses the direct measurement system and 67 percent of the fleet has ABS, the average present discounted value benefit for tread wear for Alternative 2 is $5.51 ($9.04*0.33 + $3.78*.67) and 872 miles (1,373*.33 + 625*.67) of tread life.

There are other potential unquantified benefits of increasing tread wear. Some people would not have to purchase the last set of tires for a vehicle if they were going to scrap the vehicle soon, or if it were totaled in a crash shortly before they were going to purchase new tires. So, there will be cases where the total purchase price of tires $244 ($61 per tire * 4) will be saved. However, we can't estimate the frequency of that occurrence.

Unquantifiable Benefits

Under-inflation affects many different types of crashes. These include crashes which result from:

1. an increase in stopping distance,
2. flat tires and blowouts
3. skidding and/or a loss of control of the vehicle in a curve, like an off-ramp maneuver coming off of a highway at high speed, or simply taking a curve at high speed
4. skidding and/or loss of control of the vehicle in a lane change maneuver,
5. hydroplaning on a wet surface, which can affect both stopping distance and skidding and/or loss of control.
6. overloading the vehicle

The agency can quantify the effects of under-inflation in a crash involving the reduction in stopping distance. However, it cannot quantify the effects of under-inflation in the five other types of crashes. The primary reason that the agency can't quantify these benefits is the lack of crash data indicating tire pressure and how large of a problem these conditions represent by themselves, or how often they are contributing factors to a crash. The agency does not collect tire pressure in its crash data investigations.

There are many factors that influence crashes of these types. For blowouts, there is 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 the tire pressure and doesn't know whether these blowouts occur before the crash or during the crash.

For loss of control crashes, speed is the most critical factor. Excessive speed alone can cause a loss of control in a curve or in a lane change maneuver. Tread depth, inflation pressure of the tires, and road surface condition are the most notable of a long list of factors including vehicle steering characteristics and tire cornering capabilities that affect the vehicle/tire interface with the road. So, when under-inflation is a contributing factor to a crash, it is hard to know whether correcting this one problem area could result in the collision being avoided or reduced in severity. Certainly, reducing under-inflation is an important area and a move in the right direction. The following discussions describe how inflation pressure affects these crash types to the extent known.

Skidding and/or loss of control in a curve

Low tire pressure, as a result of under-inflation, generates lower cornering stiffness because of reduced tire stiffness. When the tire pressure is low, the vehicle wants to go straight and requires a greater steering angle to generate the same cornering force in a curve. The maximum speed at which an off-ramp can be driven while staying in the lane is reduced by a few mph as tire inflation pressure is decreased. An example provided by Goodyear shows that when all four tires are at 30 psi the maximum speed on the ramp was 38 mph, at 27 psi the maximum speed was 37 mph, and at 20 psi the maximum speed was 35 mph while staying in the lane. Having only one front tire under-inflated by the same amount resulted in about the same impact on maximum speed. But, the influence of having only one rear tire under-inflated by the same amount was only about one-half of the impact on maximum speed (a 1.5 mph difference from 30 psi to 20 psi).

The agency also has run a series of tests to examine the issue of decreases in tire pressure on vehicle handling. A 2001 Toyota 4-Runner was run through 50 mph constant speed/decreasing radius circles to see the effects of inflation pressure on lateral road holding. Figure V-2 shows the results of lefthand turns plotted from 0 to 90 degrees handwheel angle for tire inflation pressures varied from 15 to 35 psi. The data indicate to us that in on-ramps/off ramps, tire inflation pressure is a critical factor in vehicle handling. The graph shows how much friction the vehicle can utilize, in terms of lateral acceleration (g's), before it slides off the road. The more lateral g's the vehicle can utilize, the better it stays on the road. So, if you are going around an off-ramp and need to turn the wheel 50 degrees at 50 mph, you can utilize 0.27 g's at 15 psi, or you can utilize 0.35 g's at 30 psi.

Skidding and/or loss of control in a lane change maneuver

In a quick lane change maneuver, under-inflated tires result in a loss of stiffness, causing poor handling. Depending upon whether the low tire(s) are on the front or rear axle impacts the vehicle's sensitivity to steering inputs, directional stability, and could result in a spin out and/or loss of control of the vehicle.

Skidding and/or loss of control from hydroplaning

The conditions that influence hydroplaning include speed, tire design, tread depth, water depth on the road, load on the tires, and inflation pressure. At low speeds (less than about 50 mph), if your tires are under-inflated, you actually have more tire touching the road. However, hydroplaning does not occur very often at speeds below 50 mph, unless there is deep water (usually standing water) on the road. As you get to about 55 mph and the water pressure going under the tire increases, an under-inflated tire has less pressure in it pushing down on the road and you have less tire-to-road contact than a properly inflated tire as the center portion of the tread gets lifted out of contact with the road. As speed increases to 70 mph and above and water depth increases due to a severe local storm with poor drainage, the under-inflated tire could lose 40 percent of the tire-to-road contact area compared to a properly inflated tire. The higher the speed (above 50 mph) and the more under-inflated the tire is, then the lower the tire-to-road contact and the higher is the chance of hydroplaning.

Tread depth has a substantial impact on the probability of hydroplaning. If you make a simplifying assumption that the water depth exceeds the capability of the tread design to remove water (which most likely would occur with very worn tires), then an approximation of the speed at which hydroplaning can occur can be estimated by the following formula:

Figure V-2

Hydroplaning speed = 10.35 x inflation pressure ⁽³⁾

Under this assumption of water depth exceeding the capability of the tread design to remove water:

At 30 psi, hydroplaning could occur at 56.7 mph

At 25 psi, hydroplaning could occur at 51.8 mph

At 20 psi, hydroplaning could occur at 46.3 mph.

This is presented to show the relative effect of inflation pressure on the possibility of hydroplaning.

Overloading the vehicle

When a vehicle is overloaded, (too much weight is added for the suspension, axle, and tire systems to carry) and the tires are under-inflated, there is an increased risk of tire failures. This can result in a loss of control of the vehicle.

Non-quantified benefits

Property Damage and Travel Delay

TPMS will impact safety by reducing both the incidence and severity of crashes. When crashes are prevented, the property damage and travel delay that would have occurred are prevented as well. In a 1996 report ⁽⁴⁾, NHTSA estimated that property damage costs averaged over $3000 per crash and travel delay averaged $260 per crash ($1994). These savings would accrue to crashes prevented by TPMS. However, most benefits from TPMS would accrue from crashes that still occur but with a reduced severity. It is unclear what the impact would be on travel delay and property damage from these reductions.

1. ¹ Evaluation of Techniques for Reducing In-use Automotive Fuel Consumption; The Aerospace Corporation, June 1978. Original reference from Goodyear, pp 3-45.

2. ² DOE Energy Information Administration, Annual Energy Outlook 2001, Table A3, Energy Prices by Sector.

3. ³ "Mechanics of Pneumatic Tires" edited by Samuel K. Clark of the University of Michigan, published by NHTSA, printed by the Government Printing Office in 1981.

4. ⁴ "The Economic Cost of Motor Vehicle Crashes", 1994, DOT HS 808 425, NHTSA, July 1996.