CHAPTER 7: INTEGRATING SKID RESISTANCE AND TIRE/ROAD NOISE
7.2.2 Quantification of Safety and Comfort Benefits
In order to quantitatively evaluate the functional benefits brought forth by the application of porous surface layers, appropriate indices have to be developed to represent the advantages of porous pavements in skid resistance enhancement and tire/road noise reduction. It is difficult to employ an clear-cut benefit estimate technique, not only because studies to establish the relationships between pavement friction and travel safety is still undergoing, but also due to the difficulty in assigning a monetary value on the comfort and health effects of traffic noise (Ahammed and Tighe, 2010). The perception of noise annoyance depends not only on physical attributes, but also to a certain extent on subjective parameters and socio-culture environment (Martín et al., 2006). Therefore, a compromise has to be reached based on empirical relationships raised from past research studies and subjective estimations on the demands of each individual project.
267 7.2.2.1 Estimation of Safety Benefit
The benefits of skid resistance improvement may be estimated according to the potential reduction in the count and severity of wet-weather crashes. Although the relationship between surface friction and crash risk is difficult to quantify, past studies have provided some useful insights on the decreasing trend of crashes with the increase of wet-pavement friction. The road research group of the Organisation for Economic Cooperation and Development (OECD, 1984) revealed a linear crash- friction relationship with a decreasing rate of about 0.045 acc/MVM per unit increase of skid number at 40 mph (see Figure 7.1). Nonlinear relationships have also been suggested from other research studies (McLean and Foley, 1998; Wallman and Astrửm, 2001) (see Figure 7.2). Because many other factors contribute to crashes as well (such as road geometry, vehicle speed and traffic condition), it should not be expected to accurately predict accident frequency from skid resistance alone (Henry, 2000). In a statistical analysis of the effect of wet-pavement friction on highway safety, Ivan et al. (2012) developed crash-friction models for various roadway and operation characteristics (curve classification, grade, intersection, driveway, shoulder width, rural/urban area, and speed limit). The negative binominal regression was adopted to estimate model coefficients. Table 7.3 illustrates some typical results of exponents of consolidated SN40 coefficients, which indicate the multiplicative factor of crash count variation for a unit increase in SN40 measurement. The above research findings may be used to estimate the safety benefits obtained from skid resistance improvement, but one should be very careful to apply such empirical relationships because no consensus have been reached among the pavement engineering practitioners.
Economic appraisal is considered as a formal practice in the roadway safety management process developed by the Federal Highway Administration (FHWA, 2013). For each potential porous mixture, the installation and maintenance costs over
268
its service life are first computed. The annual crashes on the existing pavement is then estimated based on historical observed crash frequency. AASHTO (2010) proposed several methods for predicting roadway safety, such as the long-term average value, safety performance functions and empirical Bayes method. The estimated annual crashes on the existing surface is multiplied by the crash modification factor (CMF) obtained from past researches, based on the skid resistance improvement, to predict the annual crashes on the finished porous pavement. Crash reduction is derived from the values with and without porous layer. The estimated annual crash reduction is next converted to a monetary benefit by multiplying appropriate average crash costs.
Many agencies developed their own crash costs. Table 7.4 illustrates an example, by severity level, based on an FHWA report (Council et al., 2005). The monetary benefit in a given year is converted to present value by multiplying the present value factor.
With these, the estimated annual benefit can be summed and divided by the present value of treatment cost to estimate the benefit-cost ratio. The above steps should be repeated for each candidate mixture design and the extreme cases. The extreme cases are used to define the bounds of a feasible solution range and the benefit-cost ratio for each design corresponds to a value within this range.
7.2.2.2 Estimation of Noise Reduction Benefit
The valuation of tire/road noise reduction is not as straightforward as that of skid resistance improvement. The benefit of noise abatement is a more complicated problem involving people's health and comfort, environmental concern, real estate price near the roadway, as well as impact on working productivity. It is quite difficult to include all these aspects into a single analysis and it is impracticable to represent some effects in monetary values. Therefore, there is no economical model that enables the computation of overall benefit/cost for the whole society associated with noise exposure. The existing research studies on noise reduction benefit are focused
269 on a particular aspect and quantify the benefit using either cost of abatement, cost of illness, contingent valuation or hedonic price method (Becker and Lavee, 2003).
Cost of abatement was used as the minimal estimate of noise damage in some early studies. The benefit from a noise-reducing measure must be at least equal to the price paid for this action. This approach is a very rough estimation in the case where no better alternatives are available. It actually ignores the true value of the benefit, which is probably higher than the cost involved. For a multi-purpose measure, the benefits may be much more than just noise abatement. Cost of illness is another approach to estimate noise damage. It makes use of health expenditures to measure the economic cost associated with hearing loss caused by environmental noise. It is also quite difficult to obtain reliable results by this method because there is no specific evidence providing a correlation between hearing loss and traffic noise contribution. Contingent valuation method (CVM) is a very popular technique in transportation policy analyses, which is also widely used in the valuation of noise abatement. This method is developed based on a "willing to pay (WTP)" concept, which estimates the amount of money a representative sample of the public would like to pay to achieve a given amount of noise reduction. This is a closer estimation of the subjective value on noise abatement of each individual, but problems lie in the residents' general lack of familiarity in the "given amount of noise reduction". The WTP value varies dramatically from person to person and it is difficult to cover a sufficient sample of population with extensive socioeconomic backgrounds. Another popular technique in estimating the true value of noise reduction is the hedonic price method (HPM). The real estate price is commonly used as a proxy of the WTP for a quieter apartment in this approach, taking the assumption that the attribute vector determines the price and noise is one such attribute. The separate effect of noise reduction on the real estate price is then evaluated as the monetary benefit. HPM is more objective compared to CVM because the real estate price comes from the
270
market instead of individuals. The limitations of this approach lies in the assumptions of a perfect equilibrium in the housing market and the full awareness of information among participants.
Although no existing method is able to accurately valuate the overall noise abatement benefit, many useful findings have been achieved in past research works.
Becker and Lavee (2003) used HPM to evaluate the benefits brought forth by a new noise standard in Israel with a linear noise depreciation sensitivity index, and found a 1.2% increase of average urban property value per 1 dB noise reduction. The increasing rate in rural areas was found to be 2.2%. Martín et al. (2006) examined the relationships between traffic noise exposure and annoyance in Spain and found a linear correlation between annoyance and the measured noise level (see Figure 7.3).
The CVM was also adopted in the same study and found that 50% of the population was willing to pay 7.22 € on average per person per year to reduce noise contamination. Arsenio et al. (2006) used the stated choice model to valuate the road traffic noise. Some typical results of household monthly valuations for one unit change in noise level are illustrated in Table 7.5. The Department for Environment, Food and Rural Affairs of UK published a guideline on economic analysis of noise pollution (DEFRA, 2013). The marginal value of noise impact was estimated in this guideline to convert changes in noise exposure to monetary values, which could then be incorporated into a cost benefit analysis. Table 7.6 provides the values specified in this guide.
The noise measurements in the above studies are basically the far-field sound pressure level detected in the residential along the roadways. It is difficult to calculate this far-field noise level from the CPX near-field noise level obtained from numerical simulation model because of the variations in traffic volume, traffic composition and distance between vehicle tire and sound receivers. However, the essential concept of these studies are still applicable, considering the fact that far-field noise is a
271 combination of all the near-field sources after propagation. Based on the extensive review of past researches on the valuation of noise abatement benefit (Becker and Lavee, 2003; Arsenio et al., 2006; Martín et al., 2006; DEFRA, 2013), it is concluded that linear and geometric relationships are commonly found between noise variation and its worth. Regression curves were developed and the noise level associated with each feasible mixture design can be related to a benefit value on this curve.