Optimal Pavement Design and Maintenance Strategy for Developing Countries: An Analysis using HDM-4

Abstract

This paper investigates the relationship between optimal pavement design and maintenance strategy and the level of economic development (LED), using HDM-4. The LED affects the trade-off relationships between initial design and subsequent maintenance, and between road-user costs and agency costs. It was found that pavement strategy should be more economical in developing countries, both for the initial design and the subsequent maintenance. The extent to which pavement should be designed to be sturdier, in order to counteract the insufficient maintenance practices which are so prevalent in developing countries, was also quantified. Based on these results, a formula was derived that predicts the optimal pavement strength as a function of axle loading, LED and level of maintenance insufficiency. The pavement strengths predicted by the formula and determined by applying standard pavement design guidelines were compared, and HDM-4 was judged to be an appropriate design tool for pavements, if properly calibrated.

Keywords:

• PMS
• Developing countries
• HDM-4
• Pavement design
• Appropriate technology

Acknowledgements

The authors would like to thank Dr Shishido at IMF for his advice on the relationship between social discount rate and the level of development of an economy, and Guan Changyu, deputy chief engineer of the Transport Planning and Research Institute, Ministry of Communications, P.R. China, for providing the data used in this study. The second author would like to thank the Ministry of Education, Science, Sports and Culture of the Government of Japan for supporting his studies at Saitama University under the Monbukagakusho scholarship program.

Notes

^†Tel.: [phone]. Fax: [phone] E-mail: [email protected]

^‡Note that HDM-4 does not calculate CCs internally. These and MC unit costs are provided by the user.

^¶Tsunokawa et al. (2002) discus an efficient procedure for saving much computational effort in a similar search by using common knowledge of pavement management.

^§This is an approach that should be employed when using HDM-4 as a tool for pavement design. Several authors such as Kristiansen et al. (1996) have claimed that HDM-III (Watanatada et al., 1987), and presumably its successor, should not be used as a design tool without being specific about its usage, but, as will be shown later, HDM-4 was found to be promising as a good design tool, if properly calibrated.

^∥Among these it was most annoying that the total agency cost (i.e. the sum of the initial construction cost and later maintenance cost) in current values was found to be larger in developing countries than in developed countries.

^#The traffic composition may differ greatly by country and type of roads. Readers are referred to the last paragraph of the paper.

^**It would be more realistic to include a wider range of maintenance works such as patching, crack sealing, resealing, etc. but the simplification of considering only overlays was deemed appropriate for obtaining useful insight of broad applicability. The absence of resealing in the maintenance option is partly compensated by the inclusion of very thin overlays.

^††Reliable data on social discount rates are extremely scarce for developing countries, resulting in the fact that only two such countries are listed in the table.

^‡‡In Table VIII all values are shown in current values, because the conjectures to be examined that were proposed at the outset are also in current values.

^¶¶They include traffic loading, environmental conditions, construction and maintenance costs, etc.

^§§Pavement sections were selected using cumulative ESAL, obtained by using ESAL factors and AADT listed in Table I, for a design life of 10 years and sub-grade CBR of 6%. The SNP values were calculated assuming the same material types as described in the footnotes of Tables III and IV for the three layers.

^∥∥The design Structural Number (SN) was calculated by using cumulative ESAL, obtained using ESAL factors and AADT listed in Table I, for a 10-year performance period. Typical values for other parameters (combined standard error of traffic prediction and performance prediction=0.45; initial serviceability=4.2 PSI; and terminal serviceability of 2.5, 2.25 and 2.0 PSI for high, medium and low traffic, respectively), as suggested by AASHTO (1986); Huang (1993), were used. The calculated SN number was used to select layer thicknesses by trying the combinations listed in Table V. Effective roadbed soil resilient modulus ( M_{R} = 9000 psi and layer coefficients ( a_{1} = 0.32, a_{2} = 0.25 and a_{3} = 0.15 ) were calculated for materials described in the footnotes of Tables III and IV with the aid of the charts and relationships listed in HDM-III (Watanatada et al., 1987; Huang, 1993). Material properties described in the footnotes of Tables III and IV reflect Chinese practices and may vary among different countries because of wide variations in construction materials and construction technology.

^##In addition to the fact that HDM-4 relationships were not “calibrated” in this study, part of the discrepancy in the strength values may be explained by the fact that while design strengths were obtained by assuming a 10-year pavement life (which is the shortest allowable life in these manuals), overlays were allowed to be applied from the fifth year after construction in the HDM-4 runs.

^***Readers are advised to use usual caution on the use of the statistic relationship by extrapolation.

^††† Figure 1b illustrates that a number of strategies may be found for a given MC value.

^‡‡‡Readers are advised to use usual caution on the use of the statistic relationship by extrapolation.

^¶¶¶Note that the curves are not those fitting exclusively to the points appearing to the same figures but those representing Eq. (2) that are estimated with much more data.