Road safety budget optimisation model for the state of Bahrain

• Main Author: Al-Zayani, Salman Rashid
• Other Authors: Benyon, John
• Published: University of Leicester 1992
• Subjects: 301
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.639389

The aim of this study is to develop a procedure that can be used for the optimal allocation of funds available for road safety remedial measures in the State of Bahrain. The study reviews the nature of the road accidents problem in Bahrain in sufficient detail to highlight the relative magnitude ·and seriousness of the problem. Factors comprising the total (comprehensive) cost of road safety remedial measures with special application to Bahrain (including the capital and maintenance costs of the remedial measures and accident component of road user costs) are identified and analysed. A methodology for identifying the most hazardous locations on the road network is then developed. Using this methodology preliminary and primary rankings of sites are determined. 70 hazardous locations are identified and analysed. A programme of 26 countermeasures is then drawn up and the effectiveness of 10 different types of alternative improvements is evaluated with special emphasis on the regression-to-mean effect. The other 16 countermeasures effectiveness estimates are adopted from other sources. A Budget Optimisation Model Computer Programme (BOM) is then developed using multi-stage dynamic programming for selecting an optimal set of schemes under a single-period budget constraint. Within the constraint that it must choose only one alternative for each location, BOM selects a group of alternatives that produces an optimal overall net present value (NPV) within the resource constraint level. BOM is then used to determine the optimum combination of improvement of the 70 sites with different budgets. In all budgets, BOM searches throughout the list of projects for those alternatives which would provide the greatest NPV. Projects with different budgets are clearly economically viable and viability is further demonstrated by the sensitivity tests which are subsequently carried out. Finally, it is shown that improvements selected by a dynamic programming can yield a higher return for a given budget than those chosen entirely on the basis of benefit-cost ratio.