6 Road network rehabilitation programme
114 6 Road network rehabilitation programme 6.1 Introduction 6.1.1 Strategic focus The Master Plan has concluded that roads will continue to be the dominant domestic mode only roads can provide access to all parts of the country and many trips made by other modes will inevitably use roads for their initial and final stages. The Master Plan has consequently concluded that Liberia should: rehabilitate the all-weather strategic primary network 41 that links Monrovia to the county capitals and to its main border points with Sierra Leone, Guinea and the Ivory Coast including pavement if funds are available; and construct two new links that could save substantial travel costs reduction: these are Buchanan totappita and Buchanan to Harper, on condition that these new development will not negatively impact resources to maintain existing infrastructure assets in maintainable condition. Addition links may be identified, provided that sufficient funding is provided to keep existing infrastructure in maintainable condition and these links can be justified on the basis of technical survey studying the financial, economic, social and environmental impacts. Given the high costs of current travel and the level of frustrated demand that is believed to exist, the Master Plan proposes to have this rehabilitated network completed as rapidly as possible and a five year timetable is proposed for re-establishing basic functionality. To speedily restoring the living conditions in Liberia as soon as possible, the National Transport Master Plan aims at providing access to all sections of the population and facilitating safe, comfortable transport at a design standard of 60 miles per hour or lower depending on the classification of road on the primary, secondary and feeder road network. The National Transport Master Plan adopts a flexible approach. With the focus of transport policy centred on the need for speedy improvements, Government of Liberia chooses to take optimum use of its limited resources. Provided funding is available, the National Transport Master Plan proposes to apply design standards at an economic optimum. In case of budget constraints, the National Transport Master Plan suggests to select design standards at the least life-cycle costs. A summary of the general 5 years Road Rehabilitation Plan and the 10 years Maintenance Program is shown in Table 9-1. 6.1.2 Introduction to the Roads Economic Decision (RED) Model The principal tool adopted for the prioritisation of these investments has been the World Bank s Roads Economic Decision Model (RED). The RED Model User Manual 42 (see appendix F on CD) states that the model was developed to improve the decision-making process for the development (...) of low-volume rural roads. The model performs an economic evaluation of road investment options using the consumer surplus approach and is customised to the characteristics and needs of low-volume roads such as the high uncertainty 41 The base network which has been analysed to date consists of approximately 2,300 km of mainly primary (but some secondary) roads. The full picture will, however, only emerge when data has been collected and analyses have been undertaken of the full primary and secondary networks. 42 Sub-Saharan African Transport Policy Program (SSATP) Working Paper 78, The Roads Economic Decision Model (RED) for the Economic Evaluation of Low Volume Roads Software User Guide and Case Studies ; Rodrigo Archondo-Callao.
Road network rehabilitation programme of the assessment of model inputs, particularly the traffic on and condition of unpaved roads, the importance of vehicle speeds for model validation, the need for a comprehensive analysis of generated and induced traffic (...) RED computes benefits for normal, generated, induced and diverted traffic and takes into account changes in road length, condition, geometry, type accidents and days per year when the passage of vehicles is further disrupted by a highly deteriorated road condition (wet season) (...) The model is presented on a series of Excel 2000 Workbooks that collects all user inputs, presents the results in an efficient manner and performs sensitivity, switching value and risk analyses. The model presents its results in the form of investment programme which initially assumes an Unconstrained Budget. The user has, however, the option to develop alternative investment programmes/options based on up to five constrained (fixed) budgets. With the limited data-base currently available in Liberia, RED is, thus, a highly appropriate tool for prioritising road rehabilitation works. 115 6.2 RED input data: the Liberian context 6.2.1 General RED requires: link-by-link road description data; link-by-link road traffic data; vehicle utilisation and unit price data; estimates of the values of passenger-time, cargo-holding times and accident costs; capital and maintenance costs for a do nothing base option and for a series of up to six rehabilitation options; and some additional data pertinent to general conditions in Liberia and the nature of the analyses to be undertaken. In the sections below, the project-specific approach to the gathering and analysis of this data is reported. The RED User Manual (appendix F on CD) should be referred to for definitions and for a description of how RED manipulates the data that is input. 6.2.2 Link-by-link road characteristics In the project context the most important data are: Link lengths: in most cases these were project GPS measured lengths and the information used in these RED analyses therefore supersedes information held on previous Liberian data, i.e. the National Roads Register. Existing and future widths: in most cases these were as measured during the on-site project data collection, RCS, inventory. Terrain description: RED allows three different categories to be defined for modelling purposes in the Liberian context these have been described as flat, rolling and mountainous with each link assigned to a specific category.
116 Road network rehabilitation programme Present and future surface types and conditions derived in terms of: surface type: RED allows three different categories to be defined for modelling purposes in the Liberian context these were paved, gravel and earth with each link assigned to a specific category. the International Roughness Index (IRI): which is measure of the average extents of all rises and all falls experienced by the vehicle s suspension per km of travel. As no roughness measuring devices were available to the project, the Roughness of each link was determined via a proxy. Usefully RED can apply algorithms that convert comfortable driving speed to an approximation of roughness on a link-by-link basis. 6.2.3 Link-by-Link road traffic data In general RED requires dry and wet season base present annual average daily traffic volumes for each of up to nine different vehicle categories. The user selects from sixteen predetermined motorised and four predetermined non-motorised vehicle categories that best describe the categories for which classified count data is available. For the project, eight of the RED-defined categories (all motorised) were used for modelling, viz; 1. motorcycle; 2. car small; 3. car medium (taxi); 4. bus mini; 5. bus medium 6. four-wheel drive; 7. truck medium; 8. truck articulated; In addition, the following data were necessary: separate annual traffic growth forecasts for each of the above-listed vehicle categories which can be adjusted upwards or downwards at the end of each five-years in the analysis period; and estimates for the volumes of generated and/or induced traffic. A brief description of the approach to these inputs is provided below. Traffic volumes: dry and wet season base volumes For the two seasons, the RED applied base traffic volumes derive from a number of sources each of which is described below: When both dry and wet season classified count data were available: both data sets have been used in the RED analyses the RED-estimated economic returns, in these instances, can consequently be considered most reliable. When only wet or dry season classified count data were available: the missing seasonal traffic data-set has been estimated from (i) the one data-set available and (ii) from generalised information on the ratio between wet and dry season traffic.
Road network rehabilitation programme When neither wet nor dry season data available but there was some old 1981 data (from the GIZ Economic Development Plan for Liberia, 1983) where the present traffic has been estimated by increasing the 1981 volumes by the average annualised growth rate since that date 43. When neither wet nor dry season data was available and there is no 1981 data: estimates for present traffic has been made by either: moving observer counts: as this is the most crude of all the methods used, the RED-estimated economic returns must be treated with most caution; or inputting a nominal single vehicle in each season: in this case the link is included in the analyses only in order to allow future estimates of economic returns when proper traffic data becomes available the RED estimated economic returns for these links are not meaningful. 117 Base traffic volumes: adjustments for night-time traffic The results from three night-time (18:00 to 06:00) classified count surveys are summarised on Table A.65 in Appendix A (see attached CD). On the same table, adjustment factors to convert the 12-hour counts to 24-hour estimates of AADT are also provided. Base traffic volumes: traffic growth Traffic is forecast to grow at the rates derived on Table 5-10 and takes account of the forecast growth in GDP, forecasts of population growth and the elasticity between growth in GDP and growth in travel by each vehicle type. The forecasts are, therefore, different for each vehicle type. Base traffic volumes: summarised traffic inputs A link-by-link summary of the RED-modelled dry and wet season base-traffic volumes and of the numbers of days assumed per year for each season is provided in Appendix N. Also shown on this table are the link-by-link applied: night-time correction multipliers for different vehicle types (multipliers for links in and around Monrovia and away from Monrovia are shown separately); seasonal correction multipliers: multipliers for stations with only wet season counts are separated from those for stations with only dry season counts; and multipliers for links in and around Monrovia and away from Monrovia are shown separately. when appropriate, the approaches adopted to forecasting for links with neither dry nor wet season counts are indicated. Generated traffic There is insufficient data to make separate link-by-link estimates of generated traffic. As there is believed to be much frustrated demand, an elasticity of demand with price of 1.0 has been applied. This means that for every 10% reduction in cost there will be a corresponding 10% increase in travel. On links where costs are reduced the most, the volume of generated traffic will consequently also be highest. As the model predicts reductions in cost of the order 30-40%, there is likely to be much generated traffic. This assumption in regard to generated traffic should, however, be revisited after data on traffic before and after rehabilitation becomes available. 43 This information is provided in appendix D on CD. As the traffic forecasting method is crude, the RED-estimated economic returns for these roads must be treated with caution.
118 Road network rehabilitation programme 6.2.4 Vehicle utilisation and unit price data Comprehensive vehicle utilisation and unit price data is required by RED. The data derives from a number of sources within Liberia and also from neighbouring countries. As and when better locally-derived data becomes available, this should replace the data presently being applied. 6.2.5 Estimates of values of passenger times, cargo holding times and accident costs The applied values of passenger time are all based on working time. For the purposes of the RED evaluations, it has been assumed that the majority of passenger trips have ultimate origins or destinations in the Monrovia region and that most passenger trips are generated by higher income groups. It has been assumed that small car and 4-wheel drive vehicle passenger trips are generated by the highest income groups, groups with an average annual income of about USD 2,000 (or about USD1 per hour (40 hrs per week; 50 weeks per year); taxi passenger trips are assumed to be generated by groups with an average income of about USD 1,200 per year (60 US cents per hour), bus and truck passenger trips are assumed to be generated by groups with an average income of about USD 1,000 per year (50 US cents per hour). As it is noted that the tariffs for motorcycle (taxi) trips are higher than those for motor-car taxi trips the average income has been estimated as USD 1,500 per year (75 US cents per hour). Cargo holding times are relevant only when there is much freight that has a high per tonne value (for example, manufactured electronic goods) and/or which is perishable. This Master Plan, as is the case with most economic appraisals for road investments with low traffic volumes, does not incorporate this data. Insufficient data on the present severity, cost and frequency of accidents on roads of different standards was available and the incorporation of accident benefits was unnecessary in order to justify the rehabilitation programme. The incorporation of this data should, however, be considered when data does become available. It is, however, unlikely to significantly affect the priority order of a rehabilitation programme. 6.2.6 Capital and maintenance costs for a do nothing base option and for the rehabilitation options Expected capital rehabilitation and recurrent maintenance costs are provided on Table N-1; Table N-2, Table N-3, Table N-4, Table N-5 and Table N-6 in appendix N (see attached CD). 6.2.7 Additional data pertinent to general conditions in Liberia and the nature of the analyses The analysis is based on: an opportunity cost of capital of 12% which consistent with the appraisals for other sectors; an analysis/benefit stream of ten years that is consistent with the appraisals for the other sectors and appropriate given that the appraisal focus on emergency rehabilitation ; 2010 constant prices and a 2010 base year; A Roads Agency economic cost factor of 0.9 defined as the assumed ratio between economic and financial costs applicable to the proposed investments in rehabilitation and maintenances. In its simplest form this ratio would be that which derives when taxes are removed from the costs,
Road network rehabilitation programme though it is often appropriate to value resources in over-supply (particularly unskilled labour) at rates that are much less than local-market costs and to value resources which are in short-supply, (frequently foreign-sourced items) at rates that are higher than local-market costs. Given the expected high rates of return and the present emphasis on priority ordering this initial simplification is considered acceptable. 119 6.3 Modelling conclusions - rehabilitation programmes The RED analyses have indicated that a substantial rehabilitation programme and the construction of two missing links would have substantial economic benefits. A five-year unconstrained emergency rehabilitation investment program is recommended by RED and is set out on the following pages. The required total expenditure over the emergency rehabilitation programme s projected fiveyear life would be about USD 172 million (see Table N-1 in Appendix Y). Additionally, about USD 18 million per year would have to be spent on recurrent maintenance to ensure that the network does not return to its present unacceptable condition. The construction of the two missing links would add another about USD 10 million. Five-year rehabilitation programmes based on total 5-year emergency rehabilitation reduced budgets of USD 100 million, USD 75 million, USD 50 million, USD 17 million, USD 10 million and, in the last case, just USD 5 million are shown on Table N-2, Table N-3, Table N-4, Table N-5 and Table N-6. An examination of Table N-2, Table N-3, Table N-4, Table N-5 and Table N-6 shows that as the available budgets fall: increasing proportions of the roads are recommended to be initially rehabilitated at reduced standards; and the rehabilitation of some other roads, where only basic standards are proposed fall out of the initial programme. A preliminary economic justification of the two missing links is provided in appendix G on CD. These are the routes: from Monrovia via Buchanan to Tappita (and thence to the Ivory Coast Border); and from Monrovia via Buchanan along the South Coast to Harper, Fishtown and Barclayville. The estimated cost of providing both links is USD 9.9 million. As the estimated economic returns are very high, these roads would ideally be constructed simultaneously with the emergency rehabilitation programme. Additional investments to bring the network to full, economically-justified, standards must, of course, be applied in a post-emergency rehabilitation follow-on period.
120 Road network rehabilitation programme Figure 6-1: RED scenario unconstrained budget
Ministry of Transport Parker House / Broad Street Monrovia, Liberia Ministry of Public Works Lynch Street Monrovia, Liberia