The cost of a rural road network is more difficult to estimate than that of typical rural infrastructure such as wells, buildings, or small-scale irrigation projects since roads are subject to ongoing and often severe damage by users, the natural environment and often unpredictable interactions between the twoo. Roads are also expensive to maintain for the same reasons. Furthermore, local responsibility for maintenance, effective for village or community infrastructure, and just as necessary for remote rural road networks, is difficult to promote. Since roads serve many people over a wide area, responsibility is too diluted and local people, as the history of roads management teaches us, refuse reasonably enough to pay so that others often better off than them can benefit. Unfortunately, the public access to actual rural road cost data, and therefore the ability to question and evaluate funds spent on behalf of them and for their benefit, is opaque at best and deliberately denied at worst.
Factors which determine cost
The work to be done on a road or route, and its cost, is determined by its present state as compared to the kind of road that is wanted. Present condition could be anything from a simple track to an already engineered but now seriously degraded road. Something better is required, but how much better can be afforded? Many questions clamour for answers. What size, types, and numbers of vehicles must it carry and at what speed and level of comfort? Will they be adequately served by spot improvements of the worst sections rather than complete upgrading? How much wider must it be and can we dispense, now and then with shoulders for pedestrians and parking? What are the environmental factors and available materials that will affect costs? Must the road be kept open even when it is raining heavily and at its most vulnerable or can it be occasionally closed until it dries out sufficiently to safely bear the traffic? What work must be done to ensure its stability under the combined forces of traffic, terrain and rain? Can gravel (if it is available fairly close by) be used instead of a more expensive bitumen-sealed surface? Can a gravel surface be dispensed with completely? Can other materials be substituted for gravel or bitumen?
The environment is a cruel adversary. It does not forgive poor design, and traffic is its willing ally. Furthermore, the road, once improved, will tend to revert to its original state, and mitigating measures to protect it must be built into the final cost. Rainfall annual precipitation and storm intensity are perhaps the greatest determinant of road deterioration rate and degree. Where there are slopes, rapidly flowing water gouges out deep ravines along the road and heavy traffic further deepens them. At undrained low points in the road alignment a sea of mud can be created, to be further deepened by those vehicles which make it through. When the road changes direction, water may not, and instead may carve deep trenches across the road before possibly washing it away completely. When water levels are higher on one side flow towards the other must be facilitated by culverts under the road, or drifts. If they are absent or too small, again the road will be washed away. A low-lying road floods without adequate drainage and becomes impassable, temporarily and soon permanently. In all cases, the road tends to revert to its natural state, no matter how much is spent on maintenance. Road construction and maintenance costs will also increase with topography, so that extensive earth or rock works can substantially affect prices and need mobilisation of additional resources. The classic ‘ridge road’ can be the ideal solution where earthworks, water interception and cross-drainage structures are minimised. Unfortunately, population location and other factors may prevent the choice of such lower cost routes.
Again, if a choice is to be made between labour, intermediate or large machine-based technology, unit costs and work methods must be adapted to take account of their differences. At the same time, supplementary social and economic indices to permit an informed choice between technologies must be included, such as local employment generation, foreign exchange savings, fossil fuel consumption, and gender issues. Finally, additional spending, on environmental protection and on compensation to households must be included. This may also vary according to the choice of technology since labour-based construction has a smaller footprint and is less destructive, due to the absence of heavy equipment. To get the complete picture, maintenance activities, nneed to also be costed, necessary on a routine and periodic basis to keep the new road as near as possible to its original state over time. This can usually consist of both machine and labour-based methods and must also be assessed, costed and included to provide the lifetime cost of the road.
The designers’ dilemma
The designer falls between two stools. If he under designs, trying to save on investment, the road will demand high annual maintenance costs during an unpredictable but inevitably shorter life. Overdesigned roads, on the other hand, trying to meet any eventuality of traffic or weather, waste resources to no purpose and can also be as expensive to maintain because of their high standards. To further complicate things, the time needed to measure, evaluate, design, cost, contract and supervise works is expensive. It will quickly take up a disproportionate slice of the necessarily modest road-building budget. The good-enough rural road, providing an adequate level of service for the lowest possible lifetime cost, must always be based on an informed and intelligent leap in the dark regarding future conditions and environment.
Tradeoffs are needed to arrive at an acceptable solution. If space is in short supply, as it often is in mountainous terrain, the road may have to be made narrower than the designer would like. In hilly conditions and especially if rainfall is high as well, earth or gravel surfaces may not be sufficient to protect the road foundation. The gravel surface will simply be carried away yearly (and with it the road foundation, especially if traffic is heavy and cannot be halted during and after heavy rain), blocking the drainage ditches as it goes, further accelerating deterioration. A more durable surface may be necessary at least on the steeper slopes. But what kind of surface? Again, where rainfall is high and terrain tends to being hilly or mountainous many lateral and longitudinal drainage structures will also be necessary to channel water away from or across the road. These are expensive to the point of sometimes doubling the overall cost. On the other hand, building on low-lying flat terrain subject to flooding may require an embankment to raise the level of the road, again perhaps doubling the cost.
Need for low-cost and rapid condition assessment methods
All these problems can be predicted and solved. However, as mentioned before, detailed topographical surveys necessary to pin down the quantities of work to be done on a road-by-road basis are costly and not viable in this context of low-volume rural roads. In any case, quantities of work to bedone are small and the range of estimation error high. Finally, production rates for machinery and labour are only predictable for large sites. On the small sites which characterise rural roads, productivity is difficult to predict, particularily for machinery. For this reason, rapid, low-cost assessments must be done at the planning and programming stage to determine network priorities and a priority core network.
Rapid inspection methods have been developed to measure and classify: 1) the key physical characteristics (surface condition and roughness, width available, slope, curvature, residual utility of previous investment in drainage, foundation and surface (if any); 2) its natural environment (terrain, rainfall patterns, soils, availability of materials, environmentally sensitive areas) and 3); its social environment (land use patterns, user characteristics).
Developed economy road inspection methods can be sophisticated and relatively expensive, using GPS, electronic measuring devices and video to produce a continuous image of road characteristics from a moving vehicle. They will usually provide data for estimating both costs and potential benefits. Certain simulation models, such as HDM-4, RED and RONETS will perform the cost calculations for each category. These are appropriate for high volume roads but are not usually viable for low-volume road inspection as the precision they provide is sometimes superfluous and always costly. When roads are simple and resources to apply complicated analyses are not available locally it is better to simply use recent bids or actual costs incurred in similar works. The previous comments on quality of local cost data mean that any interpretation of local previous costs need to be carefully carried out.
Simple methods are proposed using simple standardised forms or laptops on which summary data on road conditions can be coded at 100-200m intervals, relying on visual inspection of road conditions and simple measurements. TThese provide a rapid and relatively cheap basis for approximate cost estimation providing sample lengths are not too short and the volume of data collected is modest. Each short section of road is surveyed thus and located within a matrix of upgrading alternatives, and consolidated over a reasonably homogeneous length of road for costing purposes. The data collected will be sufficient to assign each length to a unique category of road to which a cost index can be assigned, which takes account of its physical and environmental characteristics. These indices can be based on recent bids or actual costs incurred in similar works.
It must be emphasized that any survey, whether automated or manual, can be expensive to carry out in remote areas. When judgement is involved as is the case here, variations among team members can also creep in. The design of procedures and forms requires clear knowledge of how the data collected will be processed. . When many inspectors are involved it requires training and daily supervision to ensure that data collected is complete and consistent among team members.
