To Pave or Not to Pave: A Framework for Systematic Decision Making in the Choice of Paving Technologies for Rural Roads

Cost–Benefit Analysis

There are several different economic and engineering models that carry out CBAs; some of the most common are the Highway Development and Management Model (HDM-4), which estimates over time the deterioration of paved and unsealed roads, and the resulting agency and user costs to perform a CBA of various alternatives; the Roads Economic Decision model (RED), an Excel model customized for the economic evaluation of low-volume roads (LVRs). RED performs a simplified CBA for road projects, based on HDM-4 principles. The CBA of a road intervention depends on a confluence of factors. These include the evaluation period and discount rate adopted; the project road terrain, condition, and traffic; the paving characteristics and costs; vehicle fleet characteristics; the expected normal traffic growth rate, and generated and/or induced traffic; and the assumptions made about the cost of road maintenance—with and without the project—over the evaluation period.

Examples traditional PEA use include the following. In one study ( 1 ), LVR surface types were analyzed using a pavement management system created for the United States Forest Service’s LVR network. Three pavement types were considered: aggregate (gravel), surface treatment (chip-sealed), and hot mix asphalt (HMA). Total life-cycle costs (LCC) for each surface type were estimated based on various traffic mixes and volumes. The results showed that the gravel and chip-sealed roads became more expensive than HMA-surfaced roads as traffic increased, owing to higher maintenance and rehabilitation costs. A study by Kerali et al. involved determining the economic viability of upgrading LVRs using the HDMIII model ( 2 ). They found that the crossover point to pave a road depends on the discount rate. At a discount rate of 5%, the crossover point was just above 100 vehicles per day (vpd). At a discount rate of 10%, it was around 200 vpd. Kerali et al. also found that of the factors beyond traffic, only rainfall and high vertical alignment (topography) had significant effects on the break-even point. They established that in mountainous areas with high rainfall, the break-even point for upgrading from gravel to sealed road ranged from 50 to 150 vpd, a lower threshold than in more typical environments. Rukashaza-Mukome et al. examined when it was economically advantageous to upgrade and pave roads ( 3 ). The researchers concluded that once average annual daily traffic (AADT) is between 100 and 200 vpd an upgrade to a gravel road should be considered.

Cost-Effectiveness Analysis

CEA compares the costs of an intervention with its expected impacts on the target beneficiaries. It is more generally applicable when the focus is on poverty or other social impacts of an intervention rather than a pure economic impact. It is also appropriate when benefits cannot easily be quantified in monetary terms. In short, it is an attempt to find the least costly way of delivering the stated objectives. A cost-effectiveness indicator (CEI) is calculated for each road link based on the following formula:

CEI = Cost of Improving the Road Link Population Served by Road Link

The CEI is then used to prioritize proposed rural road improvements. The highest priority is assigned to the road links that present the best CEI, that is, the lowest ratio of per capita capital investment. Fiscal constraints can also determine where the cutoff point is on the CEI. Given the existing resource constraints, a suitable threshold must be established. For example, it might be decided that no road project should exceed $1,000 per inhabitant, considering, for example, a gross national income per capita threshold in a low-income country context.

Multicriteria Analysis

MCA is a practical solution for balancing the factors that impinge on paving decisions, and thus helps to determine the priority roads for upgrading. Under an MCA, points are given to criteria that are thought to be important to the paving decision; then weights are applied to each of the criteria in a predetermined manner. Each road link is allocated several points corresponding to the fulfillment of the criteria, and the corresponding weights are applied. The sum of the weighted points provides a measure of the rank for that road link in the investment decision. Using this approach, it is important for policy planners and evaluators to have a clear understanding of the most critical factors, and how to justify the weights that are assigned, so as not to bias the final decision.

One study by Benchmark Engineers ( 4 ) proposes an interesting approach that combines traditional PEA with a quantification of other noneconomic factors; most of this work is based on a study by Zimmerman and Wolters ( 5 ). Benchmark Engineers undertook a detailed study to advise Laramie County in Wyoming using data from the experience of other jurisdictions in the country, proposing a step-by-step procedure for making this difficult decision based on both economic and noneconomic factors. In 2006, researchers developed the Three-Step Surfacing Alternative Decision Framework ( 6 ). The main factors considered are topography, climate/soil condition, nonmotorized traffic demand, motorized traffic volume, potential impact of dust, community impact, traffic increase after sealing, and availability of quality materials. A study by Dissanayake and Patel developed gravel road paving guidelines based on research in the state of Kansas using the MCA method ( 7 ). They identified the key factors in the decision-making process as agency cost, safety, VOCs, traffic volume, purpose of road usage, and public preference. Equations were formulated to carry out an LCC analysis along with a present worth evaluation, which provided flexibility for calculating agency costs while considering local conditions. In a study from Rwanda ( 8 ), eight criteria were selected to gauge the effects and benefits of feeder road improvements. Weights were based on the overriding policy objectives, and the socioeconomic and technical features of the area that were traversed by the roads under evaluation. Reliance of an evaluation/prioritization will depend on the quality of the socioeconomic data used in the exercise.

Challenges With the Traditional Approaches

Traditional approaches present several challenges: (i) they fail to capture important, hard-to-quantify social and other economic benefits, such as land value improvements, attracting business opportunities, health benefits from reduced dust, and considerations for nonmotorized transport users; instead, focusing exclusively on savings in road agency costs, VOCs, and travel time-savings for vehicle users; (ii) they bias investment toward more prosperous areas where vehicular traffic demand levels are already more established owing to higher motorization, thus, discriminating against investment in poorer areas; (iii) they are more suited to dealing with high- than low-traffic roads; yet many rural roads are characterized by lower levels of traffic volume; and (iv) they fail to factor in important climate change considerations and other externalities such as ever-increasing extreme weather events, and fail to properly quantify other externalities like road safety, and greenhouse gas (GHG) emissions from transport. As a result, such approaches usually yield regravelling as the solution in most rural LVR contexts, since the calculated benefits of travel time-savings and reduced operating costs are insufficient to justify the higher costs associated with bringing the road to some form of paved standard. Researchers have found that for rural road investment decision-making, the PEA approach is unsatisfactory because there is a potentially sizable share of benefits that these roads bring that cannot easily be measured in monetary terms, or aggregated in a consistent manner with the monetary values arising from the other costs and benefits ( 9 ). As a result, although traditional road project assessment tools are well-equipped to consider several parameters based on strong engineering equations and economic principles, there are other important factors that are not easily monetized, but are highly relevant and need to be considered.

Key Factors for a Systematic Paving Decision

There are seven key considerations that a systematic model needs to consider to develop a more defensible paving decision than one derived from a pure traditional economic analysis:

Ignoring these factors and concentrating solely or primarily on an economic evaluation driven by vehicular traffic volume that focuses the design on the needs of vehicles rather than the needs of all users could lead to a road being left unpaved, with a variety of negative consequences for its intended beneficiaries.

As discussed, among the three approaches—CBA, CEA, and MCA—given the preponderance of factors that impinge on paving decisions, the MCA approach offers the best hope of a meaningful and workable solution to the paving decision problem. This paper presents a novel systematic model that provides a decision framework for paving rural LVRs, based on a hybrid of MCA, CBA, and CEA: the SPADE-PLUS approach, which can be justified economically, technically, socially, and environmentally.

The SPADE-PLUS Approach

SPADE-PLUS is a two-stage, sequential approach that combines the SPADE model and either the RED model or a CEA in Stage 2 (see Figure 1). It allows for handling very complex variables in a simple manner. A sequential approach is preferred to a one-stage approach because it enables decision makers to weigh the merits of paving a road from an MCA approach. This is important given that some factors are not easily quantifiable in monetary terms. In essence, the first stage provides the paving decision (no justification to pave, medium justification to pave, and high justification to pave). The second stage provides the economic justification using the traditionally accepted CBA or CEA approaches.

After the SPADE model has been run as discussed previously, a check is made on the total weighted scores (the PPS), to determine the way forward on the paving decision. For roads with low scores (an overall PPS at or below 50%), there is no compelling justification to pave the road under evaluation. The recommendation in such cases is to do nothing, or to explore other simple alternatives like spot regravelling or drainage improvements. For roads with a medium score (a PPS between 51 and 70%), low-cost paving options like double bituminous surface treatment or Otta seals should be evaluated. For roads with a high score (a PPS over 70%), there is a compelling justification to pave, and the full menu of paving options should be considered.

An economic analysis is carried out for roads that have passed the SPADE for prioritization for paving with a medium or high priority rating, using either a RED model analysis or a CEA to complete the economic justification exercise. The objective here is not to undermine the comprehensive work done at the MCA stage, but rather to complement it. In this economic evaluation, the proposed bifurcation is to use the RED model for higher-traffic roads (>200 AADT) and the CEA for lower-traffic roads (<200 AADT). This AADT threshold has been adopted based on an extensive review of the literature that suggests that below this threshold, economic analysis models that heavily depend on VOCs and travel time-saving benefits tend to yield negative results. It is worth noting that 200 AADT is not a threshold put forward for paving decisions, it is used to define which economic justification tool is to be used in the decision-making process. It is designed to steer the economic evaluation of roads below the threshold to a CEA approach, and those above the threshold to a more traditional CBA approach. It is highly unlikely that an economic evaluation will fail for any road that has successfully passed Stage 1. In the remote event that a road fails under Stage 2, an exception could conceivably be provided by the preparation team, using the results of the MCA assigned in Stage 1 to justify moving forward. Therefore, the economic justification stage under the SPADE-PLUS approach is useful primarily for validation purposes. For this reason, it is not recommended that road ranking and prioritization decisions be made based on Stage 2, but rather the Stage 1 results.

Undertaking the traditional economic approach without making use of the SPADE model outlined in Stage 1 will result in running into the same challenges that have been fully described previously. On the other hand, undertaking the SPADE model analysis on its own, without the second step of economic justification using either the RED model or CEA, will result in an incomplete economic evaluation exercise.

OPEN IN VIEWER

Figure 1.

SPADE-PLUS framework.

Key Lessons From the Case Studies

The case studies confirmed the a priori hypothesis that traditional approaches tend to overlook critical social, environmental, and economic variables that are relevant at the time of paving. The SPADE model approach yielded a higher percentage of roads recommended for paving than traditional approaches, but remained robust and defensible since it rejected roads that did not have ample social, economic, or other justifications. The majority of cases lent themselves to a solution of “medium priority to pave,” which means using low-cost paving solutions like Otta seals, surface dressing (single or double bituminous treatments), or other stone dressings. A small sample of roads had a “high priority to pave” recommendation, which allows for all paving options including asphalt and concrete solutions. Some roads despite the new approach would still remain unjustified for paving.