Developing Cost-Effective Pavement Maintenance and Rehabilitation Schedules: Application of MEPDG-Based Distress Models and Key Performance Index

Pavement Maintenance and Rehabilitation (M&R) are the most critical and expensive components of infrastructure asset management. Increasing traffic load, climate change and resource limitations for road maintenance accelerate pavement deterioration and eventually increase the need for future maintenance treatments. Consequently, pavement management programs are increasingly complex. The complexities are attributed to the precise assessment process of the overall pavement condition, realistic distress prediction and identification of cost-effective M&R schedules. Cost-effective road M&R practices are only possible when the evaluation of pavement condition is precise, pavement deterioration models are accurate, and resources must also be available at the right time. 
\nIn a Pavement Management System (PMS), feasible M&R treatments are identified at the end of each branch of the decision trees. The decision trees are based on empirical relationships of the pavement performance index. Moreover, the predicted improvements in pavement performance for any treatment are set based on engineering experiences. Furthermore, the remaining service life of the pavement is estimated from the predicted deterioration of the overall condition. The future deterioration of the overall condition is estimated based on the initial condition and by considering only the effect of age notwithstanding the effect of traffic or materials. In assessing the overall condition of the pavement, this research overcomes the limitations of engineering judgment by incorporating a Mechanistic-Empirical (M-E) approach and estimating the improvement in performance for specific treatment types. It also considers the effect of traffic and materials on pavement performance to precisely predict its future deterioration and subsequent remaining service life.
\nThe objective of this research is to develop cost-effective pavement M&R schedules by incorporating (a) the M-E approach into the overall condition index and (b) the estimate of performance indices by considering the factors affecting pavement performance. The research objective will be accomplished by (i) incorporating variability analysis of existing performance evaluation practices and maintenance decisions of pavement, (ii) investigating estimates of existing performance indices, (iii) incorporating the M-E approach: sensitivity analysis, prediction, comparison and verification, (iv) estimating the deterioration model based on traffic characteristics and material types, and (v) identifying cost-effective M&R treatment options through Life Cycle Cost Analysis (LCCA). This study uses the pavement performance data of Ontario highways recorded in the Ministry of Transportation (MTO) pavement database.
\nPrecise assessment of pavement condition is a significant part in achieving the research goal. In a PMS, an accurate location reference system is necessary for managing pavement evaluations and maintenance. The length of the pavement section selected for evaluation may have a significant impact on the assessment irrespective of the type of performance indices. In Ontario, the highway section lengths range from 50m to 50,000m. For this reason, a variability in performance evaluation is investigated due to changes in section length. This study considers rut depth, Pavement Condition Index (PCI), and International Roughness Index (IRI) as performance indices. The distributions of these indices are compared by the following groupings of section lengths: 50m, 500m, 1,000m and 10,000m. The variations of performance assessments due to changing section lengths are investigated based on their impact on maintenance decisions. A Monte Carlo simulation is carried out by varying section lengths to estimate probabilities of maintenance work requirements. Results of such empirical investigations reveal that most of the longer sections are evaluated with low rut depth and the shorter sections are evaluated with high rut depth. This Monte Carlo simulation also reveals that 50m sections have a higher probability of maintenance requirements than 500m sections. 
\nThe method of estimating performance indices is also investigated to identify the requirement of improvement in estimation of the prediction models. Generally, in a PMS, the prediction models of Key Performance Indicators (KPIs) are estimated by using the Ordinary Least Square (OLS) approach. However, the OLS approach can be inefficient if unobserved factors influencing individual KPIs are correlated with each other. For this reason, regression models for KPI predictions are estimated by using an approach called the ‘Seemingly Unrelated Regression (SUR)' method. 
\nThe M-E approach is used in this study to predict the future distresses by employing mechanistic-empirical models to analyze the impact of traffic, climate, materials and pavement structure. The Mechanistic-Empirical Pavement Design Guide (MEPDG) software uses a three-level hierarchical input to predict performance in terms of IRI, permanent deformation (rut depth), total cracking (reflective and alligator), asphalt concrete (AC) thermal fracture, AC bottom-up fatigue cracking and AC top-down fatigue cracking. However, these inputs have different levels of accuracy, which may have a significant impact on performance prediction. It would be ineffective to put effort for obtaining accuracy at Level 1 for all inputs. For this reason, a sensitivity analysis is carried out based on an experimental design to identify the effect of the accuracy level of inputs on the distresses. Following this, a local sensitivity analysis is carried out to identify the main effect of input variables. Interaction effects are also analyzed based on a random combination of the inputs. 
\nSince the deterioration of pavement is affected by site-specific traffic, local climate and properties of materials, these variables are carefully considered during the development of the pavement deterioration model to assess overall pavement conditions. The prediction model is developed by using a regression approach considering distresses of the M-E approach. In this study, the deterioration model is estimated for three groups of Annual Average Daily Traffic (AADT) to recognize their individual impact along with properties of materials. The time required for maintenance is also estimated for these categories. The investigations reveal that the expected time to maintenance for overlay with Dense Friction Course (DFC) and Superpave mixes is higher than other Hot Laid (HL) asphalt layers. This will help pavement designers and managers to make informed decisions. The probability of failure is also investigated by a probabilistic approach. 
\nWith the increasing trend towards M&R of existing pavements, it is essential to make cost-effective use of the M&R budget. As such, identification of associated cost-effective M&R treatments is not always simple in most PMS. For this reason, a LCCA is carried out for alternate pavement treatments using the deterioration model based on traffic levels and material types. Comparing the Net Present Worth (NPW) value of alternative treatment options reveals that the overlay of pavement with DFC is the most cost-effective choice in the case of higher AADT. On the other hand, overlay with Hot Laid-1 (HL-1) is a cost-effective treatment option for highway sections with lower AADT. 
\nAlthough the results are related to the Ontario highway system, this can also be applied elsewhere with similar conditions. The outcome of the empirical investigations will result in the adoption of efficient road M&R programs for highways based on realistic performance predictions, which have significant impact on infrastructure asset management.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.