Transportation Asset Management Case Studies
PMS: The Washington State Experience

PMS Applications in Engineering and Economic Analysis

Pavement management databases hold substantial amounts of data. This data may be used for programming, economic analysis, and engineering analysis at both network and project levels. WSDOT has been using the PMS data effectively in performing various analyses to provide decisionmakers with the information they need. The following examples of engineering and economic analyses performed by WSDOT are taken from State documents.

I-5 Seattle Portland Cement Concrete Pavement Performance Study

WSDOT, the University of Washington, Parametrix, and Nichols Consulting joined together to investigate the performance of concrete pavements on I-5 in the Seattle area, [and the] report was scheduled to be completed by Fall 2007. This study will attempt to determine when [existing] concrete pavements [constructed mostly during the 1960s and far beyond their initial design life of 20 years] on I-5 will fail and how much time WSDOT has to plan and develop reconstruction projects before the pavements deteriorate to an unacceptable level. ¹³

Investigation on the Potential Impact of Increased Use of Chip Seal Pavements on Highway System

WSDOT used the WSPMS to correlate traffic thresholds for the effective use of chip seals on roadways with [annual] average daily traffic up to 4,000 instead of the [WSDOT standard of] 2,000 ADT. In 2005, WSDOT initiated a study with the University of Washington to investigate current chip seal application practices, determine whether chip seals can be applied to higher trafficked routes (greater than current practice of routes with less than 2,000 vehicles per day), and determine the statewide economic impacts [that increased] chip seal [use may have.] Since the increased use of chip seals [can] impact the performance of the state owned route system, both a structural and an economic analysis is required.

The expected results of this study are:

• Criteria on the use of chip seals as a lower cost alternative to hot-mix asphalt overlays. Specifically, which WSDOT routes can be converted to a chip seal with assurance that the structural adequacy will not be compromised.
• Criteria that examine whether WSDOT should consider alternating chip seal and hot-mix asphalt paving cycles [to preserve structure while minimizing cost].
• Insight into how to mitigate noise, roughness, performance, and construction issues.
• Improved manual on chip seal design and construction.

The economic analysis portion of this study is currently being finalized. The entire study was to be completed by Fall 2007 and shared in the December 2007 Gray Notebook. ¹⁴

Worst-First to Lowest Life Cycle Cost

This is an example of utilizing PMS to evaluate programming and funding distribution policies, and to justify the incorporation of the lowest life cycle cost concept into project selection process versus the worst-first methodology. In 1993, the Revised Code of Washington required that project selection be based on the lowest life cycle cost concept. WSDOT determined that there is [an optimal] timing [window] (a range of approximately two to three years) at which a hot-mix asphalt pavement can be rehabilitated at the lowest life cycle cost (see the figure below.) The figure was generated by determining the pavement repair, overlay and overhead costs for the rehabilitation of a hot-mix asphalt pavement at various pavement conditions. These costs were then applied to the entire state network assuming a specific rehabilitation cycle (i.e., every four, eight, ten etc. years). A pavement rehabilitated too soon will have wasted pavement life, while a pavement rehabilitated late will have higher associated repair and rehabilitation costs. ¹⁵

Figure 3. Lowest-life-cycle cost rehabilitation cycle for hot-mix asphalt pavement.

The implementation of this concept has been an easy transition for rehabilitation project selection within the WSPMS; however, for regional officials, the change has been a bit more challenging. One major challenge involves regional officials' hesitancy in selecting a pavement to be rehabilitated that was not in the worst condition….With a bit of perseverance from the Pavement Management Staff, the majority of the Regional offices (five out of the six) bought into the change. The sixth region continued to schedule pavement rehabilitation projects based on the worst first concept. It wasn't until recently (2002) that the sixth Regional office (noted as Region) acknowledged the error in the decision and has now complied with the lowest life cycle cost requirement for rehabilitation project selection. ¹⁶

Implementation of Performance Graded (PG) Binders

In 1999 WSDOT implemented PG binders (Asphalt Institute 2003) in all state highway hot-mix asphalt projects. The PG binder establishes specifications for the selection of the asphalt binder to meet the low temperature (for minimizing thermal cracking), the high temperature (for minimizing rutting), and the truck traffic volume and speed (for minimizing rutting) for a specific pavement section. For Washington State, this established two primary asphalt binder types (PG 58-22 and PG 64-28); a third binder grade is selected for mountain passes (PG 58-34). Using WSPMS, an analysis was conducted to characterize the benefits of implementing PG binders [on minimizing] rutting at signalized intersections. The WSPMS was queried to locate intersections with stopped conditions (i.e., stop sign or signalized intersection). This resulted in eight contracts that utilized PG binders, with one to seven intersections within each project. These eight contracts included three high-temperature binder grades: PG 76, PG 70, and PG 64. With three years of data, the maximum intersection rut depth using a PG [binder] was determined, and a comparison of the average rut depths by binder grade was made.

Though this analysis tended to support the use of PG binders for reducing intersection rutting, this analysis was only based on three years of performance data and the authors acknowledged that the conclusions should not be made until additional performance data was obtained. ¹⁷

Evaluation of Pavement Smoothness and Resulting Pavement Performance

In June 2002, a research study was completed for the State of Washington to investigate factors associated with driver-perceived road roughness. This study had four primary objectives. The first objective was to design an experiment that would link roughness data to public perceptions of road roughness. The second objective was to collect data on the public's general perception of pavement roughness in Washington State. The third was to compare the public's perceptions with actual measurements of road roughness and physical roadway attributes. The last objective was to compare these findings with those in other related research.

In this study, drivers were placed in [selected vehicles in] real world driving scenarios and asked to reveal their opinions about pavement roughness. A total of 56 participants each evaluated 40 highway test segments and produced 2,180 separate "observations." Driver evaluations were collected with other data, such as speed and in-vehicle noise, and matched with driver-specific socio-demographic data and pavement-specific data from the Washington State Department of Transportation and its pavement management system.

Results from [the study] indicated that the international roughness index (IRI) is the single best predictor of driver-perceived road roughness and driver acceptability. Pavements with low IRI values generally corresponded with low roughness rankings and high levels of user acceptability. Other factors statistically associated with driver-perceived measures of road roughness included the presence of pavement maintenance, the presence of joints or bridge abutments, the age of the pavement surface, the vehicle type, levels of in-vehicle noise, the speed of vehicle, and the gender and income of the driver. ¹⁸

Implementation of Superpave Mix Design

This example demonstrates the data mining capabilities of pavement management systems for conducting engineering analysis at both project and network level to [evaluate and understand positive and negative factors affecting] pavement performance.

WSDOT began placing Superpave designed mixes in 1996 and placed an increasing number each year (two percent in 1997 up to 47 percent in 2002), with full implementation [being] scheduled for 2004. Prior to 1996, WSDOT exclusively used the Hveem mix design procedure and AR4000W asphalt binder (conventional) on all hot-mix asphalt pavements.

A project-by-project comparison of the Superpave and conventional hot-mix asphalt projects was performed using the data contained in the WSPMS. Each Superpave project was compared to the previous overlay or construction (conventional mix) completed at the same location. The PSC, IRI, and rut depths were retrieved from WSPMS for both the Superpave and conventional mix projects at the same age. For all three pavement measures (PSC, IRI, and rutting), the project-by-project comparison was followed by the statewide comparison. ¹⁹

Performance of Dowel Bar Retrofits

In 1992, WSDOT constructed a test section to determine the appropriateness of dowel bar retrofit (DBR) and diamond grinding to restore the functionality of the concrete pavement as well as to provide a smooth riding surface. Due to the success of the test section, the first large-scale DBR project was constructed on Interstate 90 (Snoqualmie Pass vicinity) in 1993.

WSDOT continued to monitor this and all other sections of concrete pavement that have been retrofitted with dowel bars. Using data from the WSPMS, performance equations will be developed to relate truck volumes to faulting such that the performance life of dowel bar retrofit could be predicted. Based on the performance of the test section it is anticipated that dowel bar retrofit will extend the life of the concrete pavement by 10 to 15 years. It is estimated that over the next 20 years an additional 300 lane-miles of concrete pavement may require DBR. Since that time, WSDOT has rehabilitated over 300 miles of existing concrete pavement by dowel bar retrofitting followed by diamond grinding. The average construction costs for DBR is approximately $450,000 (2006 dollars) per lane-mile (includes all costs: PE, construction, traffic control, etc). The typical cost of a four-inch asphalt overlay, which is the minimum recommended overlay depth for rehabilitating a faulted concrete pavement, is approximately $525,000 per lane-mile (includes all costs). DBR is considered cost effective since it is only applied to the faulted lane while an asphalt overlay would be required on all lanes, shoulders, ramps, ramp tapers, etc., [significantly increasing the effective lane miles and cost for asphalt overlay.] ²⁰

Portland Cement Concrete Pavement Damage Caused by Studded Tires

In the past, it has been difficult to assign a dollar value of the damage to pavement caused by studded tires. [With] improvements in technology, it is now possible to measure the actual amount of damage caused by studded tires on PCC pavements [and hence quantify the dollar value of damage]. [Transverse profile] measurements [conducted as part of the annual pavement condition survey] on PCC pavements indicate that the current damage due to studded tires is approximately $18.2 million (cost for removing studded tire wear by diamond grinding the concrete surface.) Over the last five years, WSDOT has constructed a number of PCC pavement test sections to determine what combination of materials could be used to help offset the damage caused by studded tires. Test section approaches have included increasing the concrete strength (making the concrete surface harder would make it more resistant to studded tires), modifying the aggregate gradation (making the aggregate gradation more uniform to minimize the smaller aggregate which is more susceptible to studded tire wear), adding the Hard-Cem product (this is a product that is typically used to harden industrial floors) and modifying the surface texture (carpet drag versus tining). ²¹

SCOPER Design Method

This is an excellent example of engineering uses of pavement management data to improve network level project scoping. The availability of the pavement management database has made it possible to develop SCOPER and to produce practical, more accurate design estimates at an early date, [when project funding needs are determined, but before project specific structural evaluations are made], to result in improved pavement design and performance within the state highway system. The initial scoping design is then available to WSDOT regional engineers as a preliminary estimate for their full design process. SCOPER estimates required overlay thickness approximately 80% of the time to produce designs within 10–15% of the final required design.

The SCOPER process uses the Asphalt Institute's component analysis method with modification to layer coefficient based on Washington characteristics [Asphalt 83; WSDOT 95a ²²]. The approach requires that the total pavement structure be developed as a new design for the specified service conditions. The method takes into account pavement condition, type, and thickness of the pavement layers.

SCOPER uses a relationship between pavement structure and traffic to estimate the subgrade's stiffness. The existing structural integrity of the pavement is converted to an equivalent thickness of hot-mix asphalt, which is then subtracted from the required thickness for a new full depth hot-mix asphalt design to determine the required overlay thickness. ²³

Performance Grade (PG) Binder Specifications

The WSPMS was used to assist pavement design engineers in selecting the proper asphalt binder grade for each individual project. The PG binder selection module [of WSPMS] accesses the project information concerning state route, milepost limits, roadway speed limit, traffic condition (free, slow, or standing) and the 15-year equivalent single-axle load (ESAL) for the selected project. The user then enters the expected overlay thickness, design ESALs, and geographical area, and the module provides recommendations for appropriate PG binder designation. ²⁴