REVIEW OF OREGON DEPARTMENT OF TRANSPORTATION

ASPHALT MIX SPECIFICATION, PHASE II

Interim Report

Review of Specifications and Comparison with Other States

by

David E. Newcomb, P.E., Ph.D.

Charles Gurganus, P.E.

Haydar Al-Khayat

Maryam Sakhaeifar, Ph.D.

Jon A. Epps, P.E., Ph.D.

October 2016

REVIEW OF OREGON DEPARTMENT OF TRANSPORTATION

ASPHALT MIX SPECIFICATION, PHASE II

Introduction

Specifications of construction materials and methods are significant in the success of roadway paving projects. Many test methodsand new technologies are used in both quality control and quality assurance. Hot mix asphalt (HMA) is unique in that it consists of aggregates and asphalt binders, and other additives to meet the desired specification. The complexity of HMA requires tests, statistical analysis, and verification to reach the desired goal and balance the various elements. Each Department of Transportation (DOT) has a standard specification, typically established jointly by contractors and the DOT. These standards differ from state to state. The differences come from many factors such as climate condition, traffic loads, past experience, and the availability and varying quality of materials.

In addition to specification differences across DOTs, methods for using test results to pay the contractor also vary. DOTs usually pay contractors based on multiple material and pavement test results. Some highway agencies accept the quantity of placed HMA and calculate the pay amount based on quality control test results. Quality control tests are also referred to as contractor tests. Because these tests are performed by contractors, agencies will verify the results and compare them with quality assurance (QA) test results. Other highway agencies have enough inspectors and technicians to accept the HMA based on agency performed QA test results without considering contractor QC results(1).

Specific items of interest regarding HMA acceptance and payment include lot size, sublot size, sample location, pay factors, and tolerance limits for specification parameters. This interim report reviews the current practice within Oregon DOT (ODOT) and compares those practices with the techniques used in other state DOTs. Due to the diversity in techniques across DOTs, it is not always possible to compare each element from ODOT to each element from other DOTs.

There are two overarching goals in this study. One is to address the recommendations in the December 2013 FHWA report(2), while the other is develop techniques and practices to use test data to gain a better understanding of the mix properties and behavior. The eight recommendations from the FHWA report are summarized below:

  1. Increase the amount of verification testing.
  2. Use F- and t-testing to validate contractor test results.
  3. Change the sampling location to near the paver.
  4. Develop validation procedures for smoothness tests (not addressed in this report).
  5. Reevaluate the PWL limits.
  6. Develop processes to ensure the security of materials.
  7. Modify pay factors to include field mixed, lab compacted air voids.
  8. Implement effective electronic data collection techniques (not directly addressed in this report)

Parameters of Interest

In order to accomplish the goals within the project, the following issues are reviewed:

  • Lot and sublot definition
  • Sample location
  • Pay factors
  • Specification tolerance limits
  • Percent within limits calculations

Lot and Sublot Definition

Definitions of lot and sublot vary from one state to another. Some highway agencies consider the lot as the total quantity of HMA per job mix formula (JMF) per project, the method used by ODOT. Others consider the lot as the amount of daily produced HMA, while others specify a tonnage of HMA (3, 4). The definition of lot size impacts the sublot size, sample size, and ultimately the payment. The sublot size is a portion of a lot size, and also varies from state to state. Some highway agencies have a clear definition and set the sublot size as a fixed quantity. Some specifications consider each 1,000 tons or 750 tons of HMA as a sublot while others define the sublot size based on the number of required samples or daily production(4-6).

In general, QC and QA testing is used to determine the mixture properties at the beginning of a project and during the production process. QC tests are performed once for each sublot. A combination of sublot samples comprises testing associated with a lot. Lot and sublot size can impact the sample population size, statistical analysis, acceptability of the work, and final cost. Many states use QC and QA tests results to accept or reject the work and calculate pay factors. The sample size depends on the sublot size which depends on the lot size as defined in the specification.

Sample Location

Choosing the method of testing and sample locations for HMA plant produced mix depends on DOT specifications. Some DOTs prefer to obtain samples at the plant, whileothers consider the roadway as the best location to collect the samples. Most agencies agree that taking samples must be done randomly to perform proper statistical analyses(3-5, 7).Each highway agency has a special strategy to choose the location of samples randomly. QC and QA must follow the same process to select sample locations. “If the selection process varies for QC and QA applications, statistically proving that both data sets come from the same population may be challenging”(1).

The Federal Highway Administration (FHWA)prefers agencies to take the samples from the roadway behind the paver. This location is chosen because of the belief that it best represents the final product. This location accounts for any additional absorption during transport and placement and any changes to the mix that occur during paving operations(1).

Some contractors do not prefer sampling behind the paver because it may affect the smoothness of the finished mat, possibly leading to a penalty on ride quality. Most contractors prefer to obtain the samples from the plant. This can be advantageous because most QC laboratories are located in the plant, typically providing quick and accurate results. Sampling at the plant or from a truck might not represent the entire mixture if the samples are taken from the top of the pile or the side close to the platform. While widely varied, DOTs normally collect the samples of HMA either at the plant, at the point of delivery or behind the paver (1). Figure 1 shows the most common locations of HMA sampling adopted by a number ofhighway agencies across the country.

Figure 1. Common HMA Sampling Locations

Sampling from the truck can be done manually or by a mechanical sampling device. Each method has advantages and disadvantages. Taking samples manually from the truck may impact the technician’s safety because the high temperature of the mixture, the difficulty of using the testing tools, and difficultly reaching the required point in the truck.

Sampling from the truck does not account for any changes in the mixture that occur during transport and paving operations. Transport factors affecting the mixture include climate conditions, length of haul, and on-site waiting prior to material discharge. Truck sampling is further complicated because the technician obtaining the samples must avoid segregating the material by reaching more than one sampling point within the loaded truck (1). Sampling from the truck requires proper technician training, particularly in regards to mixture segregation and safety. It is often necessary to supervise sampling from trucks to ensure the mix is representative of the mix being placed. In general, many highway agencies obtain samples from the truck at the plant. In a previous study of 52 highway agencies, 22 agencies were using this method.

Obtaining samples of HMA at the point of delivery can be either from the windrow, paving screws, or the hopper. Sampling at this stage is more indicative of the finished product as it captures any changes that occur during transport. This method requires transporting materials back to a laboratory. Chain of custody of the sample is important to both the contractor and the owner agency. From a contractor’s perspective, it is important that the sample is protected from any damage that could occur in material transport. From an owner’s perspective, it is necessary to ensure the material being tested is the material obtained from the roadway for the correct project. Sampling in this way typically requires a standard operating procedure (SOP) for material transport back to the lab. The SOP needs to contain a detailed procedure for material custody and the responsibilities of each party. Finally, sampling at the point of delivery has safety considerations. The make-up of the paving train dictates the safety impacts of sampling at the point of delivery. When windrow paving is performed, sampling from the middle of the windrow is typically preferred. This is often the safest location and allows the sampling personnel to dig into the pile to minimize segregation effects. Sampling from the paving augers is not advised and creates a dangerous situation for the sampler and the paving crew. Sampling from the hopper can be difficult, requires appropriate safety measures to mitigate the movement of the paver and backing trucks, and due to awkwardness of reaching into the hopper can lead to segregated samples. The use of a material transfer device or vehicle (MTD or MTV) complicates sampling at the point of delivery. MTVs can often help with segregation concerns during paving, but their inclusion in the paving train prevents sampling from a windrow, unless the MTV is equipped with a windrow pick-up device. MTVs make it virtually impossible to sample from the hopper and the use of an MTV creates continuous paving, making sampling from the auger extremely dangerous.

The Federal Highway Administration (FHWA) recommends sampling behind the paver. Sampling hererepresents the final stage of the mixture before compaction. The samples are obtained behind the paver and before compaction by shovel or plate and shovel. Safety issues arise with technicians sampling behind the paver at the site frommoving equipment such as compactors. Sampling at this location must be planned with a contractor’s paving train. Obtaining samples behind the paver is adopted by many highway agencies; below are some advantages and recommendations for this method(1).

  • The samples represent the final condition of HMA.
  • Easy and quick.
  • Does not affect the surface after compaction (although this isdebated by the contracting community).
  • Use loose mixture to refill the hole.

Sample bags should be sealed and tagged for transport to the laboratory. The same chain of custody and material care concerns arise as discussed with sampling at the point of delivery. Sample locations should be provided to the contractor on a “just-in’ time” to ensure that there is no bias in sampling.

According to a previous nationwide survey, including 52 highway agencies:

  • 22 highway agencies collect samples at the plant from the loaded truck.
  • 22 highway agencies collect samples behind the paver.
  • Eight highway agencies allow both methods. Six of these eight prefer to take samples from the truck at the plant (1).

Figure 2 illustrates these survey results.

Figure 2. Survey Results for HMA Sampling Location

Pay Factors

Pay factors are those elements within a specification for which an agency defines how much of the unit price to pay the contractor based upon quality measurements. Pay factors and tolerance limits come from the owner’s specification and the contractor’s job mix formula(JMF). Before start of production, the contractor has to submit the JMF proposal to the highway agency for approval.The submission, review, and acceptance of a JMF is a fairly typical process across all DOTs. The JMF is essentially a mix design and is usually job specific. While it can be unique to a particular job, it must fit within the overall framework of the DOT specifications. This formula has target values for each constituent of the mixture such as gradation of aggregates, asphalt binder content, air voids, density, etc. Not every component within a specification is a pay factor. For example, a JMF might have gradations for nine sieves, but only consider the results of four sieves when calculating the pay factor. QC and QA tests ensure the product is within the tolerance range forall design elements and are used to calculate payment at the end of the lot.

As with most elements of HMA construction, the elements chosen as pay factors vary from DOT to DOT. Typical elements considered as pay factors include asphalt content, gradation, in-place density, and lab density. Many agencies determine the amount paid based on a percent within limits (PWL) specification. A PWL specification statistically calculates the amount of materials constructed that fall within specification limits. The statistical analysis is performed to compare contractor QC tests with agency QA tests to determine if a significant difference exists between the two. Determining if the two different types of tests can be statistically considered from the same population is important because the number of contractor tests is typically much larger than the number of agency tests. This is often a function of agency staffing, but also of agency business practices. DOTs have multiple responsibilities regarding planning, design, construction, maintenance, and operation of their facilities. It is often impractical to be completely involved with every aspect of an agency’s business. For this reason, consulting services are often used for design purposes or private sector enterprises are used for operations of tolled facility or countless other agency responsibilities. Construction and construction testing is treated similarly. The agency develops procedures to verify contractor construction by testing and reviewing a smaller sample than the contractor performs.

Agencies that do not use PWL processes to calculate pay factors typically do one of three things. First, contractor tests are verified with QA tests through simple comparisons and tolerances and payment is made based on the contractor tests and according to the specification. Another option is that an agency is structured in such a way that 100 percent QA testing can be performed and payment is made solely off of agency tests. In this case, QC testing is performed by the contractor for daily construction monitoring and as a verification of QA tests performed by the owner agency. Finally, agencies monitor all QC testing used for payment purposes and pay according to monitored QC tests.

Specification Tolerance Limits

Tolerance limits and pay factors work in conjunction within a specification. Calculation of a pay factor can only be done if a parameter of interest is compared against a requirement for that element. For example, if asphalt content is a pay factor, the obvious question is what is the upper and lower limit associated with asphalt content in the JMF? Therefore, each pay factor is accompanied with tolerance limits that dictate if a contractor will receive a bonus, a penalty, or be forced to remove and replace the material. Tolerances vary across DOTs.

Current ODOT Practices

ODOT currently verifies contractor QC tests with QA tests and makes payment based on a PWL specification. A statistical analysis is performed to determine if the QC and QA tests can be appropriately compared, thus allowing ODOT to use the more extensive contractor testing to develop pay factors. The statistical analysis was developed in response to Recommendation 2 from FHWA. Section 00745 and 00165 of ODOT’s Standard Specification addresses HMA construction. Section 00745 covers mixing and construction, while Section 00165 covers the quality of materials.

Lot and Sublot Size

ODOT defines a lot as, “the total quantity of material or work produced per job mix formula per project”. Thesublot definition is 1,000 tons of HMA, regardless of lot or project size. In Oregon, a new lot is created when:

1-a new JMF is created, when:

  • there is an asphalt cement grade change;
  • the aggregate source changes;
  • additives are added or changed; and
  • the asphalt cement source changes.

2-the method of compaction measurement changes.

3-a new compaction limit is required.

4-the procedure for determining the asphalt content changes.

5-warm mix asphalt concrete technology is used.

Sampling Location

ODOT currently samples from the plant, prior to the material being discharged into trucks.

Pay Factors

In Oregon, QC and QA samples are taken randomlyto calculate aggregate gradation, asphalt content, density, and moisture content. QC testing is performedat minimum once for each sublot (1,000 tons). If the daily production of hot mix asphalt is less than 1000 tons, QC tests are performed at least once for the day’s production. According to the quality assurance program, “In general, contractor’s quality control tests are obtained at the highest frequency. Agency verification tests are usually run at a frequency of a 10% minimum, of the QC testing frequency. While the Independent Assurance program takes steps to assure the quality of both the QC and the verification test results.”

In Oregon, the pay factors of hot mix asphalt are calculated based on test results of.

  • Aggregate gradation
  • Asphalt content
  • Moisture content
  • In place density

Each variable has a weighting factor representing the percentage of each element in the total payment. Table 1shows the weighting factors for each constituent and the tolerance limits in Oregon.

Table 1. Show the Pay Factors, Weighting Factor, and the Tolerance Limits

Constituent / Weighting factor / Tolerance Limits
Aggregate gradation / 26 / In table # 2
Asphalt content / 26 / ± 0.5 %
HMA Moisture content / 8 / 0.80 % max
In – place Density / 40 / Minimum 92%

Aggregate gradation, asphalt content, and moisture content represent tests that describe the material as a mixture, prior to final construction. In-place density captures the quality of material in its final state. From a pay factor perspective, the quality of the manufactured material comprises 60 percent of the total pay, while the quality of the constructed material comprises 40 percent of the total pay.