The Future of California’s Asphalt Binder Specifications

Asphalt concrete, often referred to as hot mix asphalt (HMA), is a designed mixture of high-quality construction aggregates and asphalt binder, and is used in the construction of a large percentage of highway pavements in California. In order to ensure HMA pavements will perform as designed over their service lives, the HMA and its components (i.e. the aggregates and binder) must meet stringent specification requirements. A binder specification should not only define the binder product, but also ensure that it will perform adequately and not cause premature pavement failure. Therefore, an ideal binder specification should contain those properties that directly affect the pavement performance when other mix components meet their respective specifications. If a particular binder property’s influence on certain types of pavement distress is only minor, then that property should probably not be part of the binder specification.

Binder specification requirements should generally be economically justified. This means the actual cost for adding a requirement should be significantly less than the average expected benefit of doing so. For example, the additional cost of testing and any related costs that might be associated with it (e.g. extra storage capacity costs because of delays in testing) should be significantly less than expected benefits such as reduced annual lane mile cost because of increased pavement life, etc.

Specifications should also guarantee uniformity. If the specification isn’t uniform it may appear in many different versions according to the experience of the state highway department, county, or city using it. Users may be tempted to change specifications so that products they have satisfactorily used in the past continue to be used. Specification proliferation can become more prevalent if the requirements do not correlate with observed pavement performance, or if they are unusual in that they test a particular binder composition or formulation.

Binder specifications might be thought of as either descriptive or performance-based. Descriptive specifications typically identify tests to be performed on the binder. Acceptable test criteria are selected to ensure a certain binder type is provided. The specification is not designed to target a performance level via the tests. Correlation with pavement performance is empirical and is typically based on field observations.

Performance-based specifications are such that test criteria are selected to correlate with pavement performance. The specification is intended to be blind, that is, the specification tests and their critical thresholds are not selected to be descriptive of a particular binder. Obviously, the in-place performance of a pavement is also influenced by other factors of the mix including the aggregates, binder content, air-void content, etc., and in the case of all major pavement distresses (i.e. rutting, fatigue, thermal cracking, stripping) the performance is not only controlled the binder component.

Much of the asphalt binder used in the late 19th and early 20th centuries was so-called Lake asphalt mined from Trinidad Lake (Trinidad and Tobago) and Bermudez Lake (Venezuela). These native asphalts were solid or semi-solid materials, and required the addition of fluxing materials to create consistencies suitable for use as HMA mix binders. The first specifications were devised to identify the source of the asphalt binders at the exclusion of other source materials.

In 1918 the Bureau of Public Roads (now the Federal Highway Administration) introduced the penetration grading system. This system utilized a penetrometer as the principal means of testing and grading binders. The penetration test measures the consistency of asphalt binder, expressed as the distance in tenths of millimeters that a standard needle vertically penetrates the asphalt binder sample under known loading, time, and temperature (77°F) conditions. At least nine penetration asphalt grades were developed which were suited to different climatic conditions and applications. These grades were established using a maximum and minimum penetration value at 77°F, the estimated mid-point of the pavement service temperature range. Standard specifications for penetration graded asphalt cements were published in 1931 by the American Association of State Highway Officials (AASHO). In 1956 the Pacific Coast Conference of Asphalt Specifications (PCCAS) was formed to reduce the number of asphalt cement grades and help further standardize the specifications. The PCCAS adopted five penetration asphalt cement grades in 1957.

The next evolutionary step in binder specifications came in the early 1960s with the viscosity (AC) grading system. This system was initiated by the FHWA, several state highway departments, and others, and is based on binder viscosity, which is by definition a fundamental or engineering property of the material. The specification was developed to replace the empirical penetration test with a more scientific viscosity test, and to measure binder consistency at 140°F rather than 77°F. 140°F was thought to approximate the HMA pavement maximum surface temperature on a hot summer day in the U.S. Various viscosity grades were developed to use under different climatic conditions and applications.

Somewhat parallel to the development of the viscosity grading system in the early 1960s, Caltrans developed an aged residue (AR) viscosity grading system. Caltrans was experiencing HMA mix setting problems (i.e. tender mixes) with some binders whose viscosities did not increase as much as others during plant mixing. Therefore, Caltrans decided to simulate plant aging of binders using the Rolling Thin Film Oven (RTFO) test. It then graded the binder residue viscosity after aging in the RTFO so that all binders would behave approximately the same during HMA construction. Caltrans and most agencies in California use the AR binder grading system today as the primary basis for specifying binders.

Both the penetration and viscosity binder grading systems have recognized limitations. The penetration test is empirical and is not directly related to HMA performance. The penetration test is conducted at 77°F, and the viscosity at 140°F even though different climatic conditions frequently exist at different geographic locations. Neither system has a test or specification requirement for binder stiffness at low temperature to control thermal cracking. Binders in service continue to stiffen over time, and therefore affect both the fatigue cracking and low-temperature cracking of HMA pavements. Long-term aging of binders in service is not considered in either system. The penetration and viscosity grading systems are also not applicable to binders modified with polymers and other additives that have become more prevalent.

In 1987 the PCCAS was charged with developing a new binder specification that would address long-term aging and low temperature cracking. The Performance-Based Asphalt (PBA) Specification that resulted from this effort was the first attempt to incorporate actual field temperatures into a performance-related specification. Seven binder grades (PBA-1 to PBA-7) were created based on the average high and low air temperature of the hottest and coldest month, respectively. For desert climates, the California Tilt Oven Aging Test (CTOAT) was used to simulate five to seven years of aging in the pavement. Modified binders were also identified by grade in the PBA grading system. Caltrans and other agencies in California continue to use the PBA grading system today to specify modified binders.

The FHWA also launched the Strategic Highway Research Program (SHRP) in 1987, which included approximately $50 million to develop performance-based specifications for both asphalt binders and HMA mixes. This research effort was called Superpave (Superior Performing Asphalt Pavements), and the Superpave binder specification adopted many of the concepts of the PBA specification. The most significant change was the move from empirical tests to testing that enabled a binder to be characterized using real engineering properties. The Dynamic Shear Rheometer (DSR), Bending Beam Rheometer (BBR), and Direct Tension Tester (DTT) replaced viscosity, penetration, and ductility, respectively. The DSR measures binder properties at high and intermediate pavement temperatures, and provides an indication of how well the binder will contribute to resistance of pavement rutting and fatigue cracking. The BBR and DTT are used to characterize binders at low pavement service temperatures and determine their propensity to thermal cracking. In order to simulate plant aging, the Superpave binder specification adopted the RTFO, and the Pressure Aging Vessel (PAV) was introduced for all climate regions instead of the CTOAT to simulate 10 to 15 years of aging in the field.

Climatic conditions, which were first introduced in the PBA specification, was expanded in a more comprehensive approach in the Superpave binder specification. Predicted pavement temperatures replaced the air temperatures in the PBA specification, and the so the Superpave Performance Graded (PG) binders are based on the average 7-day highest pavement temperature and the lowest measured air temperature. Consequently, a PG 64-10 binder means the 7-day high temperature in the pavement will with a certain level of probability not exceed 64°C, and the air temperature will with a certain level of probability not fall below −10°C.

Caltrans has announced plans to implement Superpave PG binder specifications beginning in January 2006. At that time binder suppliers will either discontinue production of AR binders, or likely have very limited supplies remaining. Therefore, all other agencies specifying asphalt binders will, of necessity, have to transition to PG binders in 2006. Caltrans is now finalizing the new PG binder specification, and will specify four or five PG “workhorse” binder grades to cover the state. These will include PG 58-22 asphalt-rubber base stock for certain locations), PG 64-10 (Central Coast, South Coast, and Inland Valley), PG 64-16 (Low Mountain, North Coast, and South Mountain), PG 64-28 (High Desert and High Mountain), and PG 70-10 (Desert). When traffic severity will exceed minimum design criteria used to select the PG binder grade, a stiffer binder may be required and the high temperature value of the PG grade may be increased or “bumped.” Traffic severity increases could be due to greater volumes of traffic and/or slower traffic speeds, which potentially increase pavement loads and rutting potential. For example, the PG binder used for certain routes along the Central Coast and Inland Valley may be “bumped” from PG 64-10 to PG 70-10 to account for additional traffic severity. Grade bumping in those areas of the state where PG binders other than PG 64-10 are typically specified will likely be accomplished by specifying an appropriate polymer modified binder (i.e. PBA 6a, 6b, 6a+, or 7) in lieu of the PG binder.

Based on the experience of binder suppliers, HMA producers, and agencies in other states, California’s transition to Superpave PG binders will require some initial training and adjustment. HMA produced with PG binders may or may not exactly mimic all the properties of mixes containing AR graded binders. HMA producers in some portions of the state may be required to increase binder storage capacity depending on the number of PG binder grades used in their market area. Specifiers will need to develop an understanding of the PG binder system, and avoid the temptation to needlessly “bump” a binder grade. Caltrans will provide at least one PG binder training session in each of its 12 Districts beginning this fall. These training sessions will be open to state and local agency personnel as well as contractors, HMA producers, binder suppliers, and commercial laboratories, and others interested in understanding how the transition will occur.