Improved Mix Design, Evaluation, and Materials Management Practices for Hot Mix Asphalt with High Reclaimed Asphalt Pavement Content (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

132 Best Practices for RAP Management I. Purpose of this Guide This document provides guidance for management of reclaimed asphalt pavement (RAP) materials from the time of collection through processing, mix design, and quality control prac- tices during production of asphalt mixtures containing RAP. This document is intended pri- marily as a guide for contractors, but contains some useful information for street and highway agencies. However, this guide is not intended to be used as a specification. This document represents the current best practices for RAP management as of 2010 and, as such, may need periodic revision. This document was prepared by the National Center for Asphalt Technology and reviewed by numerous agency and industry experts. Feedback on this document should be addressed to the author at westran@auburn.edu. The goal of this best management practices guide is to facilitate the most effective utilization of RAP. Good RAP management practices are important to ensure the greatest economic benefit for RAP and the highest quality of recycled asphalt mixtures. Historical Perspective on Recycling The asphalt paving industry has had great success with recycling asphalt pavements and other recycled materials such as shingles, glass, and ground tire rubber. Recycling of asphalt pavements dates back to 1915 (1), but it did not become a common practice until the early 1970s when asphalt binder prices skyrocketed as a result of the Arab oil embargo. Asphalt paving technolo- gists reacted to this situation by developing recycling methods to reduce the demand on asphalt binder and, thereby, reduce the costs of asphalt paving mixtures. Many practices that were initially developed during that period are still in use today and have become part of routine operations for pavement construction and rehabilitation. Motivations for recycling include economic savings and environmental benefits. Environmen- tal benefits include reduced emissions and fuel usage due to reduced extraction and transporta- tion of virgin materials, reduced demands on non-renewable resources, and reduced landfill space for disposal of used pavements. Economic benefits include materials cost savings from replacing a portion of virgin aggregates and binders with RAP as well as reduced costs associated with transporting virgin materials to a site. For over three decades, two guiding principles of asphalt recycling have been (1) mixtures containing RAP should meet the same requirements as mixes with all virgin materials and (2) mixes containing RAP should perform equal to, or better than, virgin mixtures. A P P E N D I X D

133 Recent surveys have reported that across the United States, the average RAP content in new asphalt mixes is around 12 to 15%. A goal established by the National Asphalt Pavement Associa- tion (NAPA) is to increase the average RAP content to 25% by the end of 2013. Although a few people in the pavement community have a negative perception about using reclaimed asphalt pavement materials in new asphalt mixes, mixes with moderate-to-high RAP contents are not inferior paving products. Quality recycled mixes have been successfully designed and produced for many years. The proof is in performance. A recent study comparing the per- formance of recycled versus virgin mixes based on Long-Term Pavement Performance (LTPP) data from 18 U.S. states and Canadian provinces shows that mixes containing at least 30 percent RAP are equal to virgin mixtures in all measures of pavement performance. Overview This guide is organized to follow the sequence of handling and evaluating RAP materials from the point of reclaiming RAP through quality control practices during production of asphalt mixtures containing RAP. Section II provides guidance on the reclaiming processes. Section III covers decisions and practices for processing and inventory management of RAP materials. Section IV presents best practices for sampling and testing stockpiled RAP materials. II. Managing the Reclaiming Process RAP may be obtained from several sources. The most common method is through milling oper- ations, also known as cold planning. Two other common sources of RAP are full-depth pavement demolition and wasted asphalt plant mix. This section discusses the different types of RAP sources. Milling Milling is a beneficial part of pavement rehabilitation (see Figure 1). Advantages of milling include the following: • Removes distressed pavement layers, • Maintains clearances under bridges and avoids buildup of pavement weight on bridge, Figure 1. Milling machine removes asphalt pavement layers as part of pavement rehabilitation. (Photo courtesy of Astec Industries.)

134 • Avoids filling up curbs and avoids drop-offs at drainage inlets in urban settings, • Reduces the need for the costly addition of shoulder material along the edge of pavements on rural roadways, • Restores pavement grades and profiles, which are important for smoothness, • Leaves a rough texture on the remaining surface that creates a very good bond with an over- lay, and • Is an efficient removal process that can be done within a short lane-closure with the paving operations. Selecting the Milling Depth Selection of the milling depth is a critical agency decision in planning the rehabilitation of a pavement. Often, a milling depth is based on visual examination of cores to determine the depth of surface cracks and/or the location of weak layers or interfaces. Removal of these distressed or weak layers helps achieve long-term performance of the overlay. Cores should be taken at least once every lane mile on highways and one per lane per block on city streets. It is important to check the cross-section of pavement layers across lanes, since roads have often been widened in the past with a different buildup on the added roadway width. See Figure 2. Inspecting the Milling Process Milling processes should be closely examined to make sure the milled material is not contami- nated with soil, base material, paving geotextiles, or other debris. This is particularly important for deep mills or milling on shoulders or widened roadways. Milled materials that become con- taminated should be used only as shoulder material and should be stockpiled separately from RAP to be used in asphalt mix. A recommended maximum limit of 1 percent deleterious mate- rial should be used to evaluate RAP contamination. This limit is consistent with requirements for virgin aggregates. The milled surface should also be inspected for “scabbing,” where thin, weakly bonded layers are left in place. If this is observed, the milling depth should be adjusted to remove the scab layer. If such a weakly bonded layer is allowed to remain in place, the performance of the overlay will severely diminish. Figure 2. Roadway cores showing distressed layers: top-down cracking on left, stripping damage on the right.

135 Finally, the milled surface should be inspected for uniform texture. See Figure 3. A non- uniform texture resulting from worn or broken tips on the milling drum can cause problems with compaction of thin overlays. It may also cause an unsafe surface for motorcycles if the milled surface is opened to traffic. Some agencies require a simple texture check and have a limit of ½-inch peak to valley on the milled surface. Aggregate Breakdown During Milling Milling machines consume a lot of energy in removing pavement layers by impacting the pavement with milling teeth mounted on a drum rotating at about 200 rpm. The impacts break up the pavement by ripping through the mastic and aggregate particles. Crushing of aggregate particles causes the gradation of the millings to be much finer than the gradation of the pavement layers in place. In the past, pavement cores were obtained before milling, and the layers to be milled were removed for extraction tests. Adjustment factors were then applied to the extracted gradation to estimate the gradation after milling. However, this technique is not reliable since the amount of aggregate degradation depends on the hardness and brittleness (impact resistance) of the aggregate, the stiffness of the asphalt (and, therefore, the tempera- ture of the pavement at the time of milling), the speed of the milling machine, and the depth of the cut. Milling for Removal of Specific Layers In some cases, it may be advantageous to use special milling operations to remove specific pavement layers. One example is milling to remove an open-graded friction course (OGFC) layer that is raveling. If the pavement will be resurfaced with a new OGFC or other type of very thin wearing course, it may be beneficial to remove only the existing OGFC surface without milling much into the underlying layer and produce a fine-textured milled surface on which the new sur- face course can be placed. In this case, a micro-milling drum, as shown in Figure 4, can provide a much smoother surface texture, which is better suited for achieving the desired smoothness with the new surface layer. Using a normal milling drum may result in deep and/or irregular groves that can lead to dragging when a thin layer is placed on top. Figure 3. Milled pavement surface with thin scab layer which will likely lead to premature failure of the overlay.

136 A special milling operation may also be beneficial when it is desirable to mill the surface layer in one pass and the underlying layer(s) in a second pass because the surface-course millings contain a high-value friction aggregate and/or a modified binder. Some contractors have found this type of milling operation to be economical when the cost of new friction aggregates is very high and the project specifications allow the surface-course RAP to be used in new surface layers. Pavement Demolition RAP may also be obtained from complete demolition of an existing pavement using a bull- dozer or backhoe. This process is typically limited to small areas of pavement. It is slow and results in large chunks of pavement that may be more challenging to process into a useable recycled material. When pavement rubble is contaminated with underlying layers and soil, it is better for this material to be crushed and used as a shoulder or base material than used in an asphalt mixture. See Figure 5. Plant Waste All asphalt plant operations generate some waste during plant start-up, transition between mixes, and clean out. Generally, start-up and shut-down plant wastes have very low asphalt Figure 4. Micro-milling drums have three times the number of teeth as a normal milling drum. Figure 5. Pavement rubble from full-depth demolition of a roadway.

137 contents. Another form of waste is mix rejected from a project due to incomplete coating or due to the mix temperature being too high or too low for the job. Other situations that may result in wasted mix include trucks loaded with too much mix to finish the job or mix that could not be placed due to inclement weather. These waste materials are often stockpiled for later processing into a recyclable material. Since these waste mixes have not been subjected to environmental aging from years of service, the asphalt binder is less aged than RAP recovered from the road. Waste materials also have fewer fines than other sources of RAP since it was not milled or broken up during demolition. However, waste materials must be thoroughly mixed and processed to make them into uniform, recyclable materials. Waste materials are often combined with other sources of RAP in multiple-source stockpiles. Processing RAP from multiple sources is discussed in greater detail in the next section. Contamination It is important that stockpiles be kept free of contaminants from the beginning. It is easy to understand how bad perceptions of RAP form when there is dirt, rubbish, or vegetation in RAP stockpiles (see Figure 6), or when trash is found in the mix when it shows up on the job site or pops out of the pavement a few days after paving. Treat RAP stockpiles as the most valuable material on the plant yard—because they are. Truck drivers bringing materials onto the plant yard must be clearly instructed where to dump their loads so that unwanted construction debris does not end up in the RAP stockpile and instructed that they should clean the truck beds before hauling millings or useable RAP. The plant QC personnel and the loader operator should also regularly inspect unprocessed and processed RAP stockpiles to make sure they do not contain deleterious materials. If contaminants are found, dig them out immediately so that they are not covered up with other RAP brought onto the yard. III. Inventory Management and Processing RAP Poor management of RAP stockpiles is commonly cited as one reason agencies are reluctant to increase allowable RAP contents in asphalt mixtures. This section provides guidance on inven- tory management of RAP materials and options for stockpiling, crushing, and screening RAP. Good materials management practices should always be a part of the quality control program for any asphalt mix production operation. For production of quality mixes with high RAP contents, excellent materials management practices are essential. Figure 6. Multiple-source RAP pile with dirt contami- nation on the right side of the photo.

138 Inventory Analysis RAP management should begin with a basic inventory analysis of available RAP and mix pro- duction. This analysis is important to establish realistic goals for how much RAP can be used at a particular plant. The analysis includes the following four simple steps: 1. An inventory of RAP on hand and RAP generated per year, 2. A summary of mixes produced per year by mix types and customers, 3. Determining the maximum amount of RAP that can be used, and 4. A comparison of the quantity of RAP available to the amount of RAP needed. Note that in this context, RAP contents refer to the RAP material as a percentage of the total mixture. Some agencies now have specification limitations based on the percentage of RAP binder in the total binder content. Such specifications have merit when dealing with changing the grade of the virgin binder in the recycled mixture. However, for an inventory analysis, the more com- mon expression of RAP content as a percentage of the total mixture is more appropriate. Examples are the best way to illustrate the inventory analysis. Three cases are presented. Case #1: Contractor A has an estimated 20,000 tons of RAP on his/her plant site and typically brings in about 30,000 tons per year from milling projects and other sources. The plant typically produces about 150,000 tons of HMA per year. Of that quantity, approximately 100,000 tons is produced for state projects, and the other 50,000 tons is produced for commercial work and local governments. However, the contractor generally follows DOT specifications for designing mixes for local and commercial work. It is estimated that 80 percent of the mix produced is surface mix. The state specifications currently allow up to 20 percent RAP in surface mixes and up to 30 percent in base and binder layer mixes. Contractor A currently uses the maximum allowable RAP by specification. RAP available = 20,000 tons + 30,000 tons = 50,000 tons Maximum RAP needed = 150,000 tons × [(80% surface × 20% RAP) + (20% base/binder mix × 30% RAP)] = 33,000 tons of RAP Therefore, for Contractor A to increase RAP usage, she/he will have to either 1. Get the agency specifications changed, 2. Increase the plant’s annual production, or 3. Increase rap contents in local and commercial work. If Contractor A does nothing different, she/he will have a large excess supply of RAP, which may become a storage problem. Case #2: Contractor B has 10,000 tons of RAP on site and brings in about 25,000 tons of new RAP per year. The plant typically produces 200,000 tons of HMA per year of which 80 percent is surface mix and 20 percent is non-surface mix. Production of mix for the state agency is about 120,000 tons, and the remainder is for the city, county, and private businesses. Contractor B currently uses 15 percent RAP in all DOT mixes even though the agency allows 20 percent RAP in surface mixes and 40 percent in base and leveling mixes. Mix designs are typically tweaked for local mixes to include 20 percent RAP although there is no provision on the maximum allowable RAP content for these mixes. RAP available = 10,000 tons + 25,000 tons = 35,000 tons Maximum RAP needed = 120,000 tons × [(80% surface × 20% RAP) + (20% non-surface mix × 40% RAP)] + (80,000 tons × 20% RAP) = 44,800 tons of RAP RAP currently used = 120,000 tons × 15% RAP + 80,000 tons × 20% RAP = 34,000 tons of RAP Therefore, Contractor B has about enough RAP on hand for an average year using historical RAP percentages. This contractor could increase RAP usage but will have to get more RAP. If the contractor begins to use higher RAP percentages but does not bring in additional RAP, he/she will run out of RAP before the year is over. Case #3: Contractor C has 60,000 tons of unprocessed RAP in inventory and generates nearly 40,000 tons of RAP from milling and pavement demolition each year. The contractor recently replaced the old plant and expects annual tonnage to increase from about 170,000 tons per year to 200,000 tons per year. Histori-

139 cally, the contractor was able to use only about 15 percent RAP with the old plant, but the new plant was advertised to handle up to 50 percent RAP. Annual tonnage for the city work has been about 30,000 tons, commercial work has been about 30,000 tons, and state work about 110,000 tons. All sectors are expected to grow by about 10,000 tons each. State DOT and city specs have recently changed to allow 30 percent RAP in surface mixes and 40 percent in base and binder mixes. Commercial work generally does not have limits on RAP percentages. Surface mixes generally are about 80 percent of the city and state mix produc- tion but only about 50 percent of the commercial work. RAP available = 60,000 tons + 40,000 tons = 100,000 tons Maximum RAP needed: City: 40,000 tons × [(80% surface × 30% RAP) + (20% base/binder mix × 40% RAP)] = 12,800 tons of RAP Commercial: 40,000 tons × [(50% surface × 50% RAP) + (50% base/binder mix × 50% RAP)] = 20,000 tons of RAP State: 120,000 tons × [(80% surface × 30% RAP) + (20% base/binder mix × 40% RAP)] = 38,400 tons of RAP Total: 71,200 tons of RAP If Contractor C is able to use the maximum amount of RAP for each type of mix in all sectors, this contractor will have enough RAP for the first year but will run out of RAP in the second year if he/she continues to bring in the same amount of new RAP. If Contractor C believes that 40,000 tons of new RAP is reasonable, then he/she may want to consider using 25 percent RAP in all mixes. That would consume 50,000 tons of RAP per year, which he/she would be able to sustain for 6 years. In most cases, when a contractor has a limited supply of RAP, it is logical to try to use a relatively consistent amount of RAP in all mixes rather than to use a lot of RAP in some mixes and less in other mixes. For example, if a contractor has 40,000 tons of RAP and produces 200,000 tons of HMA per year, then it is better to run 40,000/200,000 = 20 percent in all mixes. If the contractor uses 40 percent RAP in some mixes, then he/she will have to use less than 20 percent other mixes to keep the RAP in balance with the total RAP used. Running higher RAP contents could be more competitive on certain jobs, but there may be additional costs associated with higher RAP contents, such as additional materials testing, higher RAP processing costs, plant modifications, and higher plant maintenance costs. Single or Multiple Unprocessed RAP Stockpiles One of the first decisions in inventory management of RAP should be whether to put all incoming RAP materials into a single pile or to create separate stockpiles for RAP obtained from different sources. This decision will likely depend on the following factors: • Whether the state or primary local agency allows RAP from other sources in asphalt mixes produced for its agency specifications, • Whether the state or other primary local agency requires captive stockpiles or allows continu- ous replenishment of stockpiles, • The space available at the plant site for RAP processing and stockpiling, • The target RAP percentages in the asphalt mixes to be produced, and • How much RAP comes from a single project. Some agencies’ specifications allow only RAP from their projects to be used in their mixes. RAP from agency projects are often referred to as “classified RAP” since the origin of the materi- als is known. This limitation is used to assure that the aggregate and binder in the RAP were of satisfactory quality in the original pavement. Most agencies allow the use of RAP from multiple sources, including “unclassified RAP” that has been combined and processed into a single uniform RAP stockpile. Agencies typically allow

140 this practice with the stipulations that the combined blend of RAP and virgin aggregates meet the appropriate Superpave consensus aggregate requirements and the volumetric properties of the recycled mix design meet all of the standard asphalt mix specifications. When this approach is used, good processing practices of the multiple-source RAP material are necessary to create a uniform material. Since many contractors report that a substantial amount of their RAP comes from non-DOT sources, this approach enables them to best utilize RAP from different sources in a wide range of mix designs and requires the least amount of testing and mix design work. In other words, using just one RAP stockpile in many different mix designs is efficient from a testing point of view. Agencies that prohibit the use of RAP processed from multiple sources will suppress the use of RAP. In many cases, it is not cost effective to perform all the necessary tests and mix designs for small quantities of RAP. Captive or Continuously Replenishing RAP Stockpiles Another requirement some agencies impose on RAP stockpiles is that no additional material can be added to a RAP stockpile once it is built and tested. This is referred to as a “captive” RAP stockpile. A few agencies take this same approach with virgin aggregate stockpiles. The opposite and more common approach is to allow stockpiles to be continuously replenished with new material. Most agencies use this approach for virgin aggregates because there are other controls on aggregate testing at the source. This is appropriate for RAP as well if consistency can be estab- lished through a RAP quality control plan. The more conservative captive stockpile approach is based on the premise that the properties of the stockpile must be precisely known if it is to be used as a component in hot mix asphalt. However, some contractors have been able to develop RAP processing practices using continu- ously replenished stockpiles that have very consistent gradations and asphalt contents over a long period of time. Determining if the RAP processing provides a consistent material over time requires regular testing and analysis of the RAP to document the RAP stockpile variability. Guidelines for a RAP quality control plan are provided in Section IV. In some cases, limited stockpile space may constrain processing and stockpiling practices. Plant yards with limited space for stockpiles may not have sufficient room for multiple small RAP stockpiles. This has been one factor that affects how some contractors use RAP. Processing and Crushing RAP The basic goals of processing RAP are to 1. Create a uniform stockpile of material from a collection of different RAP materials from vari- ous sources, 2. Separate or break apart large agglomerations of RAP particles to a size that can be efficiently heated and broken apart during mixing with the virgin aggregates, 3. Reduce the maximum aggregate particle size in the RAP so that the RAP can be used in sur- face mixes (or other small nominal maximum aggregate size mixtures), and 4. Minimize the generation of additional P200 (i.e., dust). Processing Millings Millings from a single project are usually very consistent in gradation, asphalt content, aggre- gate properties, and binder properties. Therefore, processing millings may only be necessary to achieve Goals 2 or 3. However, as noted previously, a common limitation to increasing RAP content in asphalt mixtures is the dust content in the RAP. Since milled RAPs already contain appreciable amounts of P200 (typically between 10 and 20 percent) due to the milling of the

141 material from the roadway, it is best to minimize further crushing of milled RAP whenever pos- sible. Therefore, when a contractor obtains a large quantity of millings from a single project, it is considered a best practice not to further crush this material, but rather to use it “as-is” in mix designs or to screen the millings to remove large particles. Millings: Recommended Processing Options 1. Receive millings from project. 2. Sample and test a few locations of the millings stockpile to determine the as-received grada- tion and check the maximum aggregate size. 3. If the maximum aggregate size of the as-received millings is small enough to use in the desired mix design(s), do not further process the millings. Sample and test the millings as described in Section IV. 4. If maximum particle size is too large for desired mix(es), then either a. Fractionate the RAP over a screen equal to or smaller than the NMAS of desired mix(es). Stockpile the fine RAP (portion passing through the screen) and test for properties, as described in Section IV. Stockpile the coarse RAP fraction(s) into separate stockpile(s) for use in other, larger NMAS mixes, or b. Crush the millings so that they will pass the desired screen size. This is the least desir- able option because it will result in more uncoated faces of RAP particles and generate additional dust, which can severely hamper how much of the crushed RAP can be used in mix designs. When a contractor wants to increase RAP contents but is often limited by VMA requirements or the dust-to-binder ratio during mix designs, Goal 4 must become a primary consideration in the contractor’s RAP processing plan. Processing RAP from Multiple Sources RAP materials from multiple sources that have different compositions must be processed to create a uniform material suitable for use in a new asphalt mixture. Around the world, contrac- tors have found that they can make a uniform and high-quality RAP from a combination of pavement rubble, millings, and wasted mix. The key to achieving a consistent RAP from multiple sources is careful blending as part of the processing operations. A bulldozer, excavator, or similar equipment should be used to blend materials from different locations in the multiple-source RAP stockpile as it is fed into the screening and crushing operation. See Figure 7. This will tend to “average-out” variations in the RAP from different sources. Figure 7. Excavator feeding material into a RAP crushing and screening process.

142 Screening RAP during Processing Since crushing RAP will create more aggregate fines, it is best to set up the crushing operation so that the RAP is screened before it enters the crusher. This will allow the finer RAP particles that pass through the screen to bypass the crusher. Figure 8 shows a portable RAP crushing unit that is equipped with a screen deck in line before the crusher. Only the RAP particles retained on the screen will pass through the crusher. Some RAP crushing units are set up so that all of the RAP is conveyed from the feeder bin into the crusher, followed by a recirculation circuit after the crusher. The recirculation circuit is designed to return larger particles that do not pass through the screen back to the crusher. However, since all of the material must go through the crusher in the first pass, there is a good chance that breakdown will occur for some smaller particles that did not need to be reduced in size. Crusher Types A variety of crusher types are used for crushing RAP. Many contractors have found that the best type of RAP crushers are horizontal-shaft impactors (HSIs) and roller or mill-type breakers made specifically for processing RAP. These RAP crushers/breakers are designed to break up chunks of pavement or agglomerations of RAP rather than downsize the aggregate gradation. See Figure 9. Further information on RAP crushing equipment can be found in the National Asphalt Pavement Association’s Information Series 123, Recycling Hot-Mix Asphalt Pavements (2). Compression-type crushers such as jaw crushers and cone crushers tend to clog due to pack- ing (caking) of RAP when the RAP is warm or wet. Hammermill crushers tend to generate more fines due to the retention of the material in the chamber. The speed and clearance of Hammer- mill crushers can be adjusted to reduce aggregate crushing. Some contractors have used milling machines to crush stockpiled RAP. There may be a risk of the milling machine overturning since the stockpile is uneven and may not provide stable sup- port for the heavy machine. No data are available regarding the effectiveness of this method of processing in terms of size reduction or consistency of the RAP. Weather Moisture and temperature can affect crushing and screening of RAP. When the RAP is wet and/or temperatures are hot, RAP will be stickier and tend to build up in feeders and crushers, RAP Crusher Figure 8. RAP processing unit with a screen before the crusher. Figure 9. Illustration of HSI crusher.

143 blind screens, and RAP fines will stick to belts and accumulate under conveyors. Not only does this require more maintenance of RAP processing units and RAP feeder systems for mix produc- tion, it can also affect the gradation and asphalt content of the RAP. Fractionating Fractionating is a process gaining popularity in which RAP is screened into two or three sizes. The sizes are typically ¾″ × 3⁄8″, 3⁄8″ × 3⁄16″, and minus 3⁄16″. In some cases, the plus ¾″ size mate- rial is returned to a crusher, and the crushed material is then returned to the screening unit. The primary advantage of fractionating RAP is that having stockpiles of different RAP sizes provides more flexibility in meeting mix design requirements (see Figures 10 and 11). Producers that can answer “yes” to the following six questions should consider fractionat- ing RAP: 1. Can your plant produce mixes containing 20 percent or more RAP without emissions prob- lems or significant decline in production rate? 2. Does the market this plant supplies allow RAP contents above 20 percent (probably should be specific with a quantity of mix per year)? Figure 10. Samples of fractionated RAP. Figure 11. Portable RAP fractionation unit. This unit screens RAP into three sizes: 1¾” on right, 23⁄16” on left, and ¾”  3⁄16” in back.

144 3. Does your plant have an excess amount of RAP (i.e., the quantity of RAP stockpiled exceeds RAP usage per year)? 4. Does your plant site have at least 10,000 sq. ft. available in the stockpile area for a RAP frac- tionation plant? 5. Do you have difficulty meeting mix design requirements such as minimum VMA, dust pro- portion, or P0.075 content for mixes with over 20 percent RAP? 6. Do you have trouble keeping RAP mixes within quality control and acceptance limits? The decision of whether to fractionate RAP into different sizes should be the mix producer’s choice and not a specification. Some agencies have recently begun to require RAP fractionation for higher RAP contents. This type of method specification is not appropriate; a better approach to assure consistency of RAP is to set limits on the variability of the RAP stockpiles. This is dis- cussed in further detail in Section IV. Moving the Processed RAP Stockpiles In most cases, processed RAP will be moved from the location where it is screened and/or crushed to another location that is more convenient for feeding into the asphalt plant. This is another opportunity to remix the material and improve its consistency. Using the loader to dig into the RAP stockpile at the processing unit at different locations around the pile and remixing loads while building the stockpile at the final location can again be used to average out variations. Stockpiling to Minimize Segregation As with virgin aggregates, there is a potential for RAP materials to become segregated in stock- piles. This is a common problem when stockpiles are built using fixed conveyors that allow the RAP particles to drop long distances to the stockpile. Larger particles have more kinetic energy and will tend to roll down toward the bottom of the stockpile. This results in more coarse parti- cles with a lower asphalt content at the base of the stockpile and finer, higher asphalt content RAP in the top of the stockpile. This problem can be minimized by using indexing-type conveyors that extend and raise the end of the conveyor as the size of the stockpile increases. If segregation is evident, a front-end loader can be used to remix the stockpile. Stockpiling to Minimize Moisture Moisture content of aggregates and RAP is a primary factor affecting an asphalt plant’s pro- duction rate and drying costs. Some contractors have implemented creative approaches to reducing moisture content in stockpiles. The best practice to minimize the accumulation of moisture in stockpiles is to cover the stockpile with a shelter or building to prevent precipitation from getting to the RAP, as shown in Figure 12. Second to that, it is a good practice to use coni- cal stockpiles to naturally shed rain or snow, and to place the stockpile on a paved and sloped surface to help water drain from the pile. Irregular-shaped stockpiles with surface depressions that will pond water should be corrected by shaping the pile as it is being built with the front- end loader or a small dozer. However, the use of heavy equipment on the top of RAP stockpiles should be minimized to avoid compaction of the RAP. Likewise, it is also recommended that RAP stockpiles be limited to 20 feet in height to reduce the potential for self-consolidation of the stockpile. In-Line RAP Crushers or Crusher Circuits RAP crushers or crushing circuits that are built into the asphalt plant’s RAP feed line can change the gradation of the RAP material being fed into the mix. Gradation test results on

145 Figure 12. Covered stockpile to minimize moisture in RAP. the stockpiled RAP then become meaningless, and the quality control technician will have to make unnecessary, and probably substantial, mix adjustments to get the mix gradation and volumetric properties in specification during production start-up. In many cases, this could result in the technician reducing the RAP content in order to meet the quality control toler- ances for the mix. In-line roller crushers (also known as lump breakers) and reduced-speed impact crushers designed to break up agglomerations of RAP rather than change the gradation are used by some contractors. It is recommended to conduct a simple extracted gradation check of RAP samples before and after the in-line crusher to determine if it is breaking down the RAP aggregate (see Figure 13). Figure 13. When using in-line RAP crushers, check extracted gradations before and after the crusher to make sure the RAP aggregate gradation is not changing.

146 IV. Sampling and Testing the RAP This section provides guidance on the best methods and practices for sampling and testing RAP as part of a quality management program. A well-executed sampling and testing plan for RAP is necessary to assess the consistency of the RAP stockpiles and to obtain representative properties for use in mix designs. RAP Variability A common misconception exists that RAP stockpiles are highly variable and, thus, using higher RAP contents in new asphalt mixes will lead to more variability in the mixtures. How- ever, well-managed RAP stockpiles have a more consistent gradation than virgin aggregates (3). See Figure 14. That was the finding of a 1988 study by the International Center for Aggregate Research (4), which has been confirmed with recent data gathered by NCAT (5). Considering that RAP obtained from a single milling project in which the pavement was constructed of mix- tures subject to high quality assurance standards, it is no surprise that the millings would have a consistent gradation, asphalt content, and binder properties. Less expected is how consistent RAP processed from multiple sources can also be just as consistent in gradation and asphalt content as millings. Sampling and Testing Frequency Sampling at least one set of tests per 1,000 tons of RAP is considered a best practice. This is generally more frequent than is required for virgin aggregates, but is appropriate for a component Advantages and Disadvantages of Different RAP Processing Options Process Possible Advantages Possible Disadvantages Use of Millings without Further Processing Avoids further crushing of aggregate particles in RAP, which may allow higher RAP contents in mixes Lowest cost of RAP processing options Millings from large projects are likely to have a consistent gradation and asphalt content Requires multiple RAP stockpiles at the plant Millings from individual projects are different; therefore, when a particular millings stockpile is depleted, new mix designs must be developed with other RAP Screening RAP before Crushing Limits crushing of aggregate particles in RAP, which reduces dust generation Few RAP crushing and screening units are set up to pre-screen RAP Crushing all RAP to a Single Size Allows the processed RAP to be used in many different mix types Generally provides good uniformity from RAP materials obtained from multiple sources Increases the dust content of RAP stockpiles, which will tend to limit how much RAP can be used in mix designs Fractionating RAP Using different sized RAP stockpiles provides greater flexibility in developing mix designs Requires the most space for multiple smaller stockpiles Most expensive processing option (cost of fractionation unit plus additional RAP cold feed bins) Table 1. Advantages and disadvantages of RAP processing options.

147 that will comprise a large portion of an asphalt mixture. A minimum of 10 tests should be per- formed on a RAP stockpile to yield good statistics for consistency analyses. Sampling Method It is recommended that RAP stockpiles be sampled as they are being built at the location where they will be fed into the asphalt plant. Samples from the different locations should not be combined since the results from the different locations will be used to calculate variability statistics. Sampling at the time the stockpile is built will be easier and more representative of the stockpile compared to samples taken later, after a crust forms on the RAP stockpile. When a RAP stockpile has been in place for a while, it is generally difficult to dig into with a shovel. The best way to sample existing RAP stockpiles is with the assistance of a front-end loader, as described in Section X1.2 of AASHTO T2 or ASTM D 75-03. This method is described here and illustrated in Figure 15. 1. Use a front-end loader to dig into the ready-to-use RAP stockpile. 2. Empty the bucket on a clean surface to form a miniature sampling stockpile. 3. Use the loader to back blade across the top of the mini stockpile to create a flat surface. 4. Mini stockpile ready to be shipped. 5. Use a square-ended shovel to obtain samples from the surface of the mini stockpile. 6. Sample from three locations over the surface of the mini stockpile. 7. Combine samples taken from the same mini stockpile. This sample will later be divided into test portions. 8. Repeat this process to obtain samples at other locations around the RAP stockpile. Do not combine samples from different locations. Test Methods For mix designs using RAP, the data needed from tests on the RAP are as follows: 1. Asphalt binder content of the RAP, 2. Gradation of the aggregate recovered from the RAP, 3. Bulk specific gravity of the RAP aggregate, 4. Consensus properties of the aggregate recovered from the RAP, and 5. The RAP asphalt binder properties (for high RAP contents). In some cases, additional aggregate tests may be necessary. For example, if the RAP is to be used in a surface mix for high-speed traffic, some agencies may require tests to evaluate the polishing or mineralogical composition of the RAP aggregate. Typically, source properties such as LA abrasion and sulfate-soundness tests are not necessary since it is unlikely that the coarse aggregates in the RAP would have come from sources not originally approved by the state agency. Figure 14. Processed RAP with a uniform appearance.

148 1 2 3 4 5 6 7 8 Figure 15. Steps for the best method to sample RAP.

149 A recent joint study by the University of Nevada-Reno and NCAT examined several options for testing RAP to determine the best methods for determining many of the properties noted above. Three methods were used to determine asphalt contents and recover the aggregates for aggregate property tests: the ignition method, the centrifuge extraction method, and the reflux extraction method. Trichloroethylene was used as the solvent in the centrifuge and reflux meth- ods. The results of the study indicate that • The ignition method yielded the most accurate asphalt contents for the RAP and provided the lowest testing variability compared to the solvent extraction methods. • The centrifuge extraction method had the smallest affect on the gradations of the recovered aggregate. • The combined bulk specific gravity of the aggregates recovered by the ignition method was closest to the original materials, except for the soft limestone aggregate. In that case, the aggre- gate recovered from the centrifuge extraction was closest to the original material. • The sand-equivalent and fine-aggregate angularity values for aggregates recovered from all three methods were different from the original materials. No consistent biases were evident to warrant making adjustments to the tested results. • LA abrasion values for aggregates recovered from the centrifuge extraction were closest to the original values. Additional tests on the extracted and recovered asphalt binder from the RAP may be required for mix designs that will contain more than 25 percent RAP. Current best practices for deter- mining RAP binder properties are described in Chapter 3 of NCHRP Report 452 (6). Several research studies are currently in progress to develop alternative procedures for determining RAP binder properties and methods for selecting the grade of the virgin binder for high RAP content mixtures. Methods for Determining RAP Asphalt Contents and Recovering Aggregates for Characterization Two options are recommended for determining RAP asphalt content and recovering aggre- gates: the ignition method and solvent extractions. Both methods have advantages and disad- vantages as described in this section. Ignition Method The most popular method for determining RAP asphalt contents and recovering aggregates for other tests is the ignition method, AASHTO T 308 or ASTM D 6307. Advantages of the ignition method include quick results, little testing time, and the absence of a need for the use of solvents. One issue with this method is that in order to obtain an accurate asphalt content for a sample, it is necessary to know the aggregate-correction factor. For virgin materials, the aggregate-correction factor is determined by testing samples with a known asphalt content. The difference between the known asphalt content and the test result for the prepared samples is the aggregate-correction fac- tor. However, for RAP, it is not possible to have a sample with a known asphalt content and, there- fore, not possible to determine the aggregate-correction factor. Fortunately, aggregate-correction factors are typically consistent over time when the aggregate materials used at the location are from the same quarry or deposits. Therefore, a historical average aggregate-correction factor of the materials at a location can be used as the aggregate-correction factor for the RAP. RAP aggregates recovered from the ignition method can be used for gradation analysis and many other aggregate property tests, but not all. Some aggregate types (e.g., dolomites) can have significant changes in mass when heated to 1000°F in an ignition oven. Small natural variations in the mineralogy of these aggregates create large variations in aggregate-correction factors in the ignition oven (as high as 1 to 2 percent). Some agencies have altered the test to reduce the

150 ignition oven temperature to minimize this problem. However, in some cases, agencies have elected simply to use other methods for determining asphalt contents and recovering aggregates for asphalt mixes in their jurisdiction. In these locations, the asphalt content for RAP samples should be determined using solvent extractions. Solvent Extractions Solvent extractions with trichloroethylene or other solvents have been used for many decades to determine asphalt contents of asphalt mixtures and as a method of recovering aggregates for additional tests. However, use of the method has declined due to health and environmental concerns with the chlorinated solvents. Normal propylene bromide and some non-halogenated (terpene or d-limonene based) solvents were found to be acceptable alternative solvents and are permitted in AASHTO T 164, but some problems have been reported with the effectiveness of these solvents to remove polymer-modified asphalt binders. However, some agencies and con- tractors continue to use solvent extractions due to problems with highly variable ignition fur- nace aggregate-correction factors or with the breakdown of certain aggregate types. Depending on aggregate absorption and texture, solvency power of the solvent, and hardness of the binder, solvent extractions may not remove all of the absorbed asphalt binder from the aggregate. Based on the published precision information, the repeatability and reproducibility of the ignition method are more than four times better than the solvent extraction method. It is prudent for agencies and contractors to cooperate in establishing the best method for the materials in their region or jurisdiction. Aggregate Bulk Specific Gravity Aggregate specific gravity of the RAP aggregate is a critical property for mix design because it is used in calculating VMA. Since VMA is the primary mix design parameter to assure good durability, accurately determining the RAP aggregate Gsb is essential, especially for high RAP contents. Previous studies have recommended several options for determining the bulk specific gravity of the RAP aggregate, as follows: 1. Recovery of the RAP aggregate using the ignition method (AASHTO T 308) followed by con- ducting AASHTO T84 and T85 for specific gravity of the fine and coarse aggregate portions, respectively. 2. Recovery of the RAP aggregate using the solvent extraction (AASHTO T 164) followed by conducting AASHTO T84 and T85 for specific gravity of the fine and coarse aggregate por- tions, respectively. 3. Estimating the RAP aggregate bulk specific gravity using the following process: a. Conduct the maximum theoretical specific gravity test (i.e., the Rice method) on samples of the RAP following AASHTO T 209. b. Calculate the effective specific gravity of the RAP aggregate from the asphalt content, Gmm of the RAP, and an assumed value for specific gravity of the binder, Gb. 100 100 G RAP P P G G se b RAP b RAP mm RAP b )( = − − × ) ) ) ( ( ( c. Calculate the RAP aggregate bulk specific gravity using the following formula: G RAP G P G G sb se RAP ba se RAP b ( ) = × × + ( ) ( ) 100 1

151 where Pba (asphalt absorption) also has to be assumed based on historical records of mixes with the same raw materials. These three options were evaluated in a joint study by the University of Nevada-Reno and NCAT and in NCHRP 9-46. These studies found that the accuracy of Method 3 was highly dependent on how well the percentage of absorbed asphalt could be estimated. Even small errors in the assumed asphalt absorption value caused significant errors in VMA for the mix designs. Therefore, the author does not recommend Method 3. The flowchart shown in Figure 16 outlines the recommended process for sampling and test- ing RAP. All test results should be recorded in a spreadsheet or software program to organize and sum- marize the data. The database should include stockpile name/description, date of samples, and for each sample, the results for asphalt content, gradation of recovered aggregate, and bulk spe- cific gravity of the RAP aggregate. The spreadsheet should calculate the average and standard deviation of each property. An example spreadsheet is shown in Figure 17. It is necessary to col- lect and analyze test results of at least 10 RAP samples to estimate the statistics for the stockpile. If more RAP is added to the stockpile, sampling and testing should continue at a frequency of one set of tests per 1,000 tons of RAP. Table 2 shows guidelines for standard deviations of key properties of RAP. The standard deviation statistic is a basic measure of variability. The median sieve is the sieve closest to having an average of 50 percent passing. Typically, this is the sieve with the largest standard deviation. In the Figure 17 example spreadsheet, the median sieve is the 2.36 mm sieve. These values are based on data gathered from contractors using many of the best practices in this document. Although excellent RAP management practices are necessary to have standard devia- tions within these limits, published reports and recent surveys indicate that they are attainable. Figure 16. Recommended process for sampling and testing RAP samples. RAP Property Maximum Std. Dev. (%) Asphalt Content 0.5 % Passing Median Sieve 5.0 % Passing 0.075 mm Sieve 1.5 Table 2. Variability guidelines for RAP stockpiles.

152 If the variability of one or more properties exceeds the values in Table 2, the stockpile manage- ment guidelines in this document may be helpful in reducing the standard deviations. Also keep in mind that sampling practices can have a significant effect on variability results. References 1. Kandhal, P. S., and R. B. Mallick, Pavement Recycling Guidelines for State and Local Governments–Participant’s Reference Book, Report No. FHWA-SA-98-042, National Center for Asphalt Technology, Auburn, AL, 1997. 2. Recycling Hot-Mix Asphalt Pavements, Information Series 123, National Asphalt Pavement Association, Lanham, MD, 2007. 3. Nady, R. M., “The Quality of Random RAP: Separating Fact from Supposition,” Focus on Hot Mix Asphalt Technology, Summer 1997, Vol. 2, No. 2, National Asphalt Pavement Association. 4. Estakhri, C., C. Spiegelman, B. Gajewski, G. Yang, and D. Little, “Recycled Hot-Mix Asphalt Concrete in Florida: A Variability Study, ICAR-401-1/98, International Center for Aggregates Research, Austin, TX, 1999. 5. West, R., “Keys to Managing RAP Variability,” Better Roads, October 2009. 6. McDaniel, R., and R. M. Anderson, NCHRP Report 452: Recommended Use of Reclaimed Asphalt Pavement in the Superpave Mix Design Method—Technician’s Manual, TRB, National Research Council, Washington, D.C., 2001. Figure 17. Example spreadsheet used for organizing and analyzing RAP stockpile test results.

Abbreviations and acronyms used without definitions in TRB publications: A4A Airlines for America AAAE American Association of Airport Executives AASHO American Association of State Highway Officials AASHTO American Association of State Highway and Transportation Officials ACI–NA Airports Council International–North America ACRP Airport Cooperative Research Program ADA Americans with Disabilities Act APTA American Public Transportation Association ASCE American Society of Civil Engineers ASME American Society of Mechanical Engineers ASTM American Society for Testing and Materials ATA American Trucking Associations CTAA Community Transportation Association of America CTBSSP Commercial Truck and Bus Safety Synthesis Program DHS Department of Homeland Security DOE Department of Energy EPA Environmental Protection Agency FAA Federal Aviation Administration FHWA Federal Highway Administration FMCSA Federal Motor Carrier Safety Administration FRA Federal Railroad Administration FTA Federal Transit Administration HMCRP Hazardous Materials Cooperative Research Program IEEE Institute of Electrical and Electronics Engineers ISTEA Intermodal Surface Transportation Efficiency Act of 1991 ITE Institute of Transportation Engineers MAP-21 Moving Ahead for Progress in the 21st Century Act (2012) NASA National Aeronautics and Space Administration NASAO National Association of State Aviation Officials NCFRP National Cooperative Freight Research Program NCHRP National Cooperative Highway Research Program NHTSA National Highway Traffic Safety Administration NTSB National Transportation Safety Board PHMSA Pipeline and Hazardous Materials Safety Administration RITA Research and Innovative Technology Administration SAE Society of Automotive Engineers SAFETEA-LU Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (2005) TCRP Transit Cooperative Research Program TEA-21 Transportation Equity Act for the 21st Century (1998) TRB Transportation Research Board TSA Transportation Security Administration U.S.DOT United States Department of Transportation