Field Performance of Warm Mix Asphalt Technologies (2014)

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.

Effects of WMA on Plant Energy and Emissions and Worker Exposures to Respirable Fumes P A R T 2

205 Interest in the use of warm mix asphalt (WMA) has grown faster than any other new asphalt technology of the past sev- eral decades. WMA technologies allow the complete coating of aggregates, placement, and compaction at lower temperatures than conventional hot mix asphalt (HMA). Although the reduction in temperature varies by technol- ogy, WMA is generally produced at temperatures ranging from 25°F lower than HMA to the approximate boiling point of water (212°F). Simply put, these technologies are work- ability and compaction aids. Benefits of WMA may include reduced emissions, reduced fuel usage, reduced binder oxidation, and paving benefits such as the potential for increased densities, cool-weather paving, and longer haul distances. These purported benefits need to be better documented. Although most aspects of designing and constructing WMA are similar to HMA, lower produc- tion temperatures and binder modifications associated with some WMA technologies could result in differences in pave- ment performance relative to HMA. Regardless of the benefits of WMA, if WMA pavements do not perform as well as HMA, the benefits are likely negated. Differences in material properties of WMA compared to HMA may indicate potential problems with field perfor- mance of WMA pavement. Reduced oxidation of the binder may improve the cracking resistance of a pavement but may reduce its moisture and rutting resistance. Reduced oxida- tion and better compactability of WMA may allow for higher percentages of reclaimed asphalt pavement (RAP); however, the lower mixing temperatures may not facilitate the initial extent of blending of the aged and virgin binder typically seen with HMA. Additional guidance for producing and constructing WMA is needed. Numerous laboratory studies have attempted to demonstrate the extent to which mixing and compaction temperatures can be reduced. However, densification of the mix at a lower temperature is not the only factor that should be considered when selecting production temperatures. Flow through the plant and associated motor amperage draws also need to be considered when selecting mixing temperatures. Properly tuned burners will affect fuel combustion and the resulting emissions at lower burner settings. The first three objectives of NCHRP Project 09-47A are addressed in Part 1 of NCHRP Report 779. The fourth objective—to document the relative energy usage, emissions, and fume exposure for WMA compared to conventional HMA—is addressed in Part 2. Experimental Plan To collect the necessary data to satisfy Objective 4, the research team worked with several state highway agencies and contractors to identify appropriate field projects. The desire was to obtain plant energy, plant emissions, and field respirable fume data from several projects that included differences in plant configurations, environmental condi- tions, and WMA technologies. Three projects that included multiple WMA technologies and control HMA mixes were selected to allow for comparisons of the most important energy and emissions measurements without confounding of other factors. The three multiple WMA technology sites were located in Rapid River, Michigan, Munster, Indiana, and New York, New York. The Indiana and New York proj- ects were also used to collect data on the breathing-zone exposure to asphalt fumes for the paving crews. Limited energy usage audits were also conducted at three other WMA projects where a corresponding HMA control mix was also produced. For each project, the WMA mixes were produced with the same mix design as the corresponding HMA. Table 2.1 shows basic information for the projects used in this part of the study. C H A P T E R 1 Background and Problem Statement

206 Date Project Location Plant Site and Description Mixes July 19–21, 2010 County Road 513, Rapid River, Michigan Escanaba, Michigan, uninsulated parallel-flow drum HMA, Advera, Evotherm 3G Sept. 14–15, 2010 Calumet Avenue, Munster, Indiana Griffith, Indiana, insulated counter-flow dryer HMA, Gencor Foam, Evotherm 3G, Heritage Wax Oct. 19–22, 2010 Little Neck Parkway, New York City, New York New York City, New York, batch plant with mini drum uninsulated dryer HMA, Cecabase RT, SonneWarmix, BituTech PER April 19–20, 2010 US-12 near Walla Walla, Washington Walla Walla, Washington, portable plant, uninsulated parallel-flow drum HMA, Maxam Foam June 21–22, 2010 I-66 eastbound, near Centreville, Virginia Centreville, Virginia, double barrel, counter-flow HMA, Astec Foam Aug. 11–12, 2010 Montana Route 322, south of Baker, Montana Baker, Montana, partially insulated parallel-flow drum HMA, Evotherm DAT Table 2.1. Project information summary.

207 Background on Energy Used to Produce HMA and WMA Asphalt mixtures are produced by drying aggregate particles and mixing the dry aggregate with an asphalt binder at a tem- perature sufficient to (1) coat the aggregates and (2) allow the mixture to be properly compacted after silo storage, haul, and placement. Aggregates start at ambient temperature with mois- ture contents that vary depending on how they are produced and stored, and on the local weather conditions. The aggregate is heated in the dryer drum for a batch plant or the beginning portion of the drum for a drum plant. The fine aggregate tends to be heated by convection while showering through the hot exhaust gases. The coarse aggregate is primarily heated by conduction from the fine aggregate while lying in the bottom of the dryer. A significant amount of energy is required to turn water into steam or otherwise dry the aggregate. Theoretically, the temperature of the aggregate cannot increase above 212°F (100°C) until the aggregate is dry. Particles of different sizes will dry at different rates because of differences in their spe- cific surface (smaller particles have more surface area per unit volume), so the temperature of the aggregate particles of dif- ferent sizes is unlikely to be uniform. Once the aggregate is dry, continued heating will bring the aggregate to the mixing temperature. The energy used to dry, and then heat, aggregate is illustrated in Figure 2.1. Figure 2.2 shows a frequency distribution of fuel usage based on data collected by a member of the research team in the Mid-Atlantic states. Data were collected from both batch plants and drum plants. Fuel types include natu- ral gas, No. 2 fuel oil, and reclaimed oil. Two distributions are shown, one for data collected during stack emissions tests at 35 plants and another based on 2-year averages at the same plants. Typically, plants were operating at maxi- mum design capacity for the full 3 hours of stack emis- sion tests. The 2-year average values, however, include fuel used during plant warm-up, plant waste, for unsold mix, and during cleanout. The stack test data indicate drying and heating fuel usage for hot mix asphalt (HMA) aver- age 0.233 MMBtu/ton (million British thermal units). By comparison, fuel usage based on year-end production totals averages 0.249 MMBtu/ton, indicating 6.9% waste compared to steady-state production. This inherent differ- ence between energy use during steady-state operation and historical averages demonstrates that comparisons between HMA and warm mix asphalt (WMA) must be based on identical time intervals to be meaningful. Reduced fuel consumption saves natural resources and cost. Theoretical calculations indicate that a temperature reduction of 50°F (28°C) should result in a fuel savings of 11% (Cevarich 2007). Fuel savings reported from early European WMA projects ranged from 24% to 55% (Koenders et al. 2000, von Devivere et al. 2003, and Ventura et al. 2009), with typical values between 20% and 35% (D’Angelo et al. 2007). Fuel savings and stack emissions data were collected from several North American studies. Reported fuel savings from fifteen WMA projects, representing six technologies, range from a 15.4% increase to a 77% reduction (Harder 2008, Davidson 2005a, Davidson 2005b, Lecomte et al. 2007, Chief Environmental Group N.D., ETE 2006, Powers 2009, Ventura et al. 2009, Davidson and Pedlow 2007, and Middle- ton and Forfylow 2009). The average fuel savings was 23%. The lone increase occurred at the Ohio WMA Open House with an emulsion technology that is no longer used. The mix was produced at 277°F (136°C), which is high for that technology considering considerable energy was required to vaporize the water in the emulsion. Larger fuel savings typi- cally occurred with technologies like Low Emission Asphalt (LEA), WAM-Foam (Warm Asphalt Mix), and in some cases Evotherm™ ET (Emulsion Technology), which tend to have the lowest production temperatures. LEA and WAM-Foam production temperatures are usually close to 212°F (100°C). Casing losses and other inefficiencies are believed to account C H A P T E R 2 Energy Usage

208 Figure 2.1. Energy use as a function of aggregate heating. 0 5 10 15 20 25 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 Fr eq ue nc y (% ) MMBTU/TON Stack Test Year-End Totals Figure 2.2. Typical HMA drying and heating fuel usage in MMBtu/ton.

209 for some of the difference between theoretical and observed fuel savings (Harder et al. 2008). Asphalt plants use higher burner levels when producing HMA than when producing WMA. Improperly tuned burn- ers can result in incomplete combustion, thus wasting fuel. Elevated levels of carbon monoxide (CO) and volatile organic compounds (VOC) can be indications of incomplete combus- tion. Incomplete combustion, when using fuel oil or recycled oil, can lead to fuel contamination in the mix. Fuel contami- nation was suspected in at least two WMA projects (Hurley et al. 2010a, Hurley et al. 2010b). Fuel contamination may be less evident with HMA as compared to WMA because higher mix temperatures would tend to volatize the fuel prior to placement. The moisture contents of aggregate and recycled materi- als stockpiles can have a significant effect on fuel consump- tion when producing WMA or HMA. This is evidenced by the higher fuel savings for WMA technologies, such as LEA, which only dry a portion of the aggregate. Fuel usage reportedly increases 10% for every 1% increase in stockpile moisture content (Prowell et al. 2012). Best practices, such as sloping stockpile areas away from the side that the loader obtains material to feed the plant or covering stockpiles to reduce the moisture content of materials being dried, are recommended for both HMA and WMA to reduce fuel consumption. Research Approach Fuel usage depends on a number of factors including, but not limited to, aggregate (and recycled materials, if used), mois- ture content, production rate, mix temperature, and excess air (damper setting). For NCHRP Project 9-47A, data collection forms were developed to collect information on plant energy usage, including the above factors, during production (see the appendix). As noted in the background information, energy usage will vary depending on whether measurements are taken over a steady-state operating period, such as during a stack emissions test, or over a longer period of operation such as a day, week, or year which includes energy spent for start-up, clean out, or waste. The participating contractors were requested to tune their plants’ burners before producing asphalt for NCHRP Proj- ect 9-47A. For the three projects where stack emissions tests were performed, at Rapid River, Michigan, Griffith, Indiana, and New York, New York, burner tuning was conducted by a member of the project team with expertise in this topic. Most burners have an actuator motor that drives a mechanical link- age connected to dampers and modulating fuel valves. As the burner percentage is increased, dampers and fuel valves are opened to increase air and fuel for combustion. It can be dif- ficult to properly adjust the burner to maintain the optimum fuel-to-air ratio over the whole firing range without appropri- ate gas analyzers. If excess fuel is introduced to increase pro- duction rate, incomplete combustion will occur, wasting fuel. One plant showed a 24.8% reduction in fuel usage for HMA after burner tuning. Different plants use different fuels for heating and dry- ing aggregate. Natural gas was the most common fuel type; reclaimed fuel oil and liquid propane were also used. For plants using natural gas, data collection was based on gas meter readings. Cumulative production tonnage was collected at approximately the same time that the meter readings were taken. After collecting the data, it was found that commercial gas meters only update periodically and therefore cannot be used for accurate measurements of fuel usage over short-term periods (see further discussion in the chapter summary). The Rapid River, Michigan, project used reclaimed motor oil as fuel. The Rapid River plant did not have a fuel meter, so fuel consumption was calculated using tank charts and tanks sticks at the beginning and end of production. The Baker, Montana, project used liquid propane (LP). Fuel usage for the LP was based on percent tank volume. The Griffith, Indi- ana, New York, New York, Centreville, Virginia, and Walla Walla, Washington, projects used natural gas as burner fuel. Measurements for those projects were made based on gas meter readings. Unfortunately, precision of direct fuel measurements was questionable for a number of reasons and an alternative method to determine average heat input was investigated. Stack emission tests were conducted at Rapid River, Michi- gan, Griffith, Indiana, and New York, New York, sites with flow rate and composition of the exhaust gases measured continuously for two 1-hour runs on each WMA technology and HMA control. These stack gas data enabled backcalcula- tion of average heat input using the U.S. EPA’s Method 19 F factor. EPA developed F factors for commercially available fuels to calculate the stoichiometric volume of exhaust gases generated by burning one MMBtu of fuel. For example, burning 961 cubic feet of natural gas (1 MMBtu) results in 8,710 dry standard cubic feet of (exhaust) gas at 0% oxy- gen. Zero percent oxygen is what makes it a stoichiometric volume. Stack gas velocity was measured according to EPA Method 2. Molecular weight of stack gas and water vapor in the gas stream were measured using EPA Methods 3 and 4, respectively. Carbon dioxide (CO2) and oxygen (O2) concentrations were also determined using EPA Method 3. Stack gas velocity was converted to dry volumetric flow rate at a standard temperature and pressure based on stack area and percent water vapor in exhaust gases. These calculations are typically provided in any stack test report.

210 Fuel firing rate can be calculated from the average exhaust flow rate and oxygen concentration using the following equation: ( ) = × × − 60 20.9 % 20.9 2 Fuel Usage Q O F where: Fuel Usage = MMBtu/hr; 60 = min/hr, converts flow per minute to flow per hour; Q = average stack gas dry volumetric flow rate (dscfm) at standard temperature and pressure; 20.9 = standard O2% of air; O2% = percent stack O2 by volume, dry basis, units are percent and not decimal; ( )−20.9 %20.9 2O = correction factor to remove excess air and calculate resulting stoichiometric volume; and F = volume of combustion products per unit of heat content, dscfm/MMBtu: 8,710 dscfm/ MMBtu for natural gas and propane and 9,190 for oil (EPA Method 19). Results and Discussion Table 2.2 summarizes fuel usage based on direct measure- ment of fuel consumption and the corresponding cumula- tive production. An error was made reading the gas meter for the Virginia HMA; therefore, fuel usage for that mix is not reported. The potential error in determining fuel usage over a short time period based on tank sticks is illustrated in Table 2.2. The Michigan Advera and Evotherm 3G mixes were produced at the same average temperature. The production rates are almost identical. The aggregate moisture content was 0.2% higher for the Evotherm 3G, which would tend to increase fuel usage. However, the fuel usage calculated for the Evotherm 3G pro- duction is 0.038 MMBtu/ton (17%) less than that calculated for the Advera WMA. By comparison the fuel usage based on stoichiometric calculations of fuel usage, corrected for the slight difference in aggregate moisture, are identical. Similar inconsistencies between measured mix temperature and fuel usage were noted for the Indiana mixes. The local Indiana stack emissions contractor did not take stack velocity readings during the HMA and Heritage Wax stack emissions runs. Readings were taken only at the end of the run. There- fore, the stoichiometric calculations of fuel usage for those two mixes are suspect. The Indiana fuel usage in Table 2.2 based on gas meter readings are overall daily averages. Increased fuel usage of 0.223 MMBtu/ton for the Gencor Foam WMA was observed over the course of the day, including start-up, pre- heat, plant waste, and shutdown. The production temperature of the Gencor Foam mix was increased to HMA temperatures after stack emissions tests were completed. Fuel Savings The average fuel usage for the HMA production based on five projects was 0.249 MMBtu/ton. This compares well with the 0.233 MMBtu/ton calculated based on the data from the mid-Atlantic region that was reported in Figure 2.2. To make meaningful comparisons between the WMA and HMA, the WMA fuel usage data were corrected for the difference between the HMA and WMA aggregate moisture content at each site. By definition, it takes 1Btu to raise the temperature of 1 lb of water by 1°F. Therefore, it takes 142Btu to raise the temperature of water from an ambient temperature of approximately 70°F to 212°F and 1,000Btu to vaporize 1 lb of 212°F water. The fuel usage was corrected based on 1,142Btu/lb of moisture difference. The fuel usage for the normalized WMA data indicated an aver- age savings of 0.055 MMBtu/ton, or approximately 22.1% for an average temperature reduction of 48°F. This compares well to the average 23% savings reported in the literature. Because final mix temperatures for all mixes were greater than 212°F, the theo- retical fuel savings should be equal to differences between WMA and HMA mix temperatures multiplied by the specific heat of the aggregate. Assuming a specific heat of 0.24Btu/lb/°F for a bituminous mixture, a 48°F reduction in temperature should result in 0.0230 MMBtu/ton savings, or 9.3%. The question then becomes how to account for the additional 13% in fuel savings from WMA technologies over and above the theoretical 9.3% savings due to lower mix temperatures? Distribution of Fuel Savings Additional calculations were performed to allocate fuel sav- ings for the multi-technology sites where stack emissions tests were performed. Thermal energy generated to produce HMA or WMA is consumed by drying aggregate moisture, heating aggregate, heating stack gases, and casing losses. Casing losses are thermal energy used to heat plant iron and then radiated to the atmosphere, rather than being used to heat the aggre- gate. Components that account for the majority of casing loss include aggregate dryer, duct work, baghouse, and batch tower/mixing chamber (if applicable). The difference in fuel usage reported in Table 2.2 was allocated based on thermo- dynamic properties to three sources: (1) differences in mix temperature, (2) differences in stack exhaust temperatures, and (3) the remainder, believed to consist of casing losses. Table 2.3 shows the results of calculations to appropriately allocate energy savings. Differences in thermal energy based

Site Plant1 Mix Avg. Stock- pile Moist. (%) Avg. Prod. Rate (TPH) Avg. Mix Temp. (°F) Avg. Stack Temp. (°F) Fuel Use, (MMBtu/ton) Stoichio- metric Fuel Use (MMBtu/ton) Agg. Moisture Correction (MMBtu/ton) MMBtu/ton Corrected for Agg. Moisture Delta (MMBtu/ton) Delta (Btu/°F) Washington Uninsulated PF drum HMA 2.6% 316 325 339 0.278 NA NA 0.278 Maxam foam 3.0% 310 285 295 0.218 NA 0.009 0.209 0.069 1728 Virginia Double barrel HMA 2.3% 270 318 218 NA NA NA NA Astec foam 2.1% 221 288 191 0.203 NA -0.005 0.208 Michigan Uninsulated PF drum HMA 3.6% 310 300 330 0.271 0.2852 NA 0.285 Advera 3.9% 323 269 292 0.225 0.237 0.007 0.230 0.055 1769 Evotherm 3G 4.1% 320 269 296 0.187 0.241 0.011 0.230 0.055 1788 Montana Partially insulated PF drum HMA 1.3% 370 298 249 0.157 NA NA 0.157 Evotherm DAT 1.5% 378 252 238 0.137 NA 0.005 0.132 0.025 534 Indiana Insulated CF dryer HMA 3.2% 292 300 242 0.2262 0.2013 NA 0.226 Gencor foam 3.5% 300 277 232 0.209 0.223 0.007 0.202 0.024 1037 Evotherm 3G 3.8% 300 256 221 0.212 0.2073 0.014 0.198 0.028 630 Heritage wax 3.8% 279 268 227 0.201 0.159 0.014 0.187 0.039 1210 New York Batch- mini drum uninsulated dryer HMA 3.1% 271 332 284 0.260 0.2992 NA 0.299 Cecabase RT 3.4% 244 240 213 0.236 0.235 0.007 0.228 0.071 770 SonneWarmix 2.4% 267 252 195 0.216 0.198 -0.016 0.214 0.085 1063 BituTech PER 3.6% 268 253 202 0.253 0.211 0.011 0.200 0.099 1258 1 PF: parallel-flow; CF: counter-flow. 2 Values in bold used two measures of fuel usage. 3 Stack velocity measurements only taken at end of each stack emissions run; stoichiometric fuel usage believed to be erroneous. Table 2.2. Fuel usage.

Site Plant Mix Avg. Prod. Rate, (TPH) Avg. Mix Temp. (°F) Avg. Stack Temp. (°F) Fuel Usage, (MMBtu/ton corrected for Agg. Moisture) Delta (HMA-WMA) (MMBtu/ton) ACFM SCFM % Mois- ture MMBtu/ton up Stack (above 195°F) % Stack Temp. % Mix Temp. % Casing Loss Michigan Uninsulated PF Drum HMA 310 300 330 0.285 53,656 35,997 33.0% 0.0220 Advera 323 269 292 0.230 0.0549 50,870 35,853 33.0% 0.0151 13% 27% 60% Evo. 3G 320 269 296 0.230 0.0554 50,704 35,546 33.0% 0.0158 11% 27% 62% Indiana Insulated CF Dryer HMA 292 300 242 0.226 48,380 36,526 29.0% 0.0081 Gencor foam 300 277 232 0.202 0.0239 46,844 35,878 28.0% 0.0060 9% 46% 45% Evo. DAT 300 256 221 0.198 0.0277 49,494 38,520 33.0% 0.0047 12% 76% 12% Heritage wax 279 268 227 0.187 0.0387 44,944 34,673 33.0% 0.0056 6% 40% 54% New York Batch- Mini Drum Uninsulated Dryer HMA 271 332 284 0.299 67,820 48,313 21.0% 0.0206 Cecabase RT 244 240 213 0.228 0.0709 54,566 42,972 21.0% 0.0041 23% 62% 14% SonneWarmix 267 252 195 0.214 0.0850 54,088 43,766 16.0% 0.0000 24% 45% 31% BituTech PER 268 253 202 0.200 0.0994 53,267 42,646 14.5% 0.0014 19% 38% 43% Average 0.0550 15% 45% 40% Table 2.3. Breakdown of fuel savings.

213 on mix temperature were calculated using a specific heat of 0.24Btu/lb/°F for the asphalt mixture. The difference in each pair of average HMA and WMA mix temperatures at each site was multiplied by 0.24Btu/lb/°F, converted to MMBtu, and expressed as a percentage of the difference (delta) in MMBtu/ ton, corrected for aggregate moisture. Differences in mix tem- perature (% Mix Temp.) explained 27 to 76% of the fuel savings, with an average of 45%. Actual stack exhaust flow rates in cubic feet per minute (ACFM) were converted to standard conditions at 70°F (SCFM). The energy required to heat the air and mois- ture in the exhaust gas between the minimum observed stack gas temperature of 195°F and the average stack exhaust temper- atures was calculated for each mix (MMBtu/ton up stack). The average stack gas temperature for NY SonneWarmix was 195°F; therefore, its MMBtu/ton up the stack was 0.000. The calcula- tion used a specific heat of 0.44Btu/lb/°F for water vapor and 0.24Btu/lb/°F for dry air. Air at standard conditions has a mass of 0.0766 lb/cf. The difference between the HMA and WMA MMBtu/ton up the stack at a given site (relative to 195°F) was expressed as a percentage of the total delta in energy usage per ton (% Stack Temp.). The remaining unexplained differ- ences in the measured energy use are attributed to casing losses (% Casing Loss). These losses are heat lost through, for exam- ple, the shell of the drum and ductwork. Harder et al. (2008) reported heat loss measurements at a batch plant producing 320°F mix at an ambient tempera- ture of 59°F of 3 kg fuel oil per metric ton (approximately 0.111 MMBtu/ton). Total fuel usage for HMA production was 7 kg fuel oil per metric ton (approximately 0.259 MMBtu/ton). Therefore, casing losses were 43% (3/7) of total fuel usage. This cannot be directly compared to the NCHRP Project 9-47A data, however, because the units are not identical. Harder’s 43% cas- ing loss factor is for total casing losses while the 40% factor deter- mined in this study is the percent of energy savings from the use of WMA production attributed to reduced casing loss. The two factors have different denominators. Also, the 4 kg of fuel per metric ton appears unrealistic as it represents 0.138 MMBtu/ ton. It takes 0.125 MMBtu/ton to heat the aggregate from 59 to 320°F, leaving only 0.013 MMBtu/ton to dry moisture. 0.013 MMBtu/ton is only enough energy to dry 0.25% moisture. Comparison of Measured and Predicted Fuel Savings When analysis of the fuel usage data from this study was first presented, some plant manufacturers expressed concern that the calculated casing losses were higher than their theoretical calculations. Astec Industries developed a spreadsheet (Astec Fuel Calculate 3.0) to assist producers in evaluating efficiency of actual operations by calculating energy required to heat aggregate, evaporate stockpile moisture, and heat exhaust gases from the thermodynamic properties of the materials involved. The Astec spreadsheet was shared with the NCHRP Proj- ect 9-47A team so that comparisons could be made between Astec’s thermodynamic model and empirical data obtained during NCHRP Project 9-47A. The Astec spreadsheet uses the following inputs: plant elevation, drum diameter, asphalt content, recycled (or reclaimed) asphalt pavement (RAP) content, RAP moisture content, burner fuel, burner type, production rate, ambient temperature, aggregate tempera- ture, RAP temperature, mix temperature, stack temperature, and drum type. The spreadsheet then calculates fuel usage for a range of aggregate moisture contents. The formulas used in Astec’s spreadsheet are hidden and password protected, how- ever the thermodynamic properties used in those calculations are well established and were used to back calculate actual casing loss from NCHRP Project 9-47A data. Figure 2.3 shows a comparison between measured and cal- culated fuel usage for the three projects where stack emissions tests were conducted (Michigan, Indiana, and New York). In two cases out of three, the measured fuel usage, in terms of MMBtu/ton, exceeds Astec’s predicted fuel usage for the tem- peratures and moisture contents measured during production. Analysis suggests that Astec assumes that 12% of total fuel usage is lost through the plant casing. Figure 2.4 shows a simi- lar comparison between Astec’s calculated fuel usage and the measured fuel usage from this study. However, Astec’s 12% cas- ing loss is replaced by actual backcalculated casing loss for each site. Figure 2.4 shows good agreement with the data when the casing loss is adjusted. It appears that the Astec model generally underestimates casing losses, especially for uninsulated aggre- gate dryers. It should be noted that casing losses will vary from plant to plant, depending on plant type (parallel-flow, counter- flow, double barrel, dual drum, etc.) and level of insulation. For the three plants where stack emission tests were performed, the double barrel type has the lowest casing lost, followed by the parallel-flow type. The counter-flow batch plant with bare steel dryer had the highest casing losses. The significance of this exercise is that it demonstrates that the energy analysis used in NCHRP Project 9-47A agrees with the Astec thermodynamic model except for casing loss. Astec appears to use a uniform 12% factor to estimate casing loss. Although this appears to be a reasonable assumption for Astec double barrel plants, that factor may not be accurate for other plant types. Influence of Aggregate Moisture Content A recommended best practice for both HMA and WMA is to minimize aggregate moisture content. Average aggregate mois- ture content for the Montana project was 1.4%, 1.9% lower than the average moisture content at the other sites. Measured

0.15 0.20 0.25 0.30 0.15 0.20 0.25 0.30 N CH RP 9 -4 7A M ea su re d M M Bt u Astec Predicted MMBtu Parallel-flow Double Barrel Counter-flow Line of Equality Figure 2.3. Astec fuel calculation 3.0 (predicted) vs. NCHRP Project 9-47A (measured) fuel usage. y = 1.0052x + 0.0012 R² = 0.9425 0.15 0.20 0.25 0.30 0.15 0.20 0.25 0.30 N CH RP 9 -4 7A M ea su re d M M Bt u Astec Calculated MMBtu Parallel-flow Double Barrel Counter-flow All Linear (All) Figure 2.4. Astec-predicted casing loss corrected for NCHRP 9-47A observed casing loss.

215 fuel usage for the Montana HMA was 0.157 MMBtu/ton, com- pared to an average of 0.272 MMBtu/ton for all other HMA and 0.256 for the Michigan and Indiana HMA, which were pro- duced at the same average temperature. This indicates a savings of 0.052 MMBtu/ton per percent of moisture reduction. Thus, a 1% reduction in stockpile moisture content can produce sav- ings similar to the average savings between HMA and WMA, 0.055 MMBtu/ton. Summary • To make meaningful comparisons, fuel usage between HMA and WMA should be compared over short, steady- state runs at similar production rates. • WMA mixes were produced an average of 48°F cooler than the corresponding HMA mixes, resulting, on average, in a 22.1% fuel savings. • The measured fuel savings were higher than expected based on calculations of the energy required to heat the mix and the difference in stack gas temperatures. • The additional fuel savings are attributed to casing losses— heat radiated through the drum, ductwork and baghouse, or otherwise lost. Well-insulated plants should expect lower fuel savings than uninsulated plants. • Best practices, such as burner tuning and reduced stock- pile moisture, produced reductions of similar magnitude to the use of WMA. • A high potential for error exists when making fuel usage measurements over short intervals from tank fuel depth measurements (tank sticks), natural gas meter readings, and corresponding fuel usage with tons of mix produced. A difference of 2 minutes between measurements of fuel usage and tonnage produced can result in a 3.3% error in hourly fuel usage calculations. A 1⁄10-inch error in measur- ing the tank depth of a 20,000 gallon horizontal tank at the 10,000 gallon mark results in a 34 gallon (4.715 MMBtu) error in measured fuel usage. Recommendations • Fuel savings should be based on like comparisons between WMA and HMA at the same production rate and over the same time period. • Stoichiometric fuel measurements, in accordance with EPA Method 19, should be made in conjunction with direct measurements of fuel consumption. • Care must be taken to make fuel use and cumulative ton- nage measurements at the same time and over as long an interval as possible to minimize errors due to measurement accuracy. • Recommendations from this study are incorporated into the appendix titled Documenting Emissions and Energy Reductions of WMA and Conventional HMA, included with this report.

216 Reported Emissions Reductions from WMA Given that most pollutants of concern from asphalt plants result from combustion, they can be reduced simply by reducing fuel consumption through production of warm mix asphalt (WMA). WMA’s lower discharge temperatures should also reduce binder oxidation and volatilization loss during mixing with corresponding emission reductions. However, WMA’s ability to reduce emissions is poorly verified. Stack emissions tests have been reported from 17 projects worldwide, representing six technologies (Ventura et al. 2009, Harder 2008, Davidson 2005b, Lecomte et al. 2007, Chief Environmental Group, N.D., ETE 2006, Powers 2009, Davidson and Pedlow 2007, and Middleton and Forfylow 2009). The majority of the stack tests completed to date indi- cate that WMA reduces carbon dioxide (CO2) emissions. The only case in which CO2 emissions increased (Chief Environ- mental Group N.D.) involved an emulsion that effectively increased the moisture content of the mix and required more heat to dry even at lower mix temperatures. Emis- sions of nitrogen oxides (NOx) were reduced in all cases. Sul- fur dioxide (SO2) emissions both increased and decreased. Two projects indicate increased volatile organic compounds (VOC) with the WMA production (Harder 2008, ETE 2006). In both cases, reports attributed that increase to poor burner tuning rather than to the WMA technology. Pollutants have been reported in several different units ranging from stack concentration to pounds per hour, mak- ing meaningful comparisons of any kind difficult. Too fre- quently, reported emissions are simply uncorrected average (or worse, instantaneous) dry stack concentrations (parts per million by volume, dry; abbreviated ppmvd). Comparisons between WMA and hot mix asphalt (HMA) based on differ- ences in raw stack concentrations are suspect because of dilu- tion from excess air and may be unintentionally misleading. To make meaningful comparisons between tests or runs (e.g., to compare HMA and WMA), those results must be normal- ized to a uniform percent oxygen to correct for dilution. Reports by stack test contractors that include a mass emis- sion rate in pounds per hour, as recommended by the Warm Mix Asphalt Technical Working Group—WMA TWG (2006), still cannot be compared with other runs unless normalized for production rate and expressed as pounds pollutant per unit production. Research Approach Asphalt plant exhaust gas testing targeted emissions related to multiple areas of concern—greenhouse gases (carbon foot- print), ground-level ozone precursors, condensable particu- lates (PM-10)—and an emerging concern regarding hazardous air pollutants. Energy usage, stack emissions, and temperature reductions are interrelated but can be affected by multiple fac- tors, such as aggregate moisture, operator, plant configuration, fuel type, production rate, burner tuning, percent reclaimed asphalt pavement (RAP), ambient temperature, and so forth. Variations between mix design, fuel type, production rate, and aggregate moisture were minimized to the extent possible by testing the same mix over successive days for the same project. At the three multi-technology projects (Michigan, Indiana, and New York), stack emission tests were conducted in accor- dance with the U.S. EPA’s Title 40 Code of Federal Regula- tions (CFR) Part 60, Appendix A, and generally followed the recommendations of the WMA TWG (2006). Reported stack emissions included CO2 to assess greenhouse gas produc- tion, VOC and NOx to assess the potential for ground-level ozone, carbon monoxide (CO) to assess burner tuning, SO2, condensed particulates (a component of PM-10), and form- aldehyde emissions. Results were analyzed and reported as pounds per unit production consistent with Federal AP-42 emission factors. EPA Method 1 describes the location of sampling points to divide the cross-sectional area of the stack into a number C H A P T E R 3 Stack Emissions

217 of equal areas, each of which will be sampled using a traverse point. The number of traverse points depends on the diam- eter of the stack and the distance of the sampling points from any obstructions that may cause turbulence in the stack gas flow. EPA Method 2 describes the measurement of the aver- age gas velocity in the stack. The gas velocity for each traverse point is calculated from the density of the gas and the aver- age velocity pressure measured with a Type S pitot tube. EPA Method 4 is used to determine the temperature and moisture content of the stack gas. The stack gas flow must be corrected for moisture to a dry basis because most gas analyzers operate at ambient temperature and require dry samples. Impingers are used to condense and collect water vapor from a metered gas sample drawn continuously during measurements of the stack emissions. The emission parameters evaluated and EPA test methods are shown in Table 2.4. Local emission testing contractors experienced with these methods were used to minimize mobilization costs. The proj- ect team expert assessed their credentials and coordinated test- ing at each site to ensure that meaningful data were obtained. Because of the short notice at each project, testing contractor availability became a primary selection criterion. As noted previously, burner tuning was conducted at each of the multiple technology sites prior to stack emissions tests. In two of three cases, burner tuning reduced CO emissions tenfold, while the largest WMA reduction measured was 59%. In both Michigan and Indiana, initial CO measure- ments exceeded 10,000 parts per million (ppm). In Michigan, increasing the air-to-fuel ratio dropped this level to approxi- mately 50 ppm; in Indiana, to approximately 1,000 ppm. Further reduction in CO in Indiana would have required the natural gas ports to be cleaned and the pre-mix nozzles to be replaced. Even so, the Indiana burner adjustments resulted in a 24.8% reduction in fuel use on the same mix with no other process changes. Results and Discussion Carbon Dioxide Figure 2.5 shows average CO2 emissions for each of the mixes tested during the multi-technology projects. The shaded bars indicate the average of two tests; the whiskers show the individual test results. Similar to the fuel usage as reported in Table 2.2, CO2 production is reduced for all of the WMA mixes compared to their corresponding HMA mixtures. It was noted in Indiana that during the HMA and Heritage Wax WMA testing the local stack emissions con- tractor took stack velocity readings only at the end of the run, rather than concurrently with the other emission factor samples. Based on relatively accurate gas meter readings, this appears to have resulted in an under-reporting of the actual air-flow; hence the derived lb/ton CO2 production. CO2 emissions primarily result from fuel combustion. As such, there is a linear relationship between fuel and CO2 reductions resulting from the use of WMA. Figure 2.6 pre- sents this relationship for both the data obtained from this study and the literature. The offset of any data point from the Line of Equality reflects an inaccuracy in at least one of the two measurements. Carbon Monoxide and Volatile Organic Compounds The formation of CO and VOC is affected by burner design, maintenance, and tuning. A burner that is improperly tuned or one that is poorly maintained may result in elevated levels of CO, VOC, or both. For most burners, elevated CO and VOC emissions are not a surrogate for efficiency because the energy potential of these emissions are several orders of magnitude smaller that energy loss due to excess air, high exhaust gas tem- perature, and casing radiation. Figure 2.7 and Figure 2.8 show Table 2.4. Stack emission test parameters and methods. Emission Parameter Number of Test Runs per Technology Sampling and Analytical Methodology Volumetric flow rate * EPA Methods 1 and 2 Oxygen (O2) and carbon dioxide (CO2) * EPA Method 3A Moisture content * EPA Method 4 Sulfur dioxide (SO2) 2 EPA Method 6 Nitrogen oxides (NOx) 2 EPA Method 7E Carbon monoxide (CO) 2 EPA Method 10 Total hydrocarbons (volatile organic compounds [VOC]) 2 EPA Method 25A Particulate matter/PM-10 2 EPA Methods 5/202 Formaldehyde 2 EPA Method 316 * Determined concurrently with all emission parameter

218 Figure 2.5. CO2 emission rates. 46.0 38.3 39.7 26.8 26.0 24.2 18.6 35.9 28.4 24.4 25.4 0 5 10 15 20 25 30 35 40 45 50 H M A A dv er a Ev ot he rm 3G H M A G en co r Fo am Ev ot he rm 3G H er ita ge W ax H M A Ce ca ba se RT So nn e W ar m ix Bi tu Te ch PE R MI IN NY lb /t on Figure 2.6. Reduction in fuel usage versus reduction in CO2 emissions. -30 -20 -10 0 10 20 30 40 50 60 70 100-10-20-30 20 30 40 50 60 70 Re du c on in C O 2 Em is si on (% ) Reducon in Fuel Usage (%) Literature NCHRP 9-47A Line of Equality

219 Figure 2.7. CO emissions. 0.0255 0.013 0.038 0.392 0.3255 0.6825 0.192 0.031 0.016 0.0125 0.0145 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 H M A A dv er a Ev ot he rm 3G H M A G en co r Fo am Ev ot he rm 3G H er ita ge W ax H M A Ce ca ba se RT So nn e W ar m ix Bi tu Te ch PE R MI IN NY lb /t on Figure 2.8. VOC emissions. 0.035 0.018 0.019 0.036 0.057 0.040 0.048 0.013 0.016 0.021 0.019 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 H M A A dv er a Ev ot he rm 3G H M A G en co r Fo am Ev ot he rm 3G H er ita ge W ax H M A Ce ca ba se RT So nn e W ar m ix Bi tu Te ch PE R MI IN NY lb /t on

220 the CO and VOC emissions, respectively. Overall, CO emissions were elevated at the Indiana site compared to the other sites. As noted previously, CO emissions exceeded 10,000 ppm at the Indiana site prior to burner tuning. After tuning, CO emissions for HMA were reduced to approximately 1,000 ppm. Burner maintenance issues also resulted in elevated VOC. Additional reductions would have required cleaning out the natural gas ports and replacing the burner pre-mix nozzles, tasks that could not be performed within the time allowed for the WMA demonstration. The thin horizontal line in Figure 2.7 represents the EPA’s candidate emission factor for CO of 0.13 lb/ton for drum plants (RTI International 2004) based on stack test data from 18 drum plants. The range in data averages 89.5% of the mean. Although the CO emissions for the Michigan and Indiana Evotherm 3G appear elevated compared to their correspond- ing HMA controls, both values are within 89.5% of the HMA, indicating that they are within typical testing variability. For the Michigan parallel-flow drum plant, WMA pro- duction reduced VOC emissions by approximately 50%. A counter-flow drum plant was used in Indiana. One of the Indiana HMA VOC readings (0.012 lb/ton) appears to be an outlier. The stack test contractor had problems with the high stack moisture content and took the analyzer off-line fre- quently during the run to “dry out”. Excluding that run, the HMA reading would be 0.059 lb/ton and all of the WMA results would reflect a reduction. For the New York batch dryer, all of the VOC readings for the WMA mixes were higher than those for the HMA control. However it should be noted that VOC emissions for all mixes were among the lowest measured and reflect state-of-the-art performance in most jurisdictions. A variety of factors could explain an increase with WMA, but the uniform increase across three very different WMA tech- nologies suggests causes other than WMA itself. Sulfur Dioxide When fuels containing sulfur are burned, SO2 is produced. Sulfur content varies with fuel type. Recycled fuel oil tends to have the highest sulfur content, followed by fuel oil. Natural gas tends to have the lowest concentration of sulfur in fuels commonly used at asphalt plants. Reducing fuel consumption should reduce SO2 production. Figure 2.9 shows the SO2 stack readings for the three multi-technology projects. Overall, the SO2 emissions from Indiana and New York, both of which used natural gas as fuel, are inconsequential. The spike for the Indiana Gencor Foam could be attributed to a small amount of slag making its way into the mix. The 50% reductions in SO2 for the Michigan WMAs, in which the plant used recycled fuel oil, are significant. Discounting possible changes in the recycled oil supply, the 50% reduction suggests an increase in SO2 control efficiency at lower WMA baghouse temperatures. As might be expected, at lower baghouse temperatures more SO2 condenses out of the exhaust gas stream, is captured by 0.019 0.009 0.008 0.000 0.005 0.000 0.002 0.003 0.001 0.000 0.000 0.000 0.005 0.010 0.015 0.020 0.025 H M A A dv er a Ev ot he rm 3G H M A G en co r Fo am Ev ot he rm 3G H er ita ge W ax H M A Ce ca ba se RT So nn e W ar m ix Bi tu Te ch PE R MI IN NY lb /t on Figure 2.9. SO2 emissions.

221 the baghouse fines, and then becomes encapsulated in the WMA. For reference, the EPA’s candidate emission factor for a drum plant using recycled fuel oil is 0.058 lb/ton and for natural gas it is 0.0034 lb/ton (RTI International 2004). Nitrogen Oxides NOx emissions are a precursor to the formation of ground- level ozone. NOx emissions are higher for fuel oils compared to natural gas. The EPA’s candidate emission factor is 0.055 lb/ton for drum plants burning fuel oil and 0.026 lb/ton for drum plants burning natural gas. Figure 2.10 shows the NOx stack readings for the three multi-technology projects. For the Mich- igan tests, Advera had lower NOx emissions and the Evotherm 3G the same NOx emissions as the HMA. For the Evotherm 3G, the burner was set at an average firing rate of 26% compared to 75% for the HMA and 43% for Advera. This low firing rate may have resulted in greater excess air available to form NOx, increasing NOx emissions. For the Indiana tests, the WMA mixes produced the same or lower NOx emissions than the HMA. For the New York City tests, each of the WMA mixes yielded lower NOx emissions than did the HMA. Formaldehyde Figure 2.11 shows frequency distributions of formal- dehyde emissions reported in numerous test programs. Figure 2.10. NOx emissions. 0.064 0.058 0.064 0.023 0.022 0.021 0.015 0.031 0.024 0.019 0.022 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 H M A A dv er a Ev o th er m 3G H M A G en co r Fo am Ev o th er m 3G H er ita g e W ax H M A Ce ca ba se RT So nn e W ar m ix Bi tu Te ch PE R MI IN NY lb /t on EPA Candidate Emission Factor for Drum Plantsburning Natural Gas EPA Candidate Emission Factor for Drum Plants burning Fuel Oil Formaldehyde is a typical byproduct of combustion for all carbon-based fuels. The distribution of formaldehyde emis- sions for the WMA mixes is lower than the distribution for the HMA mixes tested as part of this study. Only four stack emissions results for formaldehyde are available in the EPA’s AP-42 database (RTI International 2004). The industry HMA data shown in Figure 2.11 represent 24 formaldehyde stack emissions tests from the mid-Atlantic region. The WMA formaldehyde emissions are similar to these levels. The results from this study also show that lower formal- dehyde concentrations were measured for WMA compared to HMA. PM-10 Particulate matter (PM), especially fine particulates (e.g., PM-10), are of increasing concern among many environ- mental agencies. Figure 2.12 shows average condensable fraction (back half of a Method 5 sample train) for HMA and WMA technologies. Filterable particulates were not measured. The condensable fraction includes organic and inorganic compounds with organics less than 1/10 of total condensables. What is striking about the NCHRP Proj- ect 9-47A PM-10 data is the scale of PM-10 emissions from limestone aggregates and parallel-flow dryers (Michigan data), and the resulting reduction achieved by WMA tech- nologies and igneous aggregate.

222 Figure 2.11. Frequency distribution of formaldehyde emissions. EPA AP 42 NCHRP 9 47A HMA Industry HMA NCHRP 9 47A WMA 0 10 20 30 40 50 60 70 0.005 0.006 0.007 0.0080.0010 0.002 0.003 0.004 O bs er ve d Fr eq ue nc y (% ) Formaldehyde Emissions (lbs/ton) Figure 2.12. PM-10 condensable fraction. Traprock Aggregate Limestone Aggregate

223 Summary Stack emissions were measured on three multi-technology projects consisting of a total of eight WMA mixes and three corresponding HMA control mixes. • As expected, reduced fuel consumption resulted in reduced CO2 emissions for all of the WMA mixes. • For the stack emissions (multi-technology) sites, on average a 52°F reduction in temperature resulted in a 21% reduction in fuel usage and a 20% reduction in CO2 emissions. • CO and VOC emissions are related to burner design, maintenance, and tuning. • CO emissions for the WMA were within normal testing variability of HMA. There appears to be no indication of reduced CO emissions with WMA. • For the Michigan parallel-flow drum plant, WMA resulted in a 50% reduction in VOC emissions. • For the New York project, VOC emissions were higher for the WMA, but comparable with the Michigan VOC emissions. • The HMA data for the Indiana project were highly vari- able; a slight reduction in VOC emissions is indicated if one point is considered an outlier. • SO2 emissions from reclaimed oil fuel were significantly reduced with WMA, suggesting better control at lower baghouse temperatures. • Use of WMA generally resulted in slight reductions of NOx. Recommendations To make meaningful comparisons between HMA and WMA, it is suggested that companion tests be performed using the same fuel, mixture, and production rate.

224 Background The primary use of asphalt has been in paving mixes for roadway infrastructures. The United States and Europe com- bined employ about 400,000 workers in the asphalt paving industry (AI, EU 2011). Asphalt is the non-distillable frac- tion of crude oil. Small amounts of volatile and semi-volatile organic compounds are trapped in this highly viscous material (Clark et al. 2011). Heating asphalt above the softening point and agitating it facilitates the release of these emissions, consti- tuting the potential for worker exposure. A large nested case control epidemiology study by Olsson et al. (2010) showed no consistent evidence of an association between indicators of either inhalation or dermal exposure to asphalt and lung cancer risk and attributed increased inci- dence in cancer to confounding issues like smoking, exposure to coal tar, and so forth. A recent 2-year skin-painting study by Clark et al. (2011) confirmed the absence of tumorigenic effects in skin regions treated with paving asphalt fume condensate. Fuhst et al. (2007) conducted an inhalation study involving exposure of Wistar rats to asphalt fumes for 2 years. Results showed that asphalt fumes are not considered to be tumorigenic to rats via the inhalation route. Asphalt-related irritant effects were, however, observed in the nasal passages and in the lungs. Despite the results of these recent studies, in October 2011, the International Agency for Research on Cancer con- cluded that “occupational exposures to straight-run bitumens (asphalts) and their emissions during road paving are ‘possibly carcinogenic to humans’ (Group 2B)” (Lauby-Secretan et al., 2011, IARC 2012). Other studies have also shown an association with vari- ous health endpoints related to irritation. A recent German study in humans by Raulf-Heimsoth et al. (2011) detected potentially sub-chronic irritative inflammatory effects in the lower airways of bitumen-exposed workers. Tepper et al. (2006) reported throat symptoms that were statistically sig- nificant compared to a control group. Similar symptoms are discussed in the Norseth et al. studies (1991) evaluating self- reported symptoms that included fatigue, reduced appetite, eye irritation, and laryngeal-pharyngeal irritation that was reported more frequently among workers exposed to asphalt fumes than among unexposed workers in a statistically sig- nificant manner. These studies emphasize the need for reducing worker exposure to asphalt emissions. The National Institute for Occupational Safety and Health (NIOSH) has recommended use of engineering controls and good work practices to mini- mize worker exposure to asphalt fumes (NIOSH 2001), including reduction of the asphalt mix temperature. First developed in Europe during the late 1990s to address worker exposure concerns for Gussasphalt placed at high tempera- tures, warm mix asphalt (WMA) mixtures typically are pro- duced at lower temperatures than hot mix asphalt (HMA) mixtures (D’Angelo et al. 2007). Recent studies (Kriech et al. 2011) show that reduced asphalt application temperature is predictive of reduced inhalation exposures (Cavallari et al. 2011) along with a reduced total absorbed dose of poly cyclic aromatic hydrocarbons (PAHs) and polycyclic aromatic compound (PAC) metabolites (McClean et al. 2012). Information currently in the public domain regarding worker exposure reduction using WMA technologies is often based on marketing or takes the form of presentations, confer- ence proceedings, or government reports. Few peer-reviewed publications specifically document this promoted benefit of WMA. D’Angelo et al. (2007) indicated 30%–50% reductions in asphalt aerosols/fumes and PAHs for WMA compared to HMA. Measurements by von Devivere et al. (2003) showed a reduction in fume emissions of 75% where zeolite had been added with an application temperature reduction of 26°C. In a study of WAM-Foam, exposure values were shown to be in the lower range when compared to exposure measure- ments conducted on paving HMA (Lecomte et al. 2007). The Ohio Department of Transportation (DOT), in conjunction C H A P T E R 4 Worker Exposure

225 with Flexible Pavements of Ohio, showed that WMA reduced emissions by 35%–65% (EES Group 2006, Powers 2009). Shifa et al. (2009) claimed a 90.2% reduction of asphalt fumes for emulsion-based WMA in a long tunnel pavement study. NCHRP Project 9-47A was designed to compare WMA technologies to traditional HMA applications under simi- lar conditions, controlling many (albeit not all) variables in the field to allow a side-by-side comparison of the worker breathing zone exposures. Three WMA technologies were compared to one HMA technology at each of two sites—one in Indiana and one in New York. Research Approach Study Population During each sampling event, four workers per crew were studied: the paver operator, two screed operators (includ- ing, in Indiana, the site foreman), and the raker. Of the entire crew, these four workers are exposed to asphalt at the hottest temperature, so they have the greatest potential for asphalt emissions exposure. Study Design The eight workers in the two crews were monitored for four consecutive days. During one day the crew performed under normal working conditions using HMA. During the other three days, the crew performed under similar condi- tions, but using a different WMA technology each day. To avoid interference with assessment of asphalt emissions, no diesel oil was used as a release/cleaning agent. Within a given site, controlled variables included asphalt source, aggregate, amount of reclaimed asphalt pavement, plant, paving equip- ment, crew, and similar traffic patterns (paved in congruent locations). Paving machines were equipped with properly functioning engineering controls. During each sampling event, meteorological data were also recorded, including ambient air temperatures, wind speed, and humidity. Whereas many studies measure mixture temperature at the production facility, for this study, application tempera- tures were monitored at the back of the screed area six times throughout the workday using an 8-inch dial stem thermom- eter in the newly placed mat. For the Indiana crew, diesel oil normally used as a release agent and to clean tools and equipment was removed from the site and replaced with B100 biodiesel oil (Bajpai and Tyagi 2006) (CAS Number: 67784-80-9). Biodiesel contains no straight chain hydrocarbons or PACs. Workers at the New York site did not use diesel oil; instead, they use a water- based product called FO™ Release II (Fine Organics Corp), also free of straight chain hydrocarbons or PACs. Collection and Analysis of Breathing Zone Samples Each worker wore two sorbent tube samplers contain- ing XAD-2 and charcoal (150 mg XAD-2 followed by 50 mg activated charcoal; see Figure 2.13). A 1-inch piece of Tygon® tubing (dichloromethane rinsed) was added to the end of tube, once broken, to protect the workers. Care was taken to break the inlet end of the tube to 4 mm to equal the NIOSH sampler. Set to a flow rate of 2.0 ± 0.2 L/min, pumps were calibrated pre-shift and re-measured post-shift. One background sample was collected each day or experiment, positioned upwind of the paving operation. A field blank was collected on each day or experiment for each crew. Sorbent tubes were eluted with 5 mL dichloromethane, charcoal-end up. Sampler Selection Justification Table 2.5 shows internal data compiled from previous studies conducted by Heritage Research Group that included breathing zone monitoring of workers during three differ- ent WMA applications. In the previous studies, each worker was monitored in a similar manner as that described above. However, in these studies the breathing zone air entered a membrane filter first, followed by the sorbent tube, allow- ing determination of total particulates (TP), benzene solu- ble fraction (BSF), and total organic matter (TOM) using NIOSH Method 5042 (NIOSH 2006). BSF levels were all below the level of detection (LOD), hindering quantita- tive comparisons to HMA. Because all samples contained detectable levels of TOM, this was selected as the primary tool for evaluating differences in exposure between HMA and WMA. Figure 2.13. A worker at the Indiana site wearing two XAD-2/charcoal sorbent tubes for collection of breathing zone exposures.

226 Total Organic Matter TOM (Kriech et al. 2002a) included hydrocarbons rang- ing from 6 to 42 carbons (C6 to C42) as determined by gas chromatography/flame ionization detection (GC/FID). A Var- ian model 3400 GC with a 1077 split/splitless injector (set at 250°C) was used, with a 5% phenyl/95% methyl-polysiloxane column (30 m × 0.33 mm ID, 0.25 µm film thickness; Restek RTX-5); hydrogen carrier gas was set at 2 mL/min. With the detector at 310°C, the oven temperature program was 40°C held for 3 minutes, increased to 120°C at 9°C/min, held for 0.5 minutes, then ramped to 305°C at 11°C/min, and held for 10.89 minutes. Calibration included kerosene standards for quantification of the TOM. Polycyclic Aromatic Compounds Forty PACs (see Table 2.6) were determined using gas chromatography/time-of-flight mass spectrometry (GC/ TOFMS) following a modified version of a published proce- dure (Kriech et al. 2002b). A Leco Pegasus II GC/TOFMS was used with a source temperature of 275°C, transfer line tem- perature of 300°C, mass range of 35–400, and five spectra/sec with a split/splitless injector (in splitless mode, set at 300°C). A Varian Select PAH column was used (30 m × 0.25 mm ID, 0.15 µm film thickness; Varian CP 7462). Helium carrier gas rate was 2.0 mL/minute. The oven temperature program was 50°C held for 0.7 minutes, ramped to 180°C at 85°C/minute and held for 0 minutes, then to 230°C at 3°C/minute and held for 7 minutes, to 280°C at 28°C/minute and held for 10 minutes, and finally taken to 350°C at 14°C/minute and held for 5 minutes. Four standards supplied by AccuStandard Inc. and three from Sigma-Aldrich were used. AccuStandard Inc. standards included a mix of 24 PACs, a custom-order standard of nine PACs, dibenzo[a,e]fluoranthene, and thi- anaphthene. Sigma-Aldrich standards included dibenz[c,h] acridine, benz[a]acridine and dibenz[c,h]acridine. Prior to injection, an internal standard mix was added to each cali- bration standard and sample (10 µL to each 100 µL aliquot). Only the samples with the highest TOM values per experi- ment were analyzed by GC/TOFMS. Supplier and catalogue number information for the products described are provided in the appendix. Nine of 13 PACs listed as agents reviewed by the Inter- national Agency for Research on Cancer (IARC) in Volume 103 for “asphalt and asphalt fumes, and some heterocyclic PACs” were included in the analysis. Four 6-ring PACs are also on the IARC list but were not tested due to lack of available standards. Results Average HMA mat temperatures for each experiment are presented in Table 2.7. New York HMA temperatures were an average of 35°C higher than those at the Indiana site. Differ- ences between the HMA and WMA experiments in Indiana were only 15°C or less, whereas the New York mat tempera- Table 2.5. Summary of data made available from prior studies. Technology Worker mg/m3 Total Particulates Benzene Soluble Fraction Total Organic Matter WMA-1 Raker left 0.69 bdl 0.73 WMA-1 Raker right 0.54 bdl 0.97 WMA-1 Screed area 1.11 bdl 0.66 WMA-1 Screed area 0.78 bdl 1.03 WMA-1 Operator left 0.91 bdl 0.91 WMA-1 Operator right 0.70 bdl 0.93 WMA-1 Operator area 0.55 bdl 1.67 WMA-1 Screed area 0.81 bdl 1.00 WMA-2 Screed operator 0.13 bdl 0.56 WMA-2 Screed operator 0.16 bdl 0.55 WMA-2 Operator 0.20 bdl 0.42 WMA-2 Raker 0.17 bdl 0.59 WMA-2 Raker 0.18 bdl 0.57 WMA-3 Screed operator 0.76 Bdl 0.99 Average 0.52 <0.04 0.81 bdl = below detection limit Source: Heritage Research Group.

227 tures were ≥44°C lower for the WMA as compared to the cor- responding HMA. In fact, the HMA at Indiana was within the normal temperature range for WMA (100–140°C). The HMA from the New York site had an average mat temperature of 161°C, well within the typical HMA range (150–180°C). Both sites used PG 64-22 asphalt for the HMA and WMA mixes. The source of asphalt was different between Indiana and New York, but was the same within each location. The paver machines were very different for the two sites. At the New York site, one paver was used the first 2 days of sampling, but it then experienced mechanical problems. On the third day, after 3–4 hours trying to fix the paver, a differ- ent paver was used. Meteorological data during each sampling event was also recorded (see Table 2.8). These data include ambient air tem- peratures, wind speed, and humidity, with the table showing the average and range of recorded data. TOM results are listed in Table 2.9 for Indiana and Table 2.10 for New York, with summary statistics shown in Table 2.11. Average data are also shown graphically in Figure 2.14 with a confidence interval of 95% (CI95%). Background and blank data were all below the LOD of ~0.04 mg/m3. Breathing zone results show that TOM concentrations for the New York site were substantially higher than those for the Indiana site. WMA arithmetic mean data compared to the correspond- ing HMA arithmetic mean data resulted in a minimum of 33% reduction in TOM exposures, with the exception of the Indiana Evotherm 3G, which was 8.4% higher. The New York TOM data showed a statistically significant difference between the HMA reference and the collective WMA technologies (95% confidence intervals [CI95%] were 1.90–2.52 mg/m3 and 1.29–1.54 mg/m3 respectively). For the Indiana data, there was not a statistically significant difference between the HMA and the collective WMA technologies (CI95% were Table 2.6. PACs investigated in eight worker breathing zone samples. Benzene Rings CAS No. PAC CAS No. PAC 1. 1+ 95-15-8 Benzothiophene 21. 5522-43-0 1-Nitropyrene 2. 2 91-20-3 Naphthalene 22. 27208-37-3 Cyclopenta[cd]pyrene 3. 2+ 83-32-9 Acenaphthene 23. 205-99-2 Benzo[b]fluoranthene 4. 2+ 208-96-8 Acenaphthylene 24. 205-82-3 Benzo[j]fluoranthene 5. 2+ 225-11-6 Benz[a]acridine 25. 207-08-9 Benzo[k]fluoranthene 6. 2+ 225-51-4 Benz[c]acridine 26. 194-59-2 7H-Dibenzo[c,g]carbazole 7. 2+ 86-74-8 Carbazole 27. 56-49-5 3-Methylcholanthrene 8. 2+ 132-65-0 Dibenzothiophene 28. 50-32-8 Benzo[a]pyrene 9. 2+ 86-73-7 Fluorene 29. 192-97-2 Benzo[e]pyrene 10. 3 120-12-7 Anthracene 30. 53-70-3 Dibenz[a,h]anthracene 11. 3 85-01-8 Phenanthrene 31. 226-36-8 Dibenz[a,h]acridine 12. 3+ 239-35-0 Benzo[b]naphtho[2,1-d]thiophene 32. 224-42-0 Dibenz[a,j]acridine 13. 3+ 206-44-0 Fluoranthene 33. 224-53-3 Dibenz[c,h]acridine 14. 3+ 243-46-9 Benzo[b]naphtho[2,3-d]thiophene 34. 2997-45-7 Dibenzo[a,e]fluoranthene 15. 4 56-55-3 Benz[a]anthracene 35. 193-39-5 Indeno[1,2,3-cd]pyrene 16. 4 3697-24-3 5-Methylchrysene 36. 191-24-2 Benzo[ghi]perylene 17. 4 218-01-9 Chrysene 37. 192-65-4 Dibenzo[a,e]pyrene 18. 4 129-00-0 Pyrene 38. 189-55-9 Benzo[rst]pentaphene 19. 4 57-97-6 7,12-Dimethylbenz[a]anthracene 39. 189-64-0 Dibenzo[a,h]pyrene 20. 4 217-59-4 Triphenylene 40. 191-30-0 Dibenzo[a,l]pyrene Benzene rings: the number of 6-membered (or 6-sided) aromatic rings in the structure—a + in this column indicates one additional 4- or 5-sided ring within the structure. CAS No.: Chemical Abstracts Service registry number. PAC: polycyclic aromatic compound. In this column, the compounds in shaded cells represent 9 of 13 PACs recently listed by IARC as their preliminary list of agents to be reviewed for asphalt and asphalt fumes. Table 2.7. Average temperature of the asphalt mat directly behind the screed for each experiment. Mix Temperature Behind Screed (°C) Difference (°C) HMA, Indiana 126 Indiana reference Gencor Foam 114 12 Evotherm 3G 111 15 Heritage Wax 116 10 HMA, New York 161 New York reference Cecabase RT 106 55 SonneWarmix 109 52 BituTech PER 117 44

228 Average High Low Average High Low Average High Low 9/14/2010 IN HMA 74.7 78.3 66.7 3.2 0.0 10.6 47.7 71.6 34.7 9/15/2010 IN WMA 73.8 83.5 61.4 2.6 1.0 4.5 49.1 67.6 35.4 9/16/2010 IN WMA 68.4 70.5 66.0 6.5 3.0 13.9 78.3 86.7 69.0 9/16/2010 IN WMA 70.3 74.4 66.9 5.0 3.0 6.9 60.6 69.0 52.9 10/19/2010 NY WMA 54.5 56.0 53.2 3.1 5.0 1.2 58.7 74.2 35.0 10/20/2010 NY HMA 56.6 61.0 51.2 1.0 1.8 0.0 48.8 60.0 35.0 10/21/2010 NY WMA 56.3 61.0 52.0 9.3 12.0 7.1 69.5 81.0 52.0 10/22/2010 NY WMA 45.8 48.0 45.0 12.5 16.0 10.0 45.8 53.0 42.0 Temp oF Wind Speed (mph) Humidity %Date Loca‡on Type Table 2.8. Meteorological data during the sampling events. Product Date Tonnage Lab ID Description Minutes 1 L Air2 TOM (mg/m3) Experiment Average TOM (mg/m3) H ot M ix , I nd ia na 9/ 14 /2 01 0 12 00 51 Operator 350 721 0.30 0.32 52 Operator 285 581 0.17 53 Raker 429 875 0.25 54 Raker 430 854 0.24 55 Screed operator 430 851 0.52 56 Screed operator 430 858 0.53 57 Foreman 430 894 0.21 58 Foreman 430 882 0.33 G en co r F oa m 9/ 15 /2 01 0 11 87 61 Operator 425 871 0.05 0.12 62 Operator 425 876 0.05 63 Screed operator 424 837 0.13 64 Screed operator 424 854 0.09 65 Raker 419 848 0.12 66 Raker 422 850 0.11 67 Foreman 432 886 0.19 68 Foreman 432 873 0.25 Ev ot he rm 3 G 9/ 16 /2 01 0 88 1 71 Operator 262 542 0.27 0.34 72 Operator 262 542 0.30 73 Screed operator 268 531 0.45 74 Screed operator 268 547 0.58 75 Raker 267 545 0.27 76 Raker 267 542 0.31 77 Foreman 264 546 0.29 78 Foreman 264 539 0.30 H er ita ge W ax 9/ 16 /2 01 0 89 0 81 Operator 225 464 0.04 0.15 82 Operator 225 467 0.05 83 Screed operator 227 452 0.24 84 Screed operator 227 462 0.30 85 Raker 228 463 0.12 86 Raker 228 462 0.12 87 Foreman 230 475 0.12 88 Foreman 230 470 0.18 1 Time sample collector running 2 Liters of air collected by sampler Table 2.9. Indiana site information and TOM data for all samples.

229 Product Date Tonnage Lab ID Description Minutes 1 L Air2 TOM (mg/m3) Experiment Average TOM (mg/m3) H ot M ix , N ew Y or k 10 /2 0/ 20 10 11 00 49 Operator 430 837 2.78 2.21 50 Operator 430 834 2.97 51 Screed operator 436 859 2.15 52 Screed operator 436 857 1.62 53 Raker 447 871 1.84 54 Raker 447 896 1.91 55 Laborer 434 862 2.21 56 Laborer 434 860 2.20 Ce ca ba se R T 10 /1 9/ 20 10 80 0 41 Operator 377 744 1.46 1.17 40 Operator 377 752 1.78 42 Screed operator 370 738 1.02 38 Screed operator 370 733 1.31 44 Raker 373 724 1.11 43 Raker 376 759 1.25 39 Laborer 387 777 0.58 37 Laborer 387 778 0.87 So nn eW ar m ix 10 /2 1/ 20 10 78 0 61 Operator 345 695 1.79 1.40 62 Operator 345 667 1.57 63 Screed operator 352 683 1.37 64 Screed operator 352 681 1.46 65 Raker 362 723 1.29 66 Raker 362 721 0.78 67 Laborer 385 765 1.41 68 Laborer 385 759 1.51 B itu Te ch P ER 10 /2 2/ 20 10 79 8 73 Operator 346 696 2.14 1.48 74 Operator 347 700 1.81 75 Screed operator 382 764 1.60 76 Screed operator 382 745 1.58 77 Raker 342 691 1.77 78 Raker 343 680 1.73 79 Laborer 388 770 1.48 80 Laborer 387 766 1.33 1 Time sample collector running 2 Liters of air collected by sampler Table 2.10. New York site information and TOM data for all samples. mg/m3 WMA, New York HMA, New York WMA, Indiana HMA, Indiana Average 1.42 2.21 0.21 0.32 Minimum 0.58 1.62 0.04 0.17 Maximum 2.14 2.97 0.58 0.53 Standard Deviation 0.36 0.46 0.13 0.14 Number 24 8 24 8 Table 2.11. Summary statistics for TOM data.

230 0.23–0.41 mg/m3 and 0.16–0.25 mg/m3, respectively). Given that the Indiana HMA was applied at WMA temperatures, this was not surprising. Evaluation of the CI95% for each indi- vidual WMA showed that, other than the Indiana Evotherm 3G, all the WMA were lower than their corresponding HMA as displayed in Figure 2.14. Overall, use of these six WMA technologies resulted in lower application temperatures that subsequently resulted in lower TOM exposures within the paving worker breathing zones. PAC results are shown in Table 2.12 for the samples with the highest TOM concentrations per experiment. Only one 4–6 ring PAC (pyrene) was detected in these eight samples. Of the 2–3 ring PACs, naphthalene was detected at the highest concentration. Because only the highest samples were tested, comparisons between HMA and WMA were not made. Table 2.12. PAC results for the samples with the highest total organic matter concentrations per site/treatment. Indiana Site New York Site Ring Size µg/m 3 HMA, Indiana Back- ground HMA, Indiana Field Blank HMA, Screed Operator Gencor Foam Fore- man Evotherm 3G Screed Operator Heritage Wax Screed Operator HMA, Operator Cecabase RT Operator Sonne- Warmix Operator Bitu- Tech PER Operator 1+ Benzothiophene 0.06 bdl 0.22 0.07 0.23 0.13 0.10 bdl bdl bdl 2 Naphthalene 0.15 bdl 3.60 2.16 5.42 2.74 2.46 2.13 1.91 4.13 2+ Acenaphthene Bdl bdl 0.28 0.10 0.16 0.15 0.06 bdl bdl bdl 2+ Acenaphthylene Bdl bdl bdl bdl bdl Bdl bdl bdl bdl 0.31 2+ Dibenzothiophene Bdl bdl 0.25 0.12 0.15 0.19 0.07 bdl bdl bdl 2+ Fluorene bdl bdl 0.35 0.13 0.19 0.18 0.10 bdl bdl 0.10 3 Anthracene bdl bdl 0.06 bdl bdl Bdl bdl bdl bdl bdl 3 Phenanthrene bdl bdl 0.75 0.27 0.38 0.39 0.11 bdl bdl 0.13 3+ Fluoranthene bdl bdl 0.14 0.06 bdl Bdl bdl bdl bdl bdl 4 Pyrene bdl bdl 0.11 bdl bdl Bdl bdl bdl bdl bdl Sum of detectable PACs* 0.45 0.40 5.80 3.04 6.80 4.09 3.07 2.56 2.32 4.97 Detection limits 0.049 0.061 0.058 0.057 0.091 0.11 0.060 0.066 0.072 0.072 * PAC: polycyclic aromatic compound. Within this data set, when bdl (below detection limit), the detection limit divided by the square root of 2 was used for the summation. Note: the Evotherm 3G in Indiana showed no statistical difference compared with HMA in Indiana Figure 2.14. TOM—95% confidence intervals.

231 Figure 2.15. TGA on the PG 64-22 asphalt binders used in New York and Indiana for this study. Discussion Average TOM data from previous HMA studies (1.69 mg/m3) (Kriech et al. 2002) showed lower results than seen at the New York site (2.21 mg/m3). Although the New York HMA temperatures were significantly higher than the Indiana HMA, the WMA temperatures were similar, yet the TOM con- centrations were seven times higher in New York. Although the asphalt grades were both PG 64-22, the sources of the asphalt were different and likely the most prominent factor contributing to the differences. To confirm that the asphalt source was the cause, a sample of each HMA obtained dur- ing the study was Soxhlet extracted to separate the binder from mineral aggregates. After evaporation of the dichloro- methane solvent, each binder was tested using thermal gravi- metric analysis (TGA). TGA is performed on samples to determine changes in weight in relation to changes in temperature. Previous studies have used this technique to evaluate various roofing asphalts (Kuszewski et al. 1997). Overlays are shown in Figure 2.15 for the two asphalts. An expanded view of the region from 100°C to 250°C shows the application temperatures used in this study. It is evident, based on its higher weight loss, that the New York binder is more volatile than the Indiana binder until the crossover at ~236°C, which is well above the applica- tion temperatures employed. It is difficult to directly compare these results with other published data. For example, Shifa et al. (2009) reported 21.1 mg/m3 bitumen fume for HMA versus 2.06 mg/m3 for WMA (a 90.2% reduction), but methods used and loca- tion of sampling are not provided. Shifa et al. also reported results for benzopyrene (HMA = 0.094 mg/m3 versus WMA = 0.019 mg/m3), whereas no benzo[a]pyrene was detected in either HMA or WMA on worker samples in this study (aver- age LOD = 0.07 µg/m3). Lecomte et al. (2007) concluded that the volatile fraction was higher (up to six times more) for HMA and represented almost all the organic emissions (up to 99%). This is consis- tent with Heritage Research Group studies in that the BSF were also below the LOD. Also consistent with internal Heri- tage Research Group data, a report by the Virginia Transpor- tation Research Council (Diefenderfer et al. 2007) showed all worker results below the LOD of 0.08 mg/m3 for BSF. It is interesting to note that the highest TOM concen- trations occurred for the screed operator/foreman in Indi- ana. However, in New York, the paver operator consistently received the highest exposure levels. This may be due to design differences between the types of paver machines, or may be related to landscape differences (i.e., with connected 2-story and 3-story buildings in New York creating an almost tunnel effect compared to the open, more rural Indiana landscape).

232 All TOM results were above the LOD, demonstrating that it is a useful measure for assessing reductions in worker breathing zone exposures with the use of WMA. Results for these two sites appeared to bracket the high and low ends of the spectrum of asphalt paving worker breathing zone exposures. Summary • Overall, use of these six WMA technologies resulted in lower application temperatures compared to their corresponding HMA; yielding an average 36% reduc- tion in TOM exposures within the paving worker breath- ing zones. • Exposures using WMA are not the same across technologies. • Twenty-two of the 40 individual PACs tested were below the LOD for the eight samples tested. – Naphthalene was detected at the highest concentration. – Only one 4–6 ring PAC (pyrene) was detected in any of these worker breathing zone samples and it was in a HMA sample. – The nine PACs tested that are part of the compounds IARC has reviewed for asphalt, asphalt fumes, and some heterocyclic PACs were all below the LOD. – Since only one 4-ring PAC was detected, it is unlikely that the 6-ring compounds not included in this study were present. – Not all asphalts are the same; in this study, the different sources resulted in significantly different breathing zone exposure levels.

233 Findings An objective of NCHRP Project 9-47A was to provide rela- tive emissions measurements of warm mix asphalt (WMA) technologies as compared to conventional hot mix asphalt (HMA). NCHRP Report 779 addresses this objective in terms of fuel usage, plant stack emissions, and worker exposure. The research conducted under this portion of NCHRP Proj- ect 9-47A included the following actions: 1. Monitoring fuel usage for six of eight projects that con- sisted of seven HMA control mixtures and 11 WMA mixtures. 2. Measuring stack emissions of duplicate production runs at three projects that had a total of three HMA (control) and eight WMA mixtures (22 total measurements). 3. Refining procedures for collecting and analyzing worker exposure based on literature review and previous testing of HMA and WMA mixtures to use total organic matter (TOM) instead of benzene soluble fraction (BSF). 4. Collecting worker exposure during a production day to TOM at two multi-technology projects that consisted of two HMA controls and six WMA mixes. 5. Developing revised recommendations for monitoring fuel usage using stack emission data to evaluate perceived energy usage for asphalt mixture production using natural gas and reclaimed oil fuels. Based on the study, the Test Framework for Documenting Emissions and Energy Reductions of WMA and Conventional HMA was revised. Fuel Usage Data were presented to show the importance of compar- ing the energy usage of new technologies, such as WMA, to HMA over similar, typically short, steady-state, time frames. Historical fuel usage data, available for HMA, typically include fuel used for plant start-up, plant waste, and end of run cleanout. In NCHRP Project 9-47A, the average reduc- tion in mix temperature of 48°F (27°C) associated with WMA production resulted in average fuel savings of 22.1%. This was higher than predictions based on thermodynamic material properties. The increased fuel savings appear to result from larger than expected casing losses—heat radiated through the plant’s metal into the surrounding environment instead of being transferred to the mix for both HMA and WMA. Potential errors were identified for direct measures of fuel usage such as tank sticks and gas meter readings by comparing measured fuel usage to fuel usage calculated from stoichiometric plant stack emissions. Gas meters were found to update usage only after large time intervals, on the order of 30 minutes for some meters, inducing error. Best practices suggest using methodologies to reduce aggregate stockpile moisture, such as sloping stockpile areas away from plant, loading on high side of sloped surface, and covering stockpiles with high fines content to reduce fuel usage. Significant fuel savings were demonstrated for one project with low stockpile moisture contents. Another recommended best practice for improving plant fuel efficiency is to conduct routine burner maintenance including nozzle cleaning and tuning of linkages to achieve proper fuel to air ratios over the plant’s normal pro- duction rates. One plant in this study had a 24.8% reduction in fuel usage after burner tuning. Stack Emissions Greenhouse gas emissions such as carbon dioxide (CO2) decreased with reduced fuel usage. Carbon monoxide (CO) and volatile organic compound (VOC) measurements appear to be more related to burner maintenance and tuning and less related to reductions in fuel usage and consequently the use of WMA. One project, with a parallel-flow dryer using reclaimed oil as fuel, indicated a reduction in VOC when producing C H A P T E R 5 Findings and Conclusions

234 WMA. Significant reductions in sulfur dioxide (SO2) were observed for the same project. The two other projects used natural gas as fuel, which has lower sulfur content. Nitrogen oxides (NOx) are a precursor to the formation of ground-level ozone. NOx emissions are also higher for fuel oils compared to natural gas. With one exception, small reductions in NOx were noted for WMA. For the exception, the burner was set at 26% of its firing rate for the WMA, compared to 75% for the corresponding HMA at the same production rate. This low firing rate may have resulted in extra excess air, con- tributing to NOx formation. Formaldehyde is classified as a hazardous air pollutant. It is a byproduct of the combustion of carbon-based fuels. The distribution of WMA formal- dehyde measurements was lower for WMA than for HMA and comparable to state-of-art performance observed in the mid-Atlantic states. Worker Exposure Worker exposure to asphalt fumes has typically been assessed by measuring BSF. In studies comparing worker exposure between HMA and WMA, most cases have found BSF below detectable limits. Thus, quantitative comparisons could not be made. For NCHRP Project 9-47A, Heritage Research Group utilized the newly developed TOM measure. Worker exposure was measured at two multi-technology sites. At one site, HMA temperatures behind the screed were within the expected temperature range for WMA; the WMA mixes were, on average, only 12°C cooler. At the other site, mat temperatures immediately behind the screed were, on average, 50°C cooler. With one exception, the WMA mixtures at both sites resulted in at least a 33% reduction in TOM; the one exception was an 8.4% increase at the site where the HMA was placed at WMA temperatures. The reduction for five of six mixes was statistically significant at the 95% con- fidence level. The asphalt at one site showed higher overall emissions in the temperature range typically associated with asphalt production. The sample with the highest overall TOM from each mix/site combination was tested for polycyclic aromatic hydrocarbons (PAHs). Naphthalene was detected in the highest concentra- tions. Only one non-carcinogenic 4–6 ring polycyclic aromatic compound (PAC), pyrene, was detected and it was from an HMA sample. All of the nine PACs listed by IARC for asphalt were below detectable limits. Conclusions WMA demonstrated reductions in fuel usage. These reduc- tions can help offset the cost of WMA technologies or equip- ment. Reductions in stack emissions of greenhouse gases corresponded to reductions in fuel usage. WMA should receive credit for reductions in greenhouse gases in life-cycle assess- ments. WMA also resulted in reductions in SO2 when using high-sulfur fuels such as reclaimed oil. The following revisions are proposed to the Test Frame- work for Documenting Emissions and Energy Reductions of WMA and Conventional HMA: • Corresponding WMA and HMA measurements should be made over similar time periods of steady-state production to compare fuel usage and stack emissions of WMA and HMA. • Direct fuel measurements (e.g., tank sticks, fuel meter, or gas meter readings) should be supplemented with stoi- chiometric fuel measurements in accordance with EPA Method 19. • Total organic matter (TOM) should replace benzene sol- uble fraction (BSF) for quantitative comparison of WMA and HMA worker exposure.

235 Asphalt Institute and Eurobitume. The Bitumen Industry—A Global Perspective, Production, Chemistry, Use, Specification, and Occu- pational Exposure. 2nd ed. (2011). Available electronically at: http:// www.asphaltinstitute.org/public/IS230_2ndedition.pdf. Bajpai D., and V. K. Tyagi. Biodiesel: Source, Production, Composition, Properties and Its Benefits. J Oleo Sci. 55(10), 2006, pp. 487–502. Cavallari, J. M., L. V. Osborn, J. E. Snawder, A. J. Kriech, L. D. Olsen, R. F. Herrick, and M. D. McClean. Predictors of Airborne Exposures to Polycyclic Aromatic Compounds and Total Organic Matter among Hot-Mix Asphalt Paving Workers and Influence of Work Condi- tions and Practices. Ann Occup Hyg. 56(2), 2011, pp. 138–147. Cervarich, M. Foaming the Asphalt: New Warm-Mix Technique Chal- lenges Conventional Wisdom. Hot Mix Asphalt Technology, 12(4), July/August 2007, National Asphalt Pavement Association, Lanham, MD, pp. 23–24, 42. Chief Environmental Group, LTD. Emission Test Results for: Warm Mix Asphalt Trial Project Mar-Zane Materials, Inc. Asphalt Plant #13, Byesville, Ohio. No Date. Clark, C. R., D. M. Burnett, C. M. Parker, E. W. Arp, M. S. Swanson, G. D. Minsavage, A. J. Kriech, L. V. Osborn, J. J. Freeman, R. A. Barter, P. E. Newton, and C. W. Stewart. Asphalt Fume Dermal Carcinogenicity Potential: I. Dermal Carcinogenicity Evaluation of Asphalt (Bitumen) Fume Condensates. Reg Tox Pharmacol, April 18, 2011 (e-pub). D’Angelo, J., E. Harm, J. Bartoszek, G. Baumgardner, M. Corrigan, J. Cowsert, T. Harman, M. Jamshidi, W. Jones, D. Newcomb, B. D. Prowell (report facilitator), R. Sines, and B. Yeaton, Warm-Mix Asphalt: European Practice. International Technology Scanning Program, Federal Highway Administration, December 2007. Davidson, J. K. Evotherm Trial in Aurora, Ontario, on August 8, 2005. McAsphalt Engineering Services, Toronto, Ontario. 2005a. Davidson, J. K. Evotherm Trial, Ramara Township, Road 46. McAsphalt Engineering Services, Toronto, Ontario, December 2005b. Davidson, J. K., and R. Pedlow. Reducing Paving Emissions Using Warm Mix Technology. Proc. of the 52nd Annual Conference of the Canadian Technical Asphalt Association, 2007, pp. 39–59. Diefenderfer, S. D., K. K. McGhee, B. M. Donaldson. Installation of Warm Mix Asphalt Projects in Virginia. Virginia Transportation Research Council and U.S. Department of Transportation, Federal Highway Administration. VTRC 07-R25, April 2007. EES Group Inc., Asphalt Emissions Study. Prepared for Shelly & Sands, Inc., October, 2006. ETE. Warm Mix Stack Emission Test, Environmental Technology & Engineering Corporation, Elm Grove, WI, June, 20, 2006. Fuhst, R., O. Creutzenberg, H. Ernst, T. Hansen, G. Pohlmann, A. Preiss, S. Rittinghausen. 24 Months Inhalation Carcinogenicity Study of Bitumen Fumes in Wistar (WU) Rats. J. Occup. Environ. Hyg. 4(S1), 2007, pp. 20–43. Harder, G. A. LEA Half-Warm Mix Paving Report, 2007 Projects for NYSDOT. McConnaughay Technologies, Cortland, NY, 2008. Harder, G., Y. Le Goff, A. Loustau, Y. Martineau, B. Heritier, and A. Romier. Energy and Environmental Gains of Warm and Half- Warm Asphalt Mix: Quantitative Approach. Paper #08-2456. Pre- sented at 87th Annual Meeting of the Transportation Research Board, Washington, D.C., 2008. Hurley, G., B. Prowell, and A. Kvasnak. Missouri Field Trial of Warm Mix Asphalt Technologies: Construction Summary. NCAT Report No. 10-02. National Center for Asphalt Technology, Auburn, AL, June 2010a. Hurley, G., B. Prowell, and A. Kvasnak. Wisconsin Field Trial of Warm Mix Asphalt Technologies: Construction Summary. NCAT Report No. 10-04. National Center for Asphalt Technology, Auburn, AL, November 2010b. International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Available online at: http://monographs.iarc.fr/. Accessed March 2012. Koenders, B., D. Stoker, C. Bowen, P. de Groot, O. Larsen, D. Hardy, K. Wilms. Innovative Process in Asphalt Production and Placement to Obtain Lower Operating Temperatures. Presented at 2nd Eurasphalt & Eurobitume Congress, Barcelona, Spain, 2000. Kriech A. J., J. T. Kurek, H. L. Wissel, L. V. Osborn, and G. R. Blackburn. Evaluation of Worker Exposure to Asphalt Paving Fumes Using Traditional and Nontraditional Techniques. AIHAJ 63(5), 2002a, pp. 628–635. Kriech A. J., J. T. Kurek, L. V. Osborn, H. L. Wissel and B. J. Sweeney. Determination of Polycyclic Aromatic Compounds in Asphalt and in Corresponding Leachate Water. Polycycl Aromat Cmpd. 22(3–4), 2002b, pp. 517–535. Kriech, A. J., J. E. Snawder, R. F. Herrick, L. D. Olsen, M. D. McClean, J. M. Cavallari, L. V. Osborn and G. R. Blackburn. Study Design and Methods to Investigate Inhalation and Dermal Exposure to Polycyclic Aromatic Compounds and Urinary Metabolites from Asphalt Paving Workers: Research Conducted Through Partner- ship. Polycycl Aromat Cmpd. 31(4), 2011, pp. 243–269. Kuszewski, J. R., W. B. Gorman, and E. G. Kane. Characterization of Asphalt Volatility Using TGA and Iatroscan Analyses. Proceedings of the 4th International Symposium on Roofing Technology, NRCA, Rosemont, IL, 1997. References

236 Lauby-Secretan, B., R. Baan, Y. Grosse, F. El Ghissassi, V. Bouvard, V. Benbrahim-Tallaa, et al. Bitumens and Bitumen Emissions, and Some Heterocyclic Polycyclic Aromatic Hydrocarbons. Lancet Oncol. 12(13), 2011, pp. 1190–1191. Lecomte, M., F. Deygout, A. Menetti. Emission and Occupational Exposure at Lower Asphalt Production and Laying Temperatures. WAM Environmental Benefits of Reducing Asphalt Production and Laying Temperature, Shell Bitumen, 2007. Accessed from www.static.shell.com/static/bitumen/downloads/wam_field_test_ results_italy.pdf. Accessed April 14, 2011. McClean, M. D., L. V. Osborn, J. E. Snawder, L. D. Olsen, A. J. Kriech, A. Sjödin, Z. Li, J. P. Smith, D. L. Sammons, R. F. Herrick, and J. M. Cavallari. Using Urinary Biomarkers of Polycyclic Aromatic Compound Exposure to Guide Exposure-Reduction Strategies Among Asphalt Paving Workers. Ann Occup Hyg., September 20, 2012 (e-pub). Middleton, B., and R. W. Forfylow. An Evaluation of Warm Mix Asphalt Produced With The Double Barrel Green Process. Transportation Research Record: Journal of the Transportation Research Board, Vol. 2126, Transportation Research Board of the National Acad- emies, 2009. National Asphalt Pavement Association and European Asphalt Pave- ment Association. The Asphalt Paving Industry, A Global Perspective. GL 101, 2nd ed., 2011. National Institute for Occupational Safety and Health. Benzene Solu- ble Fraction and Total Particulate (Asphalt Fume): Method 5042. In: Eller P. M., and M. E. Cassinelli, eds., NIOSH Manual of Ana- lytical Methods, 4th ed., 2nd supplement. DHHS (NIOSH) Pub- lication No. 98-119. Cincinnati, OH: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, 1998. National Institute for Occupational Safety and Health: Hazard Review: Health Effects of Occupational Exposure to Asphalt. DHHS (NIOSH) Publication No. 2001-110. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, 2001. Newcomb, D. An Introduction to Warm-Mix Asphalt. National Asphalt Pavement Association. Accessed from http://fs1.hotmix.org/mbc/ Introduction_to_Warm-mix_Asphalt.pdf. Accessed August 8, 2007. Norseth, T., J. Waage, and I. Dale. Acute Effects and Exposure to Organic Compounds in Road Maintenance Workers Exposed to Asphalt. Amer J Ind Med. 20, 1991, pp. 737–744. Olsson, A., H. Kromhout, M. Agostini, J. Hansen, C. F. Lassen, C. Johansen, K. Kjaerheim, S. Langård, I. Stücker, W. Ahrens, T. Behrens, M. L. Lindbohm, P. Heikkilä, D. Heederik, L. Portengen, J. Shaham, G. Ferro, F. de Vocht, I. Burstyn, and P. Boffetta. A Case- Control Study of Lung Cancer Nested in a Cohort of European Asphalt Workers. Environ. Health Persp. 118, 2010, pp. 1418–1424. Powers, D. Warm Mix Asphalt 2008 ODOT Field Trials. Presented at 2009 Ohio Asphalt Paving Conference, February 4, 2009. Prowell, B. D., G. C. Hurley, and B. Frank, Warm Mix Asphalt: Best Prac- tices, 3rd ed. (2012). Quality Improvement Series 125, National Asphalt Pavement Association, Lanham, MD, 2012. Raulf-Heimsoth, M., B. Pesch, B. Kendzia, A. Spickenheuer, R. Bramer, B. Marczynski, R. Merget, and T. Brüning. Irritative Effects of Vapours and Aerosols of Bitumen on the Airways Assessed by Non- Invasive Methods. Arch Toxicol. 1, 2011, pp. 41–52. RTI International. Emission Factor Documentation for AO-42 Section 11.1 Hot Mix Asphalt Plants. Final Report. U.S. Environmental Protection Agency, February 2004. Shell Bitumen. Asphalt Production and Laying Temperature. Accessed from www.static.shell.com/static/bitumen-en/ . . . /wrc/ . . . / wam_english.pdf. Shifa, X., X. Liting, J. Jonathan, X. Yongquing. The Application of Emul- sion Warm Mix Asphalt in Long Tunnel Pavement. Geotechnical Special Publ., No. 193, 2009. Tepper A. L., G. A. Burr, H. A. Feng, M. Singal, A. K. Miller, K. W. Hanley, and L. D. Olsen. Acute Symptoms Associated with Asphalt Fume Exposure Among Road Pavers. Am. J Ind Med. 49, 2006, pp. 728–739. Ventura, A., P. Moneron, A. Jullien, P. Tamagny, F. Olard, and D. Zavan. Environmental Comparison at Industrial Scale of Hot and Half- Warm Mix Asphalt Manufacturing Processes. Paper #09-1274, presented at 88th Annual Meeting of the Transportation Research Board, Washington, D.C., 2009. Von Devivere, M., W. Barthel, and J-P. Marchand. Warm Asphalt Mixes by Adding Aspha-min, a Synthetic Zeolite. XXIInd PIARC World Road Congress, Durban, South Africa, 2003. WMA TWG. “Documenting Emissions and Energy Reductions of WMA and Conventional HMA.” August 2006. Accessed from www. warmmixasphalt.com. Accessed November 6, 2013.

237 A P P E N D I X Documenting Emissions and Energy Reductions of WMA and Conventional HMA During Plant and Paving Operations

238 PROPOSED REVISION Project Summary Attached is a project summary data entry sheet for use in identifying warm mix asphalt (WMA) technologies and binder characteristics, aggregate, and plant type. As indicated on the project summary data entry sheet, recorded information will include: WMA technology and binder characteristics, Reclaimed asphalt pavement (RAP) usage/rate (if applicable), Recycled asphalt shingle (RAS) usage/rate (if applicable), Aggregate type(s), Aggregate moisture content, Anti-stripping agent or other additives, and Plant type. Plant Emissions Stack Testing and Energy Requirements Introduction: Stack testing should include mass emissions rate measurement of NOx, CO2, and volatile organic compounds (VOC) to compare stack emissions from WMA technologies and conventional hot mix asphalt (HMA). It is suggested that stack emissions reporting be standardized as lbs. per ton (of mix produced) and include a recording and reporting of average production rate in tons HMA or WMA produced per hour, during each test period. Testing should be performed by a certified tester and should include either two (2) or three (3) 60-minute stack sampling runs per technology, if possible. The number of runs may have to be adjusted to the available run time using the WMA technology. Production rates should be recorded every 15 minutes during each test run and used to determine average production rate in tons mix produced per hour for each run. The data from all individual test runs during a test period (conventional HMA or WMA) should be averaged to determine the overall results for each technology. Stack gas volumetric flow rate based on full traverse of stack cross section during hour run, moisture content, temperature, and a variety of other parameters should also be determined for each run, in accordance with United States Environmental Protection Agency (U.S. EPA) stack testing methodology. In order to assess fossil fuel and energy use reductions, it is suggested that beginning and end fuel usage data be recorded for each test run. This may be accomplished with direct fuel usage meter readout, where available, or by tank gauging as appropriate. To validate accuracy of direct fuel measurements, stoichiometric fuel usage calculations should be made from stack gas flow rate in accordance with U.S. EPA Method 19.

239 Stack Emissions Testing and Analytical Methods: Suggested test methods are in accordance with U.S. EPA protocol used historically in the HMA industry and are as follows: Sampling point locations per U.S. EPA Method 1, if ports have not been established during previous stack testing. If ports have been previously established, the test firm should confirm that their location is consistent with that specified by U.S. EPA Method 1. Access platforms and an appropriate power source must also be available during testing. The remaining stack emission test parameters and methods are defined in Table 2.A.1. Table 2.A.1. Stack emission test parameters and methods. Emission Parameter Minimum Number of Test Runs per Technology Sampling and Analytical Methodology Volumetric flow rate * U.S. EPA Methods 1 and 2 Oxygen (O2) and carbon dioxide (CO2) * U.S. EPA Method 3A Moisture content * U.S. EPA Method 4 Sulfur Dioxide (SO2) 2 U.S. EPA Method 6 Nitrogen oxides (NOx) 2 U.S. EPA Method 7E Carbon monoxide (CO) 2 U.S. EPA Method 10 Total hydrocarbons (VOC) — reported as molecular weight of propane 2 U.S. EPA Method 25A Particulate matter/PM-10 2 U.S. EPA Methods 5/202 Formaldehyde 2 U.S. EPA Method 316 * Determined concurrently with all emission parameters Energy Requirements and Operational Data: Attached is an operational data entry sheet for use in determining the average production rate, average mix temperature, for calculating the amount of energy required to produce the mix, for documenting burner settings, and monitoring baghouse temperature and pressure. As indicated on the operational data entry sheet, recorded information will include: Production rate recorded in 15-minute intervals. Any plant starts/stops should be noted. Mix discharge temperature. Fuel meter readings or tank dips at the beginning and end of steady-state production runs. Tank dips should be measured to the nearest 0.1 inch. Many gas meters only update periodically—up to 30 minutes between changes. Someone could monitor the meter and call the tower for cumulative production tonnage the instant the meter updates. Time lags between updates or recording tonnage result in errors. Slat conveyor voltage should be recorded in addition to amperage in order to estimate power used.

240 Suggested Reporting of Stack Emissions and Energy Results: Average mix production rate in tons/hour - Conventional mix test period - WMA test period Pounds of each pollutant per ton of mix produced - Carbon dioxide, nitrogen oxides, total hydrocarbons, carbon monoxide, sulfur oxides, fine particulates (PM-10), and formaldehyde - Conventional test period (average all runs) - WMA test period (average all runs) Fossil fuel usage—Gallons or cubic feet gas/ton mix - Type of fuel used (i.e., #2 oil, natural gas, other) - Conventional test period (average all runs) - WMA test period (average all runs) - Percent reduction corrected for differences in aggregate moisture content Include appendix for field test data and calculations summary Approximate Costs Associated with Stack Emissions Testing: Any travel costs, outside locality, are not included. Complex reporting of results will incur extra charges; this is not anticipated. Costs for developing test plans (test protocol) are not included; however, test plans are not anticipated to be needed. There are minimal differences in costs (+/- $300) associated with conducting either two or three stack tests. Baseline costs are anticipated to be approximately $3,000–$5,000 per day. - Multiple technologies (comparison with conventional HMA is an additional technology) typically require a day per technology. Includes three stack tests. - Includes simple reporting of results. - Formaldehyde, Sox, and PM-10 analysis add a small additional cost. - Costs are for a local company to conduct the emissions testing—travel costs would be incurred for non-local companies. Emissions Surrounding Laydown Operations Introduction: Ideally, placement of each mix, conventional and WMA, would use the same paving equipment; material placed oneday apart, approximately during the same time-frame. To minimize variability, it is also recommended that the paving machines utilized are equipped with properly functioning engineering controls. The recommended test period, for field emissions, is between 3 and 4 hours. More detail follows. Placement of Monitors: During the placement of each technology, conventional HMA and WMA, paving crew members will be monitored for asphalt fume emissions. The purpose of this testing is to document, with some statistical power, the reduction in field application emissions using WMA as compared with using conventional HMA. Monitoring four workers is recommended. The four workers with the greatest potential for asphalt emission exposure are: paver operator, screed operators, and raker. If diesel oil

241 is normally used as a release agent, a substitute such as B-100 (biodiesel) (CAS Number: 67784-80-9) should be used when monitoring laydown emissions. Sampling and Analytical Method: Traditional gravimetric procedures used to quantify asphalt fume emissions such as National Institute for Occupational Safety and Health (NIOSH) Method 5042 measurements of total particulates (TP) and benzene soluble fraction (BSF) generally prevent quantitative comparisons between HMA and WMA since most readings are below detectable limits. An alternate procedure to measure total organic matter (TOM) developed by Heritage Research Group, in conjunction with NIOSH Method 5042, is recommended. Each worker to be monitored can be equipped with two samplers: the NIOSH 5042 sampler if required, and a sorbent tube containing XAD-2 and charcoal (150 mg XAD-2 followed by 50 mg activated charcoal). A 1-inch piece of Tygon® tubing (dichloromethane rinsed) is added to the end of the sorbent tube, once broken, to protect the workers. Care should be taken to break the inlet end of the tube to 4-mm to equal the NIOSH sampler. Set to a flow rate of 2.0 + 0.2 L/min., pumps should be calibrated pre-shift and re-measured post-shift. One background sample should be collected each day/experiment, positioned upwind of the paving operation. Sorbent tubes were eluted with 5 mL dichloromethane; charcoal end up. A field blank should be collected on each day/experiment for each crew. If NIOSH 5042 is performed, this method requires five field blanks per day. Descriptive data should be collected on potential confounders from the site, e.g., construction dust and any other background interferences. One background sample per day, upwind of the paving operation, is highly recommended. Keep completed samples dry and cold by placing them in a cooler with ice packs and protect them from light by wrapping them with foil. This allows further chemical-specific analysis, if warranted. Minimum field sampling collection times should be between 3 and 4 hours; 6 to 8 hours would be the preferred sampling time using one single media cartridge. TOM (Kriech et al. 2002) included hydrocarbons ranging from C6 to C42 as determined by gas chromatography/flame ionization detection (GC/FID). A Varian model 3400 GC with a 1077 split/splitless injector (set at 250°C) was used, with a 5% phenyl /95% methyl-polysiloxane column (30 m x 0.33 mm ID, 0.25 µm film thickness; Restek RTX-5); hydrogen carrier gas was set at 2 mL/min. With detector at 310oC, the oven temperature program was 40°C held for 3 minutes, increased to 120oC at 9oC/min, held for 0.5 min, then ramped to 305oC at 11oC/min, and held for 10.89 min. Calibration should include kerosene standards for quantification of the TOM. The sample can also be tested for individual polycyclic aromatic compounds (PACs) and/or 4-6 ring PACs by Fluorescence spectroscopy (Osborn et al. 2001). A complete list of field equipment for monitoring lay down temperatures, collection of worker exposure samples, TOM testing, and PAC, testing (if desired), is shown in Table 2.A.2. Any equivalent or better instrumentation or supplies can be used; details are provided for convenience.

242 While sampling in the field, mix temperatures (both in the hopper and on the mat as it exits the screed strike area) should bemonitored and recorded approximately every 30 minutes, during the test period, with a dial stem thermometer; provided it can be taken safely. It is essential that weather-related information be collected and documented at least four times during the sampling period. Information would include, at minimum: wind speed and direction, air temperature, humidity, and other weather-related comments. For any personal sampling, names of all workers will be recorded along with observations during sampling including smoking habits. Workers may be asked not to smoke; if they do smoke, smoking should be documented. Pumps may be turned off while smoking. Document pertinent information regarding work positions and activities. Photographs, illustrating field application of these technologies, will be taken throughout the sampling event. Diagrams noting the area sample locations and locations of workers are also helpful. Noting the direction of the paving application is important, especially in relation to wind direction. Suggested Reporting of Results: Anomalies in sampling and results Visual observations of emissions - Conventional test period - WMA test period Mix temperature (hopper and mat) - Conventional test period - WMA test period - Percent reduction Supplier Description CAT. NO. City State HMA Lab Supply, Inc. Stainless Steel Dial Stem thermometer, with a - 18 to 204 °C range TM-4500 Richmond VA SKC, Inc. 150 mg XAD-2 followed by 50 mg activated charcoal CPM032509-001 Eighty Four PA Fisher Scientific Tygon® tubing 14-176-272 Pittsburg PA EMD Dichloromethane HPLC Grade OmniSolv® High Purity DX0831-1 Gibbstown NJ AccuStandard, Inc. Kerosene standards FU-005N Neat New Haven CT AccuStandard, Inc. Custom mix of 24 PACs H-QME-01 New Haven CT AccuStandard, Inc. Custom mix of 9 PACs S-13911-R1 New Haven CT AccuStandard, Inc. dibenzo[a,e]fluoranthene Cat. No. H-247S New Haven CT AccuStandard, Inc. thianaphthene Cat. No. H-238N New Haven CT Sigma-Aldrich dibenz[c,h]acridine BCR 156R St. Louis MO Sigma-Aldrich benz[a]acridine R308714 St. Louis MO Sigma-Aldrich dibenz[c,h]acridine BCR 156R St. Louis MO Supelco Internal standard mix 4 8902 St. Louis MO Table 2.A.2. Supply information with catalog numbers.

243 Weather data including ambient temperature and humidity Worker activities Diagrams and/or photographs documenting activities and sampling locations Notation whether paver is equipped with functioning engineering (emission reduction) controls Background-corrected asphalt fume emissions (TP, BSF, and TOM) reported in mg/m3 - Conventional test period (average all runs) - WMA test period (average all runs) - Percent reduction Approximate Costs Associated with Occupational Hygiene Field Emissions Testing: Any travel costs, outside locality, are not included. Analytical costs are approximately $100 per sample (11 samples per 3–4 hour event) x 2 events per day. Labor at approximately $110 per hour (10 hours) x 2 people. Report writing and miscellaneous at approximately $600. Baseline costs are anticipated to be approximately $5,000–$6,000. - Per technology (comparison with conventional HMA is an additional technology—i.e., a complete round of testing would be needed). - Costs are for local hygienists to conduct the field monitoring—travel costs would be incurred for non-local hygienists. - Monitoring equipment (pumps) may or may not be included in the labor rates but should not substantially affect the estimated baseline costs.

244 Data Collection Forms General Plant Information Project Identification: _______________________ Date: __________ Contractor: ______________________________________ Plant Location: __________________ GPS Coordinates: ______________ Plant Type: _______________________________________ (batch, counter-flow drum, parallel flow drum, or etc.) Plant Manufacturer: _______________________________________________ Burner Model/Type: _______________________________________________ Fuel Type: __________________________ Fuel Temperature (if oil): ______ Describe any modifications for producing WMA: ______________________ ________________________________________________________________ Binder, Aggregate, Additive Information Binder Grade: _____________ Supplier: _____________ If modified, type of modification (e.g., polymer modified SBS): ____________ Anti-stripping Additive: ________________________ Dosage: _________ Warm Mix Asphalt Technology: _____________________________________ Aggregate Type(s): ________________________________________________ (e.g., limestone, granite, or etc.) RAP/RAS Usage/Rate: _____________________________________________ Aggregate Moisture Content Date/Time Composite Moisture Content (%) Truck Release Agent: ______________________________________________ Notes:

PROCESS DATA SHEET HMA or WMA Technology: ___________________________ Date: ___________________ Page: ____ Time Production Rate Burner % Mix Temp. (°____) Aggregate Temp. (°____) Stack Temp. (°____) % Damper Baghouse Delta Pressure Tons Produced Drag Amperage Comment Fuel Reading Start of Run/Units/Time: _________________ Fuel Reading End of Run/Units/Time: _________________

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