Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 31
31 Based on the findings of the selection study, the SAFT was · Condition 4--SAFT without the lid or prototype VCS, selected for further development in NCHRP Project 9-36. airflow rate 4,000 mL/min. The SAFT was selected because the selection study found that with additional development it may be possible to extend this The mass of the volatiles was determined for Condition 2 by test to long-term aging. The MGRF, on the other hand, could flushing the volatiles from the VCS with solvent and evapo- not be extended to long-term aging because of inadequate rating the solvent. In Conditions 2, 3, and 4 mass change was mixing of air and binder in this test at the lower aging tem- determined by weighing the vessel and its components before peratures that should be used to simulate in-service conditions. and after conditioning. Table 3-11 summarizes the results of the testing. 3.4 Volatile Collection This experiment provided important insight into the System Study behavior of the prototype VCS. First, the SAFT without the lid and prototype VCS produced mass losses that are less than The objective of the volatile collection system (VCS) study was to design and evaluate an improved VCS for the SAFT. the RTFOT, but a factor of 10 higher than the mass of The prototype SAFT included a VCS, which consisted of a the volatiles collected in the prototype VCS. The mass of the copper coil condenser operated at ambient temperature. As volatiles collected in this experiment by the prototype VCS is shown previously in Figure 3-4, the mass of volatiles collected of the same size as those reported by the original developers using this system was a factor of 10 lower than the mass change of the SAFT, which ranged from 0.013 to 0.051 percent by in the RTFOT and showed little difference between binders. weight (1). Thus, it appeared that only a small percentage of The VCS study, which is described in detail in Appendix C the volatiles that are produced by the SAFT was being col- (see the project webpage on the TRB website), included an lected by the VCS. At a flow rate of 2,000 mL/min, the mass evaluation of the air-cooled condenser used in the prototype of the volatiles trapped in the collection system was only SAFT, the design of an improved VCS employing reusable 0.013 percent by mass compared to a mass loss of 0.11 per- adsorbents that are commonly used for chromatographic cent when the lid and collection system were removed. Based analyses, and evaluation of the improved VCS. This section on this finding, the design of an improved VCS that would presents key findings of the VCS study. capture a greater amount of volatiles was initiated. 3.4.1 Evaluation of Prototype SAFT VCS 3.4.2 Improved VCS Design A mass change experiment was conducted with the SAFT to identify the cause of the unexpectedly low volatile mass The design of an improved VCS was an incremental process. collected in the prototype VCS. The neat PG 58-28 binder In the first iteration, called VCS-I, silica gel and activated car- that was included in the selection study was aged under the bon filters were added after the condenser from the prototype following conditions: VCS to collect moisture and hydrocarbon material passing through the condenser. Silica gel and activated carbon filters · Condition 1--RTFOT using standard conditions; also were added before the SAFT vessel to remove moisture · Condition 2--SAFT with the prototype VCS in place, and hydrocarbon material from the incoming air. Figure 3-19 airflow rate 2,000 mL/min; is a schematic of VCS-I. This version was used in a study to · Condition 3--SAFT without the lid or prototype VCS, investigate the effect of pressure inside the SAFT on the airflow rate 2,000 mL/min; and amount of volatiles produced. All testing was performed with Table 3-11. Summary of SAFT mass change measurements. Average Average Mass of Mass Mass of Test and Test Airflow, Mass Collected Change, Collected Parameters mL/Min Change, Volatiles, wt %(a) Volatiles, wt % wt % wt % -0.357 RTFOT 4,000 -0.35 NA NA -0.349 SAFT with lid and 0.016 2,000 NA NA 0.014 collection system 0.011 SAFT without lid and -0.108 2,000 -0.11 NA NA collection system -0.112 SAFT without lid and -0.184 4,000 -0.18 NA NA collection system -0.176 Note: (a) Negative values in this table indicate a loss in mass.
OCR for page 32
32 Hydrocarbon Trap Silica Gel Moisture Trap Koby Junior Agilent Technologies MT 120-2-S Hydrocarbon Trap Koby Junior Silica Gel Moisture Trap Cole Parmer C-02908-6 2 SAFT VESSEL Collector 10 ft 1/4 in Copper Tubing BINDER Figure 3-19. Schematic of VCS-I. a PG 58-28 binder with an average RTFOT mass change of Approximately 73 percent of the total is collected in the hydro- 0.343 percent. Triplicate 250-g samples were aged for 45 min carbon trap, 17 percent in the moisture trap, and 10 percent at 163°C using an airflow rate of 4,000 mL/min. Three differ- in the condenser. The amount collected in the hydrocarbon ent conditions were used to produce the flow: slight positive trap exceeds the RTFOT mass change for this binder, which pressure, vacuum with 90-kPa absolute pressure, and vacuum is reasonable considering the mass change measurement with 70-kPa absolute pressure. The subatmospheric pressures includes mass loss due to volatilization and mass gain due to (70 kPa and 90 kPa absolute) were produced by applying a oxidation. vacuum downstream, essentially sucking the air through the VCS-I also was used to collect volatiles from successive SAFT vessel. runs of the same binder. In this study, dedicated condensers, Figure 3-20 presents the results of this testing. As shown, a silica gel filters, and activated hydrocarbon filters were used, significant mass is collected in each component of the VCS. and the components of the VCS were not cleaned or purged 0.60 0.50 Quantity Collected, wt % 0.40 RTFOT Mass Loss Condenser 0.30 Hydrocarbon Trap Moisture Trap SAFT Mass Loss 0.20 0.10 0.00 70 90 101 Absolute Pressure, kPa Figure 3-20. Mass of material collected in the improved VCS, VCS-I.
OCR for page 33
33 between successive runs. This allowed the volatiles from mul- Process Step Flow Path tiple runs to be combined to collect sufficient materials for 1 2 3 4 5 1. Purge filter system O C C C C analysis. The mass collected in each component for each run 2. Preheat binder C C O C O is shown in Figure 3-21. The mass collected in the silica gel 3. Age binder C C C O O 4. Flush H2O at end of run O C C C C decreases significantly from the first to the third run. Based on this observation and the observation that the silica gel 1 acquired a brownish color, it appeared that the silica gel 250°C adsorbs hydrocarbons as well as water. Further, the results N2 2 100°C 60-80 show that the efficiency of the silica gel decreases with succes- Tenax TA sive runs, most likely caused by interference from absorbed 3 5 polar compounds. These observations prompted the need for 4 60-80 Air a revised VCS design. Hayesep Q The second iteration of the improved VCS, called VCS-II, was designed to use filters that are commonly used in chro- SAFT 60-80 matographic analyses. The resulting system is shown in Fig- Mol Sieve 5A ure 3-22. Although not shown in Figure 3-22, the inlet silica gel and activated carbon filters from the VCS-I were retained Hydrocarbon to condition the incoming nitrogen and air. Trap* Koby Junior VCS-II used two absorbent polymer resin filters, Tenax TA and HayeSep Q to remove, respectively, the larger molecular *Hydrocarbon trap may be redundant for this system. weight polar materials (aromatics) and the remaining hydro- carbons, while the molecular sieve was selected to remove Note: Flow diagram for descriptive purposes only and not to be used to lay out system. Air is pre-dried with a commercial drier water. The activated carbon filter was included only to ensure and further cleaned by passing the air through a silica gel and the efficiency of the resin beads and molecular sieve. Nothing charcoal filter before it is introduced into the SAFT. should be collected in the activated carbon filter if the system Figure 3-22. VCS based on chromatographic filters, is properly designed. The system shown in Figure 3-22 was not VCS-II. intended for routine use but instead was selected to provide an understanding of the nature of the volatiles that are being released and to form a basis for selecting a simplified system. nitrogen through them. Estimated life of the absorbents, as Each of the absorbents was contained in a 7-in. (175-mm) predicted by the manufacturer's technical support staff, is in long, 0.5-in (12.5-mm) diameter stainless steel tube and were excess of 50 runs. connected with rubber tubing. A photograph of an assembled Data on the magnitude and repeatability of the mass changes filter is shown in Figure 3-23. The filter beds can be reused by for each of the collectors were obtained for duplicate runs of purging them of collected volatiles by passing high-temperature binders: Citgo PG 58-28, ABM-2, AAM-1, and AAD-2. The Collector, wt % Water Filter, wt % Hydrocarbon Filter, wt % Total, wt % 0.800 0.700 0.600 Weight Gain, % 0.500 0.400 0.300 0.200 0.100 0.000 1 2 3 Run Number Figure 3-21. Average amount of volatiles captured in successive runs, VCS-I.
OCR for page 34
34 mass of the volatiles collected during replicate runs for the four filter indicated that some organic material was collected on binders is shown in Table 3-12 for the SAFT with the VCS-II the molecular sieve. This indicated that the length of the and the RTFOT. Surprisingly, the majority of the volatiles col- HayeSep Q was too short. As part of the VCS-III design study, lected in the VCS-II were collected in the molecular sieve and an experiment was conducted to establish the length of not on the Tenax TA and HayeSep Q filter beds. For example, HayeSep Q collector required to minimize breakthrough. for the 58-28 binder, 80 percent of the volatiles were collected The length of the HayeSep Q collector was varied from 1.5 in. on the molecular sieve while only 20 percent were collected on (37 mm) to 4.5 in. (111) mm by adding one or two 1.5-in. the Tenax TA and HayeSep Q, with similar results for the other (37-mm) lengths to the original 1.5-in. (37-mm) length and three binders. When the Tenax TA was challenged (purged) at the mass collected in each length was then weighed. This the end of the run, considerable smoking was observed as the experiment found very little difference in the total amount volatiles were released. The odor was acrid in nature, unlike the captured when the HayeSep Q bed varies from 1.5 in. (37 mm) smell of asphalt binder. When the HayeSep Q was purged, no to 4.5 in. (111 mm). A conservative HayeSep Q filter bed smoke was observed but an asphalt-like odor was observed. The length of 3.9 in. (100 mm) was selected for the VCS-III based volatiles driven from the molecular sieve have a hydrocarbon on these results. odor somewhat similar to asphalt cement. Internal discussions as well as input from others outside Material collected during the last run for binder AAD-2 the study suggested that the Tenax TA filter could be removed was sent to Heritage Research Group to determine the com- from the system. Removing the Tenax TA filter bed would position of the volatiles collected in each filter. The procedure simplify and reduce the cost of the VCS. Replicate SAFT runs used by Heritage was to first sequentially elute each of the were conducted with various lengths of HayeSep Q with and three tubes with hexane, methylene chloride, and methanol. without the Tenax TA. This experiment found very little dif- Each eluent was analyzed by gas chromatography with flame ference in the weights captured on the HayeSep Q plus Tenax ionization detection (GC/FID) for quantification of the TA filters and on the HayeSep Q filter only. organic chromatographic material. Use of the three different The final configuration for the VCS-III is shown schemat- solvents allowed the removal of compounds of varying polari- ically in Figure 3-24. It consists of silica gel and activated ties. Water present in each of the eluents was determined by carbon filters in the inlet gas stream. The outlet stream passes Karl Fisher testing using a Mettler Toledo DL-38 unit. This through a (3.9-in.) 100-mm long HayeSep Q collector to col- analysis showed that essentially all of the water was collected in lect hydrocarbons and a 5-angstrom molecular sieve to collect the molecular sieve. The Tenax TA filter collected 91.2 percent water. The HayeSep Q collector and the molecular sieve are of the organic material, the HayeSep Q filter collected 8.2 per- challenged prior to testing by passing nitrogen gas at 2 L/min cent, and the molecular sieve collected the remaining 0.6 per- and 250°C inlet temperature for 15 minutes. cent. The mass of organic material collected on each of the filters varied somewhat from the mass changes presented in Table 3-12; however, given the tendency to lose highly volatile 3.4.3 Evaluation of Improved VCS components during handling and the nature of the GC mea- surements, the results were considered compatible. The chro- The commercial SAFT fitted with the VCS-III was used to matograms were, as expected, showing compounds similar to condition the 12 binders in the validation study. Three repli- those obtained previously in asphalt fume studies conducted cate samples of each binder were conditioned in the commer- by Heritage Research Group. cial SAFT and the RTFOT. The amount of material collected The results of the mass change and chemical analyses indi- on the HayeSep Q and the molecular sieve for each of the cated that further refinements of the VCS were warranted. asphalt binders is shown in Table 3-13 and Figure 3-25. Also The compositional analysis of the material collected in each shown in Table 3-13 is the negative value of the RTFOT mass Table 3-12. Mass change for RTFOT and SAFT with VCS-II. Aging Measured Value Filter Media CITGO ABM-2 AAM-1 AAD-2 Device 58-28 SAFT Mass Change, Percent of Tenax TA 0.025 0.020 0.007 0.034 Initial Binder Mass HayeSep 0.026 0.012 0.008 0.011 Mol. Sieve 0.204 0.179 0.128 0.166 Tenax + HayeSep 0.051 0.032 0.015 0.045 Total 0.255 0.211 0.143 0.211 Mass Change, Percent of Total Tenax + HayeSep 20 15 10 21 Volatiles Collected Mol. Sieve 80 85 90 79 RTFOT Mass Change None 0.345 0.348 -0.122 1.058