oEPA 4 United States Environmental Protection Aaencv Investigation of Gaseous Criteria Pollutant Transport Efficiency as a Function of Tubing Material Office of Research and Development Center for Environmental Measurements and Modeling ------- EP A/600/R-22/166 August 2022 Investigation of Gaseous Criteria Pollutant Transport Efficiency as a Function of Tubing Material by Kyle Digby1, Carlton Witherspoon2, Robert Yaga2, and Diya Yang2 ^SAI 2Jacobs Technology, Inc US EPA Contract: 68HERC20D0018; EP-C-15-008 Task Order: R-20F0311 Cortina Johnson, Andrew Whitehill, Russell Long, and Robert Vanderpool U.S. Environmental Protection Agency Office of Research and Development Center for Environmental Measurements and Modeling Research Triangle Park, North Carolina 27711 ------- Disclaimer The U.S. Environmental Protection Agency (EPA), through its Office of Research and Development (ORD) funded the research described herein under Contract: 68HERC20D0018; EP-C-15-008, Task Order: R-20F0311. The views expressed in this report are those of the authors and do not necessarily reflect the views or policies of EPA. This document has been subjected to Agency review and approved for publication. Mention of trade names or commercial products do not constitute an endorsement or recommendation for use. The contractor role did not include establishing Agency policy. Questions concerning this document, or its application, should be addressed to: Robert Vanderpool Ambient Air Branch Air Methods and Characterization Division Office of Research and Development U.S. Environmental Protection Agency Mail Code D475 109 T.W. Alexander Drive Research Triangle Park, NC 27711 Telephone No.: (919) 541-7877 E-mail Address: Vanderpool.Robert@epa.gov i ------- Table of Contents Disclaimer i Table of Contents ii List of Tables iii List of Figures iii List of Equations iv Acronyms and Abbreviations v Acknowledgments vii Executive Summary viii 1.0 Introduction 1 1.1 Background 1 1.2 Obj ective 2 2.0 Methods 3 2.1 Experimental Approach 3 2.1.1 Pilot Testing 3 2.1.2 Target Analytes and Test Materials 8 2.2 Experimental Setup 10 2.2.1 Experiment Equipment 10 2.2.2 Initial Sample Gas Flow Path 13 2.2.3 Sample Gas Flow Along the Rest of the Flow Path 14 2.2.4 Data Flow Path 15 2.3 Final Experimental Procedure 18 3.0 Quality Assurance (QA) and Quality Control (QC) 21 3.1 Equipment Calibration and Verification 21 3.2 QA/QC Checks 21 3.3 Data Quality Objectives 22 4.0 Results and Discussion 23 4.1 O3 Data Summary 24 4.2 SO2 Data Summary 26 4.3 NO2 Data Summary 28 4.4 CO Data Summary 30 4.5 Stati sti cal Analy si s 31 5.0 Summary and Conclusions 36 6.0 References 40 Appendix A 42 ii ------- List of Tables Table 1. Pilot Tests Experiment Objectives, Descriptions, Results, and Determinations 4 Table 2. Test Materials List 9 Table 3. Equipment List 11 Table 4. Gas Cylinder List 12 Table 5. Setpoint Concentrations 18 Table 6. DQOs for Critical Measurements 22 Table 7. O3 Transport Efficiency Results 24 Table 8. SO2 Transport Efficiency Results 26 Table 9. NO2 Transport Efficiency Results 28 Table 10. CO Transport Efficiency Results 30 Table 11. Summary of Statistical Testing 31 Table 12. Data reorganization Transport Efficiency vs NAAQS Level 32 Table 13. Results Transportation Efficiencies of Tubing vs NAAQS Level testing 32 Table 14. Data reorganization tubing type vs pollutant 33 Table 15. Results from ANOVA single factor test of tubing type vs pollutant 33 Table 16. Data reorganized for ANOVA two-way testing 34 Table 17. P-values of the ANOVA two-way testing (unconditioned transport efficiencies) 34 Table 18. Results from the ANOVA two-way testing with conditioned O3 transport efficiencies 35 Table 19. Summary of Transport Efficiencies for the Four Fluoropolymer Tubing Materials 37 Table 20. Properties of Fluoropolymer Tubing Materials 38 Appendix A Table 1. Ozone (Unconditioned) 42 Appendix A Table 2. Ozone (Conditioned) 43 Appendix A Table 3. Sulfur Dioxide (Unconditioned) 44 Appendix A Table 4. Sulfur Dioxide (Conditioned) 45 Appendix A Table 5. Nitrogen Dioxide (Unconditioned) 46 Appendix A Table 6. Nitrogen Dioxide (Conditioned) 47 Appendix A Table 7. Carbon Monoxide (Unconditioned) 48 List of Figures Figure 1 Example Graph of FEP vs Grade 316 Stainless Steel Pilot Test 6 Figure 2. Example Graph of Stainless Steel with SilcoNert® Coating vs PTFE Pilot Test 7 Figure 3. Side by Side Comparison of O3 Measurements through SS316 and SNSS Before and After Solvent Cleaning 8 Figure 4. Diagram of Flow Path Experimental Setup 13 Figure 5. Flow Chart of the Valve Switch Timing Process During Each Experiment 14 Figure 6. Diagram of the Data Flow of the System 15 Figure 7. Example Graph of an Experiment from the Envidas Reporter Application 16 Figure 8. Photograph of Flow Path Experimental Setup 17 iii ------- List of Equations Equation 1. Transportation Efficiency for a Target Analyte through Test Material Equation 2. Tubing Length Based on Residence Time of Gas through Test Material iv ------- Acronyms and Abbreviations CAA Clean Air Act CAPS Cavity Attenuated Phase Shift CASTNET Clean Air Status and Trends Network CFR Code of Federal Regulations CO Carbon Monoxide DAS Data Acquisition System EPA Environmental Protection Agency FEM Federal Equivalent Method FEP Fluorinated Ethylene Propylene FRM Federal Reference Method GFC Gas Filter Correlation GPT Gas Phase Titration hv Photon ID Inner Diameter IR Infrared LED Light Emitting Diode MFC Mass Flow Controller NAAQS National Ambient Air Quality Standards NO Nitric Oxide no2 Nitrogen Dioxide 03 Ozone OD Outer Diameter ORD Office of Research and Development ppb Parts Per Billion (by volume) ppm Parts Per Million (by volume) PFA Perfluoroalkoxy Alkane PM Particulate Matter PMT Photomultiplier Tube PTFE Poly tetrafluoroethyl ene PVC Polyvinyl Chloride PVDF Polyvinylidene Fluoride ------- QA Quality Assurance RS Recommended Standard S02 Sulfur Dioxide SNSS Grade 316 Stainless Steel tubing with Silconert® 2000 internally coated SS316 Grade 316 Stainless Steel SQL Structured Query Language TAPI Teledyne Air Pollution Instruments TCP/IP Transmission Control Protocol/Internet Protocol UV Ultraviolet vi ------- Acknowledgments The study team is grateful for the administrative and quality assurance (QA) support of Libby Nessley, QA Manager, Air Methods and Characterization Division, U.S. EPA. vii ------- Executive Summary Accurate determination of National Ambient Air Quality Standards (NAAQS) pollutant concentrations requires efficient sampling and transport of the pollutants to their respective samplers for quantitative analysis. For reactive gas sampling, Section 9 of Appendix E of 40 CFR Part 58 specifies that only borosilicate glass (Pyrex®) or fluorinated ethylene propylene (FEP Teflon®) "or their equivalent" are suitable materials for sampling and transporting to Federal Reference Method (FRM) and Federal Equivalent Method (FEM) instruments. It also specifies that the sample residence time between the inlet probe and the analyzer cannot exceed 20 seconds. These regulatory specifications have not been updated since 1979. The purpose of this research effort was to identify alternative transport materials for sampling gaseous criteria pollutants. This laboratory study determined the transport efficiency of Ozone (O3), Nitrogen Dioxide (NO2), Sulfur Dioxide (SO2), and Carbon Monoxide (CO) through six candidate materials at multiple pollutant concentration levels. The need to condition or passivate the tubing with the pollutant of interest was also tested. Based on a review of the literature, six tubing materials were selected for evaluation. These included three fluorinated polymer materials (PTFE, PVDF, and PFA), a non-fluorinated polymer (PVC), and two metals (316 stainless steel (SS316) and Silconert-2000® coated 316 stainless steel (SNSS)). All tubing had a 1/4" OD and nominal 3/16" ID, and lengths of tubing were selected to provide a 20 second residence time at the tested FRM's or FEM's design flow rate. For each pollutant and tubing type, three replicate tests were conducted at concentrations of 20%, 50%, and 120% of the NAAQS value of greatest regulatory relevance. The predetermined acceptance criteria were transport efficiencies of 97.5%) or greater (i.e., maximum loss of 2.5%), which was selected to identify potentially alternative materials to borosilicate glass and FEP Teflon®. This acceptance criterion was adopted from the through-the-probe audit program (US EPA, 2011) used for quantifying O3 line losses during field monitoring. Testing determined that O3 was the most reactive of the four gaseous NAAQS pollutants. O3 transport efficiencies for the four fluoropolymer tubing types (FEP, PTFE, PVDF, and PFA) ranged from 95.6% to 98.6%) without preconditioning. SS316, SNSS, and PVC had transport efficiencies of 0%, 9.2%, and 11.1% respectively, suggesting significant losses in the tubing without preconditioning. Following conditioning of each tubing type with 450 ppb of O3 for 1 hour, the transport efficiency of the fluoropolymer tubing types increased to an average of 99.4%. The high transport efficiency of O3 through fluoropolymer tubing was repeatable and independent of O3 concentration. Conditioning of SS316 tubing and Tygon® tubing resulted in 0% transport efficiency (i.e., 100% loss) of O3 for both tubing types. The transport efficiency of O3 through conditioned SNSS tubing improved to 87.7% with conditioning but remained below the study's 97.5% acceptance criteria. During both the unconditioned and conditioned tests, SS316, SNSS, and Tygon® test results were highly variable and dependent on the O3 concentration being tested. Unlike the tests conducted with O3, the transport efficiency of SO2, NO2, and CO through the four fluoropolymers exceeded the 97.5% acceptance criteria for transport efficiency without conditioning. The average transport efficiency of these three gases through the fluoropolymer tubing types was 99.8% and independent of test concentrations. For SS316, SNSS, and Tygon® tubing, test results with these three pollutants had higher transport efficiencies than obtained with conditioned O3 tests; although only tests conducted with CO resulted in transport efficiencies exceeding the goal of 97.5%. Based on this study's results, EPA will propose the revision of Section 9, Appendix E of 40 CFR Part 58 viii ------- to include PVDF, PTFE, and PFA as approved tubing types for regulatory sampling of all four gaseous criteria NAAQS pollutants. ix ------- 1.0 Introduction 1.1 Background Under the Clean Air Act (CAA), the National Ambient Air Quality Standards (NAAQS) were established by the Environmental Protection Agency (EPA) in 1971 for six criteria air pollutants, which are carbon monoxide (CO), nitrogen dioxide (NO2), ozone (O3), sulfur dioxide (SO2), particulate matter (PM), and lead. To ensure accurate and consistent measurements of these pollutants, specific measurements methods and techniques for ambient air instrumentation were defined and referred to as Federal Reference Methods (FRM) and Federal Equivalent Methods (FEM). The Code of Federal Regulations (CFR) Title 40 Parts 50 and 53 contain all published regulations for the implementation and verification of these methods. In order for state and local agencies to properly monitor and report these criteria pollutants (40 CFR §50.1), additional regulations were added to the CFR for the CAA in 1979. The 40 CFR Part 58 currently covers all aspects of establishing, maintaining, and reporting data from an ambient air monitoring site. In Section 9, Appendix E of Title 40 CFR Part 58, it is specified that the appropriate materials to be used as probes and intake sampling lines for measuring reactive gases from ambient air samples are Fluorinated Ethylene Propylene (FEP) Teflon®, Pyrex® borosilicate glass, or "other equivalent materials" which are not specified. Sample gases must also have a residence time within the material of less than 20 seconds (40 CFR 58, Appendix E, Section 9, QA Handbook 2017). Several inquiries have been received by the EPA Office of Research and Development's (ORD) Reference and Equivalent Methods Program regarding what commercially available materials might qualify as suitable "equivalent" materials. Research regarding other materials which could be considered equivalent and suitable for the transport of gaseous criteria pollutants has been limited during the past 40 years. The justification for limiting probe materials and sampling lines needs reinvestigation. The results in the research referenced in the CFR were not inclusive of all four gaseous criteria pollutants with each study focusing only on a subset of those pollutants. In addition, new materials have since been identified as potential alternatives, and materials previously tested may undergo new manufacturing processes which might affect the suitability as a material for NAAQS pollutant sampling. Although portions of Appendix E have been amended as recently as 2013, Section 9 has remained largely unchanged since its publication in 1979. The citations in the Section 9 publication regarding the use of Pyrex borosilicate glass or FEP Teflon® were from four publications: 1. Altshuller et al. (1961), 2. Hughes (1975), 3. Wechter (1976), and 4. U.S. EPA (1971) Field Operations Guide for Automatic Air Monitoring Equipment, No. APTD-0736, Research Triangle Park (RTP). 1. Experiments from Altshuller et al. (1961) indicated that FEP Teflon® and glass tubing were the best materials for O3 sampling. Stainless steel and polyethylene tubing were less suitable for O3 sampling, and polyvinyl chloride (PVC) tubing was unsuitable. 2. Hughes (1975) provided the standard preparations and overview of the storage pollutant reference standard gases and the proper types of cylinders to use for the storage of these gases. Hughes also investigated Teflon® permeation tubes and other reference gases including NO2 and SO2. The materials described in the Hughes article provided specific storage applications of the criteria pollutants. 1 ------- 3. Wechter (1976) reviewed the storage of gases in cylinders. The article evaluated a treated aluminum cylinder to guide the specialty gas industry to provide more stable and accurate calibration standards. Three of the four gaseous criteria pollutants, CO, SO2 and NO2 were used in Wechter's research. The treatments on the aluminum cylinder walls aided in maintaining the stability of the reactive calibration gases during the long-term performance testing of the cylinder. The treatment was described as a two-part process; the first part was to enhance the aluminum oxide layer, and the second part was listed as "proprietary in nature". 4. The U.S. EPA Field Operations Guide for Automatic Air Monitoring Equipment (1971) stated specifically that only borosilicate glass and FEP Teflon® can be used for inlet materials, citing Wohler's research from 1967. The article referenced the adsorption and desorption of glass, plastic, and metal tubing regarding CO and SO2 (Byers and Davis, 1970) and provided a schematic of the testing apparatus and test results. To update this research, a laboratory investigation was conducted to evaluate whether materials other than FEP Teflon® and Pyrex® borosilicate glass are acceptable for sampling gaseous criteria pollutants at ambient air monitoring sites. The results will inform and guide federal, state, and local agencies on decisions concerning acceptable materials for sampling criteria pollutants. Our study's results show that ambient air monitoring site managers and technicians can use a wider variety of materials when sampling gaseous criteria pollutants in order to maintain the required level of accuracy for NAAQS pollutant compliance determination. 1.2 Objective The purpose of this study was to determine what inlet materials could be considered to be functionally equivalent to FEP Teflon® and Pyrex® borosilicate glass for transporting and measuring gaseous criteria pollutants at ambient air monitoring sites. The primary objective of this work was to compare several candidate materials to FEP Teflon® to determine which materials can produce similar transport efficiencies when measuring gaseous criteria pollutants through those materials. Results were obtained for the four gaseous criteria pollutants, O3, N02, S02, and CO. Because the reactivity of these gases differ, results were different for each of the gaseous criteria pollutants. An overall recommendation of the best performing materials based on the research results has been presented at the conclusion of this report. 2 ------- 2.0 Methods 2.1 Experimental Approach To determine which materials would be suitable equivalents for FEP Teflon® and Pyrex® borosilicate glass, known concentrations of each gaseous analyte (O3, SO2, CO, and NO2) were generated by a dilution calibrator (Teledyne Air Pollution Instruments (TAPI) T700U, San Diego, CA) and delivered to an FRM or FEM analyzer that measures each specific analyte. The T700U dilution calibrator contains three mass flow controllers (MFCs) (one for diluent gas and two for target analyte gas), an O3 generator, and a photometer. The calibrator was connected to a cylinder which contained known concentrations of SO2, CO, and NO2 gases to produce accurate concentrations. The gas analyzers selected to measure the target analytes were FRM and FEM instruments that are utilized at ambient air monitoring sites. These analyzers communicated with a data acquisition system (DAS) using a data collection application Envidas (DR DAS LTD Envidas, Granville, OH). Once the analyzer was connected to Envidas, it continuously streamed data into the DAS database. The specific details of the T700U, the gas analyzers, DAS and its software, and other related equipment will be described in a later section. During each experiment, the target analyte gas was delivered through a 5-foot length of 1/4 inch OD, 3/16-inch ID FEP Teflon tubing as a control. While the flow was directed through the control material, measurements were taken by the target analyte instrument and recorded in the DAS. This control measurement represented the concentration of gas at the inlet of the test material. After obtaining the control value, the flow was then delivered through the test material and measurements were collected by the analyzers at the outlet. The results from the control material and the test material determine the transport efficiency of the test materials. _ , rrr- ¦ Concentration o f Tar qet Analyte At The Outlet of the Test Material . __ Transport Efficiency (%) = - X 100 Concentration of Target Analyte At The Inlet of the Test Material Equation 1. Transportation Efficiency for a Target Analyte through Test Material Once stable, the average concentration measurements were read by the analyzer during the final 1- minute average of data from each test and recorded into the DAS. The data from the inlet and outlet of each test material were compared, and the transport efficiency of each test material was calculated using Equation 1. Transport efficiencies close to 100% represent an experiment in which the loss of concentration of the target analyte was minimal. Appreciably lower transport efficiencies demonstrated that there was a significant loss of concentration of the gas during its contact time with the tubing material. 2.1.1 Pilot Testing In preparation for this study, several pilot tests were run to determine how to represent the transport efficiency. These pilot tests were performed using O3 as the target analyte. Previous publications indicated that O3 is the most reactive criteria pollutant of the four gases being tested (Altshuller et al., 1961). These pilot tests determined methodology used for measuring the transport efficiency through each candidate test material. 3 ------- A list of pilot test objectives, descriptions of each test, and determinations made from their results are presented in Table 1. Table 1. Pilot Tests Experiment Objectives, Descriptions, Results, and Determinations Test Pilot Test Objective Description of Test Results Determination 1 Demonstration that the flow path setup would operate correctly Delivered O3 to a T265 ozone analyzer through control material (FEP Teflon®) and subsequently through test material (SS316). Expected O3 concentration through each material matched results, demonstrating that the system was operating appropriately. This flow path setup was used for all experiments. 2 Determination of the amount of time required to stabilize the target analyte gas Delivered O3 to a T265 ozone analyzer through PTFE and SNSS material with a residence time of 20 seconds for both materials. O3 concentrations through PTFE tubing took approximately 5 minutes to stabilize, while SNSS tubing took 20 minutes or more to stabilize. Upon review of data, the minimum test stabilization time was determined to be 40 minutes in length, with the control material and test material being exposed to the target analyte for 20 minutes each. 3 Determination of whether cleaning metal tubing with solvent affected results Delivered O3 to a T265 ozone analyzer through SS316 and SNSS material and measured O3 concentrations before and after cleaning each piece of tubing with Methylene Chloride followed by an Acetone Rinse. No change in concentration was observed after either metal tubing material was cleaned with solvent. Metal tubing materials were not cleaned with solvent prior to experiments. Test 1. During pilot testing, initial tests were conducted to determine the most practical experimental setups for the research project purposes. The two solenoid valves that were connected to the ADAM 5000-series have three valve ports. The first port was a common port, which was always open. The second port was the normally open port, which was open only when the solenoids did not have a voltage delivered to them. The third port was the normally closed port, which was only open when the solenoids had a voltage delivered to them to engage the diaphragm within the valve. Voltage was delivered to the solenoids through the ADAM 5000-series system, which was controlled by the DAS. When the normally open side of both solenoid valves were open, gas flowed through the control material to the analyzer. When the normally closed side of both solenoid valves were open, gas flowed through the test material to the analyzer. The normally closed side of the solenoid valves were opened by delivering a voltage to the solenoid valves to engage the diaphragm and switch the valve flow path. For each experiment, the two solenoid valves were set to alternatively switch between the control material and the test material at five-minute intervals. These switches were automated by the ADAM- 5000 series modular I/O system (Advantech ADAM-5000-series Modular I/O, Irvine, CA) which delivered a voltage to the solenoid to activate the valve in order to open the normally closed side of the 4 ------- valve and divert flow through the test material. The valves were automated to switch at regular 5-minute intervals for the duration of the experiment, diverting flow through the control material to determine the inlet concentration and test material to determine the outlet concentration of a material for an equal time of 20-minutes while the target analyte was generated for a total of 40 minutes. By using a known control material (FEP Teflon®) and confirming that the inlet and outlet concentrations were identical, the flow path would be determined to be acceptable for use in testing. During pilot testing, manual 3-way stainless steel ball valves (Swagelok SS-43GXS4, Solon, OH) were turned by hand to change the flow path. These were replaced during the actual tests by automated PFA solenoid valves (3-Way Valve 203-3414-215, SemiTorr Group, Inc, Hillsboro, OR 97124) controlled by the ADAM-5000. After performing an initial linearity check using gas concentrations generated by the T700U to adjust the slope and intercept of the T265 O3 analyzer (TAPI T265, San Diego, CA), initial tests compared the O3 concentration values generated and measured by the T700U to those measured by the T265 when connected to the test setup. A 5-foot piece of FEP Teflon® served as the control material, while a 22-foot piece of 316 stainless steel (SS316) served as the test material. The residence time of the stainless-steel tubing was approximately 14-15 seconds for the pilot test. Because the residence time had not been determined when this test was conducted, aligning it within the CFR specifications was a key factor for the test material in this experiment. FEP Teflon® was already considered to be acceptable by the CFR, and previous research indicated that SS316 had poor transmission of O3, especially at sub-100 ppbv concentrations (Scholz, 2015). The pilot experiment tested concentrations of 400 ppb (parts per billion) and 200 ppb O3, which were chosen because they fell into the range of the T265 analyzer (max 500 ppb). The T265 O3 readings through the FEP Teflon® compared favorably to the T700U photometer measurements, while T265 O3 readings through the SS316 showed little to no O3 being delivered to the analyzer, both of which were the expected result based on this literature review. This performance can be seen in Figure 1. 5 ------- Site Report - Lab D288-A Date&Time: 24/032021 09:00 09:00:00 - 13:00:00 1 Vintrnl Matprisl Tp«;t est Material Test Linearity Test 1 J Troubleshoot r 1 1 09:00 09:10 09:20 09:30 09:40 09:50 10:00 10:10 10:20 10:30 10:40 10:50 11:00 11:10 Date & Time 11:20 11:30 11:40 11:50 12:00 12:10 12:20 12:30 12:40 12:50 13:00 Figure 1 Example Graph of FEP vs Grade 316 Stainless Steel Pilot Test. The graph in Figure 1 includes a linearity check, pilot test measuring O3 concentrations at the inlet and at the outlet of the test material (SS316), and confirmation check (troubleshoot) to confirm the dilution calibrator was still generating O3 gas as it was unknown at the time that SS316 would be rather low in transport efficiency. Time is displayed on the x-axis, while the O3 concentration is displayed in "ppb" on the y-axis. As mentioned previously, these results confirmed that the flow path system was functioning as expected, and this test setup was planned for use going forward for all test material experiments. Test 2. Tests were conducted to determine the time required for the O3 concentration to stabilize. The control material was always planned to have a very short residence time to represent the inlet concentration of the test material. However, the residence times for the test material, the target gas concentrations for each experiment, and the amount of time each experiment would run had not been determined. Therefore, tests to determine the length of time to allow O3 to stabilize during each experiment were conducted by using a flow of 450 ppb O3 through PTFE (fluoropolymer tubing) and a metal test material (grade-316 stainless steel with SilcoNert® 2000 (SilcoTek® Corporation, Bellefonte, PA) non-reactive coating [SNSS]), with an initial residence time of 20 seconds for both materials. This time was chosen as it is the maximum allowable residence time in the Section 9, Appendix E of 40 CFR part 58. During these tests, the O3 concentrations at the analyzer stabilized after approximately 5 minutes for the PTFE tubing. The SNSS required 20 minutes or more to stabilize. An example of this test can be seen in Figure 2. 6 ------- Site Report - Lab D2B8-A Date&Time : 15/04/2021 07:30 07:30:00- 13:30:00 cm _ 550- 500- 450- 400- 350- ® 300 ~ ZD >250- 200- 150- 100- 50- , 0 - ir U - _*\fl « 30 CnJ - 07 1 1 1 30 08:00 08:30 09:00 1 1 09:30 10:00 1 1 10:30 11:00 Date &Trme l 11:30 12 1 1 00 12:30 13:00 13 T2fi5_D3[ppb] Figure 2. Example Graph of Stainless Steel with SilcoNert® Coating vs PTFE Pilot Test Figure 2 displays the results from the SNSS and PTFE residence time pilot tests. Time is displayed on the x-axis while O3 concentration is displayed in "ppb" on the y-axis. The portion of the graph from 8:25 to 10:10 was when O3 gas was being measured at the outlet of the SNSS material, while the portion of the graph from 10:40 to 11:40 was when O3 gas was being measured at the outlet of the PTFE material. Each material received the same two concentration steps of O3 gas during those periods of time, which can be seen during their respective rises in concentration. Based on these stabilization pilot tests, it was decided that the total amount of time that each experiment would run would be 40 minutes, with each material being exposed to the target analyte for 20 minutes. By observing the stabilization of O3 concentrations in each material, the longest time (in SS316) was used to determine the length of the run time of the experiment. Test 3. Lastly, the original conditions prior to testing of the test material candidates were investigated, such as whether the stainless-steel tubing (SS316 or SNSS) should be pre-cleaned with solvent as there could be leftover oils or other foreign material remaining from the manufacturing process. Tests were run with stainless steel materials before and after solvent cleaning to see if any changes in O3 responsivity occurred. Two tests were run in which each of these materials was connected as the test material both before and after the solvent cleaning procedure. Figure 3 shows a side-by-side comparison of the graphs displaying the results, with the materials before solvent cleaning on the left and the materials after solvent cleaning on the right. Time is on the x-axis, while O3 concentration in "ppb" is on the y-axis. The graphs show 7 ------- that the solvent cleaning resulted in no significant change in measured O3 concentration through either material, demonstrating that addition of solvent cleaning did not significantly affect the results. For example, the percent change for SS316 at the highest concentration of O3, 400 ppb, was approximately 0.8%, while the percent change for SNSS at 400 ppb was -0.5%. Therefore, it was decided that no solvent cleaning would be performed for the experimental runs with stainless steel materials. ~~1——1——1——1——1——1——1——1——1——1——1——1——1——1——1 10:00 10:15 10:30 10:45 11.-00 11:15 11:30 11:45 12:00 12:15 12:30 12:45 13:00 13:15 13:30 13:45 Date & Time 400ppbSS316 1 400 ppb SN jK / J / / 200 ppb SNSS j 200ppbSS316 1,1,1 1 1 1 1 1 1 1 1 1 1 1,1,1 1 1 1 1 1 1 Stainless Steel and Silconert Stainless Steel Before Cleaning Stainless Steel and Silconert Stainless Steel After Cleaning Figure 3. Side by Side Comparison of Os Measurements through SS316 and SNSS Before and After Solvent Cleaning 2.1.2 Target Analytes and Test Materials The target analytes for these experiments were four gaseous criteria pollutants from the NAAQS, O3, S02, N02, and CO. The test materials that were studied and compared to FEP Teflon* during these experiments were polytetrafluoroethylene (PTFE), perfluoroalkoxy (PFA), polyvinylidene fluoride (PVDF, often referred to by its trade name Kynar®), a flexible thermoplastic tubing polyvinyl chloride (PVC, often referred to by its trade name Tygon® S-3, Formulation B-44-3 | Saint-Gobain Performance Plastic, Courbevoie France]), grade 316 stainless steel (SS316), and grade 316 stainless steel with a non-reactive amorphous silicon coating (Pac Stainless Ltd, Ftouston, TX [SNSS]). These materials, as well as the part numbers and vendors are listed in Table 2. 8 ------- Table 2. Test Materials List Material Manufacturer/Supplier OD (in) ID (in) Part Number Description FEP Teflon® Tubing St. Gobain/McMaster- Carr (Atlanta, GA) 0.250 0.188 52355K57 Fluorinated ethylene propylene tubing (control material) PFA Tubing McMaster-Carr 0.250 0.188 5773K14 Perfluoroalkoxy tubing PTFE Tubing Zeus/McMaster-Carr 0.250 0.188 5239K12 Polytetrafluoroethylene tubing PVDF Tubing McMaster-Carr 0.250 0.172 51105K22 Polyvinylidene fluoride SS316 Tubing Webco Ind./McMaster-Carr 0.250 0.194 89995K83 Stainless steel tubing of alloy 316 SNSS Tubing Pac Stainless (Houston, TX) 0.250 0.180 76C0250035BAO Stainless steel tubing of alloy 316 treated with the Silconert® 2000 non- reactive coating PVC Tygon® S3 Tubing St. Gobain/McMaster- Carr 0.250 0.188 6516T17 Thermoplastic Polyvinyl Chloride, Polymer Tubing To determine the inlet concentration, FEP Teflon® was chosen for use in all experiments as the control material tubing as seen in Figure 4. It is the most readily available CFR approved tubing, as the other CFR approved tubing, glass, must be custom made to suit the purposes of a specific project or monitoring site. The use of FEP Teflon® also provided flexibility in the set-up of the testing apparatus. For polymer tubing materials, PFA and PTFE were chosen as readily available alternatives to FEP Teflon®. While some research articles have referenced using these or similar materials for various air sampling methods (Campos et al., 2013), no specific testing of these materials for ambient air sampling has been conducted. Tygon® was also chosen although it had been tested in the past and was considered unsuitable for sampling O3. PVDF was chosen as another fluoropolymer alternative used in applications, such as filter elements for previously approved FEM instruments and tubing connecting fittings inside many types of devices. No additional research had been performed specifically with PVDF as a material in a sampling system to measure ambient air quality. For metal tubing materials, 316 stainless steel with SilcoNert® 2000 non-reactive coating (SNSS) was chosen for testing, as it is commonly used in other pollutant measurement applications in industry. In particular, the company that produces the SilcoNert® coating, SilcoTek®, markets its products for use in gas chromatographs, coatings for stainless steel-based gas cylinders, and other apparatuses for analytical equipment. Additionally, the Marshik (2015) presentation at EPA's Annual Quality Assurance Conference provided a summary of their research with guidance of materials for continuous emissions monitoring, sampling line testing, and best practices which demonstrated that the SilcoNert® coated stainless steel sampling probe and sampling line outperformed stainless steel, PTFE, and PFA for NO2, hydrogen chloride, ammonia, and formaldehyde regarding storage and retention effects. SilcoNert® was demonstrated to prevent the adsorption of gases measured from refineries in 40 CFR 60.100 Subpart Ja. The gases of interest in that research were SO2, hydrogen sulfide, nitric oxide (NO), CO, and a general array of volatile organic compounds. The coating was incorporated in the sampling probes, tubing, valves, fittings, and analyzer components. Results from testing by the Shell Corporation indicated that SilcoNert® improved sulfur detection and response compared to other stainless-steel tubing. Because SNSS was chosen for testing, grade 316 stainless steel without the SilcoNert® coating (SS316) was also chosen for comparison purposes. Additionally, testing has been performed using stainless steel in the past for other purposes. For example, the results of transport efficiency tests through 316 stainless 9 ------- steel tubing conducted by Obermiller and Charlier (1968) showed no significant difference in efficiency using "hot" and "cold" chromatography columns. Palmes et al. (2010) used stainless steel to make screens for a NO2 personal sampler, the screens were made to be efficient, usable in the field, lightweight, and unbreakable. Research performed Scholz (2015) also demonstrated that SS316 does not transport O3 efficiently at low concentrations. For each test material, the length of tubing necessary for testing was calculated by setting a target residence time of 20 seconds. In Section 9, Appendix E of 40 CFR Part 58, this is specified as the longest residence time acceptable for a sampling manifold. Using the inner diameter of each test material tubing, the target residence time of 20 seconds, and the flow rate of the analyzer used to measure gas concentrations, the length needed for each test material tubing could be calculated. The equation for tubing length based on residence time can be seen in Equation 2. Instrument Flow Rate * Residence Time Tubing Length = ^ , . ; — Tubing Cross Sectional Area Equation 2. Tubing Length Based on Residence Time of Gas through Test Material While the inner diameter (ID) of the tubing could vary somewhat depending on the manufacturer or material used, the tubing outer diameter (OD) for each material was Vi inch. All outlets from the T700U dilution calibrator and all inlets for the target analyte instruments were designed for Vi inch OD tubing. Other equipment for the flow apparatus was also designed around this Vi inch OD, including the ports for the solenoids, T-fittings used, and other hardware used. Additionally, in most cases, the test material was not pre-treated in any way, nor was it exposed to the target analytes prior to testing. However, in cases where the test material in its unconditioned state had a transport efficiency observed to be below 97.5% during the initial experiments, an additional test was performed following conditioning of the material. This conditioning and additional testing also applied to all test materials tested with O3. 2.2 Experimental Setup 2.2.1 Experiment Equipment The experimental equipment used to generate, measure, and collect data for the target analytes are shown in Table 3. 10 ------- Table 3. Equipment List Manufacturer Model Number Description Advantech ADAM-5000 Series Modular I/O Modular I/O Controller for Solenoid Valves DR DAS LTD Envidas System Data Acquisition System TAPI T700U Dilution Calibrator/03 generator/Photometer TAPI T701H Zero Air Generator TAPI T265 O3 Analyzer TAPI T500U CAPS N02 NO2 Analyzer Thermo-Fisher Scientific 43c SO2 Analyzer Thermo-Fisher Scientific 48c CO analyzer SemiTorr Group, Inc. Valve-GalTek-Solenoid 1/4 3-WA 3-Way Solenoids Swagelok® Various Stainless Steel and Teflon nuts, ferrules, and caps All tubing used for the flow path was secured to the instrumentation using Swagelok® compression fittings. The T700U was equipped with three Mass Flow Controllers (MFCs), one for diluent gas and two for target analyte gas depending on the required flow rate, an O3 generator, and a photometer to generate and quantify each of the target analyte gases at a known concentration. The T700U used an O3 generator and a calibrated and certified photometer to generate and measure accurate concentrations of O3. The T700U used a zero-air generator (TAPI T701H, San Diego, CA) to generate "zero-air" (filtered air with no contaminants) as its diluent gas and a gas cylinder with known concentrations of SO2 and CO was used to generate accurate concentrations of those gases at a particular flow rate for each experiment. For these experiments, the cylinder used contained a tri-blend of gases consisting of approximately 300 ppm CO, 10 ppm SO2, and 10 ppm NO (purchased and received from AirGas in Morrisville, NC, see Table 4). The T700U used the O3 generator and a gas cylinder (see Table 4) with known concentrations of NO in a process called Gas Phase Titration (GPT) to create the NO2 gas samples. The T700U generated user set concentrations of NO and O3 gases and mixed them in a container with a known volume and residence time within the instrument. The O3 gas was set as the limiting reactant, and the concentration of NO2 gas generated was directly related to the concentration of O3 gas generated and mixed during this process. The analyzers used to assess the target analyte concentrations were the FRM and FEM instruments used previously at ambient air monitoring sites to measure the appropriate gas concentrations. O3 was measured by a TAPI T265 (TAPI, San Diego, CA) analyzer. This analyzer operates through a chemiluminescent reaction of O3 and NO, and the analyzer requires a 1% NO cylinder (see Table 4) to be connected, as that gas reacts with the O3 in the sample gas to produce the chemiluminescent reaction. The reaction of O3 with NO results in the production of electronically excited NO2 molecules. Each of these excited NO2 molecules release its excess energy by emitting a photon (hv) and dropping to a lower 11 ------- energy level. The number of emitted photons is directly proportional to the O3 concentration in the sample stream. A photomultiplier tube (PMT) detects the hv emissions from the excited NO2 during this process. To measure SO2 concentrations, a Thermo-Fisher Scientific 43c (Thermo-Fisher Scientific, Waltham, MA) analyzer was used. This instrument operates on the principle that SO2 molecules absorb ultraviolet (UV) light and become excited at one wavelength, then decay to a lower energy state emitting UV light at a different wavelength. A PMT detects the UV light emissions from the decaying SO2 molecules during this process. For N02, a T500U (TAPI T500U Cavity Attenuated Phase Shift (CAPS), TAPI, San Diego, CA) analyzer was used. This analyzer uses light from a blue UV light emitting diode (LED) centered at a wavelength of 450 nm, a measurement cell with high reflectivity mirrors at either end, and a vacuum photodiode detector to measure the phase shift in a regular fluctuating light signal, which is proportional to the absorption cross sectional area. Lastly, a Thermo-Fisher Scientific 48c (Thermo-Fisher Scientific, Waltham, MA) analyzer was used to measure CO values. This instrument measures CO using Gas Filter Correlation (GFC). This is based on the principle that CO absorbs infrared (IR) radiation at a wavelength of 4.6 microns. The radiation from an IR source within the instrument is chopped by a chopper wheel and passed through a gas filter alternating between CO and nitrogen gas. The CO gas filter produces a reference beam, while the nitrogen gas filter produces a measurement beam, which can be absorbed by the sample CO in the cell. In this way, the GFC system responds specifically to CO. The gas cylinders used in operation are listed in Table 4. The gases in these cylinders were certified by AirGas using EPA protocol to obtain accurate concentrations (US EPA, 2012). Table 4. Gas Cylinder List Supplier Location Cylinder Type Description AirGas Morrisville, NC Tri-blend gas cylinder EPA protocol certified gases for NO, CO, and SO2 (NO and SO2 at approximately 10 ppm, CO at approximately 300 ppm) AirGas Morrisville, NC 1% NO gas cylinder EPA protocol certified 1% NO gas for T265 analyzer operation 12 ------- 2.2.2 Initial Sample Gas Flow Path => Figure 4. Diagram of Flow Path Experimental Setup Figure 4 displays a flow diagram of the final set-up of the sampling apparatus used for all experiments. During each experiment, the zero-air generator T701H, was used to generate zero air diluent gas, and the gas cylinder of the target analyte, were continuously connected to the MFCs and O3 generation, which were contained within the T700U dilution calibrator. The "exhaust point t-fitting" was installed along the flow path prior to the valves to ensure that the target analyte FRM or FEM monitor did not get over-pressurized, and to ensure that the residence time was based on the target analyte FRM or FEM monitor internal pump flow rate rather than the output from the dilution calibrator. Flow rates were adjusted to ensure that the gas analyzer was exposed only to the target analyte rather than to room air. The O3 generator within the T700U, directed O3 gas through the instrument's internal photometer, which provided a measurement by which to adjust the generator output accordingly to achieve the target O3 concentration. The O3 gas was then delivered throughout the rest of the flow path in Figure 4, where the T265 analyzer would draw in the sample gas. For NO2, the T700U dilution calibrator generated NO2 through GPT of NO and O3 gases. The two gases mixed inside a chamber of a known volume and retention time in the T700U, allowing the two gases to react to create NO2 gas with O3 being the limiting reactant. This resulted in the output generation of the target NO2 gas concentration along with excess NO gas. The NO2 gas was then delivered throughout the rest of the flow path in Figure 4 where the T500U analyzer would draw in the sample gas. For SO2 and CO, a cylinder containing known concentrations of both SO2 and CO was connected to the dilution calibrator. The MFCs within the dilution calibrator then routed appropriate flows of the diluent 13 ------- and target analyte gases to achieve the target concentration and target flow rate. These sample gases were then delivered throughout the rest of the flow path in Figure 4, where an analyzer would draw in the sample gas. For SO2, the analyzer was a Thermo 43c while for CO, the analyzer was a Thermo 48c. 2.2.3 Sample Gas Flow Along the Rest of the Flow Path The T700U dilution calibrator routed the target analyte gases into the common inlet of the first 3-way solenoid valve (SemiTorr Group, Inc., Valve-Galtek-Solenoid 1/4 3-WA, Tualatin, OR) in the setup through FEP Teflon® tubing, or control material (valve 1 in Figure 4). The first two solenoid valves have three valve ports. The first port was a common port, which was always open. The second port was the normally open port, which was open only when the solenoids did not have a voltage delivered to them. The third port was the normally closed port, which was only open when the solenoids had a voltage delivered to them to engage the diaphragm within the valve. Voltage was delivered to the solenoids through the ADAM 5000 series modular I/O system, which was connected to the DAS. The solenoid valves were automatically switched every 5-minutes by sequences programmed into the DAS during each experiment. The normally open outlet of the 3-way solenoid valve was connected to the control material, while the normally closed outlet was connected to the test material. Both materials were connected to a second 3- way solenoid valve (valve 2 in Figure 4), where the control material was connected to the normally open inlet and the test material was connected to the normally closed inlet. The common outlet of this 3-way solenoid valve was then connected to the appropriate analyzer chosen per test material. From this setup, the target analyte instruments pulled in the sample gas using their internal pumps, while any excess sample gas flow was directed into the exhaust for the laboratory. The timing of switching the valves was based on the results of the gas stabilization pilot test performed before these experiments were conducted. Based on the previously described stabilization pilot tests, it was decided that the total experiment run length was 40 minutes, with each material being exposed to the target analyte for 20 minutes. A visual representation of the timing process used for each transport efficiency experiment is presented in Figure 5. Data used in calculations (1 minute) Beginning of test Initial entry of target analyte 5-Minutes through Test Material 5-Minutes through Test Material /j\ \ i ) 20 Minutes Zero Air to Purge Both Materials 5- Minutes through Control Material Sequence Re[ Total 30 minutes, 15 seated 3 times minutes per material 5- Minutes through Control Material Last 10 minute \ Data used in calculat s of experiment n/ ons (1 minute) Figure 5. Flow Chart of the Valve Switch Timing Process During Each Experiment The flow chart in Figure 5 displays the full sequence of events as they occurred during each transport efficiency experiment. For the initial 20 minutes, the control material and test material were purged with zero air for 10 minutes each. Following that, while the dilution calibrator generated the target analyte at the specified concentration for 40 minutes, the solenoid valves switched between the control material and test material every 5 minutes. The data used in the calculations for transport efficiency were the final 1-minute averages taken when the target analyte was flowing through the control material and the 14 ------- test material, shown in the colored portions the last 10 minutes of the experiment. 2.2.4 Data Flow Path Dilution Calibrator (T700U) ADAM 5000-Series Data Flow Data Flow Data Flow Data Acquisition System (Envidas) Target Analyte Instrument Figure 6. Diagram of the Data Flow of the System Figure 6 shows the flow of data for the Envidas system and the associated equipment. The Envidas DAS was used to record the data from the target analyte analyzers (different for each analyte) and to control the solenoid valves. Each of the analyzers were connected to the DAS through either Transmission Control Protocol/Internet Protocol (TCP/IP) or Recommended Standard 232 (RS-232) serial cable connection. The analyzers continuously streamed their data to the DAS, which recorded their data into a Structured Query Language (SQL) database. The DAS also had an application that allowed users to view the data as it was streamed into the system in real time via the Envidas Viewer application. The Envidas Reporter application allowed users to generate graphs and tables of data that had already been recorded into the database. An example of a graph recorded by the DAS and displayed by the Envidas Reporter application is shown in Figure 7. 15 ------- Site Report - Lab D2SS-A Date&Time: 16/09/2021 13:4913:49:00- 14:49:00 90 -i 85- 30- 75- 70- 65- 60- 55- 50- 45- 40- 35- 30- 25- 20- 15- 10- 5- 0 - V1"-— n L .... -R - 13:50 13:55 14:00 14:05 14:10 14:15 14:20 14:25 14:30 14:35 14:40 14:45 Date & Time T265_03[ppb] Phctometer_03[ppb] Figure 7. Example Graph of an Experiment from the Envidas Reporter Application Figure 7 is an example of the graphical representation of the data recorded during an O3 experiment, as both the T700U photometer readings of the generator output and the T265 O3 analyzer readings are displayed on the graph. Time is displayed on the x-axis, while O3 concentration in "ppb" is displayed on the y-axis. The orange trace on the graph is the data recorded by the T265 during the experiment, while the red trace on the graph is the data readings of the photometer in the T700U dilution calibrator, which represents the "actual measured output" of O3 from the T700U by the O3 generator. The Envidas system recorded analyzer data at three different time resolutions. For the analyzers T265, 43c, and 48c, the time resolutions were 5 seconds (lowest resolution for these instruments) and 1 minute, while for the T500U analyzer, the time resolutions were 15 seconds (lowest resolution for this instrument) and 1 minute. 16 ------- Figure 8. Photograph of Flow Path Experimental Setup An additional piece of equipment that was connected to the Envidas DAS was an ADAM 5000-series modular I/O system. This equipment allowed the DAS to control the solenoid valves, as the DAS could give the ADAM 5000-series system a command to deliver a voltage to the solenoids to engage the valve diaphragm and divert the flow path. The ADAM 5000-series system was connected to the DAS through a RS-232 cable, and the solenoids had an analog connection to the ADAM 5000-series system. 17 ------- Figure 8 is a photograph of the real-time flow path setup used for all experiments. The valve on the right side labeled Solenoid 1 was connected to the T700U dilution calibrator through its common port. The tubing from the common port in that solenoid valve can be seen leading off the right side of the photograph, which is where the T700U is located. Solenoid 1 was switched off to direct flow through the normally open port when delivering gas through the control material, while it was switched on to deliver flow through the normally closed port when delivering gas through the test material. Regardless of which material had gas flowing through it, the gas would arrive at the next valve on the left side of the setup labeled Solenoid 2. The common port of this solenoid was connected to the respective analyzers, which were located behind the sampling system apparatus and obscured by the teal bar on the laboratory bench. For this solenoid, when gas was arriving from the control material, the solenoid was switched off to direct flow through the normally open port, and when gas was arriving from the test material, the solenoid was switched on to direct flow through the normally closed port. 2.3 Final Experimental Procedure The pilot tests provided insights into the components of the testing apparatus and each materials' ability to passivate large concentrations of O3. However, these large concentrations were based upon the range of the O3 analyzer. Following discussions with project stakeholders, a decision was made to base the experimental test concentrations upon the NAAQS concentration values. The three target concentrations were 20%, 50% and 120% of the current NAAQS concentrations, as these targets would provide a range likely to be encountered in the field. During each experiment and following QA/QC checks, the first step was to deliver zero air to the analyzer and through the control and test materials for 20 minutes (10 minutes in each material). The purpose of this was two-fold. It would ensure that the analyzer was appropriately reading zero before an experiment, and it would purge any potential outside contaminants that the flow apparatus may have prior to an experiment. Following this, the pre-determined concentrations of each target analyte were generated by the dilution calibrator, T700U, for a pre-set amount of time. These gas concentrations are presented in Table 5. Table 5. Setpoint Concentrations Setpoint O3 Cone. NO2 Cone. SO2 Cone. CO Cone. Zero Air Oppb Oppb Oppb Oppb 20% of the NAAQS 15 ppb 20 ppb 15 ppb 1,800 ppb 50% of the NAAQS 35 ppb 50 ppb 40 ppb 4,500 ppb 120% of the NAAQS 85 ppb 120 ppb 90 ppb 10,800 ppb While the dilution calibrator generated the target analyte gas for each experiment, the two 3-way solenoid valves were set to switch between the control material and test material at regular 5-minute intervals. This process was automated by an ADAM 5000-series modular I/O system that would deliver a voltage to the solenoid to engage the diaphragm and change the valve flow path to the test material, when necessary, while the diaphragm would not be engaged to have the flow path go through the control material. Through this process, the target analyte was delivered to the appropriate analyzer through the control material and the test material for an equal amount of time, which was 20 minutes total for each material. Once the target analyte was generated for the pre-determined amount of total time of 40 minutes and the analyzer data for the experiment were collected by the DAS, the system was set to run zero air through both materials before the experiment was over. This process was repeated in triplicate for all materials, for all target analyte concentrations, and for different material conditions 18 ------- when necessary. Each new experiment in the triplicate runs required a new piece of test material tubing so that every experiment had an unused, unconditioned piece of test material. Before the experiment was conducted, each test material used for any experiment was labeled with the appropriate information, such as target analyte, concentration, and other relevant information. Once the experiment was completed, the material was uninstalled from the experiment setup, capped by placing a stainless-steel Swagelok fitting and a stainless-steel Swagelok cap on both open ends of the tubing, and stored in the laboratory. Test materials that had poor transport efficiency results during the initial experiments with new, unused tubing were included in a follow-up test in which the test material tubing sections were conditioned to the target analyte before transport efficiency experiments were conducted. For the purposes of these experiments, a test material transport efficiency was classified as unacceptable if it was calculated to be below a threshold of approximately 97.5%. However, O3 experiments were an exception, as all materials had experiments in which the material was used in the conditioning process for that target analyte. For the conditioning tests, a new section of the test material was conditioned using the target analyte and was then installed into the testing setup. After the two solenoid valves were engaged so that the flow was delivered to the analyzer through the test material, the T700U dilution calibrator generated the target analyte gas at a high concentration, and the installed test material was exposed to this high concentration for at least an hour. The conditioning concentration was set to 450 ppb. The timing and the concentrations were chosen based on the Field Standard Operating Procedures (SOP) for the EPA, Through-the-Probe National Performance Audit Program (NPAP), (EPA, 2011). In that document, in Section 3.1.7.1 titled "Quarterly Ozone Line Loss Set-up", it states that the last step in the preparation for the O3 line loss test is to "Allow the system to condition with approximately 0.450 ppm of ozone for one hour". Following this conditioning process, testing commenced on this section of test material tubing in a similar manner as previously described. However, there was one primary difference between the original material procedure and the conditioned material procedure. Instead of using a new piece of tubing for each concentration, all three concentration levels were delivered through one piece of conditioned tubing. Then, once all three concentrations had been completed, a new piece of tubing was conditioned, and the process was repeated in triplicate. For all experiments, all data from the analyzers were continuously logged into a DAS, which had proprietary Envidas software installed that collected all gas analyzer data. Data being collected were received into the SQL database at a 5 second time resolution for the T265, and 43c, and the 48c, while data were received into the SQL database at a 15 second time resolution for the T500U. This difference was because the T500U could not transmit data to the SQL database at a time resolution lower than 15 seconds. Experimental data were visually inspected in the software's viewer application by the experimenter present while tests were being run. As the data were streamed from the analyzers and recorded in the DAS, the DAS stored these data in an internal SQL database. This database was accessed and visually inspected through the software's reporter application. This application was also used to export the data into Excel format and stored on the DAS. For the T265, and 43c, and the 48c, 5-second and 1-minute data were referenced and exported, while for the T500U, 15-second and 1-minute data were referenced and exported. Once exported to Excel, the data were further reviewed by the experimenter and other members of the 19 ------- research team. Following that review, Excel was used to process the data, where transport efficiencies were calculated based on the final minute average concentrations read by the analyzer when the gas passed through the control material and when the gas passed through the test material. 20 ------- 3.0 Quality Assurance (QA) and Quality Control (QC) To maintain quality assurance through quality control (QA/QC), this project was conducted under the approved category B, Applied Research Quality Assurance Project Plan (QAPP), "Reference and Equivalent Methods Program Support, QAPP-1J16-009.3" (QA Track J-AMCD-0031471-QP-1-2). 3.1 Equipment Calibration and Verification The T700U dilution calibrator was purchased from TAPI in April of 2021, and at that time, the vendor provided the certificates for the instrument's MFCs and photometer. A year after that date, the ORD/RTP Metrology Laboratory (Met Lab) performed the calibration and recertification of the photometer and MFCs within the T700U. The Met Lab also performed the verification and recertification of the DryCal flow meter used to measure flow rates. The frequency of these calibrations occurred at a minimum of annually or a year after the current certificate was performed. The photometer that measures the O3 concentration output of the instrument must be checked against a standard reference photometer, and the MFCs must be checked against a NIST-Traceable flow meter. The records of these calibrations were stored with the instrument in the laboratory, and the dates of these calibrations were recorded in the project laboratory notebook. Scanned digital copies of these records were stored on the Microsoft Teams Site. 3.2 QA/QC Checks Each target analyte analyzer (e.g., T265 O3 analyzer, Thermo 43c SO2 analyzer, T500U CAPS NO2 analyzer, and Thermo 48c CO analyzer) was required to have an intercept and slope adjustment QC check prior to each one of several regular intervals, depending on the instruments' current condition. These intervals included: an annual basis, when a new gas cylinder had been installed for the T700U dilution calibrator, or when the MFCs and photometer within T700U dilution calibrator had been calibrated by the Met Lab. QC checks were also used to verify instrument operation if the experimenter noted unexpected results during testing. Unexpected results could include the FRM instrument measuring high concentrations when delivering zero-air to the instrument or the FRM instrument reading more than +/- 5% away from the expected value of a reference measurement. These checks were used by the experimenter to troubleshoot the instrument if the problems were observed during an experiment and any observations were recorded in the project notebook. The QC verification check followed the 40 CFR 53.21(b) guidelines. These guidelines stated that a linear curve would be used to compare the target analyte instrument readings to the dilution calibrator output by plotting at least seven approximately equally spaced points, which included a setpoint at zero and a setpoint at 90 +/- 5 percent of the upper range limit of each analyzer. For the purposes of this project, the setpoints were set at 0%, 15%, 30%, 45%, 60%, 75%, and 90% of the upper range limit of each test analyzer. If the slope in this linear curve calculation was more than +/- 5% off a slope of 1.0, the slope was adjusted on the instrument accordingly. If the intercept was over +/- 0.5 ppb off zero, the intercept was adjusted on the instrument accordingly. These adjustments were made at the intervals stated above and all QC checks and adjustments made were documented in the laboratory notebook used for the research project, and the results were uploaded to the Teams site. 21 ------- 3.3 Data Quality Objectives The precision and accuracy goals for each critical measurement parameter were set based on the requirements needed to achieve the data quality objectives (DQO) of the experiment, the knowledge of and experience with sampling systems similar to the one being used in this experiment, and other similar research study parameters. The DQOs for each critical measurement parameter are listed in Table 6. If these were not met during any experiment, the settings for the critical measurement (e.g., the instrument slope or the instrument intercept) were adjusted using the QC verification check guidelines or the measurements were not reported as final data. It should be noted that the DQOs were met for the T700U photometer and MFCs during these experiments and no adjustments were performed until their yearly recertification by the Met Lab. Instances of a target analyte instrument not meeting the DQOs were observed during the QC verification checks, and any changes made were documented in the appropriate laboratory notebook. Data regarding these QC verification checks were uploaded to the Teams site. These data were reviewed by the project leads for both Jacobs and EPA. Table 6. DQOs for Critical Measurements Critical Measurement Measurement Device Accuracy/Precision Ozone Concentration T700U Photometer +/-1.1% from reading on Standard Reference Photometer; background zero of 1.0 ppb or less Other Analyte Concentration T700U Mass Flow Controllers +/- 0.25% from calibrated reference flow meter Instrument Slope Target analyte instruments +/- 5% from a slope of 1.00 during the QC Check Instrument Intercept Target analyte instruments +/-1.0 ppb from 0.0 ppb during the QC Check 22 ------- 4.0 Results and Discussion Each candidate test material was compared to the selected control material, CFR approved FEP Teflon®, by comparing target analyte concentrations when routing identified target gas analytes through both the control material and the test material. All data summaries are separated out into sections based on the target analyte of each experiment. Transport Efficiency was calculated using Equation 1 in Section 2.1, in which the final 1-minute average of the 5 second or 15 second data records of each analyzer concentration measurement through the test material (outlet) was divided by the final 1-minute average of the 5 second or 15 second data records of each analyzer concentration measurement through the control material (inlet) for each experiment. In the final data reported, the mean of all related experiments was used. "Adequate" is noted when transport efficiency results were reported as a mean efficiency of 97.5% and above. This value was based on the 2.5% threshold for O3 line loss tests as mentioned in Section 2.3 of this report (field SOP NPAP, EPA, 2011). The transport efficiencies and standard deviations for each gas type and replicate are listed in Appendix A of this report. For each of the following sections, per target analyte, each of the data tables has the same format: • The first column labeled "Tubing" represents the material sampled for the associated tests in the three columns to the right. • The next column, labeled "Cone, (ppb)" represents the specific target analyte concentration at which the transport efficiency tests were conducted. • Transport efficiency results are divided into two categories and are reported in the next two columns. One category/column is labeled "Unconditioned", in which the test material was not prepared in any way before testing, while the other category/column is "Conditioned", in which the test material was exposed to a high concentration of the target analyte for an hour prior to recording data. • The columns labeled "n", represent the number of replicate tests conducted at each concentration for each category. • Lastly, the "Mean Transport Eff (%)" columns represent the average of the "n" number of tests at each concentration for each category. 23 ------- 4.1 O3 Data Summary Table 7. O3 Transport Efficiency Results Tubing 03 Cone, (ppb) Unconditioned Conditioned n Mean Transport Eff. (%) n Mean Transport Eff. (%) 15 3 100.1 3 99.8 FEP Teflon® 35 3 98.0 3 99.8 85 3 97.6 3 99.2 98.6 99.6 15 3 94.9 3 100.5 PVDF 35 3 95.6 3 99.2 85 3 96.2 3 98.8 95.6 99.5 15 3 95.5 3 100.2 PTFE 35 3 97.3 3 98.6 85 3 98.2 3 99.7 97.0 99.5 15 3 96.5 3 98.6 PFA 35 3 97.2 3 99.0 85 3 96.8 3 99.5 96.8 99.0 15 3 -0.3 1 -0.2 SS316 35 3 -0.3 1 -0.4 85 3 -0.2 1 -0.1 -0.3 -0.2 15 3 -0.1 1 79.7 SNSS 35 3 0.2 1 89.6 85 3 27.6 1 93.8 9.2 87.7 15 3 4.1 3 0.9 Tygon® 35 3 9.0 3 0.1 85 3 20.2 3 0.1 11.1 0.4 24 ------- In their original conditions, all the polymer materials excluding FEP Teflon® did not have an adequate mean transport efficiency of above 97.5%. They were relatively close to the 97.5% threshold, excluding Tygon®, but none of them met the adequate threshold. Additionally, neither the SS316 nor the SNSS compared favorably to the FEP control material, with very low transport efficiencies of -0.3% and 9.2%, respectively. The negative transport efficiency values seen in Table 7 were due to the analyzer reading values close to the intercept for the instrument, and in these cases, readings would be very close to zero, with most averages being approximately -0.1 to -0.3 ppb. When polymer materials were conditioned, all excluding Tygon® improved in their mean transport efficiency results. Notably, this improvement meant that all the materials went from below the 97.5% transport efficiency threshold to well above it. This demonstrated that any of these fluoropolymer materials should be acceptable following their conditioned with O3 prior to use. However, for Tygon®, conditioning the material to O3 worsened its transport efficiency. For SS316, conditioning did not improve results, with its mean transport efficiency remaining around 0%. While the mean transport efficiency results for the SNSS greatly improved, the mean transport efficiency did not reach the threshold of 97.5% to be considered acceptable. 25 ------- 4.2 S02 Data Summary Table 8. SO2 Transport Efficiency Results Tubing S02 Cone, (ppb) Unconditioned Conditioned n Mean Transport Eff. (%) n Mean Transport Eff. (%) 15 3 102.0 3 99.5 FEP Teflon® 40 3 99.7 3 99.8 90 3 100.1 3 100.2 100.6 99.8 15 3 100.4 - - PVDF 40 3 99.6 - - 90 3 100.1 - - 100.0 - 15 3 100.8 - - PTFE 40 3 100.3 - - 90 3 100.2 - - 100.4 - 15 3 100.6 - - PFA 40 3 100.4 - - 90 3 100.6 - - 100.5 - 15 3 2.9 3 71.7 SS316 40 3 3.0 3 72.9 90 3 3.2 3 86.4 3.0 77.0 15 3 100.0 - - SNSS 40 3 100.0 - - 90 3 100.2 - - 100.1 - 15 3 88.4 3 87.7 Tygon® 40 3 88.6 3 87.8 90 3 89.6 3 87.6 88.9 87.7 It should be noted that for the experiments related to the target analyte SO2, conditioned tests were not 26 ------- conducted for certain test materials due to the high transport efficiencies observed in the tubing's unconditioned state. In their original conditions, all the polymer materials excluding Tygon® resulted in adequate mean transport efficiencies of above 97.5%. Specifically, Tygon® had a mean transport efficiency of 88.9%, which did not meet the 97.5% threshold. The SS316 did not compare favorably to the control material, with very low transport efficiencies of 3.0%. However, the SNSS performed very well with a mean transport efficiency of 100.1%. The only materials that were tested after being conditioned with SO2 were FEP Teflon®, SS316, and Tygon®, because the other test materials already had mean transport efficiencies above 97.5%. The FEP Teflon® was conditioned and tested as a control condition and performed very well with a mean transport efficiency of 99.8%. Conditioning SS316 greatly improved its mean transport efficiency, raising it to 77.0%>, but this result still did not meet the 97.5% acceptance criterion for acceptability. Conditioning had little to no effect on Tygon®, as the mean transport efficiency after conditioning was 87.7%), which was 1.2% less than the previous result. 27 ------- 4.3 N02 Data Summary Table 9. NO2 Transport Efficiency Results Tubing NOz Cone, (ppb) Unconditioned Conditioned n Mean Transport Eff. (%) n Mean Transport Eff. (%) 20 4 96.6 3 99.8 FEP Teflon® 50 3 99.7 - - 120 3 99.8 - - 98.7 99.8 20 4 99.5 - - PVDF 50 3 99.5 - - 120 3 99.2 - - 99.4 - 20 4 99.5 - - PTFE 50 3 99.1 - - 120 3 99.0 - - 99.2 - 20 4 99.6 - - PFA 50 3 99.0 - - 120 3 98.9 - - 99.1 - 20 4 -1.0 3 40.1 SS316 50 3 -0.7 - - 120 3 -0.2 - - -0.7 40.1 20 3 94.0 3 97.8 SNSS 50 3 95.4 3 98.6 120 3 97.3 3 99.0 95.6 98.5 20 3 79.9 3 85.1 Tygon® 50 3 79.1 - - 120 3 80.8 - - 79.9 85.1 It should be noted that for the experiments related to the target analyte NO2, conditioned tests were not 28 ------- conducted for certain test materials due to the high transport efficiencies observed in the tubing's unconditioned state. In their original conditions, all the polymer materials excluding Tygon® resulted in adequate mean transport efficiencies of above 97.5%. Specifically, Tygon® had a mean transport efficiency of 79.9%, which did not meet the 97.5% threshold. The SS316 did not compare favorably to the control material, with very low transport efficiencies of 3.0%. The SNSS had a mean transport efficiency of 95.6%, close but still below the 97.5% threshold to be considered acceptable. The only materials that were tested after being conditioned were FEP Teflon®, SS316, Tygon®, and SNSS. Additionally, the tests for FEP Teflon®, SS316, and Tygon® were only conducted at 20 ppb NO2. The SNSS was tested at all concentrations since its original condition mean transport efficiency was close to the 97.5% threshold. The FEP Teflon® was conditioned and tested as a control condition, and it performed very well with a mean transport efficiency of 99.8%. Conditioning SS316 noticeably improved its mean transport efficiency to 40.1%, but this efficiency still did not meet the threshold of 97.5% to be considered acceptable. Conditioning had little to no effect on Tygon®, as the mean transport efficiency after conditioning was 85.1%>, only 5.2% better than the previous result. Lastly, conditioning improved the mean transport efficiency of the SNSS, which had a mean transport efficiency of 98.5%, which surpassed the 97.5% threshold to be considered acceptable. 29 ------- 4.4 CO Data Summary Table 10. CO Transport Efficiency Results Tubing CO Cone, (ppm) Unconditioned Conditioned n Mean Transport Eff. (%) n Mean Transport Eff. (%) 1.8 3 100.0 - - FEP Teflon® 4.5 3 100.3 - - 10.8 3 100.1 - - 100.1 - 1.8 3 100.6 - - PVDF 4.5 3 100.5 - - 10.8 3 100.3 - - 100.5 - 1.8 3 99.0 - - PTFE 4.5 3 99.7 - - 10.8 3 100.2 - - 99.6 - 1.8 3 100.9 - - PFA 4.5 3 100.3 - - 10.8 3 99.6 - - 100.3 - 1.8 3 98.7 - - SS316 4.5 3 100.3 - - 10.8 3 99.6 - - 99.5 - 1.8 3 100.5 - - SNSS 4.5 3 99.9 - - 10.8 3 99.8 - - 100.1 - 1.8 3 100.9 - - Tygon® 4.5 3 100.3 - - 10.8 3 100.1 - - 100.4 - It should be noted that for the experiments related to the target analyte CO, no conditioned tests were 30 ------- conducted, therefore no data results were collected. In their original conditions, all the materials resulted in adequate mean transport efficiencies of above 97.5%. In fact, all materials had between 99% and 100% mean transport efficiency, confirming that CO transport efficiency was acceptable regardless of the material that it traveled through. None of the materials were conditioned and retested because all materials displayed acceptable transport efficiencies during the initial experiments. 4.5 Statistical Analysis Transport efficiencies were compared using Excel's Data Analysis tool pack for ease of communication and transparency. Three statistical tests, Analysis of variance (ANOVA) Single Factor, ANOVA Single Factor with Ad hoc testing and ANOVA Two Factor with replicates, were performed upon multiple arrangements of the transport efficiencies, replicates, and gas concentrations to calculate the statistical evidence that the results are repeatable. The single factor ANOVA was chosen to test the means between two or more groups of values. The two factor ANOVA with replicates analysis was chosen to evaluate the difference between the means of more than two groups and to determine the interaction between both the replicates and groups. Table 11 provides a brief overview of the statistical questions, the tests performed, if ad hoc analysis was performed, and the name of the respective data table. Table 11. Summary of Statistical Testing Question Test performed Ad hoc Analysis Performed Table Name and Document Reference Does concentration (20%, 50% and 120% of NAAQS value) change the transport efficiency for a certain gas vs the tubing type? ANOVA-Single Factor Not performed Transport Efficiency vs NAAQS Level testing (Table 13) Does Tubing Material Matter vs Particular Gas? ANOVA-Single Factor Performed, Tukey ANOVA Single Factor, with ad hoc Results - Tubing vs Gas type, unconditioned tubing Table 15) Are there any statistical differences between the groups of data? ANOVA - Two Factor with Replicates Not performed ANOVA - Two Way with Replication Results for unconditioned tubing (p- value) Table 17) Will conditioning the fluoropolymer tubing with Ozone increase the transport efficiency to match the transport efficiencies of the remaining three gases? ANOVA - Two Factor with Replicates Not performed ANOVA - Two Way with Replication Results (p- values) for unconditioned tubing (S02, N02 and CO) vs conditioned Tubing (03) (Table 18) Table 11 provides a brief overview of the three statistical testing methods used to support the overall observations provided in the aforementioned data summaries. To review the statistical testing methods, Excel bases the null hypothesis of the single factor ANOVA 31 ------- test on whether the mean is the same for all groups or the columns of data. Calculated p-values are compared to the requested alpha value, which is 0.05 for all ANOVA testing in this report. If a p-Value is greater than 0.05, we cannot conclude that a significant difference exists. The Tukey statistical ad hoc analysis uses a Critical Value of Studentized Range Distribution or Q-value combined with the number of treatments, replicates number and calculated standard error values to determine the statistical differences of the data means. Table 12 provides an example of how the transport efficiencies were reorganized from Table 7 to execute the ANOVA Single Factor analysis. The analysis was performed to determine if the concentration value changed transport efficiencies vs tubing type. Table 12. Data reorganization Transport Efficiency vs NAAQS Level Data corresponding to question 1 in Table 11 for Ozone 15ppb Ozone 35ppb Ozone 85ppb Ozone FEP 100.05% 98.01% 97.64% PTFE 95.52% 97.31% 98.15% PFA 96.53% 97.20% 96.80% PVDF 94.92% 95.62% 96.15% Tygon® 4.14% 9.01% 20.18% SS316 -0.07% -0.32% -0.14% SNSS -0.08% 0.17% 27.56% Table 12 displays the transport efficiency data arrangement to compare the three NAAQS concentration values against the type of tubing. Transport efficiency data for unconditioned data can be found in Appendix A Table 1, Appendix A Table 3, Appendix A Table 5 and Appendix A Table 7. Table 13. Results Transportation Efficiencies of Tubing vs NAAQS Level testing Transport Efficiency of Tubing vs NAAQS Level testing Gas p-value Ozone 0.96412 Sulfur Dioxide 0.99983 Nitrogen Dioxide 0.99951 Carbon Monoxide 0.78306 Table 13 displays the p-values for single factor ANOVA test comparing the mean transport efficiencies of the different concentrations for the four NAAQS gases. The p-values for the four gases are greater than the alpha level of 0.05, indicating that there aren't statistical differences. In other words, the transport efficiencies are not dependent on the tested NAAQS concentrations. To provide insight into the transport efficiency interaction between each particular gas and tubing material, a single factor ANOVA test with a Tukey ad hoc analysis was performed. Table 14 is an example of the data reorganization of the data sets for the unconditioned mean transport efficiencies. 32 ------- Table 14. Data reorganization tubing type vs pollutant NAAQS level Ozone -FEP Sulfur Dioxide - FEP Nitrogen Dioxide - FEP Carbon Monoxide - FEP 20% 100.05% 102.00% 99.61% 100.01% 50% 98.01% 99.66% 99.72% 100.27% 120% 97.64% 100.10% 99.79% 100.05% Table 15. Results from ANOVA single factor test of tubing type vs pollutant ANOVA Single Factor, with ad hoc Results - Tubing vs Gas type, unconditioned tubing P-values of Statistical Difference of Transport Efficiencies Between Gases Statistical Testing Tubing vs Gas Tubing Type 03 S02 N02 CO FEP 0.112357 Not Different Not Different Not Different Not Different Different from S02, PTFE 1.02E-06 N02 and CO Not Different Not Different Not Different Different from S02, PFA 3.19E-03 N02 and CO Not Different Not Different Not Different Different from S02, PVDF 1.02E-06 N02 and CO Not Different Different from CO Different from N02 Different from S02, Different from CO Different from S02, Tygon® 6.97E-17 N02 and CO Different from CO and 03 N02 and 03 Different from S02 Different from CO Different from S02 Different from S02, SS316 1.41E-06 and CO and 03 and CO N02 and 03 Different from S02, SNSS 1.90E-08 N02 and CO Not Different Not Different Not Different Table 15 displays the p-values from the ANOVA single factor test comparing the individual tubing versus the four target analyte gases. The p-values listed in Table 15 indicate that the transport efficiencies are statistically different across all four gases for the tubing materials tested besides FEP. Results of ad hoc analysis via Tukey analysis indicate that O3 is the most reactive gas and transport efficiencies statistically differ from the majority of other gases. The remaining gases, SO2, NO2 and CO were generally not as reactive with the fluoropolymer tubing; except for PVDF. As previously mentioned, transport efficiencies for Tygon® and Stainless Steel only reached the 97.5% threshold for CO. The ad hoc analysis statistically confirmed the transport efficiency observations reported. To provide insight into whether there was an interaction between the concentration groups and the replicates, an ANOVA two-way statistical test with replicates was performed. Briefly, the two-way ANOVA tests in excel are based upon evaluating three null hypotheses. The three hypotheses of concern are as follows. Hypothesis 1: the means of observations grouped by one factor are the same (i.e., no difference between in the 20%, 50% and 120% groups [each group is the testing replicates]) Hypothesis 2: the means of observations grouped by the other factor are the same (i.e., no difference between the gases) 33 ------- Hypothesis 3: there is no interaction between the two factors (i.e., no difference between the 20%, 50% and 120% vs the gases). Table 16 displays how the data was reorganized for the ANOVA two-way testing, while Table 17 provides the p-values. Table 16. Data reorganized for ANOVA two-way testing Sulfur Nitrogen Carbon Dioxide - Dioxide - Monoxide - NAAQS Ozone -FEP FEP FEP FEP 20% 98.31% 103.97% 99.75% 99.48% 20% 100.29% 102.80% 99.48% 99.83% 20% 101.55% 99.23% 99.48% 100.72% 50% 97.24% 98.91% 99.70% 100.19% 50% 96.72% 100.02% 99.75% 100.39% 50% 100.07% 100.04% 99.70% 100.23% 120% 97.96% 100.68% 99.65% 100.11% 120% 95.90% 99.67% 99.82% 100.26% 120% 99.07% 99.96% 99.90% 99.78% Table 17. P-values of the ANOVA two-way testing (unconditioned transport efficiencies) ANOVA - Two Way with Replication Results for unconditioned tubing (p-value) FEP PTFE PFA PVDF Tygon® SS316 SNSS NAAQS value comparison 0.06630384 0.082876 0.910589 0.974245 0.000907 0.04358 3.3E-08 Gases Comparison 0.00727193 6.8E-09 6.21E-05 6.03E-09 1.34E-29 8.91E-49 3.43E-31 Interaction 0.2369388 0.017502 0.783657 0.837164 0.000116 0.089765 3.47E-10 Table 17 contains the p-values from the ANOVA two-way testing with unconditioned transport efficiencies. The p-values greater than 0.05 were shaded in a light gray color while p-values less than 0.05 were shaded in light blue. (The color shading is only provided to aid in the viewing of the data.) The fluoropolymer tubing materials have p-values greater than 0.05, indicating that there aren't significant differences between the NAAQS concentrations groups. Tygon®, SS316 and SNSS have p- values less than 0.05 indicating there are significant differences. All tubing materials have p-values less than 0.05 for the gas comparison portion, indicating that there are significant differences between the gases. This provides further statistical proof that O3 is the most reactive of the gases as the averages were lower for each tubing material result. The interaction p-values were less than 0.05 for PTFE, Tygon® and SNSS, and therefore indicate that the transport efficiency is dependent on NAAQS level concentrations. Additionally, this indicates that the effect of the efficiencies at the different NAAQS levels is dependent on gas type. 34 ------- Alternatively, the interaction between p-values greater than 0.05 for FEP, PFA, PVDF and SS316 indicates that the effect of different gas on the transport efficiency is not dependent on the concentration. The transport efficiencies for O3 were statistically lower than the other gases as seen in Tables 16 and 17. Only FEP tubing achieved the 97.5% transport efficiency threshold for O3, and experiments were performed with conditioned tubing to investigate if conditioning tubing would increase the transport efficiencies over the previously mentioned threshold. In an effort to compare if conditioning would increase the transportation efficiencies to the bulk average of unconditioned transport efficiencies, the ANOVA two-way test was reperformed with the conditioned O3 values. Table 18. Results from the ANOVA two-way testing with conditioned O3 transport efficiencies. ANOVA - Two Way with Replication Results (p- values) for unconditioned tubing (S02, N02 and CO) vs conditioned Tubing (03) FEP PTFE PFA PVDF NAAQS value comparison Gases Comparison Interaction 0.30793611 0.264976 0.975098 0.188877 0.13553282 0.004551 0.022389 0.085892 0.23736644 0.096694 0.33202 0.730326 Table 18 contains the p-values from the ANOVA two-way testing with both unconditioned transport efficiencies and conditioned transport efficiencies. Results indicate that conditioned fluoropolymer tubing with O3 increased the transport efficiencies to be statistically similar to the remaining unconditioned tubing. Although the p-values remained under 0.05 for PTFE and PFA, the p-values greatly increased from the previous test results (in Table 17) providing further support to confirm the positive aspect of the conditioning process. 35 ------- 5.0 Summary and Conclusions Ambient air monitoring sites operating under the guidelines given in Section 9 of Appendix E of 40 CFR 58 are currently required to use FEP Teflon® or borosilicate glass for their sampling probes and sample lines. The requirement to use these materials was made to ensure that during testing, the concentration of the criteria pollutants at the inlet of each analyzer was as close as possible to the ambient air's concentration. While research has been conducted in the past regarding other materials that might be considered equivalent, recent advances in the production process of tubing materials and the development of the Silconert® non-reactive coating for stainless steel material provided an opportunity to further investigate potentially acceptable alternative materials. A review of pertinent literature was conducted to determine what tubing materials would be the best potential candidates for the equivalence tests. Through this review, several commercially available tubing materials were identified, which included several polymer materials, stainless steel with the Silconert® non-reactive coating (SNSS), and untreated grade 316 stainless steel (SS316). The pollutant types to be tested were the four gaseous criteria pollutants: O3, NO2, SO2, and CO. Following the literature review, an experimental test setup was designed to determine the fractional transport efficiency as a function of material type. During each experiment, known concentrations of criteria pollutant were generated by a dilution calibrator and delivered to the FRM or FEM instrument through the control material and through the test material for an equal amount of time. The control material was a 5-foot section of FEP and the length of the candidate tubing material was adjusted to provide a 20 second residence time to correspond with Section 9 of Appendix E. The lengths of candidate tubing varied as a function of actual tubing ID and the operating flow rate of the FRM or FEM analyzers used during the tests. While the pollutant generator delivered each criteria pollutant for 40 minutes, valves automatically switched between the control and candidate materials every 5 minutes. The fractional transport efficiency was calculated by dividing the final minute average of the output concentration by the final minute average of the input concentration. For the laboratory tests, all candidate tubing types had a Vi" OD and a nominal 3/16 ID. To determine the influence of concentration on transport efficiency, experiments were conducted for each material at 20%, 50%, and 120% of the NAAQS value for each criteria pollutant being tested. The repeatability of the test results was determined by conducting three replicate tests under identical test conditions. This procedure thus resulted in nine separate experiments for each candidate test material for each of the three criteria pollutant concentrations. Each of these replicate tests was conducted using a new, unconditioned section of candidate tubing. Research regarding increase of transport efficiency through sections of repeated use tubing until that material's surface became physically and chemically conditioned was conducted in the past (Altshuller et al. (1961), Scholz and Wallner (2015)). This need for conditioning of the tubing's surface was demonstrated for very reactive pollutants such as O3. As a result, during this study, additional tests were conducted in several cases where previously unused tubing was conditioned to a concentration of 450 ppb of the criteria pollutant for 1 hour before being tested at the three concentrations. This procedure was adopted from the O3 line loss test procedure in the field SOP Through-the-Probe NPAP (EPA, 2011). Section 9 of Appendix E of 40 CFR 58 indicates that alternate tubing materials can be used if they are deemed equivalent by EPA. However, the exact definition of what constitutes "equivalent" was never specified in the CFR. Therefore, ORD adopted NPAP's O3 line loss acceptance criteria of 97.5%, or a pollutant loss of 2.5%, as the acceptance criteria for an equivalent candidate material in the CFR. 36 ------- As summarized in Section 4.0, this study confirmed the FEP Teflon® was a highly efficient material for transporting all four criteria pollutants, regardless of pre-conditioning. Specifically, for the most reactive gas, O3, its mean transport efficiency was 98.4% in its unconditioned state, which represents a 0.9% efficiency above the 97.5% acceptance threshold. Comparatively, the transport efficiency of O3 through the other fluoropolymers (PVDF, PTFE, PFA) was between 95.6% and 97.0%, meaning they were all slightly below the 97.5% acceptance threshold. However, when any of the fluoropolymers was conditioned to O3, each material's individual transport efficiency of O3 increased, with all results observed between 99.0% and 99.6%. The repeatability of the three replicates tests for each of these materials was very favorable and the transport efficiency was found to be independent of the concentration being tested. Additionally, the three fluoropolymer test materials and FEP Teflon® all had excellent transport efficiency for SO2, NO2, and CO in their unconditioned states. The repeatability for these pollutant tests was very favorable, and the transport efficiency was once again independent of the concentration being tested, similar to the O3 test results. For O3, unconditioned SNSS displayed a mean transport efficiency of 9.2%. While its transport efficiency improved to 87.7% when conditioned to O3, the results were still below the 97.5% acceptance criteria. For NO2, SNSS had a mean transport efficiency of 95.6% in its unconditioned state. This improved to 98.5% when conditioned with 450 ppb NO2 for one hour, raising its results to within the acceptance criterion. The repeatability of these tests was favorable, and the results were determined to be independent of the concentration being tested. Additionally, SNSS performed well for SO2 and CO, with mean transport efficiencies with unconditioned tubing of 100.1% for both gases. SS316 and Tygon® performed poorly for O3, NO2, and SO2, with all associated transport efficiencies falling well below the 97.5% threshold, regardless of the conditioning of the material. However, both materials resulted in a high transport efficiency of CO, independent of the concentration being tested. Based on these test results, a revision is proposed for Section 9 of Appendix E of 40 CFR 58 to include PVDF, PTFE, and PFA in the list of approved materials for use in sample probes and sample lines, as they have successfully demonstrated their ability to effectively transport gaseous criteria pollutants. The recommendation to condition all sample probe materials using O3 will remain for all materials. However, even with conditioning, SS316SS, SNSS, and Tygon® are still considered to be unacceptable materials for efficient sampling of the criteria gaseous pollutants. A summary of the transport efficiency test results for the four approved fluoropolymers is provided in Table 19. Reported values for each pollutant and tubing material represent the mean of three replicate tests conducted using the three pollutant concentrations. All tests were conducted using 20 second residence times. For conditioned O3 and unconditioned NO2, SO2, and CO, mean transport efficiencies for FEP Teflon®, PVDF, PTFE, and PFA were measured to be 99.8%, 99.9%, 99.1%, and 99.1%, respectively. Table 19. Summary of Transport Efficiencies for the Four Fluoropolymer Tubing Materials Tubing Material Mean Trans Mil l Efficiency for 20 Second Residence Time (%) 03 (Unconditioned) 03 (Conditioned) so2 (Unconditioned) no2 (Unconditioned) CO (Unconditioned) Mean of the Four Gases FEP Teflon® 98.6 99.6 100.6 98.7 100.1 99.8 PVDF (Kvnar) 95.6 99.5 100.0 99.4 100.5 99.9 PTFE 97.0 99.5 100.4 99.2 99.6 99.7 PFA 96.8 99.0 100.5 99.1 100.3 99.7 Mean of Four Fluoropolymers 97.0 99.4 100.4 99.1 100.1 CV (%) 1.27 0.27 0.26 0.30 0.39 37 ------- Selection of which fluoropolymer to incorporate into a given NAAQS monitoring station is largely a subjective decision. As previously mentioned and documented in Table 19, the transfer efficiencies of the gaseous pollutants through each fluoropolymer types are statistically identical independent of pollutant type and pollutant concentration. All four fluoropolymer types display high resistance to UV light and are thus suitable for indoor and outdoor use. FEP Teflon® and PFA have one significant advantage over PVDF and PTFE in that they are transparent and thus provide a better visual indication when they've become internally contaminated and require replacement. Table 20. Properties of Fluoropolymer Tubing Materials Tubing Material Trade Name & Owner 3/16" ID Cost/ft UV Resistance Optical Properties FEP Teflon® FEP (DuPont) Neoflon (Daikin) Hyflon (Solvay Specialty Polymers) $3.00 High Transparent PVDF Kynar (Arkema, Inc.) Solef (Solvay Specialty Polymers) Hylar (Solvay Specialty Polymers) $2.13 High Opaque PTFE Teflon® (DuPont) Algoflon (Solvay Specialty Polymers) Polyflon (Daikan) $1.85 High Opaque PFA Fluon PFA (AGC Chemicals) $3.17 High Transparent Note: Costs accessed July 2022 from https://mcmaster.com All fluoropolymer materials are inherently flexible, although overall flexibility and strength are a function of each tubing's wall thickness. In some cases, monitoring organizations have reported crimping problems with 3/16" ID tubing and therefore have adopted 1/8" ID tubing to minimize crimping. Because an equivalent length of 1/8" ID tubing has one-half the surface area and 45% of the pollutant residence time as that of 3/16" ID tubing at the same flow rate, the loss of pollutants through 1/8" ID tubing would be predicted to be even lower than that of 3/16" ID tubing. Due to the greater amount of material required to fabricate 1/8" ID tubing, however, its cost is approximately twice that of 38 ------- the cost listed in Table 20 for 3/16" ID tubing. It's also important to ensure that flow systems can provide the proper flow rate given the 1/8" ID tubing's higher pressure drop. This report has identified and documented three additional fluoropolymers (PVDF, PTFE, and PFA) which have statistically identical pollutant transport performance as that of FEP Teflon® and thus meet the regulatory requirements for efficient NAAQS pollutant sampling. It was also determined that sampling lines intended for O3 sampling require conditioning prior to use. No conditioning is required for sampling lines intended for sampling of SO2, NO2 and CO. EPA intends to revise Section 9 of Appendix E of 40 CFR Part 58 to include these three fluoropolymer materials. In conjunction with the previously approved borosilicate glass and FEP Teflon® materials, including PVDF, PTFE, and PFA will provide monitoring organizations with a larger variety of efficient sampling and transport materials with which to meet their Congressionally mandated obligations to conduct accurate NAAQS compliance monitoring. 39 ------- 6.0 References Altshuller, A. P., Wartburg, A., Taft, R.A. (1961). The interaction of ozone with plastic and metallic materials in a dynamic flow system. International Journal of Air and Water Pollution, 4, 70-78. Byers, R.L., Davis, J.W. (1970) Sulfur Dioxide Adsorption and Desorption on Various Filter Media, J. Air Pollution Control Assoc., Vol. 20, pp. 236-238. Campos, V. P., Cruz, L. P. S., Godoi, R. H. M., Godoi, A. F. L., Tavares, T. M. (2020). Development and validation of passive samplers for atmospheric monitoring of SO2, NO2, O3, and H2S in tropical areas. Microchemical Journal, 96, 132-138. https://doi.Org/10.1016/i.microc.2010.02.015 Code of Federal Regulations. (2018). Probe and monitoring path siting criteria for ambient air quality monitoring, 40 CFR 58, Appendix E, § 9. Hughes, E.E. (1975). Development of standard reference materials for air quality measurement. ISA Transaction, 14(4), 281-291. Marshik, B. (2015, October 19-23). Material and process conditions for successful use of extractive sampling techniques. EPA Region 6 - 25th Annual Quality Assurance Conference, Dallas, TX, United States. Obermiller, E. L., Charlier, (1968). Gas chromatographic separation of nitrogen, oxygen, argon, carbon monoxide, carbon dioxide, hydrogen sulfide, and sulfur dioxide. Journal of Gas Chromatography, 6, 446-447. https://doi.Org/10.1093/chromsci/6.8.446 Palmes, E. D., Gunnison, A. F., Dimattio, J., & Tomczyk, C. (1976). Personal sampler for nitrogen dioxide. American Industrial Hygiene Association Journal, 37(10), 570-577. https://doi.org/10.1080/00028897685Q7522 Scholz, A (2015) "The Influence of Sample Tubing Material on Accuracy of Low Ozone Concentration and Measurement", IOA-EA3G Conference, Barcelona, Spain 28 June - 3 July, 2015 40 ------- SilcoTek. (2020). SilcoNert® 2000 Coating Data Sheet. https://www.silcotek.com/hubfs/Literature%20Cataloe/Data%20Sheets/DATA-SILCQNERT2.pdf US EPA (2012) "Traceability Protocol for Assay and Certification of Gaseous Calibration Standards" (PDF) (174 pp, 1.7 M, About PDF) Publication No. EPA/600/R-12/53, available at https://www.epa.gov/air- research/epa-traceability-protocol-assay-and-certification-gaseous-calibration-standards US EPA (2011) "Field Standard Operating Procedures for the EPA Through-the Probe National Performance Audit Program" available via https://www.epa.gov/sites/default/files/2020- 10/documents/20110719npapttpdraftoperatorsfieldsop O.pdf Wechter, S. (1976). Preparation of stable pollution gas standards using treated aluminum cylinders. STP598-EB Calibration in Air Monitoring, 40-54. https://doi.org/10.1520/STP32353S Wohlers, H. C., Newstein, H., Daunis, D. (1967). Carbon monoxide and sulfur dioxide adsorption on — and desorption from glass, plastic, and metal tubings. Journal of the Air Pollution Control Association, 17(11), 753-756. https://doi.org/lQ.lQ8Q/00022470.1967.10469Q68 41 ------- Appendix A Transport efficiencies and standard deviations for each tubing type and gas concentration. Appendix A Table 1. Ozone (Unconditioned) Ozone 15ppb Repl Rep2 Rep3 Tubing Transport Standard Transport Standard Transport Standard Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 98.31% 0.67% 100.29% 0.41% 101.55% 0.43% PTFE 94.77% 0.25% 96.56% 0.25% 95.24% 0.39% PFA 93.88% 0.00% 99.49% 0.55% 96.22% 0.39% PVDF 93.58% 0.42% 96.06% 0.31% 95.12% 0.42% Tygon® 0.12% 0.27% 8.18% 0.03% 4.12% 0.01% SS316 0.58% 0.27% -0.28% -0.35% -0.50% -0.30% SNSS -0.23% -0.33% 0.67% 0.00% -0.69% 0.00% Ozone 35ppb Repl Rep2 Rep3 Tubing Transport Standard Transport Standard Transport Standard Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 97.24% 0.27% 96.72% 0.34% 100.07% 0.57% PTFE 96.56% 0.41% 97.21% 0.19% 98.15% 0.27% PFA 95.43% 0.27% 98.44% 0.17% 97.71% 0.26% PVDF 93.09% 0.27% 95.13% 0.36% 98.65% 0.25% Tygon® 1.10% 0.13% 14.17% 0.02% 11.77% 0.12% SS316 -0.58% 0.00% 0.00% 0.00% -0.37% -0.13% SNSS -0.36% -0.13% 0.58% 0.00% 0.29% 0.00% Ozone 85ppb Repl Rep2 Rep3 Tubing Transport Standard Transport Standard Transport Standard Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 97.96% 0.30% 95.90% 0.16% 99.07% 0.14% PTFE 97.34% 0.19% 98.35% 0.16% 98.77% 0.11% PFA 95.44% 0.14% 97.55% 0.14% 97.42% 0.22% PVDF 95.76% 0.07% 96.38% 0.11% 96.30% 0.10% Tygon® 19.44% 0.02% 21.55% 0.05% 19.56% 0.01% SS316 -0.24% 0.00% -0.12% 0.00% -0.05% -0.06% SNSS 31.76% 1.20% 18.83% 1.04% 32.10% 1.08% 42 ------- Appendix A Table 2. Ozone (Conditioned) Ozone 15ppb Conditioned Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 97.89% 0.38% 100.23% 0.38% 101.29% 0.54% PTFE 99.94% 0.34% 100.00% 0.58% 100.64% 0.46% PFA 99.22% 0.48% 98.32% 0.35% 98.24% 0.44% PVDF 98.16% 0.43% 101.69% 0.43% 101.72% 0.44% Tygon® 1.03% 0.36% 0.22% 0.33% 1.48% 0.27% SS316 -0.05% -0.06% N/A N/A N/A N/A SNSS 79.66% 0.49% N/A N/A N/A N/A Ozone 35ppb Conditioned Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 100.31% 0.25% 100.05% 0.26% 99.01% 0.21% PTFE 99.19% 0.21% 98.60% 0.11% 98.13% 0.22% PFA 98.97% 0.19% 99.23% 0.20% 98.71% 0.18% PVDF 98.96% 0.49% 98.82% 0.19% 99.73% 0.23% Tygon® -0.27% -0.09% 0.00% 0.00% 0.47% 0.15% SS316 -0.36% -0.13% N/A N/A N/A N/A SNSS 89.57% 0.17% N/A N/A N/A N/A Ozone 85ppb Conditioned Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 98.78% 0.09% 99.15% 0.17% 99.59% 0.09% PTFE 99.59% 0.14% 100.31% 0.18% 99.04% 0.10% PFA 99.50% 0.07% 99.51% 0.16% 99.59% 0.14% PVDF 98.30% 0.11% 99.41% 0.11% 98.59% 0.17% Tygon® -0.19% -0.06% 0.12% 0.00% 0.47% 0.06% SS316 -0.17% -0.31% N/A N/A N/A N/A SNSS 93.84% 0.17% N/A N/A N/A N/A \Tote that blank values indicate that test was not performed due to lack in of tubing availability (SNSS) or lack of efficiency increase due to Conditioning (SS316). 43 ------- Appendix A Table 3. Sulfur Dioxide (Unconditioned) Sulfur 15ppb Dioxide Repl Rep2 Rep3 Tubing Type Transport Efficiency Standard Deviation Transport Efficiency Standard Deviation Transport Efficiency Standard Deviation FEP 103.97% 1.08% 102.80% 0.51% 99.23% 0.49% PTFE 101.12% 0.63% 101.84% 1.14% 99.37% 1.07% PFA 101.44% 0.91% 96.01% 0.51% 100.59% 0.61% PVDF 98.95% 0.74% 102.92% 0.49% 99.42% 0.77% Tygon® 88.21% 0.93% 86.14% 0.93% 90.98% 0.71% SS316 2.96% 0.01% 3.33% 0.39% 2.29% 0.28% SNSS 98.68% 0.98% 100.24% 0.77% 101.20% 0.47% Sulfur 40ppb Dioxide Repl Rep2 Rep3 Tubing Type Transport Efficiency Standard Deviation Transport Efficiency Standard Deviation Transport Efficiency Standard Deviation FEP 98.91% 0.35% 100.02% 0.29% 100.04% 0.37% PTFE 99.51% 0.34% 100.00% 0.19% 101.48% 0.43% PFA 100.68% 0.50% 100.83% 0.23% 99.66% 0.25% PVDF 99.39% 0.43% 99.70% 0.29% 99.83% 0.56% Tygon® 89.70% 0.48% 88.35% 0.78% 87.87% 0.21% SS316 3.07% 36.10% 2.95% 0.14% 3.07% 0.12% SNSS 99.76% 0.39% 100.36% 0.35% 99.96% 0.46% Sulfur 90ppb Dioxide Repl Rep2 Rep3 Tubing Type Transport Efficiency Standard Deviation Transport Efficiency Standard Deviation Transport Efficiency Standard Deviation FEP 100.68% 0.31% 99.67% 0.17% 99.96% 0.18% PTFE 100.10% 0.44% 100.42% 0.15% 100.15% 0.24% PFA 101.27% 0.28% 100.86% 0.17% 99.64% 0.34% PVDF 100.54% 0.16% 99.86% 0.22% 99.75% 0.13% Tygon® 90.52% 0.39% 89.86% 0.27% 88.33% 0.24% SS316 3.16% 0.01% 3.24% 0.05% 3.05% 0.00% SNSS 100.46% 0.28% 100.37% 0.21% 99.69% 0.18% 44 ------- Appendix A Table 4. Sulfur Dioxide (Conditioned) Sulfur 15ppb Dioxide (Conditioned) Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 99.64% 0.62% 100.82% 0.49% 98.09% 0.42% PTFE N/A N/A N/A N/A N/A N/A PFA N/A N/A N/A N/A N/A N/A PVDF N/A N/A N/A N/A N/A N/A Tygon® 87.90% 1.13% 85.46% 0.68% 87.90% 0.61% SS316 79.26% 0.53% 63.46% 1.98% 72.34% 0.97% SNSS N/A N/A N/A N/A N/A N/A Sulfur 40ppb Dioxide (Conditioned) Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 100.47% 0.22% 99.68% 0.40% 99.14% 0.43% PTFE N/A N/A N/A N/A N/A N/A PFA N/A N/A N/A N/A N/A N/A PVDF N/A N/A N/A N/A N/A N/A Tygon® 88.30% 0.44% 87.13% 0.41% 87.94% 0.36% SS316 73.05% 1.14% 72.29% 0.59% 73.22% 0.56% SNSS N/A N/A N/A N/A N/A N/A Sulfur 90ppb Dioxide (Conditioned) Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 99.36% 0.33% 100.27% 0.15% 100.90% 0.37% PTFE N/A N/A N/A N/A N/A N/A PFA N/A N/A N/A N/A N/A N/A PVDF N/A N/A N/A N/A N/A N/A Tygon® 88.22% 0.18% 87.70% 0.21% 87.01% 0.23% SS316 85.91% 0.50% 87.09% 0.34% 86.16% 0.44% SNSS N/A N/A N/A N/A N/A N/A Note that blank values indicate that test was not performed due to original set of data reaching the 97.5% transport efficiency threshold. 45 ------- Appendix A Table 5. Nitrogen Dioxide (Unconditioned) Nitrogen 20ppb Dioxide Verification Repl Rep2 Rep3 Tubing Type Transport Standard Efficiency Deviation Transport Standard Efficiency Deviation Transport Efficiency Standard Deviation Transport Efficiency Standard Deviation FEP 99.75% 0.29% 99.48% 0.00% 99.48% 0.00% 99.74% 0.31% PTFE 99.50% 0.00% 99.47% 0.00% 99.47% 0.00% 99.47% 0.00% PFA 99.50% 0.00% 99.48% 0.00% 99.74% 0.31% 99.47% 0.00% PVDF 99.38% 0.25% 99.61% 0.40% 99.48% 0.00% 99.47% 0.00% Tygon® 78.90% 0.31% 80.80% 0.23% 80.66% 0.38% 79.35% 0.33% SS316 -1.51% -0.02% -0.52% 0.00% -1.04% 0.00% -1.04% 0.00% SNSS 96.77% 0.23% 92.39% 0.24% 92.85% 0.36% Nitrogen 50ppb Dioxide N/A Repl Rep2 Rep3 Tubing Type Transport Standard Efficiency Deviation Transport Efficiency Standard Deviation Transport Efficiency Standard Deviation FEP N/A N/A 99.70% 0.14% 99.75% 0.15% 99.70% 0.12% PTFE N/A N/A 98.90% 0.12% 99.40% 0.00% 98.85% 0.10% PFA N/A N/A 99.15% 0.10% 98.99% 0.00% 98.79% 0.00% PVDF N/A N/A 99.50% 0.14% 99.45% 0.15% 99.65% 0.15% Tygon® N/A N/A 79.80% 0.09% 79.12% 0.08% 78.36% 0.00% SS316 N/A N/A -0.61% 0.00% -0.75% -0.10% -0.81% 0.00% SNSS N/A N/A 94.82% 0.16% 95.92% 0.15% 95.49% 0.11% Nitrogen Dioxide N/A Repl Rep2 Rep3 Tubing Type Transport Standard Efficiency Deviation Transport Efficiency Standard Deviation Transport Efficiency Standard Deviation FEP N/A N/A 99.65% 0.06% 99.82% 0.30% 99.90% 0.13% PTFE N/A N/A 98.85% 0.15% 99.32% 0.10% 98.92% 0.08% PFA N/A N/A 98.81% 0.09% 98.85% 0.09% 99.06% 0.09% PVDF N/A N/A 99.09% 0.07% 99.73% 0.15% 98.78% 0.06% Tygon® N/A N/A 80.66% 0.05% 80.30% 0.09% 81.47% 0.08% SS316 N/A N/A -0.25% 0.00% -0.25% 0.00% -0.16% 0.00% SNSS N/A N/A 96.98% 0.09% 97.51% 0.04% 97.37% 0.09% An extra test or verification test was performed due to an error in data recording by the Envidas System. The verification data set was used in the overall transport efficiencies table in Table 9. 46 ------- Appendix A Table 6. Nitrogen Dioxide (Conditioned) Nitrogen Dioxide 20ppb (Conditioned) Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP N/A N/A N/A N/A N/A N/A PTFE N/A N/A N/A N/A N/A N/A PFA N/A N/A N/A N/A N/A N/A PVDF N/A N/A N/A N/A N/A N/A Tygon® 85.49% 0.00% 84.90% 0.00% 84.82% 0.00% SS316 34.27% 1.61% 42.66% 0.70% 43.41% 0.27% SNSS 98.53% 0.00% 97.28% 0.28% 97.54% 0.00% Nitrogen Dioxide 50ppb (Conditioned) Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP N/A N/A N/A N/A N/A N/A PTFE N/A N/A N/A N/A N/A N/A PFA N/A N/A N/A N/A N/A N/A PVDF N/A N/A N/A N/A N/A N/A Tygon® N/A N/A N/A N/A N/A N/A SS316 N/A N/A N/A N/A N/A N/A SNSS 99.10% 0.12% 98.25% 0.15% 98.35% 0.10% Nitrogen Dioxide 120ppb (Conditioned) Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP N/A N/A N/A N/A N/A N/A PTFE N/A N/A N/A N/A N/A N/A PFA N/A N/A N/A N/A N/A N/A PVDF N/A N/A N/A N/A N/A N/A Tygon® N/A N/A N/A N/A N/A N/A SS316 N/A N/A N/A N/A N/A N/A SNSS 99.18% 0.09% 99.00% 0.09% 98.95% 0.08% \Tote that blank values indicate that test was not performed due to original set of data reaching the 97.5% transport efficiency threshold. Tygon® and SS316 were not tested at higher concentrations due to lower concentrations not increasing over the 97.5% transport efficiency threshold. 47 ------- Appendix A Table 7. Carbon Monoxide (Unconditioned) Carbon 1.8ppb Monoxide Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 99.48% 0.79% 99.83% 0.71% 100.72% 1.47% PTFE 100.03% 1.12% 99.06% 2.16% 97.79% 0.96% PFA 101.81% 0.98% 100.91% 1.89% 99.82% 1.05% PVDF 101.14% 2.50% 100.02% 2.04% 100.65% 1.17% Tygon® 101.52% 0.93% 100.75% 1.90% 100.50% 1.01% SS316 97.32% 1.07% 98.88% 1.41% 99.93% 1.11% SNSS 100.40% 1.15% 98.77% 2.23% 102.27% 1.43% Carbon 4.5ppb Monoxide Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 100.19% 0.36% 100.39% 1.00% 100.23% 0.45% PTFE 99.12% 0.40% 99.35% 0.74% 100.51% 0.73% PFA 99.65% 0.41% 100.64% 0.55% 100.71% 0.49% PVDF 100.52% 0.66% 101.00% 0.24% 99.91% 0.43% Tygon® 99.76% 0.51% 100.60% 0.42% 100.64% 0.52% SS316 100.09% 0.58% 100.71% 0.61% 99.95% 0.59% SNSS 100.00% 0.41% 100.55% 0.74% 99.14% 0.95% Carbon 10.8ppb Monoxide Repl Rep2 Rep3 Transport Standard Transport Standard Transport Standard Tubing Type Efficiency Deviation Efficiency Deviation Efficiency Deviation FEP 100.11% 0.20% 100.26% 0.17% 99.78% 0.36% PTFE 100.02% 0.46% 100.22% 0.37% 100.40% 0.34% PFA 99.42% 0.42% 99.56% 0.41% 99.92% 0.46% PVDF 100.04% 0.36% 100.46% 0.20% 100.38% 0.26% Tygon® 100.08% 0.30% 100.18% 0.48% 100.00% 0.29% SS316 99.91% 0.29% 99.42% 0.54% 99.58% 0.21% SNSS 100.06% 0.39% 99.33% 0.56% 100.06% 0.33% 48 ------- oEPA United States Environmental Protection Agency PRESORTED STANDARD POSTAGE & FEES PAID EPA PERMIT NO.G-35 Office of Research and Development (8101R) Washington, DC 20460 Official Business Penalty for Private Use $300 ------- |