Document 2JNbE52bXxOO0N8rVwdO65eON
Project: Investigate Municipal Incineration of Polytetrafluoroethylene (PTFE) to Evaluate the Release of Per- and Poly-fluorinated Chemicals of Environmental Concern (PFCOEC) thru Flue Gas Emissions
November 6, 2017
Study Location: Karlsruhe Institute of Technology (KIT)
Study Director: Dr.-Ing. Hans-Joachim Gehrmann: Institute for Technical Chemistry (ITC) at KIT
Combustion Technology Group
Study Coordinator: Mr. Dan Pigeon, W.L. Gore & Associates, Inc.
Study Sponsor: W.L. Gore & Associates, Inc. (Gore)
1. Study Objective: To determine if significant levels of specific compounds of environmental concern (Table 1) are emitted during the incineration of PTFE at conditions consistent with European Union (EU) municipal waste incineration regulations1,2.
2. Study Plan Summary: Under the supervision Dr.-Ing. Hans-Joachim Gehrmann the study will utilize the BRENDA (Brennkammer mit Dampfkessel) incineration facility at KIT to incinerate wood pellet blanks and wood pellet/ PTFE sample mixtures at municipal incineration conditions prescribed in EU Directive 2000/76/EC on the Incineration of Waste3. During the sample incineration, flue gas emissions will be sampled using isokinetic gas sampling techniques and sent for analysis by a third party lab to determine the concentration of the compounds being investigated. The isokinetic gas sampling will be performed by the staff of ITC under the technical oversight of Mr. Dan Pigeon. The staff will be trained in the sampling techniques and will have completed the capture validation study prior to the start of the incineration campaign. To analyze inorganic fluorine species, an additional sampling train will be used to sample hydrogen fluoride (HF) and capture fly ash. The results of the analyses will be sent back to the ITC team at KIT for compilation with the experimental parameters and validation data to create a paper for publication.
3. Equipment: 3.1. BRENDA (Brennkammer mit Dampfkessel): The Institute for Technical Chemistry at KIT operates a modern rotary kiln test facility (BRENDA) equipped with a boiler for heat recovery and a
1 IED (2010): "Directive 2010/75/EU Of The EuropeanParliament And Of The Council Of 24 November 2010 on industrial emissions (integrated pollution preventionandcontrol)", in: Official Journal of the European Communities, 17.12.2010, L 334/17-119. [hereinafter Industrial Emissions Directive (IED)]. 2 17. BImSchV (2003): Siebzehnte Verordnung zur Durchfhrung des BundesImmissionsschutzgesetzes (Verordnung ber die Verbrennung und die Mitverbrennungvon Abfllen - 17. BImSchV), Germany.Seventeenth [Ordinance on the Implementationof the Federal Immission Control Act (Ordinance on Incineration and CoIncinerationof Waste]. 3 WID (2000): "Directive 2000/76/EcCof The European Parliament And of The Councilof 4 December 2000 on the incineration of waste", in: Official Journal of the EuropeanCommunities, 28.12.2000, L 332/91-111.
complete flue gas cleaning system which complies with the German regulations of emissions2 (Figure 3.1.1), and IED regulations1.
Location #2
3.1.1 Operational Description of BRENDA (Brennkammer mit Dampfkessel): Solid fuel (e.g. wood pellets up to 40 mm grain size) are fed into the kiln through the bulk materials inlet and combined with a burner for natural gas. The rotary kiln revolves at 0.1 to 2 rpm at an inclination adjustable between 0.5-3 to cause the solid materials to travel slowly downwards through the rotary kiln (length 8.4 m, inside diameter 1.4m). The solid fuel is degasified, incinerated and at the end discharged into the wet de-slagger system. Depending on the temperature in the rotary kiln (up to 1300C), the combustion residues may be either solid or molten when discharged. The adjustment of the angle of inclination and the revolution speed of the kiln controls the residence time of the solid waste in the rotary kiln (fuel residence time: 2-6 h, gas residence time: 4-6 sec). The combustion gases from the rotary kiln are fed into the post combustion chamber which is equipped with two combined burners for natural gas. The burners are staggered anti-parallel to each other. This configuration induces high turbulence and, consequently, improved mixing of the combustion gases. The post-combustion chamber is designed for gas temperatures up to 1300C. The hot flue gas leaving the post-combustion chamber enters the boiler (residence time in the boiler: 22-25 sec) and generates saturated steam of 40 bar and 250C. The complete gas cleaning line (pollution control system) comprises a spray dryer, a fabric filter, two scrubbers and an SCR reactor.
The total thermal power is about 2.5 MW, while the rotary kiln contributes about 1.5 MW and the burners in the PCC about 1 MW.
3.1.1.For further information: http://www.itc.kit.edu/downloads/2PDF_2016_01_11_BRENDA_en.pdf
3.2. Isokinetic sampling train: The sampling train will be in a modified United States Environmental Protection Agency (EPA) Method 54 configuration. (Figures 3.2.1 & 3.2.2) Sample collection will be consistent with EPA Methods 1 thru 55,6,7,8,4 for stationary source sampling with one exception. The flue traverse collections points will be limited to one axis due to equipment limitations. (See Section 7.6.1) 3.2.1.The isokinetic sampling train will consist of the following parts: 3.2.1.1. Sampling probe consisting of a nozzle, sample barrel, pitot tube and temperature probe to allow for simultaneous measurement of flue gas velocity and temperature during sampling. 3.2.1.2. Heated filter assembly using a quartz fiber filter to capture fly ash particulate. 3.2.1.3. Impinger and adsorbent train consisting of the following parts: 3.2.1.3.1. Impinger #1 -Plain Stem 500ml Greenburg-Smith impinger containing between 200-250ml 0.15 molar NaOH to cool the gas stream and capture the water soluble compounds of interest 3.2.1.3.2. Impinger #2 - Stem with Orifice & Plate 500ml Greenburg-Smith impinger containing between 200-250ml 0.15 molar NaOH to capture any carried over water soluble compounds of interest from Impinger #1 3.2.1.3.3. Impinger #3 -Plain Stem 500ml Greenburg-Smith impinger kept empty to capture aerosolized water droplets from Impingers #1 and #2 3.2.1.3.4. Chilled Adsorbent Bed containing between 20-30 grams of adsorbent material (XAD-2) to capture the water insoluble compounds of interest 3.2.1.3.5. Impinger #4 -Plain Stem 500ml Greenburg-Smith impinger containing 200300 grams silica gel to collect remaining water vapor for flue gas humidity calculations. 3.2.1.4. Air tight pump to draw the sample 3.2.1.5. Dry Gas meter to measure total gas captured 3.2.1.6. Manometer to measure the sample capture velocity to match with the stack velocity to ensure isokinetic sampling conditions. 3.2.2.Process for collecting samples and rinsing the sampling train for a given test run: 3.2.2.1. Sample A will consist of the quartz fiber filter (3.2.1.2.) 3.2.2.2. Sample B will consist of the adsorbent from the chilled trap (3.2.1.3.4)
4 EPA Method 5 - Determinationof Particulate Matter Emissions from Stationary Sources. 40 C.F.R. Part 60, Appendix A-3. 5 EPA Method 1 - Sample andVelocity Traverses for Stationary Sources. 40 C.F.R. Part 60, Appendix A-1. 6 EPA Method 2 - Determinationof Stack Gas Velocity andVolumetric Flow Rate (Type S Pitot Tube), 40 C.F.R. Part 60, Appendix A-1. 7 EPA Method 3 - Gas Analysisfor the Determination of Dry Molecular Weight. 40 C.F.R. Part 60, Appendix A-2. 8 EPA Method 4 - Moisture Content. 40 C.F.R. Part 60, Appendix A-3.
3.2.2.3. Sample C will consist of the contents of Impinger #1 (3.2.1.3.1) and Impinger #2 (3.2.1.3.2)
3.2.2.4. Sample D will consist of the methanol rinse of all glassware between probe and the chilled Adsorbent trap.
3.2.3. All samples will be collected in clean polyethylene bottles and stored at 3-5C prior to shipping to the third party lab for analysis.
3.3. Validation Gas Generator (VGG): The validation gas generator (Figure 3.3.1 & 3.3.2) is designed to preheat a nitrogen gas flow of 0.5 - 0.8 SCFM (14-20 liters/minute) to between 150C - 250C. The preheated gas flows into a heated vaporization block where it is mixed with a vaporized stream of solvent and standard to fortify the gas stream with the standard vapor. The fortified gas stream is then sent to the sampling train to measure capture efficiency.
3.4. Gravimetric aerosol mass concentrations and sampling of HF: 3.4.1. The fly ash collection will be based on VDI 20669 (Figure 3.4.1a and 3.4.1b) 3.4.2. For collection of HCl and HF, an impinger filled with NaOH will be placed between the condenser/cooling system and the dryer of the VDI 2066 sampling unit (Figure 3.4.1a). The contents of the impinger and the condenser/ cooling system will be combined for analysis.
4. Incineration Materials: In order to maximize the signal to noise ratio, clean untreated wood pellets will be used as the solid fuel for incineration as well as the blank material. For the sample runs, the fuel will consist of the same wood pellets combined with PTFE pellets to a concentration of 0.3% (w/w) PTFE. Please note, the European Commission reference document "Integrated Pollution Prevention and Control Reference Document on the Best Available Techniques for Waste Incineration, August 200610" states the range of Fluorine in typical municipal waste is 0.010% 0.035% (w/w dry solids). 4.1. The 0.3% PTFE will be achieved through the use of a 0-10 kg/hour mass feeder system that will add the PTFE to the incoming solid fuel (wood pellets) conveyor just prior to addition to the incinerator. 4.2. Representative samples of blank wood pellet material and PTFE will be sent to a third party lab for Total Fluorine Analysis and PTFE identification respectively.
5. Laboratories for Analysis: Recognized third party commercial laboratories will be used for analysis of all validation and incineration collection train samples, as well as the incineration fuel materials. 5.1. All collection train samples (Section 3.2.1) will be analyzed via LC-MS/MS and GC-MS to quantify the compounds of interest (Table 1). 5.1.1. Each shipment of samples sent to the laboratories for analysis will contain at least one field blank sample set, one spiked sample set, and one solvent blank. Additional field blank sample sets will be added for each day of incineration testing.
9 VDI 2066 (2006): Particulate matter measurement, dust measurement in flowing gases, gravimetric determinationof dust load. [VereinDeutscher Ingenieure (2006)Messen vonPartikeln, Staubmessung in stromenden Gasen, Gravimetrische Bestimmung der Staubbeladung. VDI 2066] 10 European Commission (2006) Integrated pollution prevention andcontrol - reference document on the best available techniques for waste incineration.
5.2. The incineration materials (Section 4) will be analyzed via IR to confirm the identity of PTFE and Total Fluorine Analysis to determine the concentration of Fluorine in the blank wood pellets.
5.3. Chemicals selected for analysis were chosen based on the prevalence of use in environmental research for fluoro-organic chemistries and the existence of validated third party laboratory methods with appropriate standards and quality control.
6. Capture Validation: The Validation Gas Generator (VGG) (Section 3.3) will be used to validate the capture efficiency of the sampling train. A standard solution containing even number Perfluorocarboxylic acids (C4-C14) will be made in methanol. This solution will be vaporized in the VGG and delivered to the gas sampling train (Section 3.2.1). 6.1. The decision to validate the sampling system with perfluorocarboxylic acids was driven by the greater likelihood of their generation in the combustion process. Further it was decided to utilize the even number acids across the boiling range to measure the capture efficiency. (See Appendix 1: Stakeholder feedback - ANALYTES.) 6.1.1. Please note that while the validation will be done with the carboxylic acids, all of the compounds in Table 1 will be measured at Intertek for all samples tested. 6.2. Final operating conditions of the VGG will be determined experimentally, but the initial work will be done using the following procedure: 6.2.1. Set N2 Flow to 0.7 SCFM (20 l/min) Temperature controller #1 and #2 to 150C and allow system to equilibrate. 6.2.2. With the system heated, fully purging both syringe pump liquid feed lines (standard solution to waste and rinse solvent to system), if the temperature varies significantly at thermocouple #3 (TC3) adjust liquid flow rate or temperatures to compensate. Please note, TC3 is a thin wire thermocouple (0.015in / 0.381mm) capable of reading small temperature fluctuations in the gas stream with response time of 0.8 seconds. The stability of TC3 temperature will be used to determine if the vaporization is stable. 6.2.2.1. As a secondary check for consistent vaporization, a cool glass disk will be held near the gas outlet. A fog will form on the disk due to re-condensation, if the fog size and shape are constant the delivery rate is constant. 6.2.3. Start gas collecting using the impinger train. 6.2.4. Switch the 4-way valve (Figure 3.3.2) to send the standard to vaporization block, start the standard solution syringe pump, deliver the prescribed volume, and stop the syringe pump. 6.2.5. Switch the 4-way valve to send the rinse solvent to vaporization block, start the rinse solvent solution syringe pump, deliver adequate volume of rinse solvent to fully purge the line between the valve and the vaporization block, and stop the syringe pump. 6.2.6.Increase the set point of temperature controllers #1 and #2 to 250C. 6.2.7.Replace the solvent rinse syringe with a water rinse syringe and pump > 10ml through the system, as a steam cleaning step, then hold at temperature for 20 additional minutes. 6.2.8. Stop gas collection and gather samples for analysis. (Section 3.2.2) 6.3. For a given compound, collection capture efficiently must be greater than 50% and less than 150% to be acceptable for inclusion in during the incineration testing.
7. Incinerator Testing Plan 7.1. Incineration testing will be carried out at the KIT utilizing the BRENDA (Section 3.1) rotary kiln incinerator. 7.2. Target conditions for the post combustion chamber are 850C with at least 2 second retention time. 7.2.1. Target temperature and residence time were chosen based on municipal incineration regulations in the regions where the vast majority of waterproof outerwear sales occur.
Table 7.2.1
Country China
Post Combustion Chamber Temperature C
> 8501
Post Combustion Chamber Retention Time (s)
> 2
European Union
> 850
> 2
Japan
> 800 (900-1000*)
> 2
South Korea
> 850
> 2
United States
980-1100 C**
>2**
* Industry common practice.
** Industry common practice. No set temperature or retention time but
must meet emission regulations
1: for wastes with the sum of halogens < 1 wt.%
7.3. Together with natural gas, the feed rate of the wood is expected to be between 100-150 kg/hr to maintain the 1MW minimum thermal output in the main burner system located in the rotary kiln (primary combustion unit). Variations in residence time and temperature will be done by load variations of the natural gas burners in the post combustion chamber. 7.3.1.Measurements will be taken at the lowest load feasible, that best fulfill the target criteria of 850 C and 2 seconds residence time. The lowest load has to be determined experimentally during the incinerator campaign. If determination of the load is required, a higher load (e.g. 2 MW) will be used first and the load will be reduced incrementally to achieve the closest conditions to the target criteria. For low loads, the critical parameter is the flue gas temperature, for high loads the residence time. Both cases will be investigated. 7.3.2. Table 7.3.2 shows as an example the procedure and test program for one day.
7.4. The residence time will be determined by the volume of the combustion chamber divided by the volume flow of the flue gas. The balance for this determination is the volume after the (last) secondary air injection to the inlet of the boiler (see Figure 7.4 below). In this experiment, this is the level of the dust burner and the first boiler path. The flue gas temperatures are measured in the top of the combustion chamber. The flue gas volume flow is measured after the boiler. To validate these flows a mass and material balance is necessary.
7.5. The incinerator will be allowed to equilibrate after each fuel feed change to ensure steady state conditions prior to sampling. Equilibration time will be determine based on feed and burn rates, this time will be at least twice of the average residence time of the solid material in the rotary kiln (estimated to be 1 h).
7.6. The flue gas will be captured using the isokinetic sampling train (Section 3.2.1) following stack sampling procedures consistent with EPA Methods 1 thru 5 with one exception. The flue traverse collections points will be limited to one axis due to equipment limitations. 7.6.1. In the primary sampling location, after the boiler/ heat exchanger, access is limited to the bottom of the duct. Access is blocked by piping and infrastructure in the other axis. Since the sampling location is more than 8 duct diameters downstream and more than 2 duct diameters upstream of a potential flow disturbance, the variance should be minimal. As a further confirmation, separate fixed stack velocity measure points will be utilized to ensure an accurate representation of the duct velocity and flow.
7.7. For each incineration run, two locations listed in Figure 3.1.1 will be sampled. The first (Location #1) after the boiler, but before any of the pollution control systems and the second (Location #2) at the emission stack. Please note, if the samples from Location #1 do not show
significant levels of the compounds of concern (Table1) the samples from Location #2 may not be analyzed. 7.8. A representative sample of fly ash will be captured at Location #1 in the particulate filter assembly (Figure 3.2.2 - #2). There will be no direct sampling of the bottom ash due to the inability to adequately clean the system to collect unbiased samples or representative blanks. 7.8.1.Additional details on the decision to sample fly ash are provided in Appendix 1. Typical
municipal incinerator bottom ash consists of a mix of inert materials such as sand, stone, glass, porcelain, metals, and ash from burnt materials. 7.8.2.Additional fixed bed incineration tests will be performed to obtain information on the total fluorine in the bottom ash. 7.9. As part of various projects, exhaust gas measurements at large-scale plants are carried out by the ITC. The sampling method is based on VDI 2066 with filter units outside the flue gas duct9. In the following, the applied measuring equipment is shown in Figure 3.4.1a and 3.4.1b. This unit will be used for the fly ash collecting as well as the determination of HCl and HF. 7.9.1.A partial flow is sampled isokinetically at one measuring point (often the center of the pipe) from the chimney with a stainless steel sampling tube and the particles are deposited on 47 mm binderless quartz plane filters. The particle mass concentration is determined gravimetrically, whereby the filters (QMA / Whatman) have an empty weight of approx. 150 mg. For the determination of the particle mass concentration a loading of about 10mg is necessary. Planar filter housings from Tecora are used for sampling, which are heated to exhaust gas temperature by a heating sleeve. Gas sampling systems (Desaga Type 316) are used for the sampling, which have a maximum volume flow of 12 SLPM and the collection is carried out as standard with a sample flow of 10 SLPM. The sampling tube diameter needed is calculated for each plant separately. Using the available data for the temperature, the moisture content and the volume flow of the flue gas, as well as the diameter of the duct, the gas velocity in the duct can be calculated. By varying the diameter of the probe, the partial flow through the probe can be adjusted to isokinetic conditions. 7.9.2.For collection of HCl and HF, an impinger filled with NaOH will be placed between the condenser/cooling system and the dryer of the VDI 2066 sampling unit (Figure 3.4.1a). The contents of the impinger and the condenser/ cooling system will be combined for analysis. 7.10. To minimize any interference from external sources, all fluoropolymer runs will be paired with blank runs performed either immediately before or after the sample run. 7.11. The testing will take place over multiple days to obtain 4-6 paired samples. 7.12. Additional conditions will be investigated, if time permits. These conditions may include different temperatures and fuel loadings, which may better represent waste to energy production. These conditions are expected to produce higher temperatures and residence times than the minimums found in EU Directive 2000/76/EC on the Incineration of Waste3
8. Data Analysis: Results of the third party laboratories analyses will be sent directly to Dr.-Ing. HansJoachim Gehrmann (KIT) for compilation with the experimental parameters and validation data to create a paper for publication.
8.1. For each compound of each run the amount found will be the total of the amount found in samples A, B, C, and D (Section 3.2.2) of the sampling train multiplied by the standard total volume of gas emitted divided by the standard volume of the gas sampled. 8.1.1. For any sample, if a compound is not quantifiable it will be added at half the LOQ to the total of the other samples in the run set.
8.2. For every PTFE incineration sample run there will be a paired blank incineration run. Four to six paired tests will be run.
8.3. To determine if a significant signal for any given compound was found, a one sided Paired Ttest with a 95% confidence and power interval will be employed.
8.4. Tabulated results will be written into a report and a paper will be submitted for publication.
9. Quantitation Limits: Analytical laboratory quantitation limits will be determined by the quality control procedures of the third party lab (Intertek). Procedural quantitation limits will be calculated using the third party laboratory quantification limits and the relevant experimental data. 9.1. To calculate procedural LOQ in nanograms (ng) for a given compound collected in a sampling run, the sum of the quantification limits from each sample type will be divided by the weight fraction of the total sample used for the analysis. (Equation 9.1)
9.2. The calculation for the fraction of a given compound potentially generated due to PTFE incineration is a function of the amount captured multiplied by the total standard volume of flue gas emissions divided by the standard volume of the gas collected then divided by the amount of PTFE incinerated during the collection run. This value will be reported in parts per million (ppm) (Equation 9.2).
Table 1.
Substance name
Perfluorobutanoic acid Perfluoropentanoic acid Perfluorohexanoic acid Perfluoroheptanoic acid Perfluorooctanoic acid Perfluorononanoic acid Perfluorodecanoic acid Perfluoroundecanoic acid Perfluorododecanoic acid Perfluoro-tridecanoic acid Perfluorotetradecanoic acid Perfluorobutanesulfonic acid Perfluorohexanesulfonic acid Perfluoroheptanesulfonic acid Perfluorooctanesulfonic acid Perfluordecanesulfonic acid Perfluorooctanesulfonamide N-Methyl-Perfluorooctanesulfonamide N-Ethyl-Perfluorooctanesulfonamide
N-Methyl-Perfluorooctane-sulfonamidoethanol
N-Ethyl-Perfluorooctanesulfonamidoethanol 1H,1H,2H,2H-Perfluorooctanesulphonic acid 2H,2H,3H,3H-Perfluoro-undecanoic acid Perfluoro-3-7-dimethyl octane carboxylate 7H-Dodecafluoro heptane carboxylate 2H,2H-Perfluoro decan carboxylate 1H,1H,2H,2H-Perfluorohexan-1-ol 1H,1H,2H,2H-Perflourooctan-1-ol 1H,1H,2H,2H-Perflourodecan-1-ol 1H,1H,2H,2H-Perflourododecan-1-ol 1H,1H,2H,2H-Perflouro octyl acrylate 1H,1H,2H,2H-Perfluoro decyl acrylate 1H,1H,2H,2H-Perfluoro dodecyl acrylate 1H,1H,2H,2H-Perfluoro hexyl methacrylate 1H,1H,2H,2H-Perfluoro octyl methacrylate 1H,1H,2H,2H-Heptadecafluorodecyl methacrylate
Abbrev.
PFBA [PFC C4] PFPeA [PFC C5] PFHxA [PFC C6] PFHpA [PFC C7] PFOA [PFC C8] PFNA [PFC C9] PFDA [PFC C10] PFUdA [PFC C11] PFDoA [PFC C12] PFTrDA [PFC C13] PFTeDA [PFC C14] PFBS [PFS C4] PFHxS [PFS C6] PFHpS [PFS C7] PFOS [PFS C8] PFDS [PFS C10]
PFOSA N-Me-FOSA N-Et-FOSA N-Me-FOSE
alcohol N-Et-FOSE alcohol 1H,1H,2H,2H-PFOS
4HPFUnA PF-3,7-DMOA
HPFHpA H2PFDA 4:2 FTOH 6:2 FTOH 8:2 FTOH 10:2 FTOH 6:2 FTA 8:2 FTA 10:2 FTA 4:2 FTMA 6:2 FTMA
8:2 FTMA
CAS-No.
Capture Validation
375-22-4
Yes
2706-90-3
307-24-4
Yes
375-85-9
335-67-1
Yes
375-95-1
335-76-2
Yes
2058-94-8
307-55-1
Yes
72629-94-8
376-06-7
Yes
375-73-5
355-46-4
375-92-8
1763-23-1
335-77-3
754-91-6
31506-32-8
4151-50-2
24448-09-7
1691-99-2 27619-97-2 34598-33-9
2043-47-2 647-42-7 678-39-7 865-86-1 17527-29-6 27905-45-9 17741-60-5 1799-84-4 2144-53-8
1996-88-9
Analysis Group
Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside
Ionic Tenside
Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Ionic Tenside Non-ionic Tenside Non-ionic Tenside Non-ionic Tenside Non-ionic Tenside Non-ionic Tenside Non-ionic Tenside Non-ionic Tenside Non-ionic Tenside Non-ionic Tenside
Non-ionic Tenside
Figure 3.2.1 Isokinetic Sampling Train, Pump and Meter
cS i. et| Erpl innLa fo
=
we
Ted Hel
Vacuum
orifice
wByipeass jMaoisn Gauge
=t--1 [b
[1]
OC
DryGas Meter
LR,
8)
eeNeauseNE T
Regteiig 8
VlTok
1%
d iE
[=
SITt [l: i
= tH
2
. = 2)
. =r
mnie
x = =--+r 77D3s-
Figure 3.3.1 Schematic of Validation Gas Generator
=p Tempers
`Vaporization
TCeomnpterroaltleurre 1
Block
Samping Sytem pling Syst
Coma osloed sox
F( iSgtuarned3a.r3d.2InStcrhoedmuacttiiocnoSfuVbasleicdtaitoino)n Gas Generator
`SStyarnidngaerdPuSomlput#i1on
[| ttm
4-Way Valve
RSyirnisnegSeopluutmiopn62
rComrmoller |
Heer core
Thermocouple #2
|
[ standard
|;
EVioporization ||| Gas Outlet
|
Pina) |
erremePrreenhteateodoo. tiatBeoxd| Gos nex
Figure 3.4.1a Gravimetric measurement of particle mass concentration based on VDI2066
Figure 3.4.1b Sampling on 47 mm plan filters with heated TECORA filter housing
350
051
Sampling tube
Sonde 6 mm
Ball valve with heating sleeve
Kugelhahn mit Heizmanschette
SampPlrionbgefnialthemrefwilteitrh
mit Heizmanschette
heating sleeve
Figure 7.4 Post Combustion Chamber
2 calibrated thermocouples
Area of 2 - seconds residence tim e
Start post combustion zone
Level of natural gas burners
363E
Appendix 1: Stakeholder Feedback
GENERAL
Comment: The term "PFC of Environmental Concern" was challenged in the context of combustion studies due to the possible perception of intended derogation. The term "PIC" or products of incomplete combustion was noted as a more commonly used phrase.
Response: The term "PFC of Environmental Concern" was established in Gore's Fabrics' Goal and Roadmap for Eliminating PFCs of Environmental Concern1 prior to the commission of any studies. In order to maintain consistency with the roadmap document, "PFC of Environmental Concern" will be used throughout this study.
Comment: "Heavy focus on stack sampling, less emphasis on good solid waste combustion."
Response: The purpose of this study is to determine specific compounds of environmental concern emitted during the incineration of PTFE at conditions consistent with EU municipal waste incineration regulations. Collecting stack emissions generated during PTFE combustion/incineration during typical municipal waste incineration conditions was identified as a data gap. Stack testing is the primary method of collecting emissions generated throughout the incineration process to address current data gaps and is the focus of this experiment.
The incineration conditions will be governed by the laws of operation for the incinerator system according to the German and EU municipal waste incineration regulations2,3.
PTFE
Comment: Four commenters requested additional details on the type of PTFE material chosen for this study.
Response: To best represent the component layer of an outerwear garment, a high molecular weight PTFE extrudate form will be used. It was produced through a standard PTFE process consistent with early stages of ePTFE film manufacture. The PTFE pellets were manufactured from fine powder PTFE homopolymer that used a fluorinated emulsifier/surfactant (PFOA) as a polymer production aid. This particular type of PTFE is no longer purchased by Gore, but it was selected to represent a conservative case for the presence of long chain perfluoroalkyl acids.
1 Kiehl, B. (2017). Gore Fabrics' Goal andRoadmap for Eliminating PFCs of Environmental Concern. https ://www.gor e-tex.c om/pfc goal. 2 IED (2010): "Directive 2010/75/EU of The European Parliament And Of The Council of 24 November 2010 on Industrial Emissions (Integrated Pollution PreventionandControl)", in: Official Journal of the European Communities, 17.12.2010, L 334/17-119. [hereinafter Industrial Emissions Directive (IED)]. 3 17. BImSchV (2003): Siebzehnte Verordnung zur Durchfhrung des BundesImmissionsschutzgesetzes (Verordnung ber die Verbrennung und die Mitverbrennungvon Abfllen - 17. BImSchV), Germany.Seventeenth [Ordinance on the Implementationof the Federal Immission Control Act (Ordinance on Incineration and CoIncinerationof Waste)].
1
Laminated materials were not selected for this study due strong recommendations from subject matter experts (SMEs). The difficulty to definitively identify the PTFE contribution due to a lack of validated analytical techniques for complex matrices was cited.
INCINERATION MATERIALS
Comment: Several commenters questioned the choice of wood (sawdust) as a fuel for incineration as compared to natural gas, ethanol or a municipal waste mixture. Additionally, several commenters recommended analyzing the PTFE-fortified and blank wood pellets off-site for Wickbold Torch Total F Analysis.
Response: Through further investigation of the equipment capabilities at the KIT Incineration facility, the initial plan of creating fortified wood pellets has been modified. The 0.3% PTFE will be achieved through the use of a 0-10 kg/hour mass feeder system that will add pure PTFE pellets to the incoming solid fuel (wood pellets) conveyor just prior to addition to the incinerator.
The wood pellets will be analyzed at a third party lab for total Fluorine content using a Wickbold Torch Analysis or a comparable method to determine the total fluorine in the solid feed as well as in the fly ash collected after the boiler. The choice of wood pellets as the incineration material was discussed with subject matter experts (SMEs) and it was concluded that the final determination of fuel will be made to minimize impurities, ensure good integration of fuel with the PTFE and optimize burnout levels. The best solution at this time is wood pellets.
Comment: The SAACKE SSB swirl burner can apparently feed various liquid and gaseous fuels. Please confirm elsewhere in the plan that natural gas is the fuel to be used. Please also consider noting (later) the conditions under which natural gas is to be fired through the swirl burner.
Response: There are three burner systems to consider.
System 1 2 3
Location
Rotary Kiln
Entrance of Post Combustion Chamber
Entrance of Post Combustion Chamber
Burner system Lances for
natural gas and oil, dosing for
solids Post Combustion Chamber
SAACKE SSB Swirl
Possible fuels Light Oil, Natural Gas,
Paste, Solid Waste
Light Oil, Natural Gas
Dust, Natural Gas
Planned Fuel Natural Gas, Wood pellets
Natural Gas
Natural Gas
In the current plan the 2nd and the 3rd system will be used in order to achieve at least 850 C. Natural gas and solid fuel mixture will be used in the rotary kiln and the post combustion chamber (PCC) burner will use natural gas.
Comment: Wood may burn too quickly to allow PTFE to pyrolize. Mixing PTFE with a higher caloric content material was suggested. CO and CO2 should be measured to characterize burn rates.
2
Response: Both CO and CO2 will be measured after the rotary kiln and after the supply of natural gas in the PCC throughout the testing to ensure good combustion efficiency.
Comment: Have you reviewed this Reference Document ["Integrated Pollution Prevention and Control Reference Document on the Best Available Techniques for Waste Incineration, August 2006"]? Have you calculated the associated [HF] prior to control? Note that these numbers suggest F loading during the test (0.3%) will be at least 7 times higher than typical.
Response: Yes, the document referenced above has been reviewed. The HF concentration in the document was the starting point. It was concluded that 0.3% (w/w) PTFE is the best balance between the worst case scenario and safely running the incinerator within legal limits of 1% total halogen content according to the EU IED - operating conditions, No. 1 and 2 (IED, 2010)2. Although Directive 2000/76/EC (WID, 2000)4 refers to "halogenated organic substances expressed as chlorine", this percentage is being used as a reference for the study calculations, since fluorine is not specified.
STACK TESTING/SAMPLING
Comment: Two commenters described the potential for gaseous PFASs to sorb strongly to surfaces, including, but not limited to, unburned wood chips, chamber walls, glass fiber filters, plastic surfaces and glassware.
Response: The potential for PFASs to sorb to surfaces is understood and steps will be taken to limit the likelihood of this occurring where feasible. Measures to reduce the potential of PFAS absorption to surfaces include using methanol to clean various surfaces prior to and after the experiment and extracts will be analyzed. In addition, validation work scheduled in September and November 2017 should address some of these concerns.
Comment: "Relying on a thermocouple to demonstrate continuity of flow and vaporization may not be adequate. Preliminary testing with water and a relative humidity monitoring device could work. If the solvent is something like methanol, UDRI has been shown to work as a method to demonstrate consistent vapor flow rate in lab-scale systems."
Response: As a secondary check for consistent vaporization a cool glass disk will be held near the gas outlet. A fog will form on the disk due to re-condensation, if the fog shape is constant, the delivery rate is constant.
Comment: One commenter suggested analyzing a filter from the filter lot to be used and review the analytical results for the filter prior to conducting the test program.
Comment: "What is the planned basis for adsorbent selection? Is this to be determined as part of testing in preparation for (or part of) the capture validation?"
"Adsorbent and silica gel blanks are advisable. It may be prudent to analyze a sample from the lot of each to be used prior to beginning the test program akin to the note above about the quartz filter.
4 WID (2000): "Directive 2000/76/EC of the EuropeanParliament and of the Council of 4 December 2000 on the incineration of waste" in: Official Journalof the EuropeanCommunities, 28.12.2000, L 332/91-111.
3
Additionally, as with the quartz filter, reagent blanks of each material used in the Method 5 train are generally collected." Response: All materials from sampling train including filters will be analyzed with exception of the silica gel since it will be used to capture moisture and is introduced after XAD-2 Adsorbent bed. Comment: Several commenters suggested modifying the capture efficiency from 50% - 150% to 70% 130% or 80% - 120%. Response: While the aim is to achieve better recoveries, it was concluded that compounds should not be eliminated simply because they are difficult to capture. SMEs agreed that achieving capture efficiency outside of this range is difficult outside of a laboratory environment. Comment: What is the basis for 4-6 runs? Response: The availability of this incinerator is limited. In addition, significant time is allotted each day for preparing the incinerator for this experiment. Due to these time constraints, only 4-6 paired runs per condition can be performed over the 6 day course of this experiment. For a paired sample t-test with 95% confidence and 95% power a positive difference of, 2.8 to 4 times the standard deviation, from the blank with 6 to 4 paired samples, respectively, would be required to show significance.
Comment: If all 4 separate samples are <LOQ it would be good to also combine these four samples, extract and analyze them together. If you are still <LOQ then you can be more sure that no PFASs are being generated from the PTFE. Response: For this study, standard laboratory quality control procedures will be followed. The 4 separate samples will be extracted and analyzed individually. Extracting and combining four samples after the initial analytical work increases the potential introduction of contaminants and falls outside of standard laboratory quality control procedures. In order to achieve the lowest procedural LOQ for a given test run, a third party laboratory (Intertek) is being collaborated with to determine the most efficient sample collection grouping to minimizeeach laboratory LOQ ..
4
ANALYTICAL / LABORATORY
Comment: Several commenters requested additional information on the selected laboratory. In addition, one commenter suggested sending the results to a data validation firm to ensure the results meet quality control criteria.
Response: A reputable laboratory, Intertek, located in Germany has been selected. Intertek holds many industry and customer specific quality certifications including ISO certificates, Nadcap, and Good Laboratory Practice (GLP)5. In addition, Intertek's testing and certification services support the quality, performance, regulatory compliance, safety, benchmarking, evaluation, validation, analysis, and other requirements for products, components, raw materials, sites, and facilities. Based on the selected laboratory's qualifications, a third party data validation firm will not be used. Intertek is two hour drive from incineration site and KIT staff will be trained on proper handling and transport of samples to the laboratory.
Comment: Confirm IR apparatus is capable of detecting C-F at the planned PTFE concentration for the wood pellets to assure IR suitability.
Response: Because the decision has been made to use a secondary mass feeder for pure PTFE instead of the fortified wood pellets, IR spectroscopy should be adequate to identify the pure material.
INCINERATOR
Comment: One commenter asked why a rotary kiln was chosen over a grate type kiln. The commenter further explained via teleconference that in the EU, rotary kilns are typically used for hazardous waste incineration and grate type kilns are used for municipal or non-hazardous waste incineration.
Response: The rotary kiln (BRENDA) can be used for municipal waste incineration and is the best suited system for the purposes of this study. Different systems were evaluated including a smaller grate system (fixed bed), but those systems could not achieve the 2 second residence time at 850 C as required for the purposes of this study. The BRENDA system is able to meet the constraints and requirements of this study while complying with EU Municipal Waste regulations4.
Comment: How will you minimize the potential impact of residual build-up in the incinerator system?
Response: Prior to the study, the incinerator will be cleaned by attempting to break free any buildup on the walls of the system. Further, 60-75% of each sample day will be spent running blank material (wood pellets). The rotary kiln will be continuously supplied with wood pellets to ensure a steady state mode operation.
Comment: The average temperature and gas-phase residence time in the post combustion chamber is quite low for burning such recalcitrant materials. I think partial combustion could be an issue, especially new surfaces and boundary layers.
5 http://www.intertek.com/testing/ 5
Response: The control temperature will be measured at the exit of the post combustion zone according Figure 1. These temperature measurements were calibrated by the TV Rheinland Group (TV, 2007)6.
2 calibrated thermocouples
Area of 2 - seconds residence time
Start post combustion zone
Level of natural gas burners
Figure. 1: Schematic of the BRENDA PCC Both CO and CO2 will be measured throughout the testing to ensure good combustion efficiency. For more detail on the operation of the Incinerator please see section 3.1.1 of the Study Plan. Comment: What excess air level will exist in the chamber? This is not specified. In the UDRI study, it was 150% EA. Response: Based on previous experiments, Dr. Gehrmann expects the excess O2 level to be between to 160-190% during the testing.
6 TV Rheinland Group (2007). Calibration Study(unpublished) No.: 829057_2007_60032857 1/1419/03. 6
ANALYTES
Comment: Several commenters noted certain chemicals listed on Table 1 of the study plan were not likely to form. Specifically, fluorotelomer alcohols (FTOHs), perfluoroalkane sulfonic acids, perfluoroalkane sulfonamide, perfluoroalkane sulfonamido alcohols (PFAS), and branched PFOA - perflouro-3-7-dimethyl octane carboxylate.
Response: Greenpeace's Priority Group 11 (Detox MRSL 08062015) list of PFCs (Table 1) are the targeted analytes for the purposes of this study. While the compounds mentioned above are not likely to form (as explained in detail below), Intertek has proven methodologies for all of the compounds listed in Table 1. To allay any potential concerns, samples will be analyzed for the full series of compounds in Table 1. However, based on feedback from several SME's, validation activities will be focused on straight chain carboxylic acids. The system will be validated using even numbered acids (C4-C14).
Dr. Greg Fritz of Greg Fritz Consulting, LLC (Senior Chemist (retired) Office of Pollution Prevention and
Toxics, USEPA - and former drafting committee member of the Technical Subgroup on Incineration
Test Methods for Fluoropolymers) provided the following explanation:
"The conversion of waste to energy is the primary function of Municipal Waste Incinerators (MWI), as opposed to the function of hazardous waste incineration which is to destroy hazardous materials or remediate contaminated substrate. The first consideration in the list of chemicals in Table 1 must be their fuel value. The wood pellets [or in some cases fossil natural gas / oil supplements] used as fuel are chemicals which are composed of predominantly covalently bound C, H, O atoms. The energy from bond cleavage is released during the exothermic reactions of oxidative cleavage of these fuels depending on the temperature used and the oxygen content maintained in the rotary kiln and incineration chamber. This makes the initial examination of the chemicals in Table 1 relatively easy. Chemicals having C-H and C-O bonds in the structure will be converted into water [H2O] and carbon dioxide [CO2]. In the case of the fluorotelomer alcohols (FTOHs) and their derivatives listed in Table 1 [Rf-CH2-CH2-O-R' where Rf is a perfluoroalkyl moiety and R' is a H atom or other alkyl /acyl derivative] the initial reaction in the incineration process will be the consumption of the fuel value chemical bonds. The remaining Rf moiety will be subject to mineralization if the incineration temperatures are high enough to completely cleave the remaining C-F bonds. Alternatively, the remaining Rf moiety will react to produce fluorine containing degradants which would be similar to the degradation products produced from reaction of the structurally similar PTFE fluorochemical chain during low temperature incomplete incineration. This analysis also applies the 1H,1H,2H,2H-tetrahydro PFAS derivatives listed on Table 1. The two methylene CH2 groups will be converted to CO2 and H2O like any other fuel leaving an Rf moiety as described above. The Norwegian Institute for Air Research (NILU) "Emissions from incineration of fluoropolymer materials" (Huber, 2007)7 literature survey cited a combustion study (Garcia, 2007)8 that found some PFAC [< 0.01%] and other analytes as minor products from PTFE fuel rich combustion.
7 Huber, S., Moe, M. K., Schmidbauer, N., Hansen, G. H., & Herzke, D. (2009). Emissions from incinerationof fluoropolymer materials. 24-25. 8 Garcia, A.N., Viciano, N. and Font, R. (2007). Products obtained inthe fuel-richcombustionof PTFE at high temperature. J. Anal. Appl. Pyrol., 80, 85-91.
7
The analysis of chemicals from the Table 1 that are PFAS based is a bit more complicated and centers on the question whether fluorine containing moieties under the MWI conditions used for the Incineration Pilot protocols could form C-S bonds generating PFAS [Rf-SO3H] or PFAS derivatives. The derivatives are addressed in the fuel discussion above. The derivatives compounds listed will be consumed to PFAS or Rf as the CH, CO, C-N, and S-N bonds are oxidatively cleaved, not formed, under the incineration conditions used. Several authors and 3M [in support of the Minnesota state emissions permit] note the lower temperature 600 C cleavage of the C-S bond as support that PFAS will not be released from stack emissions (Concawe, 2016)9. The NILU literature survey cited a study that concluded "The concentration was below 1 g/m3 in the stack, and the concentrations were relatively independent of kiln feed. According to Lemieux et al. (2004)10the results indicated that the perfluorinated substances were effectively destroyed even under mild combustion conditions".
That leaves the list of PFAC C4 - C14 perfluoroalkyl carboxylates as the chemicals of interest, although the above literature would suggest that MWI emissions of this is also low from PTFE combustion / incineration. It should be pointed out that the one PFAC, perfluoro-3,7dimethyloctane carboxylate, would not be a direct product from PTFE decomposition, but would have to be formed from alkene coupling [a pyrolysis product mechanism] followed by oxidation. These reactions would be in competition with the oxidative process forming the linear acids which as stated above were only present in < 0.01% from PTFE combustion."
Comment: Non-listed chemicals should be considered - HF from any mineralization pathway - 1-Hydroperfluorohydrocarbons from decarboxylation.
Comment: I think it would be really good to look if polyfluorinated dibenzodioxins and dibenzofurans are formed (openly, this is the only concern that people would have if F-containing waste is burned). Some would go for mixed fluorinated-chlorinated DD/DF (as they do for the brominated and mixed DD/DF).
Comment (General): Why aren't you looking at non-listed chemicals such as polyfluorinated dibenzodioxins, dibenzofurans, 1-hydro-perfluorohydrocarbons or HF?
Response: While it is acknowledged that other PFCs of Environmental Concern not listed on Table 1 could potentially form during the incineration process, only proven and commercially available methods to capture and validate PFCs of environmental concern will be investigated. In a previous study, formation of polyfluorinated dibenzodioxins (PFDDs) / polyfluorinated dibenzofurans (PFDFs) could not be detected during thermal treatment of PTFE11(Weber, 1995). Additionally, no commercially validated methodologies for sampling or analysis are currently available for PFDFs, PFDDs, or 1hydroperfluorohydrocarbons.
9 Concawe (2016). Environmental fate andeffects of poly- andperfluoroalkyl substances (PFAS). Report no. 8/16. 10 Lemieux, P. M., M. Strynar, D. G. Tabor, Joe Wood, M. Cooke, B. Rayfield, and P. Kariher. (2004). Emissions of Fluorinated Compounds from the Combustion of Carpeting. In Proceedings, 2007 International Conference on IncinerationandThermal Treatment Technologies, Phoenix, AZ, May 14 - 18, 2007. Air & Waste Management Association, Pittsburgh, PA. 11 Weber, Roland & Schrenk, Dieter & Schmitz, Hans-Joachim & Hagenmaier, A & Hagenmaier, H. (1995). Polyfluorinated dibenzodioxins and dibenzofurans -- synthesis, analysis, formation and toxicology. Chemosphere. 30. 629-39. 10.1016/0045-6535(94)00429-X.
8
Samples will be collected for HF analysis using a NaOH filled impinger incorporated into the VDI 2066 particulate sampling system (see section 3.4.2); however, due to the reactivity of HF and potential losses in the system, closure of the mass balance on Fluorine is not expected.
Comment: The scrubber water is the critical matrix for the PFAS; therefore, if any organic fluorinated compound may survive (or those that are in Table 1) would be found in water and not in the air. Thus, I recommend at least a check of the scrubber water for PFAS; perhaps the whole Table 1.
Response: It is understood that many of the compounds are very likely to be captured in an aqueous scrubber which is why testing will be conducted prior to the pollution control devices (e.g. scrubber, etc.) utilizing a modified Method 5 train containing two 0.15M NaOH filled impingers, which will act as scrubbers, to capture the air stream.
Comment: Several commenters asked if the fly ash and/or bottom ash will be sampled and analyzed for organic compounds.
Comment: In a similar way: I do not think it is necessary to analyse the fly ash or any organic fluorine molecule; rather suggest to look into the bottom ash I the incineration process went well in the primary chamber (but this may be difficult with wood and with a rotary kiln.
It appears there will be no attempt to analyze the fly ash for organic compounds? This would seem to be an important potential sink for PTFE degradation products.
Response: Fly ash will be sampled and analyzed in this study, but bottom ash will not. Typical municipal incinerator bottom ash consists of a mixture of inert materials such as sand, stone, glass, porcelain, metals, and ash from burnt materials. Acquiring independent samples of bottom ash during this study is not feasible. BRENDA has a wet ash discharger, so bottom ashes will be discharged into a water bath with several cubic meters of water. It would be impossible to determine ionic fluorine in the water or the mass of bottom ash for each specific test.
Additionally, fly ash is expected to be the predominant ash in this study. Due to the minimal amount of ash in the wood (< 1 wt.-%) and assuming 100 kg/h mass flow, a rather low mass flow of bottom ash (< 1 kg/h) is expected.
Further, some information is already available on PFAA content in fly ash and bottom ash (slag). In the thesis "Waste Incineration as a Possible Source of Perfluoroalkyl Acids to the Environment - Method Development and Screening", fly ash and slag (bottom ash) were sampled and analyzed for PFAAs. The results from sample analysis revealed that PFAAs were present in all solid samples at concentrations in the low to sub ng/g range and in all but one condensate and wastewater samples at concentrations in the low to sub ng/L range. PFAAs in bottom ash (slag) were variable between samples, but generally lower than PFAAs in fly ash. (Sandblom, 2014)12
12 Sandblom, O. (2014). Waste Incinerationas a Possible Source of Perfluoroalkyl Acids (PFAAs) to the Environment - Method Development andScreening. Master's Thesis. University of Stockholm. Unpublished.
9
