United States Environmental Protection Agency National Risk Management Research Laboratory Research Triangle Park NC 27711 Research and Development EPA/600/SR-96/007 March 1996 4>EPA Project Summary Experimental Investigation of PIC Formation During CFC Incineration G. Kryder and B. Springsteen Experiments were conducted to as- sess (1) the effect of residual copper retained in an incineration facility on polychlorinated dibenzo-p-dioxin and dibenzofuran (PCDD/PCDF) formation during incineration of non-copper-con- taining chlorofluorocarbons (CFCs); and (2) the formation of chlorinated and aromatic products of incomplete com- bustion (PICs), including PCDD/PCDFs, during incineration of CFC recycling residue and hydrochlorofluorocarbons (HCFCs). High concentrations of PCDD/ PCDFs (23,800 ng/dscm at 7% O2) mea- sured in a fiscal year (FY) 91 study during incineration of dichlorodifluo- romethane (CFC-12) in the turbulent flame reactor (TFR) could not be re- peated in the present study. Repetition tests conducted in the same facility under similar operating conditions re- sulted in PCDD/PCDF concentrations of 118 ng/dscm at 7% O2. However, results of the present study suggest that residual copper retained in an in- cineration facility possibly promotes the formation of PCDD/PCDFs during in- cineration of CFC-12 which does not contain copper. Tests conducted in the TFR facility resulted in measured PCDD/ PCDF concentrations of 386 - 454 ng/ dscm at 7% O2 during incineration of CFC-12 which followed incineration of copper-containing compounds. These results suggest that CFCs may best be incinerated in incinerators that do not treat any copper-containing waste prior to CFC incineration, in order to elimi- nate the possibility that residual cop- per retained in the incineration systems could promote PCDD/PCDF formation during subsequent CFC incineration. In this study, 1,1-dichloro-1-fluoroethane (HCFC-141b) and the oily residue gen- erated during CFC recycling processes were thermally destroyed without sig- nificant emissions of volatile organic PICs and PCDD/PCDFs. This Project Summary was developed by EPA's National Risk Management Research Laboratory, Research Triangle Park, NC, to announce key findings of the research project that is fully docu- mented in a separate report of the same title (see Project Report ordering infor- mation at back). Introduction A bench-scale study of the incineration of CFCs, CFC-12 and trichlorofluorometh- ane (CFC-11), was conducted in FY 91. For tests performed in the TRF, a water- cooled furnace (rated at 20.5 kW) at rela- tively low flame temperatures (790°C for the primary flame and 980°C for the sec- ondary flame), the CFCs were shown to be consistently destroyed at very high ef- ficiency (greater than 99.999%); however, significant levels of chlorinated and aro- matic PICs were detected. The test facility in which the FY 91 tests were conducted was previously exposed to trace metals including copper during incineration of metals-containing waste prior to the CFC incineration study. PCDD/PCDFs were sampled for and detected at high levels during a single test condition in this study. The formation of high levels of PCDD/ PCDFs could have been the result of the catalytic effect of residual copper in the test facility remaining after incineration of the metals-containing waste. Of particular concern, PCDD/PCDFs were sampled for and detected at high levels in one test condition. Subsequently, additional CFC incinera- tion tests were performed in T-Thermal's pilot-scale incinerator (rated at 290 kW) in FY 92. These tests were conducted at a high flame temperature (1,090°C) with water injection into the flame zone for temperature control. PCDD/PCDFs were ------- found at moderate levels in only one test with a high input of a CFC (69% CFC-11 by weight in fuel oil). PCDD/PCDFs were not found in other tests with lower CFC inputs. The avoidance of PCDD/PCDF for- mation in these tests is possibly due to the higher temperature and/or water injec- tion. Also, the pilot-scale incinerator is equipped with a water-quench tank di- rectly downstream of the burner to pro- vide rapid flue gas quenching; flue gases were quenched from 1,090 to 120°C in about 0.5 sec. The high flue gas quench- ing rate may have limited chemical reac- tions that lead to the formation of PICs in the flue gas. Follow-up bench-scale tests were con- ducted in FY 92, using the controlled tem- perature tower (CTT) to determine the effect of flame zone temperature on gas- phase flame formation and destruction of PICs during CFC-12 incineration. The ef- fect of water injection into the flame zone was also studied. PCDD/PCDFs were not detected at a high flame zone tempera- ture (1,200°C), while moderate levels of PCDD/PCDFs were detected at a lower flame zone temperature (900°C). Low lev- els of PCDD/PCDFs were also detected at the lower temperature with water injec- tion into the flame zone. Flame zone wa- ter injection may have a reducing effect on PCDD/PCDF formation during CFC-12 incineration. The effect of metal contamination of CFCs on the incineration emissions was studied in FY 93. The CFCs evaluated during all of the previous incineration tests were unused, reagent grade products; waste and recycled CFCs were not exam- ined. CFC refrigerants are likely to have had long-term contact with heat exchang- ers made of copper-based alloys. The pos- sibility exists that some copper may have been leached from copper alloy tub- ing by acids which may be formed as the result of a CFC's degradation. The cata- lytic properties of copper in PCDD/PCDF formation are well documented. Test re- sults indicated that incineration of waste CFC-11 produced low levels of PCDD/ PCDFs. Significant levels of PCDD/PCDFs were found when the waste CFC-11 which was tested was spiked with 300 ppm cop- per. Purpose The purposes of this work were (1) to further evaluate incineration as one of the appropriate technologies for the safe dis- posal of CFCs, including HCFCs and CFC recycling residues; and (2) to compare current results with those from earlier work in order to determine if the earlier high- dioxin emission results are typical for CFC incineration or a one-time event. The first objective of this study was to assess the effect of residual copper retained in an incineration facility on PCDD/PCDF for- mation during incineration of non-copper- containing CFCs. Therefore, initial work of the present study was a repetition of the PCDD/PCDF formation measured in FY 91. This repetition test was performed in the same water-cooled furnace (TFR) at relatively low flame temperatures in which the FY 91 tests were conducted. In fact, the bench-scale test facility was not in use between the FY 91 and the present tests so any residual copper that was present in FY 91 should have been present in FY 94. Subsequent to the repetition test, two additional tests were performed with more copper injected into the TFR prior to the CFC-12 incineration to evalu- ate the effect of residual copper retained in the TFR on subsequent CFC incinera- tion. At the beginning of the test, copper- containing fuel was incinerated in the test facility, followed by incineration of non- copper-containing CFC-12 during which the emissions of PCDD/PCDFs were mea- sured. The second objective of this study was to measure the formation of chlorinated and aromatic PICs, including PCDD/ PCDFs, during incineration of a CFC re- cycling residue and a representative HCFC. The production of CFCs is se- verely restricted under international agree- ments and federal regulations. HCFCs have become more popular as CFC sub- stitute refrigerants. Incineration may be an appropriate disposal technology for HCFCs. Therefore, in the second part of this study, the formation of PICs and PCDD/PCDFs during incineration of a HCFC was investigated. In addition to sub- stitution of HCFC for CFC, recycling of CFCs is becoming more popular. Signifi- cant quantities of residues generated dur- ing the recycling process require disposal. Incineration of these residues may be an appropriate disposal method. However, preliminary analysis of CFC recycling resi- dues has shown that they may contain up to 15 ppm copper, which has been shown to have a catalytic effect on the formation of PCDD/PCDFs. Therefore, in the last part of this study, the formation of PICs and PCDD/PCDFs during incineration of a CFC recycling residue was investigated. Procedure All experimental testing was conducted at a test site in Irvine, California. Two bench-scale combustion research facilities were utilized in this study. The experi- mental tests to assess the effect of re- sidual copper retained in an incineration facility on PCDD/PCDF formation during incineration of non-copper-containing CFCs, including the repetition of the FY 91 tests, were conducted in the TFR used for the FY 91 study. The experimental tests to measure the formation of chlori- nated and aromatic PICs, including PCDD/ PCDFs, during incineration of a CFC re- cycling residue and a HCFC were con- ducted in the CTT used for the FY 92 and FY 93 studies. The experimental tests in this study con- sisted of three tasks shown in Table 1. The primary focus of the experimental sampling was the measurement of PCDD/ PCDF formation during incineration of CFC, HCFC, and a CFC recycling resi- due. During each test, flue gas samples were collected for analysis of semi-vola- tile PCDD/PCDFs using EPA Method 23. During incineration of a HCFC and a CFC recycling residue, flue gas samples were collected and analyzed for volatile haloge- nated and non-halogenated organic PICs using EPA SW-846 Method 0030 (volatile organic sampling train). Also, the flue gas was monitored for other combustion prod- ucts (O2, CO2, CO, and NO) using a con- tinuous emission monitor. Task 1 was a repetition test of the FY 91 high PCDD/PCDF emission experimen- tal results. The tests were conducted in the TFR. The TFR facility was not used for experimental testing after completion of the CFC incineration study in FY 91; however, the refractory quarl in the pri- mary combustion zone had been replaced. For these tests, the TFR was configured similarly to the FY 91 configuration. The primary combustion zone was maintained at 760 - 870°C, and the secondary was maintained at 980 - 1,090°C. The Task 2 tests were also conducted in the TFR facility which was operated in the same configuration as in the Task 1 tests. Fuel oil containing copper was fired in the facil- ity, and an aqueous solution of copper salt was injected into the facility to simu- late incineration of copper-containing wastes. Following this, CFC-12 (~ 9% by volume in propane) was incinerated, and the formation of PCDD/PCDFs in the flue gas was measured at two sampling loca- tions. Additional copper-containing fuel was fired in the facility, and then the CFC-12 incineration test was repeated. Task 3 was designed to measure formation of chlori- nated and aromatic PICs, including PCDD/ PCDFs, during incineration of a HCFC and a CFC recycling residue. ------- Results and Discussion Task 1 consisted of three repetition tests, Tests 1 to 3 (see Table 1). PCDD/ PCDFs were not detected in Test 1, the TFR system blank, indicating no back- ground contamination in the propane fuel, sampling train, recovery agents, or result- ing from the analytical procedure. Tests 2 and 3 were conducted under similar con- ditions as the FY 91 test in which high levels of PCDD/PCDFs were detected. The two tests were conducted on consecutive days under similar facility operating condi- tions in order to measure the repeatability of the experiment. The high temperature sample (Location 1 at the afterburner out- let) for Test 3 was not analyzed due to loss of sample prior to laboratory analy- sis. Low levels of PCDD/PCDFs (5.4 ng/ dscm) were detected at the high tempera- ture sampling location sample collected during Test 2. The PCDD/PCDF concen- trations measured at the downstream sam- pling location (Location 2) were higher (117.8 ng/dscm for Test 3). However, these measured PCDD/PCDF concentrations were much lower than the high levels mea- sured in FY 91 (23,830 ng/dscm). Excel- lent combustion conditions were achieved for all test conditions. In all cases, less than 50 ppmv of CO, corrected to 7% O2, was detected in the combustion flue gas. Prior to Task 2, the TFR facility was conditioned by incinerating a copper-con- taining fuel and by injecting a copper- containing aqueous solution. The copper-containing fuel was fired in the TFR primary combustor, and the copper-con- taining solution was injected into the pri- mary zone during natural gas combustion. During combustion of the copper-contain- ing fuel and injection of the copper-con- taining solution, the primary combustion zone of the TFR was fired at approxi- mately 30 kWand the afterburner was not fired. For Task 2 (see Table 1), Tests 7 and 8 were CFC-12 incineration tests after cop- per had been added into the test facility. Results show that higher levels of PCDD/ PCDFs were formed under these condi- tions. In Test 7, the PCDD/PCDF concen- trations at the afterburner outlet (Location 1) and at the low-temperature sampling location (Location 2) were 118 and 454 ng/dscm, respectively. During Test 8, PCDD/PCDF formation was not quite as high: results indicated 53.8 and 386 ng/ dscm for the afterburner outlet and down- stream locations, respectively. The major difference between Tests 7 and 8 was the elapsed time between copper injection and CFC-12 incineration which allowed residual copper to be purged from the TFR. Test 7 commenced 3 hours of propane firing af- ter injecting copper into the TFR, while for Test 8, 54 hours of propane firing elapsed following copper injection prior to CFC-12 incineration. For Task 3, Tests 4 to 6 were con- ducted in the CTT (see Table 1). Test 4 was a facility system blank in which only propane was fired in the facility. During Tests, HCFC-141bwas incinerated in the facility at a HCFC/propane concentration of 7.4%. Finally, during Test 6, CFC recy- cling residue was incinerated without any auxiliary fuel. Excellent combustion condi- tions were achieved for all test conditions, including incineration of CFC recycling resi- due with no auxiliary fuel. In all cases, less than 35 ppmv of CO, corrected to 7% O2, was detected in the combustion flue gas. Volatile PICs, determined from the EPA SW 846 Method 0030 sampling trains, were measured for Tests 4 to 6. The con- centrations of volatile organics in the flue gas for all three tests were very low. Dur- ing Test 4 (facility system blank), no vola- tile organic compounds were measured at levels significantly higher than the back- ground levels measured in the field blank. During Test 5 (HCFC-141b incineration), chloroform was measured at an average concentration of 14.5 |ig/dscm. The con- centrations of acetone and methylene chlo- ride were 67.9 and 129 |ig/dscm, respectively; however, the concentrations of these compounds in the associated field blank were nearly the same. During Test 6 (CFC recycling residue incineration), no compounds were measured at significant concentrations. For the analysis in all cases, surrogate recoveries were within acceptable limits. Test 4 was a facility blank test to mea- sure the background contamination of PCDD/PCDFs during propane combustion in the CTT. A moderate concentration of (~ 3 ng/dscm) PCDD/PCDF was detected in the facility blank test, suggesting back- ground contamination from the propane fuel, sampling train, reagents, or analyti- cal procedure. Duplicate analysis was per- formed to confirm these facility blank results. The presence of background con- tamination in the Test 4 measurements suggests that the Test 5 and 6 results may also contain background contamina- tion. Two Method 23 sampling trains collected samples simultaneously at the downstream sampling location during Test 5. Results of the analysis of these samples indicate PCDD/PCDF concentrations of 2 and 11 ng/dscm for the two samples. These con- centrations are relatively low compared to those from CFC incineration. CFC recycling residue was incinerated in Test 6. Two simultaneous Method 23 samples were collected at the downstream sampling location. The concentrations of PCDD/PCDFs during these tests were 52 and 61 ng/dscm. These measured con- centrations are similar to those observed from incineration of pure CFCs in the pre- vious study. Note that the flue gas temperature for Tests 4 to 6 at the Method 23 sampling location (Location 2) was approximately 225°C, which is within the PCDD/PCDF formation "temperature window" of about 200 to 450°C. The residence time be- tween the 1,450°C flame zone tempera- ture and this sampling location was approximately 6.2 sec, more than suffi- cient time for the de novo synthesis of dioxins and furans. However, results from the present study indicate that incinera- tion of HCFC and CFC recycling residue generates very low levels of PCDD/PCDF emissions during this temperature window. The lack of a sufficient level of copper- contaminant present in the HCFC and the CFC recycling residue to promote PCDD/ PCDF formation may be the reason for the observed low PCDD/PCDF emissions. Comparable levels of PCDD/PCDF emis- sions were also observed from incinera- tion of pure CFC-12 in the FY 92 study. Conclusions And Recommendations High concentrations of PCDD/PCDFs (23,800 ng/dscm) measured in a FY 91 study during incineration of CFC-12 in the TFR could not be repeated in the present study. Repetition tests conducted in the same facility under similar operating con- ditions resulted in PCDD/PCDF concen- trations of 118 ng/dscm at 7% O2. Residual copper retained in an incin- eration facility can possibly promote the emission of PCDD/PCDFs during incin- eration of CFC-12 which does not contain copper. Tests conducted in the TFR facil- ity resulted in measured PCDD/PCDF con- centrations of 386 - 454 ng/dscm at 7% O2 during incineration of CFC-12 which followed incineration of copper-containing compounds. Previous studies have shown evidence of PCDD/PCDF formation dur- ing incineration of chlorinated wastes in the presence of trace copper in systems with sufficient residence time in the PCDD/ PCDF temperature formation window. This promotional effect of copper may limit the types of waste materials which can be incinerated prior to incineration of highly chlorinated CFCs. The present results sug- ------- gest that CFCs may best be incinerated in incinerators that do not treat any copper- containing waste materials prior to CFC incineration, in order to eliminate the pos- sibility that residual copper retained in the incineration systems could promote PCDD/PCDF formation during subsequent CFC incineration. Incineration can be used to destroy 1,1- dichloro-1-fluoroethane (HCFC-141 b) with- out generating significant quantities of volatile organic PICs and PCDD/PCDFs. Incineration can also be used to destroy the oily residue (which contains 3 ppm copper) generated during CFC recycling without generating significant quantities of volatile organic PICs or PCDD/PCDFs. Table 1. Test Matrix Task 1 Task 2 Task3 Test Number 1 2 3 7 8 4 5 6 Test Facility TFR TFR TFR TFR TFR CTT CTT CTT Test Material Facility Blank CFC -12 CFC -12 Fuel oil doped with Cu followed by CFC-12* Fuel oil doped with Cu followed by CFC-12* Facility Blank HCFC-141b CFC Residue PCDD/F Sampling Location 1 # Replicates 0 1 1 1 1 0 0 0 PCDD/F Sampling Location 2 # Replicates 2 1 1 1 1 2 2 2 CFC/PIC Sampling # Replicates 0 0 0 0 0 3 3 3 Fuel oil doped with copper was fired in the test facility prior to testing. Immediately following, CFC-12 was incinerated in the facility. Manual sampling for PCDD/PCDF occurred during CFC-12 incineration. G. Kryderand B. Springsteen are with Energy and Environmental Research Corp., Irvine, CA 92718. C.W. Lee is the EPA Project Officer (see below). The complete report, entitled "Experimental Investigation of PIC Formation During CFC Incineration," (Order No. PB96-152186; Cost: $38.00, subject to change) will be available only from National Technical Information Service 5285 Port Royal Road Springfield, VA22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at National Risk Management Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 United States Environmental Protection Agency National Risk Management Research Laboratory (G-72) Cincinnati, OH 45268 BULK RATE POSTAGE & FEES PAID EPA PERMIT No. G-35 Official Business Penalty for Private Use $300 EPA/600/SR-96/007 ------- |