Report

New UEF Emission Factors for SMC Production and Compression Molding David Lipiro Environmental Compliance & Risk Management [email protected] ECRM Terminology • SMC = Sheet Molding Compound • BMC = Bulk Molding Compound • LCM = Liquid Composites Molding • UEF = ANSI/ACMA/ICPA UEF-1-2009, Estimating Emission Factors from Open Molding and Other Composite Processes (Current Version) Factors for VOC Emissions • Compounding (in UEF) – from SMC machines – based on ECRM study • Compression molding (proposed) – SMC, BMC, LCM – based on EECS studies (R. Haberlein) Estimating Emissions • Emission Factor = ratio – Lb emitted/lb compound used = E/C = 0.002 – Lb emitted/lb styrene used = E/S = 0.02 • Equivalent Emission Models – E = 0.002*C – E = 0.02*S – Simplest model constant factor • Derive models and factors from test data Emissions vs Styrene Run 25 E, lb/hr 20 E = 0.2S 15 10 S E E/S 100 19.55 0.20 80 15.44 0.19 60 12.28 0.20 40 8.44 0.21 20 3.89 0.19 10 1.98 0.20 Average EF = 5 0 0 50 100 S, lb/hr 150 0.20 S E E/S 100 25.80 0.26 80 21.29 0.27 60 17.59 0.29 40 13.65 0.34 25 20 8.85 0.44 20 10 7.79 0.78 Emissions vs Styrene Run 40 E = 0.4S 35 E, lb/hr 30 E = 0.2S + 5 Average EF = 15 10 5 0 0 50 100 S, lb/hr 150 0.40 Compounding Emission Model • Theory • Lab weight loss experiments at MFG Research – C. Piper • Evaluate recent VOC emission tests of 7 SMC machines Theory • Areas where paste is not covered by carrier film = window to environment Upper Lower • Exposed area A ft2 is constant for each SMC machine • Simplest model: E lb/hr = kA, k = constant Styrene Content? • Affects skinover? • Expected to be important based on UEF for open molding Line Speed? • Skinover effect at longer travel times lower emissions at lower linespeeds? • Induced air velocity as paste moves under air impact depends on vent airflow patterns? MFG Lab Tests • Wt loss by analytical balance + datalogger, 10 readings per second, 6-8 reps • Constant sample pan area = 0.16379 ft2 • Range of S.G. and styrene % in pastes used • “Wind tunnel” w anemometer – simulate motion-induced airflow • Run times cover realistic line travel times simulate skinover Conclusions/Predictions • Not much skinover effect – most weight loss curves linear or near • Styrene content will probably have little impact on emissions • Line speed probably won’t effect emissions • Emissions will vary with open area, temperature and air velocity - only area practical for permitting Field Verification • Method 24 TTE emission tests • 7 uncontrolled machines – 24”, 48”, 60” • Wide range of open areas, styrene contents, linespeeds • No temperature control at doctor boxes • Doctor boxes open to nearly closed Field Data Analysis • Regression: Emissions vs Predictor • Predictors compared: – Area – Styrene throughput (traditional) – SMC throughput Emission vs Total Wet Area, with Confidence and Prediction Limits (k=3 for Prediction Limits) E = 0.1457A - 0.1454 R2 = 0.9917 18 Emissions, lb/hr 16 Actual 14 12 LCL 10 8 UCL 6 LPL 4 UPL 2 0 0 50 100 Area, ft2 150 Figure 2. Throughput Model 1 y = 0.0009x - 1.773 R2 = 0.4 Styrene Emitted, lb/hr 16 1 14 12 2 10 3 8 4 6 5 4 6 2 7 0 0 5,000 10,000 SMC Run, lb/hr 15,000 Compounding Emission Model in UEF • Er = 0.1457*A-0.1454, where: – Er = emission rate of SMC machine, lb/hr – A = open wet area of machine (including uncovered doctor box area), ft2 • Emissions = Er * t (hrs) • Permitting: E is constant when paste on line - throughput irrelevant Compression Molding • SMC, BMC, LCM • Emissions from charge prep only Once press closes, negligible emissions • EECS Study: Charge prep emissions measured in Method 204 TTE SMC Compression Molding • Two “traditional” EFs proposed based on test data • Which is best predictor? • What are underlying models? Compound Throughput Model Emissions vs SMC Molded E = 0.00148C + 0.10955 R2 = 0.74233 1.6 E = 0.002C Emissions lb/hr 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 100 200 300 400 500 SMC molded, lb/hr 600 700 800 C, lb/hr 16 86 115 229 191 189 306 388 237 319 279 142 628 602 690 690 E, lb/hr 0.027 0.250 0.231 0.540 0.264 0.627 0.651 0.750 0.169 0.672 0.715 0.335 0.736 0.682 1.472 1.200 Average EF = E/C 0.00164 0.00292 0.00201 0.00236 0.00138 0.00332 0.00213 0.00193 0.00071 0.00211 0.00256 0.00235 0.00117 0.00113 0.00213 0.00174 0.0020 Styrene Throughput Model Emissions vs Styrene Molded E = 0.0129S + 0.0646 R2 = 0.8093 1.6 E = 0.0152S 1.4 Emissions lb/hr 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 20 40 60 Styrene molded, lb/hr 80 100 S, lb/hr 2.16 12.75 15.51 31.80 23.52 28.31 49.35 54.00 27.56 38.22 33.55 15.40 67.16 74.07 95.20 73.84 E, lb/hr 0.027 0.250 0.231 0.540 0.264 0.627 0.651 0.750 0.169 0.672 0.715 0.335 0.736 0.682 1.472 1.200 Average EF = E/C 0.01250 0.01961 0.01489 0.01698 0.01122 0.02215 0.01319 0.01389 0.00613 0.01758 0.02131 0.02175 0.01096 0.00921 0.01546 0.01625 0.0152 EFs Compared • E/C = 0.002 is poorer predictor (R2=0.74, 10% total overestimate) • E/S = 0.0152 is better predictor (R2=0.81, 5% total overestimate), same form as published AP-42 factor • UEF: E lb = 1.5% styrene lb run vs 2% AP-42 factor BMC and LCM • UEF: BMC Molding Emissions = 1.15% of styrene in compound molded • UEF: LCM EF, lb styrene / lb paste molded = – Spread prep: 0.0072*S% + 0.008 – Pour prep: 0.0022*S% + 0.008 – LCM Emissions = EF * lb paste molded Why Use These Factors? • Best available methods to estimate emissions w/o site testing • Most credible factors available – adopted into UEF by transparent ANSI consensus process involving industry and state/fed regulators Thank you for attending Questions ECRM