The Impact of Oil Consumption Mechanisms on Diesel Exhaust Particle Size Distributions and Detailed Exhaust Chemical Composition

Transcription

1 The Impact of Oil Consumption Mechanisms on Diesel Exhaust Particle Size Distributions and Detailed Exhaust Chemical Composition John Stetter, Nate Forster Jaal Ghandhi, David Foster University of Wisconsin-Madison DEER Conference, August 24, 23, Newport RI

2 Acknowledgements Army Research Office (ARO) Nippon-Mitsubishi Oil Co. BP-Amoco Co. Yanmar Co. Cummins Engine Co. Lubrizol Co.

3 Introduction Strategies for Reducing PM Aftertreatment Systems Catalytic Diesel Particulate Traps Particulate oxidation rate depends on its composition and history Fuel Composition Ultra Low Sulfur Diesel Fuel Synthetic Diesel Fuel (i.e. Fischer-Tropsch Fuel) Impact on EC, OC, trace metals, sulfates Lubrication Oil Composition and Variables Viscosity, Volatility, and Sulfur Content Engine Design Electronic Engine Control Fuel Injection System Intake Air System (Turbo-charging, etc.) Combustion Chamber Modification Lubrication Oil Consumption Reduction This study Studies on the impact of engine operating conditions on particulate characteristics are underway

4 Introduction Oil Consumption Mechanisms 1. Blowby Return PCV (Effect not studied) 2. Migration of oil past valve stem seals 3. Migration of oil past piston rings 4. Turbocharger leakage (Effect not studied) Adapted from SAE

5 EXHAUST Experimental Setup Test Engine Engine Bench Setup Engine Specification Engine Type Cummins N14 Single Cylinder Diesel REGULATOR BUILDING AIR ARROW PNEUMATICS LARGE CAPACITY FILTER OILESCER FILTER CRITICAL FLOW ORIFICES BUILDING AIR MICROMOTION FILTER PUMP FILTER REGULATOR HEAT EXCHANGER H2O TANK STRAIN GAGE AMPLIFIER FUEL TANK CHARGE AMPLIFIER PUMP HEATER INTAKE SURGE TANK ECM EXHAUST SURGE TANK FLOW ORIFICE TO FTIR HEATED FILTER FROM ENCODER FROM OPTICAL INTERRUPTER ENGINE AND DYNAMOMETER EXHAUST FULL DILUTION TUNNEL Cycle Combustion Chamber Piston Chamber Number of Intake Valves Number of Exhaust Valves Compression Ratio Swirl Ratio Displacement Bore Stroke Combustion Chamber Diameter Connection Rod Length 4-stroke Quiescent Shallow Dish : cc mm mm 97.8 mm 34.8 mm NI DAQ ELECTRONIC INJECTION CONTROL SYSTEM (SHAFT8) Piston Pin Offset Injection System None Unit Injector, Direct Injection (DI) Nozzle Dimension 8 Φ.2 mm Length/Diameter of holes (l/d) 4.1 Spray Angle 152

6 Experimental Design Experimental Operating Conditions CARB 8-Mode Test Points Steady-State Operation Fuel was a 26 EPA low sulfur fuel, 14 PPM Engine Torque/Cylinder [Nm] Mode 8 (Idling) Peak Torque Mode 5 (1% load) Mode 6 (75% load) Mode 7 (5% load) Mode 4 (25% load) Mode 1 (1% load) Mode 2 (75% load) Mode 3 (5% load) Engine Speed [rpm] Rated Speed Mode 1 Mode 2 Mode 3 Mode 4 Mode 5 Mode 6 Mode 7 Mode 8 Speed [rpm] Load [%] (idle) Remark Rated Peak speed torque Intake T [ C] Intake P [kpa] Exhaust P [kpa] SOI, CA atdc [degrees]

7 LOC Rate Measurement Method Calcium: a viable LOC tracer (SAE and SAE ) Calcium compounds are common additives Other metals in oil (Fe, Al, Cu, etc.) can also be attributed to engine wear. Ca and Zn show good agreement. LOC Rate Calculated by Zinc [g/hr] LOC Rate Calculated by Calcium [g/hr]

8 Lube Oil Consumption 12% 1% Exhaust Lube Oil Consumption Variation from Base Condition Mode 5 - Peak Torque 12 rpm Mode 1 - Peak Power 18 rpm 8% 6% 4% 2% % -2% -4% No Intake VSS No E or I VSS No E or I VSS No Exhaust VSS Base 5% OCR 5% OCR -6% -8%

9 Lube Oil Consumption Measured LOC rates: g/hr (Mode 4 LOC <.1 g/hr) Typical LOC (High Speed Diesels):.1-.5% of full load fuel which would correspond to 7 4 g/hr Test points are steady state conditions

10 Transient Lube Oil Consumption EC OC P- Fe+ Ca+ Mn+ 11:47: 11:47:3 11:48: 11:48:3 11:49: 11:49:3 11:5: 11:5:3 11:51: 11:51:3 11:52: 11:52:3 11:53: 11:53:3 11:54: 11:54:3 11:55: 11:55:3 11:56: 11:56: Timeline 3152m415TR BP-ARCO fuel Time EC OC P- Fe+ Ca+ Mn+ # of particles (Ca+, Fe+, Mn+, P-, EC) 17:1: 17:1:3 17:2: 17:2:3 17:3: 17:3:3 17:4: 17:4:3 17:5: 17:5:3 17:6: 17:6:3 17:1: 17:7: 17:7:3 17:8: 17:8:3 17:9: 17:9:3 17:1:3 Transient calcium spikes on order of 1x steady state operation from raw ATOFMS data Timeline 3152m325TR Mn+ 4 P- Fe+ 35 Ca+ EC 3 Ca OC 25 Ca BP-ARCO fuel # of particles (Ca+, Fe+, Mn+, P-, EC) 1 5 Time Mn+ P- Fe+ Ca+ EC OC

11 Lube Oil Consumption Measured LOC rates: g/hr Typical LOC (High Speed Diesels):.1.5% of full load fuel which would correspond to 7 4 g/hr Test points are steady state conditions Transient operation critical to understanding overall LOC, but What can be learned from less complex, fundamental steady state operation?

12 Engine Out PM 5% Engine Out Particulate Matter Variation from Base Condition Mode 5 - Peak Torque 12 rpm Mode 1 - Peak Power 18 rpm % -5% No Intake VSS No E or I VSS No E or I VSS No Exhaust VSS Base 5% OCR 5% OCR -1% -15% -2%

13 Reconstructed Mass vs. PM2.5 Mass Very Good Agreement 8 Reconstructed Mass = EC+1.2*OC+sulfates Recon > Measured for low loading This may be due to sorption of semivolatile gas-phase organic compounds to the filter. Reconstructed Mass [mg/m3] :1 Slope Filter PM2.5 Mass [mg/m3]

14 Particulate Matter Composition Mode 4 25% Load 12 rpm EC 16% SO4.33% Ash.1% OM 83% Varying composition with engine mode Fairly consistent (±5%) composition among variations in LOC for each engine mode Mode 5 1% Load 12 rpm Mode 1 1% Load 18 rpm EC 61% SO4.13% Ash.13% OM 38% EC 85% SO4.7% Ash.2% OM 14%

15 EC OC LOC SO Sulfate Detection Limit Emission Rate [g/hr] No Intake VSS No E or I VSS No E or I VSS No Exhaust VSS Base 5% OCR 5% OCR m4 EC, OC, and Sulfates with LOC Experimental Results EC OC LOC SO Emission Rate [g/hr].75 m5.5 m1.25 Modes 5 and 1 show significant differences in OC emissions Differences do not correspond with variations in lube oil consumption EC OC LOC SO Emission Rate [g/hr] No Intake VSS No E or I VSS No E or I VSS No Exhaust VSS Base 5% OCR 5% OCR No Intake VSS No E or I VSS No E or I VSS No Exhaust VSS Base 5% OCR 5% OCR

16 OC vs. LOC for Mode 5 and Mode 1 Mode Mode Organic Carbon, OC [g/hr] No EVSS No E or I No E or I No IVSS Base 5% OCR 5% OCR Organic Carbon, OC [g/hr] No E or I No E or I No EVSS Base No IVSS 5% OCR 5% OCR Lube Oil Consumption, LOC [g/hr] Lube Oil Consumption, LOC [g/hr] No obvious relationship between Organic Carbon and Lube Oil Consumption for Modes 5 and 1

17 Lube Oil Contribution to OC 2% 18% 16% 14% 12% 1% 8% 6% 4% 2% % Potential Lube Oil Contribution to Organic Carbon (assuming no lube oil combustion) No Intake VSS - m4 No E or I VSS - m4 No E or I VSS - m4 No Exhaust VSS - m4 Base - m4 5% OCR - m4 5% OCR - m4 No Intake VSS - m5 No E or I VSS - m5 No Exhaust VSS - m5 No Exhaust VSS - m5 Base - m5 5% OCR - m5 5% OCR - m5 No Intake VSS - m1 No E or I VSS - m1 No E or I VSS - m1 No Intake VSS - m1 Base - m1 5% OCR - m1 LOC [g/hr]/oc [g/hr] 5% OCR - m1 Mode 4 Up to 42% Mode 5 Up to 32% Mode 1 Up to 1 + %

18 Filter and SMPS Correlations No Clear Correlation PM [mg/m3] OC [mg/m3] EC [mg/m3] SO4 [mg/m3] Lube Oil [mg/m3] No Clear Correlation PM [mg/m3] OC [mg/m3] EC [mg/m3] SO4 [mg/m3] Lube Oil [mg/m3]

19 Filter and SMPS Correlations No Clear Correlation -.25 PM [mg/m3] OC [mg/m3] EC [mg/m3] SO4 [mg/m3] Lube Oil [mg/m3] E+7 1.E+7 1.E+7 1.E+7 9.8E+6 9.6E+6 9.4E+6 9.2E+6 y = 1.18E+6x + 1.6E+6 R 2 = E+7 1.E+7 1.E+7 1.E+7 9.8E+6 9.6E+6 9.4E+6 9.2E+6 y = 1.7E+6x E+6 R 2 = E+6 9.E Dilution Tunnel PM Concentration [mg/m3] Dilution Tunnel OC Concentration [mg/m3] Particle 146 nm dn/dlog(dp) [#/cm3] Particle 146 nm dn/dlog(dp) [#/cm3]

20 Filter and SMPS Correlations No Clear Correlation PM [mg/m3] OC [mg/m3] EC [mg/m3] SO4 [mg/m3] Lube Oil [mg/m3] Particle 181 nm dn/dlog(dp) [#/cm3] 5.8E+6 5.8E+6 5.7E+6 5.7E+6 5.6E+6 5.6E+6 5.5E+6 5.5E+6 5.4E+6 5.4E+6 y = 8.32E+5x + 2.2E+6 R 2 =.762 Particle 29 nm dn/dlog(dp) [#/cm3] 4.1E+6 4.1E+6 4.E+6 4.E+6 3.9E+6 3.9E+6 3.8E+6 3.8E+6 y = 8.17E+5x + 1.4E+6 R 2 = E Dilution Tunnel PM Concentration [mg/m3] 3.7E Dilution Tunnel EC Concentration [mg/m3]

21 Filter and SMPS Correlations No Clear Correlation PM [mg/m3] OC [mg/m3] EC [mg/m3] SO4 [mg/m3] Lube Oil [mg/m3] E+6 1.6E+6 Particle 289 nm dn/dlog(dp) [#/cm3] 1.6E+6 1.5E+6 1.5E+6 1.5E+6 1.5E+6 1.5E+6 1.4E+6 1.4E+6 1.4E+6 y = -2.18E+7x E+6 R 2 =.797 Particle 289 nm dn/dlog(dp) [#/cm3] 1.6E+6 1.5E+6 1.5E+6 1.5E+6 1.5E+6 1.5E+6 1.4E+6 1.4E+6 1.4E+6 1.4E+6 y = -2.67E+5x E+6 R 2 = E+6 1.4E Dilution Tunnel SO4 Concentration [mg/m3] Dilution Tunnel Lube Oil Concentration [mg/m3]

22 Particulate Matter Effective Density Some effective density data available from various instruments Assume empirical density function of the similar functional form Spherical particles PM = Conc * Vol * ρ eff Effective Density [g/cm3] Mode 5 1% Load Mode 1 1% Load DMA-ELPI DMA-ELPI DMA-ELPI 2 DMA-ELPI 22 DMA-APM 22 DMA-APM 23 1% Load DMA-APM 23 5% Load DMA-APM 23 75% Load Mobility Diameter, Dp [nm]

23 Particulate Matter Effective Density Effective Density [g/cm3] Mode 5 1% Load Mode 1 1% Load DMA-ELPI DMA-ELPI DMA-ELPI 2 DMA-ELPI 22 DMA-APM 22 DMA-APM 23 1% Load DMA-APM 23 5% Load DMA-APM 23 75% Load Mobility Diameter, Dp [nm] 2 nm 2 nm

24 SMPS Size Distributions 23529m53-5% Oil Control Ring 12.8 dn/dlogdp [#/cm3] Average Conc [#/cm3] Mass Conc p=var [mg/m3] Mass Conc p=1.2 [mg/m3] m Mass Concentration [mg/m3] Highest LOC but no significant nanoparticle formation (nuclei mode) Constant density (SMPS) overestimates filter mass m125-5% Oil Control Ring.1 dn/dlogdp [#/cm3] Midpoint Diameter, Dp [nm] 1 12 samples shown for each hour test with average and 1 standard deviation shown Data for comparison to gravimetric filters Average Conc [#/cm3] Mass Conc p=var [mg/m3] Mass Conc p=1.2 [mg/m3] m Mass Concentration [mg/m3] Midpoint Diameter, Dp [nm]

25 Conclusions (Stead State Operation) Lube oil consumption for steady state operation has insignificant effect on diesel particulate matter Engine operating conditions have significant effects on the detailed particulate matter composition Lube oil has a small contribution to overall organic carbon for some operating conditions Effective density calculations provide additional insight in understanding the details of diesel PM. No apparent changes in particle size distributions with changes in oil consumption

26 Questions?