Challenges of Ultra-low NOx Measurements

Transcription

1 Challenges of Ultra-low NOx Measurements 7th Portable Emissions Measurement System (PEMS) Workshop March 30-31, 2017 Presented By: Dr. Kent Johnson Co-Authors Yuwei Han, Thomas D. Durbin, Georgios Karavalakis, Yu Jiang, Jiacheng Yang, and David Cocker III (951) University of California, Riverside Center for Environmental Research and Technology (CE-CERT)

2 Outline Background Approach Results Summary 2

3 Advanced Natural Gas Engines are a Success California Optional Standards: 0.02 g/bhp-hr SCAQMD-managed $5M program to develop a 0.02 g/bhp-hr engine Cummins-Westport ISL G NZ 8.9L Engine: 3

4 Approach Refuse Truck: 0.02 g/bhp-hr Low NOx HD NG Engine Rated Power Disp. Adv NO x Std PM Std. Mfg Model Year Eng. Family rpm) (liters) g/bhp-h 1 g/bhp-h CWI ISL G NZ 2014 ECEXH0540LBH The family ECEXH0540LBH is on the engine label given its year of manufacture. The engine tested was produced under the ECEX label but was later certified and upgraded to the GCEX label. The engine tested is thus, based on the GCEX label and represents a 0.02 g/bhp-hr NO x standard, see Appendix F Figure 4 for details. Model year 2016 certified to 0.02 g/bhp-hr Vehicle equipped with a automatic transmission Day Distance Average Speed Duration Near Dock Local Regional UDDSx CBDx AQMD Refuse

5 Test Performed On UCRs Chassis Dyno With the Mobile Emissions Laboratory (MEL) Added acid treated PM filters for high ammonia emissions Added three advanced NOx measurement methods Measured Global Warming Potential Measured PM and fines

6 Approach: Measurement Upgrades Type Analyzer Meth. ID Description Traditional 600 HCLD dil 600 HCLD amb M1 Modal NO x with ambient bag correction Traditional 600 HCLD dil Dilute bag NO x with ambient bag M2 600 HCLD amb correction Upgrade 300 HCLD raw M3 Raw NO x no ambient bag correction Upgrade Upgrade 600 HCLD dil TECO amb TECO dil TECO amb M4 M5 Modal dilute NOx with ambient real time correction Trace analyzer dilute bag with trace ambient bag correction TECO - Trace level chemiluminescence NO-NO2-NOx analyzer modal 42C manufactured by Thermo Environmental Instruments Inc. Calibrated to 600 ppb. nn NNNN xx_mm1 ccccssii _ii ) ρρ mm CC aa 1 1 _mm2 CC aa 1 1 xx_mm3 xx_mm4 (QQ eeeehii ssii _ii ) ρρ NNNNxx CC CC rr_ii mm_ii CC DDDD _ii 1 xx_mm5 = cccccc_aaaaaa tt) ρρ NNNNxx CC dd_aaaaaa CC aa_aaaaaa 1 1 DDDD _aaaaaaii DDDD DDDD _aaaaaaii ii ii=1 From Johnson et al (2016), Final report to Cummins West Port via SC-AQMD funding In-Use Ultra-Low NOx Natural Gas Vehicle Evaluation ISL G NZ 8.9, Feb

7 Comparison of Method Calculations nn NNNN xx_mm1 = (QQ ccccssii tt _ii ) ρρ NNNNxx CC mm _ii CC aa 1 1 DDDD ii ii =1 NNNN xx _mm2 = (QQ cccccc _aaaaaa tt) ρρ NNNNxx CC dd CC aa 1 1 DDDD _aaaaaa nn NNNN xx_mm3 = (QQ eeeehii tt _ii ) ρρ NNNNxx CC rr_ii ii =1 nn NNNN xx_mm4 = (QQ cccc ssii tt _ii ) ρρ NNNNxx CC mm_ii CC aa_aaaaaa _ii 1 1 DDDD ii ii=1 NNNN xx_mm5 = (QQ cccccc_aaaaaa tt) ρρ NNNNxx CC dd_aaaaaa CC aa_aaaaaa 1 1 DDDD _aaaaaa 7

8 Source and Ambient Concentrations Are Converging Percentile Amb Dilute Raw Amb/Dil % 10th % 50th % 90th % 8

9 Method 3 (Raw Sampling) Shows Lowest Variability Method t-test f-test M M M M o Traditional methods lower signal/noise, reasonable means o M4 showed poor repeatability and M5 low response 9

10 Meets Optional CA Standard And Even Lower For Lower Power Duty Cycles o NOx emissions drop with decreasing load o The weighted NO x hot/cold emissions g/bhp-hr 10

11 Emissions Spikes are Significant o Spikes show most of the emission increases o Variability is not the measurement, but the vehicle 11

12 Up to 90% of The NO x in One Acceleration 12

13 Can PEMS Capture Deviations In the Field? I/M? 13

14 If We Zoom In 90% What Does It Look Like 14

15 Lets Zoom In Another 90% (99%) What Do We See 15

16 Okay Lets Zoom in Another 75% (99.47%) 16

17 The Variability Is Real and Results from Shifting 17

18 Accumulated NOx (g) Cold Start Emissions are Significant 0915 g 0719_CS g 0915 ExhTemp C 0719 ExhTemp C Time (seconds) o 90% of the NO x emission result in fist 100 seconds. o PEMS can easily capture this large NOx spike. RSD? o Hybrid applications need to manage cold starts Exhaust Temperature (C) 18

19 With On Board Sensors Can we Estimate NO x from Concentration? NOx table Vehicle Cycle ppm bsnox Diesel 2008 FTP Diesel SCR FTP NZ NG CS_UDDSx NZ NG UDDSx NZ NG DPT

20 Summary Thoughts: Measurements o Raw NOx concentrations stabilized between 0.3 and 0.1 ppm where ambient NOx ranged from 0.23 to 21 ppm (10 th and 90 th percentile) o NOx spikes ranged from 700 to 60 ppm. One spike > 90% of the emission level at 0.02 g/bhp-hr. o NG engines with acceleration issues may not be captured by I/M, but would be captured by PEMS or an telematics reporting system (Air-To-Fuel) ratio o Vehicle ATS remained active after light off (unlike diesel ATS). Great for hybrid applications. 20

21 Summary Thoughts: Emission Levels o The 0.02 g/bhp-hr Optional NOx standard was a success. o The NG ISLG NZ vehicle showed lower emissions as load decreased (an additional 90% g/bhp-hr). o Real world driving, aged systems, and fuel variability may show higher emission levels. Great system to hybridize. o Other Emissions (see the report) o o o o HC emissions below certification (closed crank case design) Particulate mass emissions are very low (at method detection) Black carbon emissions are very low (at method detection) Particle number emissions are low 21