A Study of the Emissions from Diesel Vehicles Operating in Beijing
Contents
1. ooo ea flash gps data samplel 155554 log flash gps data sample2 155570 loq 0 SAVE RESULTS Time Offset hrs CALCULATE Display Powerin Bins Caicuiation Hesults 4 3 4 5 6 7 8 Se e BE eee Te ke Fee a ale STG SE 001 0 0 01 0 01 0 06 0 07 044 062 154 292 592 55 69 15 25 975 5 26 1 54 International Vehicle Emissions Mod 31 Calculation Location Z Fleet Base Adjustments Location Fleet InUseTestCycle x Enter the data manually for the 60 Day Month en Day ofthqweek Al bins and the average velocity in 14 w September v 2004 Tuesday v the Location Page of the IVE Al dt 27 C SD F cus APP model 0 0 Lead Pb Benzene Oxygenate none v moderate 150 0 r Fuel Characteristics Overall Sulfur S Gasoline moderate premixed moderate 300 Overall Sulfur S Diesel moderate y moderate 68 r Hour 0 00jall Y W Use this hour Driving Characteristics Humidity Brenn pe VSP Bins Soak Bins 7 50 0j 0 0 kilometers 0 0 fi re re Temperature l Group 1 Group 2 I i VSP Bhi VSP Bin2 VSP Bind VSP Bind VSP BAS VSP Bin6 VSP Bin 7 VSP Bind VSP Bind VSP Bin 10 01 J i 07 44 62 1 54 2 92 a VSP E VSP Bin 12 VSP Bin 13 VSP Bin 16 VSP Bin 17 VSP Bin 18 VSP Bin 19 VSP Bin 20 Average Velocity 55 69 15 25 26 1 54 12 03 og 17 20 kmih2. Diesel emissions are important contributors to air quality degradation in urban areas Diesel particulates are considered to be carcinogenic or likely carcinogens in the United States and diesels are often the prime source of nitrogen oxide emissions Itis thus important that diesel emissions be well understood and that air quality planners be able to predict the impact of diesels in the present time and at times in the future based on specific control scenarios To support these efforts a process of on road measurement of diesel emissions has been devised and the International Vehicle Emissions IVE model was developed to estimate emissions from diesel vehicles under different driving and control scenarios The IVE model is designed to make estimates of in use vehicle emissions in the full range of global urban areas At the point in time of IVE model development data to establish base emission factors and driving pattern adjustments were of necessity based on vehicle studies carried out primarily in the United States This has raised questions as to the applicability of the base emission rates and driving pattern adjustments used in the model to developing countries The IVE in use vehicle emissions study is designed to test the hypothesis that similar vehicle technologies will produce equivalent emission results in a given location and to provide some rudimentary data for creating improved emission factors The IVE modeling framework provide
3. device was developed by ISSRC to use in on road emissions testing with the DMM Figure 1 illustrates the design of the ISSRC field dilution unit Pressure Tap for Flow Measurement NE From Exhaust d Micro Filter 99 9 PM2 5 removal Diluted Exhaust to Dekati Sampler 10 liters minute Dilution Control Figure 1 ISSRC Field Dilution Device The exhaust flow in the dilution device is measured using a Dwyer differential pressure transducer which is accurate to 0 25 of full scale The differential pressure gauge is used to measure the pressure difference between P2 and P3 shown in Figure 1 A micro filter produces particle free air to be diluted with the exhaust sample The exhaust sample and dilution air are heated to 110 degrees C to avoid water and organic condensation The dilution level reaches values from 20 to 1 to 30 to 1 depending on the vehicle tested and the exhaust flow rate Figure 2 presents a flow diagram for the overall emissions testing system The data collected by Vehicle Exhaust peated Sample Line Sensor D Gas Measurement Flow Measurement Device Device Dekati Partical Measurement Device Figure 2 Flow Diagram for the Overall Emissions Testing System The flow measurement device and the Sensor D are recorded to a flash card on the Sensor D unit The Dekati information is recorded to a laptop computer that is connected to the Dekati by a serial cable Figures 3 and 4 show the exterior of
4. n 1 I 1 where Tyo 70 confidence interval o standard deviation n sample size In the case of vehicular emissions o is often close to the sample mean although diesel vehicles show a little less variation that do gasoline fueled vehicles In this case Equation II 1 becomes Tom M V n 1 11 2 where M measured sample mean In this special case to insure that the measured mean has a 70 probability of being within 15 ofthe actual mean requires a sample size of 44 vehicles The actual variation in emissions from similar technologies is unknown but based on the previous discussion it is clear that a relatively large number of vehicles of a given technology class should be tested For purposes of this study at least 10 vehicles of each important technology should be measured to insure even a moderate level of confidence in the results Once the data is collected the variance in the data of similar vehicles will establish the true confidence that can be placed in the results of the tests 23 DE DALE With present vehicle testing technology about 1 5 hours are required to remove the equipment from one vehicle and install the sampling equipment in a second vehicle To collect data for of an hour for a given vehicle will thus require a total of 2 25 hours per vehicle In a 9 hour day about 4 vehicles can be sampled In a two week sampling period 10 sampling days about 40 vehicles can be studied Thus based on the precedi
5. 0 buses appear to be skewing the results There is some relationship between mileage and emissions NOX emissions seems to go down as the mileage increase but MP and HC emissions are higher as the mileage increases CO emissions show higher emissions on the middle range of mileage and lower in the high range of mileage 500 000 5 450 000 400 000 x lt 80 000 80 000 lt x lt 160 000 350 000 4 300 000 250 000 4 200 000 Mileage Km 150 000 4 100 000 4 50 000 ween 0 T T T T T T T T T 1 Figure 13 Vehicles Distribution According to Mileage 13 12 0 mx lt 80 000 x lt 160 000 lt 80 000 x lt 160 000 10 0 8 0 8 0 x QD oa 6 3 m 6 0 c 60 in 5 2 4 9 47 LLI 4 0 3 5 2 8 2 4 2 4 2 3 2 0 q 1 6 1 6 1 4 0 9 0 0 CO THC NOx PM 10 CO2 100 Figure 14 FTP Normalized Emissions Averaged Over Vehicle Mileage Vehicle Emissions Under Different Driving Conditions Another purpose of this study is to determine how emissions vary under different driving conditions These conditions can be represented by the IVE driving bin designations The IVE model divides the range of driving situations into 20 vehicle energy demand situations and 3 engine stress situations Figures 6 7 and 8 present emissions from the large diesel vehicles as a function of IVE driving bin The energy demand on a ve
6. Average Velocity for Group 2 Vehicles Driving Style Distribution Facility Cycle Distribution Group 2 Vehicles VSP Bin I VSP Bin 2 VSP Bin 3 VSP Bin4 VSPBin5 VSP Bin6 VSP Bin 7 Sec seconds min minutes hr hours Mhr 1000 s of hours Ikm kilometers Mkm 1000s of kilometers mi miles Mmi thousands of miles I Soak Time Distribution Group 2 Vehicles IS single units M 1000 s I IC degrees Centigrade F degrees Fahrenheit I m s meters second mph miles per hour km hr kilometers hour I Description Driving Style Distribution Facility Cycle Distribution Group 1 Vehicles Time Period VSP Bin I VSP Bin 2 VSP Bin 3 VSP Bin 4 VSP Bin 5 VSP Bin6 VSP Bin 7 1 hour 2 hour 3 hours 4 hours 6 hours Total Distance or Time Driven Number of Statups Soak Time Distribution Group 1 Vehicles Temperature Relative Humidity 1 hour 2 hour 3 hours 4 hours 6 hours Average Velocity for Group 1 Vehicles Average Velocity for Group 2 Vehicles Driving Style Distribution Facility Cycle Distribution Group 2 Vehicles POTTI GITT nts TTTTTET VSP Bin1 VSPBin2 _VSPBin3 __VSPBin4__VSPBin5 VSP Bin 6 VSP Bin 7 sec seconds min minutes hr hours Mhr 1000 s of hours km kilometers Mkm 1 000s of kilometers mi miles Mmi thousands of miles IS single units M 1000 s IC degrees Centigrade F degrees Fahrenheit m s meters second mph miles per hour km hr kilometers hour I Figure VIII 3 Location File Template in Excel Soak Time Distribu
7. Base Correction Factors in the IVE model You can use the emissions data collected in the field study to derive base or emission correction factors in the IVE model These emission factors will apply a correction to the IVE emission rates already used in the model To calculate the change in emissions from the IVE default emissions to the new measured emissions emissions will need to be predicted on the same driving trace as the emissions measurements were made on This means a driving trace for the overall driving conducted during the emission measurement test procedure will need to be created and input into the IVE model The GPSEvaluate Program has already made a composite set of data with fractions of driving in each bin that add up to 100 This data can be entered into the IVE model in one of two ways The data can simply be entered into the IVE model directly into the location page Figure VI 2 While this is a simple option for entering in data for a single file this can be time consuming for many data files or many hours of the day For multiple files the import function in the Location File Template may be used Figure VI 3 To use the Location File Template follow the instructions on the first spreadsheet in the workbook or refer to the GPS Operating Instructions document 32 EVALUATE GPS DATA JOE File Edit Select Files to Evaluate BusDataExample txt flash gps data 040825 233826 loa M e v vw
8. EXPECTED VARIATION IN TEST DATA FOR REPEATED DRIVING CYCLES rrarrersrsssssrssvsvvnnnnvvnrveseseee 5 TABLE 3 EMISSION MEASUREMENT RESULTS FOR THE TESTED BEIJING DIESEL FLEET nennen 6 TABLE 4 EMISSION MEASUREMENT RESULTS FOR THE TESTED BEIJING DIESEL FLEET CONTINUED scccssseeseeeseeeeseeaes 7 TABLE 5 POTENTIAL ADJUSTMENT VALUES TO BE USED IN IVE MODEL w cccescccccssssseeceeseseceecsssscececeessseeeeceesssssecesseeees 17 TABLE 6 RECOMMENDED ADJUSTMENT VALUES FOR USE IN THE IVE MODEL ccessssccceesssseeeceesseececeessseeecceesseeeecesesaees 18 TABLE 7 EMISSIONS RESULTS FOR AN AVERAGE FLEET IN SEVERAL CITIES cccssssccceessssececcesescecesssseeeeceesssceeeceesnsssesenssaees 18 List of Figures FIGURE 1 ISSRC FIELD DILUTION DEVICE rrrrrnnnvrrorrrennnnrrrsrernnnereennnnrressennnnrrnssensnnnrnsennnnrsnssensnnrsnssensnnrnssseennssssennnnnsensnennnnsenene 3 FIGURE 2 FLOW DIAGRAM FOR THE OVERALL EMISSIONS TESTING SYSTEM c cccccccssssccccesesseccesessseeccesessseeeeeseesaaees 3 FIGURE 3 EXTERIOR OF BUS OUTFITTED FOR EMISSIONS TESTING cccssssscecceeesssceecesssceccesessseecccesesseeeeccesseeeeeesssseeeeeeesaaees 4 FIGURE 4 EXTERIOR OF A TRUCK OUTFITTED FOR EMISSIONS TESTING cc ccccccssssssceceesssseccessssescecessseececessssececesseeseeeensseeees 4 FIGURE 5 EMISSIONS AVERAGED OVER SELECTED MODEL YEARS FOR LIGHT DUTY TRUCKS G KM nennen 8 FIGURE 6 EMISSIONS AVERAGED OVER SELECTED MODEL YEAR
9. Medium Size Trucks and 6 18 are Buses In addition 3 of the vehicles 9 complies the Euro 0 standard 7 vehicles 18 complies Euro 1 standard 17 52 complies Euro 2 and 6 18 complies the Euro 3 standard 20 17 16 15 v 2 10 2 7 gt 6 5 4 4 3 3 3 2 1 0 0 of 00 0 a LD Trucks Mtruck Buses Overall Figure 7 Vehicles Distribution According to Emissions Standard Figure 8 9 and 10 show the emissions results for CO HC NOX PM and CO2 according to the vehicle engine standard for Overall Trucks and finally Buses The Overall results show a reasonable correlation with the standard looking from Euro 1 through Euro 3 standards indicating a reduction between them The Euro 0 vehicles show low levels of emissions for every pollutant but it should be noted that all Euro 0 vehicles were buses It appears that the Euro 0 buses were much better maintained compared to the other vehicles Observing Trucks results there is a relationship between emissions and engine standard except for NOx where there is little improvement between Euro 1 and Euro 3 The sampled buses belongs to Euro 0 and Euro 1 standards groups the results in emissions show that buses complying with the Euro 0 standard were cleaner than the Euro 1 buses In summary the results show a good relationship between emissions and standard except for NOx it has to be notice that Euro 0 buses show a very low range of emissions in comparison with
10. a bus and a truck outfitted for testing Figure 4 Exterior of a Truck Outfitted for Emissions Testing In order to simulate a loaded truck 10 Kilograms of sand in sand bags were placed on the bus or truck according to the weight capacity Depending upon the size of the vehicles sand bags were loaded to simulate 50 of the total load capacity to produce consistent measurements Appendix A contains a description of the overall testing procedure and data processing steps HI Measurement Error No measurement process is free from measurement and operating error Referring to Figure 2 there is the potential for measurement error in the flow measurement process the dilution rate measurement the gas concentrations measurement Table 2 outlines potential error assuming that each process can be held to produce only 2 error to help understand the expected variations in results from the repeated testing Table 2 Estimation of Expected Variation in Test Data for Repeated Driving Cycles Impact on Gaseous Impact on Particle Measurement Process Measurements Measurements Exhaust Volume Flow Measurement 2 2 Dilution Measurement 2 Emissions Concentration Measurement 2 2 Total Potential Variation 4 6 To further complicate the process data is collected on normal city streets which results in different driving patterns from test to test Daily variations in traffic flow have a major impact on how the vehicles
11. ten inches of water column is maintained during the testing The mass flow measurement device will then be attached to the rear of the vehicle using high vacuum suction cups high temperature silicone sleeve sized to match the OD of the tailpipe will be attached to the tailpipe with a hose clamp The silicone sleeve will be attached to flexible silicone transport tubing A second silicone sleeve will be used to attach the other end of the transport tubing to the mass flow measurement device The PEM will be disconnected from the line serviced 12 volt power supply and placed in the trunk or back seat of the test vehicle along with the deep cycle 12 volt battery A 18 foot sample line will be used to connect the mass flow measurement device to the sample input system of the PEM The GPS unit will be magnetically attached to the roof of the vehicle and connected to the PEM The temperature and humidity probe will be located near the front of the vehicle and connected to the PEM Ifthe PEM is placed in the trunk a special piece of hardware will be used to allow the lid of the trunk to be latched but still allow space for the sample line and other lines to be connected to the external devices At this point the PEM will be switched to the measurement mode and the engine of the test vehicle will be started Following completion of the vehicle driving procedure described earlier in Table IV 1 the installation process will be reversed to remove the PEM from
12. that do not represent the usual driving pattern Start hour 0 n a Indicates which hour of the day to start the processing according to the time column or if calculations for each hour should be performed End hour 23 n a Indicates which hour of the day to end the data processing according to the time in the time column This does not apply if hourly calculation has been selected in the start hour column CO column n a Mass sec Indicates which column the CO emissions are in if they exist CO2 column n a Mass sec Indicates which column the CO2 emissions are in if they exist NOx column n a Mass sec Indicates which column the NOx emissions are in if they exist VOC column n a Mass sec Indicates which column the VOC emissions are in if they exist PM column n a Mass sec Indicates which column the PM emissions are in if they exist No Limit on No minutes Indicates the maximum time to allow for idling in the program If Idle Limit this is set to 10 minutes the program would set any idle time defined as a velocity of less than 5m s of longer than 10 minutes to 10 minutes Satellite n a Integer Indicates the column that contains the number of satellites This column through column is optional and is not used in the default configuration If this 14 option is selected program will ignore data that has less than 3 satellites Straight Speed Straight n a Indicates whether to average the current row of data and the previo
13. the vehicle The data collected will be downloaded to a laptop computer at the end of each vehicle test At the end of each day a span check will be conducted to observe the sustained linearity of the system Calibration gases are a critical component of insuring that the measurements by the Semtech D are correct The following table recommends gas concentrations for calibrating and auditing the Semtech D unit The audit gases may be skipped if it is difficult to get gases or if the cost is beyond that budgeted for the project Gas For Unit Calibration For Unit Auditing co2 12 6 co 1200 ppmv 200 ppmv NO 1500 ppmv 300 ppmv Total Hydrocarbons as Propane 200 ppmv 50 ppmv VIII Local Support Requirements The most difficult job for the local partnering agencies is the procurement of the needed 40 vehicles This procedure needs to be started one month or more before the beginning oftesting Each vehicle owner is required to bring their vehicle for testing at least exactly atthe scheduled time In addition a secure location must be found where two large vehicles can be parked and sampling equipment removed from one and installed on another 28 Once the testing is begun a person is needed to insure that the next vehicle will arrive on time and to help with installation of equipment and data downloading An experienced driver is needed to drive the vehicle This should be supplied by the vehicle owner An expe
14. vehicles resulted in 90 confidence interval of plus or minus 20 When all potential errors are combined it should be anticipated that this study will produce results which are to a 90 probability to within 25 30 of the actual emissions produced by the local fleet While this potential error is larger than preferred it is still better than using emission estimates derived from studies in the U S and Europe In the long run more emissions tests are A vehicle will typically not operate in all of the defined Power Bins during a given driving test Since this data is only used to calculate emissions projected for an FTP cycle it is only important to have values for the bins that occur in the FTP cycle This usually occurs In the few cases were a FTP bin is missing then the data is interpolated to fill in values required in order to reduce the mobile source inventory uncertainty to the more preferred 10 range IV Overall Results Table 3 and Table 4 present the average emissions measured for the various vehicles tested in the program They are listed in the order tested An average value and 90 confidence limits are also included at the end of Table 4 The 90 confidence interval in Table 4 indicates the range of emissions for which there is a 90 probability that the true mean emission rate of the fleet would exist if the tested vehicles were randomly selected from the Beijing fleet The vehicles tested should be somewhat representat
15. 1 41 2 59 HC g km 1 06 0 51 0 78 0 04 0 48 0 55 NOx g km 5 67 3 58 5 91 1 10 3 13 5 30 PM 10 g km 2 37 0 92 3 81 0 09 0 42 1 48 CO2 100 g km 4 82 3 36 5 40 2 21 2 85 4 84 In Figure 19 the results are shown with the 90 confidence interval so that one can better understand the variation in the measured data for each city and likely representativeness of the data for each city 18 10 00 5 9 00 8 00 7 00 Beijin E 6 00 1 3 m MexicoCity 2 5 00 im IStanbul 2 E SantiagoLD a 4 00 E SantiagoHD m m Sao Paulo 3 00 2 00 1 00 0 00 co HC NOx PM 10 CO2 100 Figure 19 Comparison of Measured Emission in Several Cities Normalized to the FTP LA4 Cycle and Grouped to Similar Size Distributions After studying the results it can be concluded that CO emissions from the trucks and buses in Beijing are comparable to those emissions found in vehicles measured in Mexico City and Sao Paulo However the vehicles measured in Mexico City and Sao Paulo were significantly larger than the Beijing vehicles tested This indicates that the Beijing vehicles are not very fuel efficient for their size CO HC and NOx emissions most resemble Sao Paulo although the Sao Paulo vehicles tested were significantly larger than the Beijing vehicles tested PM emissions are also higher in Beijing than in any other city than Mexico City This is especially worrisome considering that the Mexico City vehic
16. 2 16 2 18 3 00 3 02 304 3 06 3 08 3 10 3 12 3 14 3 16 3 18 Figure 17 PM Emissions by IVE Bin There is a consistent result for low bins in PM results they have high emissions in bins 1 02 and a 1 05 probably for the same cause that was already discussed with respect to the NOx results Other than that these cases emissions looks normal compared to measurements made in other cities VI Emission Comparisons with the IVE Model As noted earlier 33 buses and trucks were successfully tested in Beijing This limited data does not provide large enough samples of individual technologies to do an analysis of emission comparisons by technology type Instead the IVE model was run using an FTP driving pattern the pattern used to develop base emission factors for the IVE model and using the overall distribution of vehicles tested in Beijing The average measured values normalized to FTP driving cycles were then divided by the IVE predicted values to evaluate the comparisons Figure 19 provides the results of this analysis As can be seen the CO2 emission projections were accurate producing a ratio close to 1 The other predictions however showed a wide variance The model appears to be overestimating the emissions for Euro 0 vehicles it has to be noticed that this vehicles were the buses that showed low values of emissions in every pollutant For Euro 1 Euro 2 and Euro 3 the model is underestimating he emissions for CO HC and NOx The res
17. 2 35 0 53 10 Light Truck 2007 Euro3 519 6 42 0 076 2 31 1 49 616 8 99 0 093 2 84 1 66 12 Light Truck 2007 Euro3 1024 9 24 0 062 0 00 1 72 841 9 44 0 052 0 00 1 25 13 Light Truck 2006 Euro3 421 3 23 0 048 1 35 0 24 563 4 73 0 065 1 84 0 33 14 Bus 1999 Euro 0 475 1 77 0 090 2 51 0 31 574 2 26 0 110 3 07 0 37 15 Bus 1999 Euro0 299 2 43 0 083 0 65 0 05 420 3 36 0 120 0 93 0 08 16 Light Truck 2006 Euro 3 413 7 72 0 042 1 96 1 39 492 10 32 0 052 2 42 1 62 17 Light Truck 2007 Euro3 360 2 80 0 035 1 35 0 23 502 4 13 0 050 1 95 0 34 Table 4 Emission Measurement Results for the Tested Beijing Diesel Fleet continued Measured Emissions FTP Normalized Emissions N he Vehicle Type Year Std grams kilometer grams kilometer Be co2 Nox pm co THC coz Nox PM co THC 18 Bus 2002 Euro 1 503 6 39 0 511 2 43 0 35 673 8 30 0 702 3 33 0 50 19 Bus 1999 Euro 0 391 1 78 0 029 0 62 0 11 512 2 41 0 040 0 83 0 14 20 Light Truck 2004 Euro2 393 4 74 0 127 2 15 1 18 496 6 54 0 165 2 80 1 46 21 Light Truck 2002 Euro 1 405 3 92 0 485 2 98 1 38 551 6 02 0 684 4 20 1 86 22 Light Truck 2001 Euro 1 471 5 96 0 209 3 12 1 47 470 7 51 0 213 3 20 1 30 23 Light Tru
18. A Study of the Emissions from Diesel Vehicles Operating in Beijing China December 2007 Sebastian Tolvett University of Chile Huan Liu Tsinghua University James Lents ISSRC Mauricio Osses University of Chile Kebin He Tsinghua University Table of Contents I INTRODUCTION wisevcccscscstescteesconsesactciaccvessossecescstscvatsovuacesesstesudeecesusdesssdessoesessavesectecssecddveccseseleteetascnosdndedeecesessccess 1 IL TESTING PROCEDURE iccccesscdsssecssscisccsenscsinasceccvesevssclsseeseasscesectedsevessascdbsuesdsucsessesecddsesessccedvesecsesetesvesdccsscoede 1 IN MEASUREMENT ERROR iivsiccstecissiscccssceccleccsesstecseutessucnvessvecssnvecesudcsccetelesvesesscsoovescrecsoudesudesvecsesoncstescusesesvesess 5 IV OVERALL RESULTS siseiscccicccalsssccissnsstcsastsstucsescssessevsctsnccoessoncscetevdsecsessausecssesssucscesesevesvesessccessesssdesccnessececsscesse 6 V VEHICLE EMISSIONS UNDER DIFFERENT DRIVING CONDITIONS eesessesssvvevssesesssnvsessseessesesssnsee 14 VI EMISSION COMPARISONS WITH THE IVE MODEL ssssseeseveessvvvevssnnevsneessnnsessnnsesnsnsessnnsessnnseenssesnnnsee 16 VIL CONCLUSIONS wiccciccscccsscetecosesecesscvecevescstsesevessscusteounsssbectesesteveHscebscosessueceteseusssvecssbecccevestescvecdendecessostecscossssces 18 List of Tables TABLE 1 DIESEL VEHICLES TESTED DURING THE STUDY ccsccccccsssssseeccessssecesessceececesssseececessseeeceessaeeeceeseseeecesenssaeeceeeaeeees 1 TABLE 2 ESTIMATION OF
19. Light Truck Dong Feng Euro 3 2007 128 061 6 1K kg 18 Bus Jiang Huai Euro 1 2002 158 299 6 2K kg 19 Bus Jing Long Euro 0 1999 255 835 6 2K kg 20 Light Truck FORLAND Euro 2 2004 122 446 3 9K kg 21 Light Truck FORLAND Euro 1 2002 146 027 3 9K kg 22 Light Truck FORLAND Euro 1 2001 54 000 3 9K kg 23 Light Truck FORLAND Euro 2 2003 3 9K kg 24 Light Truck FORLAND Euro 2 2005 108 640 3 9K kg 25 Light Truck Dong Feng Euro 2 2005 59 759 4 4K kg 26 Light Truck FORLAND Euro 2 2004 98 169 3 9K kg 27 Light Truck FORLAND Euro 2 2003 207 053 3 9K kg 28 Light Truck FORLAND Euro 2 2003 130 084 3 9K kg 29 Light Truck FORLAND Euro 2 2003 83 392 3 4K kg 30 Light Truck Dong Feng Euro 2 2005 51 754 2 9K kg 31 Light Truck BLAC Euro 1 2001 99 314 4 2K kg 32 Light Truck Jiefang Euro 3 2007 7622 4 0K kg 33 Light Truck FORLAND Euro 2 2004 133 671 3 9K kg 34 Bus Kinglong Euro 1 2000 173 464 6 3K kg 35 Bus JAC Euro 1 2002 177 727 6 2K kg This study was a joint effort of the Tshinghua University and ISSRC II Testing Procedure Vehicles were brought to the test site by drivers supplied by the owners of the vehicles for test equipment installation These vehicles were warmed up at the time of the testing Once emissions testing equipment was installed the vehicles were driven over a prescribed driving circuit by the original vehicle drivers that allowed the vehicles to be operated over as wide a range of operating conditions as could be achieved within the city limits of B
20. S FOR LIGHT DUTY G KM nennen 8 FIGURE 7 VEHICLES DISTRIBUTION ACCORDING TO EMISSIONS STANDARD ccssssscccessssceeceeeseccesenseeeceesessseeeeceessseeeessseeees 9 FIGURE 8 FTP NORMALIZED EMISSIONS AVERAGED OVER EMISSIONS STANDARDS OVERALL s c cccesssssseeceeesseeeesssseees 10 FIGURE 9 FTP NORMALIZED EMISSIONS AVERAGED OVER EMISSIONS STANDARDS TRUCKS csssscceceesssseeeceessssceeesssseees 10 FIGURE 10 FTP NORMALIZED EMISSIONS AVERAGED OVER EMISSIONS STANDARDS BUSES cssescccceessssececeeeseseceessseeees 11 FIGURE 11 VEHICLES DISTRIBUTION ACCORDING TO WEIGHT cccsscccccessssscececcssscecessseeecceseseeeecesessseeceesaseeeecessaseeceesesaees 12 FIGURE 12 FTP NORMALIZED EMISSIONS AVERAGED OVER VEHICLE WEIGHT ccccsscccceesesscecceesseeceeeessseeecceessseeeeesesaees 12 FIGURE 13 VEHICLES DISTRIBUTION ACCORDING TO MILEAGE ccssccccccssssseececcesseccesssseeecceesseeeccecssseecesssseeeecessaeeeceeseauaes 13 FIGURE 14 FTP NORMALIZED EMISSIONS AVERAGED OVER VEHICLE MILEAGE ccsssscccessesssecccesseececesssseeccceensseeeceessaeee 14 FIGURE 15 CO2 EMISSIONS FROM HEAVY DUTY VEHICLES BY IVE BIN ccccceccccccesssssceceeesceceecesssceeeceessssseeceeessseeeeetseeees 15 FIGURE 16 NOX EMISSIONS FROM HEAVY DUTY VEHICLES BY IVE BIN cccceccccccesssssecceeescecceeessceeeceessssseeceeessseeeesseeees 15 FIGURE 17 PM EMISSIONS FROM HEAVY DUTY VEHICLES BY IVE BIN ccccessccccessssscecenssceecec
21. according to Model Year Figure 5 shows the Model Year distribution of the sampled vehicles in this study There were 6 18 vehicles in the range of years 1999 2002 13 39 were in the 2002 2005 range and 14 43 were in the 2005 2007 year range Figure 6 shows the CO HC NOx PM and CO2 emissions from the Light Duty Trucks and emissions for three groupings of model years As can be seen in Figure 6 there is little or no overall relation between emissions and model year for the vehicles tested in China NOx shows a growing trend while PM shows an increase then a decrease in emissions Figure 5 Emissions Averaged over Selected Model Years for Light Duty Trucks g km Model Year 2008 7 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1999 2002 EEE EEE SE EE EE EEE ETE SIGN 2002 2005 2005 2007 99909090 99299 4 9 12 15 18 21 24 27 30 33 Emissions g km 12 0 10 0 8 0 6 0 4 0 2 0 0 0 m 1999 2001 2002 2004 4 2005 2007 7 2 6 3 1 5 6 5 3 5 0 4 6 3 2 3 4 2 7 2 9 4 15 1 7 a i 0 9 CO HC NOx PM 10 CO2 100 Figure 6 Emissions Averaged over Selected Model Years for Light Duty g km Emission Standard Figure 7 shows the distribution of the sampled vehicles 24 vehicles 73 are Light Duty Trucks 3 9 are
22. actor worksheet Figure VI 4 Any time this base correction factor file is used it will correct the emissions predicted by a factor of 1 3 for CO for this specific technology 34 Base Adjustments Base Adjustment file Example Add Technology anFuELTYPES v all FUEL TYPES allAIR FUEL 2 AIR FUEL 118 Pt Auto SmTk Lt MPFI 3Wy PCV L FT my ee 118 Pt Auto SmTk Lt MPFI 3Wy PCV 80 161K km 1 3 t 1 1 1l EE 2 cc KIKI Figure VIII 4 Applying Location Emission Correction Factors in the IVE model 35
23. can be operated Thus emissions can vary considerably from test to test even using the same vehicle To correct for this variation the data is divided into different power demand categories 60 power demand categories are used for this purpose These categories are typically referred to as Power Bins and are numbered from 0 to 59 The amount of emissions that occurs when the vehicle is operated in each of the 60 Power Bins is determined This bin emission rate is multiplied by the driving distribution that would have occurred had the vehicle driven an FTP driving cycle to produce a standardized estimate of emissions that would have resulted if the vehicle had been operated on an FTP cycle This approach was found in an already published study in Brazil to produce estimates of emissions on the actual driving cycle within 6 While this is good in many respects it must be added to measurement uncertainty indicated in Table 2 which results in a potential emissions error of 10 12 overall The error discussed in the previous paragraphs should be random and thus should average out to some degree over multiple tests Finally the number of vehicles that can be tested in a 2 week period is limited This limited testing further decreases the certainty of how well the tested fleet actually represents the actual urban fleet Based on ISSRC data collected in similar gasoline emissions studies the collection of data from a fleet of randomly selected gasoline fueled
24. ck 2003 Euro2 372 6 67 0 172 3 02 1 38 467 9 83 0 224 3 93 1 70 24 Light Truck 2005 Euro2 419 2 90 0 405 3 73 1 21 546 4 07 0 544 5 02 1 59 25 Light Truck 2005 Euro2 443 8 09 0 299 4 24 0 66 538 11 07 0 374 5 30 0 77 26 Light Truck 2004 Euro 2 401 4 17 0 177 2 36 1 45 370 4 82 0 154 2 05 1 18 27 Light Truck 2003 Euro2 306 2 89 0 184 2 00 1 43 399 4 35 0 249 2 71 1 84 28 Light Truck 2003 Euro 2 309 4 57 0 044 1 27 1 08 422 6 72 0 063 1 79 1 49 29 Light Truck 2003 Euro2 272 2 57 0 131 0 49 0 44 367 3 77 0 183 0 69 0 60 30 Light Truck 2005 Euro 2 310 2 81 0 351 1 02 0 40 395 3 92 0 462 1 34 0 51 31 Light Truck 2001 Euro 1 466 4 22 0 047 3 81 2 68 671 6 59 0 070 5 70 4 02 32 Light Truck 2007 Euro3 271 2 66 0 033 0 86 0 30 360 3 87 0 046 1 18 0 41 33 Light Truck 2004 Euro2 303 3 28 0 050 0 93 1 01 391 4 32 0 065 1 20 1 30 34 Bus 2000 Euro1 344 5 42 0 229 2 85 0 50 435 6 99 0 290 3 58 0 63 35 Bus 2002 Euro1 316 4 68 0 467 3 77 2 14 422 6 40 0 623 5 03 2 85 co2 NOx PM co THC co2 NOx PM co THC Average of Tests 403 4 74 0 19 2 04 0 91 497 6 44 0 24 2 64 1 11 90 Confidence Interval 9 14 24 15 20 6 13 25 15 21 Emissions
25. eijing The driving circuit required from 30 to 40 minutes to complete depending upon the traffic situation and included a moderate hill that the vehicles drove over For emission measurement purposes a Semtech Sensor D gas emissions testing unit was used to measure the emissions of CO CO total Hydrocarbons THC NOx and NO The Sensor D unit uses infrared absorption technology to measure CO and CO ultraviolet absorption technology to measure NOx and NO and a flame ionization detector to measure total hydrocarbon emissions The Sensor D testing unit is an integrated emissions testing device designed to be used in on road testing programs The Sensor D measures emission concentrations and must be provided with exhaust flow rates and ambient temperatures and pressures in order to determine mass emission rates The Sensor D is equipped with a temperature pressure sensor A Semtech manufactured 4 inch 10 cm exhaust flow measurement device was used to measure the exhaust flow rate from the vehicles This device uses standard dynamic and static pressure measurement techniques to calculate exhaust flow The Sensor D was also equipped with a GPS device to measure location and speed All data were collected at one second intervals For further information on the Sensor D test unit and the Semtech exhaust flow device please go to www sensors inc com The Sensor D test unit was zeroed and spanned at each set of test cycles The unit was found to be ver
26. en averaged over the vehicles tested Based on the results emission adjustment files can be generated for the IVE model for the location of interest A Viewing GPS and Emissions Output The SEMTECH system outputs mass emissions second by second into a spreadsheet that can be opened in excel There is a template made called Raw Emissions Data xls that can be opened and the data post processed to create the proper emissions files from the SEMTECH unit Also in this worksheet there are some graphs to view the emissions data in some common formats B Binning GPS and Emissions Output After processing the data in excel the file should be saved as a text file and used in the GPSEvaluate program This program will take all of the data files and compile them to create the emissions correction for each of the 60 bins as well as the driving fraction for each of the 60 bins The output file will be a text file that can be used in the IVE model The text files that are to be processed should be placed in the Data folder in the same folder with the GPSEvaluate program The program should then be started Figure VI 1 The program will list all files found in the Data folder at the time of program start These appear in the upper left hand window Clicking in the box to the left of the file names will cause the checked file to be evaluated when the Calculate button is 29 clicked For each group all data associated with that group is
27. er fee may be necessary unless representative buses can be obtained from government sources supportive of the testing program It is recommended that 200 300 be set aside for payment to bus truck owners for providing a vehicle and driver for the approximate 3 hour testing program However this is a local decision that should be made VII Vehicle Sampling Equipment For purposes of these studies a SEMTECH D portable emissions monitor PEM manufactured by Sensors Inc will be used for gaseous emissions This unit employs a flame ionization detector FID to measure THC a Non Dispersive Ultraviolet NDUV analyzer to measure NO and NO a Non Dispersive Infrared NDIR analyzer to measure CO and CO and an Electrochemical sensor to measure O2 Fuel for the FID is provided via a high pressure canister mounted within the PEM For measuring particulate 26 emissions a Finnish Dekati testing unit will be used to collect second by second particulate emissions data The Dekati test unit makes use of technology that ionizes the particulates and collects them by size A whole exhaust mass flow measurement device will measure the exhaust flow rate based on static and dynamic pressure differentials A partial stream of the exhaust taken from within the mass flow measurement device is routed through the analyzer system at a constant rate of 10 liters per minute The concentration and flow rate data is input to the onboard data logger on a sec
28. esssseeecesessseeeceesneeeceeensaeees 16 FIGURE 18 COMPARISON OF MEASURED EMISSION RATES WITH IVE PREDICTED EMISSION RATES 17 FIGURE 19 COMPARISON OF MEASURED EMISSION IN SEVERAL CITIES NORMALIZED TO THE FTP LA4 CYCLE AND GROUPED TO SIMILAR SIZE DISTRIBUTIONS cccsssccccessssscececssseeccessssseeccecessseeccesensseecesssseeeceeesueeeceeeesaeeecessseeeeeesssaeees 19 I Introduction From July 4 to July 15 2007 a series of 33 diesel vehicles were tested in Beijing China 24 of these vehicles were classified as light heavy duty vehicles most of them were trucks The tests were carried out in Beijing at a private service garage Table 1 indicates the vehicles that were tested Table 1 Diesel Vehicles Tested During the Study Odometer get Vehicle Type Model Emission Standard Year km kg en 1 Light Truck FORLAND Euro 2 2004 71 656 unknown 2 Light Truck FORLAND Euro 1 2002 216 047 4 0K kg 3 Light Truck BLAC Euro 2 2005 98 531 3 7 It 4 Light Truck Dong Feng Euro 2 2005 91 566 4 0K kg 6 Light Truck Qing Qi Euro 2 2003 449 558 4 5K kg 7 Light Truck Euro 2 2005 3 7 It 8 Light Truck Dong Feng Euro 2 2005 6 0K kg 9 Light Truck FORLAND Euro 2 2005 100 000 4 8K kg 10 Light Truck Bei Jing 180T Euro 3 2007 3 6K kg 12 Light Truck FORLAND Euro 3 2007 3 6K kg 13 Light Truck Dong Feng Euro 3 2006 6 0K kg 14 Bus Jing Long Euro 0 1999 323 578 6 2K kg 15 Bus Jing Long Euro 0 1999 307 717 6 2K kg 16 Light Truck DFAC Euro 3 2006 27 385 4 3K kg 17
29. he emissions being classified into too high bins Steps are taken to filter these events out of the data however a few data points slip by The data in bins 1 15 1 19 in Figure 6 represent only 0 0007 ofthe collected data and are included only for the sake of completeness We believe that these data should be ignored Figure 7 presents data from the same vehicles but this data is the NOx data from those vehicles 0 16 0 14 0 12 0 1 0 08 0 06 0 04 6 0 02 0 4 1 00 102 1 04 1 06 1 08 1 10 112 1 14 1 16 1 18 200 2 02 2 04 2 06 2 08 2 10 2 12 214 2 16 2 18 300 3 02 3 04 3 06 3 08 3 10 312 3 14 3 16 3 18 Figure 16 NOx Emissions by IVE Bin The main difference compared to previous studies of diesel vehicles is that there is a high emissions rate in bins 1 04 and 1 05 those bins represented the emissions on vehicles when they are decelerating and one of the possible causes of this behavior is that the drivers are using the engine to brake the trucks and buses placing a load on the engines when the vehicles are being rapidly slowed down 15 Figure 8 presents the binned data for particulate matter from the large diesel vehicles The standard form of the emissions curve can still be seen in the data 0 012 0 008 x 0 006 A 0 004 3 0 002 e gt 0 Bm aaa aa aaa 1 00 102 104 106 108 1 10 1 12 1 14 116 1 18 200 202 2 04 2 06 208 2 10 2 12 2 14
30. hicle is the result of engine and rolling friction wind resistance acceleration energy and road grade For a further discussion the reader is referred to the user s manual for the IVE model which can be obtained at www issrc org ive gt Engine stress relates to engine rpm and the average energy demand on the vehicle in the most recent 15 seconds The reader is referred to the user s manual for the IVE model which can be obtained at www issrc org ive 14 14 _ 12 2 10 b 8 v id 6 v Q 4 2 b Kj e s 0 1 00 102 1 04 1 06 108 110 1 12 1 14 116 1 18 2 00 202 204 2 06 2 08 2 10 2 12 2 14 2 16 2 18 3 00 3 02 304 3 06 3 08 3 10 3 12 3 14 3 16 3 18 Figure 15 CO2 Emissions From Heavy Duty Vehicles by IVE Bin The emissions look typical with the exception of the apparent fall off in emission rate in stress category 1 in the case of bins 16 18 i e 1 16 to 1 18 in Figure 6 These data points are marked in red Due to traffic congestion in Beijing it is not likely that driving occurred in bins 1 15 1 19 Thus these data points must result from erroneous classification of data into bins It has been found that the GPS unit will loose signal freeze and then jump to the correct speed a few seconds to a minute later when the signal returns Altitude can also be mis measured by the GPS These errors cause an improperly high acceleration or road slope calculation which results in t
31. ich column the hour min and second is in the GPS text files The GPS units report this information in the column 1 when counting from 0 see Table 1 in the section above Max Time 1 seconds When there is a time gap of more than the entry in this column it Jump will discard the data during that time period and pick up when the time gap is over Speed column 9 Mph Indicates which column the velocity is in the GPS text files The GPS units report this information in the column 9 when counting from 0 see Table I in the section above Min Time 1 seconds When there is a time gap of less than the entry in this column it will Jump ignore the data during that time period This is to protect against data that is collected at a faster resolution than the GPS reports For example if the GPS is 1 Hz 1 measurement per second but the data is collected at every half second the output file will report a line of data every half second with the information only changing every second In this situation you would want to make sure the Min Time Jump is set to 1 second Altitude 7 Meters Indicates which column the altitude is in the GPS text files The GPS column above units report this information in the column 7 when counting from 0 sea level see Table 1 in the section above Start Rows to 3 integer This is the number of rows after the data starts that is not used in the Skip analysis Usually once the recording begins there are several seconds
32. ive of the Beijing fleet thus the measured values are likely within 20 25 of the true mean of the Beijing fleet It should be noted that any zero emissions shown in Tables 3 or 4 indicates that emissions measured for the vehicles were below the detection limits of the equipment This only occurs in the case of CO and THC due to the fact that diesel vehicles run with high amounts of air compared to the fuel and a well running engine can have low CO and THC Table 3 Emission Measurement Results for the Tested Beijing Diesel Fleet Measured Emissions FTP Normalized Emissions Foss Vehicle Type Year Std grams kilometer grams kilometer co2 NOx PM CO THC co2 NOx PM CO THC 1 Light Truck 2004 Euro2 418 5 46 0 314 3 07 0 00 431 6 92 0 332 3 25 0 00 2 Light Truck 2002 Euro1 425 7 64 0 184 0 52 1 10 524 10 38 0 234 0 66 1 34 3 Light Truck 2005 Euro2 404 7 57 0 249 1 71 0 56 539 10 51 0 341 2 35 0 77 4 Light Truck 2005 Euro2 366 1 88 0 275 1 89 0 88 480 2 77 0 372 2 57 1 15 6 Light Truck 2003 Euro2 348 1 92 0 507 2 17 1 41 468 2 99 0 708 3 02 1 89 7 Light Truck 2005 Euro2 434 10 66 0 088 1 62 0 88 472 12 87 0 097 1 81 0 91 8 Light Truck 2005 Euro2 328 4 69 0 029 0 96 0 17 491 7 62 0 046 1 50 0 28 9 Light Truck 2005 Euro2 372 5 19 0 090 1 67 0 38 504 7 82 0 125
33. les were significantly larger than the Beijing vehicles An analysis was also done to derive the rates of emission increase with vehicle use However as was the case with model year comparisons there does not seem to be a trend of vehicle emissions increasing with use This is an unusual result but may be because the fleet tested in Beijing was relatively homogeneous The fact that emissions from the Beijing diesel vehicles seem to run higher than comparable vehicles in the other cities tested may suggest that maintenance in Beijing may be lax It might be good to establish a traditional Inspection and Maintenance program for vehicles or to use remote sensing to measure the emission rates of trucks as they transverse the city to identify and 19 repair high emitters In addition no significant differences of emissions were found among vehicles supposedly conforming to different vehicle emission standards which mean there is no effective emission improvement for diesel trucks in Beijing during in recent years with tightening emission standards Particulate and NOx control devices are becoming available for reducing diesel emissions Beijing may want to take advantage of these controls by setting more stringent new vehicle standards and by considering some form of retrofit program for diesel vehicles 20 Appendix A Field Manual For Diesel Vehicle Testing 21 IVE In Use Vehicle Emissions Study for Diesel Vehicles I Introduction
34. m less than I to greater than 1 17 Table 6 presents the recommended adjustment values for diesel vehicles in Beijing These values should of course be improved as more data is collected Table 6 Recommended Adjustment Values for Use in the IVE model Class CO Ratio VOC Ratio NOx Ratio PM Ratio Overall Diesel Heavy Duty Vehicle 1 59 2 40 1 32 1 00 VII Conclusions In summary the results of this report explain the relationship between the real emissions and the estimated emissions as determined by the IVE model However this information is not complete without comparing the emissions of the light duty fleet in Beijing with others fleets around the world The comparisons in this chapter include five cities Beijing China Istanbul Turkey Mexico City Mexico Santiago Chile and Sao Paulo Brazil In order to normalize the results and make them comparable the Bin methodology has been used to evaluate the emissions of each city under the FTP LA4 cycle The comparison includes the overall emissions for each campaign in each city It has to be noticed that in the case of Santiago the Light Duty vehicles are small diesel vehicle with less than 2 5 liter engines The Table 7 below displays the results Table 7 Emissions results for an average fleet in several cities Pollutant Beijing Istanbul Mexico Santiago Santiago Sao Paulo City Light Duty Heavy Duty CO g km 2 47 1 35 6 52 0 40
35. mass per unit distance to be determined in real driving situations To date this on board emissions measurement 22 II technology can be used to measure carbon monoxide CO carbon dioxide CO2 total hydrocarbons THC and nitrogen oxides NOx A different testing device to measure particulate emissions is also used during the study to establish real time particulate emissions from the tested diesel vehicles The IVE in use vehicle emissions study is built around a two week study period where on road vehicle emissions data is collected Equipment Needed to Complete Study A wide range of equipment is required to carry out the diesel emissions study The key pieces of equipment will be shipped from the United States However a significant amount of equipment and supplies must be provided locally An Excel spreadsheet program is provided with this write up that indicates the equipment that will be sent from the United States and the equipment that must be procured locally HI Sample Size and Impact on Emissions Measurement Reliability Unfortunately for the researcher studying the emissions from on road vehicles the variance in emissions among vehicles with similar technologies is quite large This means that multiple tests on different vehicles are required to accurately establish the true fleet wide average for a given technology Equation II 1 indicates the 70 confidence interval for data in a Gaussian distribution boy oN
36. n the vehicle to the starting point at the desired time The roads selected should also allow for as great a variety of driving as feasible for the location It is critical that vehicles arrive on time A testing facilitator should be in touch with the vehicle suppliers to insure that the vehicles will arrive on schedule Table IV 1 Vehicle Driving Procedure Step Procedure Time Interval 1 A vehicle arrives at the test setup facility at the designated ER appointment time while previous vehicle is being tested 2 Vehicle will be parked in the next vehicle test setup location 5 minutes Vehicle will be studied and a decision will be made as to the best 3 way to attach testing equipment to the just arrived vehicle Sand 20 minutes bags are loaded onto vehicle to be tested 4 Tested vehicle returns to the test setup location and parks near the 5 minutes 25 next vehicle setup location 5 Download data from the just tested vehicle 5 minutes 6 Test equipment is removed from tested vehicle and transferred to Sonnige new vehicle and tested vehicle released for return to the owner 7 Equipment installed on new vehicle 25 minutes Data collection initiated and vehicle started and driven over the 2 8 3 45 minutes designated one hour test route 9 Total time to test one vehicle 135 minutes Traffic will of course impact the distances that will be covered during the driving phase Thu
37. ng discussion only four general vehicle technology groups should be studied over the two week sampling period in order to insure at least 10 vehicles per technology class are studied The variety of diesel engine technologies operating in most urban areas today are relatively small and the study of the four most predominate vehicle technologies in an area will provide useful data for understanding the overall diesel emissions in an urban area IV Sampling Program Based on Section II it is desired to study four vehicle technology classes during the two week study The vehicle technology classes that should be studied are those groups which dominate the local diesel fleet being studied and which will continue to be important in the next 5 years The vehicles selected may vary from location to location but based on previous experience it is suggested that the following four technology classes be considered Euro 1 type technologies Euro 2 type technologies Euro 3 type technologies Euro 4 type technologies if available or extend testing of most common of the three previous groups The previous suggested technology classes should not be rigidly adhered to in cases where the local fleet does not contain significant numbers of vehicles in any of the vehicle classes listed Itis best to tailor the study to the makeup of the local fleet For this paper the four suggested classes will be used for discussion purposes Start emissions are importa
38. nt for gasoline vehicles but are not considered as important for diesel vehicles Thus the testing program that will be used will not collect diesel vehicle start emissions Vehicles will be scheduled to be brought to the testing area at 2 25 hour intervals beginning at 08 00 Thus a second vehicle will be due at 10 15 On the first day only two tests should be scheduled to allow for proper on site training It is critical that vehicles not arrive late Itis best to schedule them to arrive a little early The new vehicles will be parked in the test area leaving room for the returning vehicle to be parked Once the vehicle being tested arrives at the test setup facility the equipment is moved from the tested vehicle and installed in the new vehicle that has arrived for testing Buses are expected to arrive with only a driver Sand bags will be loaded onto the bus to simulate passenger weight Trucks should arrive with a half to a full load for purposes of 24 testing The truck or bus will be at the facility for about 3 hours for equipment installation testing and equipment removal It can then be returned to its owner Figure 1 Truck and Bus with Flow Measurement Devices Connected to the Exhaust V Vehicle Driving Procedure Table IV 1 indicates the approach that will be used to test the vehicle The driving roads selected for the study should provide for convenient stopping locations that will occur at the desired time intervals and retur
39. ond by second basis Internal filters carbon absorbers and chillers are strategically located in the sample stream to minimize interferences A temperature and humidity measurement device also provides second by second data on the temperature and humidity of the engine intake air Algorithm s in the processing software provide the necessary adjustment to the NO and NO results from the NDUV based upon the humidity ofthe intake air Data also collected on a 1 hertz cycle from an onboard GPS unit allows the measured mass of each pollutant to be matched up with the driving activities of the vehicle The PEM including protruding knobs and connectors measures 404 mm in height by 516 mm in width by 622 mm in depth It weighs approximately 35 kg The ID of the mass flow measurement device will have a diameter of 10 2cm me Figure 2 A Bus and Truck with Equipment and Sandbags Loaded Except during actual testing the internal temperatures of the PEM will be maintained using a line serviced 12 volt DC power supply A Y connector allows the PEM to be simultaneously connected to the line serviced power supply as well as a deep cycle 12 volt battery Prior to starting the first test each day the PEM will undergo a leak test as 27 well as a zero span and if necessary calibration test The proper size mass flow measurement device will be selected based upon the engine size of the test vehicle This will insure that a backpressure of less than
40. r VSP E VSP Bin 22 VSP Bin 23 VSP Bin 24 VSPBin28 WSPBin28 VSP Bin 30 vsPB VSPBN3 VSPBAS VSPBN34 VSPBin38 VSPEN39 _ VSPBn40 04 02 vSP a VSPBin48 VSPBin49 VSP Bin 50 VSP VSPBInS8 VSPEnSa _ VEPEnED Vehicle Spec Power Distribution 15 4hours 6 hours 12 hours 18 hours og Soak Time Figure VIII 2 VE Model User Interface for Entering in Driving Pattern Data 33 Location Various Input Latitude Longitude Altitude Units Location Info Template gt 500 m m meters ft feet MM DD YYYY Date 8 28 2002 Units Road Grade positive value is uphill negative number is downhill UM Class Percent AC In Use at 80 F 27 C Fleet File to Use Interpolation File to Use none enter text for one of five options percent of public with AC on vehicle using AC at 80F 27C ambient temperature A blank will be interpreted to use a linear fit for missing hours Overall Lead Pb Sulfur S Benzene Oxygenate Gasoline Enter gasoline related data Diesel Enter diesel related data Description Driving Style Distribution Facility Cycle Distribution Group 1 Vehicles Time Period VSP Bin I VSP Bin 2 VSP Bin 3 VSP Bin 4 VSP Bin 5 VSPBin6 VSP Bin7 Total Distance or Time Driven Number of Statups Soak Time Distribution Group 1 Vehicles 1 hour 2 hour 3 hours 4 hours 6 hours Average Velocity for Group 1 Vehicles
41. re 12 the results of emissions for these three ranges are shown In general there isn t a relationship between the weight and emissions the emissions of the high weight range vehicles are slightly lower than the lighter ones these can be explain because the heavy vehicles has better standard in average and there are some of the buses which were cleaner than trucks 11 Weigth tons 7 0 6 0 5 0 1 0 0 0 I 9 i 93999990 I 2 47 I 4 5T 5 7T I I I I 09000000 0 099 ur km e HE HE HE a ee G 3 6 9 12 15 18 21 24 27 30 33 Figure 11 Vehicles Distribution According to Weight Emissions g km 8 00 7 00 6 00 oi 2 gt 2 w 2 S 2 00 4 1 00 4 0 00 m 2 4 Tons E 4 5 Tons 05 7 Tons 69 70 5 1 50 51 49 29 2 6 25 24 25 23 0 6 co HC NOX PM 10 C02 100 Figure 12 FTP Normalized Emissions Averaged Over Vehicle Weight 12 Emissions according to mileage In Figure 13 the distribution according to mileage is shown the range on mileage is going between 7 500 to 450 000 kilometers The distribution shows 10 of the vehicles in the low range 0 to 80 000 kilometers 14 in the middle range 80 000 to 160 000 kilometers and 9 in the high range 160 000 and more The results in Figure 14 show the mileage distribution of the trucks only since the Euro
42. rienced mechanic is needed to help install the test equipment and to identify the engine type and technology Necessary ladders to enable the installers to reach the exhaust for connection will be required In the case of buses sand bags representing a the weight of a 2 3 full bus must be available along with two persons to load the bags onto the bus The ISSRC team will supply one person to work with the group to carry out the testing The local partner must also arrange for a zero gas and a calibration gas that is guaranteed to be within a 2 tolerance of specified values The testing program is begun at 06 30 when all personnel must arrive at the testing location and typically continues until 18 00 although on good days the testing may finish by 17 00 and on bad days the testing may take until 19 00 The daily work crew must be prepared to stay until testing is completed IX Analysis of Emissions Data Once collected the data needs to be divided into the appropriate 60 driving bins required by the IVE model based on the GPS data collected along with the emissions data The GPS evaluation program contains the necessary algorithms to estimate average emissions in each power bin to look at the relative emissions in the various power bins The binned GPS data can also be entered into the IVE model and emissions predicted for that driving pattern and vehicle technology These two comparisons will then indicate how well the IVE model is performing wh
43. s the test route should be designed so that there are alternate routes to be taken so that the vehicle can complete its test run in the 45 minutes that are allocated Since traffic will flow better at various times of the day the test may be completed in 30 minutes in one case and in 60 minutes in another case The test route should be selected so that the vehicle can complete the test run in 60 minutes in the worst traffic Itis also critical that the driving course selected contain street sections where higher speeds and accelerations can be achieved as well as slower speeds and lower accelerations The driver should operate the vehicle in a manner typical of the traffic that is occurring at the time of testing and in the manner that the vehicle is normally used i e a bus will stop at bus stops even though it does not pick up or discharge passengers Buses and trucks should be marked with signs taped to the vehicle indicating that the vehicle is participating in a testing program VI Vehicle Procurement Procuring 40 large diesel vehicles for testing with a driver and in the case of trucks can be a challenge Bus companies must be contacted as well as trucking companies to find the desired vehicles In both Mexico City and Sao Paulo a US50 per vehicle fee was paid for gasoline fueled vehicle This fee plus the use of contacts at the partnering agencies provided all of the needed vehicles in Sao Paulo In the case of buses and trucks a high
44. s the user with the ability to enter adjustments to the base emission factors that are specific to a location in case the supplied factors are found to be in error This capability was built into the model to support emission measurement studies that would be made in developing countries as local capacity increases The IVE In Use vehicle emissions study is not designed to fully develop correction factors for the IVE model The original data base used to developed the IVE model correction factors was based on more than 500 vehicle tests each involving three driving cycles carried out at the University of California CE CERT research facility This information was combined with summarizations of thousands of in use vehicle tests provided by the United States Environmental Protection Agency The IVE In Use emissions study discussed in the following section results in emissions data for 40 diesel fueled vehicles Information from 40 vehicles while significant does not provide the range of data for the development of a full range of new emission factors and adjustments Nonetheless the IVE In Use emissions study can be used to make rudimentary adjustments to the IVE model that will certainly improve its performance for developing countries The actual IVE In Use vehicle emissions study makes use of recently developed emissions measurement technology that can be carried on board vehicles while they are driven on urban streets This technology allows emission
45. the rest of the vehicles this result isn t logical and should be check in the future 12 0 E Euro 0 m Euro 1 10 0 Euro 2 Euro 3 8 0 7 5 5 6 9 D 6 5 n 4 5 gt 5 4 Pe a 5 0 E 46 LU 4 0 4 0 37 26 2 7 2 6 2 0 201 1 18 11 0 9 0 9 0 6 0 0 co HC NOx PM 10 02 100 Figure 8 FTP Normalized Emissions Averaged Over Emissions Standards Overall 12 0 E Euro 0 m Euro 1 Euro 2 10 0 Euro 3 8 0 4 76 E 6 9 3 6 5 2 60 55 35 8 E 4 6 LU 4 0 3 4 3 0 2 6 2 6 2 0 2 1 2 0 11 0 9 0 0 co HC NOX PM 10 C02 100 Figure 9 FTP Normalized Emissions Averaged Over Emissions Standards Trucks 10 12 0 5 E Euro 0 Euro 1 10 0 4 8 0 MA Vie a 2 6 0 5 5 5 3 4 0 4 04 3 9 3 9 3 0 2 0 2 0 1 3 1 0 0 7 0 2 0 0 r r r CO HC NOx PM 10 CO2 100 Figure 10 FTP Normalized Emissions Averaged Over Emissions Standards Buses Emissions according to Weight On Figure 11 the Weight distribution of the vehicles tested is shown the range for the overall sample is only 3 tons between 3 to 6 Tons In the range of 2 to 4 tons there were 18 vehicles 16 ofthem between 3 5 and 4 tons In the range of 4 to 5 tons there are only 5 vehicles In the higher range there are 9 vehicles mostly in the range of 6 to 6 5 tons In Figu
46. tion Group 2 Vehicles 15 min 30 min 1 hour 2 hour 3 hours 4 hours Once the driving fractions have been entered select all other information as close as possible to match the emissions testing conditions in the Location Page This includes selecting the ambient temperature and humidity fuel specifications and fleet The user will have to create a fleet file to represent the type of vehicle tested in this study For deriving correction factors only one technology should be used at a time To more information on how to fill out the location file and creating a fleet file please refer to the IVE user s manual Once the fleet and location file have been properly filled out the user can run the model and record the emission rate per distance for each pollutant Once the emissions values have been predicted by the IVE model these values can then be compared with the actual emissions values that were collected in the study To derive the correction factors simply divide the measured emission value over the predicted value to get the correction factor for that specific technology Then in the IVE model this correction should be entered and used when predicting emissions from this area For example if the IVE model predicts a CO emission rate of 10 g mi and the emissions measurements indicated on average an emission rate of 13 g mi of CO the correction factor for this technology would be 1 3 This information is entered in the base correction f
47. ults for PM emissions are below the model value in Euro 0 Euro 1 and Euro 3 the model is underestimating the emissions for Euro 2 4 The exact value for all the vehicles is 1 29 16 6 0 5 0 4 4 0 amp Euro 0 m Euro 1 3 0 4 m Euro 2 Euro 3 Overall 2 0 4 0 0 J CO Ratio HC Ratio NOX Ratio PM Ratio CO2 Ratio Figure 18 Comparison of Measured Emission Rates with IVE Predicted Emission Rates The default base emission factors in the IVE model are based primarily on emission measurements made in the United States and Europe and represent test results from thousands of vehicles It is difficult to know how much weight to give to emission results from only 33 vehicles in Beijing However the confidence limits shown in Table 3 suggest that the averages of the results should be in the ballpark of 30 of the actual values Table 5 shows the actual ratios and the ratios of the measured emission results were modified to be 30 closer to the IVE projected values Table 5 Potential Adjustment Values To Be Used in IVE Model NOX Ratio CO2 Ratio Because of the limited number of tests it is suggested that the values closer to the IVE default values the 30 adjusted values be used until more in use emissions data is collected in Beijing A value of I is used in the cases where the 30 adjustment takes the values fro
48. us row average speed or to use each data point separately straight speed Save Settings n a n a This button will save the current settings Settings should be saved in Button the Settings Folder as a txt file Load Settings n a n a This button will load the file named GPSEvaluateSettings txt that Button is located in the Settings Folder Time Offset 0 0 hours This is to enter the time difference from UTC GMT time reported Button in the GPS data to local time If location is 6 hours from GMT enter 6 If location is 6 hours from GMT enter 18 31 Once the appropriate settings have been selected or loaded and the files to process have been checked in the boxes the user is ready to process the data by pressing Calculate The bar above the Calculate button will show you the status of the calculation The calculation may take a few minutes especially with many data files or long data files Once the calculation has been finished the results will be displayed in the bottom half of the program There will be the percentage of travel in each of the 60 bins along with the total number of rows processed and the average speed over all the files It is also possible to save the output of the file analysis Click on the Save Output button and a text file with the information contained in the Results box can be saved All data on the driving and emissions will be saved in a text file C Applying
49. usually selected for the calculation EVALUATE GPS DATA JE EG Select Files to Evaluate vin Tae surp eget Start Rows to Skip 3 ozcan O Lists the files found in the data folder that might be processed 7 ee Del SAVE RESULTS CALCULATE 00 Display Power in Bins _ Time Offsst nis ee Calculation Results o o flash gps data sa ple1 155554 log Low 1 Stress med MZ 24 Stress arte Stress pl Displays results including percentage in each bin emissions in each bin overall number of data Indicates what information will be points processed and average displayed in lower portion of ee speed program Options are Driving Dropdown menus indicating the fractions CO CO2 PM VOC or columns for the time and speed data NOx The display does not impact the start rows to skip columns for each output files all data is always pollutant and other user options Average Speet provided in the output files Figure VIII 1 GPSEvaluate Program The upper right portion of the program displays the current settings for the program A description of each setting is described in the Table VI 1 below 30 Table VIIL1 Description of the Options in the GPSEvaluate Program Parameter Default Units Description Value Time column 1 hh mm ss Indicates wh
50. y stable from day to day with the zero and span holding within 1 of the calibration gases values from day to day Particulates were measured on a second by second basis using a Dekati DMM testing unit This unit uses a particle charging process and six stage impactor setup to determine particle mass The DMM measures particle concentration The exhaust flow rates collected by the Sensor D unit must be used with the Dekati measurements to determine particulate mass flow rates The DMM measures particles in the 0 to 1 5 micron range which is the size range where virtually all diesel particulates reside The DMM has been found to produce results comparable to the reference particulate sampling methods for diesel particulates although it was found to produce readings about 30 high in one published study Dekati experts believe that this is due to the fact that the Dekati measurement process can measure volatile particulate matter that is lost in the case of filter based particulate sampling devices For further information on the DMM please see www dekati com The DMM was zeroed at the beginning of each testing cycle The charging and impactor units become covered with particulates and must be cleaned after each 2 3 hours of testing to keep the unit operating properly The DMM can not handle the mass concentrations found in uncontrolled diesel exhaust Thus the diesel exhaust must be diluted at a controlled rate in order to use the DMM A field dilution
Download Pdf Manuals
Related Search