Environmental Systems Products Holdings Inc. 1 Presented to ARAI on March 18 th /19 th, 2004 Dr. Donald Stedman, Niranjan Vescio, Gary Full
Agenda 2 Instrumentation Gary Full RSD Gas Calculations Dr. Stedman Summary Statistics Dr. Stedman 2-Wheeler Improvement Gary Full Other Discussion - All
RSD4000 Technology Device Measurement Overview 3 Presented by Gary Full
Optical Path 4 TMM SDM Return beam RCVR Outgoing beam SRC Direction of vehicle travel
Optical Beam 5 IR filament source and UV light source are combined and positioned so as to appear at the focus of the source mirror. Result is a collimated (nearly parallel rays) exit light beam. Transfer mirror offsets the beam and returns it to the receiver optical section of SDM. Receiver mirror collects and focuses the received light. UV light is split out via a dichroic mirror and distributed to a spectrometer via fiber optic cable for NO and SF measurements. IR light is sent to a rotating (12,000 RPM) spinning mirror scanner assembly which distributes light to individual IR detectors for CO, CO2, REF, and HC detection. Scanner produced 2400 light pulses/second at each detector. Thermoelectrically cooled detectors (~-5 degree C) produce stable, highly sensitive, low noise electrical signal outputs.
Optical Beam Path 6
Signal Conditioning/Firmware 7 Each detector output is AC-coupled via a preamplifier. Discrete gain selection (0.5, 1, 2, 4, or 8). Q=10 band-pass filter produces sine waves at carrier frequency of 2400 Hz. Gas information is represented by the amplitude modulation of the carrier. Sine waves are then rectified. A synchronous integrator provides an average over 24 sine wave periods to produce a result that is sampled by a 16-bit A/D converter every 10-milliseconds (100 samples/second). IR gas channel voltages (CO,CO2, and HC) are ratioed to the REF channel to provide a signal proportional to gas transmittance (ratio of emitted light to received light).
Gas Measurement Processing 8 Curves of gas transmittance versus gas amount (number of molecules/unit beam cross-section) are stored in the SDM. Measured transmittances are converted to gas amounts. Gas amount measurements (ambient gas readings) at front of the vehicle are recorded for CO, CO2, and HC. 50-samples of information are recorded at the rear of the vehicle. Amount of gas attributable to the vehicle is found by subtracting the ambient value from each of the 50 samples.
CO2 Response Curve 9 3b-1 350 350 P2 CO2 (%-cm), std-day 300 250 Amount of gas %-cm estco2plot 200 150 100 50 0 50 50 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85.2 ratioco2plot kr CO2 Normalized ratio (CO2/REF).9
Plume Decay Graphs 10
11 Gas Measurement Processing continued Aggregate gas ratios (a single value for the possible 50 point sets) are calculated for CO/CO2, HC/CO2, and NO/CO2. The aggregate values are least squares linear regression fits of the gas value array with respect to the CO2 array. E.g. CO array versus CO2 array. For the aggregate slope calculations we reject data points immediately before and after beam blocks/unblocks, and where the sum of the gas amounts are below a threshold. We consider a gas measurement valid if there are at least 5- points available in the slope calculation invalid otherwise.
12 THE BASIC INSTRUMENT Detectors/filters CO2 Software Software Gas Ratios n CO2 Source(s) CO HC NOx Using calibration curves (results of coefficienting) determine the number of molecules in the path for each of the gas species. n CO n HC n NOx Determine gas ratios n CO /n CO2 n HC /n CO2 n NOx /n CO2 REF Results Reporting n CO /n CO2 n HC /n CO2 n NOx /n CO2 COMBUSTION EQUATION %CO2 = 1/ [ C1 * n CO / n CO2 + C2 +C3 * n HC / n CO2 + C4 * n NOx / n CO2 ] %CO = %CO2 * n CO / n CO2 %HC = %CO2 * n HC / n CO2 %NOx = %CO2 * n NOx / n CO2 %CO %HC %NOx %CO2 ASSUMPTIONS that impact determination of C1, C2, C3, and C4 1. The measured gases are the results of combustion and only combustion gases. 2. The fuel hydrogen to carbon ratio is known. 3. The oxidizer is air.
13 Gas Measurement Processing continued We also report stoichiometric (meaning the concentration that would be measured by a tail pipe probe if there were no excess air in the exhaust beyond what is needed for combustion) gas concentrations. These gas concentrations are the result of combustion chemistry where the basic assumptions are that the aggregate carbon to hydrogen content of the fuel is known and that air is the oxidizer in the combustion process.
Other RSD Subsystems/Options 14 Speed/acceleration Vehicle/license plate picture Weather measurement GPS for time base as well as position Wireless communications links - 300 meters - Cellular links PC storage and transfer media Smart Signs / motorist displays
RSD Calculations 15 Presented by Dr. Donald Stedman
On-road Results from Denver 16 Sept 2003 Slides prepared by Donald H. Stedman, March 18, 2004 for ESP www.feat.biochem.du.edu
IR Signal vs Time 17 IR Signal versus time 9 8 7 6 IR signal 5 4 Ref CO CO2 HC 3 2 1 0 0 100 200 300 400 500 600 Time ms
Gas Readings vs Time 18 Gas readings versus time 8 7 6 5 % in 8 cm. 4 3 CO CO2 HC 2 1 0 0 100 200 300 400 500 600-1 Time ms
CO vs CO2 19 CO versus CO2 0.1 0.08 0.06 0.04 0.02 CO 0 0 1 2 3 4 5 6 7 8-0.02-0.04-0.06-0.08-0.1 CO2
HC vs CO2 20 HC versus CO2 0.1 0.08 0.06 0.04 0.02 %CO2 in 8 cm 0 0 1 2 3 4 5 6 7 8 HC -0.02-0.04-0.06-0.08-0.1 % HC in 8 cm
IR Signal vs Time 21 Figure 2b: IR Signal versus time 8 7 6 Signal V 5 4 3 Ref CO CO2 HC 2 1 0 0 50 100 150 200 250 300 350 400 450 500 Time ms
Gas Readings vs Time 22 Figure 3b: Gas readings versus time 7 6 5 % in 8 cm. 4 3 2 CO CO2 HC 1 0-1 0 50 100 150 200 250 300 350 400 450 500 Time ms
23 CO vs CO2 Figure 4b: CO versus CO2 1 0.8 y = 0.7999x - 3.7888 R 2 = 0.9976 0.6 CO 0.4 0.2 0 0 1 2 3 4 5 6 7 CO2
Proof that the on-road readings are correct 24 Excellent Correlation by MY to IM240, published and repeated in Chicago and Phoenix and other years.
Denver 1999 CO 25 300 Average CO from RSD (g/kg fuel) 250 200 150 100 50 95 y = 0.6785x + 14.003 R 2 = 0.9736 88 82 0 0 50 100 150 200 250 300 Average CO from IM (g/kg fuel)
Denver 1999 HC 26 Average HC from RSD (g/kg fuel) 20 18 16 14 12 10 8 6 4 2 0 y = 0.4108x + 0.8305 82 R 2 = 0.9783 88 95 0 5 10 15 20 Average HC from IM (g/kg fuel)
Denver 1999 NO 27 Average NO from RSD (g/kg fuel) 20 18 16 14 12 10 8 6 4 2 0 82 y = 1.0848x + 1.7888 R 2 = 0.9783 88 96 0 5 10 15 20 Average NOx from IM (g/kg fuel)
Proof from Roadside Pullovers 28 Using a single remote sensor to pull over on-road high emitters, BAR found 83%-88% ASM failure rates on CO or HC or NO depending on the remote sensing cut point chosen. At a 10,000 vehicle per day site, a very high cut point (top 1%) will fail 100 vehicles per day! California Bureau of Automotive Repair, Remote Sensing Device high emitter identification with confirmatory roadside inspection Final Report 2001-06.
29 A few broken vehicles cause most of the emissions. RSD is ideal for emissions inventory and to identify broken vehicles
Repeatability 30 Low emitters with good air/fuel ratio control have very repeatable emissions regardless of the test method. High emitters with poor air/fuel ratio control have very erratic emissions regardless of the test method. "Motor Vehicle Emissions Variability", G.A. Bishop, D.H. Stedman, L. Ashbaugh, J. Air Waste Manage. Assoc., 46:667-675, 1996.
31 Grams/mile AQIRP data 21 FTP tests on each vehicle.
Hit rates 32 Exact validity criteria for University of Denver data see CRC reports on our web site www.feat.biochem.du.edu Hit-rate depends upon the exhaust plume magnitude (vehicle load, speed and engine size and measurement site) and software conservatism. With good sites we routinely observe a hit rate >96% on automobiles. With the folded optical beam system we observe >95% hit rate for 95 cc motorcycles.
Data analysis from India 33 Presented by Dr. Donald Stedman
Gas Flag 34 Type C V x Total V % Fleet% 2WM 2095 5031 7126 29% 39.18% 2WS 1326 3378 4704 28% 25.86% 3WV 2065 1630 3695 56% 20.31% 4WD 759 216 975 78% 5.36% 4WV 1558 131 1689 92% 9.29% 7803 10386 18189
45% 40% 35% 30% 25% 20% 15% 10% 5% 0% CO CO Contribution by Vehicle Type/km 35 CO fleet% Two wheel motorcycle Two wheel motor scooter Three wheeler Heavy diesel vehicles Four wheel vehicles Percentage of Total CO
45% 40% 35% 30% 25% 20% 15% 10% 5% 0% HC HC Contribution by Vehicle Type/km HC fleet % 36 Two wheel motorcycle Two wheel motor scooter Three wheeler Heavy diesel vehicles Four wheel vehicles Percentage of Total HC
60% 50% 40% 30% 20% 10% 0% NOx NOx Contribution by Vehicle Type/km 37 NO Fleet% Two wheel motorcycle Two wheel motor scooter Three wheeler Heavy diesel vehicles Four wheel vehicles Percentage of Total NOx
50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Smoke Smoke Contribution by Vehicle Type/km 38 Smoke (<25) Fleet% Two wheel motorcycle Two wheel motor scooter Three wheeler Heavy diesel vehicles Four wheel vehicles Percentage of Total Smoke
400 350 300 250 200 150 100 50 0 CO Emissions CO Emissions/kg of fuel 39 Two wheel motor scooter Three wheeler Heavy diesel vehicles Four wheel vehicles Two wheel motorcycle
300 250 200 150 100 50 0 HC Emissions HC Emissions/kg of fuel 40 Two wheel motor scooter Three wheeler Heavy diesel vehicles Four wheel vehicles Two wheel motorcycle
25.00 20.00 15.00 10.00 5.00 0.00 6.59 NO Emissions NO emissions/kg of fuel 0.16 0.51 19.52 11.68 41 Two wheel motor scooter Three wheeler Heavy diesel vehicles Four wheel vehicles Two wheel motorcycle
Smoke Smoke gm/kg of fuel (assumes all black) 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00 42 Two wheel motorcycle Two wheel motor scooter Three wheeler Heavy diesel vehicles Four wheel vehicles
Improvements Necessary for Improved 2-wheeler measurement 43 Presented by Gary Full
1. Traffic Diversion 44 2-wheeler and 3-wheeler traffic must be diverted to a single file lane in which measurement equipment can be deployed.
2. Equipment Modification 45 Present technology equipment must be operable for measurements for path lengths of 3 to 5 meters. Present equipment is calibrated for measurement path lengths of 5 to 10 meters.
3. Loading Mechanism 46 Suggest that a portable ramp be deployed at the measurement station. 0.3 meter rise over a 3-meter length. Similar ramp down.
4. Multiple Analyzers 47 Two analyzers may be required deployed at two different heights to assure plume intersection with measurement beams.
Other Discussion 48 Cutpoints Correlations