Patent Application
20050274169
The present invention generally relates to exhaust gas analysis, and more particularly to an exhaust gas analysis method and apparatus having a constant sampling rate.
Frequently in exhaust gas analysis, the absolute level of a component in the gas phase sample is of interest. For example, it is often necessary to measure the real-time mass emission rates of certain regulated components in vehicle exhaust. In order to calculate the mass of the component from a relative measurement such as concentration, a precise time alignment of the concentration value and the volume or flow rate of the exhaust must be made. Delay times associated with transporting a sample of the exhaust gas from the source to the gas analyzer must remain constant despite changes in the system environment. However, due to the nature of the sampling arrangements employed in conventional systems, sample delay times are affected by changes in atmospheric pressure due to weather or altitude encountered during testing, especially during real world driving. Thus, as changes in atmospheric pressure occur, the differential between the instrument sample pressure and the atmospheric pressure will change. These changes may result in changes in the transport delay time of the sample and thus, the misalignment of the data used to calculate the mass emission rates, thereby making the system inaccurate.
Therefore, there remains a need for an improved method and apparatus for an exhaust gas analysis system.
The present invention meets this need by providing an apparatus and method which provide a controlled, representative sample of vehicle exhaust to a variety of analytical instruments independent of temperature, pressure, and engine operating conditions at altitudes up to 14,000 feet.
To compensate for atmospheric pressures changes, the present invention utilizes a reduced pressure (relative to atmospheric pressure) sampling technique to maintain a gas sample at constant temperature and pressure, which are independent of the vehicle's exhaust temperature and pressure. Consistency in sample delay times is maintained preferably by regulating the power supply to the sample pump, thereby regulating pump speed, in order to maintain constant gas sample delay times at various atmospheric pressures.
In one embodiment, the exhaust gas analysis system includes a temperature control system maintaining a gas sample at constant temperature; a pressure control system maintaining the gas sample at constant pressure; a gas analyzer analyzing the gas sample; a vacuum pump transferring the gas sample to the gas analyzer; and a delay time control system maintaining a constant gas sample delay time through the exhaust gas analysis system.
Another aspect of the invention is a method for exhaust gas analysis. The method includes providing a gas sample for analysis; maintaining the gas sample at constant temperature; maintaining the gas sample at constant pressure; providing a gas analyzer; maintaining a constant gas sample delay time through the exhaust gas analysis system; and analyzing the gas sample in the gas analyzer to provide a concentration of a gas species.
The present invention can be carried into practice in various ways but several embodiments will now be described by way of example, and with reference to the accompanying drawings in which:
The exhaust gas analysis system of the present invention samples and analyzes vehicle exhaust. The exhaust gas analysis can optionally be combined with vehicle parameter data from the Powertrain Control Module (PCM). The exhaust gas analyses, in terms of concentrations, can be combined with the data from the PCM to calculate mass emission rates. Precise time alignment of the exhaust gas concentrations measured by the exhaust gas analysis system and the PCM data is desirable to calculate these emission rates accurately.
In order to calculate the mass of a component from a relative measurement, such as concentration, a precise time alignment of the concentration value and the exhaust volume or exhaust flow rate of the sample is made. Delay times associated with transporting the sample from the source to the gas analyzer should remain constant despite changes in the system environment. When real-time mass emission rates of an emission component in vehicle exhaust are measured while the vehicle is driven in mountainous areas where changes in altitude result in ambient pressure changes, the sample delay times associated with the gas analysis system should remain constant so that a correct product of concentration and exhaust flow rate is obtained.
A series of studies was undertaken to evaluate the changes in the sample delay times that occur as a result of atmospheric pressure changes that might be encountered during real world driving. The pressure study encompassed atmospheric pressures ranging from 725 mbar to 975 mbar, which correspond to altitudes ranging from slightly above sea level (about 1000 ft) to as high as 8500 ft. The effects of changes in instrument (sample) pressures on sample delay times were also examined. These studies were conducted under various atmospheric pressure conditions which indicated that atmospheric pressure changes on the order of about 100 mbar can change sample delay times through the instrument by as much as one second or more. The accuracy of the mass emission rates generated during real-world driving studies conducted in mountainous regions could suffer measurably from such deviations in sample gas delay times. Therefore, a delay time control system is used to maintain constant sample gas delay times. The delay time control system may be a power supply regulator controlling power supplied to the pump. Circuitry to do so was developed. The circuits were validated via laboratory testing. The circuit diagrams and the results of the validation testing are presented and discussed below, although other circuitry and other delay time control systems could be used.
Experiment 1
As depicted in
Experiment 1 examined the behavior of the exhaust gas analysis system at three different atmospheric pressure levels, at the actual atmospheric pressure of about 990 mbar, and two simulated pressures of about 900 and about 800 mbar. Adjustments to the pump voltage were made while sample flow rates and the ability of the system to maintain the desired sample pressure were monitored and recorded. The data acquired are presented in Tables 1, 2, and 3.
*C = Controlling NC = Not Controlling
At an atmospheric pressure of about 990 mbar and a pump duty cycle setting of 100% (12.3V), the exhaust gas analysis system pressure control circuit failed to control the sample pressure at pressure board settings greater than 38 (corresponding to a sample pressure of about 891 mbar). At this atmospheric pressure, pump duty cycle setting, and sample pressure, the flow rate through the exhaust gas analysis system was about 6.6 L/min.
At an atmospheric pressure of about 990 mbar and a pump duty cycle setting of 69% (10.3V), the exhaust gas analysis system pressure control circuit failed to control the sample pressure at pressure board settings greater than 42 (corresponding to a sample pressure of about 907 mbar). At this atmospheric pressure, pump duty cycle setting, and sample pressure, the flow rate through the exhaust gas analysis system was about 5.7 L/min.
At an atmospheric pressure of about 990 mbar and a pump duty cycle setting of 55% (8.3V), the exhaust gas analysis system pressure control circuit failed to control the sample pressure at pressure board settings greater than 45 (corresponding to a sample pressure of about 919 mbar). At this atmospheric pressure, pump duty cycle setting, and sample pressure, the flow rate through the exhaust gas analysis system was about 4.8 L/min.
*C = Controlling NC = Not Controlling
At a simulated atmospheric pressure of about 900 mbar and a pump duty cycle setting of 100% (12.3V), the exhaust gas analysis system pressure control circuit failed to control the sample pressure at pressure board settings greater than 16 (corresponding to a sample pressure of about 803 mbar). At this simulated atmospheric pressure, pump duty cycle setting, and sample pressure, the flow rate through the exhaust gas analysis system was about 6.0 L/min.
At a simulated atmospheric pressure of about 900 mbar and a pump duty cycle setting of 69% (10.3V), the exhaust gas analysis system pressure control circuit failed to control the sample pressure at pressure board settings greater than 20 (corresponding to an sample pressure of about 820 mbar). At this simulated atmospheric pressure, pump duty cycle setting, and sample pressure, the flow rate through the exhaust gas analysis system was about 5.1 L/min.
At a simulated atmospheric pressure of about 900 mbar and a pump duty cycle setting of 55% (8.3V), the exhaust gas analysis system pressure control circuit failed to control the sample pressure at pressure board settings greater than 25 (corresponding to an sample pressure of about 838 mbar). At this simulated atmospheric pressure, pump duty cycle setting, and sample pressure, the flow rate through the exhaust gas analysis system was about 4.4 L/min.
*C = Controlling NC = Not Controlling
At a simulated atmospheric pressure of about 800 mbar and at pump duty cycle settings of 100% (12.3V) and 69% (10.3V), the exhaust gas analysis system pressure control circuit failed to control the sample pressure. At a simulated atmospheric pressure of about 800 mbar, and a pump duty cycle of 55% (8.3V), the exhaust gas analysis system pressure control circuit failed to control the sample pressure at pressure board settings greater than 3 (corresponding to a sample pressure of about 751 mbar). At this simulated atmospheric pressure, pump duty cycle setting, and sample pressure, the flow rate through the exhaust gas analysis system was about 3.8 L/min.
At a simulated atmospheric pressure of about 800 mbar and a pump duty cycle setting of 50% (7.5V), the exhaust gas analysis system pressure control circuit failed to control sample pressure at pressure board settings greater than 4 (corresponding to a sample pressure of about 755 mbar). At this simulated atmospheric pressure, pump duty cycle setting, and sample pressure, the flow rate through the exhaust gas analysis system was about 3.5 L/min.
Several observations can be made about the performance of the pressure controlled sampling system under varying atmospheric pressure conditions. First, at a given atmospheric pressure, lower pump duty cycle (pump voltage) results in lower flow rate through the exhaust gas analysis system. This suggests that as the power supply voltage to the pump changes, the sample delay times through the exhaust gas analysis system will also change. The addition of a regulated power supply for the sample pumps should ensure consistent sample delay times at a given atmospheric pressure.
In addition, for a given pump duty cycle, as the atmospheric pressure decreases, the differential required between atmospheric pressure and the sample pressure for control to be maintained by the pressure control circuit decreases. Under the experimental set-up used for Experiment 1, the minimum pressure differential required with a minimum flow rate of about 4 l/min is about 50 mbar. To accommodate moderate changes in atmospheric pressure that might occur due to either atmospheric conditions or altitude changes, a minimum pressure differential of 100 mbar is recommended.
Experiment 2
The equipment for the exhaust gas analysis system which was used in Experiment 1 did not permit simulating pressures below approximately 740 mbar. However, lower sample pressures might be required to accommodate atmospheric pressures typically experienced at higher altitudes, such as Denver, Colo. (atmospheric pressure about 825 mbar). Therefore, the equipment was modified so that lower simulated pressures could be obtained.
On the inlet side of the exhaust gas analysis system 120, toggle valve 155 was used to select either a sample of ambient air or a sample of a standard gas. On the outlet side of the exhaust gas analysis system 120, ballast 160 was positioned after pressure control valve 150 to dampen pressure fluctuations created by the 12V vacuum pump 165. Mass flow meter 170 was employed on the outlet side of the exhaust gas analysis system 120, prior to pressure control valve 150, to monitor sample flow rates. Dryer 175 was placed inline prior to mass flow meter 170 because the particular exhaust gas analysis system 120 used in this experiment had been used previously in actual vehicle studies and still contained small amounts of water in the filtering system.
Software (MBWin.exe software obtained from Sensor, Inc. and modified in-house (version 1-97F) to accommodate certain algorithms for the IR analysis), was used to acquire the time delay data. The software was used to obtain voltage values from the gas analyzer and convert those values into concentration units. It was modified to display the concentration units calculated using calibration equations to correct the analysis of a specific compound for the effects of other compounds present in the exhaust stream.
The data acquisition rate was determined by the rate at which the gas analyzer measured the voltages for each of the data channels, and it was determined to be about 1.6 Hz. For each trial conducted, a data file of 100 data points was acquired. Injections of a standard sample gas were used to determine the actual delay times. Toggle valve 155 was switched between a sample line carrying a gas consisting of about 30% CO2 (in N2) and a sample line carrying filtered ambient air. The NDIR (non-dispersive infrared) CO2 detector voltage was monitored using the MBWin.exe software to determine the delay times associated with the system.
The sample delay time is the time required for the sample gas to travel from the introduction point to the gas analyzer. The voltage of the NDIR channel pertaining to the compound of interest was monitored to determine the time it took for the sample to travel from the introduction point to the NDIR cell. The longer the sample line, the greater the sample delay time. It is desirable to minimize the distance between the sample introduction point and the gas analyzer so that the sample delay time is a fairly accurate estimate of the sample delay through the gas analyzer itself.
The data acquisition was initiated with the toggle valve 155 sampling the filtered air. After exactly 40 data points (about 25 seconds), the toggle valve 155 was switched to the CO2 sample line. The CO2 was allowed to flow for exactly 30 data points (about 19 seconds), at which time, the toggle valve 155 was switched back for the remainder of the data file. Thus, each trial provided a delay time corresponding to a falling edge (voltage drop due to the introduction of CO2) and a rising edge (voltage increase due to the reintroduction of air). A minimum of 30 data points for each toggle valve setting provided ample time for the voltage to reach and maintain a plateau. Thus, the time delay associated with the rise and fall in the channel voltage could be accurately measured. Four trials were performed at each atmospheric pressure/sample pressure setting. The delay times were recorded, and the average of the delay times measured for all valid trials was calculated.
The simulated atmospheric pressure was varied in increments of 50 mbar from a setting of about 975 mbar down to the lowest atmospheric pressure at which the exhaust gas analysis system could control the sample pressure properly, which is about 100 mbar greater than the desired sample pressure. The study was conducted at three instrument (sample) pressure settings: about 600 mbar, about 725 mbar and about 860 mbar. The individual and the average delay times measured at the various atmospheric pressure settings for each of the three instrument settings are presented in Tables 4, 5, and 6.
*Falling Edge = First occurrence of a CO2 channel voltage below the average voltage value (+/−3 σ) corresponding to air sample.
**Rising Edge = First occurrence of a CO2 channel voltage above the average voltage value (+/−3 σ) corresponding to CO2 sample.
The results of this experiment indicate that with a consistent voltage of 11.5V applied to the sample pumps, the mass flow rates measured at a given instrument pressure increased with decreasing atmospheric pressure, while the average of the sample delay times decreased. This suggests that if the exhaust gas analysis system were used to evaluate emissions from a vehicle traveling from a location at or slightly above sea level to a location at much higher altitude, the sample delays times would vary measurably during the travel period. In the extreme case, if the instrument pressure were set at about 600 mbar to accommodate all altitude/atmospheric pressures changes likely to occur during real world driving, the maximum change in the time delay resulting from a 250 mbar change in pressure (corresponding to a change in altitude of about 7500 feet) would be about 2.5 seconds.
Even in a more moderate case, if the instrument pressure were set at about 725 mbar, to accommodate atmospheric pressure changes of about 100 mbar, corresponding to altitudes of up to about 5500 feet, changes in delay times would be on the order of 1 second. Such changes in delay times are significant and would result in the incorrect alignment of the vehicle parameter and emissions data, creating errors in the mass emission values generated. In order to maximize the accuracy of the real-time mass emission data acquired using the exhaust gas analysis system, it is recommended that changes in altitude resulting in atmospheric pressure changes of greater than 50 mbar be avoided, unless the data is acquired in altitude “bins” with atmospheric pressures differences of 50 mbar or less. This could be done by adjusting the instrument pressure for every 50 mbar change in ambient pressure. For example, for ambient pressures between 900 and 950 mbar, one instrument pressure setting would be used to achieve a given sample delay time. For ambient pressures of 850 to 900 mbar, a slightly different instrument pressure would be used to achieve the same sample delay time. Similar adjustments would be made for each 50 mbar pressure change. Thus, to maintain system accuracy, sample delay times can be adjusted appropriately prior to data acquisition.
Experiment 3
The experimental set-up for Experiment 3, shown in
On the inlet side of the exhaust gas analysis system 220, toggle valve 255 was used to select either a sample of ambient air or a sample of a standard gas. On the outlet side of the exhaust gas analysis system 220, ballast 260 was positioned after pressure control valve 250 to dampen pressure fluctuations created by the 12V vacuum pump 265. Mass flow meter 270 was employed on the outlet side of the exhaust gas analysis system 220, prior to pressure control valve 250, to monitor sample flow rates. Dryer 275 was placed inline prior to mass flow meter 270 because the particular exhaust gas analysis system 220 used in this experiment had been used previously in actual vehicle studies and still contained small amounts of water in the filtering system.
To eliminate changes in sample pump speed (and hence changes in sample gas delay times) due to variations in the available battery power, pulse width modulated pump control (PWMPC) circuitry 280 was added to regulate the power supplied to the sample pumps in the exhaust gas analysis system 220. A circuit diagram for one embodiment of PWMPC circuitry is shown in
The validation study was conducted at a single instrument pressure of about 700 mbar. The CDTMC circuitry was used to set delay times at 3.5 and 5.0 seconds. Atmospheric pressures from 975 mbar to 775 mbar at 50 mbar increments were studied. The actual sample delay times were determined using the same procedure described in Experiment 2. Four trials were performed at each atmospheric pressure setting. A falling edge and a rising edge time delay value was determined for each valid trial. The average of the sample delay times was calculated for each atmospheric pressure level. The results of the circuit validation study for the settings of 3.5 and 5.0 seconds delay times are presented in Tables 7 and 8, respectively.
*(FE) = Falling Edge; First occurrence of a CO2 channel voltage below the average voltage value (+/−3 σ) corresponding to N2sample.
**(RE) = Rising Edge; First occurrence of a CO2 channel voltage above the average voltage value (+/−3 σ) corresponding to CO2 sample.
The data presented in Tables 7 and 8 indicate that the CDTMC circuit controls to the requested delay times quite well. In the case of the 5.0 second delay time setting, the average of the average measured delay times for all four atmospheric pressure levels is 4.88 second with a standard deviation of 0.08 second. For the 3.5 second delay setting, the average of the average measured delay times for all four atmospheric pressure levels is 3.43 second with a standard deviation of 0.17 second. For purposes of comparison, in a previous study which was conducted at the same atmospheric pressures and under a similar instrument pressure (about 725 mbar) and in which the CDTMC circuit was not employed (hence pump voltage remained at a constant value of 11.5 volts), the average of the average measured delay time at all four atmospheric pressure settings was 3.53 seconds with a standard deviation of 0.41 second.
Experiments 1 and 2 showed that changes in atmospheric pressure due to atmospheric conditions or changes in altitude can cause variations in the sample delay times associated with the exhaust gas analysis system. Since precise time alignment of the exhaust gas concentrations measured by the exhaust gas analysis system and the vehicle parameter data obtained from the PCM is desirable in order to ensure the accuracy of the system, significant variations in the sample delay times are undesirable. As shown in this experiment, the addition of the pulse width modulated pump control circuitry and the constant delay time motor control circuitry to the exhaust gas analysis system will enable reduction in the variation of sample delay times associated with changes in atmospheric pressure and improve the accuracy of the system.
Although the gas sample does not have to be diluted in this system, it can be diluted if desired. Dilution of the sample gas would affect the selection of the gas analyzer because of the effect it has on the detection limits of the gas analyzer.
To ensure the accurate analyses of the exhaust gas using these techniques, the gas sample should be maintained at a constant temperature and pressure as it passes through the gas analyzer. This is achieved independent of the vehicle exhaust gas temperature and pressure. The analyzer compartment 320 should be maintained at a constant temperature to prevent condensation of any water and water-soluble components contained in the exhaust gas sample. One or more temperature sensors 330 measure the temperature of the gas sample. Suitable temperature sensors can be any small and fast sensor having an appropriate temperature range. Temperature sensors having a temperature range of about 25° C. to about 100° C. would be useful, although other temperature ranges could be used, as is well known in the art. Examples of suitable temperature sensors include, but are not limited to, diodes, thermistors, and thermocouples. If desired, temperature sensors incorporating temperature control circuitry, such as a thermostat, can be used to control one or more heaters 335. Alternatively, the temperature control circuitry can be separate from the temperature sensors. A suitable constant temperature is 60° C., although other temperatures can be used. The analyzer compartment 320 can optionally include a fan 337 to circulate air in the analyzer compartment to assist in maintaining the constant temperature.
Constant pressure should be maintained in the analyzer compartment 320. This can be accomplished using one or more pressure sensors 345, pressure control valve 350, and associated pressure control circuitry 355. A leak check pressure sensor 347 can be included if desired. Pressure control valve 350 is situated in the sample line 315 upstream of the gas analyzer 325, and is responsive to pressure sensor 345. Suitable pressure control valves 350 include, but are not limited to, variable flow pressure control valves. One pressure sensor 345 can be located in the sample line 315 upstream of the gas analyzer 325, and a second pressure sensor can be located at the gas analyzer 325. The pressure sensor 345 upstream of the gas analyzer 325 is desirably positioned downstream of any filters or valves. It measures the absolute pressure of the sample gas. The second pressure sensor 345 at the gas analyzer 325 measures the pressure in the gas cell. Suitable pressure sensors 345 include, but are not limited to, absolute pressure gauges. The signal from pressure sensor 345 is compared to a fixed reference signal that corresponds to the desired absolute pressure. If the pressure is lower than desired, pressure control valve 350 is opened until the proper pressure is achieved. If the pressure is higher than desired, pressure control valve 350 is closed to maintain the correct pressure at the gas analyzer 325. This is a fairly simple system which can correct for a broad range of pressure variations that might occur in the sample gas. For example, it controls the pressure at the gas analyzer 325 effectively when sudden increases or decreases in the pressure of a vehicle exhaust sample occur during rapid acceleration or deceleration. Suitable pressure control circuitry is shown in
One or more sample pumps 340 draw the exhaust gas sample at constant reduced pressure (relative to atmospheric pressure) into and through the gas analyzer 325. The sample pumps 340 include a regulated power supply which is controlled by pulse width modulated pump control circuitry 341 and constant delay time motor control circuitry 343, examples of which are shown in
Constant sample delay times may be achieved by regulating the power supplied to the sample pumps. To determine the pump voltage required for a given sample flow rate (sample delay time) at a given ambient pressure, a plot of pump voltage v. ambient pressure at a given sample inlet pressure (as measured at the gas analyzer) can be generated. The slope of the plot will be inversely proportional to the sample delay time. This relationship can be represented by the following equation:
where C1 and C2 are constants that are determined by the nature of the analyzer, the type of sample pump, and the inlet pressure of the system.
Once the desired delay time is determined, and the ambient pressure is measured, the required pump voltage can be calculated and maintained using constant delay time motor control circuitry.
The exhaust gas analysis system 310 can optionally include a gas calibration component 370. The gas calibration component 370 can include one or more sources of calibration gas 375, 380. Calibration gases include gases which are components of the gas sample to be analyzed, such as carbon dioxide, various hydrocarbons, and oxides of nitrogen, among others. Calibration gases can also include ambient air, which is used to determine the voltage values for the baseline corrections for the gas analyzer 325. If more than one source of calibration gas 375, 380 is present, they can be connected by a valve 385. Valve 385 can be a three-way valve, or another suitable valve, so that either of the sources of calibration gas 375, 380 can be connected. The gas calibration component 370 can be connected to the sample line 315 by a valve 390. Valve 390 can also be a three-way valve or another suitable valve. Any type of valve can be used if the material it is made of will not react with the species in the sample.
The exhaust gas analysis system 310 can optionally include one or more filters 395, such as to remove particulate matter from the gas sample, as is well known in the art.
A microprocessor can optionally be included to calculate the mass emission rates of the species using the concentration data from the gas analyzer and data from the PCM including vehicle speed, load, and air/fuel ratio. The vehicle speed, load, and air/fuel ratio data could be obtained from a source other than the PCM if desired.
The exhaust gas analysis system of the present invention can be placed on-board a vehicle for real world driving analysis. It can also be used in stationary applications, such as in a laboratory.
While certain representative embodiments and details have been shown for purposes of illustrating the invention, it will be apparent to those skilled in the art that various changes in the compositions and methods disclosed herein may be made without departing from the scope of the invention, which is defined in the appended claims.
