CA2326469A1

CA2326469A1 - Engine emission analyzer

Info

Publication number: CA2326469A1
Application number: CA002326469A
Authority: CA
Inventors: Christopher N. Knott; Norman S. Knott
Original assignee: Oasis Emission Consultants Inc
Current assignee: Oasis Emission Consultants Inc
Priority date: 2000-11-20
Filing date: 2000-11-20
Publication date: 2002-05-20

Abstract

A method and apparatus for determining, in real time: the emission rates, for a specified engine load, of different gases in the exhaust of large industrial engines is disclosed. The apparatus comprises: gas analyzers; intake manifold temperature and pressure sensors; fuel gas temperature, pressure and flow rate sensors; a data collection buffer; a programmed computer; and a printer. Another aspect of the invention is a method and apparatus for determining, in real time, the effectiveness of catalytic converters. Another aspect of the invention is a method and apparatus for determining, in real time, the brake specific fuel consumption of large industrial engines.

Description

.rk. w ENGINE EMISSION ANALYZER
FIELD OF THE INVENTION
This invention relates to apparatus and methods for the testing of engine exhaust, particularly the exhaust from large industrial engines.
BACKGROUND OF THE INVENTION
Large, industrial engines are used for a variety of purposes, including:
electrical power generation; to drive pumps; and to drive compressors for the compression of natural gas in pipelines. In use, these engines emit a variety of gases, including carbon monoxide ("CO"), carbon dioxide ("C02") and nitrogenloxygen compounds ("NO"
and "NOz"). Concern about the environmental effect of the exhaust from these engines has resulted in widespread regulation of the operation of these engines. In many countries, these engines may not be operated without a permit granted by the relevant regulatory body.
Typically, such permits set out maximum emission limits for certain specified gases.
The permit for a particular engine typically sets out a maximum emission rate for each gas at a specified engine load. For example, a permit might specify a maximum emission rate for each gas in terms of weight of gas per unit of time and unit of engine load (the latter conventionally expressed as brake horsepower or "BHP"), such as some specified number of gramsIBHP hour. To ensure that the engine complies with the permitted emission rate, such permits also typically require that the engine emissions be monitored using a specified testing protocol. The permit may require that the emissions be monitored continuously, but, more commonly, such permits require that the engine be tested periodically, such as every year.

r The test protocols for periodic engine emission testing typically require that a series of tests of set duration be conducted. As well, the test protocols typically specify pre-test and post-test calibration procedures for the gas sensors used to measure the concentration of the test gases. Typically, when an industrial engine is tested for compliance with the permitted emission rate, neither the emission rates of the test gases nor the engine load can be easily measured directly. Rather, the test protocols provide for a variety of different measurements to be taken so as to enable the testers to estimate the emission rates of the test gases and the engine load.
In simple terms, the emission rate of a test gas is determined by: measuring the concentration of the test gas, typically in parts per million; determining the exhaust gas volumetric flow (that is, the rate of exhaust gas emission as indicated by a unit of volume over a unit of time); and using these two numbers to estimate the emission rate of the test gas.
It is, however, difficult to accurately directly measure the volumetric flow of the hot, turbulent exhaust gas. Therefore, the exhaust gas volumetric flow is also estimated.
For an engine powered by natural gas, the exhaust gas volumetric flow can be estimated from: the volumetric flow of the fuel gas; a fuel factor constant;
and the concentration of 02 in the exhaust gas. The volumetric flow of the fuel gas can be measured directly with a flowmeter, but it must be corrected for temperature and pressure to be of use in estimating the exhaust gas volumetric flow. The fuel factor constant is determined from the concentrations of the constituent compounds of the fuel gas. In simple terms, the exhaust gas volumetric flow is estimated by determining the corrected volume of fuel gas and calculating, on the basis of the fuel gas composition, what the volume will be after combustion, with a correction for the concentration of 02 in the exhaust gas.
The concentrations of the test gases can be measured directly with any of a variety of commercially available gas analyzers, including electrochemical, non-dispersive infrared and chemiluminescence gas analyzers. Typically, these gas analyzers are configured to measure the concentration (in parts per million) of the gases specified in the engine permit (usually CO, C02, NO and N02.) As well, the gas analyzers typically also measure the concentration of Oz. In the known procedures for analyzing engine emissions, the 02 measurements are used as indicators of whether the engine is running in a rich or lean combustion state.
These sensors may be cross sensitive in that their accuracy may be affected by the presence of non-target gases (referred to as "interfering gases"). The tendency of the measurements from a sensor to be affected by interfering gases is also referred to as the interference response. Cross sensitivity is tested for by exposing the sensor to be tested, and a second sensor having a known interference response, to a gas containing the test gas and the interfering gas in known concentrations.
The measurements from the gas sensors may not be stable, in that they may have a tendency to drift over time when the sensor is exposed to a gas with a constant concentration of the relevant test gas. This quality of the sensors is referred to as stability or sensor drift. Stability may be evaluated by exposing the sensor to a calibration gas and noting how the sensor measurements vary over time.
Stability is often stated as the maximum absolute percentage deviation from an average measurement recorded shortly after the measured response time of the sensor.
Further, the accuracy of the measurements from a sensor may not be consistent over a range of concentrations, particularly when the sensor is subject to rapidly changing concentrations of the test gas. This quality of the sensors is referred to as linearity.
Linearity is tested for by exposing a sensor to a known concentration of the test gas and observing the response of the sensor.
The test protocols typically require that the sensors be calibrated within a specified period before and after the relevant test. The test protocols typically require that the L
sensors are tested for calibration error and cross sensitivity before and after each test run. The calibration error test results may be used to correct the sensor's measurements, or if they fall outside of the required parameters, they may be cause to reject the results from the test run as unreliable.
The known methods of periodic testing of industrial engine emissions involve:
transporting a gas analyzer to the engine location; connecting it to the exhaust stream;
running the required tests and recording the test data; disconnecting the gas analyzer;
removing it from the test location; and processing the data to generate the test results at some later date.
The delay in processing the test data means that it is not known whether the engine has met the required emission standards until after the testing equipment has been removed from the engine site. If an engine fails a test it is necessary to reinstall the testing equipment in order to rerun the test. In some cases, tuning the engine might make the difference between meeting the permit requirements and failing the test.
However, the known testing procedures do not provide feedback of data on the engine emissions in real time and therefore offer no guidance with respect to tuning the engine.
The information required to determine the emission rates of the test gases at a certain engine load includes: the concentration of the test gases in the exhaust; the concentration of 02 in the exhaust; the fuel gas volumetric flow; the fuel gas temperature; the fuel gas pressure; and the engine load. The concentrations of the relevant gases are recorded electronically. However, with the known procedures for performing emission testing, the fuel gas volumetric flow, the fuel gas pressure; the fuel gas temperature, and the engine load are merely written down by the person conducting the test. Typically, this handwritten information is later manually entered into a computer database or spreadsheet for processing with other information a recorded during the test. It is clear that errors can occur at both the initial note taking and when the information is subsequently entered into the computer.
As well, with the known procedures for engine emission testing, the engine load is usually estimated from the work done by whatever the engine is driving. For example, if the engine is driving a compressor, the work done by the compressor may be determined by measuring the pressure and volumetric flow of gas upstream of the compressor, and the pressure of the gas downstream of the compressor. Such measurements can be used to determine the work done by the compressor, but, due to power losses in the compressor, and in the linkage between the engine and the compressor, they may not be an accurate indicator of the engine load.
Depending on these power losses, the actual engine load may be up to 12% greater than the engine load estimated by this method, resulting in errors in the emission test results.
What is needed is an engine emission analyzer which: produces engine emission information in real time; permits the generation of a test report immediately after an emission test is conducted ; reduces the risk of operator error; and which is used in combination with a more accurate source of engine load information.

M
BRIEF SUMMARY OF THE INVENTION
The invention relates to an apparatus and method for analyzing engine emissions which produce engine emission information in real time; permit the generation of test results immediately after a test is conducted; automate the acquisition and processing of fuel gas information; and use engine load information approximated in real time, from a relatively reliable source and not by estimate from the work done by the device that the engine is driving.
The various features of novelty which characterize the invention are pointed out with more particularity in the claims annexed to and forming a part of this disclosure. For a better understanding of the invention, its operating advantages and specific objects attained by its use, reference should be made to the accompanying drawings and descriptive matter in which there are illustrated and described preferred embodiments of the invention.
SUMMARY OF THE DRAWINGS
Figure 1 is a schematic view of the engine emission analyzer connected to an engine.
Figure 2 is a schematic view of the data collection bufFer.
Figure 3 is a flow chart of steps performed by the data collection buffer in receiving, organizing and sending information.
Figure 4 is a schematic view of the engine emission analyzer configured for evaluating the effectiveness of a catalytic converter.

~.
Figure 5 is a schematic view of the engine emission analyzer configured for optimizing engine performance by determining the brake specific fuel consumption.
Figure 6 is a flowchart showing initial steps of the WinStackT"~ software.
Figure 7 is a flowchart showing steps of the Emission Sourcel Compliance test mode of the WinStackT~~ software.
Figure 8 is a flowchart showing fuel measurement and fuel factor calculation steps of the Source test mode of the WinStackT~" software.
Figure 9 is a flowchart showing fuel measurement and fuel factor calculation steps of the Compliance test mode of the WinStackT"~ software.
Figure 10 is a flowchart showing engine load approximation steps of the WinStackT"~
software.
Figure 11 is a flowchart showing calibration error test steps of the WinStackT~~ software.
Figure 12 is a flowchart showing Catalyst Efficiency mode steps of the WinStackT~"
software.
Figure 13 is a flowchart showing Engine Optimization mode steps of the WinStackT""
software.
DESCRIPTION OF A SPECIFIC EMBODIMENT
In ordinary engineering parlance, and in this specification, "in real time"
means that data and actions on data are time-correlated to the sequence of physical events to which the data relate. In a more narrow sense, "in real time" means that such data and actions on data occur or are available at the actual time at which these physical events occur. In this specification, "in real time equivalent" means data and actions on data occur or are available in real time, or are so time-correlated to the sequence of physical events to which the data relate so as to provide the same benefit, for all practical purposes, as if they had occurred or were available in real time.
Figure 1 shows a conventional industrial, gas-fuelled engine (20) and a preferred embodiment of the engine emission analyzer (22) according to the invention.
The engine includes: a fuel gas inlet (24) for coupling to a fuel gas line (30);
intake manifold (26); and exhaust stack (28). The working parts of the engine (20) are of no interest and are omitted from the drawing in the interest of simplicity. The fuel gas line (30) provides fuel gas to the engine (20).
The engine emission analyzer (22), includes: a gas analyzer (40); a fuel gas flowmeter (42); a fuel gas pressure sensor (44); a fuel gas temperature sensor (46); an intake manifold pressure sensor (48); an intake manifold temperature sensor (50); a data collection buffer (52); a programmed computer (54); and a printer (56). The meters and sensors are selected for suitability in measuring selected parameters associated with an industrial gas-fuelled engine. The data collection buffer (52) is connected by suitable communication links (58) to the gas analyzer (40); the fuel gas flowmeter (42);
the fuel gas pressure sensor (44); the fuel gas pressure sensor (44); the intake manifold pressure sensor (48); the intake manifold temperature sensor (50);
and the programmed computer (54). The programmed computer (54) is connected by suitable communication link (58) to the printer (56).
In a preferred embodiment, the gas analyzer (40) is an electrochemical five-gas analyzer, such as the ECOM SG PLUST"", manufactured by ECOM America. It will be clear to those skilled in the art of exhaust gas analysis that alternative gas analyzers, such as non-dispersive infrared gas analyzers and chemiluminescence gas analyzers, may also or instead be used. The gas analyzer (40) contains sensors (not shown), also _8_ referred to as cells, for measuring the concentrations of CO, NO, NOz and 02.
These sensors are typically capable of measuring concentrations of the relevant test gases as small as a few parts per million. The measurements from NO cells are affected by changes in temperature, so the gas analyzer (40) also includes a NO cell temperature sensor (a thermocouple) (not shown).
In the embodiment shown in Figure 1, the gas analyzer (40) is a typical extractive type, in that it draws gas from the exhaust stack (28) via an internal vacuum pump (not shown). Also typically, the exhaust gas flows from the exhaust stack (28) through a heated sample line (70) (to prevent condensation) to a sample conditioner (72). The sample conditioner (72) dries the exhaust gas with a desiccant and then heats it to avoid condensation. The heated exhaust gas then enters the gas analyzer (40) where it is immediately cooled below the dew point by an internal cooler (not shown). This process is the usual means of extracting exhaust gas when using an extractive-type gas analyzer and is intended to ensure that the exhaust gas analyzed by the gas analyzer (40) is dry.
The fuel gas flowmeter (42) is a commercially available turbine-type meter, such as the 7400 Series (TM) turbine meter manufactured by Barton Instruments Systems Ltd.
The fuel gas flowmeter (42) is positioned within a rigid pipe, referred to as a meter run (74).
Typically, the manufacturer's specifications require that the turbine meter be installed at least 10 pipe diameters downstream and 5 pipe diameters upstream of any flow disturbances such as elbows or sudden expansions. In use, a turbine-type flowmeter emits a voltage-pulse output, with the pulse rate proportional to the velocity of the fuel flow. A number provided by the flowmeter manufacturer, referred to as the "K-factor", is used to convert the frequency of the pulsed output to a volumetric flow rate.
The fuel gas pressure sensor (44) is a commercially available pressure transducer, such as the PT-400 model (TM), manufactured by SRP Controls. The fuel gas pressure sensor (44) is usually positioned in the meter run (74) proximate to the fuel gas _g_ t flowmeter (42). Typically, the fuel gas pressure sensor (44) emits a fluctuating current output (4-20 mA) with the current being proportional to the fuel gas pressure.
The fuel gas temperature sensor (46) is a commercially available thermocouple, such as the type "J"T"" thermocouple manufactured by Alltemp Sensors. The fuel gas temperature sensor (46) is usually positioned in the meter run (74) proximate to the fuel flowmeter. Typically, the fuel gas temperature sensor (46) emits a fluctuating voltage output with the voltage being proportional to the fuel gas temperature.
As shown in Figure 1, the meter run (74) is connected to the fuel gas line (30) so that the fuel gas can be diverted to pass through the meter run (74). The meter run is typically connected to the fuel gas line with flexible high-pressure, TeflonT""-lined hose.
When the fuel gas bypass valves (76) are open and the fuel gas block valve (78) is closed, the fuel gas flows through the meter run (74) en route to the fuel gas inlet (24).
The intake manifold pressure sensor (48) is a commercially available pressure transducer, such as the PT-400 model T"", manufactured by SRP Controls. The intake manifold pressure sensor (48) is installed where it can sense the pressure within the intake manifold (26). Typically, the intake manifold pressure sensor (48) emits a fluctuating current output (4-20 mA) with the current being proportional to the intake manifold pressure.
The intake manifold temperature sensor (50) is a commercially available thermocouple, such as the type "J"T"" thermocouple manufactured by Alltemp Sensors. The intake manifold temperature sensor (50) is installed where it can sense the temperature of the gases within the intake manifold (26). Typically, the intake manifold temperature sensor (50) emits a fluctuating voltage output with the voltage being proportional to the intake manifold temperature.
-io-In a preferred embodiment shown in Figure 2, the data collection buffer (52) includes:
a microprocessor (90) with internal Flash EPROM (not shown) and Static RAM
(not shown); a UART (91 ); two thermocouple amplifiers (92); two current to voltage precision resistor circuits (94); a pulse signal amplifier (96); a pulse counter (97);
an analog to digital converter (98); a gas analyzer serial port (100); a programmed computer serial port (102); two temperature sensor communication ports (104); and two pressure sensor communication ports (106).
The Flash EPROM is an electrically programmable read-only memory device wherein a program can be read into the memory and then made permanent by sending a higher than normal voltage (the "flash" voltage) to the device. Such Flash EPROMs are well known in the art and are commercially available. It will be clear to those skilled in the computer art that other memory devices could be used.
The Static RAM is a form of random access memory device. Such Static RAMs are well known in the art and are commercially available. It will be clear to those skilled in the art that other random access memory devices could be used instead.
The DART (91 ) is a universal, asynchronous receiver/transmitter which controls communication between the microprocessor (90) and the serial ports. Such devices are well known in the art and are commercially available.
The thermocouple amplifiers (92) amplify the signals from the intake manifold temperature sensor (50) and the fuel gas temperature sensor (46). In a preferred embodiment, the thermocouple amplifiers (92) amplify the signal from the temperature sensors so that each 10mV of the amplified signal corresponds to 1 degree Celsius.
Such thermocouple amplifiers (92) are well known in the art and are commercially available.
-il-The current to voltage precision resistor circuits (94) convert the electric current fluctuations in the signals received from the intake manifold pressure sensor (48), and the fuel gas pressure sensor (44), into voltage fluctuations. In a preferred embodiment, the voltage precision resistor circuits (94) convert a 4 to 20 mA amperage fluctuation into a .4 to 2 V voltage fluctuation. Such current to voltage precision resistor circuits (94) are well known in the electrical art and are commercially available.
The pulse signal amplifier (96) conditions the voltage pulse signals from the fuel gas flowmeter (42). Typically each voltage pulse generated by a turbine-type flowmeter is a sine wave. The pulse signal amplifier (96) converts the sine waves of the voltage pulses into square waves, which facilitates counting the pulses. The pulse signal amplifier (96) conditions the voltage pulse signals by over amplifying the sine waves and squaring off the resulting peaks and valleys. Such pulse signal amplifiers (96) are well known in the art and are commercially available. The signal from some flow meters is preamplified by the flowmeter device, in which case the pulse signal amplifier (96) may not be necessary.
The pulse counter (97) counts the voltage pulses it receives and produces digital data representing the pulse count. The pulse count data is sent to the microprocessor (90).
Such pulse counters (97) are well known in the art and are commercially available.
The analog to digital converter (98) receives: the amplified signals from the intake manifold temperature sensor (50) and the fuel gas temperature sensor (46); the converted signals from the intake manifold pressure sensor (48) and the fuel gas pressure sensor (44); and the conditioned signal from the fuel gas flowmeter (42), all of which code information by way of absolute voltage or voltage differentials.
The analog to digital converter (98) converts these analog voltages into digital data interpretable by the microprocessor (90) and the programmed computer (54).

In a preferred embodiment, the gas analyzer (40) produces digital signals interpretable by the microprocessor (90) and the programmed computer (54). The gas analyzer (40) sends the data collection buffer (52) data strings, containing data from the sensors, one after another. Each data string has an identifier which indicates the beginning of the data string or the end of the data string. For example, the Ecom SG PLUST""
gas analyzer uses two characters comprised of bit sequences, hexadecimal '00' and hexadecimal 'FO', to identify the beginning of the data strings which it produces. Each data string is of a set length. The data strings are comprised of several fields, each of set size and set order within the data string. Each field has an identifier which distinguishes it from the other fields in a particular data string, but which is the same for that particular field in the other data strings. The data from each particular sensor in the gas analyzer is contained in the same particular field in every data string. For example the data from the NO sensor might be contained in the fourth field in the data strings. The fourth field could be identified by counting bits from the identifier which indicates the beginning of the data string and by recognizing the fourth fields distinguishing identifier. This identification could be confirmed by counting the bits or characters comprising the data identified as the fourth field.
When power is applied to the data collection buffer (52), the microprocessor (90) starts and runs a program stored in the Flash EPROM (108). As shown in Figure 3, the program instructs the microprocessor (90) to receive data from the gas analyzer serial port (110). Each bit sequence received is compared with the bit sequence known to indicate the beginning or end of a data string. This process continues until two sequential input bit sequences match the special bit sequences which signify the start or end of a data string. This sequence of bit sequences, confirmed by counting the bits in the assumed data string, and checking the field identifiers in the data, are used by the microprocessor (90) to verify that a valid data string has been received.
The bit sequences identifying the beginning or end of the data string are used as a reference point, and the required information is extracted from the data string using the known field length and order. In a preferred embodiment using the Ecom SG PLUS T""
gas analyzer, the microprocessor (90) compares a received character to "00"
hexadecimal (112). If the received character is "00" hexadecimal, the microprocessor (90) compares the next character to "FO" hexadecimal (114). If the next character is "FO"
hexadecimal then the microprocessor (90) counts the characters (115) in the data string and reviews the string for the proper field identifiers (116).
Once the relevant gas analyzer data have been extracted from the data string, the data are stored in the SRAM (118) and the microprocessor (90) stops receiving data from the gas analyzer serial port (100). The microprocessor (90) then retrieves dat representing the measurements from the fuel gas flowmeter (42); fuel gas pressure sensor (44); fuel gas temperature sensor (46); intake manifold pressure sensor (48);
and the intake manifold temperature sensor (50). The microprocessor (90) then stores these data in SRAM. The microprocessor (90) then formats the gas analyzer and other sensor information and sends it the programmed computer (122). The microprocessor (90) then resumes receiving data from the gas analyzer serial port (110) and the cycle is repeated continually.
In use, the programmed computer (54) receives digitized information from the data collection buffer (52), including data representing the following:
measurements of the concentrations of NO, NOZ CO, COZ and O2; measurements of the NO cell temperature;
measurements of the intake manifold pressure; measurements of the intake manifold temperature; measurements of the fuel gas volumetric flow rate; measurements of the fuel gas temperature; and measurements of the fuel gas pressure.
In one embodiment of the invention, the person conducting the test measures the ambient air pressure and the ambient temperature and enters these measurements into the programmed computer (54). It will be clear to those skilled in the art that suitable thermometers and barometers (not shown) can be connected to the programmed computer (54) by suitable communication link so as to transmit the ambient temperature and pressure data continuously to the programmed computer (54).

The programmed computer (54) performs a series of calculations to determine the emission rates of the test gases for the engine load, including:
Gross Calorific Value (GCVw) Of Fuel Gas (Per GPSA Standard 2172-72):
XnHn Eqn. 1 GCVw = ------------------------------- Where: Xn= Mole fraction of each component.
1 - ( ~ Xn 'J bn ) 2 Hn= Gas Heating Value of each component.
bn= Summation Factor of each component.
Fuel Gas Compressibility Factor (GPSA Method, Section 16, Using Standing &
Kantz Compressibility Curves):
Although not generally required by EPA methodologies, the programmed computer calculates the compressibility factor of the fuel gas each time it performs the calculations (pursuant to the user's instructions regarding the number of emission analyses to be performed in a particular test and the time interval between each analysis), to provide a corrected volumetric flow rate.
The gas compressibility Fis calculated instantaneously by first calculating the critical pressure and temperature of the fuel gas as follows:
PC = ~ PCn * Xn TC = ~ TCn * Xn Eqn 2 Where: Pcn = Critical Pressure of each component.
Tcn = Critical Temperature of each component.
Xn= Mole fraction of each component.
The actual pressure and actual temperature of the fuel gas are then divided by the critical pressure and critical temperature yielding reduced pressure and temperature.
Pr - Pact Tr = Tact Eqn 3 -Where: Pr = Reduced Pressure of fuel gas.
Tr = Reduced Temperature of fuel gas.
Pc = Critical Pressure of fuel gas.
Tc = Critical Temperature of fuel gas.
Reduced pressure and temperature are correlated with the Standing & Kantz compressibility curves using an internal software subroutine to produce the actual compressibility factor.

Dry Fuel F Factor (Per EPA 60 CFR 40, Method 19, Eqn. 19-13):
Fd = 10E6[(3.64 * %H ) + (1.53 * %C) + (0.57 * %S) + (0.14 * %N) - (0.46 *
%O)]/GCVO, Eqn 4 Where: %H= Weight Percentage Of Hydrogen From Ultimate Analysis.
%C= Weight Percentage Of Carbon From Ultimate Analysis.
%S= Weight Percentage Of Sulfur From Ultimate Analysis.
%N= Weight Percentage Of Nitrogen From Ultimate Analysis.
%O= Weight Percentage Of Oxygen From Ultimate Analysis.
Dry Volumetric Flow rate of Fuel Corrected For Standard Conditions And Compressibility:
DSCFMfie~ 60 * f *~(528/(460 + Tfiel )) * (Pa,"b + pfuel ) / 29.92] Eqn s z*K
Where: f = Frequency Of Pulses Generated By Turbine Meter.
Tfue1 = Instantaneous Temperature Of Fuel Gas.
Pamb = Ambient Pressure.
Pfuel = Instantaneous Pressure Of Fuel Gas.
z = Instantaneous Compressibility Factor Of Fuel Gas.
K = Pulse Conversion Factor Provided By Turbine Meter Manufacturer The meter run which is inserted into the fuel gas stream is placed downstream of conventional fuel gas scrubbers (not shown) thus rendering the fuel gas measurement on a dry basis. The instantaneous dry volumetric flow rate of fuel is calculated for each reading using the equation above.
Conversion of Dry Volumetric Flow rate of Fuel To Exhaust Effluent (Extension Of EPA 40 CFR 60 Method 19):
Qsd Fd * Qh * 20.9 Eqn 6 20.9 - % 02 Where: Qsd = Dry Effluent Volumetric Flowrate.
Fd = Dry Fuel F Factor.
.0 Qh = Heat Input Rate (Fuel Heat Content * Fuel Usage Rate).
02 = Oxygen Content Of Effluent Gas (Used For Excess Air Correction).
The dry fuel factor and fuel heat content are calculated from data obtained from the ultimate fuel analysis. The fuel usage rate and oxygen content are instantaneous measurements.
Calculation Of Emission Rates (CARB 100 Method):
50 In order to calculate emission rates , the following equation is used.

E = 1.56E-7 * PPM * Qsd * MW Eqn ~
Where: E = Emission Rate Of Pollutant (lbs/hr).
PPM = Concentration Of Pollutant.
Qsd = Dry Effluent Volumetric Flowrate.
MW = Molar Weight Of Pollutant.
The standard engineering unit for E is lbs/hr. This may be changed depending on the user's preference and may also be divided by the engine power to obtain units of grBHP-hr or gr/KW-hr. Qsd does not require a correction for standard temperature and pressure as this has already been performed in the volumetric fuel flow calculation, equation 5.
CO Interference Response (CTM030 Method, Section 6.3.1):
lco L(Rco-no ~ Cnog * Cnos ~ Ccos) + (Rco-not ~ Cno2g * Cno2s ~ Ccos)~ * 1 ~~
Eqn 8 Where: lco = CO interference response (%) Rco-no CO response to NO span gas (ppm CO).
Cnog concentration of NO span gas (ppm NO).
Cnos concentration of NO in stack gas (ppm NO).
Ccos concentration of CO in stack gas (ppm CO).
Rco-not CO response to N02 span gas (ppm CO).
Cno2g concentration of N02 span gas (ppm N02).
Cno2s concentration of N02 in stack gas (ppm N02).
NO Interference Response (CTM030 Method, Section 6.3.2):
lno = (Rno-not ~ Cno2g ) * ( Cno2s ~ Cnoxs) * 1 ~~ Eqn 9 Where: Ino = NO interference response (%).
Rno-not NO response to N02 span gas (ppm NO).
Cno2g concentration of N02 span gas (ppm N02).
Cno2s concentration of N02 in stack gas (ppm N02).
Cnoxs concentration of NOx in stack gas (ppm NOx).
Concentration Correction For Sensor Drift (CTM030 Method, Section 8.1):
CGAS (CR - CO ) * CMA ~ (CM - CO ) Eqn 10 Where: CGAS = corrected flue gas concentration (ppm).
CR = flue gas concentration indicated by gas analyzer (ppm ).
C~ = average of initial and final zero checks (ppm ).
CM = average of initial and final span checks (ppm).
CMA = actual concentration of span gas (ppm).

Load Approximation Using Manufacturer's Published BSFC Value:
BHP = HIR / BSFC Eqn a Where: HIR = Heat Input Rate (BTU / hr).
BSFC = Brake Specific Fuel Consumption Published By Engine Manufacturer (BTU /
BHP-hr).
As shown in Figure 4, another embodiment of the engine emission analyzer (22), comprising two gas analyzers and two exhaust stack temperature sensors, is useful for testing the effectiveness of catalytic converters (130) in reducing pollution.
Catalytic converters (130) are often installed in-line in the exhaust stack (28). In use, exhaust gas enters the catalytic convertor chamber (132) and passes through a catalyst element (134) of either ceramic or metallic composition. An exothermic chemical reaction occurs through the catalyst element (134) which reduces the levels of NO, N02 CO and C02, and increases the downstream exhaust temperature. The catalytic convertor (130) may not function at peak reduction efficiency due to a number of reasons, such as: an incorrect air/fuel ratio setting;
masking of the catalyst element (134) with sulphated ash from the engine lubricating oil; an exhaust temperature that is too low; or partial destruction of the catalyst element (134) due to an engine backfire. It is often desirable to be able to test the efficiency of the catalytic converter (130).
For the purpose of testing the efficiency of a catalytic converter (130), it is useful to simultaneously measure the concentrations of the test gases, and the temperature of the exhaust gas, upstream and downstream of the catalytic converter (130).
A
preferred embodiment utilizes two gas analyzers (40), an upstream gas analyzer (136), which draws exhaust gas from upstream of the catalytic converter (130) and a downstream gas analyzer (138) which draws exhaust gas from downstream of the catalytic converter (130). As well, it is useful to obtain temperature measurements -is-upstream and downstream of the catalytic converter (130). A preferred embodiment utilizes two temperature sensors: an upstream temperature sensor (140) which sense the exhaust temperature upstream of the catalytic converter (130) and a downstream temperature sensor (142) which senses the exhaust temperature downstream of the catalytic converter (142).
In one embodiment, suitable for testing catalytic converters (130), the data collection buffer (52) includes a second gas analyzer serial port (100) as shown in Figure 2.
The data collection buffer (52) uses the same procedure to recognize and extract data from each of the upstream gas analyzer (136) and the downstream gas analyzer (138), as it uses when only one gas analyzer (40) is present. When the data collection buffer (52) is connected to two gas analyzers for the purpose of testing a catalytic converter (130), the microprocessor (90) repeatedly extracts data from the gas analyzers one after the other (typically, first the upstream gas analyzer (136) and then the downstream gas analyzer(138)) and stores the data in the SRAM. Once the microprocessor (90) has retrieved and stored the data from both gas analyzers, the microprocessor (90) stops receiving data from the gas analyzer serial ports (100). The microprocessor (90) then retrieves the data representing measurements from the upstream temperature sensor (140) and the downstream temperature sensor (142). The microprocessor (90) then stores these data in SRAM. The microprocessor (90) then formats the gas analyzer and other sensor information and sends it the programmed computer (122). The microprocessor (90) then resumes receiving data from one of the gas analyzer serial ports (110) and the cycle is repeated.
One measure of an engine's efficiency is its brake specific fuel consumption (BSFC) rating. The BSFC indicates an engines rate of fuel consumption per unit of engine load. A lower BSFC indicates that an engine is more fuel efficient than an engine with a higher BSFC. The BSFC is typically calculated by dividing the fuel consumption rate by an approximated engine load. A feature of a preferred embodiment of the engine emission analyzer is that it determines in real time the fuel usage rate and the engine load as part of the engine emission analysis.
In one embodiment the invention calculates and displays the engine BSFC in real time, which helps a user tune an engine (20) in an attempt to reduce the BSFC.
During an emission test in which the engine load is being approximated on the basis of intake manifold pressure and temperature, the programmed computer (54) is receiving the information necessary to approximate the BSFC, and can be programmed to approximate the BSFC as part of the emission test.
Alternatively, as shown in Figure 5, only those sensors making the measurements necessary for calculating the BSFC need be connected to the engine, being: an intake manifold pressure sensor (48); an intake manifold temperature sensor (50); a fuel gas flowmeter (42); a fuel gas temperature sensor (46); and a fuel gas pressure sensor (44).
In a preferred embodiment of the invention, the programmed computer has a display screen and runs a program, WinStackT"", created by the inventors, which incorporates the relevant testing protocol and uses the emission limits from the relevant permit to ensure that the test results comply with the permit requirements and the test protocol. WinStackT"~ is a ~ndowsT"~ based system containing executable files and incorporating a 32 bit Sybase database platform.
WnStackT"~
will run on WindowsT"~ 32 bit platforms, such as Windows 95T"~ & Windows 98T"~, but WinStackT"~ hasn't been certified on Windows 2000T"" or Windows NTT~".
WinStackT"~ requires a computer having at least a PentiumT"~ 133MHz processor and 32 Mb of RAM, and with an SVGA type monitor. WinStackT"~ requires at least 50 Mb of hard disk space in order to operate correctly.
As shown in Figure 6, after WinStackT"~ is started and entered (148), it performs system diagnostic checks (150).

Engine emission test protocols typically specify that each gas sensor be calibrated using a gas, referred to as the span gas, with a known concentration of the gas which the sensor is designed to detect, the target gas. The test protocols also typically specify that the concentration of the target gas in the span gas used in the calibration checks must be within a set range. The bounds of this range are defined in terms of the concentration of the target gas in the actual exhaust being tested.
For example, the upper range of the allowable span gas concentration is roughly three times the concentration of the target gas in the exhaust stream. The typical practice is for the exhaust tester to take several span gas bottles with different concentrations of target gases to the test site and then guess what span gases to use.
The system diagnostic checks (150) performed by WnStackT"" include checking the measurements being received from the CO, NO, N02 and 02 sensors for the purpose of determining an appropriate span gas for the calibration of the sensors.
WinStackT~~ calculates an approved span gas range for each sensor.
WinStack T"" then prompts the user to select a test mode (152), either:
Emission Source/ Compliance (154); Catalyst Efficiency (156) or Engine Optimization (158).
"Source" and "compliance" are terms used by regulators to distinguish different test criteria. A source test generally has a more rigorous test protocol than a compliance test. A permit for a particular engine might specify that a compliance test be conducted every six months and a source test every four years.
The Catalyst Efficiency test mode (156) is used to test the effectiveness of catalytic converters. The Engine Optimization test mode (158) is used to approximate the BSFC for the purpose of tuning an engine.

As shown in Figure 7, when the user selects the Emission SourcelCompliance test mode (154), the user is prompted to enter the relevant permit information, or select the permit information from the WinStackTM database if the permit information has been previously entered into the computer (160). The permit information stored by the WinStackT"" database includes the state or province the permit is issued in, when the last test was performed, the permitted emission levels and units, and the maximum permitted time between each source or compliance test. WinStackTM
compares this data with the measured emission levels during the test and indicates any breach of the permit requirements. WnStackT"~ also notifies the user of any upcoming emission tests, based on the permit number, the state or province, and the time since the last test.
WinStackTM then prompts the user to enter, into the appropriate fields, the following information: the facility location, the relevant environmental board; the gas analyzer serial number; the facility operator; and an identifier for the person performing the emission test (162). The user then indicates whether a Source or Compliance test will be performed (164). The user then enters the ambient conditions at the test site (166). The ambient conditions are the ambient barometric pressure and ambient temperature which are measured by any suitable means.
WinStackT"~ then prompts the user to select a method for determining the fuel gas volumetric flow (168).
If the user selected a Source test in a previous step (169), then WinStackT~~
will require real-time fuel gas volumetric flow, fuel gas temperature and fuel gas pressure measurements. Therefore, for a Source test, it is necessary to divert the fuel gas through a fuel metering system. As shown in Figure 8, WinStackT~~
prompts the user to indicate the model of flowmeter being used to measure the fuel gas volumetric flow and to either enter a K-factor or edit the K-factor displayed by WinStackT"~ based on the most recent calibration of the flowmeter (170).

WinStackT"" then prompts the user to enter a recent fuel gas composition (172).
WnStackT"~ then permits the user to either: enter the gross calorific value and the molecular weight of the fuel, typically from a fuel gas composition sheet; or enter the various mole fractions of the components of the fuel gas (obtained from an earlier analysis of the fuel gas) (174) in which case WinStackT"" will calculate gross calorific value and the molecular weight of the fuel. WinStackT"~ then calculates the mass percentages of Carbon, Oxygen, Sulfur, Nitrogen, Hydrogen, the pseudo critical properties and the fuel F Factor (176). All data calculated and recorded during this step is saved to the WinStackT"~ database for later retrieval and report generation.
As shown in Figure 9, if a compliance test was selected in a previous step (177), then WinStackT"" prompts the user to enter a static value for the fuel volumetric flow rate, generally based on a measurement from a flowmeter or a similar estimation.
The volumetric flow rate must be corrected for temperature and pressure.
WinStackT"~ permits the user to enter either an actual fuel flow rate (178) as directly measured by the relevant metering device; or an already corrected fuel flow rate (180). If the user elects to enter a corrected flow rate, the user simply types in the corrected volumetric flow rate. If the user elects to enter an actual flow rate, then the user must enter the fuel flow, pressure and temperature (182). WinStackT~~
then converts the actual flow to a corrected flow based on the entered temperature and pressure. The user then selects (183) between entering a generic fuel F factor (184) or having WinStackT~~ calculate the fuel F factor based on the fuel gas composition. If the user enters a generic fuel F factor (184), WinStackTM
estimates a generic value for the gross calorific value of the fuel (1000 Btulcf). If the user elects to calculate the fuel F factor based on the fuel gas composition, then the user enters the fuel gas composition information (186). The user may then select for the programmed computer to calculate the fuel parameters or the user may enter the parameters from a gas analysis, if they are available (188). WinStackTM then calculates the parameters required by EPA method 19 (190).

As shown in Figure 7, for both the Compliance and Source test modes, the user is then prompted to set the total number of measurements recorded during each test and the time interval between each reading (200).
WinStackT"~ then prompts the user to select the relevant engine model for the test (202). WinStackT"~ lists engine models based on the manufacturer, aspiration (natural or turbo-charged) and combustion (rich burn or lean burn).
Depending on the selected engine model, WinStackT"" then permits the user to select from three methods for approximating the engine load (204). Load can be approximated based on the manifold conditions (208), the manufacturer's rated brake specific fuel consumption (BSFC, typically in BTUIBHP-hr) (214), or the manufacturer's rated engine load for specified engine RPM (226).
As shown in Figure 10, if the user elects to approximate the engine load from manifold conditions (208) in data acquisition mode, W'rnStackT"" waits for the chosen delay period and then accepts engine manifold pressure and temperature readings (210). Then WinStackT"" correlates the measured intake manifold temperature and pressure with the engine manufacturer's load curves (which relate intake manifold pressure and temperature with engine load) and corrects for ambient temperature and barometric pressure (already entered by the user) per the manufacturer's guidelines (211). WinStackTM uses Newton's Interpolation Method to approximate the engine load at temperature and pressure measurements that do not fall at the discrete points defined by the load curve representations. The user may tune the engine (212). WinStackT"~ repeatedly accepts intake manifold temperature and pressure measurements and repeats the above calculations based on the number of measurements selected by the user (213).
As shown in Figure 10, if the user elects to approximate the engine load with the manufacturer's brake specific fuel consumption (BSFC) (214), in data acquisition mode, WinStackT"" waits for the chosen delay period and then accepts fuel flow, fuel temperature and fuel pressure measurements (216). WinStackT"" then calculates the heat input rate from the corrected fuel volumetric flow rate and the gross calorific value of the fuel (218). WinStackT"~ then approximates the engine load (typically BHP) based on the heat input rate (typically Btulhr) and the manufacturer's published BSFC values (typically BtuIBHP-hr) (220). The user may tune the engine (222). WinStackT"~ repeatedly accepts fuel flow, fuel temperature and fuel pressure measurements and repeats the above calculations based on the number of measurements selected by the user (224).
As shown in Figure 10, if the user elects to approximate the engine load from the manufacturer's engine load for specified engine RPM ratings (226), then the user enters the engine RPM. In data acquisition mode, WinStackT~~ waits for the chosen delay period and then accepts 02 measurements (228). WinStackT"" then retrieves the rated load from the database and correlates the entered RPM and the measured 02 with the manufacuture's ratings to approximate the engine load (230).
WinStackT"~ repeatedly accepts 02 measurements and repeats the above approximation based on the number of measurements selected by the user (232).
As shown in Figure 7, WinStackT~~ then checks for blank entries or erroneous inputs, and displays the previously entered testing parameters (240), so as to permit the user to adjust or correct any values through a 'Preferences' section of the menu.
Once the user indicates that he or she is satisfied with the entries, ~nStackT"~ goes into the data acquisition mode.
WinStackT"" then displays all the data channels coming from the data collection buffer (242). The display is configured so as to make a viewer aware of any unexpected inputs (ie: inputs that are not within the expected channel ranges) (244), so as to permit the user to remedy a faulty sensor, or correct a situation where a sensor has not been installed or connected properly.

As shown in Figure 7, when the user is satisfied that all sensors are reading correctly, he or she instructs WinStackT"" to commence the pre-test calibration error phase (246). WinStackT~~ requires the user to follow standard United States Environmental Protection Agency ("EPA") procedures to ensure that the gas analyzer correctly reads the gas concentration levels for NO, N02, CO & 02..
The sensors in the gas analyzer are also calibrated after the emission test.
Figure 11 shows the calibration error procedure used for both the pre-test calibration error phase and the post-test calibration error phase.
The individual gas sensors in the gas analyzer are each tested by exposing them to a gas, referred to as a span gas or calibration gas, containing a known concentration of the gas which each is designed to detect. The testing of each gas sensor involves two phases: the zero to span phase; and the span to zero phase. In the zero to span phase, the gas sensor is exposed to ambient air and then to the span gas. In the span to zero phase, the sensor is exposed to the span gas and then to the ambient air.
The user first selects one of the gas sensors to test (250). The user connects the appropriate span gas cylinder to the gas analyzer and then performs a zero to span analysis (252). During the zero to span analysis, WinStackT~~ compares the actual data received from the gas analyzer, with the known span gas concentration, typically stamped on the calibration gas cylinder. The time it takes a gas sensor to respond to 95% of the step change from zero to span or span to zero, is referred to as the response time of the gas sensor. After the gas sensor has sensed 95% of the step change from zero to span, WinStackT"~ will compare the gas sensor measurements to the known concentration of the target gas in the span gas to determine if the gas sensor measurement is within the EPA tolerance (254). If the reading is not within the EPA tolerance, WinStackT"~ will require the user to conduct another zero to span test of the sensor (264).

Once the sensor has passed the zero to span test, WnStackT~~ requires the user to perform a span to zero test on the sensor (256). The user must disconnect the gas analyzer from the span gas cylinder, allowing the analyzer to sample ambient air, which is the zero reference. After the gas sensor has sensed 95% of the step change from span to zero, WinStackT"" will compare the gas sensor measurements to the known concentration of the target gas in the ambient air to determine if the gas sensor measurement is within the EPA tolerance (258). If the reading is not within the EPA tolerance, WinStackT"~ will require the user to re-test the sensor (including the zero to span test). If the reading is within the required EPA
tolerance, WinStackT"~ saves all the raw data for both the span to zero and zero to span tests to the database (260). The user then conducts the same tests of the remaining sensors. WinStackT"" checks that all the sensors have been tested (262). Once all the sensors have passed the pre-calibration error test, the user proceeds to the testing phase.
When the testing phase is entered, WinStackT"~ prompts the user to select either line or bar graph format for real-time graphical display purposes. WinStackT"" then prompts the user to select between two modes: "Tune Engine" and "Start Test".
The "Tune Engine" option allows the user to view all real-time levels (engine emission, engine load, engine exhaust and fuel consumption levels) without saving any data to the database. The purpose of this mode is to provide an opportunity to the user to tune or adjust the engine to be compliant with the permit emission limits.
The "Start Test" option allows the user to view all levels, and records all raw data to the database for eventual report generation. Prior to initiating the "Start Test" mode, most users will generally have already made an attempt to tune the engine to meet the compliance requirements of the permit.
The only significant difference between the "Tune Engine" and "Start Test"
modes is that in the "Start Test" mode all raw data is saved to the database, whereas in the ~ , "Tune Engine" mode no data is saved to the database. The description that follows refers to the "Start Test" mode, but it also applies to the "Tune Engine"
mode.
As shown in Figure 7, in the "Start Test" mode, raw data is obtained from the data collection at the time intervals previously stipulated by the user (270). The raw data is forwarded to WinStackT"" from the data collection buffer in a specific format and sequence. The data is transferred to the computer memory for analysis and plotting (271). This raw data is converted to standard engineering units by WinStackT"".
Depending on the load approximation method previously selected by the user [28], WinStackT"~ approximates the engine load using: pressure and temperature measurements (engine manifold method); corrected fuel flow measurements (engine BSFC method); or the manufacturer's rated load. WinStackT"" calculates the real-time fuel gas compressibility with a subroutine that uses the Standing and Kantz compressibility curve approximation method utilizing calculated pseudo critical gas properties and the measured fuel gas temperature and pressure.
WinStackT~" calculates the exhaust flow rate based on the type of testing that was previously selected by the user. If the user selected a Source test, WinStackT~~ takes the actual fuel flow and corrects it for pressure, temperature and compressibility to arrive at a corrected fuel flow. This value, coupled with other previously calculated values (ie: the Fuel F factor, the Gross Calorific Value etc.) is then used to calculate the real-time exhaust flowrate If the user selected a Compliance test, WinStackT"~
uses the previously entered static value for the fuel flow.
WinStackTM calculates the emission levels based on the user selected engineering unit, the measured concentration levels, and the exhaust flow calculation.
WinStackT"~ calculates the engine fuel consumption and the engine BSFC in real-time, based on the fuel gas volumetric flow measurement and the engine load approximation. WinStackT~" presents the user with: real-time emission levels, the engineering units of which may be changed at any time through a menu selection; a corrected fuel flow measurement; an approximated engine load; a calculated exhaust flow; the engine brake specific fuel consumption (BSFC); and all raw data readings on individualized windows-style tab pages. The user may make adjustments to the engine and view resultant levels in real-time (272).
WinStackT~~
repeats the data collection based on the number of readings selected by the user.
(273). Then the post-calibration error check is performed (274), using the same procedure as the pre-calibration error check (Figure 11). WinStackT"~ notifies the user if any sensors fail the post-test calibration check (275). Once the test is complete, all raw data is saved to the database to allow report generation at a time of the user's choosing.
As shown in Figure 4, when the user wishes to use the Catalyst Efficiency test mode, two gas analyzers are used: an upstream gas analyzer (136), which draws exhaust gas from upstream of the catalyst element (134) and a downstream gas analyzer (138) which draws exhaust gas from downstream of the catalyst element (134). As well, two temperature sensors are used: an upstream temperature sensor (140) which sense the exhaust temperature upstream of the catalyst element (134) and a downstream temperature sensor (138) which senses the exhaust temperature downstream of the catalyst element (134).
As shown in Figure 12, when the user selects the Catalyst Efficiency test mode, WinStackT~" prompts the user to enter, into the appropriate fields, the following information : the facility location, the relevant environmental board; the gas analyzer serial number; the facility operator; an identifier for the person performing the emission test; the ambient conditions at the test site; and whether a Source or Compliance test will be performed (280). The ambient conditions are the ambient barometric pressure and ambient temperature, which are measured by any suitable means.

._ WinStackTM then requires the user to select an engine model to test (282).
WinStack T"" lists engine models based on the manufacturer, aspiration (natural or turbo-charged) and the combustion (rich burn or lean burn). WinStackT~" then permits the user to set the total number of readings to be recorded during a test, and the time interval between each reading (284).
WinStackT"~ then displays the previously entered information; checks for blank fields or erroneous inputs and notifies the user if there is an error in the inputted data; and permits the user to adjust or correct any values through the 'Preferences' section of the menu (286). The test information and ambient condition data is saved to the WinStackT~~ database for later retrieval and report generation. Once the user indicates that he or she is satisfied with the entries, WinStackT"" goes into the data acquisition mode.
WinStackT~" then displays all the data channels coming from the data collection buffer (288). The display is configured so as to make a viewer aware of any unexpected inputs (ie: inputs that are not within the expected channel ranges) (290), so as to permit the user to remedy a faulty sensor, or correct a situation where a sensor has not been installed or connected properly.
When the user is satisfied that all sensors are reading correctly, the user then instructs WinStackT"" to commence the pre-test calibration error phase. In the Catalyst Efficiency test mode, WinStackT"" goes through the same pre-test calibration error test as it does in the Emission Source/ Compliance test mode, except that the calibration error tests are performed on two gas analyzers (292, 294, 296 and 298).
Once the user is satisfied with the pre-test calibration error tests, the user instructs WinStackT"" to commence testing. WinStackT~~ waits for the chosen delay period and then accepts measurements from the data collection buffer (300).
WinStackT""

, ~_ transfers the sensed data to the computer memory for analysis and plotting (302).
WinStackT"" calculates and displays, in real time, any difference in the upstream and downstream levels of NO, N02 CO the C02. WinStackT"" compares the upstream temperature to the temperature necessary to stimulate the desired chemical reaction between the catalyst element and the exhaust gas. WinStackT~~ also displays the temperature differential between the upstream and downstream exhaust gas, which is an indicator of the extent to which the desired exothermic reaction is occurring.
The user may tune the engine if required (304). WinStackT"~ repeats the data acquisition and processing steps for the previously entered test duration (306), providing real time feedback for any tuning or adjustments made by the user.
As shown in Figure 13, when the user selects the Engine Optimization test mode, WinStackT"" prompts the user to enter, into the appropriate fields, the following information: the facility location, the relevant environmental board; the gas analyzer serial number; the facility operator; and an identifier for the person performing the emission test (310). Then the user enters the ambient conditions at the test site (312), being the ambient barometric pressure and ambient temperature, which are measured by any suitable means.
WinStackT"" then prompts the user to select an engine model to test (314).
WinStackT~~ lists engine models based on the manufacturer, aspiration (natural or turbo-charged) and the combustion (rich burn or lean burn). WinStackT"~ will present the user with a set of load curves that pertain to the engine model selected.
The user selects the curve that most closely matches current field conditions (316).
To approximate the engine load, WinStackT"" correlates the measured intake manifold temperature and pressure with the engine manufacturer's load curves (which relate intake manifold pressure and temperature with engine load) and corrects for ambient temperature and barometric pressure (already entered by the user) per the manufacturer's guidelines. WinStackT"~ uses Newton's Interpolation Method to approximate the engine load at temperature and pressure measurements that do not fall at the discrete points defined by the load curve representations.
In order to approximate the engine brake specific fuel consumption (BSFC), WinStackT"~ must obtain a corrected fuel flow rate. WinStackT"" prompts the user to validate the model of flow meter that will be used to measure the fuel and enter a K-factor based on the most recent calibration of the flow meter [45]. The user then enters a recent fuel gas composition (318). WinStackT"~ can calculate the gross calorific value and the molecular weight of the fuel based on the fuel gas composition, or have the user directly enter these parameters from a fuel gas composition sheet. WinStackT"" then calculates the pseudo critical properties of the fuel gas (320), which are required for the fuel gas compressibility calculation.
WinStackT"" then permits the user to set the total number of readings to be recorded during a test, and the time interval between each reading (322). WinStackT~~
then displays the previously entered information (324), and permits the user to adjust or correct any values through the 'Preferences' section of the menu. Once the user indicates that he or she is satisfied with the entries, all data calculated and recorded during this step is saved to the WinStackT"" database for later retrieval and report generation; and WinStackT"~ goes into the data acquisition mode.
WinStackT"" then displays all the data channels coming from the data collection buffer (326). The display is configured so as to make a viewer aware of any unexpected inputs (ie: inputs that are not within the expected channel ranges) (328), so as to permit the user to remedy a faulty sensor, or correct a situation where a sensor has not been installed or connected properly.
WinStackTM then prompts the user to select either line or bar graph format for real-time graphical display purposes. When the user instructs WinStackT"~ to start the test, WinStackT"" waits for the chosen delay period and then accepts raw data from ._ , the data collection buffer at the time interval previously stipulated by the user (330).
WinStackT"~ transfers the data to the computer memory for analysis and plotting (332). WinStackT"~ calculates an approximated brake specific fuel consumption (BSFC) from the approximated engine load determined from the intake manifold pressure and temperature measurements; the corrected fuel gas flow; and the gross calorific value of the fuel. WinStackT~~ displays the real-time BSFC, corrected fuel flow measurement, and approximated engine load, on individualized windows-style tab pages. The user may make adjustments to the engine (334) and view the effects in real-time. WinStackTM repeats the data collection based on the number of readings selected by the user (336). Once the test is complete, all the raw data is saved to the database to allow report generation at a later date.
The foregoing is a description of a preferred embodiment of the invention which is given here by way of example. The invention is not to be taken as limited to any of the specific features as described, but comprehends all such variations thereof as come within the scope of the appended claims.

Claims

1. An apparatus for determining the emission rate of a selected test gas, or selected test gases, emitted by a gas-fuelled engine, the apparatus comprising:
(A) means for making ongoing measurements of the engine fuel gas volumetric flow rate;
(B) a gas analyzer capable of making ongoing measurements of the concentration of at least one test gas;
(C) a programmed computer having an information display means; and (D) means for communicating the fuel gas volumetric flow rate measurements and the concentration measurements, to the programmed computer in real time;
wherein when the apparatus is in use, the fuel gas volumetric flow rate measurements and the concentration measurements are communicated to the programmed computer, which calculates and displays, in real time equivalent, the emission rate of the at least one gas.

2. The apparatus of claim 1 wherein the means for communicating the measured volumetric flow rate and the measured concentration of the at least one gas, comprises a data collection buffer having communication links to: the gas analyzer;
the means for measuring the fuel gas volumetric flow rate; and the programmed computer, wherein when the apparatus is in use, the data collection buffer organizes the measurements it receives so that the measurements can be understood by the programmed computer and communicates the organized measurements to the programmed computer.

3. The apparatus of claim 1 further comprising:
(A) a fuel gas temperature sensor capable of making ongoing measurements of the fuel gas temperature;

(B) a fuel gas pressure sensor capable of making ongoing measurements of the fuel gas pressure; and (C) means for communicating the fuel gas temperature measurements and the fuel gas pressure measurements, to the programmed computer;
wherein when the apparatus is in use, the measured fuel gas temperature measurements and the fuel gas pressure measurements are communicated to the programmed computer, which uses the fuel gas temperature measurements and the fuel gas pressure measurements to correct the fuel gas volumetric flow rate measurements for compressibility and deviations from standard atmospheres.

4. The apparatus of claim 3 wherein the means for communicating the fuel gas temperature measurements and the fuel gas pressure measurements, to the programmed computer, comprises the data collection buffer connected by communication links to the fuel gas temperature sensor and the fuel gas pressure sensor.

5. The apparatus of claim 1 wherein the means for measuring the fuel gas volumetric flow rate comprises a turbine-type flowmeter, and wherein the fuel gas temperature sensor and the fuel gas pressure sensor are positioned proximate to the turbine-type flowmeter.

6. The apparatus of claim 1 further comprising means for determining engine load in real time equivalent.

7. The apparatus of claim 6 wherein the means for determining engine load comprises:
(A) means for making ongoing measurements of the engine intake manifold pressure;
(B) means for making ongoing measurements of the engine intake manifold temperature;

(C) means for communicating the engine intake manifold pressure measurements, and the engine intake manifold temperature measurements, to the programmed computer in real time equivalent;
(D) a computer memory, readable by the programmed computer, containing power information for the engine, relating engine load to engine intake manifold pressure and temperature;
wherein, when the apparatus is in use, the programmed computer receives engine intake manifold pressure measurements and engine intake manifold temperature measurements and calculates the engine load in real time equivalent.

8. The apparatus of claim 7 wherein the means for communicating the engine intake manifold pressure measurements, and the engine intake manifold temperature measurements, to the programmed computer in real time equivalent, comprises: the data collection buffer connected by communication links to the means for making ongoing measurements of the engine intake manifold pressure and means for making ongoing measurements of the engine intake manifold temperature.

9. The apparatus of claim 6 wherein the means for determining engine load comprises:
(A) a computer memory, readable by the programmed computer, containing brake specific fuel consumption information for the engine, relating the fuel heat input rate to the engine load;
(B) means for determining the gross calorific value of the fuel;
(C) means for providing the gross calorific value of the fuel to the programmed computer;
(D) means for determining the corrected fuel gas flow rate; and (E) means for providing the corrected fuel gas flow rate to the programmed computer;

wherein, when the apparatus is in use, the programmed computer uses the brake specific fuel consumption information, the gross calorific value of the fuel and the corrected fuel gas flow rate to calculate the engine load.

10. The apparatus of claim 9 further comprising:
(A) a computer memory, readable by the programmed computer, containing power information for the engine, relating engine load to engine RPM; and (B) means for communicating the engine RPM to the programmed computer wherein, when the apparatus is in use, the programmed computer receives the engine RPM and calculates the engine load.

11. An apparatus for determining the emission rate, with respect to engine load, of a selected test gas, or selected test gases, emitted by a gas-fuelled engine, the apparatus comprising (A) a data collection buffer;
(B) A turbine-type flowmeter capable of making ongoing measurements of the engine fuel gas volumetric flow rate, connected to the data collection buffer by communication link;
(C) a fuel gas temperature sensor capable of making ongoing measurements of the fuel gas temperature, connected to the data collection buffer by communication link;
(D) a fuel gas pressure sensor capable of making ongoing measurements of the fuel gas pressure, connected to the data collection buffer by communication link;
(E) a gas analyzer capable of making ongoing measurements of at least one test gas, connected to the data collection buffer by communication link;

(F) an intake manifold pressure sensor capable of making ongoing measurements of the engine intake manifold pressure, connected to the data collection buffer by communication link;
(G) an intake manifold temperature sensor capable of making ongoing measurements of the engine intake manifold temperature, connected to the data collection buffer by communication link;
(H) a programmed computer including an information display and connected to the data collection buffer by communication link;
(I) means for providing the engine load to the programmed computer;
wherein, when the apparatus is in use, the data collection buffer organizes the measurements it receives so that the measurements can be understood by the programmed computer and communicates the organized measurements to the programmed computer, which calculates and displays, in real time equivalent, the engine load and the emission rate of the at least one test gas.

12. An apparatus for determining the emission rate of a selected test gas, or selected test gases, emitted by a gas-fuelled engine, the apparatus comprising:
(A) a flow rate sensor for making a series of measurements of the engine fuel gas volumetric flow rate;
(B) a gas analyzer for making a series of measurements of the concentration of at least one test gas in the engine exhaust;
(C) means responsive to the flow rate sensor and to the gas analyzer for providing as real-time outputs a flow rate signal and a test gas concentration; and (D) a programmed computer for receiving in digital format as inputs the flow rate signal and test gas concentration signal and for providing in response thereto an output representative of the real-time emission rate of the test gas.

13. The apparatus as defined in claim 12, wherein the sensors provide the flow rate signal and test gas concentration signal as respective analog output signals, and additionally comprising a buffer connected to and interposed between the sensors and the computer, the buffer including an analog/digital converter for converting the analog output signals to digital signals and transmitting the digital signals to the computer.

14. A method of analyzing engine emissions, comprising the steps of:
(A) sensing the volumetric flow rate of the engine fuel gas;
(B) sensing the concentration of at least one test gas in the engine exhaust; and (C) in response to the sensed volumetric flow rate and concentration, providing an output representative of the emission rate of the test gas.