Note: Descriptions are shown in the official language in which they were submitted.
CA 02363378 2001-11-20
ENGINE EMISSION ANALYZER
FIELD OF THE INVENTION
This invention relates to methods and portable apparatus for testing engine
exhaust, particularly the exhaust from large industrial engines.
BACKGROUND OF THE INVENTION
Large, industrial engines are used for a variety of purposes, including: to
generate electrical power; 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 ("COZ") and nitrogen/oxygen
compounds ("NO" and "NO2") Concern about the environmental effect of the
exhaust
from these engines has resulted in widespread regulation of the operation of
these
engines, and particularly regulation of exhaust emissions. 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 specified gases.
The permit for a particular engine may merely set out a maximum emission rate
for
each specified gas or it may specify a maximum emission rate for each
specified gas
at a specified engine load. 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.
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
-1-
CA 02363378 2001-11-20
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.
It is difficult to measure the weight per unit time of a given regulated
effluent
gas (test gas) directly, so it is conventional to measure the concentration of
the test
gas in the exhaust and the volume of the exhaust gas and from those
measurements
compute the rate of emission of the test gas in pounds per hour (Ibs/hr) or
other
designated units of measurement. 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 or at least estimating 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, further difficult to accurately directly measure the
volumetric flow
of the hot, turbulent exhaust gas. Therefore, conventionally, 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 oxygen (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.
-2-
CA 02363378 2001-11-20
As well, using previously known procedures and conventional portable
apparatus for engine emission testing, the engine load is usually estimated
from the
work done by whatever equipment 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. While some tolerance
for such
errors can be taken into account when the regulatory authority sets emission
standards, it would be preferable to obtain more accurate measurements of
engine
load.
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 contain sensors (also referred to in the trade as "cells") for
measuring 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 02. In the known procedures for analyzing engine emissions,
the OZ
measurements are used as indicators of whether the engine is running in a rich
or lean
combustion state.
The sensors may be cross-sensitive in that their accuracy may be affected by
the presence of non-target gases (referred to as "interfering gases"). Cross-
sensitivity
is also referred to as the interference response. A sensor's cross-sensitivity
to a
particular interference gas is tested by exposing the sensor and a sensor
targeted to
the interference gas, to a test gas containing the interference gas but not
containing
the target gas of the sensor being tested for cross-sensitivity. For example,
a N02
-3-
CA 02363378 2001-11-20
sensor's cross-sensitivity to NO would be tested for by exposing the NOZ
sensor and
a NO sensor to a test gas containing NO but not containing N02. Any response
by the
N02 sensor to the test gas would be due to cross-sensitivity. Cross-
sensitivity may be
quantified by comparing the interference response of the sensor being tested
(the NOZ
sensor in the example) with the response of the interference-gas-targeted
sensor (the
NO sensor in the example).
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, the two terms implying opposite characteristics.
Sensor drift
may be evaluated by exposing the sensor to a calibration gas and noting how
the
sensor measurements vary over time. The extent of sensor drift 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 a sensor is
referred to
as degree of linearity of the sensor, or simply "linearity". Linearity is
tested by first
exposing a sensor to at least two gases having different concentration of the
test gas,
one after the other, and observing the response of the sensor over time to the
different
concentrations of the test gas.
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 sensors be 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, and possibly to
indicate the need
to replace the sensor.
-4-
CA 02363378 2001-11-20
The testing procedure typically involves 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 conventional delay in processing the test data means that it is not known
whether the engine has met the required emission standards until the testing
is
complete and the data can be processed, which typically does not occur until
after the
testing equipment has been removed from the engine site. If it turns out from
the later
data processing that an engine has failed a test, it is typically necessary to
re-transport
the testing apparatus to the test site and 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, various
previously
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.
For an engine powered by natural gas, the data required to determine the
emission rates of the test gases at a certain engine load include: the
concentration
of the test gases in the engine 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 specified test 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 recorded during the test. It is clear that
errors can
occur both at the initial note-taking and later when the information is
subsequently
entered into the computer.
-5-
CA 02363378 2001-11-20
What is needed is a portable engine emission analyzer that: 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
is used in
combination with a more accurate source of engine load information.
-6-
CA 02363378 2001-11-20
BRIEF SUMMARY OF THE INVENTION
In ordinary engineering parlance, and in this specification, "in real time"
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
simultaneously with the physical events to which they relate.
The method of emissions testing and determination of the present invention
differs from most conventional prior methods in that it comprises a method of
testing
a engine and analyzing the exhaust emissions of the engine and making the
necessary computations to determine the emission rate of a specified test gas
or test
gases in real time. Most conventional prior methods are incapable of providing
real-
time results but instead require a delay between the measurement stage of the
method and at least part of the computation stage of the method.
The preferred method of analyzing the exhaust emissions of gas-fueled engines
in real time according to the invention is capable of providing a relatively
accurate
determination of the emission rate for the test gas (to the extent that the
sensors used
are reliable and that the test equipment is accurately calibrated).
In one aspect of the invention, suitable for use for determining the amount of
a specified test gas in the exhaust of an engine powered by natural gas fuel
or other
gaseous fuel, the method includes the steps of:
(a) measuring the flowrate, temperature and pressure of the fuel gas;
(b) measuring the relative concentration of the test gas in the engine
exhaust;
(c) computing a volumetric flowrate of the exhaust gas from the measurement
data
representing the flowrate, temperature and pressure of the fuel gas;
(d) computing an emission rate (conventionally expressed in the units Ibs/hr)
of the test
gas from the exhaust gas volumetric flowrate data and the data representing
the
relative concentration of the test gas;
CA 02363378 2001-11-20
(e) determining or measuring the engine load at which steps (a) and (b) occur,
and
computing an emission rate per engine load (conventionally expressed in the
units
Ibs/BHP-hr) from the computed emission rate and engine load;
(f) optionally, recording or displaying the calculated test gas emission rate;
and
(g) optionally performing further calculations and tabulations of the gas
emission rate
data, e.g. comparing the calculated test gas emission rate to the maximum
permitted
emission rate and, if the calculated emission rate is greater than the
permitted
emission rate, displaying a warning or alarm or record or display of this
result.
In the foregoing summary, and in this specification generally, the step of
"measuring" may be a composite step involving measurement of one or more given
parameters and then performing a calculation on it to derive an estimated
value for the
parameter whose value is sought. In other words, measured values include
estimated
values where direct measurement of a parameter is difficult. Further, in the
above
summary, and in this specification generally, the step of "computing" or
"calculating"
includes providing as an interim or final output the results of the
computation in digital
data format. Further, the step of "measuring" is to be taken as including, as
necessary, the conversion of any analog measurement data to digital format.
All of
the foregoing computation steps may be performed in real time by a programmed
computer and may be repeated a pre-selected number of times at pre-selected
time
intervals.
As discussed above, the method normally includes the step of determining the
engine load (typically expressed in BHP) so that an emission rate per engine
load
(typically Ibs/BHP-hr) may be calculated and recorded or displayed as
required.
However, in some cases this step may not be necessary. For example, if the
maximum emission levels in a permit are not in terms of emission rate per
engine load,
there may be no need to determine the engine load or compute the emission rate
per
engine load.
_g_
CA 02363378 2001-11-20
Preferably, the step of determining the engine load comprises correlating in
real
time engine data, including intake manifold temperature and pressure, and the
engine
RPM (usually manually entered as an input into the programmed computer) with a
load
curve for the engine. A load curve appropriate to a particular engine is
typically
prepared by, and obtained from, the manufacturer of the engine. A load curve
is
typically specific to a particular model, but a load curve may instead be
specific to a
particular conformation (e.g. turbo-charged or naturally aspirated) of a
particular model,
or to a particular conformation of a particular model under particular
operating
conditions (e.g. intercooler water temperature or manifold temperature).
Correlating
the engine data to the load curve involves: selecting the appropriate load
curve, which
entails comparing some of the known engine data with the load curve selection
criteria;
and then using the load curve data and additional engine data to calculate the
engine
load. Preferably, discrete values not found on the engine manufacturer's load
curve
are calculated using Newton's Method of Interpolation. Preferably the data
from each
load curve is incorporated into a computer routine (sometimes referred to in
the trade
as a "function") along with Newton's Method of Interpolation, such that the
programmed computer may "call the function", that is, provide a particular
routine with
values for the required variables and instruct the routine to calculate the
engine load.
Alternatively, the engine load may be approximated by correlating the engine
RPM with the engine manufacturer's RPM-engine load specifications(preferably
stored
in the computer-readable database).
In another aspect of the invention, an engine emission analyzer comprises a
programmed computer connected to: a data collection buffer, a computer-
readable
database and a display device. The data collection buffer is configured to
connect to,
and receive data from: a gas analyzer for sensing the relative concentration
of the
specified test gas or gases and Oz; an intake manifold temperature sensor; an
intake
manifold pressure sensor; a fuel gas flowmeter; a fuel gas pressure sensor;
and a fuel
gas temperature sensor. The data buffer is configured to: accept the sensed
data
(some in analog and some in digital form), digitize the analog data, organize
the data
-9-
CA 02363378 2001-11-20
into batches that can be recognized by the programmed computer, and send the
batches to the programmed computer. The computer-readable database is for
storing
data used in performing the engine emission analysis, including: the
specifications of
the engine being tested; the maximum emission limits for each of the test
gases
(pollutants) specified in the relevant permit; the testing parameters; and
various
calculation factors. The display device is preferably a display screen, but
may be a
printer or any other suitable means for indicating, recording or storing test
results for
the benefit of the user. The programmed computer is programmed to receive
batches
of data from the data buffer; to call up and receive data from the database as
required;
to perform the engine emission analysis calculations; and to send data
representing
computed emission rates and other relevant data to the display device.
Preferably, the data collection buffer is configured to connect to, and
receive
data from, a second gas analyzer and two exhaust temperature sensors, so that
the
apparatus may also be used to test the effectiveness of in-line catalyst
elements
(catalytic converter) for treating exhaust. In such use, the data collection
buffer is
connected to an upstream gas analyzer that draws exhaust gas from upstream of
the
catalytic converter, an upstream temperature sensor that senses the exhaust
temperature upstream of the catalytic converter; a downstream gas analyzer
that
draws in exhaust gas from downstream of the catalytic converter; and a
downstream
temperature sensor that senses the exhaust temperature downstream of the
catalytic
converter. The data collection buffer digitizes this data, organizes it into
batches that
can be recognized by the programmed computer, and sends the batches to the
programmed computer. The programmed computer calculates the differences
between the upstream and downstream levels of the relevant test gas or gases;
calculates the difference between the upstream temperature and the temperature
necessary to stimulate the desired chemical reaction between the catalyst and
the
exhaust gas; calculates the difference between the upstream and downstream
temperature of the exhaust gas (an indicator of the extent to which the
desired
catalyzed reaction, which is exothermic, is occurring); and sends the
calculated results
to the display device in real time.
-lo-
CA 02363378 2001-11-20
Preferably, the programmed computer is programmed to calculate, in real time,
and send to the display device, brake-specific fuel consumption (BSFC) for use
as a
guide to engine tuning. The BSFC, typically expressed as rate of fuel
consumption per
engine load, is an indication of engine efficiency, with a lower BSFC
indicating greater
efficiency than a higher BSFC. In use for tuning an engine, the programmed
computer
calculates BSFC from the measurements of intake manifold pressure and
temperature,
and fuel-gas flow, pressure and temperature, and sends the digitized result to
the
display device or other suitable recording device.
The various features of novelty that characterize the invention are pointed
out
with more particularity in the appended claims. 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.
-11-
CA 02363378 2001-11-20
SUMMARY OF THE DRAWINGS
Figure 1 is a schematic block diagram of the engine emission analyzer
connected to an engine, in accordance with the principles of the present
invention.
Figure 2 is a schematic block diagram of the data collection buffer of Figure
1.
Figure 3 is a flow chart of steps performed by the data collection buffer of
Figure 2 in receiving, organizing and sending data.
Figure 4 is a schematic block diagram of the engine emission analyzer of Fig.
1 configured for evaluating the effectiveness of a catalytic converter.
Figure 5 is a schematic block diagram of the engine emission analyzer of Fig.
1 configured for determining the brake specific fuel consumption of the engine
being
tested, for the purpose of optimizing engine performance.
Figure 6 is a flowchart showing initial steps of a preferred implementation of
a
method according to the invention, preferably using WinStackT~~ software that
is
suitable for use in performing computations according to the method of the
invention.
Figure 7 is a flowchart showing steps of the Emission Source/ Compliance test
mode of use of the WinStackT"~ software, in accordance with a preferred
implementation of a method according to the invention.
Figure 8 is a flowchart showing fuel measurement and fuel factor calculation
steps of the Source test mode of use of the WinStackT"" software, in
accordance with
a preferred implementation of a method according to the invention.
-12-
CA 02363378 2001-11-20
Figure 9 is a flowchart showing fuel measurement and fuel factor calculation
steps of the Compliance test mode of the use of the WinStackTM software, in
accordance with a preferred implementation of a method according to the
invention.
Figure 10 is a flowchart showing engine load approximation steps of the use of
the WnStackT"~ software, in accordance with a preferred implementation of a
method
according to the invention.
Figure 11 is a flowchart showing calibration error test steps of the use of
the
WinStackT~" software, in accordance with a preferred implementation of a
method
according to the invention.
Figure 12 is a flowchart showing Catalyst Efficiency mode steps of the use of
the WinStackT~~ software, in accordance with a preferred implementation of a
method
according to the invention.
Figure 13 is a flowchart showing Engine Optimization mode steps of the use of
the WinStackT~~ software, in accordance with a preferred implementation of a
method
according to the invention.
Figure 14 is a generalized flowchart showing the steps of an emission test (as
opposed to the preliminary inputting of data and calibration of sensors)
performed with
the use of a preferred method according to the invention.
-13-
CA 02363378 2001-11-20
DESCRIPTION OF A SPECIFIC EMBODIMENT
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 fuel gas line (30) provides fuel
gas to the
engine (20). The working parts of the engine (20) are of no interest for the
purposes
of the description of the invention and are omitted from the drawing in the
interest of
simplicity.
The engine emission analyzer (22), includes a data collection buffer (52); a
programmed computer (54) incorporating a display device (typically a video
monitor)
(not shown) and a computer readable database (not shown). The engine emission
analyzer (22) is connected to a printer (56); 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); and an intake manifold temperature sensor (50).
The
meter 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 referred to as cells, for measuring the concentrations of CO, NO,
N02
and Oz. These sensors are typically capable of measuring concentrations of the
-14-
CA 02363378 2001-11-20
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 (e.g., 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 method 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 any suitable selected 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 numeric value
provided by the
flowmeter manufacturer, referred to as the "K-factor", is used as a conversion
factor
to convert the frequency of the pulsed output to a volumetric flow rate.
The fuel gas pressure sensor (44) is any suitable selected 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 flowmeter (42). Typically, the fuel gas pressure
sensor (44)
-15-
CA 02363378 2001-11-20
emits a fluctuating current output (typically in the range 4-20 mA) with the
current
being proportional to the fuel gas pressure.
The fuel gas temperature sensor (46) is any suitable selected commercially
available thermocouple, such as the type "J"TM 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
(74) is typically connected to the fuel gas line with flexible high-pressure,
TefIonT""-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). Once the fuel gas bypass valves (76), fuel gas block valve (78)
and
associated T junctions (tees) are installed in the fuel gas line (30), the
meter run (74)
may be installed and removed without stopping the engine (20). The meter
run(74)
is generally installed downstream of conventional fuel gas scrubbers (not
shown) so
that the fuel gas is dry when the temperature, pressure and flow measurements
are
made.
The intake manifold pressure sensor (48) is any suitable selected 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.
-16-
CA 02363378 2001-11-20
The intake manifold temperature sensor (50) is preferably any suitable
selected
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.
As 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).
An example of a suitable microprocessor is model number MC68HC16Z1
manufactured by Motorola. For ease of understanding, the pulse counter (97)
and
analog to digital converter (98) are shown as components separate from the
microprocessor in Figure 2. However, one pulse counter input and eight analog-
to-
digital converter inputs are integral parts of the Motorola MC68HC16Z1
microprocessor.
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 readily commercially available. An
example of a suitable Flash EPROM is model number AT49F001 manufactured by
Atmel. It will be clear to those skilled in the computer art that other memory
storage
devices could be used.
-m-
CA 02363378 2001-11-20
The Static RAM is a form of random access memory device. Such Static
RAMs are well known in the art and are commercially available. An example of a
suitable static Ram is model number AS7C1026 manufactured by Alliance
Semiconductors. It will be clear to those skilled in the art that other random
access
memory devices could be used instead.
The UART (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. An example of a
suitable
UART is model number TL16C5541 FN manufactured by Texas Instruments.
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. An example of a suitable thermocouple amplifier is the "Monolithic
Thermocouple Amplifier with Cold Junction Compensation", model number AD594
version 'C', manufactured by Analog Devices Inc.
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
readily
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
-ls-
CA 02363378 2001-11-20
the voltage pulses into square waves, which conversion 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 readily 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
readily 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 data 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
-19-
CA 02363378 2001-11-20
which is the same for that particular field in the other data strings. The
data from each
particular sensor in the gas analyzer are 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 data are extracted from the data string
using the
known field length and order. In a preferred embodiment using the Ecom SG PLUS
TM 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
data representing the measurements from the fuel gas flowmeter (42); fuel gas
pressure sensor (44); fuel gas temperature sensor (46); intake manifold
pressure
-20-
CA 02363378 2001-11-20
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 data and sends the data to 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 data from the data
collection buffer (52), including data representing the following:
measurements of the
concentrations of NO, N02 CO, C02 and 02; 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) has associated with it a computer-readable
database (not shown). Typically, the database is a component of the computing
device incorporating the programmed computer, but the database may also be a
peripheral device connected to the programmed computer; a remote database
accessed via a long-distance communication means such as the Internet; or a
combination of these, in that different parts of the data used by the
programmed
computer in determining emission rates may be in different databases. For
simplicity,
-21-
CA 02363378 2001-11-20
this description refers to one database though it is understood that more than
one
database may be used.
The following data are stored in the database so as to be accessible to the
programmed computer (54) during a test:
a) the specifications of the engine to be tested, including typically: the
model
number, compression ratio, timing, carburetor setting (e.g. lean), engine
displacement, intercooler water temperature (if appropriate), rated BHP, and
rated BSFC;
b) the maximum emission limits for each of the pollutants specified in the
relevant
permit, typically in gms/BHP hour, tons/yr and Ibs/hr;
c) the selected testing parameters, typically the time interval between
readings,
the number of readings and the emission rate units;
d) various calculation factors, including the ambient pressure, ambient
temperature, the K-factor for the fuel gas flowmeter (42), the molar weight of
the fuel gas, and the dry fuel F factor, gross calorific value, critical
temperature
and critical pressure of the fuel gas.
The engine specifications are typically obtained from the engine manufacturer
and the
engine operator, and typically are manually input by the person conducting the
test.
The emission limits are typically manually input when the engine is first
tested and
thereafter the related data are stored in a database record associated with
the relevant
engine and permit. The testing parameters are either selected by the user and
manually input, or are set by a testing protocol associated with the relevant
permit, in
which case they may be stored in a database record associated with the
protocol or
permit.
-22-
CA 02363378 2001-11-20
As set out above, the ambient pressure and ambient temperature are typically
manually input into the programmed computer (54) which stores the ambient
pressure
and ambient temperature in the database. The K-factor is either provided by
the
manufacturer fuel gas flowmeter (42) based on the manufacturer's initial
calibration of
the fuel gas flowmeter (42) or is manually input by the person performing the
test to
reflect a subsequent recalibration of the fuel gas flowmeter (42).
The molar weight, dry fuel F factor, gross calorific value, critical
temperature
and critical pressure of the fuel gas are calculated from an ultimate analysis
of the fuel
gas, typically obtained from the engine operator. Typically, the calculations
are
performed by the programmed computer (54) which then stores the results in the
database. The dry fuel F factor (Fd) is calculated (per EPA 60 CFR 40, Method
19,
Eqn. 19-13) as follows:
Fd = I OE6((3.64 * %H ) + ( 1,$3 * %C) + (Q,$'7 * %S) + (0,14 * %N) - (0.46 *
%O)]/GCVH, 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.
The gross calorific value (GCVw) of the fuel gas is calculated (per GPSA
Standard
2172-72) as follows:
E xnHn
Eqn. 1
GCVw = _______________ _____ __
1-(Exn~bn)2
Where: xn = Mole fraction of each component
Hn = Gas Heating Value of each component.
bn= Summation Factor of each component
The critical temperature T~ and critical pressure P~ of the fuel gas are
calculated as
follows:
-23-
CA 02363378 2001-11-20
pC = ~ PCn * xn TC - ~ TCn * xn E~qn 2
Where: Pcn = Critical Pressure of each component.
Tcn = Critical Temperature of each component.
xn= Mole fraction of each component.
The programmed computer (54) typically also performs pre-test and post-test
calculations related to the pre-test and post-test calibration of the sensors,
to
determine interference response and correction for sensor drift for the
sensors. These
calculations include:
1. CO Interference Response calculated (per CTM030 Method, Section
6.3.1 ) as follows:
Ico URco-no ~ Cnog * Cnos ~ Ccos) + \Rco-not ~ Cno2g * Cno2s ~ Ccos~~ * 1 ~~
Eqn 8
Where: Ico = 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).
2. NO Interference Response calculated (per CTM030 Method, Section
6.3.2) as follows:
Ino - (Rno-not ~ Cno2g ) * ( Cno2s ~ Cnoxs) * 1 ~~ Eqn 9
where: Ino = NO interference response (%).
Rno-not- NO response to N02 span gas (ppm NO).
Cno29= concentration of N02 span gas (ppm N02).
-24-
CA 02363378 2001-11-20
Cno2s= concentration of N02 in stack gas (ppm N02).
Cnoxs- concentration of NOx in stack gas (ppm NOx).
3. Concentration Correction For Sensor Drift calculated (per CTM030 Method,
Section 8.1 ) as follows:
CGAS - (CR - CO ) * CMA ~ (CM - CO ) Eqn 10
Where: CGAS = corrected flue gas concentration (ppm).
CR = flue gas concentration indicated by gas analyzer (ppm ).
Cp = average of initial and final zero checks (ppm ).
CM = average of initial and final span checks (ppm).
CMA = actual concentration of span gas (ppm).
As shown in Figure 14, when the actual emission test (as opposed to the
preliminary inputting of data and calibration of the sensors) commences, the
programmed computer (54) performs the step of INITIALIZE VARIABLES AND
COUNTER, which involves initializing (or clearing) the relevant variables and
setting
a counter (referred to as RECORD) to 0 (350). The programmed computer (54)
then
performs the step of RETRIEVE INFORMATION FROM DATABASE (352), which
entails retrieving the engine specifications, emission limits, testing
parameters
(including the number of readings, referred to as NUM-READ) and calculation
factors
from the database).
The programmed computer (54) then starts the TIMER (354) and instructs the
TIMER (354) to commence execution of the steps starting with INCREMENT
RECORD (356). The TIMER will thereafter periodically commence execution of the
steps starting with INCREMENT RECORD (356) at the predetermined time interval
between readings (a setting previously obtained from the database) until the
TIMER
receives instructions to halt.
-25-
CA 02363378 2001-11-20
The step of INCREMENT RECORD (356) entails adding 1 to the RECORD
counter.
The programmed computer (54) then performs the step of COLLECT RAW
DATA (358), which entails collecting from the data collection buffer (52): the
gas
analyzer data (typically the sensed relative concentrations of O2, CO, NO, NOZ
and
C02, and the NO cell temperature); the intake manifold temperature and
pressure; the
fuel gas temperature and pressure; and the frequency of pulse counts from the
fuel
gas flowmeter (42).
The programmed computer then performs the step of CALCULATE FUEL GAS
COMPRESSIBILITY (360) which entails calculating the Fuel Gas Compressibility
Factor (per GPSA Method, Section 16, Using Standing & Kantz Compressibility
Curves). The sensed pressure (Pay) is divided by the critical pressure of the
fuel gas
(P~) (previously calculated and obtained from the database) to determine the
reduced
(corrected) pressure of fuel gas Pr. The sensed temperature of the fuel gas
(Tact) is
divided by the critical temperature of the fuel gas (T~) ) (previously
calculated and
obtained from the database) to determine the reduced (corrected) temperature
of fuel
gas (Tr).
Pr Pact Tr Tact Eqn 3
PC ___ _TC _
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.
The reduced pressure and temperature are correlated with the Standing & Kantz
compressibility curves using an internal software subroutine to produce the
actual
compressibility factor.
-26-
CA 02363378 2001-11-20
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 programmed computer (54) then performs the step of CALCULATE FUEL
FLOW (362), which entails calculating the dry volumetric flow rate of the fuel
gas
corrected for standard conditions and compressibility (DCSFMf~ey as follows:
DSCFM fuel= 60 * f * [(528/(460 + Tfuel O * (Pamb + Pfuel ~ / 29.92] Eqn 5
z*K
Where: f = Frequency Of Pulses Generated By Turbine Meter.
Tfuel = Sensed Temperature Of Fuel Gas.
Pamb = Ambient Pressure.
Pfuel = Sensed Pressure Of Fuel Gas.
z = Compressibility Factor Of Fuel Gas.
K = Pulse Conversion Factor Provided By Turbine Meter Manufacturer
The programmed computer (54) then performs the step of CALCULATE EXHAUST
FLOW (364), which entails calculating the dry effluent volumetric flow rate
(per an
extension of EPA 40 CFR 60 Method 19) as follows:
Qsd= Fd * HIR * 20.9 Eqn 6
20.9 - % OZ
Where: Qsd = Dry Effluent Volumetric Flowrate.
Fd = Dry Fuel F Factor.
HIR = Heat Input Rate (Fuel Heat Content (GCVw) * Fuel Usage Rate (DSCFMfuel))
-27-
CA 02363378 2001-11-20
02 = Oxygen Content Of Effluent Gas (Used For Excess Air Correction).
The programmed computer (54) then (or possibly before any or all of steps
360, 362 and 364, or, alternatively, concurrently with any or all of steps
360, 362
and 364) performs the step of CALCULATE ENGINE HORESPOWER (366), by
one of three possible alternative methods, previously selected by the person
performing the test, as follows:
1. By correlating engine data , typically the values of intake manifold
pressure,
intake manifold temperature, engine model and engine RPM with the engine
manufacturer's load curve. Discrete values not found on the engine
manufacturer's load curve are calculated using Newton's Method of
Interpolation. The data from each load curve are incorporated into a computer
routine (referred to as a function) along with Newton's Method of
Interpolation,
such that the programmed computer may "call the function", that is, provide a
particular routine with values for the required variables and instruct the
routine
to calculate the engine load.
2. By using the engine manufacturer's brake specific fuel consumption (BSFC)
(previously obtained from the database) as follows:
BHP = HIR / BSFC Eqn 1 ~
Where: HIR = Heat Input Rate (BTU / hr) (Fuel Heat Content ~CrCVw) * Fuel
Usage Rate
(DSCFMfuel))
BsFC = Brake Specific Fuel Consumption Published By Engine Manufacturer (BTU /
BHP-hr).
3. By correlating the engine RPM with the engine manufacturer's
specifications.
If this method is selected, then, since the RPM does not vary during the test,
the engine load determined by this method remains constant throughout the
-28-
CA 02363378 2001-11-20
test and need not be determined for each iteration of the test steps. The
engine load is merely retrieved from the database, held in the programmed
computer's (54) local memory and assigned as the result of the
CALCULATE ENGINE HORESPOWER (366) step.
The programmed computer (54) then performs the step of CALCULATE
EMISSION RATE (368) which entails calculating the emission rate (per CARB 100
Method) of each of the relevant pollutants, as follows:
ERP = 1.56E-7 * PPM * Qsd * MW sqn ~
Where: ERP = Emission Rate Of Pollutant (lbs/hr).
PPM = Concentration Of Pollutant.
Qsd = Dry Effluent Volumetric Flowrate.
MW = Molar Weight Of Pollutant.
The CALCULATE EMISSION RATE (368) step yields units of Ibs/hr. The molar
weight
of each pollutant is a constant. The molar weight of each pollutant may be
stored in
the database or, preferably, may be incorporated in the program of the
programmed
computer.
The programmed computer (54) then performs the step of CHECK EMISSION
RATE UNIT (370), which entails: checking if the preferred emission rate unit
(either
previously obtained from the database or entered by the user during the test)
is Ibs/hr;
and, if the preferred unit is not Ibs/hr, converting the emission rate to the
preferred unit
or dividing the emission rate by the engine load to obtain gr/BHP-hr or gr/KW-
hr, if
such is preferred.
The programmed computer (54) then performs the step of POPULATE TAB
PAGES (372), which entails adding the raw and calculated data to various tab
pages
viewable on a display screen (not shown) associated with the programmed
computer
-29-
CA 02363378 2001-11-20
(54) to enable the user to view the data in real time. Alternatively, or
concurrently, the
programmed computer could send the data to a printer so as to continuously
print the
data during the test.
The programmed computer (54) then performs the step of COMPARE TO
PERMIT LEVELS (374), which entails comparing the calculated emission rate of
each
pollutant with the maximum permitted emission rate (previously obtained from
the
database), and, if the calculated emission rate for a pollutant is above the
maximum
permitted emission rate, notifying the user of this, preferably by an
indication on the
display screen or alternatively by generating an audible alarm. Preferably the
programmed computer (54) also notifies the user, in a similar though
distinguishable
manner, when a calculated emission rate is 90% or more of the maximum
permitted
emission rate.
The programmed computer (54) then performs the step of COMPARE RECORD
TO NUM_READ (376), which entails comparing the RECORD counter to the preferred
number of readings (NUM_READ) (previously obtained from the database). If
RECORD is less than NUM_READ then the TIMER is permitted to continue
commencing execution of the steps commencing with INCREMENT RECORD (356).
If RECORD is equal to NUM_READ then the TIMER is halted.
The programmed computer (54) then performs the step of SAVE RAW DATA
(378), which entails saving the raw data to the database for subsequent
preparation
of a formal emission report.
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
-30-
CA 02363378 2001-11-20
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 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
-31-
CA 02363378 2001-11-20
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 data and sends the data to 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 data 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).
-32-
CA 02363378 2001-11-20
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. WinStackTM is a WindowsT"~ based system containing
executable
files and incorporating a 32 bit Sybase database platform. WinStackT~~ will
run on
WindowsT~~ 32 bit platforms, such as Windows 95TM & Windows 98TM, but
WinStackT~~
hasn't been certified on Windows 2000T~~ or Windows NTT~~. WinStackTM requires
a
computer having at least a PentiumTM 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 that the sensor is designed to detect, that is, 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.
Conventionally, the person conducting the test obtains the concentrations of
the target
gases with un-calibrated sensors and uses the un-calibrated sensed
concentrations
of the target gases as guides to the selection of appropriate span gases.
Typically,
the person testing the exhaust has several span gas bottles with different
concentrations of target gases to choose from.
-33-
CA 02363378 2001-11-20
The system diagnostic checks (150) performed by WinStackT"~ include checking
the measurements being received from the CO, NO, N02 and OZ 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 Source/Compliance
test mode (154), the user is prompted to enter the relevant permit information
data, or
select the permit information data from the WinStackT~" database if the permit
information data have been previously entered into the computer (160). The
permit
information data stored by the WinStackT"" database include the state or
province in
or for which the permit has been issued, the date upon which the most recent
previous
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. WinStackT~~ 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.
-34-
CA 02363378 2001-11-20
WinStackT~" then prompts the user to enter, into the appropriate fields, the
following data : 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). WinStackT~~ 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. WinStackTM 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
-35-
CA 02363378 2001-11-20
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.
WinStackTM
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). WinStackTM 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), WinStackT~~ estimates a generic value for the
gross calorific
value of the fuel (1000 Btu/cf). If the user elects to calculate the fuel F
factor based
on the fuel gas composition, then the user enters the fuel gas composition
data (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). ~nStackT~~ 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
-36-
CA 02363378 2001-11-20
specific fuel consumption (BSFC, typically in BTU/BHP-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, WinStackT"~ 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 ). 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. The user may tune the engine (212).
WinStackTM 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 Btu/hr) and the manufacturer's published
BSFC values
(typically Btu/BHP-hr) (220). The user may tune the engine (222). WinStackTM
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).
-37-
CA 02363378 2001-11-20
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, WinStackTM waits for the
chosen
delay period and then selects and retrieves a rated load from the
manufacturer's
engine load in the database, corresponding to the entered RPM (230).
WinStackT""
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 confirms via the computer program interface that the entries are
satisfactory, WinStackT~~ goes into the data acquisition mode.
~nStackT"" 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, the user instructs WinStackTM to commence the pre-test calibration
error
phase (246). WinStackTM 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.
-38-
CA 02363378 2001-11-20
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, WinStackTM 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, WinStackT~~ 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
-39-
CA 02363378 2001-11-20
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. WnStackT"~
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, ~nStackT~~ 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 WinStackTM.
Depending on the load approximation method previously selected by the user
[28], WinStackT~~ approximates the engine load using: pressure and temperature
-40-
CA 02363378 2001-11-20
measurements (engine manifold method); corrected fuel flow measurements
(engine
BSFC method); or the manufacturer's rated load. WinStackTM 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. Persons skilled
in
the art of petroleum engineering will be familiar with the Standing and Kantz
method.
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.
WinStackTM 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.
-41-
CA 02363378 2001-11-20
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,
WinStackTM prompts the user to enter, into the appropriate fields, the
following data
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.
WinStackT"~ 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). WnStackT~~ 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 data ; checks for blank
fields
or erroneous inputs and notifies the user if there is an error in the input
data; and
permits the user to adjust or correct any values through the 'Preferences'
section of
the menu (286). The test data and ambient condition data are saved to the
WinStackT~~
database for later retrieval and report generation. Once the user indicates
that the user
is satisfied with the entries, WinStackT~" goes into the data acquisition
mode.
-42-
CA 02363378 2001-11-20
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, WinStackTM 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. WinStackTM compares the
upstream temperature to the temperature necessary to stimulate the desired
chemical
reaction between the catalyst element and the exhaust gas. WinStackTM 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). WinStackTM 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
data: the facility location, the relevant environmental board; the gas
analyzer serial
-43-
CA 02363378 2001-11-20
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.
WinStackTM 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). WinStackTM 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. WinStackTM 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). WinStackTM then displays the previously
entered
-44-
CA 02363378 2001-11-20
data (324), and permits the user to adjust or correct any values through the
'Preferences' section of the menu. Once the user indicates that the user is
satisfied
with the entries, all data calculated and recorded during this step is saved
to the
WinStackTM database for later retrieval and report generation; and WinStackTM
goes
into the data acquisition mode.
WnStackT"~ 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.
WinStackT~~ then prompts the user to select either line or bar graph format
for
real-time graphical display purposes. When the user instructs WinStackTM 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).
WnStackT~~ 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
-45-
CA 02363378 2001-11-20
any of the specific features as described, but comprehends all such variations
thereof
as come within the scope of the appended claims.
-46-
