U.S. patent application number 11/789410 was filed with the patent office on 2008-10-30 for advanced engine control.
Invention is credited to Nicholas J. Bellovary.
Application Number | 20080264036 11/789410 |
Document ID | / |
Family ID | 39885374 |
Filed Date | 2008-10-30 |
United States Patent
Application |
20080264036 |
Kind Code |
A1 |
Bellovary; Nicholas J. |
October 30, 2008 |
Advanced engine control
Abstract
A method for controlling an internal combustion engine to
maintain a predetermined combustion efficiency, comprises:
measuring the concentration of CO in an engine-out emission;
determining whether the measured concentration of CO is above or
below a target CO concentration for the engine under the operating
conditions in use; and if the concentration of CO is above the
target CO concentration, adjusting one or more engine and/or
combustion parameters so as to reduce the concentration of CO in
the engine-out emission, and if the concentration of CO is below
the target CO concentration, adjusting one or more engine and/or
combustion parameters so as to increase the concentration of CO in
the engine-out emission; wherein the target CO concentration is
determined on the basis of a correlation with indicators of
combustion efficiency under known engine operating conditions.
Exhaust gas recirculation (EGR) is one such engine and/or
combustion parameter. A method for controlling an internal
combustion engine to obtain an engine-out emission level of NOx
within a predetermined concentration range is also described. The
methods may be implemented by an engine control unit.
Inventors: |
Bellovary; Nicholas J.;
(Brighton, MI) |
Correspondence
Address: |
DELPHI TECHNOLOGIES, INC.
M/C 480-410-202, PO BOX 5052
TROY
MI
48007
US
|
Family ID: |
39885374 |
Appl. No.: |
11/789410 |
Filed: |
April 24, 2007 |
Current U.S.
Class: |
60/274 ; 123/679;
60/276; 60/278 |
Current CPC
Class: |
Y02A 50/2325 20180101;
F01N 3/00 20130101; Y02T 10/12 20130101; Y02T 10/20 20130101; F01N
2560/022 20130101 |
Class at
Publication: |
60/274 ; 123/679;
60/276; 60/278 |
International
Class: |
F01N 3/00 20060101
F01N003/00 |
Claims
1. A method for controlling an internal combustion engine to
maintain a predetermined combustion efficiency, the method
comprising: measuring the concentration of CO in an engine-out
emission; determining whether the measured concentration of CO is
above or below a target CO concentration for the engine under the
operating conditions in use; and if the concentration of CO is
above the target CO concentration, adjusting one or more engine
and/or combustion parameters so as to reduce the concentration of
CO in the engine-out emission, and if the concentration of CO is
below the target CO concentration, adjusting one or more engine
and/or combustion parameters so as to increase the concentration of
CO in the engine-out emission; wherein the target CO concentration
is determined on the basis of a correlation with indicators of
combustion efficiency under known engine operating conditions.
2. The method of claim 1, wherein said one or more engine and/or
combustion parameters is EGR.
3. The method of claim 1, wherein the known engine operating
conditions include engine speed and load.
4. The method of claim 1, wherein the predetermined combustion
efficiency is selected during engine development and/or
testing.
5. The method of claim 1, wherein the target CO concentration is
determined on the basis of a correlation with one or more engine
and/or combustion parameters which are indicators of combustion
efficiency; and wherein the correlation is by a means of data
comparison, the data relating to measured concentrations of CO and
the one or more engine and/or combustion parameters.
6. The method of claim 1, wherein the target CO concentration is
predetermined on the basis of a correlation with one or more engine
and/or combustion parameters; wherein the correlation is by a means
of data comparison, the data relating to measured concentrations of
CO and the one or more engine and/or combustion parameters; and
wherein the means of data comparison is a look-up table,
calibration curve or equation obtained for the engine under known
engine operating conditions.
7. The method of claim 1, wherein the correlation with indicators
of combustion efficiency under known engine operating conditions is
an approximation of the correlation existing under the operating
conditions in use.
8. The method of claim 1, wherein the correlation with indicators
of combustion efficiency under known engine operating conditions is
an estimated correlation for the operating conditions in use, the
estimated correlation being obtained by adjusting the correlation
obtained under the known engine operating conditions by means of an
appropriate regression analysis, equation or algorithm.
9. The method of claim 1, wherein the target CO concentration is
predetermined on the basis of a correlation between CO
concentration and hydrocarbon (HC) concentration under known engine
operating conditions, the target CO concentration being that which
corresponds to a predetermined HC concentration at which a
predetermined acceptable level of combustion stability is
achieved.
10. The method of claim 1, wherein the target CO concentration is
predetermined on the basis of a correlation between CO
concentration and hydrocarbon (HC) concentration under known engine
operating conditions, said correlation being an approximation of
the correlation existing under the operating conditions in use; and
wherein the target CO concentration is that which corresponds to a
predetermined HC concentration.
11. The method of claim 1, wherein the target CO concentration is
predetermined on the basis of a correlation between CO
concentration and hydrocarbon (HC) concentration under known engine
operating conditions, the correlation being an estimated
correlation for the operating conditions in use, the estimated
correlation being obtained by adjusting the correlation obtained
under the known engine operating conditions by means of a
regression analysis, equation or algorithm; and wherein the target
CO concentration is that which corresponds to a predetermined HC
concentration.
12. The method of claim 1, wherein the target CO concentration is
predetermined on the basis of a correlation between CO
concentration and hydrocarbon (HC) concentration under known engine
operating conditions, the target CO concentration being that which
corresponds to an HC concentration of approximately 400 ppm.
13. An engine control unit for a motor vehicle that, in use,
performs the method of claim 1.
14. A motor vehicle comprising an engine control unit that, in use,
performs the method of claim 1.
15. A method for controlling an internal combustion engine to
obtain an engine-out emission level of NOx within a predetermined
concentration range, the method comprising: measuring the
concentration of CO in an engine-out emission; calculating the
concentration of NOx in said engine-out emission based on a
correlation with the measured concentration of CO; determining
whether the calculated concentration of NOx for the engine under
the operating conditions in use is within the predetermined
concentration range; and where the calculated concentration of NOx
is outside of the predetermined concentration range, adjusting one
or more engine and/or combustion parameters so as to achieve an
engine-out emission level of NOx within the predetermined
concentration range.
16. The method of claim 15, wherein said one or more engine and/or
combustion parameters is EGR.
17. The method of claim 15, wherein said one or more engine and/or
combustion parameters is EGR, and wherein EGR is increased when the
calculated concentration of NOx is higher than the predetermined
concentration range.
18. The method of claim 15, wherein said one or more engine and/or
combustion parameters is fuel injection timing.
19. The method of claim 15, wherein said one or more engine and/or
combustion parameters is injected fuel quantity.
20. The method of claim 15, wherein an engine control unit is used
to: calculate the concentration of NOx on the basis of the measured
concentration of CO; and adjust the one or more engine and/or
combustion parameters using a feedback signal.
21. The method of claim 15, wherein the correlation is by a means
of data comparison, the data relating to measured concentrations of
CO and NOx in engine-out emissions under known engine and/or
combustion parameters.
22. The method of claim 15, wherein the correlation is by a means
of data comparison, the data relating to measured concentrations of
CO and NOx in engine-out emissions under known engine and/or
combustion parameters, and wherein the means of data comparison is
a look-up table, calibration curve or equation.
23. The method of claim 15, wherein the correlation is by a means
of data comparison, the data relating to measured concentrations of
CO and NOx in engine-out emissions under known engine and/or
combustion parameters, and wherein the correlation is an
approximation of the correlation existing under the operating
conditions in use.
24. The method of claim 15, wherein the correlation is by a means
of data comparison, the data relating to measured concentrations of
CO and NOx in engine-out emissions under known engine and/or
combustion parameters, and wherein the correlation is an estimated
correlation for the operating conditions in use, the estimated
correlation being obtained by adjusting the correlation obtained
under the known engine and/or combustion parameters by means of an
appropriate regression analysis, equation or algorithm.
25. An engine control unit for a motor vehicle that, in use,
performs the method of claim 15.
26. A motor vehicle comprising an engine control unit that, in use,
performs the method of claim 15.
27. A method for controlling an internal combustion engine to
obtain an engine-out emission level of NOx within a predetermined
concentration range, the method comprising: measuring the
concentration of CO in an engine-out emission; calculating the
concentration of NOx in said engine-out emission based on a
correlation with the measured concentration of CO; determining
whether the calculated concentration of NOx for the engine under
the operating conditions in use is above a predetermined
concentration level; and where the calculated concentration of NOx
is above the predetermined concentration level, operating one or
more after-treatments to reduce the concentration of NOx in said
engine-out emission to a level no greater than said predetermined
concentration level.
28. The method of claim 27, wherein an engine control unit is used
to: calculate the difference between the calculated concentration
of NOx and the predetermined concentration level; and adjust the
one or more after-treatments to reduce the concentration of NOx in
said engine-out emission to a level no greater than the
predetermined concentration level, and wherein the adjustment to
the one or more after-treatments is dependent on the difference
between the calculated concentration of NOx and the predetermined
concentration level.
29. The method of claim 27, wherein the one or more
after-treatments is selected from the group consisting of: a diesel
particulate filter, a catalytic converter, selective catalytic
reduction and an NOx trap.
30. The method of claim 27, wherein the correlation is by a means
of data comparison, the data relating to measured concentrations of
CO and NOx in engine-out emissions under known engine and/or
combustion parameters.
31. The method of claim 27, wherein the correlation is by a means
of data comparison, the data relating to measured concentrations of
CO and NOx in engine-out emissions under known engine and/or
combustion parameters, and wherein the means of data comparison is
a look-up table, calibration curve or equation.
32. The method of claim 27, wherein the correlation is by a means
of data comparison, the data relating to measured concentrations of
CO and NOx in engine-out emissions under known engine and/or
combustion parameters, and wherein the correlation is an
approximation of the correlation existing under the operating
conditions in use.
33. The method of claim 27, wherein the correlation is by a means
of data comparison, the data relating to measured concentrations of
CO and NOx in engine-out emissions under known engine and/or
combustion parameters, and wherein the correlation is an estimated
correlation for the operating conditions in use, the estimated
correlation being obtained by adjusting the correlation obtained
under the known engine and/or combustion parameters by means of an
appropriate regression analysis, equation or algorithm.
34. An engine control unit for a motor vehicle that, in use,
performs the method of claim 27.
35. A motor vehicle comprising an engine control unit that, in use,
performs the method of claim 27.
36. A method for controlling an internal combustion engine to
obtain an engine-out emission level of NOx within a predetermined
concentration range, the method comprising: measuring the
concentration of CO in an engine-out emission; calculating the
concentration of NOx in said engine-out emission based on a
correlation with the measured concentration of CO; determining
whether the calculated concentration of NOx for the engine under
the operating conditions in use is within the predetermined
concentration range; and where the calculated concentration of NOx
is outside of the predetermined concentration range, adjusting one
or more engine and/or combustion parameters so as to achieve an
engine-out emission level of NOx within the predetermined
concentration range; and operating one or more after-treatments to
reduce the concentration of NOx in said engine-out emission.
37. The method of claim 36, which comprises the steps of:
determining whether the calculated concentration of NOx is above a
predetermined concentration level; and where the calculated
concentration of NOx is above the predetermined concentration
level, operating the one or more after-treatments to reduce the
concentration of NOx in the engine-out emission to a level no
greater than the predetermined concentration level.
38. An engine control unit for a motor vehicle that, in use,
performs the method of claim 36.
39. A motor vehicle comprising an engine control unit that, in use,
performs the method of claim 36.
40. A method for calculating the concentration of NOx in an
engine-out emission from an engine, the method comprising:
measuring the concentration of CO in said engine-out emission; and
calculating the concentration of NOx in said engine-out emission
based on a correlation with the measured concentration of CO under
the engine operating conditions in use; wherein the correlation is
by a means of data comparison, the data relating to measured
concentrations of CO and NOx in engine-out emissions under known
engine and/or combustion parameters.
41. The method of claim 40, wherein the correlation is an
approximation of the correlation existing under the operating
conditions in use.
42. The method of claim 40, wherein the correlation is an estimated
correlation for the operating conditions in use, the estimated
correlation being obtained by adjusting the correlation obtained
under the known engine and/or combustion parameters by means of an
appropriate regression analysis, equation or algorithm.
43. The method of claim 40, wherein the means of data comparison is
a look-up table, calibration curve or equation.
44. The method of claim 40, wherein the correlation is performed by
an engine control unit.
45. An engine control unit for a motor vehicle that, in use,
performs the method of claim 40.
46. A motor vehicle comprising an engine control unit that, in use,
performs the method of claim 40.
47. A method for calculating the engine operating Air-to-Fuel Ratio
(AFR) in an engine, the method comprising: measuring the
concentration of CO in the engine-out emission from said engine;
and calculating the AFR of said engine based on a correlation with
the measured concentration of CO under the operating conditions in
use; wherein said correlation is by a means of data comparison,
said data relating to measured AFR levels and concentrations of CO
in engine-out emissions under known engine and/or combustion
parameters.
48. The method of claim 47, wherein the correlation is an
approximation of the correlation existing under the operating
conditions in use.
49. The method of claim 47, wherein the correlation is an estimated
correlation for the operating conditions in use, the estimated
correlation being obtained by adjusting the correlation obtained
under the known engine and/or combustion parameters by means of an
appropriate regression analysis, equation or algorithm.
50. The method of claim 47, wherein the means of data comparison is
a look-up table, calibration curve or equation.
51. The method of claim 47, wherein the correlation is performed by
an engine control unit.
52. An engine control unit for a motor vehicle that, in use,
performs the method of claim 47.
53. A motor vehicle comprising an engine control unit that, in use,
performs the method of claim 47.
54. Use of a CO sensor in a system for controlling or adjusting one
or more engine and/or combustion parameters in an internal
combustion engine, wherein the CO sensor is arranged to measure the
concentration of CO in an engine-out emission, and the system
includes a means for correlating the measured concentration of CO
with one or more of: an indicator of combustion efficiency under
the operating conditions in use; the concentration of NOx in the
engine-out emission under the operating conditions in use; and/or
the level of AFR in the engine under the operating conditions in
use.
55. The use of claim 54, wherein the indicator of combustion
efficiency is hydrocarbon (HC) concentration in the engine-out
emission.
56. The use of claim 54, wherein the system controls or adjusts one
or more engine and/or combustion parameters in order to bring the
calculated concentration of NOx in the engine-out emission to
within a predetermined concentration range.
57. The use of claim 54, wherein the system controls or adjusts one
or more engine and/or combustion parameters in order to bring the
calculated level of AFR in the engine to within a predetermined
range.
58. The use of claim 54, wherein the indicator of combustion
efficiency is hydrocarbon (HC) concentration in the engine-out
emission, and wherein the system controls or adjusts one or more
engine and/or combustion parameters in order to bring the
combustion efficiency in the engine to within a predetermined
range.
Description
FIELD OF THE INVENTION
[0001] This invention relates to methods for controlling an
internal combustion engine to control combustion stability and/or
reduce engine-out nitrogen oxides (NOx) emissions. In particular,
the invention relates to methods for using the measurement of
carbon monoxide (CO) in the engine-out emissions to control
combustion stability and/or indicate the level of NOx in the engine
out emissions.
BACKGROUND OF THE INVENTION
[0002] Current emissions legislation in, for example, the US and
Europe, set strict limits on the acceptable levels of polluting
gases that are tolerated in the exhaust gas emissions of
compression ignition engines; and engines using lean burn (oxygen
rich) combustion, including diesel and gasoline direct injection
engines. However, future legislation is set to pose a far more
serious challenge to the manufacturers of such engines.
[0003] To overcome the challenge of reducing exhaust gas emissions,
intense research and development is necessary in the areas of
combustion, after-treatment or both.
[0004] The typical approach taken to improving the combustion
process and reducing the production of undesirable engine gas
emissions is to utilise some form of "advanced combustion".
Advanced combustion can be considered to be an umbrella term that
encompasses any type of combustion process which aims to minimise
the traditional flame burn of a combustion engine and replace it
with simultaneous ignition of well mixed air and fuel. Examples of
advanced combustion processes include Homogeneous Charge
Compression Ignition (HCCI), Low-Temperature Combustion (LTC) and
Pre-mixed Charge Compression Ignition (PCCI). Regardless of the
name given to the process, all forms of advanced combustion
generally attempt to achieve very low levels of NOx and soot
emissions through improved fuel-air mixing and reduced combustion
temperatures.
[0005] Advanced combustion typically involves one or more of the
following aspects: fuel injection much advanced of Top Dead Centre
(TDC); multiple fuel injections; high amounts of Exhaust Gas
Recirculation (EGR); and high injection pressures.
[0006] It is generally accepted that some form of advanced
combustion is necessary as a practical and cost effective means for
internal combustion engines to meet present and certainly future
emissions regulations. However, it has been found that during an
advanced combustion process, even small errors (or changes) in air
charge or combustion chamber temperature can lead to dramatic
changes in engine emissions, combustion stability and engine
noise.
[0007] Moreover, it should be appreciated that the control of
advanced combustion is not merely a matter of air charge and engine
controls, but also of fuel properties. The most significant of
these fuels properties being cetane number, which is a measure of a
fuel's ignitability. For instance, it has been shown that changes
in cetane number can cause significant variations in both
combustion and engine NOx emissions. In short, all of the available
advanced combustion processes face a number of challenges, in terms
of: combustion controls, system to system dispersion, combustion
stability, variation in fuel properties, and estimating/modelling
engine-out emissions.
[0008] Therefore, several systems under development seek to monitor
engine performance during advanced combustion, and ultimately to
reduce exhaust gas emissions. For example, engine control processes
may utilise NOx sensors, Air-Fuel Ratio (AFR) sensors, ammonia
sensors, and/or cylinder pressure sensors.
[0009] NOx sensors provide the most direct means of monitoring the
engine-out NOx emissions from an internal combustion engine. Once
the level of NOx has been determined, a feedback signal may be used
to control EGR, in an attempt to adjust the amount of NOx produced.
However, this system has several drawbacks. For instance, the
concentration of NOx is typically very low during advanced
combustion (e.g. approximately 4 ppm), so any NOx sensor must be
extremely sensitive and accurate in order to provide a useful
measurement. Even then, it is difficult to accurately control EGR,
for example, to within 5%, so that even if a feedback mechanism is
used to subsequently adjust EGR, any changes in NOx may be too
crude to produce the desired result. Hence, despite the requirement
for extremely highly sensitive measuring equipment, such a method,
as implemented, lacks resolution. To compound the situation, it has
also been found that NOx levels are not a good indicator of
combustion stability, which means that NOx levels could be adjusted
to the detriment of combustion stability and, thus, at the expense
of hydrocarbon emissions, soot and fuel economy.
[0010] AFR sensors provide a fairly useful system for calculating
the amount of EGR. Having determined the approximate level of EGR,
the amount of NOx in the engine-out emissions can be estimated. If
deemed necessary, a feedback signal can then be used to adjust the
AFR through changes in both EGR and fuel injection quantity, so as
ultimately to adjust the (estimated) NOx levels. A particular
problem with this system, however, is that NOx is not measured
directly and, therefore, the calculated values of NOx may be
inaccurate. Also, as noted above, there is no direct link between
NOx emissions and combustion stability.
[0011] In-cylinder pressure based controls can be used to indicate
whether or not to remove EGR. For example, when the measured values
of in-cylinder pressure become too variable (despite that NOx
levels may be low), it is a good indication that EGR is too high
and should be reduced. However, there are numerous disadvantages
associated with reliance on in-cylinder pressure control.
Primarily, such a system relies intensively on micro-processor
power, and also the sensors used are not yet robust enough for
reliability. Further disadvantages are that: there is no direct
link between cylinder in-pressure and NOx levels; such a system may
require multiple cylinder sensors (e.g. one in each cylinder) even
to attempt to achieve a desirable effect; and a very rapid response
from the sensors are needed to provide any meaningful data.
[0012] Hydrocarbon (HC) sensors can provide a means for determining
the efficiency of the combustion process. In this regard, the
higher the concentration of HCs in the engine-out emissions, the
more unburned fuel is in those emissions. The level of HCs can then
be used as a feedback signal to alter fuel injection quantity,
injection timing and EGR, so as to improve the efficiency of
combustion. However, although there is a correlation between HC and
NOx concentration in the engine-out emission, this link can be
broken in the event of misfire or wall wetting. In addition, HC
sensors are sensitive to hydrogen, which means that in some
circumstances measurements of HC concentration can be
inaccurate.
[0013] Ammonia (NH.sub.3) sensors are useful within a feedforward
system for controlling Selective Catalytic Reduction (SCR), i.e.
using a catalytic converter. For example, the amount of NH.sub.3
measured in the engine-out emissions can be used to determine the
amount of urea that must be injected into the catalytic converter
in order to remove NOx in the engine-out emission by catalytic
reduction. The amount of ammonia emitted should preferably be
negligible. Therefore, is ammonia is detected in the engine-out
emission, then too much urea was added during the SCR process.
However, NH.sub.3 levels cannot be directly linked to levels of
engine-out NOx. Thus, although an NH.sub.3 sensor may serve a
useful purpose, e.g. as an indicator for optimising after-treatment
of the engine-out emissions, it cannot be used as a means for
controlling the combustion process itself.
[0014] Thus, there are significant problems associated with each of
the known systems for monitoring engine-out emissions, and
accordingly, none provides a completely satisfactory solution to
the problem of reducing undesirable gas emissions, and particularly
NOx emissions, in the exhaust gas of an internal combustion
engine.
[0015] In recognition of the above shortfalls in the prior art, as
already noted, some form of after-treatment is typically necessary
to reduce exhaust gas emissions to the level required by emissions
regulations.
[0016] Typical the after-treatment is provided by one or more
device, such as a catalytic converter and/or a particulate filter.
A catalytic converter may, for example, be designed to reduce NOx
gases into O.sub.2 and N.sub.2. Alternatively, selective catalytic
reduction, for instance, using urea injection, may be used to
concert NOx gases into N.sub.2 and water.
[0017] However, after-treatments, such as catalytic converters, can
be economically undesirable, because of the costs of manufacture,
the need to regularly monitor and replace old systems, and the
running costs associated with the consumables, such as fuel, which
are required for their operation.
[0018] Therefore, especially in view of the more stringent
emissions criteria that will soon enter into force, there is a need
for a system for accurately determining, and if necessary reducing,
the levels of pollutants in engine-out emissions from internal
combustion engines. More specifically, there is a need for a method
for accurately determining the level of NOx in such engine-out
emissions, and/or for adjusting an engine management system (which
may include control over the combustion process, after-treatments,
or both), so as to reduce the amount of NOx that is released into
the atmosphere from the exhaust gas emissions.
[0019] One way in which levels of NOx (and other undesirable gases)
in exhaust gas emissions may be controlled is by the control of
combustion stability. However, the presently available sensors and
systems do not provide a convenient mechanism by which combustion
stability can be controlled.
[0020] Accordingly, there is an additional need for a system and
method for controlling engine combustion, in particular, combustion
stability, such that it is not necessary to additionally employ
after-treatments, or such that any after-treatments can be used at
a reduced level compared to current systems.
[0021] It would also be an advantage to have a method that achieves
some of the above aims without significantly adding to the cost of
manufacturing or running an internal combustion engine within the
exhaust gas emissions regulations.
[0022] Present means of controlling combustion, and particularly
means of controlling advanced combustion processes in a motor
vehicle are based on monitoring various engine/combustion
parameters, such as NOx production, and using an engine control
unit to make adjustments to engine parameters to push the engine
towards a set of conditions under which it is perceived that NOx
production during the combustion process will be minimised.
However, a difficulty with adjusting such a sensitive process as
advanced combustion, especially when it is pushed in the direction
of minimal NOx production levels (under which conditions combustion
can become unstable), is that it is necessary to have the ability
to: (i) rapidly monitor the combustion process in real time; (ii)
accurately assess the combustion process; and (iii) sensitively
adjust the combustion process, thereby to reduce NOx levels.
Measuring NOx concentration in engine-out emissions is one method
used in the art because, in theory, it provides a direct
measurement of the concentration of NOx in the engine-out
emissions. However, for the reasons discussed herein, the
measurement of engine-out NOx emissions can be technically
challenging, lacking in resolution, and the necessary equipment can
be either impractical or expensive. In addition, it has been found
that NOx emissions do not appear to link well to combustion
stability. Therefore, a direct focus on NOx emissions can be at the
detriment of combustion stability. Measuring CO levels in
engine-out emissions, by comparison, is relatively straightforward,
accurate and, as described herein, provides a link to both
combustion stability and engine-out levels of NOx.
[0023] Thus, this invention aims to overcome or alleviate some of
the problems associated with the prior art.
SUMMARY OF THE INVENTION
[0024] In a first aspect of the invention there is provided a
method for controlling an internal combustion engine to maintain a
predetermined combustion efficiency, the method comprising:
measuring the concentration of CO in an engine-out emission;
determining whether the measured concentration of CO is above or
below a target CO concentration for the engine under the operating
conditions in use; and if the concentration of CO is above the
target CO concentration, adjusting one or more engine and/or
combustion parameters so as to reduce the concentration of CO in
the engine-out emission, and if the concentration of CO is below
the target CO concentration, adjusting one or more engine and/or
combustion parameters so as to increase the concentration of CO in
the engine-out emission; wherein the target CO concentration is
determined on the basis of a correlation with indicators of
combustion efficiency under known engine operating conditions.
[0025] Preferably the one or more engine and/or combustion
parameters is EGR, which, in accordance with the data provided
herein, may be adjusted to change combustion conditions in a
predictable manner.
[0026] Typically, the known engine operating conditions include
engine speed and load, which can affect combustion parameters and
hence, engine-out emissions. Conveniently, the predetermined
combustion efficiency is selected during engine development and/or
testing. In this way an engine control unit (ECU), for example, may
be programmed to control engine and/or combustion parameters to
achieve/maintain the desired level of combustion stability, and/or
fuel economy and/or NOx emission level.
[0027] In one embodiment, the target CO concentration is determined
on the basis of a correlation with one or more engine and/or
combustion parameters which are indicators of combustion
efficiency; wherein the correlation is by a means of data
comparison, the data relating to measured concentrations of CO and
the one or more engine and/or combustion parameters. In preferred
embodiments the means of data comparison is a look-up table,
calibration curve or equation (or similar means) that has been
obtained for the engine under known engine operating conditions,
for example, conditions of engine speed and load. Most preferably,
a plurality (or series) of look-up tables, calibration curves or
equations are used so that a suitable data correlation can be
conducted for any engine operating conditions in use.
[0028] In any or all aspects of the invention, there are some
circumstances wherein the one or more known engine conditions that
were used to obtain the one or more means of data comparison may
not exactly correspond to the engine operating conditions (such as
speed and load) when the method is in use. In such circumstances
the correlation between engine-out CO concentration and indicators
of combustion efficiency obtained under the known engine conditions
may be considered to be an approximation of the correlation that
exists under the engine operating conditions in use. When there is
more than one correlation, i.e. where there is a plurality of means
of data comparison, then the correlation obtained under the known
operating conditions that are closest to the engine operating
conditions in use will preferably provide the appropriate
correlation for use.
[0029] Alternatively, or in addition, the correlation with
indicators of combustion efficiency under known engine operating
conditions may be based on an estimated correlation for the
operating conditions in use. In this case, the estimated
correlation may be obtained by adjusting the correlation obtained
under the known engine operating conditions by means of inter alia
an appropriate regression analysis, equation or algorithm. In other
words, the trend in the correlation between engine-out CO
concentration and the indicator of combustion efficiency (such as
engine-out CO concentration) can be used to anticipate (estimate)
the appropriate correlation under the conditions of use, based on
the available correlation under the known operating conditions.
Again, where there is more than one means of data comparison
(obtained under different engine operating conditions), then
preferably at least the correlation obtained under the conditions
most similar to those in use are used as a basis for the
estimation. As already noted, such a system may apply to any of the
aspects of the invention mentioned herein.
[0030] In a preferred embodiment, the target CO concentration is
predetermined on the basis of a correlation between CO
concentration and hydrocarbon (HC) concentration under known engine
operating conditions, the target CO concentration being that which
corresponds to a predetermined HC concentration at which a
predetermined acceptable level of combustion stability is achieved.
As described below, HC concentration is a good indicator of
combustion efficiency and stability, which have predictable effects
on fuel economy. The preferred fuel economy and/or combustion
stability may be selected or determined during engine testing and
development, for example; and may be selected in accordance with
any required engine specifications or performance characteristics
considered appropriate. Thus, the target CO concentration may, for
example, be the concentration of CO in the engine-out emission that
corresponds to an engine-out HC concentration of 350 to 450 ppm. In
a preferred embodiment the target CO concentration is that which
corresponds to an HC concentration of approximately 400 ppm.
[0031] In a second aspect there is provided a method for
controlling an internal combustion engine to obtain an engine-out
emission level of NOx within a predetermined concentration range,
the method comprising: measuring the concentration of CO in an
engine-out emission; calculating the concentration of NOx in said
engine-out emission based on a correlation with the measured
concentration of CO; determining whether the calculated
concentration of NOx for the engine under the operating conditions
in use is within the predetermined concentration range; and where
the calculated concentration of NOx is outside of the predetermined
concentration range, adjusting one or more engine and/or combustion
parameters so as to achieve an engine-out emission level of NOx
within the predetermined concentration range.
[0032] The predetermined concentration range of NOx may be any
appropriate range, for example, a range that is specified by an
engine or motor vehicle manufacturer, an end user, or a legal
regulation or guideline. Preferably, the predetermined
concentration range is sufficiently low that by use of exhaust
after-treatments, the engine-out emission level of NOx may be
reduced to below national or international NOx emissions criteria.
More preferably, the predetermined concentration range is no
greater than one or more national or international NOx emissions
criteria, so that after-treatments are not required.
[0033] Thus, in preferred embodiments the predetermined
concentration range of NOx may be in the range: 0 to 0.20 g/mile
for Tier II Bin 8; 0 to 0.07 g/mile for Tier II Bin 5; 0 to 0.25
g/km for Euro IV; 0 to approximately 1.25 g/hp*hr for 2007 and
later US On-Highway HD Engines; and 0 to 0.20 g/hp*hr for 2010 and
later US On-Highway HD Engines. However, these preferred
embodiments are in no way intended to restrict the scope of the
invention, which is also suitable for vehicles required to meet
more stringent emissions regulations such as Tier II Bin 3, or Euro
V, which will be known to the skilled person in the art.
[0034] In preferred embodiments the one or more engine and/or
combustion parameters is EGR and, for example, EGR is increased
when the calculated concentration of NOx is higher than the
predetermined concentration range.
[0035] In the alternative, or in addition, the one or more engine
and/or combustion parameters is fuel injection timing. In a further
embodiment, the one or more engine and/or combustion parameters is
injected fuel quantity.
[0036] In preferred embodiments of the invention, an engine control
unit is used to: calculate the concentration of NOx on the basis of
the measured concentration of CO; and adjust one or more engine
and/or combustion parameters using a feedback signal.
[0037] Preferably in this aspect of the invention the correlation
is by a means of data comparison, the data relating to measured
concentrations of CO and NOx in engine-out emissions under known
engine and/or combustion parameters. As in all aspect and methods
of the invention, the means of data comparison may be one or more
look-up tables, calibration curves or equations.
[0038] In some embodiments this method of the invention may further
comprise the steps of: determining whether the calculated
concentration of NOx is below a predetermined concentration level;
and where the calculated concentration of NOx is above the
predetermined concentration level, operating one or more
after-treatments to reduce the concentration of NOx in said
engine-out emission to a level no greater than said predetermined
concentration level. Thus, the method of the invention may, for
example, be used to control SCR (e.g. to avoid ammonia slip), or to
model NOx storage in an NOx adsorber.
[0039] Accordingly, in a third aspect of the invention there is
provided a method for controlling an internal combustion engine to
obtain an engine-out emission level of NOx within a predetermined
concentration range, the method comprising: measuring the
concentration of CO in an engine-out emission; calculating the
concentration of NOx in said engine-out emission based on a
correlation with the measured concentration of CO; determining
whether the calculated concentration of NOx for the engine under
the operating conditions in use is above a predetermined
concentration level; and where the calculated concentration of NOx
is above the predetermined concentration level, operating one or
more after-treatments to reduce the concentration of NOx in said
engine-out emission to a level no greater than said predetermined
concentration level.
[0040] In one embodiment of this aspect of the invention an engine
control unit is used to: calculate the difference between the
calculated concentration of NOx and the predetermined concentration
level; and adjust the one or more after-treatments to reduce the
concentration of NOx in said engine-out emission to a level no
greater than the predetermined concentration level, and wherein the
adjustment to the one or more after-treatments is dependent on the
difference between the calculated concentration of NOx and the
predetermined concentration level.
[0041] Any suitable after treatment may be used. Preferably, the
one or more after-treatments is selected from the group consisting
of: a diesel particulate filter, a catalytic converter, selective
catalytic reduction and an NOx trap or NOx adsorber.
[0042] As in other aspects of the invention, the correlation is by
a means of data comparison, the data relating to measured
concentrations of CO and NOx in engine-out emissions under known
engine and/or combustion parameters; and the means of data
comparison is a look-up table, calibration curve or equation.
Preferably a plurality or series of such look-up tables,
calibration curves or equations is used. Alternatively, the means
of data comparison may be a computer model or models.
[0043] Accordingly, a fourth aspect of the invention provides a
method for controlling an internal combustion engine to obtain an
engine-out emission level of NOx within a predetermined
concentration range, the method comprising: measuring the
concentration of CO in an engine-out emission; calculating the
concentration of NOx in said engine-out emission based on a
correlation with the measured concentration of CO; determining
whether the calculated concentration of NOx for the engine under
the operating conditions in use is within the predetermined
concentration range; and where the calculated concentration of NOx
is outside of the predetermined concentration range, adjusting one
or more engine and/or combustion parameters so as to achieve an
engine-out emission level of NOx within the predetermined
concentration range; and operating one or more after-treatments to
reduce the concentration of NOx in said engine-out emission.
[0044] In a preferred embodiments, the method of this aspect of the
invention further comprises the steps of: determining whether the
calculated concentration of NOx is above a predetermined
concentration level; and where the calculated concentration of NOx
is above the predetermined concentration level, operating the one
or more after-treatments to reduce the concentration of NOx in the
engine-out emission to a level no greater than the predetermined
concentration level.
[0045] In a fifth aspect there is provided a method for calculating
the concentration of NOx in an engine-out emission from an engine,
the method comprising: measuring the concentration of CO in said
engine-out emission; and calculating the concentration of NOx in
said engine-out emission based on a correlation with the measured
concentration of CO under the engine operating conditions in use;
wherein the correlation is by a means of data comparison, the data
relating to measured concentrations of CO and NOx in engine-out
emissions under known engine and/or combustion parameters.
[0046] Similarly, in a sixth aspect of the invention, there is
provided a method for calculating the engine operating Air-to-Fuel
Ratio (AFR) in an engine, the method comprising: measuring the
concentration of CO in the engine-out emission from said engine;
and calculating the AFR of said engine based on a correlation with
the measured concentration of CO under the operating conditions in
use; wherein said correlation is by a means of data comparison,
said data relating to measured AFR levels and concentrations of CO
in engine-out emissions under known engine and/or combustion
parameters.
[0047] Typically, in either or both of the fifth and sixth aspects
of the invention (as in other aspect of the invention), the means
of data comparison is by means of one or more look-up tables,
calibration curves, equations or computer models; and where
necessary, the correlation may be based on an approximation of the
correlation under the engine operating conditions in use, or an
estimation of that correlation. Preferably, the correlation is
performed by an engine control unit.
[0048] In a seventh aspect there is provides the use of a CO sensor
in a system for controlling or adjusting one or more engine and/or
combustion parameters in an internal combustion engine, wherein the
CO sensor is arranged to measure the concentration of CO in an
engine-out emission, and the system includes a means for
correlating the measured concentration of CO with one or more of:
an indicator of combustion efficiency under the operating
conditions in use; the concentration of NOx in the engine-out
emission under the operating conditions in use; and/or the level of
AFR in the engine under the operating conditions in use.
[0049] In one embodiment, the system controls or adjusts one or
more engine and/or combustion parameters in order to bring the
combustion efficiency to within a predetermined concentration
range. In another embodiment, the system controls or adjusts one or
more engine and/or combustion parameters in order to bring the
calculated concentration of NOx in the engine-out emission to
within a predetermined concentration range. In a further embodiment
the system controls or adjusts one or more engine and/or combustion
parameters in order to bring the calculated level of AFR in the
engine to within a predetermined range. Preferably, in one aspect
of the invention, the indicator of combustion efficiency is
hydrocarbon (HC) concentration in the engine-out emission.
[0050] The invention also provides an engine control unit for a
motor vehicle that, in use, performs according to the method of any
aspect or embodiment of the invention. The invention further
provides a motor vehicle comprising such an engine control
unit.
BRIEF DESCRIPTION OF THE FIGURES
[0051] FIG. 1 demonstrates the relationship between: the
coefficient of variance (COV) of "mean effective pressure" (MEP)
within a combustion cylinder (left-hand axis, filled triangles);
and engine-out HC emissions (right-hand axis, filled diamonds), as
a function of NOx emissions for a North American V8 Diesel engine
under steady state operating conditions of 1200 rpm, 59 Nm;
[0052] FIG. 2 demonstrates the relationship between engine-out
emissions of NOx and the amount of EGR used during advanced
combustion in a V8 diesel engine operating at an engine speed and
load of 1600 rpm, 130 Nm (filled diamonds, dotted line); and 1200
rpm, 59 Nm (filled square, solid line);
[0053] FIG. 3 demonstrates the variation in NOx emissions from an
advanced combustion diesel engine in relation to fuel injection
timing for three different fuels, with cetane numbers (CN) of: 25
(open circles); 40 (open triangles); 53 (filled circles);
[0054] FIG. 4 demonstrates the relationship between engine-out
emissions of CO and engine-out emissions of NOx during advanced
combustion under three different engine speeds and loads: 1200 rpm,
59 Nm (filled squares, solid line); 1600 rpm, 50 Nm (filled
triangles, dashed line); and 1600 rpm, 130 Nm (filled diamonds,
dotted line);
[0055] FIG. 5 demonstrates the relationship between engine-out
emissions of CO and engine-out emissions of HCs during advanced
combustion under three different engine speeds and loads: 1200 rpm,
59 Nm (filled squares, solid line); 1600 rpm, 50 Nm (filled
triangles, dashed line); and 1600 rpm, 130 Nm (filled diamonds,
dotted line);
[0056] FIG. 6 demonstrates the relationship between engine-out
emissions of CO and air to fuel ration (AFR) for a North American
V8 Diesel engine capable of Tier II Bin 8 emissions (in a 6000 lb
vehicle), under three different engine speeds and loads: 1200 rpm,
59 Nm (filled squares, left-hand vertical axis); 1600 rpm, 173 Nm
(filled diamonds, left-hand vertical axis); and 1600 rpm, 297 Nm
(filled triangles, right-hand vertical axis);
[0057] FIG. 7 is a schematic representation of a feedback algorithm
for adjusting EGR on the basis of the difference between a target
engine-out CO concentration and a measured engine-out CO
concentration;
[0058] FIG. 8 is a second schematic representation of a feedback
algorithm for adjusting fuel injection timing on the basis of the
difference between a target engine-out CO concentration and a
measured engine-out CO concentration in relation to a maximum
allowable engine-out CO concentration;
[0059] FIG. 9 is a schematic representation of an algorithm for
calculating engine-out NOx levels on the basis of the difference
between a target engine-out CO concentration (correlating to a
target NOx level) and a measured engine-out CO concentration
(correlating to a calculated NOx level);
[0060] FIG. 10 is a schematic representation of an algorithm for
calculating AFR on the basis of engine-out CO concentration.
DETAILED DESCRIPTION OF THE INVENTION
[0061] As used herein, the term "engine-out emission(s)" refers to
the emissions, which may be in the form of gas, liquid or
particulate matter, from an engine combustion chamber itself. Thus,
an engine-out emission has not undergone any so-called
"after-treatments", which are positioned downstream of the
combustion chamber. By way of contrast the term "exhaust gas
emission(s)" is used to refer to the emissions, which again may be
gas, liquid or particulate, that are released from the exhaust pipe
into the atmosphere. Thus, an "exhaust gas emission" may have
undergone one or more after-treatments (e.g. in a catalytic
converter) to remove undesirable molecules contained in an
engine-out emission. Where no after-treatments are used, the
engine-out emissions can be assumed to be the same as the exhaust
gas emissions.
Advanced Combustion
[0062] As noted above, there are a number of internal combustion
processes that aim to reduce engine-out and exhaust gas emissions,
and can be classified as operating a form of "advanced
combustion".
Homogeneous Charge Compression Ignition
[0063] One form of advanced combustion is Homogeneous Charge
Compression Ignition (HCCI), in which well mixed fuel and air
(although another oxidiser could replace the air) are compressed to
a pressure at which they auto-ignite. HCCI combines aspects of the
internal combustion processes of both gasoline engines, i.e.
homogeneous charge spark ignition, and diesel engines, i.e.
stratified charge compression ignition. However, in contrast to the
gasoline system, rather than using a spark to ignite the fuel-air
mixture, compression is used to raise the temperature and
concentration of the mixture until the entire fuel-air mixture
reacts (simultaneously). In a typical diesel injection engine,
compression is also used to increase temperature and gas
concentration, but the combustion event begins at the interface
between the injected fuel and the compressed combustion chamber
gases.
[0064] Since in HCCI, ignition of the fuel-air mixture can occur in
several places at the same time, the entire gas content of the
combustion chamber can combust almost simultaneously. This fact,
combined with the lack of a direct ignition event (e.g. from a
spark at a predetermined time), makes the HCCI process inherently
difficult to control. For example, if auto-ignition occurs too
early, or with too much ferocity, there is the risk that the
extremely high gas pressure inside the cylinder could destroy an
engine. Therefore, to avoid this situation, it is typical for HCCI
to be carried out with lean fuel mixtures.
[0065] The homogenous mixing of fuel and air before ignition and
the lean fuel mixtures in the HCCI process achieve a number of
advantages over alternative combustion processes, including
excellent fuel economy and cleaner combustion, which results in low
engine emissions. In particular, due to the lower combustion
temperatures, HCCI engines are known to achieve very low levels of
NOx emissions, which under some emissions regulations may obviate
the need for catalytic converter after-treatments.
[0066] However, the low temperature combustion in HCCI has the
disadvantage that CO and HC emissions are typically relatively
high, so that appropriate after-treatments may be required to meet
those respective emissions criteria.
[0067] Two of the main disadvantages of current HCCI processes are:
(i) the difficulty in controlling the process (as already noted);
and (ii) the limited power range available. These issues are
discussed below.
[0068] In contrast to the tightly controlled ignition events of
other popular combustion systems, such as gasoline engines (which
ignite with a spark) and direct-injection diesel engines (which
ignite when fuel in injected into compressed air); in HHCI, the
homogeneous mixture of fuel and air is compressed, and combustion
begins whenever the appropriate conditions of fuel-air pressure and
temperature are reached. Hence, there is no well-defined combustion
initiation event that can be directly controlled. Control of
ignition in HCCI is, therefore, a major hurdle that needs to be
overcome.
[0069] The concentration and/or temperature of the fuel and air
mixture in an HCCI engine can be increased (i.e. to trigger
combustion) in several different ways, for example, by using: (i) a
high compression ratio; (ii) pre-heat induction gases; (iii) forced
induction; and (iv) exhaust gas retention/recirculation.
[0070] These processes can be adjusted so that the appropriate
ignition conditions occur at a desirable timing. However, these
conditions only apply under specific operating parameters of, e.g.
engine work, torque and fuel composition.
[0071] In order to dynamically control an HCCI engine, the engine
management/control system must be capable of adjusting engine
conditions, for example, to control combustion timing, in response
to real-time signals (or data) obtained from the engine system.
Thus, in response to an appropriate signal, the engine management
system would beneficially adjust (and monitor) one or more of: the
compression ratio, the inducted gas temperature, the inducted gas
pressure, or the quantity of retained or re-inducted exhaust.
Variable Compression Ratio
[0072] There are various methods known to the person of skill in
the art for adjusting the compression ratio in a combustion
cylinder. By way of example, the "effective" compression ratio can
be altered by changing the timing of closure of the fuel/air intake
valve(s). Thus, by closing the intake valve(s) late in the
combustion cycle, the effective compression ratio will be reduced,
and by closing the intake valve(s) early, the effecting compression
ratio will be increased. This may be known as variable valve
actuation. Alternatively or in addition, the "geometric"
compression ratio may be varied using a movable plunger at the top
of the cylinder head, to relatively increase or decrease the volume
of the combustion chamber. These methods are technically quite
difficult to implement and can be expensive.
Variable Induction Temperature
[0073] This technique requires a means of rapidly changing the
temperature of the intake fuel (charge), from one combustion cycle
to the next. Such a system requires extremely fast thermal
management, may be expensive, and has a limited operational
range.
Variable Exhaust Gas Percentage
[0074] Exhaust gas can be very hot if retained from the preceding
combustion cycle. In this case, the hot combustion will increase
the temperature of the gases in the combustion cylinder, and will
typically advance the combustion event. In contrast, if EGR is
recirculated through the intake (as in conventional systems), the
gases entering the combustion chamber are cooler and the fresh
charge is diluted. As a consequence, ignition is delayed.
[0075] EGR is perhaps the most popular form of NOx reduction
technique, since it is used in many engine types. Dilution and
mixing of engine-in fuel and air with recirculated exhaust gas
dilutes the fuel with some inert gas, thereby lowering the
adiabatic flame temperature and (in diesel engines) reducing excess
oxygen. Another benefit is that the exhaust gas increases the
specific heat capacity of the mix, which lowers the peak combustion
temperature, and reduces the generation of NOx.
Variable Valve Actuation
[0076] Variable valve actuation can be a complicated and expensive
system, but advantageously, it can be used to control both the
compression ratio and the exhaust gas percentage.
High Peak Pressures and Heat Release Rates
[0077] In HCCI, because the entire fuel/air mixture ignites and
burns almost simultaneously, the process involves higher peak
pressures within the combustion cylinder, and higher energy release
rates, than in other combustion processes. Thus, to physically
withstand these higher pressures, an HCCI engine must be
structurally stronger, and consequently heavier. To reduce the need
for heavier engines, it has been proposed to reduce the combustion
rate. One such proposal is to mix two different fuel, having
different combustabilities, such that combustion occurs over a
longer period. However, it is then necessary to provide an entirely
different engine structure to accommodate the different fuel
supplies. In addition, the lower combustion pressure and rate leads
to reduced engine power.
Power
[0078] In non-advanced combustion engines, power output is simply
increased by increasing the amount of fuel (or fuel/air charge)
that is injected in to the combustion chamber. In HCCI combustion,
however, the problem of the high peak pressures (discussed above)
is also relevant to maximum power output. Since the entire fuel/air
mixture combusts almost simultaneously, to increase power by
increasing the fuel to air ratio will result in higher peak
pressures and higher heat release rates. In addition, some of the
common processes used in HCCI for controlling combustion, such as
exhaust gas retention, mean that the fuel and air mix is preheated,
and accordingly less dense. Therefore, the amount of fuel contained
in the combustion chamber is actually reduced. Finally, "knocking",
which is generally reduced or eliminated in HCCI engines, can
become a problem at high fuel to air ratios.
[0079] For these reasons, increasing the power output from a HCCI
engine is quite difficult.
[0080] One way of increasing power output can be to use different
types or blends of fuels, as discussed above, by extending the
combustion period. Alternatively, it may be possible to produce
thermal gradients in the combustion cylinder, such that different
regions combust at slightly different times. Another method is to
run the engine in HCCI mode only for those periods during which the
engine load is within the optimal conditions for HCCI, and to run
it by the conventional spark or diesel ignition modes during other
periods. [0081] All of these issues and potential solutions can
have serious consequences in terms of engine-out emissions, and
these must also be addressed in order for advanced combustion
engines to remain within emissions regulations.
Premixed Charge Compression Ignition
[0082] Premixed charge compression ignition (PCCI) aims to address
the problems of combustion control and specific power output in
HCCI engines. PCCI can be considered to be a generalisation of
HCCI, in which the fuel and air mixture may be partially stratified
at the moment of ignition (some of these measures have already been
discussed under "Homogeneous Charge Compression Ignition",
above).
[0083] Thus, PCCI engines can include direct injected engines with
early injection, and controlled auto-ignition engines that use
variable valve timing and high residual exhaust gas fraction to
control combustion. In PCCI engines, the stratification of the fuel
to air ratio (equivalence ratio) may be used for lengthening the
burn duration, thereby allowing the engine to operate at higher
specific power. Engine control can also be aided by direct
injection of fuel into the combustion chamber, especially if the
injected fuel is highly reactive, such as diesel fuel. However, as
already noted, a significant disadvantage of fuel stratification is
that high NOx and particulate matter emissions may result.
Fuel Injection Timing
[0084] As indicated above, fuel injection timing can be an
important aspect of advanced combustion, both in terms of producing
a relatively homogeneous fuel/air mixture to achieve the benefits
of HCCI, but also in terms of controlling the combustion and power
output.
Port Injection
[0085] The easiest way of creating a homogenous in-cylinder
fuel/air mixture is by injecting fuel upstream of the intake
valves, and inducting the mixture into the cylinder during the
intake stroke, such that the turbulence generated during intake
promotes mixing. This approach is the most common system used in
HCCI engines. However, it has some disadvantages. First, because
fuel is necessarily injected early, there is no option for
adjusting injection timing in order to control the start of
combustion. Secondly, it can result in high HC and CO emissions, as
well as in increased fuel consumption and oil dilution. These
problems are particularly relevant for less volatile diesel
fuel.
Early In-Cylinder Injection
[0086] Along similar lines to the above method, early in-cylinder
injection (i.e. well in advance of TDC), of all or part of the
fuel, can help promote a homogeneous fuel/air mixture. As above,
this system does not allow effective control of the combustion
process through varying the fuel introduction time. Another concern
is the amount of fuel wall impingement, particularly when heavy
fuels such as diesel are injected into the low-density environment
of the combustion chamber. To address this concern, low-penetration
fuel injectors are being developed. Despite its drawbacks, early
in-cylinder injection is gaining popularity in HCCI engines.
Late In-Cylinder Injection
[0087] As the name suggests, in this system, fuel is injected
directly into the combustion chamber near or after TDC. However,
the system is able to delay the ignition event by: (i) using large
amounts of cooled EGR; (ii) reducing the engine compression ratio;
and (iii) generating vigorous swirl. In this way, combustion can be
delayed until well after the end of injection. Advantageously, low
NOx emissions are achieved despite the lack of mixture homogeneity
within the cylinder. Moreover, in the case of diesel fuel, the
problems associated with fuel wall impingement (e.g. of early
in-cylinder injection) are reduced or avoided, and some control
over the combustion timing is enabled. However, at high engine
loads it may not always be possible to delay ignition until after
the end of fuel injection.
[0088] Thus, a preferred mode of operating a diesel HCCI engine in
accordance with the invention involves the use of late in-cylinder
injection of fuel.
Exhaust Gas Emissions
[0089] Although HCCI can be extremely effective in reducing
engine-out emissions, a combustion system has not yet been
developed that will be able to meet future exhaust gas emissions
regulations over a broad engine operating range. The engine-out
emissions relevant in HCCI are briefly discussed below.
Nitrogen Oxides (NOx)
[0090] The greatest general benefit of HCCI is the reductions that
can be achieved in engine-out NOx emissions. For example, typical
NOx emissions from HCCI may be over 90% lower than in conventional
Diesel combustion engines. As already noted, this benefit is
achieved through the lower combustion temperatures reached in an
HCCI engine. Since combustion temperature is proportional to the
fuel/air ratio, as engine load increases, NOx emissions typically
also increase (Bergman & Golovitchev, 2005). Thus, HCCI
combustion is practical only at low engine loads and low fuel/air
ratios. NOx emission from HCCI combustion of Diesel fuel has been
modelled (Dodge et al., 1998; Dickey et al., 1999), and it was
found that the beneficial reduction in NOx emissions from HCCI at
low engine loads, in comparison to conventional direct-injection
diesel combustion engines, was lost at high engine loads. In a
similar study on the effect of combustion phasing, it was also
found that at high engine loads, premature ignition resulted in a
large increase in NOx emission levels; while under low and medium
loads, the NOx levels remained acceptably low.
Particulate Matter (PM)
[0091] HCCI combustion has also been reported to produce low levels
of smoke and PM emissions (e.g. Suzuki et al., 1997; Mase et al.,
1998). In this regard, it is likely that the homogeneous charge
removes fuel-rich regions and encourages even combustion. However,
the processes operating within HCCI to reduce NOx emissions, such
as high EGR, can lead to increased PM formation, due to lower
combustion temperatures and lower oxygen availability. Also, under
some operating conditions, for example, with late in-cylinder
injection, deposition of liquid fuel on the combustion chamber wall
may cause localised fuel-rich regions, which promote soot
formation. In general, PM emissions tend to increase in HCCI
combustion as the fuel/air ratio is increased.
Hydrocarbons (HC) and Carbon Monoxide (CO)
[0092] As already mentioned, the lean fuel mixtures (i.e. low fuel
to air ratio) used in HCCI engines means that the combustion
process typically operates at lower peak temperatures than in other
engines. These lower peak temperatures are beneficial in reducing
the amount of engine-out NOx gases produced, which is a primary
concern in regard to emissions regulations. However, in contrast to
the generally advantageous NOx emissions in HCCI combustion, HC and
CO emissions are typically higher than in conventional
direct-injection diesel engines (Suzuki et al., 1997).
[0093] Probably, the most significant causal factor for the high HC
and CO emissions is the low combustion temperature, which leads to
lower oxidation rates within the combustion chamber. As discussed,
this is an inevitable consequence of the lean fuel mixtures and
high levels of EGR that are necessary to perform HCCI with low NOx
emission. Again, liquid fuel deposition on combustion chamber walls
(a particular problem with diesel fuel) may also result in
increased HC emission levels (Stanglmaier et al., 1999).
[0094] In addition, in combustion models it has been found that CO
concentration increases with the fuel/air ratio (Bergman et al.,
2005), probably because the oxidation of CO to CO.sub.2 is
inhibited by lack of oxygen at high equivalence ratios. Meanwhile,
at low fuel to air ratios, CO emissions have been found to increase
at low temperatures.
[0095] CO emissions can, therefore, be considered to be a side
effect of both low soot and NOx emissions. The present invention
benefits from the higher levels of CO emissions in HCCI combustion,
which is demonstrated by the data provided herein, to provide a
means of not only monitoring engine performance, but also
monitoring NOx emissions.
After-Treatments
[0096] As discussed, in order to meet emissions regulations in the
EU and the US, diesel engines typically require one or more
after-treatment device, to concert the engine-out emissions into
exhaust gas emissions having acceptable levels of pollutants. The
most common forms of after-treatments for diesel engines are a
diesel particulate filter and a catalytic converter. In fact, both
of these devices will typically be used, and may be positioned
sequentially within the same housing, for example, within an
expansion chamber.
Diesel Particulate Filter (DPF)
[0097] A DPF is designed to remove particulate matter (e.g. soot)
from the engine-out emissions of a diesel engine. The efficiency of
approved DPF devices tends to be in the region of 85-90%, such
that, when fitted, no visible smoke should be emitted from an
exhaust pipe.
[0098] By its nature a DPF requires maintenance, because otherwise
the filtered soot (and other particles) that builds up on the inlet
face of the filter can eventually lead to clogging of the filter.
If too much PM is retained by the DPF, the pressure differential
(back pressure) created across the filter can lead to engine
damage. Therefore, it is necessary to either replace old DPFs or
employ a means of regenerating the DPF to remove the PM from the
filter.
[0099] There are several varieties of filter, which may be
disposable, or in some cases can be regenerated.
[0100] Cordierite wall flow filters, which contain a ceramic
material are the most common type of diesel filter. These filters
are simple to use, relatively inexpensive and are highly efficient.
It is possible to regenerate cordierite filters by burning off the
captured soot particles. However, they have a relatively low
melting point (approximately 1200.degree. C.), which means that
there is a danger of them melting during some regeneration
procedures.
[0101] Silicon carbide wall flow filters are another popular filter
type, which can be interchanged with cordierite wall flow filters
in an engine exhaust. Although the silicon carbide cores have a
higher melting point (1700.degree. C.) than cordierite, they are
thermally less stable and can be more expensive to use.
[0102] Metal fibre flow-through filters have cores of metal fibres,
which are generally woven. These have the advantage that they can
be readily heated for regeneration purposes by passing an electric
current through the core. However, a disadvantage is that they are
relatively expensive and not interchangeable with the above
filters.
[0103] Disposable paper cores are not commonly used. However, they
are used in certain circumstances where vehicles are used in-doors.
The engine-out emissions are usually cooled through a water trap
before contacting the paper.
[0104] Partial filters achieve between 50% and 85% PM filtration.
The most common application of these filters is for retrofit; i.e.
when a vehicle was not initially manufactured with a PM filter.
Regeneration of DPFs
[0105] DPFs may be regenerated by heating the filter to
temperatures of approximately 600.degree. C. or above, so as to
burn off any PM. By adding a catalyst to the DPF, the regeneration
temperature can be reduced, for example, to about 300.degree. C. In
catalysed DPFs regeneration can be aided by increased amounts of
NO.sub.2, while the process may be hindered by sulphur.
Regeneration can be "passive"; or when the DPF is heated, for
example, using a fuel burner, it is considered to be an "active"
mechanism.
[0106] An active mechanism of regeneration may be triggered by an
engine management system (computer) that responds to the load of PM
on the DPF, for example, by measuring the back pressure and/or the
temperature of the filter. Active DPF management may use one of a
variety of systems to increase the exhaust temperature, such as: a
fuel burner; engine management; a catalytic oxidiser; resistive
heating coils; or microwave energy. All of these systems require
fuel for operation. Therefore, it is uneconomical to run a
regeneration cycle too frequently, but potentially damaging to the
engine to allow too much build up of PM.
[0107] The methods and engine control systems of the present
invention encompass active regeneration systems for DPF, when
used.
Catalytic Converter (CC)
[0108] A catalytic converter (CC) can be used to reduce or
eliminate levels of CO, HC and NOx in engine-out emissions. The
conversion reactions that are catalysed by the CC are as
follows:
oxidation of CO: 2CO+O.sub.2.fwdarw.2CO.sub.2 (i)
oxidation of HCs: CxHy+nO.sub.2.fwdarw.xCO.sub.2+mH.sub.2O (ii)
reduction of NOx: 2NOx.fwdarw.xO.sub.2+N.sub.2 (iii)
[0109] Under earlier emissions regulations it was possible to use a
"two-way" CC for treating emissions from a diesel engine, because
it was only necessary to reduce the exhaust emission of CO and HCs.
However, under current emissions regulations (and probably also
future legislation) it is preferable to employ a "three-way" CC,
which is able to remove CO, HCs and NOx from the exhaust emissions.
That said, typically, three-way CCs are not practical for use with
a diesel engine. Therefore, it is more likely that two different
types of catalyst will be employed concurrently, such as an
oxidations catalyst and an NOx adsorber or SCR. Furthermore, it is
possible that where an advanced combustion system is operated in
such a way as to sufficiently reduce engine-out NOx emissions (for
example, using the systems and methods of the invention), it may
again be possible to use only a two-way CC. It is particularly
advantageous to reduce the engine-out emissions of NOx (i.e.
pre-after-treatment) in advanced combustion systems, because the
lean fuel mixtures used in advanced combustion lead to high oxygen
levels (e.g. up to 15%) in the engine-out emissions, which can
greatly inhibit the ability of a CC after-treatment to reduce NOx
to N.sub.2.
[0110] The catalytic component of a CC is typically a precious
metal, such as platinum. However, due to its expense and the
possibility of unwanted side-reactions, palladium and/or rhodium
may alternatively be used. Other materials such as cerium, iron,
manganese and nickel could also be used. Further components of a CC
are the core (or substrate), which is generally a ceramic or
stainless steel honeycomb and is used to support the catalyst; and
the washcoat, which is usually a mixture of silicon and aluminium,
and is used to provide a rough surface (i.e. a large surface area
for reaction).
[0111] An example of a CC that may be used is the Plasma-Assisted
Catalytic Reduction process (PACR), which oxidises NO and HCs to
NO.sub.2 and partially oxidised HCs, respectively, which are then
reduced over a catalyst to N.sub.2, CO.sub.2 and water.
[0112] In a diesel engine, it is common to use an "oxidation
catalyst" to oxidise CO and HCs as already described above. An
oxidation catalyst can remove 90% of these pollutants. However,
levels of NOx cannot be reduced in such an oxidising
environment.
[0113] Hence, to remove NOx from engine-out emissions, it is
generally necessary to employ an alternative technology, such as an
NOx trap or "selective catalytic reduction" (SCR).
Selective Catalytic Reduction (SCR)
[0114] In SCR a reductant, such as ammonia or urea, is added to the
engine-out emission stream so that it is absorbed onto a catalyst
in a CC. The reductant then reacts with any NOx in the engine-out
emissions, converting it into N.sub.2 and water. Thus, SCR requires
both a reductant and a catalyst. An engine control unit is
typically used to control the amount of reductant that is released
into the engine-out emission stream. For the SCR process to be
successful, special types of catalyst are required, for instance,
catalysts based on vanadium, or having zeolites in the washcoat. In
addition, the reaction process can be temperature dependent, so it
can be necessary for the engine control unit to also regulate the
temperature of the engine-out emissions that contact the CC. A
disadvantage of the process is that the catalyst can require
regular cleaning/maintenance, because the reductant can solidify
onto the catalyst, particularly when the engine-out emissions are
too cold, and especially if the reductant contains any impurities.
In particular, the urea injection system in SCR can be difficult to
design robustly for use in cold conditions.
[0115] The choice of reductant depends on various factors. Urea
tends to be a preferred reductant in motor vehicles because it is
less toxic than ammonia. However, the use of urea generates ammonia
in the exhaust system. This can be a particular concern when
unreacted reductants are released in the exhaust gas emissions
(e.g. ammonia slip). The benefit of using ammonia, however, is that
it is a more effective reductant of NOx. Different SCR equipment
may be required, depending on the selection of reductant.
NOx Trap
[0116] A NOx trap provides a supplement (or an alternative) to the
use of techniques, such as EGR and SCR, which are designed to
reduce NOx in exhaust gas emissions.
[0117] An NOx trap (or NOx adsorber) requires a zeolite material,
which can be included into a CC by applying it with a washcoat. As
the name suggests, an NOx trap simply traps NO and NO.sub.2
molecules, preventing them from being released in the exhaust gas
emissions. Of course, once the trap is full, it must either be
replaced or regenerated. Regeneration ("purging") can be achieved,
by injecting diesel (or another reactive agent), into the
engine-out emissions up-stream of the NOx trap. The HCs in the
reactant will react with the NOx to produce water and N.sub.2.
[0118] Currently, NOx traps are relatively expensive.
[0119] In accordance with the methods of the present invention, a
three-way CC may be employed to reduce/eliminate engine-out
emissions of CO, HC and NOx. Preferably, a two-way CC is used where
the engine-out emissions are of a composition that would inhibit
the reduction of NOx gases in a three-way CC. In alternative
embodiments, an NOx trap or SCR may be employed to reduce NOx
emissions, in addition to or instead of either a three-way or an
two-way CC. More preferably, no CC is used, for instance, in cases
where the combustion process can be operated under conditions where
the engine-out emissions of each of the undesirable pollutants
(such as NOx) are reduced to an acceptable level in the engine-out
emissions. Most preferably, no after-treatments are used when the
levels of undesirable pollutants are sufficiently low in the
engine-out emissions. In certain preferred embodiments, an engine
control unit is used to control the release of reductant in SCR
and/or the release of reactant for an NOx trap, according to the
requirements of the after-treatment processes used.
Sensors
[0120] To regulate and/or reduce the exhaust gas emissions from an
internal combustion engine, it is preferable to attempt to minimise
(or eliminate) the production of undesirable emissions at the
combustion stage, so that reliance on after-treatments can be
reduced or even obviated. In this regard, although an internal
combustion engine can be pre-programmed to operate according to
parameters shown to reduce engine emissions, e.g. during engine
testing; it is known that engine-out emissions can vary
considerably during actual use. For example, according to the type
of fuel used, the engine operating conditions (e.g. engine load),
and the age of the engine (e.g. due to engine wear).
[0121] Therefore, the most effective means of maintaining low
engine-out emissions would appear to involve a reactive mechanism,
in which the emissions are regularly (or continuously) monitored,
and an engine control unit is provided to feedback signals to
change/control the engine operating conditions, according to the
levels of pollutants detected. In this way, one or more engine
parameters can be adjusted, in real time, to optimise the engine
running conditions (i.e. attempt to minimise the production of
undesirable engine-out emission gases), at that point in time.
[0122] A further advantage of monitoring the engine-out emissions
is that an engine control unit can be provided to feedforward
signals to change/control after-treatment processes, such that any
after-treatments, when used, are operated under the optimal
conditions in relation to the levels of the various engine-out
emission gases at that point in time.
[0123] To employ an engine control unit in this manner, it is, of
course, necessary to know the levels (concentration) of one or more
of the gases of interest in the engine-out emissions. It is,
therefore, convenient to provide one or more sensors in the exhaust
system downstream of the combustion cylinder, but upstream of any
after-treatments, if used. Such a sensor can detect or measure gas,
liquid or particle concentration in the engine-out emissions,
before those emissions are released into the atmosphere (as exhaust
gas emissions), and before they have been exposed to any
after-treatments.
[0124] In addition, one or more of the same or different sensors
may be provided in the exhaust system downstream of any
after-treatments that may be used, so that the amount of any
particular component of the exhaust gas emission can be
determined.
[0125] As has been mentioned above, there are a number of sensors
that may advantageously be used to detect or measure the
concentrations of one or more components of interest in the
engine-out emissions from an internal combustion engine (and
particularly from an advanced combustion engine). These sensors
include: NOx sensors, Oxygen sensors (in conjunction with AFR
meters), HC sensors, and ammonia sensors; some of which are
reviewed by Sheikh et al., "Ceramic sensors for industrial
applications"
(http://www.mse.eng.ohio-state.edu/.about.akbar/paper.htm) In
addition, some sensors, such as in-cylinder pressure sensors and
oxygen sensors may be used to indirectly estimate the concentration
of relevant engine-out emission gases, e.g. NOx.
Oxides of Nitrogen (NOx) Sensors
[0126] NOx sensors provide the only direct means of monitoring the
engine-out NOx emissions from an internal combustion engine. The
term "NOx" relates to all of the oxides of nitrogen, including but
not limited to: NO (the most common nitrogen oxide in engine-out
emissions), NO.sub.2, and N.sub.2O.
[0127] Most of the NOx sensors that are currently available are
constructed from metal oxides, such as yttria-stabilized zirconia
(YSZ), which are compacted into a dense ceramic material that can
conduct oxygen ions (O2-). High temperature electrodes made from
relatively inert metals (such as platinum, gold, or palladium), or
certain metal oxides, are placed on the ceramic material and an
electrical signal (for example, resulting from a change in current
or voltage within the sensor) is measured as a function of NOx
concentration.
[0128] Sensitivity and selectivity of NOx sensors currently
available is of particular concern given the low levels of NOx in
the engine-out emissions from internal combustion engines. For
example, it may be expected that a typical gasoline combustion
engine would produce levels of NO in the region of up to 2000 ppm,
and levels of NO.sub.2 in the region of up to 200 ppm. However, to
be effective within an advanced combustion system, which may
generate NOx emissions of less than 10 ppm, such a sensor must be
even more sensitive to NOx. Other drawbacks of NOx sensors
currently available for use in motor vehicle exhaust systems,
relate to: response time, in view of the frequency with which
engine conditions and engine emissions may vary; reliability; and
cost.
[0129] In use, if an NOx sensor detects too high a concentration
level in the engine-out emissions: (i) a feedback signal may be
used to control e.g. EGR, in an attempt to adjust the amount of NOx
produced; and (ii) a feedforward signal may be used to optimise the
catalytic conversion (reduction) of NOx and/or purging
(regeneration) of any NOx traps that may be used.
Oxygen (O.sub.2) Sensors
[0130] Oxygen sensors (O.sub.2 sensors, lambda probes, or lambda
sensors) can be fitted into the exhaust system of a motor vehicle
to measure the concentration of oxygen in the engine-out emissions.
It is then possible to calculate the AFR in the engine-out
emissions. In this regard, an AFR meter and a wide band O.sub.2
sensors are one in the same, since AFR is directly related to
excess O.sub.2 in the exhaust. The AFR can be used as an indication
of whether the charge entering the combustion chamber is too rich
or too lean for efficient combustion. Thus, once the AFR in the
engine-out emissions is known, an engine control unit may be used
to control the efficiency of the combustion process (i.e. to give
the best fuel economy and lowest exhaust emissions). For example,
by measuring the proportion of oxygen in the exhaust gas (and by
knowing other relevant parameters, such as the volume and
temperature of the air entering the combustion cylinders), an
engine control unit may, for example, use look-up tables or a
computer model to determine the optimal amount of fuel that should
be injected into the combustion cylinder to completely combust with
the available oxygen (i.e. the optimal stoichiometric ratio of air
to fuel can be determined).
[0131] A typical oxygen sensor is made from a ceramic cylinder
(such as zirconium dioxide), which is plated inside and out with
porous platinum electrodes. This may then be encased in metal
gauze. The sensor provides an electrical signal in proportion to
the difference in O.sub.2 concentration between the engine-out
emissions (i.e. the environment to be measured) and the external
air. Generally, the sensors are required to be at a temperature of
approximately 300.degree. C. in order to work effectively, and
therefore, the sensor may include its own heating element.
[0132] The above type of oxygen sensor is known as a "narrow band"
oxygen sensors, which means that the measurement of oxygen
concentration is extremely crude. In fact, the measurement is
essentially binary, allowing only a reading of "lean" (i.e. high
AFR), or "rich" (i.e. low AFR); so that the engine control unit
tends to adjust the combustion process cyclically between these two
extremes. In addition, these sensors can be slow to respond,
because they average the signals over a period of time. Another
type of narrow band oxygen sensor is made of titanium oxide. These
type of sensors are largely irrelevant to diesel engine control
systems, because, by way of example, "rich" and "lean" are relative
terms, and the term rich when used to describe diesel AFRs is
likely to refer to an AFR that is still quite lean in relation to
other engine types.
[0133] An improved type of oxygen sensor is known as a "wide band"
oxygen sensor. It comprises a planar zirconium oxide element and an
electrochemical gas pump, which (by way of a feedback loop)
controls the gas pump current to keep the output of the
electrochemical cell constant. In this way, the gas pump current
directly indicates the oxygen content of the exhaust gas.
Accordingly, the sensor can determine the oxygen content far more
rapidly (without the need for averaging), and allows the engine
control unit to more effective control the combustion process.
These sensors are not yet in wide use in motor vehicles.
[0134] The optimal position for an oxygen sensor in a motor vehicle
is between the combustion cylinder and the CC, when used. However,
a sensor may also be used after the CC.
[0135] As previously noted, these sensors cannot be used to
accurately determine or reduce engine-out emissions of NOx.
Hydrocarbon (HC) sensors
[0136] As already noted, an HC sensor can provide a means for
determining the efficiency of the combustion process, since the
higher the concentration of HCs in the engine-out emissions, the
more unburned fuel is present, and therefore, the less efficient is
the combustion process. Once the concentration of HCs in the
engine-out emissions is known, an engine control unit can provide a
feedback signal to alter one or more of fuel injection quantity,
fuel injection timing and EGR, so as to improve the efficiency of
combustion.
[0137] A conventional hydrocarbon sensor comprises: a thin, solid
electrolyte layer having high proton conductivity; a pair of
electrodes, which sandwich the proton conducting layer; and a
catalytic surface on the anode, which liberates protons from HCs.
Protons liberated from HCs travel through the electrolyte layer
under the influence of a (constant) potential difference across the
electrodes, and the current measured is proportional to the
concentration of HC at the anodic surface. HC sensors are
described, for example, in U.S. Pat. No. 6,103,080, U.S. Pt. No.
6,238,535 and U.S. Pat. No. 6,037,183.
[0138] Various chemical reactions can be used for the generation of
hydrogen from HCs, including dehydrogenation, cracking, and steam
reforming.
[0139] Commercially available compounds such as Fe.sub.2O.sub.3 or
FeO(OH) may be used as dehydrogenation catalysts. Another suitable
dehydrogenation catalyst is LaFeO.sub.3, which has better stability
than some other Fe-containing catalysts. Precious metals, such as
Pt or Pd may also be used as dehydrogenation catalysts (when
supported on porous ceramic materials, such as MgO,
Al.sub.2O.sub.3, or a silica gel). However, these materials may be
less selective that the Fe-containing catalysts. Cracking catalysts
are compounds such as La.sub.1-xCe.sub.xFeO.sub.3 (e.g.
La.sub.0.9Ce.sub.0.1FeO.sub.3), while steam reforming catalysts
includes materials such as NiO, nickel metal and precious
metals.
[0140] The catalyst material in the form of a powder is typically
mixed with an organic solvent to form a slurry and applied to the
surface of the sensor electrode.
[0141] As electrodes in HC sensors, numerous materials may be used,
including Pt, Pd, Ag, Au, or their alloys, and certain metal
oxides.
[0142] The thin solid electrolyte layer can be a Ba-Ce-based oxide
layer.
[0143] The selectivity (i.e. cross-reactivity) of HC sensors is of
some concern. For instance, although a conventional HC sensor can
be capable of linearly responding to HC concentration in the range
of several percent down to a few ppm; at low concentrations of HCs
(e.g. less than 10 ppm), the sensors may be significantly confused
by oxygen in the atmosphere. This problem is of particular concern
when monitoring HC concentration in the engine-out emissions of
advanced combustion engines, which operate with lean fuel-air
mixtures. In addition to oxygen effects, HC sensors tend to be
sensitive to hydrogen. Therefore, especially at low concentrations
of HC, an HC sensor may give quite inaccurate measurements of HC
level.
[0144] Alternative types of HC sensors, such as fibre optic
devices, are under development (e.g. US 2006/165344). These sensors
should be less sensitive to non-HC contaminants.
[0145] A further limitation of HC sensors for use in controlling
engine combustion is that, as already noted, there is not such a
reliable relationship between HC and NOx concentration in
engine-out emissions, as there is between NOx and CO, as
demonstrated herein. Therefore, although an engine control unit may
adjust the combustion cycle in response to engine-out HC levels,
this system may not be capable of accurately adjusting NOx
levels.
Ammonia (NH.sub.3) sensors
[0146] Ammonia (NH.sub.3) sensors may be used in a feedback system
for controlling Selective Catalytic Reduction (SCR); in particular,
such a sensor may be used downstream of a CC to monitor the amount
of NH.sub.3 in exhaust gas emissions, and report to an engine
control unit in the event of "ammonia slip" (i.e. too much urea
injected during SCR). There are a wide variety of NH.sub.3 sensors,
including: solid-state sensors, semi-conductor sensors,
electrochemical sensors and optical (e.g. infra-red) sensors (some
of which are discussed in US 2006/194330 and U.S. Pat. No.
6,069,013, for example).
[0147] Solid-state sensors provide high reliability, relatively low
cost, and can measure NH.sub.3 in the range from low ppm range to
several percent. However, some can be quite sensitive to other
gases (e.g. O.sub.2) in engine-out emissions, and are, therefore,
impractical for monitoring engine-out emissions in motor vehicles.
Semi-conductor sensors are typically based on materials such as
metal oxides or polymers, and measure the change in resistance or
capacitance of the coating as a function of adsorbed molecules.
They are capable of measuring NH.sub.3 concentrations in the low
ppm range. However, a problem with semi-conductor oxides is the
issue of cross-contamination, since most gases adsorb onto
high-surface area ceramic substrates to some extent. In engine-out
or exhaust gas streams the main problem of cross-contamination is
with CO and NOx. Meanwhile, a disadvantage of polymer-based sensors
is that they can only be used at relatively low temperatures, at
which they are chemically stable. A further problem with these
sensors is that gas absorption is kinetically relatively slow and
can be inhibited at high temperatures, which means that these
sensors can have a slow response time and are not best suited for
monitoring engine-out emissions.
[0148] Electrochemical sensors can provide a high degree of gas
specificity. An example of an electrochemical sensor is a
mixed-potential based ceramic sensors, which comprise an oxygen ion
conducting membrane and metal, metal oxide or perovskite sensing
electrodes. Their high temperature operating capabilities can make
them suitable for use in exhaust/engine-out emission streams.
Optical sensors (such as infra-red sensors) provide high gas
specificity (i.e. little cross-sensitivity), stability and
durability. However, disadvantages include a potentially long lag
period for obtaining NH.sub.3 measurements, expense and low
operating temperature range.
[0149] A limitation of all NH.sub.3 sensors is that levels of
NH.sub.3 cannot be directly linked to levels of NOx, and so
NH.sub.3 sensors have little use in feedback systems for
controlling the combustion process.
Measuring CO in Engine-Out and Exhaust Gas Emissions
[0150] As described herein, it has now been found that the
combustion process of an internal combustion engine, and
particularly in an advanced combustion system, can be monitored,
modelled and controlled by measuring the concentration of CO in
engine-out emissions. Most advantageously, it has now been found
that CO concentration in engine-out emissions from an advanced
combustion engine can be used to: (i) monitor the combustion
process and, therefore, by appropriate adjustments of engine
parameters, to control the combustion process; and (ii) estimate or
determine the concentration of NOx in engine-out emissions.
[0151] With particular regard to point (i) above, by using the
measurement of CO in engine-out emissions to control an advanced
combustion system (for example, using an engine control unit), it
is also possible to: (a) control combustion efficiency; (b)
optimise the combustion process for the minimisation of NOx
production (i.e. the level of NOx in engine-out emissions directly
resulting from the combustion process) using a feedback mechanism;
and (c) optimise the removal of NOx from engine-out emissions, for
example, using feedforward mechanisms to control exhaust
after-treatment systems.
[0152] In further regard to point (ii) above, the use of CO
concentration as an indication of NOx levels, and to adjust the
combustion process to reduce NOx levels, offers significant
advantages over prior art systems, because inter alia the
measurement of CO in engine-out emissions can be more reliable,
less expensive and more accurate than the available means for
measuring the concentration of alternative gases in engine-out
emissions.
Carbon Monoxide (CO) Sensors
[0153] Carbon monoxide (CO) sensors can be made from a range of
materials, including base metals such as titanium, aluminium and
copper. The configurations of CO sensors currently in industrial
use are relatively simple and straightforward, and therefore, the
manufacture of a CO sensor for use in a control system for an
internal combustion engine can be considered to be technologically
unchallenging. CO sensors in common usage are primarily used from
monitoring atmospheric CO in the home or workplace, can be broadly
classified into the following groups: (i) biomimetic sensors; (ii)
semiconductor sensors; and (iii) electrolytic/electrochemical
sensors; and (iv) spectroscopic sensors.
[0154] Biomimetic (or chemical) CO sensors (e.g. U.S. Pat. No.
5,063,164; U.S. Pat. No. 5,618,493) mimic the effect that CO has on
haemoglobin in the human body. In gel cell biomimetic sensors, a
gel-coated disc will change colour (e.g. darken) in the presence of
CO. For example, these devices are usually based on palladium or
iodine salts, which exhibit a colour change on exposure to CO. The
presence of CO gas may then be determined by monitoring the optical
properties of the sensor, for example, by visual inspection, or
using a sensor that detects colour change. Typically, the detector
system includes a housing containing a photon source, which emits
photons in a region of the electromagnetic spectrum that the sensor
absorbs in response to CO exposure, and a photodetector that
measures light at the corresponding wavelength. These detectors can
be inexpensive, particularly with respect to the lower technology
devices, but tend to have slow response times. Although, at
present, they may be more useful for monitoring CO emissions in the
home, some systems may also be suitable or adaptable for measuring
CO in engine-out emissions.
[0155] Semiconductor sensors use an electrically powered sensing
element, which can be monitored by a computer. The sensing element
typically comprises a thin layer of SnO.sub.2 overlaid onto a
ceramic base, which contains electrically conductive wires. The
ceramic base does not conduct electricity, but electrons are able
to travel over the surface of the SnO.sub.2 layer. Absorption of CO
onto the surface of the SnO.sub.2 causes a flow of electrons
between the electrical wires, which is proportional to the
concentration of CO. Generally, this type of sensor operates in
cycles in which CO is first detected and then the CO absorbed in
that stage is burned off the sensor.
[0156] These solid-state semiconductor sensors can be highly
accurate and reliable, making them suitable for monitoring CO
concentration in motor vehicle engine-out emissions and/or exhaust
gas emissions. Although these sensor typically require heating to
operate, this is less of a disadvantage in the context of measuring
CO levels in a hot engine-out emission stream.
[0157] Another type of semiconductor CO sensor contains both a
p-type semiconductor (CuO) and an n-type semiconductor (ZnO) in
physical contact with each other (Japanese Patent Laid-open No.
Sho-62-90529). This type of sensor is reported to have improved
selectivity for CO, but its mechanism of action can mean that its
sensitivity is prone to variability.
[0158] It will be appreciated that the methods of the present
invention benefit from the use of a CO sensor that is highly
selective for CO over other gases in engine-out emissions. Sheikh
et al., (supra) briefly discuss a CO sensor that may have practical
application for monitoring CO levels in combustion emissions, which
contain NOx, CH and O.sub.2. This sensor is based on p-n
hetero-junctions between anatase (n) and rutile (p) with a
TiO.sub.2 base material. In addition, the anatase portion was doped
with CuO (to catalyse CO combustion) and La.sub.2O.sub.3 (to
stabilise the anatase phase of TiO.sub.2).
[0159] Electrolytic (or electrochemical) sensors usually contain
three electrodes (typically made from platinum) that are placed
into an electrolytic solution. CO is adsorbed into the solution and
oxidised at an electrode, generating an electrical current that is
proportion with the CO concentration. In some cases, the
sensitivity of these devices can be affected by the adsorption of
other gases, such as NO and O.sub.2 into the electrolytic solution.
However, generally, these sensors provide a high degree of accuracy
and can be extremely specific for CO.
[0160] Spectroscopic sensors take advantage of the fact that CO
absorbs infrared radiation. Thus, infrared CO sensors measure the
differences between infrared radiation absorption in a test cell
containing a gas to be tested (e.g. an engine-out emission gas) and
a reference cell containing a Known composition of gas. A
disadvantage of these sensors is that, at present, they tend to be
expensive and bulky, requiring a long path length, heated IR
sources, and expensive detectors. However, they are accurate and
sensitive, so that they may offer potential for use in motor
vehicles.
[0161] In summary, the electrolytic and spectroscopic detectors
provide the advantages of rapid response times, high resolution and
high accuracy. However, compared to biomimetic and semi-conductor
sensors, they tend to be expensive and less suitable for domestic
use, such as in a motor vehicle. In the methods and systems
described herein, the CO sensor used for measuring CO concentration
in engine-out emissions and/or exhaust gas emission streams may be
any suitable CO sensor known to the person of skill in the art, and
not limited to the devices described herein. Preferably, however,
the CO sensor used is a semi-conductor sensor, because these
sensors are well known in the art and, presently, offer the most
optimal combination of features for domestic use (such as in a
motor vehicle). As the technology of CO sensors develops, however,
it is anticipated that other types of CO sensors may become more
useful for measuring CO levels within the engine-out emissions and
exhaust gas emissions from an internal combustion engine. This is
particularly the case, because, until now, there has been no demand
for a CO sensor for monitoring engine-out emissions within the
exhaust system of a motor vehicle. However, it is anticipated that
a suitable CO sensor could be readily designed and manufactured by
a skilled person in the art. The methods and systems described
herein encompass the use of any such CO sensing equipment that may
become available.
Engine-Out CO Concentration in Control of Advanced Combustion
[0162] As would perhaps be expected in a finely tuned
chemical-mechanical process, such as advanced combustion; small
changes in any of the important variables could have profound
effects on the result of the process. Advanced combustion is a
finely tuned process, the primary aim of which is to minimise the
production of NOx, thereby to reduce levels of atmospheric
pollution from internal combustion engines.
Advanced Combustion Instability
[0163] As already discussed, present and future emissions
regulations on exhaust gas emissions are extremely challenging to
meet without the use of effective after-treatments, such as CCs. It
is desirable to reduce or eliminate the reliance on these
after-treatments and, therefore, engine manufacturers also seek to
reduce those emissions at source, i.e. by optimising the combustion
process itself. However, as demonstrated in FIG. 1, the pursuit of
low engine-out NOx emissions can challenge the thresholds for
stable combustion.
[0164] FIG. 1 demonstrates the effects on: (i) the coefficient of
variance (COV) of "mean effective pressure" (MEP) within a
combustion cylinder, and (ii) engine-out HC emissions, which result
from pushing the combustion process in the direction of minimal NOx
emissions. In the example shown, minimal NOx emissions for a
particular engine load (in this case: steady state operating
condition of 1200 rpm, 59 Nm on a North American V8 Diesel engine
capable of Tier II Bin 8 emissions levels for a 6000 lb vehicle),
may be achieved by, inter alia, a combination of EGR rate,
injection timing, and boost pressure.
[0165] By way of explanation, the MEP within a combustion cylinder
provides a measure of the percentage load at which the engine is
operating (i.e. it is an indication of the torque that is being
delivered by the engine). Thus, a variation in MEP from one event
to the next is an indication that the engine is misfiring and that
the combustion process is not operating efficiently. Accordingly, a
consistent/reproducible combustion process should lead to a
predictable MEP for each combustion cycle and a low COV. In the
example depicted it can be seen that at low NOx levels the COV of
MEP increases dramatically, which indicates that the combustion
cycle is lacking control and reproducibility under the conditions
of EGR etc. that are necessary to reduce NOx. Similarly, it is
apparent that engine-out HC emissions also increase rapidly under
these conditions of low NOx emissions, which demonstrates that a
highly inefficient combustion process is taking place.
[0166] From this data it appears that even a slight error in EGR
rate, injection timing, boost pressure, or other
combustion-controlling processes could lead to highly inefficient
combustion, which may reduce NOx emissions, but with unacceptable
losses in terms of fuel economy and other relevant factors.
Relationship Between NOx and EGR
[0167] A key challenge with advanced combustion is consistency of
combustion from one combustion cycle to the next. Without this
capability, the engine-out emissions of NOx, and other undesirable
gases, could fluctuate dramatically and severely hamper attempts to
provide consistently low NOx emissions. For example, it is known
that even small errors in air charge (i.e. the mass of air in a
cylinder immediately preceding combustion), or combustion chamber
temperature can lead to drastic differences in engine emissions,
combustion stability, and combustion noise. As described above,
changes in air charge and combustion chamber/fuel temperature can
be achieved by changes in EGR; i.e. the amount of exhaust gas that
is retained in or recirculated to the combustion chamber from one
combustion cycle to the next.
[0168] FIG. 2 graphically demonstrates the relationship between NOx
emissions from an advanced combustion diesel engine as a function
of the amount of EGR for two different engine loads (data from a
North American V8 diesel engine capable of Tier II Bin 8 emissions
in a 6000 lb vehicle). From the graph it can be seen that
relatively large changes in NOx emissions are caused by relatively
small deviations in EGR. For example, an approximate 5% change in
EGR may double, or even triple, the engine-out NOx emissions. This
effect is especially pronounced at high engine loads.
[0169] As discussed above, there is no precise method for
controlling the amount of EGR during advanced combustion, although
improved systems of controlling EGR are in development. Hence, it
can be difficult to control EGR to within 5%. (Even in the case
that EGR is accurately controlled, NOx emissions can still be
variable due to AFR, temperatures, pressures and general
engine-to-engine discrepancies, if not properly controlled).
Therefore, it can be quite challenging to control and maintain
minimal levels of NOx by adjusting EGR exclusively. It is
particularly difficult to achieve consistent combustion parameters
by directly measuring NOx and adjusting EGR in response (so as to
control NOx), because of the highly sensitive relationship between
NOx and EGR, and the relatively low levels of NOx that are
produced. In this regard, at levels of NOx of e.g. 5 ppm, an NOx
sensor may be near to the limit of its resolution and, therefore,
prone to error. However, if a consequential small change in another
combustion parameter, such as EGR, is made in response to an
apparently small error in an NOx measurement, this could cause an
overly large compensation in actual NOx levels, because as
indicated in FIG. 4, a small change in EGR results in a much larger
proportional change in NOx.
[0170] It would be advantageous, therefore, to have a system for
controlling a combustion process in an internal combustion engine
that was not reliant on such highly sensitive and variable
parameters such as measuring NOx. Measurement of CO levels in
engine-out emissions and its use for controlling combustion
provides such a system as described below with reference to FIGS. 5
and 7.
Cetane Variance in Advanced Combustion
[0171] As has been discussed briefly above, it is known that the
type of fuel used in an internal combustion engine can have a
significant impact on the results of the combustion process itself.
Thus, the control of advanced combustion is not merely a matter of
adjusting air charge and other engine controls, but is also
dependent on fuel properties. The most notable of these fuel
properties is cetane number, which represents the fuels capacity to
ignite (i.e. combustibility).
[0172] FIG. 3 demonstrates the variation in combustion--indicated
as a function of NOx emissions--in relation to fuel injection
timing, for three fuels with different cetane numbers (CN). Results
are shown for fuels having CNs of 25 (open circles), 40 (open
triangles), and 53 (filled circles), and the injection timing is
indicated as a function of crankshaft angle in advance of
top-dead-centre (ATDC).
[0173] First, the data depicted clearly demonstrate that changes in
CN can cause significant variations in engine-out NOx emissions;
i.e. at higher CNs the fuel produces higher levels of NOx
emissions. In addition, it was shown from measurements of
in-cylinder pressure and heat release rate, that the more ignitable
fuels (i.e. those having higher CNs), combusted earlier in the
combustion cycle, for an identically timed fuel injection event,
than the less combustible fuels (data not shown). Thus, it is
apparent that fuel CN affects both combustion and engine-out
emissions of NOx.
[0174] Further in this regard, it is notable that, generally, the
CN of commercially available diesel fuel in North America can vary
from 40 to 50, depending on the region and the season. Thus, the
effects observed for changes in CN are relevant to motor vehicles
in common usage. Moreover, where consumers are unable to control
the specific parameters of the fuel type (e.g. CN) that is used, it
is a separate and significant challenge for an engine to detect the
type of fuel present, and to adjust its combustion process
accordingly.
[0175] It would be advantageous, therefore, to have a method and
system for controlling combustion in an internal combustion engine
(and especially in an advanced combustion engine), which is not
sensitive to, or affected by, variations in fuel CN. The methods
described herein provide such as system, as will be discussed.
Relationship Between CO and NOx
[0176] The process of advanced combustion is designed primarily to
improve fuel efficiency and to reduce engine-out NOx levels,
because emissions regulations related to NOx are extremely strict
(limiting allowable NOx levels to virtually zero). The emissions
levels of other undesirable pollutants, such as CH and CO tend,
therefore, to be a secondary issue at the combustion stage, such
that any reduction in CO that is required may be carried out mainly
at the stage of after-treatment. In fact, it seems that the
advanced combustion conditions that favour a reduction in NOx, tend
to favour the production of CO. In this study, the relationship
between engine-out NOx and engine-out CO levels was investigated in
an advanced combustion engine.
[0177] As shown in FIG. 4, it has now been found that there is a
clear relationship between engine-out NOx concentration and
engine-out CO concentration in an advanced combustion engine. In
more detail, at NOx concentrations of between approximately 40 ppm
and 20 ppm, a decrease in NOx concentration is reflected by a
gradual increase in CO concentration. At lower levels of NOx
emissions, for example, between 15 ppm and 5 ppm, a reduction in
NOx concentration is reflected by a more rapid increase in CO
emission levels. The absolute concentration of CO in engine-out
emissions is also dependent on engine load, as indicated by the
data for engine loads of 1200 rpm, 59 Nm (solid line); 1600 rpm, 50
Nm (dashed line); and 1600 rpm, 130 Nm (dotted line). Thus, these
data demonstrate that for a given engine load (set of engine
operating conditions), the concentration of engine-out NOx can be
determined from the concentration of engine-out CO.
[0178] Furthermore, it is notable that whilst the concentration of
NOx varies between approximately 5 ppm and 50 ppm (0.0005% to
0.005%), which range can be extremely difficult to measure
accurately; the concentration of CO tends to vary between 0.1% and
1% (1000 ppm to 10000 ppm). Hence, concentrations of CO in
engine-out emissions may be in the order of 200-fold higher than
concentrations of NOx under the same conditions. Moreover, from the
graph of FIG. 4, it is apparent that a change in NOx concentration
of just 1 ppm (i.e. 0.0001%) can lead to a change in CO
concentration of 0.1% (i.e. 1000-fold greater). Therefore, it
appears that measurement of CO concentration can offer much
improved resolution and accuracy over NOx measurement.
[0179] In summary, using conventional gas sensing equipment, it may
be less technically challenging and less expensive to measure
accurately the concentration of CO in engine-out emissions, rather
than the concentration of NOx. Moreover, as demonstrated, the
measured concentration of CO can provide a direct link to NOx
concentration, so that the concentration of NOx can be accurately
assessed.
[0180] Accordingly, the use of CO sensors to monitor the engine-out
(and exhaust gas) emissions of an internal combustion engine, and
particularly a diesel advanced combustion engine, may provide the
means for a reliable system for optimising advanced combustion to
reduce engine-out NOx levels. In short, CO sensors may be used to:
(i) rapidly monitor the combustion process in real time; (ii)
accurately assess the combustion process in regard to the
production of NOx; and (iii) sensitively adjust the combustion
process to reduce NOx levels.
[0181] Beneficially, in the methods and systems of the invention, a
CO sensor will be arranged so as to measure the concentration of CO
in engine-out emissions from an internal combustion engine. Thus, a
CO sensor will preferably be located in the exhaust system
downstream of one or more combustion cylinders and upstream of any
after-treatments, where used. Preferably, the methods provided
herein obviate or reduce the need for after-treatments to reduced
engine-out levels of NOx. Where one or more after-treatments are
employed, a CO sensor may optionally be located in the exhaust
system downstream of said after-treatments, so as to monitor the
exhaust gas emissions released into the atmosphere.
[0182] Similarly, in the methods and systems described herein,
other means of monitoring exhaust gas streams may be used. For
example, one or more sensors for measuring concentrations of CH,
O.sub.2 and/or NOx may be arranged to measure the concentration of
said gas in the engine-out or exhaust gas emissions. Preferably,
the methods and systems of the invention obviate the need for such
sensors to monitor engine-out emissions. However, it may be
advantageous to measure the concentrations of certain exhaust
gases, such as NOx, in the exhaust gas emissions, e.g. downstream
of any after-treatments, so as to assess the functioning of said
after-treatments, when used. When after-treatments are not used, or
are used sporadically, it may then be advantageous to include one
or more auxiliary gas sensor, such as an NOx sensor, in the exhaust
system to measure engine-out emission levels of the gas of
interest, so as to validate the correct functioning of the
combustion process. It will be appreciated that the concentration
of CO measured downstream of one or more after-treatments may have
no correlation with, or an entirely different correlation with the
concentration of NOx either upstream or downstream of the
after-treatment.
[0183] Preferably, the concentration of CO in the engine-out
emissions is measured continuously, or semi-continuously over the
period during which the engine is running. By semi-continuously, it
is meant that the CO sensor monitors CO concentration as frequently
as its mode or operation and technology allows. For example,
separate measurements of CO concentration may be taken every minute
or more frequently, such as 2, 3, 4, 5, 6, 10, 20, 30, 60 or more
times per minute.
[0184] The concentration of CO that is measured by the one or more
CO sensors is preferably recorded/monitored and processed by a
computer, such as an engine control unit. An engine control unit
(or ECU) preferably obtains a set of data for each CO measurement
recorded and, by way of a feedback system, adjusts one or more of
the engine parameters associated with the control of combustion,
accordingly.
[0185] For example, the data may provide a set of combustion
parameters for optimising the combustion cycle in view of the
measured CO concentration. Preferably, the engine load is also
taken into account for each CO concentration measured. In
particular, the measured concentration of CO is preferably used by
an engine control unit to determine: the optimal EGR, and/or the
optimal injection timing, and/or the optimal fuel quantity to be
injected in each combustion cycle, so as to minimise the production
of engine-out NOx. In addition (or alternatively), the CO
concentration measurements can be used to predict or monitor the
NOx levels in the engine-out emissions. Such NOx measurements can,
for example, be used in a feedforward system to control/adjust any
after-treatments used to reduce optimally the NOx gases in the
engine-out emissions. Most preferably, the engine control unit also
takes account of the relevant engine load for each CO
measurement.
[0186] Any suitable means of correlating the measured concentration
of CO with one or more other engine and/or combustion parameters
may be used. Thus, data in the form of a graph, such as that
depicted in FIG. 4, may provide a "calibration curve", from which a
concentration of NOx can be determined for any particular value of
CO concentration. However, in all cases where one form of data is
used to calculate, predict or select other parameters, it is not
intended that the means of data comparison should be restricted to
a graph or other means of pictorial display. Instead, it is
intended that any means for data correlation; including graphs,
tables and models, and any suitable means by which such data may be
digitally stored and processed (e.g. by way of a mathematical
equation), which provides a means of converting a measurement of CO
concentration into a corresponding NOx concentration or one or more
optimal combustion parameters, is encompassed. Preferably, the data
correlating means has been previously generated, for example, by
carrying out a prior series of tests or models to obtain sample
measurements, to provide an existing set of reference data.
[0187] Preferably, the means of data correlation is electronically
stored by way of a "look-up table", mathematical equation or model
that provides, for example, a preferred fuel injection timing
and/or level of EGR for an inputted CO concentration measurement.
Advantageously, more than one means of data correlation is
available and has been previously generated, for each of a
plurality of different engine loads. For instance, it is beneficial
to produce a series of (standard) data sets of engine-out CO
concentration against NOx concentration and/or preferred (optimal)
EGR, and/or optimal fuel injection timing, and/or optimal fuel
injection quantity for a series of different engine loads.
[0188] More preferably, the means of data correlation is one or
more look-up tables (preferably in electronic form) that correlate
CO concentration to: (i) levels of engine-out NOx at particular
engine loads; and (ii) optimal engine parameters for controlling
engine combustion so as to reduce engine-out NOx levels, where
possible.
[0189] Feedback means for adjusting one or more engine parameters
is preferably provided, and/or feedforward means for adjusting one
or more after-treatments is also provided.
[0190] By way of non-limiting example, in one aspect of the
invention, a series of combustion experiments is performed on a
particular engine type under known engine and combustion
conditions/parameters. The amount of EGR is varied in 1 or 2%
intervals from a minimum level of 0% EGR (or approaching 0%) to a
maximum level of e.g. 75%. It will be appreciated, however, that a
maximum level of EGR of e.g. 60%, 80% or 90% may also be suitable.
For each EGR level the concentration of CO and the concentration of
NOx in the engine-out emissions is measured. These data points can
be used to produce a calibration curve, look-up table, or model to
correlate/link a particular CO emission level to a particular NOx
emission level. Thereafter, by measuring CO concentration under the
test set (or a closely related set) of conditions, the
concentration of NOx in the emissions can be calculated (or
estimated). The experiment can be repeated under different sets of
conditions to obtain further data sets (look-up tables, calibration
curves or models etc.), which can be used to calculate engine-out
NOx concentration from a measured engine-out CO concentration.
[0191] Conditions that may be varied in order to build up a
substantial set of data, include: oil temperature, in-cylinder
pressure, EGR, AFR, injection timing, fuel injection quantity,
engine type, fuel type (e.g. petrol/diesel), engine speed, engine
load and other parameters known to the person of skill in the
art.
[0192] These experiments have the further advantage of providing
information of the engine and/or combustion parameters that are
required in order to produce specific concentrations (or
concentration ranges) of NOx in the engine-out emissions.
[0193] Thereafter, when an engine is in use, by measuring the
concentration of CO in the engine-out emissions, not only can the
level of NOx in the engine-out emissions be calculated, but also, a
set of engine and/or combustion conditions may be calculated (or
determined) to adjust the engine-out concentration of NOx to a
particular (predetermined) concentration or concentration range. An
engine control unit may then use this information to feedback
instructions to control or adjust the engine and/or combustion
parameters, for example, to reduce the concentration of NOx
produced in the combustion process.
[0194] Furthermore, an engine control unit may use the calculated
concentration of NOx to feedforward instructions to adjust or
control the activity of one or more exhaust after-treatments (when
present), to reduce the level of NOx in the engine-out emission,
before the exhaust gases are released into the atmosphere. This
feedforward system is particularly preferred in cases where the
engine control unit is not able to reduce the engine-out emissions
of NOx to below an allowable emissions level, without the use of
one or more after-treatment.
[0195] In addition, it is an option to disable the correlation
means between engine-out CO and engine-out NOx, when the engine is
operating under conditions in which the correlation between CO and
NOx concentration breaks down. For example, this may be appropriate
when the engine is operating outside of the conditions for advanced
combustion. In such circumstances, it is preferred that the engine
control unit of a motor vehicle adjusts the activity of
after-treatments, where used, to ensure that exhaust gas emissions
of NOx do not exceed allowable emissions levels.
Relationship Between CO and Combustion Stability
[0196] As advanced combustion is pushed in the direction of minimal
NOx production, combustion stability can become a significant issue
(FIG. 1). Combustion instability can lead to poor combustion and
therefore, a significant amount of unburned fuel at the end of a
combustion cycle. This unburned fuel is manifested by the presence
of HCs in engine-out emissions. Thus, as combustion becomes less
stable, higher levels of HCs can be measured in engine-out
emissions. The disadvantages of unstable combustion may include a
reduction in fuel economy, a loss in engine power and increased
engine noise.
[0197] FIG. 5 demonstrates that during advanced combustion there is
a correlation between engine-out CO emissions and engine-out HC
emissions. Thus, the level of CO emissions also provides a link to
combustion stability. Moreover, it can be seen from FIG. 5 that
measuring CO emissions provides excellent resolution with respect
to engine-out HC levels. In this regard, a change in HC emission of
100 ppm (0.01%) can result in a change in CO concentration of
approximately 0.1% CO. In other words, once the relationship of CO
emissions to HC emissions is known, measurement of CO concentration
in engine-out emissions provides a highly sensitive measure of
engine-out HC concentration and hence, provides a sensitive measure
of combustion stability and fuel economy.
[0198] Thus, in one aspect of the invention, the measurement of CO
concentration in an engine-out emission is regarded as a indication
of engine combustion efficiency, and this measurement is used in a
feedback system to adjust one or more engine parameters in order to
control the combustion process. In this way, measurement of CO
concentration provides a means for ensuring that combustion
parameters are not selected that would result in unstable
combustion, and moreover, the control of the combustion process
does not rely on a knowledge or even a prediction of the actual
concentration of NOx in the engine-out emissions. A further
advantage of this aspect of the invention is that optimising the
combustion process on the basis of engine-out CO levels has the
benefit that engine-out NOx levels are inherently reduced (see FIG.
4).
[0199] In more detail, the relationship between the amount of EGR
and the level of NOx emissions demonstrated in FIG. 2 provides the
possibility of minimising engine-out NOx levels irrespective of
whether the absolute concentration of NOx in the engine-out
emission is known, by achieving an optimal EGR level. Furthermore,
at a particular engine speed and load there is a maximum EGR that
can be used before combustion stability breaks down, as indicated
in FIG. 1 by the rapidly increasing COV of IMEP with decreasing NOx
concentration (resulting from increasing EGR). However, to
effectively control the combustion process by relying on these
relationships alone, it is necessary to make direct measurements of
NOx and/or EGR and to implement sensitive adjustments to each
parameter.
[0200] In the advantageous method of the invention, the
relationship demonstrated in FIG. 5 and the measurement of CO
concentration in engine-out emissions is used to assess when
combustion stability is at the verge of degrading to the point at
which fuel economy and HC emissions have been unacceptably
compromised, and control the combustion process accordingly.
[0201] As a first step, it is advantageous to provide one or more
calibration curves, look-up tables, or models to correlate/link a
particular CO emission level to a corresponding HC emission level.
Since CO levels in engine-out emissions can be dependent on engine
speed and load, it is advantageous to determine a desirable
"target" CO concentration level (or a corresponding maximum
desirable HC emission level, or COV of MEP) for one or more engine
speed and load conditions. The target CO concentration is the level
of CO which provides the minimum possible NOx concentration while
combustion stability and fuel economy are maintained within
acceptable parameters. These determinations, for example, the
minimum desirable fuel consumption/economy at a particular engine
speed and load, and a particular acceptable level of combustion
stability can be carried out during engine development and testing
(or at any point thereafter). The appropriate parameters may then
be recorded in an engine control/management unit for implementation
when the engine is in use.
[0202] Expressed in another way, it may have been predetermined
that, for a particular engine speed and load, the maximum level of
HC in the engine-out emission before combustion stability and fuel
economy are unsatisfactorily compromised is 400 ppm. By
cross-reference to an appropriate means of data comparison, for
example, a look-up table, calibration curve or equation (such as a
graph similar to that shown in FIG. 5), the concentration of CO in
the engine-out emission that will achieve the predetermined level
of HC can be readily obtained. The corresponding CO concentration
can be considered to be a "target" CO concentration for the same
(and similar) engine conditions.
[0203] Where engine testing to obtain a suitable means of data
comparison (calibration curves, look-up tables and so on) have not
been carried out under the specific engine conditions in use, then
an engine management unit may operate in one (or a combination) of
two ways. For example, the data obtained under the most similar
engine operating conditions (e.g. speed and load) to those in use
may be taken as an approximation of the combustion conditions in
use. In another method, a regression analysis or algorithm may be
used to adjust the data obtained under the closest engine operating
conditions in order to obtain a modified set of data that is an
estimate of the combustion parameters in use. In this way, an
engine management unit and the methods of the invention can be used
under all operating conditions of engine speed and load, despite
that a means of data comparison, obtained for example, during
engine development and testing, is not available for every possible
set of engine operating conditions. These methods of approximating
and/or estimating the engine combustion parameters in use apply
similarly to all of the methods and systems of the invention, which
involve the use of a means of data comparison.
[0204] Thus, in accordance with the invention, a CO sensor is used
to measure the engine-out CO concentration. The measured
concentration of CO is then compared to the target CO
concentration. If the measured CO concentration is lower than the
target CO concentration, EGR is increased (to increase HC
concentration and inherently reduce NOx concentration); whereas if
the measured CO concentration is higher than the target CO
concentration, combustion is unstable (or inefficient) and EGR is
reduced. By performing these steps advanced combustion can be
controlled by measuring engine-out CO concentration to minimise NOx
concentration within acceptable levels of, for example, combustion
stability, engine-out HC concentration and fuel economy.
[0205] This aspect of the invention is further illustrated by
reference to FIG. 7, described below.
[0206] In summary, CO is used to maintain combustion efficiency,
allowing an engine control unit (ECU) or engine management unit to
add maximum EGR without causing misfire, and consequently minimal
NOx emissions are produced for the engine operating conditions in
use. In fact, using this system it may not even be necessary to
estimate (or measure) NOx concentration in engine-out emissions (in
order to know that NOx levels are optimally reduced), but just to
control EGR based on CO. Of course, estimating NOx based on CO
levels may still be useful, as discussed below and elsewhere
herein.
[0207] In an alternative method, the correlation between engine-out
CO concentration and engine-out NOx concentration (FIG. 4) may be
used to achieve a particular desirable NOx concentration by
adjusting EGR (further described with reference to FIG. 9).
[0208] In this aspect, a series of combustion experiments is
performed on a particular engine type under known engine and
combustion conditions/parameters. The amount of EGR is varied in 1
or 2% intervals from a minimum level of 0% EGR (or approaching 0%)
to a maximum level of e.g. 75%. It will be appreciated, however,
that a maximum level of EGR of e.g. 60%, 80% or 90% may also be
suitable. For each EGR level the concentration of CO and the
concentration of NOx in the engine-out emissions is measured. These
data points can be used to produce a calibration curve, look-up
table, or model to correlate/link a particular CO emission level to
a particular NOx emission level. Thereafter, by measuring CO
concentration under the test set (or a closely related set) of
conditions, the concentration of NOx in the emissions can be
calculated (or estimated). The experiment can be repeated under
different sets of conditions to obtain further data sets (look-up
tables, calibration curves or models etc.), which can be used to
calculate engine-out NOx concentration from a measured engine-out
CO concentration.
[0209] Conditions that may be varied in order to build up a
substantial set of data, include: oil temperature, in-cylinder
pressure, EGR, AFR, injection timing, fuel injection quantity,
engine type, fuel type (e.g. petrol/diesel), engine speed, engine
load and other parameters known to the person of skill in the
art.
[0210] These experiments have the further advantage of providing
information of the engine and/or combustion parameters that are
required in order to produce specific concentrations (or
concentration ranges) of NOx in the engine-out emissions.
[0211] Thereafter, when an engine is in use, by measuring the
concentration of CO in the engine-out emissions, not only can the
level of NOx in the engine-out emissions be calculated, but also, a
set of engine and/or combustion conditions may be calculated (or
determined) to adjust the engine-out concentration of NOx to a
particular (predetermined) concentration or concentration range. An
engine control unit may then use this information to feedback
instructions to control or adjust the engine and/or combustion
parameters to correspondingly adjust the concentration of NOx
produced in the combustion process. For example, on the basis of
the measured CO concentration the amount of EGR may be increased to
achieve a particular (lower) level of NOx in the engine-out
emission.
Effect of Cetane Number (CN) on the Relationship Between CO and
NOx
[0212] In FIG. 3 it is demonstrated that fuel CN can have a
significant effect on the advanced combustion process. In
particular, it has been demonstrated that, the higher the CN, the
earlier the fuel--air mixture combusts and the higher the
concentration of NOx in the engine-out emission. For this reason,
and the fact that commercially available fuel can vary considerably
in CN, it is widely believed that a method for CN detection/fuel
analysis is necessary to adjust combustion parameters (e.g. fuel
injection timing and fuel quantity), in order to maintain
acceptable engine emissions and performance.
[0213] However, data from a recent study by the Oak Ridge National
Laboratory has indicated that engine-out NOx emissions share a
common relationship to engine-out CO emissions regardless of CN
(Table 1). In brief, these data appear to demonstrate that for a
constant engine-out CO emission level of approximately 2500 ppm,
the engine-out NOx emissions will range only from 0.5 ppm to 1.7
ppm, for a range of different fuels having CNs of 34 to 76.
TABLE-US-00001 TABLE 1 Correlation between CN, engine-out CO
emissions and engine-out NOx emissions in advanced combustion
(Szybist & Bunting, "Cetane Number and Engine Speed Effects on
Diesel HCCI Performance and Emissions," SAE Technical Paper Series,
2005-01-3723). Cetane Number CO NOx CA50 Fuel (CN) (ppm) (ppm) (CA
deg.) 1 76 2500 0.5 354 2 62 2500 1 358 3 48 2500 1 361 4 34 2500
1.7 372 5 19 2500 -- --
[0214] These data suggest that if engine-out CO emissions are
regulated, for example, using a feedback system controlled by an
engine control unit, it will be possible to control engine-out NOx
emission levels irrespective of the CN of the fuel. Significantly,
such a system will obviate the need to know or estimate the CN of
the fuel used, and thus, provides a further advantage of the
methods and systems described herein.
Relationship Between CO and AFR
[0215] Advanced combustion is typically characterised by the use of
high levels of EGR in comparison to other forms of internal
combustion; although some work has been carried out on HCCI without
EGR. EGR also effects the AFR in the combustion
chamber/cylinder.
[0216] In the present investigations it has also now been found
that the level of CO in engine-out emissions correlates to AFR,
particularly as engine load increases. This correlation is
demonstrated in FIG. 6, which provides values of AFR against CO
emissions for a North American V8 Diesel engine capable of Tier II
Bin 8 emissions (in a 6000 lb vehicle), under three different
engine loads. The engine was rated at 720 Nm. In FIG. 6, the AFR is
reported in units of AFRC, wherein the "C" stands for "carbon
balance". Several methods for measuring AFR are known to the
skilled person in the art, and the method employed in this example
used several emission analysers in a laboratory to count molecules
of carbon ("C", for example, CO, CH, CO.sub.2, etc.). An
alternative method is AFRO, which measures molecules of oxygen
(wherein "O" stands for "oxygen balance").
[0217] In this regard, at low loads of 1200 rpm, 59 Nm (filled
squares, left-hand vertical axis) and 1600 rpm, 173 Nm (filled
diamonds, left-hand vertical axis), it was found that the
engine-out CO emissions were not only a function of AFR, but also
of injection timing, injection pressure and EGR. Thus, the plots of
AFR against CO emissions appear relatively more scattered,
depending on these other variables. However, at 1600 rpm, 297 Nm
(filled triangles, right-hand vertical axis), which is still less
than 50% load, a strong correlation between engine-out CO emissions
and AFR emerged, regardless of the other combustion variables.
[0218] Accordingly, the methods and systems described herein
provide a means for determining the AFR in a combustion cylinder,
and particularly during an advanced combustion process, on the
basis of the measured concentration of CO in the corresponding
engine-out emissions. The measured concentration of CO in the
engine-out emissions may be correlated to the corresponding AFR
using any appropriate means, as described above. Briefly, a
calibration curve, look-up table, or mathematical
algorithm/equation may be used to convert a CO concentration into
an AFR value. Preferably, a plurality of "standard" sets of data is
provided for correlating CO concentration to AFR for any of a
number of different engine and/or combustion parameters, such as
engine speed and load. The means of data correlation is preferably
stored electronically, for example, so that an engine control unit
can process the CO data measured, in real time, to determine AFR
for the measured CO concentration.
[0219] Still more preferably, the methods and systems include a
feedback means for adjusting combustion parameters, in view of the
measured CO concentration and the corresponding AFR, to optimise
the combustion process for: reduction of NOx emissions; reduction
in particulate matter; injector corrections; and improvement of
combustion efficiency.
Feedback and Feedforward Control of Advanced Combustion
[0220] As has been discussed above in detail, the methods and
systems described herein provide important information on the
combustion process in an internal combustion engine. This
information, which is derived from the concentration of CO in
engine-out emissions, can be used advantageously to: improve
combustion efficiency (by feedback control); reduce the production
of undesirable gases, such as NOx (by feedback control); and
effectively remove undesirable gases, such as NOx, using
after-treatments (by feedforward control).
[0221] FIGS. 7 to 10 exemplify means by which CO concentration may
be used by an engine control unit to adjust: EGR; fuel injection
timing; engine-out NOx emissions; and AFR, respectively, in order
to optimise advanced combustion.
[0222] Turning to FIG. 7, for a given engine speed and load, a
"target" engine-out NOx emission also equates to a "target"
engine-out CO emission (as discussed hereinbefore, see e.g. FIG.
4). Using the feedback from an appropriately arranged CO emissions
sensor, the error/difference between the target CO concentration
and the actual (measured) CO concentration can feed a Proportional,
Integral, Derivative [PID: a sum of 3 different corrections
depending on error (Proportional), cumulative error (Integral), and
change in error (Derivative)] for an appropriate modification to
EGR. For example, where the measured CO concentration is lower than
the target CO, the EGR would be increased to raise CO production.
As has already been shown with reference to FIG. 4, the production
of NOx in advanced combustion is reduced as the level of CO is
raised. Therefore, it is not necessary to know or even estimate the
concentration of NOx in the engine-out emissions at a specific
point in time; it is sufficient to control EGR so as to achieve a
predetermined CO engine-out emission level, knowing that the
predetermined CO level will correspond to a desirable (optimal or
target) NOx concentration.
[0223] It should be noted that EGR and advanced combustion may only
apply to a portion of a given engine speed and load range; i.e. at
very high engine loads it is known that present advanced combustion
processes break down. Therefore, the P, I and D terms should
preferably also depend on engine speed and load.
[0224] Likewise, advanced combustion may only be achieved
effectively for a specific range of coolant and air temperatures.
Therefore, the control system described may beneficially be
disabled when the engine parameters are outside of the appropriate
range.
[0225] The control loop does not have to be a PID. For instance,
there could be a number of ways to modify EGR based on CO
emissions. Furthermore, additional parameters may also be included
for enabling/disabling the control, such as engine run time, oil
temperature, or atmospheric pressure. Suitable additional
parameters will be known to the person of skill in the art.
[0226] In addition, or in the alternative to the feedback system
described above, the engine control unit may process CO
concentration data for feedforward EGR demand. Thus, when the
feedback system is disabled, e.g. due to parameters that fall
outside of those required for advanced combustion, the algorithm of
the engine control unit may preferably resort to a feedforward EGR
demand, which may be based, for example, on engine speed, engine
load, and various temperatures, pressures and other parameters,
such as those mentioned above.
[0227] FIG. 8 is a schematic representation of an algorithm for
adjusting fuel injection timing on the basis of a measured
engine-out CO concentration.
[0228] If the CO emissions are high, there is an increased
likelihood of poor combustion efficiency or misfire. One way of
avoiding misfire is to advance fuel injection timing. Reducing EGR
can also help avoid misfire, however, control of EGR has been
discussed with respect to FIG. 7. Thus, to avoid misfire, a maximum
engine-out CO emission level is identified (based, for example, on
control measurements or modelling), preferably, for a range of
engine speeds and loads. The maximum allowable CO level can be
subtracted from the base target CO level, to obtain the maximum
allowable "error" (or difference). Depending on the measured CO
level, the exemplified algorithm is designed to advance injection
timing in the event that the maximum allowed error is breached.
[0229] In a similar manner to the EGR control algorithm (FIG. 7),
the engine-out CO emission measured by a CO sensor is fed into a
PID or other control system. As before, the control terms are
preferably a function of engine speed and load. Also, the control
system may preferably be enabled or disabled according to engine
speed and load, and various other combustion parameters, such as
temperature and pressure (as discussed above).
[0230] Turning to FIG. 9, a schematic representation of an
algorithm for estimating engine-out NOx levels is depicted.
[0231] It is important to know the engine-out NOx emissions from an
internal combustion engine. Using the methods and systems described
herein, the NOx emission levels can be calculated on the basis of
the engine-out CO emissions. Various routes by which this
calculation can be achieved may be apparent to the person of skill
in the art. One such system is described herein. For example, first
a "base" (or target) NOx emission level may be identified;
preferably, for each engine speed and load. This base NOx emission
level can then be correlated to a target CO emission level; and the
"error" in the NOx emission level can be determined from the
corresponding error between the target CO emission level and the
measured (or actual) CO emission level. It may also be possible to
represent NOx emissions entirely from correlated CO emissions.
[0232] As before, the algorithm may be disabled under certain
conditions, for example, in cold conditions, or high altitude, and
for specific engine speeds and loads where advanced combustion is
not possible. Under these conditions where the algorithm cannot be
operated, the engine-out NOx emissions may instead be estimated
using open loop tables.
[0233] For after-treatment models it is also important to know the
engine-out NOx emissions. Therefore, a similar algorithm to that in
FIG. 9 may be used to provide a feedforward signal to control
after-treatments. Typically, however, there would be one such
algorithm, such as that exemplified in FIG. 9 which estimates
engine out NOx, and those results would be used globally, for
example: to adjust EGR in a feedback mechanism; and also in a
feedforward after-treatment algorithm to control one or more
exhaust gas after treatments. Such a feedforward system could, for
example, up- or down-regulate an NOx-removing after-treatment on
the basis of the calculated level of NOx in the engine-out
emissions. In this way, the after-treatment could be operated
optimally accordingly to the amount of NOx present, so that fuel
efficiency is optimised and waste products/unnecessary exhaust
emissions are minimised.
[0234] FIG. 10 depicts an algorithm for calculating AFR on the
basis of engine-out CO concentration.
[0235] As discussed in regard to FIG. 6, AFR can be calculated
directly from engine-out CO emissions. Preferably, the engine speed
and load is taken into consideration when determining the AFR based
on CO concentration. Once the AFR has been calculated, feedback
systems may be employed to adjust the AFR, according to preferred
parameters.
[0236] It will be appreciated that in any of the feedback systems
described herein, the feedback signal from an engine control unit
may adjust the engine and/or combustion parameters associated with
one or more engine cylinder. Preferably, the feedback signal
adjusts the engine and/or combustion parameters associated with
each of the engine cylinders.
[0237] As before, there may be engine speeds and loads where the
relationship between CO emission levels and AFR breaks down, for
example, outside of the advanced combustion zone. Similarly, there
may be low temperatures and atmospheric pressures under which the
correlation between CO concentration and AFR is not appropriate.
Under these conditions, preferably the algorithm is disabled, and
in those circumstances, the levels of AFR are advantageously based
solely on mass air flow (MAF-measured with an MAR sensor) and
commanded fuel levels.
* * * * *
References