Applied Energy 99 (2012) 534–544

Contents lists available at SciVerse ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Gasoline engine exhaust gas recirculation – A review

Haiqiao Wei, Tianyu Zhu, Gequn Shu ⇑, Linlin Tan, Yuesen Wang
State Key Laboratory of Engines, Tianjin University, 92 Weijin Road, Tianjin 300072, China

a r t i c l e i n f o a b s t r a c t

Article history: EGR technique, as one of effective measures to reduce NOX formation, was firstly adopted in diesel
Received 25 November 2011 engines. But with the growing energy and environment problems, EGR has been commonly used also
Received in revised form 24 March 2012 in gasoline engines together with other advanced techniques. Since there hasn’t been any comprehensive
Accepted 6 May 2012
review on gasoline engine using exhaust gas recirculation, this paper is made by its motivation. Recircu-
Available online 10 June 2012
lating exhaust gas on gasoline engines is employed primarily to reduce throttling loss at part load range
in order to reduce fuel consumption, and secondarily, to reduce NOX emission levels. In addition, EGR can
Keywords:
replace fuel enrichment in gasoline engine to inhibit knock. The aim of this paper is to review the influ-
Gasoline engine
EGR
ence of EGR on the performance and emission of gasoline engine as well as to compare the application of
NOX EGR on GDI engines and on PFI engines. Furthermore, a detailed analysis of comparison between cooled
Stratified EGR EGR and hot EGR, the effect of EGR on knock suppression and the implementation of EGR on turbocharged
Knock gasoline engine are introduced. From the deep analysis, EGR can improve fuel economy, reduce NOX
emission and inhibit the tendency of engine towards knock. However, the maximum possible EGR rate
is limited by high cyclic variations, misfire, the decrease of total efficiency and the increase of HC emis-
sion. This problem can be solved by stratified EGR system. The combination of EGR technique with super-
charged direct injection engine is the trend of today’s gasoline engine.
Ó 2012 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
2. EGR system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
3. EGR vs. NOX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
4. EGR in GDI engines and in PFI engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
5. Stratified EGR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
6. Comparison between cooled EGR and hot EGR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
7. Knock suppression using cooled EGR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
8. Implementations of EGR for turbocharged gasoline engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
9. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543

1. Introduction ious countries in the world have developed more stringent

regulations on vehicle emission and fuel economy, which drive
In engine evolution, EGR technique was firstly adopted in diesel gasoline engine towards downsizing direction [1–3]. Engines with
engines in order to limit thermal NOX formation rate by reducing turbo-charged spark ignition are becoming increasingly popular in
combustion chamber temperature thanks to the dilution of fresh the world market due to their compactness and high power
charge with a certain amount of exhaust gas recycled at engine in- density. However, due to the high power density of turbo-charged
take. But with the growing issues of energy and environment, var- engine, knock combustion and high temperature of exhaust gas at
high loads constitute problems [4]. The increasing temperature
may also cause the problem of NOX emission increase. Accordingly,
⇑ Corresponding author. Tel./fax: +86 022 27891285. EGR is commonly used also in gasoline engines together with other
E-mail address: tjusgq@163.com (G. Shu). advanced techniques [5,6].

0306-2619/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.apenergy.2012.05.011
 H. Wei et al. / Applied Energy 99 (2012) 534–544 535

Nomenclature

AFR air fuel ratio ISHC indicated specific hydrocarbons

BMEP break mean effective pressure ISNOX indicated specific oxides of nitrogen
CA crank angle LPL low pressure loop
CA10 crank angle where 10% of MFB MFB mass fraction burned
CA90 crank angle where 90% of MFB N2 nitrogen
CO carbon monoxide NOX oxides of nitrogen
CO2 carbon dioxide NVO negative valve overlap
COV coefficient of variations O2 oxygen
EGR exhaust gas recirculation PFI port fuel injection
GDI gasoline direct injection SI spark ignition
HC hydrocarbon TDC top dead center
H2O water TWC three-way catalytic converter
HPL high pressure loop VGT variable geometry turbine
IMEP indicated mean effective pressure WOT wide open throttle
k air excess ratio

Unlike diesel engines, k can be actually used in gasoline engines bility issue, thus two implementation methods are reviewed in this
to reduce NOX emission effectively. Therefore, recirculating ex- paper, and both the advantages and disadvantages of these
haust gas on gasoline engines is used primarily to reduce throttling methods are compared. Furthermore, the problem of turbo match-
loss at part load range, thus reduce fuel consumption, and second- ing when introducing EGR is described briefly.
arily, to reduce NOX emission levels [4]. In order to keep the same
torque and power output after introducing EGR in gasoline engine,
further opening of engine throttle is necessary to increase the 2. EGR system
trapped charge density, which can reduce pumping loss and to in-
crease fuel economy compared with that when no EGR is used. On Cylinder charge dilution with exhaust gas can be classified into
the other hand, gasoline engine works with a stoichiometric air– internal EGR and external EGR. With external EGR, exhaust gas is
fuel mixture in order to meet the need of TWC, so the O2 concen- taken from the exhaust port and supplied into the inlet port. Inter-
tration in the exhaust is very low, which will cause a corresponding nal EGR is achieved by increasing NVO during exhaust stroke,
lower NOX emission. which requires an improved cam that can rapidly switch cam pro-
For a GDI engine that is designed to work with lean or even ul- files to achieve any variable valve timing, otherwise it’s impossible
tra-lean mixtures, it is unfortunately true that a conventional TWC to independently and effectively control EGR ratio. This greatly
cannot be used to remove NOX, therefore other techniques for in- limits the application of internal EGR. As a result, external EGR
cylinder NOX reduction or exhaust after-treatment must be em- has become widely used on today’s automobile engines. External
ployed. EGR is an effective measure. Compared with PFI engine, EGR has a relatively low cost. It only needs to use dedicated EGR
EGR rates as high as possible are required by GDI engine to reduce control valve, which can control EGR rate effectively under all work
overall NOX emission levels. EGR is considered as a principle meth- conditions of engine [9–12]. Only external EGR is discussed in this
od to reduce NOX emission of GDI engine [7]. The application of paper.
EGR in GDI engine and in PFI engine is discussed in detail in this External EGR system consists of EGR pipe, EGR valve and EGR
paper. cooler (cooled EGR). Exhaust goes through EGR valve and EGR cool-
When hot EGR is used in engine, the exhaust can be used to heat er, and then enter intake manifold. Constant coolant goes through
the intake, so combustion quality and thermal efficiency are im- EGR cooler. EGR valve can be adjusted to get various EGR rates.
proved. However, cooled EGR increases inlet density and thus in- Pipe material is stainless steel to avoid transfer of engine vibration
creases volumetric efficiency of the engine. At the same time, the to exhaust system and then to measuring instruments [9].
reduced temperature can further reduce NOX emission, but the cy-
cle-by-cycle variations will be increased compared with hot EGR.
Cooled EGR is a key technology enabling downsized SI engines
and providing a means to meet the market requirement of ‘‘doing
more with less’’ [8]. The comparison of the advantages and disad-
vantages between hot EGR and cooled EGR is given in this paper. As
it is well known, the crucial point of gasoline engine behavior is the
dramatic fall of energy conversion efficiency in part load operation.
So, much effort has been made to decrease pumping loss and to
optimize engine thermodynamic efficiency. However, knock risks
impose strong limitations to the performance and efficiency of gas-
oline engine. So, in full load operation, fuel enrichment was usually
adopted to suppress knock. EGR can reduce the combustion pres-
sure and inhibit knock as well, and additional benefit is fuel econ-
omy as well as the decrease of HC and CO emission.
Turbocharged gasoline engine is a typical representative of
downsized engines. EGR technique combining with supercharged
direct injection engine is the trend of today’s gasoline engine.
Using EGR in turbocharged gasoline engine may involve compati- Fig. 1. Centralized and decentralized EGR [7].
536 H. Wei et al. / Applied Energy 99 (2012) 534–544
For multi-cylinder engine, important parameters in designing
EGR systems are good homogeneous EGR distribution to each cyl-
inder and good dynamic response. The recirculated exhaust gas can
be supplied to the engines either centralized or decentralized as
schematically shown in Fig. 1.
For centralized exhaust gas recirculation systems, EGR valve is
far away from inlet valves, normally at the entrance to intake man-
ifold, where collector provides a good mixture of exhaust gas and
fresh air and thus optimal distribution to each cylinder. The decen-
tralized EGR system directly supplies exhaust gas to inlet valves.
Two different applications are possible: one application is one valve
for each cylinder, and the application is using a multiple valve. With
the two applications, it is ensured that, with closed EGR valve, cross
flow is impossible between intake manifold runners. The benefits of
decentralized EGR system are optimized EGR distribution and good
dynamic response. However, the location and orientation of EGR
port and valve may influence in-cylinder EGR distribution signifi-
cantly, and thus influence fuel stratification by producing an extra
flow of recirculated exhaust gas within combustion chamber. This
influence must be carefully evaluated when using this application.
In addition, the sealing of EGR valve or port and assembly
constraints in the vicinity of cylinder cover is also important factors
that constrain the application of this EGR strategy [13].
When EGR is applied, engine intake consists of fresh air and
recycled exhaust gas. EGR (%) usually represents the percentage
of the recirculated exhaust gas. The percentage of exhaust gas
recirculation is defined as the percentage of recirculated exhaust
in total intake mixture [14]. Where [mi = ma + mf + mEGR] is the
mass of total intake mixture and [mEGR] is the mass of the EGR.
EGRð%Þ ¼ ðmEGR =mi  100Þ ð1Þ
The calculation of recirculation degree was defined by relating
the amount of CO2 in intake manifold with that in exhaust
pipe, through Eq. (2). Where the CO2 concentrations used in the
expression are such that [CO2]Ambient was measured in the environ-
ment, [CO2]Intake gas was checked in the intake manifold and
[CO2]Exhaust gas was measured in the exhaust gases [15,16].
½CO2 Intake gas  ½CO2 Ambient
EGRð%Þ ¼ ð2Þ
½CO2 Exhaust gas  ½CO2 Ambient
3. EGR vs. NOX
As it is well known, NOX is generated under the condition of
high-temperature and oxygen –rich environment. The NOX emis-
sion of internal combustion engine is mainly NO, but it will be oxi-
dized into NO2 quickly after entering air. The formation of NO is
increased dramatically in exponential function with temperature
increase. When temperature is below 1800 K, the rate of NO forma-
tion is very low, but it will reach a very high speed when the tem-
perature reaches 2000 K. It can be generally considered that for
every 100 K increase of temperature, NO formation rate will nearly
doubles. The formation mechanism of NO is different from that of
HC and CO. No is not the result of incomplete combustion of mix-
ture. Its generation relates with the proliferation of mixed combus-
tion, concentration distribution of flame and heat transfer, thus
reaction mechanism is very complex. Basically, nitrogen hardly ex-
ists in gasoline. The formation of NOX originates from thermal reac-
tion of N2 and O2 in the air in high temperature combustion. NO is
formed inside combustion chamber in post-flame combustion pro-
cess in high temperature region. NO formation and decomposition
inside combustion chamber can be described by extended Zeldob-
ich Mechanism [15,17,18]. The principal reactions near stoichiom-
etric fuel–air mixture governing NO formation from molecular
nitrogen are:
N2 þ O () NO þ N
N þ O2 () NO þ O ð3Þ
N þ OH () NO þ H
The initial rate controlling NO formation can be described with
Eq. (4). In the expression, [NO] denotes molar concentration, and
[O2]e and [N2]e denote equilibrium concentrations [19]. The sensi-
tivity of NO formation rate to temperature and oxygen concentra-
tion is evident from this equation. Hence, in order to reduce NOX
formation inside combustion chamber, the temperature and the
oxygen concentration in combustion chamber need to be reduced.
!  
d½NO 6  1016 69; 096
¼ exp ½O2 0:5
e ½N 2 e mol s=cm
3
ð4Þ
dt T 0:5 T
In gasoline engine, EGR is the most principal technique in
reducing NOX emission. The main constituents of EGR are N2,
H2O, O2 and CO2. It is well known that CO2 exerts three effects
when being introduced into combustion process. They are:
(1) Thermal effect: Exhaust gas consists of gases of two-atom and
three-atom, and the heat capacity of three-atom gas
increases faster in combustion process. Due to the increase
in heat capacity of the oxidizer, flame temperature is
reduced;
(2) Dilution effect, which results from the reduction of oxygen
concentration in the main stream of the oxidizer and from
the reduction of reactive species in combustion process,
which, in turns, reduces their collision frequency;
(3) Chemical effect, since CO2 is an active species and thus partic-
ipates chemically in the combustion process;
These three effects have been studied in the literature [20–22].
Engine tests have demonstrated that NOX is greatly suppressed
when the O2 concentration in combustion chamber is reduced.
Fig. 2 shows that the reduction in O2 concentration in the cylinder
suppresses NOX emission. The figure shows the combustion in two
GDI fuel sprays, one with and the other without EGR. In GDI engine,
when using EGR, some O2 in the cylinder is replaced with exhaust
gas, and thus local O2 concentration in the cylinder becomes lower.
With the local O2 concentration reduced, a given amount of fuel
will have to diffuse over a wider area before encountering suffi-
cient O2 for a stoichiometric mixture to be formed. Now, for a given
amount of fuel, this larger area contains not only stoichiometric
mixture but also an additional quantity of CO2, H2O and N2. These
additional gases absorb the energy released by the combustion,
leading to lower flame temperature and then lower NOX generation
[9,23].
4. EGR in GDI engines and in PFI engines
PFI engines work under stoichiometric condition, and EGR is
employed primarily to reduce throttling loss at part load range,
thus to reduce fuel consumption, and secondarily, to reduce NOX
emission levels. In order to keep the same torque and power
Fig. 2. Increase in volume occupied by spray flame with use of EGR.
 H. Wei et al. / Applied Energy 99 (2012) 534–544
output after introducing EGR in gasoline engine, further opening of
engine throttle is necessary to raise the trapped charge density.
This can reduce pumping loss and improvement of fuel economy
compared with that when no EGR is used. The improvement in fuel
consumption with increasing EGR is due to three factors: first, re-
duced pumping work, as EGR increases at constant brake load (fuel
and air flows remain almost constant, hence, intake pressure in-
creases); second, reduced loss of heat transferred to cylinder wall
since burned gas temperature is decreased significantly; third, a
reduction in the degree of dissociation in the high temperature
burned gases, which allows more fuel’s chemical energy to be con-
verted to sensible energy near TDC [24].
It is worth to mention that when PFI engine works at full load,
the throttle has reached the WOT condition. In this case, the throt-
tle cannot be opened more to increase intake density. Therefore,
boosting intake pressure is necessary to gain the same level of tor-
que and power output. If supercharge is not used to increase intake
density, the power loss will increase with the increase of EGR ratio
at full load [4]. Furthermore, in a conventional PFI engine, EGR
trends to influence combustion stability and investigations have
confirmed that both ignition delay and combustion period are
extended with EGR due to the associated decrease in laminar flame
speed. Consequently, it may be appropriate to use EGR in combina-
tion with other techniques [25,26].
In general, a larger NOX reduction can be realized with EGR in
GDI engines than that in either PFI engines or diesel engines. For
GDI engines, the available fuel–air mixing time is comparatively
longer than that of diesel engines, and, as a result, EGR role is more
effective and NOX emission can be further reduced. In a GDI engine,
since there is no throttling effect, the engine works with lean or
even ultra-lean mixtures and the in-cylinder AFR is higher, and
the introduction of exhaust gas replaces part of fresh air directly.
For a given torque and power output, the amount of fuel that
engine supplies must be constant. Therefore, the in-cylinder AFR
decreases with the increase of exhaust gas, which can reduce NOX
emission effectively. Although in GDI engine, lean-NOX after-treat-
ment technologies can be used to reduce NOX emission, EGR was
considered the only feasible way to reduce NOX [7]. In addition ,
EGR ratio as high as possible are required to keep GDI engine work-
ing under stoichiometric condition, so that TWC can work normally,
and then we can use it to reduce HC and CO emission. Table 1 shows
the basic differences between GDI engine EGR system and PFI
engine system for different applications.
The EGR application ranges of PFI engines and of GDI engines
are different. Fig. 3 shows the ranges of conventional PFI engines
in comparison with the various working modes of GDI engines. In
Fig. 3, in homogeneous mode of GDI engines, the use of recircu-
lated exhaust gas corresponds to that of conventional PFI engines.
Due to the higher EGR tolerance of the engine, EGR rates are also
higher in the case of homogeneous work. In stratified mode, EGR
rates comparable to those of diesel engines are attainable and
much higher than that of PFI engines. For a GDI engine using
Table 1
The basic differences between EGR systems for the different applications.
Target
Max. EGR rate
Max. exhaust temperature in the operating range
EGR cooling
Other requirements
537
charge stratification, the mixture near spark gap is ideally either
stoichiometric or slightly richer. Furthermore, mixture prepara-
tion can be improved due to EGR heating effects. Robust combus-
tion is therefore possible with a much higher level of EGR than
that with homogeneous combustion. However, there is an associ-
ated compromise between a significant NOX reduction and a
simultaneous degradation in both HC emission and in fuel
consumption due to the degradation of combustion quality [13].
Moreover, in a good EGR system, dynamic response is mandatory
to avoid drivability problems in transition from one work mode
to the other.
When EGR is used in GDI engine, providing appropriate amount
of EGR has always been a design challenge, since a relative higher
EGR mass flow rate must be metered and the flow must be distrib-
uted uniformly to individual cylinders under a much lower
pressure difference with that of traditional PFI engines [27]. The
flow of so much EGR may require a moderate level of intake
vacuum, which will cause great pumping loss, while GDI engines
is supposed to reduce such loss substantially.
5. Stratified EGR
Whether in PFI engine or in GDI engine, homogeneous EGR
system reduces laminar flame speed, which will cause reduce of
burning speed, HC emission increase, cycle-by-cycle variations
aggravation, and steady-state combustion being difficult to be
achieved, and even causing fire [14,28]. High-diluted stratified
EGR charge to separate air/fuel mixture in intake and compression
strokes can be a good practice to generally overcome the above dif-
ficulties. The stratified exhaust gas recirculation is characterized by
separating EGR air and fresh air in combustion chamber. Due to a
minimized exhaust gas concentration at spark plug region, flame
propagation is improved compared with homogeneous EGR, EGR
compatibility is increased. However, the flow structure in combus-
tion chamber is extremely complex, so to realize complete separa-
tion of air and exhaust gas is very difficult. Another challenge in
this system is how to realize stratification in intake stroke and
how to maintain the stratification in compression stroke prior
ignition.
Different possibilities of exhaust gas stratification are shown in
Fig. 4, including radial stratification, lateral stratification and axial
stratification. Lateral EGR stratification, consisting of air/fuel mix-
ture at the side of cylinder intake and of only EGR at the exhaust
side, may separate them well. However, this method requires
strong tumble intensity to maintain vertical flow momentum.
When the piston moves upwards to TDC, the tumble flow is easily
destroyed by turbulent and/or squeeze flow. Axial EGR stratifica-
tion divides the cylinder into top air zone and bottom EGR zone.
The injected fuel can penetrate into the pure air zone, mix with
the air, and then be ignited and burned at this zone. In this method,
a vertical flow, such as tumble flow or squeeze flow, may make the
EGR System for GDI Engines EGR System for PFI Engines
First, reduction in nitrogen oxides First, reduction in fuel consumption
Second, reduction in fuel consumption Second, reduction in nitrogen oxides
 50%, Stratified mode 25%
 25%, Homogeneous mode
 450 °C 650 °C
 (650 °C, homogeneous mode)
Under discussion Required
 Good dynamics Reduction in power losses at high EGR ratios
 Good resolution capability
 Good distribution
538 H. Wei et al. / Applied Energy 99 (2012) 534–544

Fig. 3. EGR ranges within the characteristic map [7].

Fig. 4. EGR-stratification modes.
gases easily mix. As a contrast, radial EGR stratification seems to be
the most appropriate flow structure for compression stroke. Since
the central air–fuel cylinder and the outer EGR tubular cylinder are
concentric with the engine cylinder, as the piston moves upwards
in compression stroke, the two cylinders will be compressed in ax-
ial direction. Particularly, if both EGR and air are swirling in the
same direction, the conservation of angular momentum may make
the two zones readjust their interface location to reach new force
balance. Radial stratification can sustain much longer time towards
the end of compression stroke [29–31]. In summary, radial stratifi-
cation is the most appropriate method for getting stratification of
intake and exhaust.
6. Comparison between cooled EGR and hot EGR
The operation that exhaust gas is recycled to intake directly is
called hot EGR, and the operation that EGR after cooling is applied
to recycled exhaust is called cooled EGR [15]. The engine using hot
EGR can use the high temperature exhaust to heat the intake, pro-
mote combustion and thus improve the thermal efficiency. While
cooled EGR increases intake density, thereby increases volumetric
efficiency of engine. At the same time, the decreased temperature
can further reduce NOX emission, but HC emission and cycle-by-
cycle variations are increased compared with that of hot EGR.
The comparison of the characteristics between cooled EGR and
hot EGR is given in Table 2.
Fig. 5a shows COVIMEP as a function of cooled and hot EGR in per-
centage. The maximum EGR tolerance is defined as 10% COVIMEP. As
can be seen from Fig. 5a, increasing cooled EGR fraction will in-
crease COVIMEP. This might be because of non-uniform mixture
composition consisting of recirculated EGR, air and residual gases.
Better mixing is expected when only hot EGR is used, so the COVIMEP
is lowered than that of cooled EGR.
Fig. 5b and c shows 0–10% and 0–90% MFB durations. Fig. 5b
shows the data of 0–10% MFB duration, indicating the process of
flame initiation in the mixture. One can see that, with increasing
EGR, EGR replaces air, the MFB duration becomes longer and the
mixture becomes more difficult to be ignited. Fig. 5c shows the in-
creased temperature caused by hot EGR enhances flame propaga-
tion, while cooled EGR does not affect the overall burn duration.
Fig. 5d and e shows HC and CO emissions as function of cooled
EGR and hot EGR in percentage. HC emission map shows, the frac-
tion of cooled EGR and hot EGR has great effect on HC emission.
Generally, increasing cooled fraction increases HC emission, while
increasing hot fraction reduces the emission. This effect is mainly
attributable to the decrease of combustion stability with increased
cooled EGR. The heating effect of hot EGR improves the combustion
temperature and then gains improvements in fuel–air mixing,
therefore HC emission will be reduced. With further increasing
EGR ratio, HC emission began to deteriorate, as the introduction
of EGR narrows flammability limits. Fig. 5e shows, NOX emission
decreases significantly with increasing percentages of cooled EGR
as a result of lower combustion temperature presented in this case.
While increasing percentage of hot EGR tends to leading to slight
increase in NOX emission.
Fig. 5f shows relative efficiency as a function of cooled and hot
EGR in percentage. The efficiency decreases with increased cooled
EGR. This can be attributed to the changes in combustion temper-
ature, and reduced temperature leads to a reduction in combustion
speed and then in thermal efficiency.
At present, despite of the advanced combustion systems and
emission control systems, cooled EGR is still the most effective
measure to reduce NOX emission [33]. When using EGR, the intake
composition changes in a large extent. Since the exhaust tempera-
ture reduces the intake density, the exhaust gas must be cooled to
maintain a high volumetric efficiency to avoid the increase of heat
loss. Cooled EGR is a key technology in the trend of downsized SI
engines, and it provides a means to meet the market requirement
of doing more with less [8]. Despite of cooling process, high tem-
perature exhaust will increase the intake temperature at high load.
Therefore, volumetric efficiency is reduced.

Table 2
A comparison of the characteristics between cooled EGR and hot EGR.
7. Knock suppression using cooled EGR

Cooled EGR Hot EGR

Characteristics  Lower NOX values
 Better knock suppression
 Complex structure
 Higher cost
In engine revolution, EGR technique was firstly adopted in
 Lower combustion duration diesel engines in order to limit thermal NOX formation rate by
 Lower HC values reducing combustion chamber temperature thanks to the dilution
 Simple structure
of fresh charge with a certain amount of exhaust gases recycled at
engine intake. Some studies also suggest that appropriate EGR ratio
 H. Wei et al. / Applied Energy 99 (2012) 534–544 539

Fig. 5. Comparison between hot EGR and cold EGR (the maximum EGR tolerance is indicated by the highlighted 10% COVIMEP contour. All data is for lean conditions with
constant fuel flow and constant diluents (air + EGR) flow) [32].

can reduce combustion noise by lower the pressure rise rate and
the pressure high frequency oscillation magnitude [34,35]. But
with the growing energy and environmental issues, various coun-
tries in the world have developed more stringent regulations on
vehicle emission and fuel economy to drive gasoline engine to be
developed towards downsizing direction. Turbo-charged spark
ignition engines are becoming increasingly popular in the world
market due to their compactness and high power density. How-
ever, due to the high power density of turbo-charged engine, knock
combustion and high exhaust gas temperatures constitute prob-
lems at high loads [19]. As a result, knock control is becoming
increasingly important.
In the past century, many scholars conducted studies on knock
[36–39]. Knocking combustion is generally accepted as the effect of
auto-ignition of the portion of cylinder charge where propagating
turbulent flame hasn’t reached. This view had been proved by
experimental evidence. The pressure/temperature of end-gas end
is controlled by the displacement of piston and the expansion of
combusted gases. With advanced facing of the combustion, end-
gas temperature increases, and finally, the reaction rate in end-
gas reaches a level at which spontaneous ignition occurs. Another
view is based on the interpretation of flame propagation theory. It’s
saying that knock is the result of flame acceleration in homoge-
neous mixture. When flame speed is higher than sound speed,
knock occurs [40]. When knock occurs, pressure and heat transfer
increase sharply, which will cause serious damage to engine. Nor-
mally, the number of knocking cycles is on-line determined using
band-pass filtered cylinder pressure signal. For each cycle, the
540 H. Wei et al. / Applied Energy 99 (2012) 534–544

amplitude of the signal from oscillating cylinder pressure is com-

pared with a reference voltage. If the amplitude exceeds reference
voltage, the cycle indicates knocking [41]. The paper presented
here does not care about the causes of knock, but only about knock
suppression methods.
Several knock suppression methods have been studied in re-
lated research. These methods can be divided into three categories
[42,43]. The first is to decrease effective compression ratio or delay
ignition timing to limit cylinder pressure. The second is to inject
excessive fuel in the mixture to decrease AFR. The third, an alterna-
tive and efficient solution, emerging as a promising technology to
limit knocking, is through dilution, such as cooled EGR.
To reduce exhaust temperature and to limit knock, the most
commonly used method is excessive fuel to decrease combustion
temperature. Under the condition of fuel enrichment, HC and CO
emissions will increase, and TWC can only work under stoichiom-
etric condition, which causes problems of fuel economy and emis-
sion. However, using EGR as a diluents allows the use of an overall
stoichiometric composition of the charge, thus insures a good con-
version of all emissions. Therefore, using cooled EGR instead of
excessive fuel to inhibit knock and to reduce emission is an effec-
tive measure in gasoline engine [41–43].
In Fig. 6a, maximum BMEP, as a function of the number of the
amount of EGR and ignition angle, is presented. For these tests
maximum BMEP was limited by knocking combustion and exhaust Fig. 7. Comparison of combustion characteristics for 11% fuel enrichment and 10%
temperature. The knock detection was set to detect cycles with EGR (engine speed of 4000 rpm, boost pressure of 90 kPa, spark ignition timing of
amplitudes of the oscillating cylinder pressure above 250 kPa, 6.6 CAD BTDC) [42].
and the exhaust temperature was limited to 960 °C [39]. We can
see clearly that after increasing EGR ratio from 7% to 13%, ignition
angle and maximum BMEP are increasing. But with further ad-
vanced ignition, BMEP will not increase, which is limited by knock-
ing combustion. The increase of EGR ratio will reduce this
limitation. Because an increase of EGR increases the mass in the
cylinder, and therefore lowers gas temperature, and subsequently,
more fuel can be combusted before reaching temperature limit.
Therefore, it can further widen ignition advance angle.
Fig. 6b shows cylinder pressure traces limited by knock as a
function of EGR ratio. In the figure, all ignition angles are the angle
before top dead center. The peak pressure tolerance of the engine is
slightly increased with EGR increases. The possible reason is that
the dilution and thermal effect of EGR reduce combustion temper-
ature and inhibit knocking combustion, therefore allow a greater
cylinder pressure.
With high load, the major role of EGR is to replace fuel enrich- Fig. 8. Engine emissions before catalyst from various EGR rates (engine speed of
ment to inhibit knock. When EGR ratio is 10% with a k of 0.9, the 4000 rpm, stoichiometric fuel–air mixtures) [42].

Fig. 6. Maximum BMEP and cylinder pressure as a function of EGR and the ignition angle (engine speed of 4000 rpm, EGR temperature of 120 ± 4 °C, charge air temperature of
40 ± 2 °C) [42].
 H. Wei et al. / Applied Energy 99 (2012) 534–544
Fig. 9. Knock intensity of knocking cycles (engine speed of 5000 rpm, full load
condition, 8 °CA BTDC spark ignition timing of k = 0.88 and 20 °CA BTDC spark
ignition timing of 13% EGR) [43].
engine has BMEP values of 1645 KPa and 1673 KPa respectively
(calculated from Fig. 7). That is to say, if knock doesn’t occur, using
high EGR ratio can get the same level of power output compared
with that of fuel enrichment. And using EGR can ensure stoichiom-
etric combustion, so this will greatly improve fuel economy and
lower emission.
Fig. 8 shows CO, HC and NOX emissions before catalyst at vari-
ous EGR rates. The values from fuel enrichment case are in the
upper part of the figure. The major effect using EGR instead of fuel
enrichment is that CO is significantly decreased from over 4% with
fuel enrichment to a more reasonable value of 0.8% with a stoichi-
ometric mixture. The negative aspect is that NOX is increased with
a stoichiometric mixture. But with the increase of EGR rate, NOX
decreases gradually.
Fig. 9 shows intensity of knocking cycles at 5000 rpm and under
full load conditions. The fuel rich cases show consistently high
average values of knock intensity, while the EGR cases display
quite a number of high knock intensity, but average knock inten-
sity is low. The possible explanation is that, the greater instability
of the cycles has a rapid initial combustion phase, leading to higher
cylinder pressure and consequently to higher end-gas temperature.
When knock limit is set to 50 KPa, the average knock intensity of
the knocking cycles in the two cases are the same, but in the fuel
rich cases, 87% of the cycles are knocking, while only 49% exhibits
knock with EGR. Consequently, using EGR to replace fuel enrich-
ment achieves relatively good results.
Fig. 10. Comparison of fuel consumption at high load (IMEP  1650 kPa, knock
intensity 62 kPa at ignition timing 20 °CA BTDC in case of EGR and 81 kPa at ignition
timing 8 °CA BTDC in case of fuel enrichment) [43].
541
In the full load cases with a rich mixture, fuel consumption is
very high. Naturally the consumption is reduced when fuel enrich-
ment is replaced by EGR. This is shown in Fig. 10, where fuel con-
sumption of a case with 13% EGR is compared with a case with
k = 0.9. With EGR, the fuel consumption at a comparable knocking
intensity reduces by more than 10%.
8. Implementations of EGR for turbocharged gasoline engines
The implementation of EGR is simple for naturally aspirated
gasoline engines, since exhaust tailpipe backpressure is normally
higher than intake pressure. With the development of downsized
gasoline engine, turbocharged gasoline engine is becoming
increasingly popular in the market. The implementation of EGR
is, therefore, more difficult, which leads to a coexistent problem
in engines. Two different schemas are shown in Fig. 11 [44].
The first structure is LPL EGR, as shown in Fig. 11a. Exhaust gas
goes into turbine at first, and then into compressor together with
fresh air. Tailpipe pressure P4 is larger than compressor inlet pres-
sure P1, so a positive difference between turbine outlet and com-
pressor inlet is generally available. In the structure with low
pressure, pressure difference in EGR circuits is very low, therefore
a high permeability for the EGR valve and EGR cooler is required to
ensure sufficient EGR flow. Furthermore, EGR cooler must have
high efficiency to ensure the efficiency of compressor, otherwise
it will cause intake air overheat and reduce volumetric efficiency.
However, conventional compressors and inter-coolers are not de-
signed to ensure the temperature and fouling of exhaust, which
limits the application of LPL EGR.
LPL EGR is divided into two structures: clean EGR and dirty EGR,
as shown in Fig. 12a and b. The impact of the two structures on
compressor is quite different. When using clean EGR, exhaust goes
into TWC and then goes through compressor together with fresh
air. The ceramic particles from TWC may corrode compressor
wheel, therefore additional coating is required to protect the
wheel. When using dirty EGR, a major difference is the presence
of sticky hydrocarbons, which might result in additional necessity
of component protection. On the other hand, dirty EGR will also
bring some additional benefit [45]:
 Improved rate and efficiency of combustion due to re-burn of
hydrocarbons, H2 and CO in the cylinder, and improved fuel
economy.
 Reduced pumping loss as a result of improvement of pressure
drop across the turbine.
 Improved EGR cooling due to lower enthalpy pre-catalyst
(based on exothermic reactions in the catalyst).
In order to ensure compressor durability, it is necessary to con-
duct more in-depth theoretical and experimental studies to resist
corrosion in compressor.
The one that has wide applications is the second structure, as in
Fig. 11b. In this structure, exhaust gas is from the upstream of tur-
bine to the downstream of compressor or of intercooler. However,
such HPL EGR must ensure that turbine upstream pressure is suf-
ficiently higher than boost pressure. That is to say, P3 > P2, other-
wise we cannot acquire enough exhaust gas being introduced
into intake air. Especially under low-speed and high-load condi-
tion, exhaust backpressure is lower than intake pressure, as shown
in Fig. 13. Therefore, in this case, exhaust cannot go into intake nat-
urally, thus remedies must be made by either increasing turbine
upstream pressure P3 or reducing boost pressure P2. The first meth-
od is to intake throttling or exhaust throttling that installs a throt-
tle valve in the downstream of compressor or upstream of turbine.
Another widely used method is to use VGT to provide desired EGR
542 H. Wei et al. / Applied Energy 99 (2012) 534–544

Fig. 11. Low pressure loop EGR and high pressure loop EGR.

Fig. 12. Clean vs. dirty EGR circuits.

Fig. 13. Inlet and exhaust pressure vs. speed of the engine [41].

driving pressure without substantially sacrificing the performance

of turbocharged engine. The shrinking of the flow passage of tur-
bine nozzles will increase turbine upstream pressure P3 and reduce
boost pressure P2.
When using EGR in turbocharged engine, the matching of tur-
bocharger relies on selected EGR structure. Fig. 14 shows the com-
pressor map without EGR, HPL EGR and LPL EGR. For each
structure, EGR variation rate has a different impact on the evolu-
tion of initially set point in compressor map [41]:

 In high pressure configuration, increasing EGR rate increases

only pressure ratio. This can induce surge line problems at
low speed and high load.
 In low pressure configuration, increasing EGR rate increases
simultaneously pressure ratio and air mass flow. This can
induce over-speed problem at high speed and high load. Fig. 14. Compressor MAP analysis [41].
 H. Wei et al. / Applied Energy 99 (2012) 534–544
In summary, it is possible to naturally recirculate exhaust into
intake in LPL EGR, but high temperature exhaust might corrode
compressor and inter-cooler. LPL EGR is divided into two struc-
tures: clean EGR and dirty EGR. The impact of the two structures
on compressor is quite different. The pressure difference over
HPL EGR circuit is very low. Therefore, it is necessary to use VGT
or throttle valve to generate a driving force to ensure sufficient
EGR flow. In HPL EGR configuration, EGR gases are taken before
turbine, thus before after treatment. EGR gases contain HC and
CO, which could be a source of corrosion in EGR circuit. Further-
more, using EGR in turbocharged gasoline engine must consider
turbo matching problem.
9. Conclusion
(1) EGR technique can reduce fuel consumption of gasoline
engine and meet more stringent emission regulations in
the future together with other advanced techniques.
(2) The engine using hot EGR can use exhaust to heat intake,
promote combustion, and thus improve thermal efficiency.
While cooled EGR increases intake density, thereby
increases volumetric efficiency of engine. At the same time,
decreased temperature can further reduce NOX emission, but
HC emission and cycle-by-cycle variations are increased
compared with that of hot EGR.
(3) Knock is a main problem to improve BMEP of gasoline
engine. Commonly used method to inhibit knock is fuel
enrichment. Using cooled EGR instead of excessive fuel to
inhibit knock and to reduce emission is an effective measure
in gasoline engine.
(4) With the development of downsized gasoline engine, turbo-
charged gasoline engine is becoming increasingly popular in
the market, and the implementation of EGR is, therefore,
more difficult, which lead to a coexistence problem in
engines. Furthermore, using EGR in turbocharged gasoline
engine must consider turbo matching problem.
Acknowledgment
Authors wish to express their sincere appreciation to the finan-
cial support of the National Natural Science Found of China NSFC
(No. 51176138) and the key project of Tianjin’s application infra-
structure with cutting-edge technology research program (Natural
Science Foundation 12TJZDTJ2800).
References
[1] Najafi G, Ghobadian B, Tavakoli T, et al. Performance and exhaust emissions of
a gasoline engine with ethanol blended gasoline fuels using artificial neural
network. Appl Energy 2009;86:630–9.
[2] Suh Hyun Kyu. Investigations of multiple injection strategies for the
improvement of combustion and exhaust emissions characteristics in a low
compression ratio engine. Appl Energy 2011;88:5013–9.
[3] Lee Dae Hee, Lee Jun Sik, Park Jae Suk. Effects of secondary combustion on
efficiencies and emission reduction in the diesel engine exhaust heat recovery
system. Appl Energy 2010;87:1716–21.
[4] Fontana G, Galloni E. Experimental analysis of a spark-ignition engine using
exhaust gas recycle at WOT operation. Appl Energy 2010;87:2187–93.
[5] Fathi Morteza, Saray R Khoshbakhti, Checkel M David. The influence of Exhaust
Gas Recirculation (EGR) on combustion and emissions of n-heptane/natural
gas fueled Homogeneous Charge Compression Ignition (HCCI) engines. Appl
Energy 2011;88:4719–24.
[6] Diana S, Giglio V, lorio B, Police G. A strategy to improve the efficiency of
stoichiometric spark ignition engines. SAE paper 961953; 1996.
[7] Blank Helmut, Dismon Heinrich, Kochs Michael W, Sanders Michael. EGR and
Air Management for Direct Injection Gasoline Engines. SAE paper 2002-01-
0707; 2002.
543
[8] Kaiser Michele, Krueger Uwe, Harris Roderick, Cruff Luke. Doing more with less
– the fuel economy benefits of cooled egr on a direct injected spark ignited
boosted engine. SAE paper 2010-01-0589; 2010.
[9] Abd-Alla GH. Using exhaust gas recirculation in internal combustion engines: a
review. Energy Convers Manage 2002;43:1027–42.
[10] Sarikoc Fatih, Kettner Maurice, Velji Amin, Spicher Ulrich. Potential of reducing
the NOX emission in a spray guided DI gasoline engine by stratified exhaust gas
recirculation (EGR). SAE paper 2006-01-1261; 2006.
[11] Fontana G, Galloni E. Variable valve timing for fuel economy improvement in a
small spark-ignition engine. Appl Energy 2009;86:96–105.
[12] Bai Yun-long, Wang⁄ Zhi, Wang Jian-xin. Part-load characteristics of direct
injection spark ignition engine using exhaust gas trap. Appl Energy
2010;87:2640–6.
[13] Zhao Fuquan, Harrington David L, Lai Ming-Chia. Automotive gasoline direct-
injection engines. SAE book; 2002.
[14] Ladommatos N, Abdelhalim SM, Zhao H, Hu Z. The dilution, chemical and
thermal effects of exhaust gas recirculation on diesel engine emission-part 4:
effect of carbon dioxide and water vapour. SAE paper 971660; 1997.
[15] Agarwal Deepak, Singh Shrawan Kumar, Agarwal Avinash Kumar. Effect of
Exhaust Gas Recirculation (EGR) on performance, emission, deposits and
durability of a constant speed compression ignition engine. Appl Energy
2011;88:2900–7.
[16] Vianna JN de S, Reis A do V, Oliveira AB de S. Reduction of pollutants emission
on SI engines – accomplishments with efficiency increase. ABCM; 2005.
[17] Benson RS, Whitehouse ND. Internal combustion engines. Oxford: Pergamon;
1979.
[18] Lavoie GA, Heywood JB, Keck JC. Experimental and theoretical study of nitric
oxide formation in internal combustion engines. Combust Sci Technol
1970;1:313–26.
[19] Levendis YA, Pavlatos I, Abrams R. Control of diesel soot, hydrocarbon and NOX
emission with a particulate trap and EGR. SAE paper 940460; 1994.
[20] Ghazikhani⁄ M, Feyz ME, Joharchi A. Experimental investigation of the exhaust
gas recirculation effects on irreversibility and brake specific fuel consumption
of indirect injection diesel engines. Appl Therm Eng 2010;30:1711–8.
[21] Al-Qurashi Khalid, Lueking Angela D, Boehman André L. The deconvolution of
the thermal, dilution, and chemical effects of exhaust gas recirculation (EGR)
on the reactivity of engine and flame soot. Combust Flame
2011;158:1696–704.
[22] Ladommatos N, Abdelhalim SM, Zhao H, Hu Z. Effects of EGR on heat release in
diesel combustion. SAE paper 980184; 1998.
[23] Ladommatos N, Abdelhalim SM, Zhao H, Hu Z. The effects of carbon dioxide in
exhaust gas recirculation on diesel engine emission. Proc Inst Mech Eng Part D
– J Autom Eng 1998;212:25–42.
[24] Heywood JB. Internal combustion engine fundamentals. New York: MGraw-
Hill; 1998.
[25] Lake T. Simulation and development experience of a stratified charge gasoline
direct injection engine. SAE paper 962014; 1996.
[26] Sasaki S. Effects of EGR on direct injection gasoline combustion. J Am Soc
Echocardiogr 1997;971:333–6.
[27] Reuter W. Innovative EGR systems for direct-injection gasoline engines. In: 8th
Aachen Colloquium-Auto Eng Tech.; 1999. p. 793–814.
[28] Huang Bin, Hu Erjiang, Huang⁄ Zuohua, Zheng Jianjun, Liu Bing, Jiang Deming.
Cycle-by-cycle variations in a spark ignition engine fueled with natural gas–
hydrogen blends combined with EGR. Int J Hydrogen Energy
2009;34:8405–14.
[29] Xu Min, Chen Grant, Daniels Chao, Dong Mike. Numerical Study on Swirl-Type
High-Dilution Stratified EGR Combustion System. SAE paper 2000-01-1949;
2000.
[30] Dong Mike, Chen Grant, Xu Min, Daniels Chao. A preliminary CFD investigation
of in-cylinder stratified EGR for spark ignition engines. SAE paper 2002-01-
1734; 2002.
[31] Jackson NS. Understanding the CCVS stratified EGR combustion system, SAE
Paper 960837; 1996.
[32] Toulson Elisa, Watson Harry C, Attard William P. The effects of hot and cool
EGR with hydrogen assisted jet ignition. SAE paper 2007-01-3627; 2007.
[33] Tomazic Dean, Pfeifer Andreas. Cooled EGR – a must or an option for 2002/04.
SAE paper 2002-01-0962; 2002.
[34] Li Zhaowen, Shu Gequn, Wei Haiqiao, Liang Xingyu. Experimental
investigation on effect of EGR on combustion noise in transient conditions of
DI – diesel engine. Chinese Int Combust Engine Eng 2010;31:17–22.
[35] Wei Haiqiao, Liang Xingyu, Shu Gequn. Influence of the accelerating operation
mechanism on the combustion noise in DI – diesel engines. Int J Automotive
Technol 2012;13:373–88.
[36] Merola⁄ Simona S, Vaglieco Bianca M. Knock investigation by flame and radical
species detection in spark ignition engine for different fuels. Energy Convers
Manage 2007;48:2897–910.
[37] Brunt Michael FJ. Gasoline engine knock analysis using cylinder pressure data.
SAE paper 980896; 1998.
[38] Leppard WR. Individual-cylinder knock occurrence and intensity in
multicylinder engines. SAE Paper 820074; 1982.
[39] Livengood JC, Wu PC. Correlation of autoignition phenomena in internal
combustion engines and rapid compression machines. In: 5th Int symp on
combustion. The Combustion Institute; 1955. p. 347–56.
[40] Curry S.A three dimensional study of flame propagation in a spark ignition
engine, SAE paper; 1963.
544 H. Wei et al. / Applied Energy 99 (2012) 534–544

[41] Potteau Sebastien, Lutz Philippe, Leroux Samuel, Moroz Stephanie, Tomas Eva.
Cooled EGR for a turbo SI engine to reduce knocking and fuel consumption.
SAE paper 2007-01-3978; 2007.
[42] Grandin Börje, Ångström Hans-Erik, Stålhammar Per, Olofsson Eric. Knock
suppression in a turbocharged SI engine by using cooled EGR. SAE paper
982476; 1998.
[43] Grandin Börje, Ångström Hans-Erik. Replacing fuel enrichment in a turbo
charged SI engine: lean burn or cooled EGR. SAE paper 1999-01-3505; 1999.
[44] Zheng Ming, Reader Graham T, Gary Hawley J. Diesel engine exhaust gas
recirculation––a review on advanced and novel concepts. Energy Convers
Manage 2004;45:883–900.
[45] Roth David B, Keller Philip, Becker Michael. Requirements of external EGR
systems for dual cam phase turbo GDI engines. SAE paper 2010-01-0588;
2010.