2014-01-1308

8/12/2019 2014-01-1308

http://slidepdf.com/reader/full/2014-01-1308 1/14

AbstractThe advantages of applying Compressed Natural Gas (CNG)

as a fuel for internal combustion engines are well known. In

addition to a signicant operating cost savings due to a lower

fuel price relative to diesel, there is an opportunity to reduce

the engine's emissions. With CNG combustion, some

emissions, such as Particulate Matter (PM) and Carbon

Dioxide (CO2), are inherently reduced relative to diesel fueled

engines due to the nature of the combustion and the molecular

makeup of the fuel. However, it is important to consider the

impact on all emissions, including Total Hydrocarbons (THC)

and Carbon Monoxide (CO), which can increase with the use

of CNG. Nitrogen Oxides (NOx) emission is often reported to

decrease with the use of CNG, but the ability to realize this

benet is signicantly impacted by the control strategy and

calibration applied.

FEV has investigated the emissions and performance impact

of operating a heavy-duty diesel engine with CNG in a dual fuel

mode. The CNG was introduced via injectors mounted to an

inlet pipe located upstream of the intake manifold. The

fumigation approach included a mixer to improve the

distribution of gas prior to delivery to the cylinder. The initial

investigations sought to determine how the performance of a

heavy-duty diesel engine would be affected by the introduction

of CNG. For this effort there was no change to the base enginecalibration, and the ability to maximize substitution of diesel

with CNG was investigated. It was observed that the ability to

maximize substitution of diesel with CNG across the operating

map was limited by extremely high THC levels, combustion

instability and limitations in peak cylinder pressure and exhaust

gas temperature.

With the application of a simplied engine calibration with a

single diesel injection and Exhaust Gas Recirculation (EGR),

timing adjustments allowed higher CNG substitution levels in

several areas of the operating map. A further increase in gas

substitution along with higher fuel conversion efciency,

improved combustion stability and even lower emissions couldbe achieved through Reactivity Controlled Compression

Ignition (RCCI) combustion. This approach required a unique

injection strategy along with a careful balance of EGR rates

and boost pressure. Under this combustion regime it was

possible to observe a simultaneous reduction of NOx and PM

emissions, approaching engine-out emission levels that could

avoid, or signicantly minimize, aftertreatment of these

species.

With the desire to quickly apply CNG systems to existing diese

engine architecture in an effort to reap the benet in fuel cost

savings, manufacturers and system developers must be carefu

to understand the full impact on the engine's performance andemissions. Tests conducted as part of this investigation have

revealed that an un-optimized approach to CNG introduction

can lead to extreme THC emissions that mostly consist of

Methane (CH4). In addition, the maximum gas substitution

level is signicantly limited in most regions of the engine

operating map. Thus, the ability to specically tune the

calibration for operation with CNG is essential to achieving the

maximum benet in fuel cost savings and emission control.

Introduction

The use of CNG in heavy-duty applications has seen a sharp

growth in the last couple of years as reported in studiesconducted by the Energy Information Administration (EIA) [1].

EIA studies project CNG consumption in heavy-duty

applications to increase at 11.4% per year through 2040. The

growth is mainly seen in the off-road market, but other sectors

such as on-highway, marine and locomotive are also

embracing a switch to CNG at a rapid rate.

There are two main technology approaches through which

CNG can be applied for heavy-duty engines. A common

approach for existing diesel engines is to operate in a dual fuel

regime where diesel and CNG are combusted simultaneously.

Investigation of Diesel and CNG Combustion in aDual Fuel Regime and as an Enabler to Achieve RCCI

Combustion

2014-01-1308

Published 04/01/2014

Mufaddel Dahodwala, Satyum Joshi, Erik W. Koehler, and Michael FrankeFEV Inc.

CITATION: Dahodwala, M., Joshi, S., Koehler, E., and Franke, M., "Investigation of Diesel and CNG Combustion in a Dual

Fuel Regime and as an Enabler to Achieve RCCI Combustion," SAE Technical Paper 2014-01-1308, 2014,

doi:10.4271/2014-01-1308.

Copyright © 2014 SAE International

Downloaded from SAE International by Erik Koehler, Wednesday, March 26, 2014 11:09:34 AM

8/12/2019 2014-01-1308


A second approach considers a dedicated CNG combustion

system that requires the addition of an ignition source. The

main technological challenges that apply to this type of engine

conversion are studied in detail by Ribas [2].

As an additional consideration, past studies have shown that

due to throttling losses and lower compression ratio, efciency

of a Spark Ignited (SI) CNG engine is lower compared to a dual

fuel diesel/CNG engine, particularly at part loads [3]. Brake

Mean Effective Pressure (BMEP) capability is also reduced

with an SI CNG engine, leading to lower torque at low speeds

when compared to a dual fuel diesel/CNG engine. An SI CNG

engine does however have an advantage on the exhaust gas

aftertreatment, as stoichiometric operation allows the

application of a conventional and comparatively simple 3-way

catalyst.

In the case of dual fuel diesel/CNG engines, two types of

technologies are currently considered depending on the

method used for CNG induction. In the rst type, both the

diesel fuel and natural gas are directly injected into the

combustion chamber using either two separate injectors or a

special injector with a dual-concentric needle design [4]. In thesecond type, natural gas is either fumigated at a single point

into the intake path of the engine before or after the

turbocharger compressor and is premixed with air and EGR or

injected at multiple points in the intake port of the engine.

Intake fumigation at single point is the most widely used

method of CNG induction for on-highway applications, due to

its simplicity, and is the approach applied for this work.

Much research has been conducted to understand the

combustion behavior of dual fuel diesel/CNG engines [4, 5, 6,

7, 8, 9, 10]. As reported by most researchers, the major

difculty with dual fuel operation is the challenge of providing

high levels of CNG substitution, especially at low and mediumloads. Attempts to apply high substitution levels lead to lower

engine efciency and higher concentrations of CO and

unburned hydrocarbon emissions. At higher loads, the

improvement in CNG utilization leads to improvements in both

engine performance and exhaust emissions. To date a large

portion of the research has been focused on either single

cylinder research engines or light-duty production engines.

Relatively few studies look at engine performance and

corresponding emission impact of dual fuel concepts applied to

medium-duty and heavy-duty production engines. One study

on a heavy-duty engine was conducted by Guzman et. al. [4],

wherein the authors recalibrated the base diesel engine to

operate in dual fuel mode to quantify the CO2 and PM

emissions during the transient portion of the Heavy-Duty

Federal Test Procedure (HD FTP) and World Harmonized

Transient Cycle (WHTC). Their work found that, with a properly

calibrated dual fuel engine signicant reductions in PM and

CO2 emissions could be achieved in transient operation when

compared to a baseline diesel engine.

In recent years, the heavy-duty market has seen the strongest

growth in the application of dual fuel technologies, with refuse

trucks and buses leading the way. In terms of on-highway

applications, most dual fuel engines currently in use are End

Of Life (EOL) retrots that do not need to comply with stringent

emission standards. To date, few companies have certied

dual fuel engines to US2010 emission standards. One of those

that has certied to US2010 has stated a fuel cost savings of

up to 30% and a maximum CNG substitution of 70%.

One motivation behind this study was to understand the

limitations imposed by a US2010-compliant on-highway diesel

engine in terms of maximum CNG substitution. Additional

complexity to this approach results from the management of

cylinder-to-cylinder variations, which in some cases makes the

most promising strategy developed on a single cylinder engine

not feasible for a production engine without applying signicant

design changes. One of the main requirements for employing a

dual fuel approach is to have the exibility of operating in

diesel-only mode. Design changes to specically allow better

operation with CNG will not be attractive to the end user if they

compromise the performance of the engine in diesel-only

mode.

The rst part of the study aims at understanding engine

behavior with introduction of CNG without changes to the basecalibration, as is typically done by retrot companies. To

accomplish this, 13 load points were selected, and a maximum

CNG substitution sweep was conducted at each of these

points. The second part of the study investigated the impact of

diesel calibration changes on the allowable maximum CNG

substitution and thus the achievable fuel cost savings. The

calibration optimization was completed at all 13 load points.

The nal part of this study is focused on exploring advanced

combustion concepts with CNG and seeks to maximize

substitution and lower emissions.

Experimental Setup A production heavy-duty diesel engine was used for this

investigation. The 2010 inline 6-cylinder engine has a

displacement of 13 liters, is rated at 425 HP and is On-Board

Diagnostics (OBD) - II compliant. It is equipped with a high

pressure common rail fuel injection system, cooled high

pressure EGR, and a compression ratio of 16.5:1. CNG was

introduced into the intake using eight CNG injectors located

downstream of the charge air cooler. A mixer was installed

downstream of the CNG introduction location to support equal

distribution of the CNG in the intake manifold. Figure 1 depicts

the test cell setup applied for this investigation. A Rapid

Controller Prototyping (RCP) system was used for controlling

the amount of CNG induction. The production engine was

equipped with a Diesel Oxidation Catalyst (DOC), Diesel

Particulate Filter (DPF) and Selective Catalytic Reduction

(SCR) catalyst, however for this investigation only a DOC was

installed and the backpressure was adjusted via a back

pressure valve to simulate the absent aftertreatment

components. The aftertreatment module in the Engine Control

Unit (ECU) was deactivated to prevent it from affecting the

base engine performance.


8/12/2019 2014-01-1308


Figure 1. Test Cell Schematic

The engine was coupled to a 560 kW Alternating Current (AC)

dynamometer. CNG and diesel fuel ow measurements were

accomplished using Micro Motion ow meters while the air ow

was measured using an ABB air ow meter. Engine-out and

tailpipe gaseous emissions were measured with a dual channel

Horiba MEXA 7500 DEGR emission bench. The intake

manifold was instrumented for CO2 concentration

measurement, which was used for calculation of the EGR rate.

PM emission was determined through the Motor Industry

Research Association (MIRA) calculation, which utilized data

provided by a smoke meter drawing sample upstream of the

DOC. The engine was instrumented with Kistler 6061B

in-cylinder pressure transducers to allow cylinder pressure

measurements on all six cylinders.

Test Matrix

Based on operating points from the Ramped Modal Cycle

(RMC), HD FTP and highway cycles for this engine, 13 specic

test points were selected for this study. The points wereselected to provide a good balance between certication cycles

and real-world operation. The program test points are overlaid

upon the cycle operating points in Figure 2. For the baseline

substitution study, the CNG substitution was increased in 10%

increments until the engine became unstable (Indicated Mean

Effective Pressure (IMEP) Coefcient of Variation (COV) less

than 10%) or a limit of peak pressure or maximum exhaust

temperature was reached.

Figure 2. Test points selected based on the operating points on the

RMC, HD FTP and highway cycles

For optimization studies, the effect of main injection timing and

CNG substitution was studied at each of the selected operating

points. Figure 3 shows an example test matrix for one

particular engine speed. The main injection timing was swept

at different substitution levels across the three load points. At

high loads the timing could not be advanced because of peak

pressure limitations, and it was found that higher substitution

could only be achieved by retarding the main timing. RCCI

combustion was evaluated at the low load points. The test

matrix for RCCI combustion at low loads was an extension of

the test matrix shown in Figure 3. The timings were advanced

to 50-80 deg Crank Angle (CA) before Top Dead Center

(bTDC) at substitution levels of 60-80%. The CNG substitution

was calculated on an energy basis. The details of the

calculation are outlined in the Appendix.

Figure 3. Test matrix applied for optimization at a fixed speed

Result and Discussion

Baseline Maximum Substitution A review was conducted to understand the methods employed

by retrot manufacturers to substitute CNG for diesel without

changes to the base calibration. Two popular approaches are

to manipulate the boost signal and to bleed off part of the fuel

supplied to the injectors back to the fuel tank to reduce the

injected diesel fuel for a particular load. These approaches are

not directly possible when considering an engine equipped with

OBD. Without OBD calibration changes, such approaches

would cause a low boost or injector fault and result in a de-rate

of the engine. Thus, to avoid this inuence for the current work,

the reduction in diesel quantity was obtained by reducing the

pedal demand and maintaining load by introducing CNG into

the system. With this approach, the control parameter setpoints, including the injection events, EGR rates and rail

pressures, were not consistent for different substitution levels

within the same load point. These variables were controlled by

the ECU based on the commanded pedal (fueling) and the

engine speed.

In an effort to maintain this paper within a reasonable length

the results presented without calibration changes will be limited

to 1500 rpm and three different load points; 6 bar, 14 bar and

24 bar. These points will be analyzed to understand the effect

of CNG substitution on engine performance and emissions.


8/12/2019 2014-01-1308


Figure 4 shows the Apparent Heat Release Rate (AHRR) and

cylinder pressure traces for 6 bar BMEP at three different

substitution levels. The emission impact and corresponding

control variables for different substitution levels at this load

point are provided in Figure 5. From these gures, deterioration

in combustion efciency can be seen with an increase in CNG

substitution, possibly due to the reduction in injected diesel fuel

leading to lower combustion chamber temperatures. This led to

signicant increase in hydrocarbon emissions, Brake Specic

Fuel Consumption (BSFC) (calculated on a diesel equivalence

fuel ow rate) and brake specic cost, which is dened as the

net fuel cost per kilowatt hour of energy ( Appendix). An

increase in NOx emission was observed up to 40% CNG

substitution, due to reduction in EGR, but then dropped as the

combustion efciency deteriorated. The EGR reduction can be

attributed to the richer lambda values observed with increasing

CNG substitution as the fresh air charge is displaced with

CNG. Increased PM emissions at higher substitution were

possibly due to the increased soluble organic fraction at higher

loads and a richer lambda value.

Figure 4. AHRR and in-cylinder pressure for 6 bar BMEP load at 1500

rpm at three different substitution levels

A reduced peak in the heat release trace and a late burning of

CNG, also observed by Maxey et al. [5], was clearly visible with

an increase in substitution. Observations regarding the

combustion behavior are in agreement with those of Maxey et

al.; wherein the reduction in diesel fuel quantity reduces the

amount of diesel available to ignite the less reactive CNG [5].

Figures 6 outlines the AHRR and cylinder pressure trace for 14

bar BMEP, while Figure 7 shows the emission impact with

increasing CNG substitution at the same load. Similar trends

for NOx, PM and THC emissions were observed for 14 bar

BMEP as seen at 6 bar BMEP, however, here the combustion

efciency remained above 90% leading to a net benet in

brake specic cost. At this load, a maximum substitution of

98% could be reached where CNG was ignited by the diesel

pilot injection and an SI type heat release trace was observed.

Figure 5. Emissions and performance parameters at 6 bar BMEP,

plotted against CNG substitution

A NOx emission increase was closely coupled to a reduced

EGR rate and an earlier heat release. In general, the

combustion COV, as also observed by Sun et al. [6], was

affected by CNG substitution percentage and the EGR quantity

The cylinder pressure rise rate and the peak cylinder pressure

increased signicantly at 24 bar BMEP, shown in Figure 8, due

to the simultaneous combustion of diesel and CNG. This

limited CNG substitution to 36% at this load point. Combustion

efciency remained above 98.8% due to the higher combustion

temperatures. The NOx, PM and THC emissions trends

presented in Figure 9 for 24 bar BMEP are similar to those

observed at 6 bar and 14 bar BMEP.


8/12/2019 2014-01-1308



rpm

Figure 7. Emissions and performance variables at 14 bar BMEP,



rpm

Figure 9. Emissions and performance variables at 24 bar BMEP,



8/12/2019 2014-01-1308


Figure 10 shows the practical CNG substitution map when

access to the calibration is not available and thus no calibration

changes are applied. This map is determined by maintaining

the engine-out emissions within a level that would allow the

engine to meet US2010 emission legislation when conventional

diesel aftertreatment is applied. The average NOx emission for

the baseline diesel calibration was approximately 2.0 g/bhp-hr,

therefore a NOx limit of 2.0 g/bhp-hr, along with a NMHC limit

of 0.238 g/bhp-hr, was applied as limiting criteria.

From Figure 10 we can see that the average substitution would

be limited to less than 10% if the base calibration is carried

over and no changes are applied. There are three major

reasons for the limited substitution shown in Figure 10. First, at

low loads the amount of unburned THC is very high. Second,

at mid loads the NOx emissions tend to increase due to a

reduction in EGR caused by running richer lambda values.

Lastly, the peak pressures increase with increase in CNG

substitution at higher loads, leading to the necessity to observe

mechanical limits. Most dual fuel retrots on the market today

are applied to EOL vehicles that do not require emission

certication. These applications can therefore accept increases

in emission levels and apply higher CNG substitution levels.

Figure 10. Practical CNG substitution map without base engine

calibration changes and meeting US2010 emission standards

Calibration Optimization

As shown in the previous section, only very low CNG

substitution levels can be achieved when changes to the base

diesel calibration are not considered. Therefore, to increase the

CNG substitution levels, a modication of the base calibration is

required. The base calibration used three injections along with

EGR to reach the desired NOx, PM and Non-Methane

Hydrocarbons (NMHC) targets when operating on diesel fuel. Tounderstand the effect of each injection event, a series of tests

were conducted at 1500 rpm and 6 bar BMEP and multiple

substitution levels. Three injection strategies were explored:

main injection only, main + pilot injection and main + post

injection. In order to eliminate the effect of EGR, the EGR rate

was reduced and maintained at a near-zero level throughout

modication of the base calibration. A summary of the injection

strategy employed for each case is provided in Table 1.

Figure 11 shows the impact of each of these strategies on the

NOx and THC emissions at various CNG substitution levels.

The NOx emissions decrease with an increase in CNG

substitution levels for all three strategies, but none of the

strategies show a signicant advantage in their ability to contro

THC emissions. Therefore, to simplify the diesel injection

control, a main injection only strategy was applied for the

calibration optimization.

Table 1. Overview of the three injection strategies employed at 1500

rpm, 6 bar BMEP and different CNG substitution

Figure 11. Comparison of NOx and THC emission for multiple CNG

substitution levels at 1500 rpm and 6 bar BMEP

It was also realized that EGR was required to control the NOxemissions, as shown in Figure 12.

A study was then conducted with multiple CNG substitution

levels and diesel injection timings for the main injection only

case. At each speed and load point the EGR rate and injection

pressure were held constant. The boost pressure was

controlled by the ECU based on the existing calibration. The

diesel injection timing advance was restricted to 20 deg bTDC,

above which the possibility of spraying the diesel fuel directly

on the cylinder liner was high. The CNG substitution was

limited to 80%; above this level the combustion was unstable


8/12/2019 2014-01-1308


and resulted in very high THC emissions. The optimization

study was conducted at all 13 points as outlined in the Test

Matrix section. To describe the impact of injection timing and

CNG substitution on emission and performance, the following

text focuses on the results obtained at 1500 rpm and three

different load points.

Figure 12. Comparison of NOx emissions for multiple CNG substitution

levels at 1500 rpm and 6 bar BMEP

Figure 13 shows the effect of diesel injection timing and CNG

substitution levels on the NOx, PM and THC emissions at 1500

rpm and 6 bar BMEP. A decrease in NOx emissions and an

increase in THC emissions can be observed as CNG

substitution levels are increased at constant injection timing.

Figure 13. Injection timing study for different CNG substitution levels at

1500 rpm, 6 bar BMEP with 35% EGR

Figure 14 compares the AHRR and pressure traces for three

different substitution levels at a xed injection timing of 11 deg

bTDC. As shown in Figure 14, the initial peak of the AHRR

trace reduces as CNG substitution is increased, while the

longer burn durations lead to lower combustion temperatures

and lower NOx emissions. However, at a xed substitution

level the NOx emissions increase as the injection timing is

advanced, but the THC emissions decrease.

Figure 14. AHRR and cylinder pressure trace for 1500 rpm, 6 bar

BMEP at 11 deg bTDC with three different substitution levels

Comparing the AHRR traces in Figure 15, it can be observed

that, as the injection timing is advanced, the AHRR tracesmove toward Top Dead Center (TDC). This leads to higher

in-cylinder temperatures causing higher NOx emission

formation but at the same time allowing more time for the CNG

to burn at higher temperatures and reducing the THC

emissions. Although the best NOx emission results are

obtained between 60% and 80% substitution levels and

approximately 12 deg bTDC injection timing, at this point, the

unburned THC emissions are very high. Thus, a compromise

must be identied between the maximum substitution and the

allowable NOx and THC emissions. The PM emissions, on the

other hand, are very low across the complete optimization

range, with a minimum achieved at 20 deg bTDC between 20%

and 60% CNG substitution levels. Unfortunately, the NOXemissions in this range are very high.

Figure 16 shows the NOx, PM and THC emissions at 1500 rpm

and 13 bar BMEP. The NOx emission results are similar in

terms of a decreasing trend with increase in CNG substitution

levels at the same injection timing and an increase in NOx

emissions with injection timing advance at a constant

substitution level. Again, as in the previous case, the regions

where the NOx emissions are lowest have very high THC


8/12/2019 2014-01-1308


emissions and vice-versa. Thus, at mid loads as well, the

amount of CNG substitution is dependent upon the trade-off

between NOx and THC emissions.


BMEP at 60% CNG substitution with three different injection timings



Figure 17 shows the impact of injection timing at three different

substitution levels for 24 bar BMEP. Contrary to the approach

at the lower loads, the injection timing was retarded at high

load to allow for higher CNG substitution levels. By retarding

the injection timing, the mechanical limits of maximum cylinder

pressure were avoided. As the injection timing was retarded

with increasing CNG substitution levels, the NOx and PM

emissions were decreasing with a familiar increase in THC

emissions. Comparing the AHRR traces in Figure 18, it is

observed that as the CNG substitution is increased the peak of

the AHRR trace moves away from TDC, matching the earlier

nding of a longer burn duration with less reactive CNG leading

to reduced NOx formation. At a xed substitution level, a

retarded timing reduced the NOx emission as expected, but the

THC emissions increased. The ability to further retard the

injection timing is limited beyond a certain crank angle since

the longer CNG burn duration leads to high THC emissions.



Based on the main timing and CNG substitution studies

conducted at the 13 speed and load points, a map was

generated showing the maximum CNG substitution that is

possible when allowing changes to the base calibration. The

map is again determined by observing the limits imposed by

the US2010 emission legislation.


8/12/2019 2014-01-1308


As shown in Figure 19 the average substitution achieved is

approximately 50%, with higher substitution achieved at the

higher load points where the retarded injection timing strategy

was employed to control the pressure rise rates. With this

substitution level, there was a penalty in the CH4 emissions,

but these emissions are not regulated by the US2010 emission

legislation, and therefore, no limit was imposed on the CH4

levels while generating the optimized map. At low loads the

substitution is mainly limited by the unburned hydrocarbons

due to a limited diesel quantity available for igniting the

premixed CNG charge.

One possible approach to overcoming these limitations is to

change from conventional diesel combustion to reactivity

controlled combustion. RCCI is a dual fuel strategy in which a

high reactivity fuel (diesel) is injected early enough to allow

complete mixing with a low reactivity fuel (CNG) to create a

diverse reactivity map within the chamber, allowing for

controlled and complete auto ignition of the premixed charge.

The RCCI combustion strategy not only allows higher gas

substitution levels but also has the potential to simultaneously

reduce NOx and PM emissions.


BMEP with three different substitution levels

Figure 19. Optimized CNG substitution map with calibration changes

and meeting US2010 emission standards

Investigation of RCCI Combustion

Figure 20. AHRR and cylinder pressure traces for three different main

injection timings at 60% substitution (6 bar BMEP at 1500rpm)

The existence of very high hydrocarbon emissions at low loads

impeded higher substitution levels during the optimization

effort. As a step toward further optimization, it was considered

that the higher fuel reactivity difference between CNG and

diesel could potentially enable RCCI combustion, despite a


8/12/2019 2014-01-1308


relatively high compression ratio. To explore this concept, the

main injection timing for diesel was advanced well beyond

conventional diesel injection timings at a low load point.

Figures 20 and 21 show the AHRR and the corresponding

emissions at four different main injection timings and a 60%

CNG substitution level. Up to a main injection timing of 32 deg

bTDC, the peak pressures and temperatures continue to

increase, leading to high levels of NOx formation. THC, CO

and smoke were reduced due to an increased peak

combustion temperature.

Figure 21. Emissions and performance variables at 1500 rpm and 6

bar BMEP, plotted against main injection timing

However, with further advancement in timing, beyond 32 deg

bTDC, the AHRR trace moved closer to TDC. This resulted in

lower peak combustion temperatures and pressures causing a

sharp reduction in NOx emissions. The heat release trace is

wider with a lower peak and clearly visible two-stage heat

release that includes a cool ame region. The heat release

prole closely resembled that of RCCI combustion [11], and it

was concluded that the spatial fuel reactivity gradient led to the

controlled heat release [11].

Although the reactivity gradient governed the heat release rate,

EGR was necessary to control the combustion phasing for this

high compression ratio engine. At 1500 rpm and 6.0 bar BMEP,

the maximum possible EGR level was 37%. This EGR level

was held constant for each injection timing investigated at this

operating point. The NOx emissions at the most advanced

injection timing of 65.2 bTDC was below the US2010 emission

legislation, indicating the potential of this concept in eliminating

NOx aftertreatment. A summary of the operating condition and

observed emissions at the injection timing of 65.2 bTDC has

been provided in Table 2.

Table 2. Operating conditions and performance for RCCI at 6 bar

BMEP and 1500 rpm

It could be shown that with further advanced timings of up to

80 deg bTDC the combustion phasing did not exhibit further

change. However, an increase in CO emissions was observed

with further timing advancement, possibly due to the

impingement of liquid fuel on the cylinder walls.

Figure 22 shows an optimized CNG substitution map when

RCCI combustion was included at the low load points. The

average CNG substitution was improved from approximately

50% to above 65% for the complete engine map.

Figure 23 shows the cost savings based on the optimized CNG

substitution map outlined in Figure 22. A maximum cost

savings of up to 40% can be achieved with advanced

combustion concepts tested at the lower speed and load

points.

As outlined in Table 3, operation in dual fuel mode (that makes

use of RCCI at low loads) allows, on average, a 43% reduction

in NOx emissions, a 68% reduction in PM emissions and a

22% reduction in fuel costs.


8/12/2019 2014-01-1308


Additional work is ongoing to identify the factors that contribute

toward achieving RCCI combustion at different speed and load

points. Results from these investigations will be further

explored and discussed in a future publication.

Figure 22. Optimized CNG substitution map with calibration changes,

including RCCI combustion at low loads and meeting US2010 emission

standards

Figure 23. Cost savings in percentage from baseline diesel operation

Table 3. Average cost savings and emissions reduction in percent,

obtained from the optimized CNG substitution map that includes RCCI

combustion at low loads

Conclusion

1. Assuming no access to the base engine calibration, a

baseline maximum substitution map with practical emissions

was created. It was found that the cost and emissionbenets of running on CNG were limited at lower loads due

to high hydrocarbon emissions and at higher loads due to

peak cylinder pressures and turbine inlet temperatures.

2. In an optimization step, strategies were developed

primarily involving diesel injection timing change and CNG

substitution level. Advancing the diesel injection timing at

lower loads led to a reduction in THC emissions. However,

at higher loads, slightly retarded injection timing helped

in achieving higher CNG substitution. The optimized

substitution map achieved an average of 49% substitution

across the test point map with 15% reduction in NOx and

43% reduction in PM emissions. At higher engine speeds,

it appears the reduced time for CNG/air mixing results in a

varying cylinder-to-cylinder distribution of CNG, and thus

deteriorated combustion stability.

3. RCCI combustion could be achieved at low loads,

which enabled even higher CNG substitution and lower

emissions. A maximum of 50% net indicated thermal

efciency was observed at 6 bar BMEP load point along

with 75% reduction in both NOx and PM emission. The

potential benets of RCCI combustion were limited dueto the un-optimized combustion chamber design and high

compression ratio.

Future Work

This work shows that an RCCI combustion strategy for burning

CNG and diesel fuel is most promising in terms of NOx

emission control and fuel consumption reduction. However,

challenges remain that must be resolved to allow extended use

of RCCI combustion at higher loads. Studies are currently

ongoing to explore the factors that affect the ability to achieve

RCCI combustion at different speed and load points. Along with

calibration optimization, changes in engine design, includingpiston bowl design, compression ratio and nozzle

conguration, are being considered. Port fuel injection of CNG

in an effort to reduce the cylinder-to-cylinder variation observed

with RCCI is also being investigated. Solutions enabling RCCI

combustion and maximizing CNG substitution in dual fuel

mode must also respect the option for diesel-only operation

when CNG is not available.

References

1. US Energy Information Administration, “Natural Gas,” http://

www.eia.gov/naturalgas/, Aug 2013

2. Ribas X., “Heavy Duty liquefied Natural Gas engine

developments to meet future emissions requirements,

methodology and real application,” FISITA paper

F2010F013, 2010.

3. Ouellette, P., “Cummins Westport Spark-Ignited (SI)

and High Pressure Direct Injection (HPDI) Natural Gas

Engines,” presented at NGVTF 2003, USA, January 28-29,

2003.

4. Barroso Guzman, P., Ribas, X., García, J.Sr., and Pita,

M.Sr., “PM and CO2 Reduction in a Dual-fuel Heavy-

duty Diesel Engine during the Freeway Part of Transient

Worldwide Emission Tests,” SAE Technical Paper 2013-01-

2759, 2013.

5. Maxey, C., Kalaskar, V., Kang, D., and Boehman, A.,

“Impact of Supplemental Natural Gas on Engine Efficiency,

Performance, and Emissions,” SAE Technical Paper 2013-

01-0847, 2013, doi:10.4271/2013-01-0847.

6. Sun, L., liu, Y., Zhou, L., and Zeng, K., “Experimental

Investigation of Cycle-by-Cycle Variations in a Natural Gas/

Diesel Dual Fuel Engine with EGR,” SAE Technical Paper

2013-01-0853, 2013, doi:10.4271/2013-01-0853.

7. Aroonsrisopon, T., Salad, M., Wirojsakunchai, E.,

Wannatong, K. et al., “Injection Strategies for Operational


http://www.eia.gov/naturalgas/


http://www.sae.org/technical/papers/2013-01-2759




http://dx.doi.org/10.4271/2013-01-0847


http://dx.doi.org/10.4271/2013-01-0853

http://dx.doi.org/10.4271/2013-01-0853


http://dx.doi.org/10.4271/2013-01-0847







8/12/2019 2014-01-1308


Improvement of Diesel Dual Fuel Engines under Low Load

Conditions,” SAE Technical Paper 2009-01-1855, 2009,

doi:10.4271/2009-01-1855.

8. Kowalewicz A., and Woloszyn R., “Comparison of

performance end emissions of turbocharged CI engine

fuelled either with diesel fuel or CNG and diesel fuel,”

Combustion Engines, PTNSS-2011-SC-117, 2011.

9. Papagiannakis, R., Hountalas, D., and Kotsiopoulos, P.,

“Experimental and Theoretical Analysis of the Combustion

and Pollutants Formation Mechanisms in Dual Fuel DIDiesel Engines,” SAE Technical Paper 2005-01-1726,

2005, doi:10.4271/2005-01-1726.

10. Papagiannakis, R., Hountalas, D., Rakopoulos, C.,

and Rakopoulos, D., “Combustion and Performance

Characteristics of a DI Diesel Engine Operating from Low to

High Natural Gas Supplement Ratios at Various Operating

Conditions,” SAE Technical Paper 2008-01-1392, 2008,

doi:10.4271/2008-01-1392.

11. Nieman, D., Dempsey, A., and Reitz, R., “Heavy-Duty

RCCI Operation Using Natural Gas and Diesel,” SAE Int. J.

Engines 5(2):270-285, 2012, doi:10.4271/2012-01-0379.

Contact Information

Koehler, Erik

Manager, Performance and Emissions

Commercial Engines; FEV, Inc.

[email protected]

Dahodwala, Mufaddel

Senior Engineer, Performance and Emissions

Commercial Engines; FEV, Inc.

[email protected]

Acknowledgments

The authors would like to thank FEV, Inc. management for

encouraging this research effort and providing the resources

necessary to accomplish its goals. The authors would also like

to thank those who took the time to review this paper and

provided valuable feedback.

Abbreviations

AC - Alternating Current

AHRR - Apparent Heat Release Rate

BMEP - Brake Mean Effective Pressure

BSC - Brake Specic CostBSFC - Brake Specic Fuel Consumption

bTDC - Before Top Dead Center

CH4 - Methane

CNG - Compressed Natural Gas

CO - Carbon Monoxide

CO2 - Carbon Dioxide

COV - Coefcient of Variance

DOC - Diesel Oxidation Catalyst

DPF - Diesel Particulate Filter

ECU - Engine Control Unit

EGR - Exhaust Gas Recirculation

EIA - Energy Information Administration

EOL - End Of Life

FTP - Federal Test Procedure

GGE - Gasoline Gallon Equivalent

HD - Heavy-Duty

IMEP - Indicated Mean Effective Pressure

MIRA - Motor Industry Research Association

NMHC - Non-Methane Hydrocarbon

NOx - Nitrogen Oxides

OBD - On Board Diagnostics

PM - Particulate Matter

RCCI - Reactivity Controlled Compression Ignition

RCP - Rapid Controller Prototyping

RMC - Ramped Modal Cycle

SCR - Selective Catalytic Reduction

SI - Spark Ignited

TDC - Top Dead Center

THC - Total Hydrocarbons

WHTC - World Harmonized Transient Cycle



http://dx.doi.org/10.4271/2009-01-1855


http://dx.doi.org/10.4271/2005-01-1726


http://dx.doi.org/10.4271/2008-01-1392

http://dx.doi.org/10.4271/2012-01-0379

http://dx.doi.org/10.4271/2012-01-0379

http://dx.doi.org/10.4271/2008-01-1392


http://dx.doi.org/10.4271/2005-01-1726


http://dx.doi.org/10.4271/2009-01-1855


8/12/2019 2014-01-1308


APPENDIX

CNG SUBSTITUTION CALCULATION

(1

Where,

= Mass ow rate of CNG

QLHV cng

= Lower heating value of CNG

= Mass ow rate of diesel

BRAKE SPECIFIC COST CALCULATION

Brake specic cost is a parameter that accounts for the amount of cost savings that one can get through CNG substitution. The average

national cost of CNG (2.09$/GGE) and Diesel (3.84$/Gallon) in May 2013 were used for this calculation.

(2

Where,

C cng

= Average National Cost of CNG in $/Kg (1 GGE CNG = 2.567 Kg CNG)

C diesel

= Average National Cost of Diesel in $/Kg (1 US Gallon Diesel = 3.149 Kg Diesel)

P = Engine Power (W)

CNG SUBSTITUTION OPTIMIZEROptimum CNG substitution at different operating points was calculated using the composite desirability index for multiple response

optimization. The target, maximum values and weight factors were selected for NOx, PM, NMHC, CO, BSC and ringing intensity

according to regulation limits.

Desirability can be calculated for individual responses by the following formula:

(3

The composite desirability can then be expressed as:

(4

Where,

Y = Response (NOx, PM, BSC…etc)

w = Weighing factor

n = Number of Responses


8/12/2019 2014-01-1308


CD = Composite Desirability Index

d = Desirability

The Engineering Meetings Board has approved this paper for publication. It has successfully completed SAE’s peer review process under the supervision of the sessionorganizer. The process requires a minimum of three (3) reviews by industry experts.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical,photocopying, recording, or otherwise, without the prior written permission of SAE International.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE International. The author is solely responsible for the content of thepaper.

ISSN 0148-7191

http://papers.sae.org/2014-01-1308




2014-01-1308

Documents