The Contribution of Oil to Carbon Particle
Emissions from Diesel Engines
Technical Report
PETTER TORNEHED
Stockholm 2010
Trita-MMK 2010:04
ISSN 1400-1179
ISRN/KTH/MMK/R-10/04-SE
Department of Machine Design
School of Industrial Engineering and Management
Royal Institute of Technology, KTH
SE-100 44 Stockholm
www.kth.se
The Contribution of Oil to Carbon Particle Emissions from
Diesel Engines
Petter Tornehed
Scania CV AB, Engine Development, NMBP
SE-151 87 Södertälje, Sweden
E-mail: [email protected]
Abstract
Reducing particle emissions from diesel engines has been an important issue in recent
decades. This study assesses the contribution of oil to carbon particle emissions by reviewing
the literature and theory on fuel soot formation, and combining it with estimates based on a
maximum and minimum approach.
The study shows that the minimum oil contribution to carbon particle emissions for a Euro 5
heavy-duty diesel without exhaust aftertreatment is around 0.4 mg/kWh. The corresponding
maximum estimate is 1.1 mg/kWh.
1(10)
1 Introduction
Particle emissions from heavy-duty diesel engines have been greatly reduced in the last few
decades. For example, in Europe the allowable emissions were reduced from 0.36 g/kWh in
1992 (Euro 1) [1] to 0.02 g/kWh in 2005 (Euro 4) [2]. They will be reduced to 0.01 g/kWh for
2013 (Euro 6) [3]. The Euro 6 legislation will probably also include particle number
legislation.
Traditionally, particle emissions have been reduced by reducing the particles arising from the
fuel, for example by improving the fuel injection system and air management, and by
reducing fuel sulphur content. However, as the fuel-related particles have been effectively
reduced, the role of oil-related particles has become more and more prominent.
The author is currently working in a project to increase understanding of oil-related particles.
The project has shown that when best practices are used, the cylinder system is normally the
largest contributor of oil-related particles [4]. Other contributors are the turbocharger
bearings, valve stem seals, and crank case ventilation.
In-cylinder oil consumption can be categorized as follows [4]:
• Throw-off, when oil is driven towards the combustion chamber by inertia forces (Figure
1a)
• Reverse blow-by, when gas containing oil is driven by pressure to flow towards the
combustion chamber (Figure 1b)
• Evaporation from hot surfaces (Figure 1c)
• Top land scraping, when oil is scraped from the cylinder liner by the top land of the piston
or, more likely, by carbon deposits on the top land (Figure 1d).
a) Throw-off
b) Reverse blow-by
c) Evaporation
d) Top land
scraping
Figure 1. In-cylinder oil consumption, from Johansson and Andersson [5]
This categorization of oil consumption is derived from Yilmaz et al. [6] and Herbst and
Priebsch [7]. Yilmaz et al. use the terms throw-off, transport with reverse gas flow and
evaporation. Herbst and Priebsch [7] divide in-cylinder oil consumption into evaporation, oil
throw-off, reverse oil blow and oil scraping by piston top land. An alternative categorization
by Mihara et al. [8] depends on the state of the oil when it is consumed. They divide the incylinder oil consumption into vapour oil consumption (VOC) and liquid oil consumption
2(10)
(LOC). The SAE Piston and Rings Standards Committee [9] uses the categories liquid oil that
is scraped, flowing or being squeezed around the ring; vaporization of oil; and oil in a mist.
The inter-ring pressure is often a concern when oil consumption occurs, and has previously
been addressed by Johansson and Andersson [5].
The definition of a particle can differ from field to field, but within in the automotive
industry, particle emissions are defined by the measuring methods and how the particles are
collected on a filter. Particle emissions from diesel engines are often divided into a solid
fraction consisting of carbon and ash; a liquid fraction that is often referred to as SOF (soluble
organic fraction) or VOF (volatile organic fraction); and sulphates with water. The solid
carbon is formed during rich combustion (and is often referred to as soot); ash is formed from
additives in the lubricant or fuel. The liquid fraction is fuel or oil (hydrocarbons) that is not
fully oxidized. Oxidation of sulphur to SO3 leads to sulphuric acid/sulphates [10].
The contribution of oil to particle emissions is modelled using transfer functions (Figure 2).
The transfer functions for the oil hydrocarbons, ash, and sulphur-associated particles are
described in [11], [12] and [13].
PMoil
Carbon Ash HC Sulphates+H2O
•
•
•
•
Temperature
Sulphur content
Oil ash content
Other factors
Transfer function
In cylinder oil consumption
Figure 2. Transfer function model from oil consumption to particle emission, from Tornehed
and Olofsson [11]
This study aims to assess the contribution of oil to carbon particle emission. It is done by
using maximum and minimum estimates based on a literature review of theory and
measurements within the field. The influence of the increasing use of exhaust aftertreatment
devices is left outside of this study. Thus the current (Euro 5) legislation is used as reference
for calculating examples of emissions, since these requirements can be met without exhaust
aftertreatment [14].
2 Review of results from engine tests
Buchholz et al. [15] report a study in which they used 14C to distinguish between oil and fuel
carbon particle emissions. They used a 100% biodiesel with a known 14C content and a 14Cfree lubrication oil. The lubricant contributed 4% of the carbon particles, with a standard
deviation of 1%. The engine used was a 1993 Cummins B5.9, a six-cylinder diesel engine
with the total capacity of 5.9 L and a maximum power of 175 hp at 2500 rpm. The test was
run at 1600 rpm and 285 Nm (50% of maximum torque). The engine was equipped with a
Bosch B7100 PE inline fuel injection pump capable of pressures up to 115 MPa.
Hilden and Mayer [16] studied the contribution of the oil to total particulate matter and to
SOF. Their data can be used to assess the oil’s contribution to non-SOF particles. They used
14
C to label the oil in 5 diesel engines with displacements between 1.8 L and 5.7 L
3(10)
manufactured between 1978 and 1983. The vast majority of the oil particulates were found in
the SOF; 2% of the non-SOF (soot) was from the oil, with a standard deviation of 2%.
Schneider at al. [17] studied oil deposits in the exhaust system and particles in the exhaust
stream of a gasoline engine by labelling the oil with 14C. The engine was a 1997 4.6 L V8
engine. The oil’s contribution to the soot deposits in the exhaust manifold was 2%, in the
exhaust port 6%, and on the oxygen sensor 3%. The results from the sample filtered out from
the exhaust gas stream showed a clear influence of engine speed. The oil’s contribution to
soot was ~3% at 750 rpm and ~20% at 2500 rpm. The average calculated from all measuring
points tabled by Schneider at al. [17] is 6% with a standard deviation of 5%. For the samples
taken in the exhaust stream, the corresponding values were 11% with a standard deviation of
7%.
Mayer et al. [18] studied a 5.7 L diesel engine from 1978. The engine oil was labelled with
C. The recovery rate, defined as radioactivity in the SOF particles in relation to radioactivity
in the total particulates, was found to be 101% with a standard deviation of 4%. Hence the
oil’s contribution to non-SOF particles is very low (on average negative) in this study.
14
3 Soot particle formation
Studies by Haynes and Wagner [19] and Frenklach [20] indicate that soot particle formation
originating from diesel fuel combustion can be summarized as:
• Particle inception, when the first soot particles arise from condensed material,
typically with a diameter < 20 Å.
• Surface growth, when gas phase species adhere to the surface and increases particle
mass. This is where most of the solid phase is generated; the number of particles is
constant but the mass increases.
• Particle coagulation, when the size of the particles increases either by coalescence or
agglomeration. The sum of the particle mass is constant and the number of particles
decreases.
• Soot oxidation.
Several principles for soot inception have been discussed, including polyacetylene, ionic
species, or polyaromatic hydrocarbons as a gaseous precursor of soot. The present majority
opinion is that soot particles form via PAH (polycyclic aromatic hydrocarbons) [20]. The
small size of the initial soot particles, often referred to as soot nuclei, makes the soot loading
from the initial particles negligible. The bulk of the (solid) soot particles are formed during
the surface growth phase [19]. The end product of soot leaving the cylinder will be dependant
on the balance between the formation and subsequent oxidation of soot particles.
Akihama et al. [21] studied the influence of the local temperature and equivalence ratio (1/λ)
on the soot formation rate for n-hexane. A rich mixture is not enough to promote soot
production, as the right temperature is also needed (Figure 3). They found that the maximum
soot production for n-hexane occurred at 2000 K and that soot formation is suppressed at
temperatures below ~1700 K. Other authors using the same approach as Akihama et al. [21]
with temperature and equivalence ratio maps have shown very similar maps, but the soot
formation region is often displaced towards lower temperatures. For example, Charlton [22]
published a map in which the maximum soot formation region is around 1500 K.
4(10)
Equivalence ratio φ
Max soot
formation
Temperature (K)
Figure 3. Soot formation rate dependency on local temperature and equivalence ratio,
principle after Akihama et al. [21]
The clear connection between local temperature and soot in diesel diffusion combustion has
been shown by, for example, Lindström and Ångström [23] (see Figure 4).
Figure 4. Visual spectrum, temperature analysis, and soot intensity during combustion in an
optical diesel engine, from Lindström and Ångström [23]
Heywood [24] describes three paths for carbon particle formation:
• Low temperature path, below 1700 K. Only the aromatics and highly unsaturated
aliphatic particles with high molar weight are prone to form solid carbon; this occurs
by pyrolysis.
• Medium temperature path, above about 1800 K (typical diesel temperatures). If the
combustion is rich enough, all hydrocarbons normally appearing in diesel fuel produce
soot.
• High temperature path, above typical diesel temperatures. Nucleation of carbon
vapour occurs.
5(10)
4 Estimation of oil-related carbon particles
The data presented earlier in this paper were used to estimate the oil’s contribution to carbon
particle emissions. The following assumptions were used:
• Oil consumption, 0.2 g/kWh (0.1 % of the fuel consumption)
• Fuel consumption, 200 g/kWh
• Hydrogen to carbon relationship, 2 for the lubrication oil
• Hydrogen to carbon relationship, 1.85 for fuel
• The lubrication oil consists only of hydrocarbons. Ash and other species are
disregarded.
1. A maximum limit for the specific oil-related carbon particle emissions was calculated
by assuming that all carbon in the oil consumed creates carbon particles. The
calculations indicate maximum carbon particle emissions of 0.17 g/kWh.
2. The assumption that all carbon in the consumed oil ends up as carbon particles is not
plausible, since the oil carbon particles alone would clearly exceed the Euro 5 particle
emission limit (0.02 g/kWh [2]). However, experience has shown that the rules can be
met without exhaust aftertreatment [14].
3. Another approach is to assume the same ratio for carbon particle formation for the oil
as for the fuel. The current Euro 5 emission legislation allows particle emissions of up
to 0.02 g/kWh [2]. On this basis one could assume that carbon particles arising from
the fuel produce emissions of 0.01 g/kWh. This leads to a fuel carbon transfer rate of
5.8 ⋅ 10−5 . The transfer rate is defined as the carbon particle mass divided by the
consumed amount of carbon. Applying this transfer rate to the maximum limit
estimate for the oil carbon particles yields a specific oil-carbon particle emission of
9.9 ⋅ 10−6 g/kWh. This is considerably lower than 0.02 g/kWh.
4. Buchholz et al. [15] showed that 4% of the carbon particles from a 1993 diesel engine
were oil-related. Applying the 4% factor to a modern Euro 4/5 engine leads to oilrelated carbon particle emissions of 0.4 mg/kWh, using an assumed total carbon
particle emission of 0.01 g/kWh (50% of the Euro 4/5 limit).
5. The engine tested by Buchholz et al. [15] was produced in 1993. The US emission
standard at that time allowed PM emissions of up to 0.34 g/kWh (0.25 g/bHPh) [25].
Applying the 4% value and assuming that 50% of the particles are carbon leads to oilcarbon particulate emissions of 6.7 mg/kWh.
6. Schneider at al. [17] studied oil’s contribution to soot particles on a port injected petrol
engine. They showed an 11% oil contribution to soot in the exhaust. Applying 11% in
the same manner as the 4% from Buchholz et al. [15], together with the current
emission legislation, leads to an estimate of 1.1 mg/kWh.
6(10)
5 Discussion
The quantity of oil-related carbon particles has been estimated in a number of ways in this
study and the results are summarized in Figure 5. The oil’s contribution to soot particles
derived from the measurements reviewed is shown in Figure 6.
0
20
-1
1. All carbon in oil TR=100%, maximum estimate
-2
2. Euro 4/5, maximum estimate
10
10
-3
10
Oil contribution to soot (%)
Estimated oil carbon
particle emissions (g/kWh)
10
5. 4% oil carbon 1993, maximum estimate
6. 11% oil carbon E4/E5, maximum estimate
4. 4% oil carbon E4/E5, minimum estimate
-4
10
-5
10
3. TR as for fuel (Euro 4/5), minimum estimate
-6
Diesel - Exh.
Gasoline - Deposits & Exh.
Gasoline - Exh.
10
Buchholz et al.
5 Hilden and Mayer
0
1980
10
Figure 5. Oil carbon particle emission
estimates
15
Schneider et al.
1985
1990
1995
Engine build year
2000
Figure 6. Oil contribution, with one standard
deviation, to carbon particle emission
depending on engine build year and fuel
The maximum estimate, by assuming all carbon in the lubrication oil produces carbon
particles, TR=100 %, indicates a maximum emission level of 0.17 g/kWh. This is a gross over
estimation since the oil-carbon particles alone would clearly exceed the Euro 5 particle
emission limit (0.02 g/kWh [2]), which can be met without exhaust aftertreatment [14].
Although this maximum estimate is far from the real value, it can be used for comparative
purposes and to determine how large fraction of the available carbon finally produced carbon
particles.
Using the assumption that oil has the same transfer rate into carbon particles (TR = 5.8 ⋅ 10−8 )
as fuel would give oil carbon particle emissions of approximately 10 μg/kWh for a Euro 5
engine. This is likely to be an underestimate since all the studies summarised in Figure 6 [15,
16, 17] show an oil contribution to carbon particles above reasonable levels of oil to fuel
consumption. The lowest average oil contribution was 2% (Hilden and Mayer [16]). The oil
consumption was not stated, but normal oil consumption is well below 1% of the fuel
consumption.
The Buchholz et al. [15] study showed that the oil contributed 4% oil of the carbon particles.
For a modern engine (Euro 4/5) this would correspond to 0.4 mg/kWh. This is 2% of the Euro
4/5 particulate limit. For a US diesel engine from 1993, the corresponding value is
6.7 mg/kWh, which is 34% of the Euro 4/5 limit.
The total soot particulates could be assumed to be low in the Schneider at al. [17] study,
because the engine was a port-injected gasoline engine. Port-injected gasoline engines are
generally assumed to be soot free, and Heywood [26] states “In properly adjusted sparkignited engines, soot in the exhaust is not a significant problem”. The absence of soot is
generally due to the pre-cylinder mixing of the gasoline and air prior to combustion, making
the equivalence (1/λ) approximately 1. This to a large extent avoids the local soot-rich island
7(10)
in Figure 3. Remarkably, even though the fuel is considered to burn almost soot-free, the fuel
clearly dominated as soot source.
The reason why the majority of the soot particles are caused by the fuel even in engines
considered not to have soot problems might be found in the soot creation/oxidation process.
According to Akihama et al. [21], soot formation is promoted not only by a rich mixture, but
also by the right local temperature (maximum soot formation rate at 2000 K [21] or 1500 K
[22]). Fuel consumption is in the range of 1000 times the oil consumption, and the cylinder
temperature will therefore be dominated by the fuel heat release. Thus the conditions to create
soot are dominated by the fuel, and reducing the soot from the fuel automatically reduces the
soot from the oil. It is likely not only the creation process, but also the burnout of the particles
that are affected. Since the fuel governs the soot emissions, the continuing reduction of fuel
soot from a 1993 vintage engine (Buchholz et al. [15]) up to today would also reduce the oil
soot particles. Therefore the 4% reported by Buchholz et al. [15] is likely to be the oil’s
contribution to soot particles even for modern diesel engines. This figure should be seen as a
minimum estimate, and the 11% extracted from Schneider at al. [17] could be used as a
maximum estimate.
The fact that the standard deviations of the oil’s contribution to soot (Figure 6) range from
29–90 % of the average values ought also to be kept in mind. Perhaps more studies containing
a range of engine speed and load combinations could help to explain the variations in data.
For example, Schneider et al. [17] found that the oil’s contribution to soot particles increased
with higher engine speeds. It would be interesting to repeat the 14C measurements to see how
much the oil contributes to soot in a modern diesel engine. Another alternative to improve our
understanding of oil soot particle formation could be to introduce lubrication oil into a model
originally designed for fuel soot formation and oxidation. An example of such a model is the
one presented by Westlund et al. [27].
6 Conclusions
Oil-related carbon particle emissions may be estimated by multiplying total carbon particle
emissions by a factor of 4 % as a minimum and 11 % as a maximum.
This study indicates that the contribution of the oil to carbon particles for a Euro 5 heavy-duty
diesel without exhaust aftertreatment can be estimated to be between 0.4 mg/kWh and
1.1 mg/kWh.
7 Acknowledgements
I would like to thank Scania CV AB and the Swedish Energy Agency for financing this work.
In addition, I would like to express my gratitude to all those who have shown interest in the
project, especially Ulf Olofsson, Hans-Erik Ångström, Jonas Holmborn, Håkan Malmstad,
Matts Bormark, and Sören Andersson.
8(10)
8 References
[1]
Council Directive (EEC) 91/542 on measures to be taken against the emission of
gaseous pollutants from diesel engines for use in vehicles [1991] OJ L295.
[2]
Council Directive (EC) 2005/55 on measures to be taken against the emission of
gaseous and particulate pollutants from engines [2005] OJ L275.
[3]
Regulation (EC) No 595/2009 on type-approval of motor vehicles and engines with
respect to emissions from heavy duty vehicles (Euro VI) and on access to vehicle repair
and maintenance information and amending regulation (EC) No 715/2007 and directive
2007/46/EC and repealing directives 80/1269/ECC, 2005/55/EC and 2005/78/EC [2009]
OJ L 188.
[4]
Johansson, P., Oil-Related Particle Emissions from Diesel Engines, Licentiate thesis.
Royal Institute of Technology, Department of Machine Design, Sweden, Trita-MMK
2008:08, ISSN 1400-1179, ISRN/KTH/MMK/R-08/08-SE, Stockholm 2008.
[5]
Johansson, P. and Andersson, S., Variations in Piston Second Land Pressure as a
Function of Ring Gap Position, Int. J. Engine Res. (11)(2), 153-161, 2010.
[6]
Yilmaz, Y., Tian, T., Wong, V. W. and Heywood, J. B., The Contribution of Different
Oil Consumption Sources to Total Oil Consumption in a Spark Ignition Engine, SAE
Technical Paper 2004-01-2909, 2004.
[7]
Herbst, H. M. and Priebsch, H. H., Simulation of Piston Ring Dynamics and Their
Effect on Oil Consumption, SAE Technical Paper 2000-01-0919, 2001.
[8]
Mihara, K., Iwakata, K., Naitou, Y. and Naitou, K., Effect of Lubricating Oil on Oil
Consumption in Diesel Engines, JSAE Review, 16(1), 97-97(1), 1995.
[9]
SAE Piston and Ring Standards Committee, Power Cylinder Oil Consumption:
Transport Mechanisms, J2794, February 2007.
[10] Kittelson, D. B., Engines and Nanoparticles: A Review, J Aerosol Sci. 1998: 29(5/6),
575–588.
[11] Tornehed, P. and Olofsson, U., Towards a model for engine oil hydrocarbon particulate
matter. Submitted to SAE 2010 Powertrains, Fuels and Lubricants Meeting October 2527, 2010, San Diego, California, USA, submission date 04/09/2010. Manuscript number
2010-01-10FFL-002.
[12] Tornehed, P. and Olofsson, U., Modelling of lubricant ash particles in diesel engine
exhaust. Submitted to Journal of Automobile Engineering, submission date 07 Jun 2010
Manuscript number JAUTO1667.
[13] Johansson, P., Impact of Sulphur on Particulate Matter Paying Special Attention to the
Lubricant: Based on a Literature Review, Royal Institute of Technology, KTH, TritaMMK 2009:14, ISSN 1400-1179, ISRN/KTH/MMK/R-09/14-SE.
[14] Scania Press release: Scania to launch Euro 5 engine with EGR in 2007, 6 December
2006, available at http://www.scania.com/media/pressreleases/n06051en.aspx accessed
2010-04-23.
[15] Buchholz, B.A., Dibble, R.W. and Rich, D., Quantifying the Contribution of Lubrication
Oil Carbon to Particulate Emissions from a Diesel Engine. SAE Paper No 2003-011987, 2003.
9(10)
[16] Hilden, D.L. and Mayer, W.J., The Contribution of Engine Oil to Particulate Exhaust
Emissions from Light Duty Diesel Powered Vehicles, SAE Paper No. 841395, 1984.
[17] Schneider E.W., Sell J.A. and Siekkinen J.W., The Contribution of Lubricating Oil to
Exhaust Deposits and Exhaust Particulates from Gasoline Engines: A Radiotracer
Method, SAE Paper No 982580, 1998.
[18] Mayer W.J., Lechman, D.C. and Hilden, D.L., The Contribution of Engine Oil to Diesel
Exhaust Particulate Emissions, SAE Paper No. 800256, 1980.
[19] Haynes, B.S., and Wagner, H.G., Soot Formation. Progress in Energy and Combustion
Science, 7 (4), 229-273, (1981).
[20] Frenklach, M., Reaction Mechanism of Soot Formation in Flames, Phys Chem Chem
Phys, 4 (2002), 2028–2037.
[21] Akihama, K., Takatori, Y., Inagaki, K., Sasaki, S. and Dean A.M., (2001) Mechanism of
the Smokeless Rich Diesel Combustion by Reducing Temperature, SAE Paper No 200101-0655, 2001.
[22] Charlton, S., Heavy-Duty Diesel Emission Control Symposium, presentation 2007.
[23] Lindström, M. and Ångström H.-E., “In-Flame Evaluation of Emission Formation in
Optical and Metal Engine Using High Speed Camera and Endoscope”, THIESEL 2010
Conference on Thermo- and Fluid Dynamic Processes in Diesel Engines, 2010.
[24] Heywood, J.B., Internal Combustion Engine Fundamentals, McGraw-Hill, p. 638, 1988.
[25] United States Environmental Protection Agency, EPA “Emission Standards Reference
Guide Heavy-Duty and Nonroad Engines” EPA420-F-97-014 September 1997,
available at http://www.epa.gov/otaq/cert/hd-cert/stds-eng.pdf accessed 2010-04-23.
[26] Heywood, J.B., Internal Combustion Engine Fundamentals, McGraw-Hill, p. 626, 1988.
[27] Westlund, A., Ångström, H.-E. and Lindström M., Validation of a Simplified Model for
Combustion and Emission Formation in Diesel Engines Based on Correlations for Spray
Penetration and Dispersion, Gas Entrainment into Sprays and Flame Lift-Off, SAE
Paper No 2010-01-1494, 2010.
10(10)