4. Experimental results for diesel engine combustion
4.6. W ATER - IN - FUEL EMULSIONS
A series of back diffused laser images with light cycle oil in ambient air as well as in 15.5%
oxygen is given in Figure 51. At -2 deg. ATDC, it can be seen that light cycle oil in 21% oxygen is ignited, while in 15.5% oxygen it has not done yet. In 21% oxygen, light cycle oil forms a typical soot cloud, visible until close to the injector nozzle. In 15.5% oxygen, the soot cloud is visible more distant from the injector nozzle with a cap between the liquid spray and the soot cloud. Moreover the soot cloud is wider, as expected from the photographs in Figure 50.
However, a clear increase in soot formation within the luminous flame by lowering the oxygen content is not recognised in these experiments.
Figure 51: Back diffused laser images of light cycle oil combustion in 21 and 15.5%
oxygen
immediately after injection, thus causing micro-explosions which result in smaller fuel droplets and hence in a better fuel and oxygen mix [71]. However, this theory has never been proven. The momentum-theory approach is widely accepted [72] [73]: this calculates the air / fuel ratio in the spray at the end of the fuel injection. An example is shown in Figure 52, in which the spray cone at the end of injection is calculated for fuel only and for 40% water added to the fuel, respectively. The latter forms the larger spray cone as a consequence of the larger momentum of the injected liquid. As the quantity of fuel is the same in both cases, the added water leads to a higher air / fuel ratio. The air / fuel ratio at this latest injection time point strongly influences the after-burning behaviour. Furthermore, it is expected that H2O dissociates to OH- and H- radicals at the high temperatures in the combustion chamber and the OH-radicals oxidise soot particles.
Figure 52: Improvement of air / fuel ratio by adding water [72]
In [71], emulsions with 10% and 20% water added to the fuel were tested in a single cylinder YANMAR TF120 engine with a 92 mm bore. The water-in-oil droplets of the emulsions were measured to be within a range of 20 - 30 µm. A decrease in NOx emission of 10% was achieved and carbon monoxide emission was reduced, compared to pure diesel fuel. A slightly narrower spray angle by applying emulsified fuel is reported in [74], where emulsions on the basis of MDO and additional water contents of 10, 20 and 30% are investigated in a combustion chamber. The authors observe a longer ignition delay and a shorter burning period with increasing water content. A Hilton’s Continuous Combustion Unit was applied to investigate emulsions in [75];
the highest combustion efficiency is achieved by adding 5% water to the diesel fuel. Carbon monoxide emission does not change in these experiments, and NOx emission is not reported. A combination of nitrogen-enriched air and emulsified fuel was investigated on a seagoing ship [62].
As smoke emission and carbon monoxide emission increased with nitrogen-enriched air, emulsified fuel was successfully applied to reduce smoke emission. Moreover, a further decrease of 10% in NOx emission by adding 10% water is stated. Kawasaki Heavy Industries has tested emulsions on a 58,000 DWT bulk carrier [76]. The main engine is a Kawasaki MAN B&W 6S50MC-C7. During the shop trial, emulsions with 25% and 50% water added to 100% MDO were tested. NOx emission was reduced by 40% with 50% water at 75% load. Fuel oil
consumption was slightly reduced and carbon monoxide emission was almost unchanged.
During the sea trial, emulsions on the basis of heavy fuel oil (380 cSt at 50 ˚C) and 15%, respectively, with 25% water were tested. NOx emission was reduced by 16 – 17% at 75% and 100% load; interestingly, carbon monoxide emission increased by 150 – 200%.
In the current study, emulsions are investigated with diesel fuel and light cycle oil. Based on the author’s knowledge, this is the first time that emulsified light cycle oil has been studied. The aims are primarily to test emulsions in the rapid compression expansion machine in order to investigate the potential to lower emissions and improve after-burning behaviour. Moreover, the study’s objectives are to gain an insight into the combustion process; to define suitable optical measurement methods and to produce validation data for an on-going simulation. As a next step, experimental results are validated with the optical two-stroke engine. In the second part of the study, emulsions are combined with the technique of nitrogen-enriched air to examine the potential of combining advantages. Finally, the approach is applied to the low-grade fuel light cycle oil to assess the potential of emulsions to improve its combustion behaviour.
Figure 53: Notation of emulsified fuel
The following notations are applied in this study and illustrated in Figure 53: DO-EMF25 and LCO-EMF25 refer to 100% fuel and 25% water, while DO-EMF40 and LCO-EMF40 consist of 100 % fuel and 40% water.
Figure 54: Formulation of emulsified fuels
The procedure to formulate the emulsions is as follows. The fuel is premixed with 0.1%
glycerine and the water with 0.9% RHEODOL SP-L 10; the percentages refer to the total
emulsion volume. Afterwards, the fuel is slowly mixed into the water before a pump starts circulating the emulsion through a homogeniser, where the emulsion is forced by high pressure into narrow channels, leading to a sharp increase in flow velocity while the pressure declines below the saturation steam pressure. The apparatus is shown in Figure 54.
Samples of each emulsion are taken after the homogenisation process and after injecting through a nozzle diameter of 0.5 mm with 150 MPa injection pressure. The results are shown in Figure 55. The injection process itself further refines the emulsion, as the injection nozzle acts like a homogeniser: the emulsion is forced by high pressure through a narrow nozzle followed by an expansion. After the injection process, as can be seen in Figure 55, the water droplets are too small to be clearly visible through a microscope.
Figure 55: Microscopic view of an emulsion after the homogeniser and after injection The experimental conditions to analyse diesel emulsions in the rapid compression expansion machine are listed in Table 13. In order to compare the three fuels, the injection duration is adjusted to the same total heat release. Again, a relatively high injection pressure of 150 MPa is chosen. Besides the injection duration, the experimental conditions are the same as in Sections 4.4 and 4.5.
Table 13: Experimental conditions for diesel emulsions
Fuel Diesel DO-EMF25 DO-EMF40
Compression pressure at injection start 10 MPa
Compression temperature at injection start 820 K
Nozzle diameter 0.5 mm
Injection pressure 150 MPa
Injection start setting [deg. ATDC] -5.0 -5.0 -5.0
Injection end setting [deg. ATDC] 15.0 22.0 25.5
The fuel injection pressure curves are displayed in Figure 56. They serve to clarify the injection strategy to adjust the total heat release by the enhanced injection duration.
After homogeniser After injection
Figure 56: Injection pressure curves for emulsified fuels
A comparison of the rate of heat release and the integrated heat releases from diesel, DO-EMF25 and DO-EMF40, is given in Figure 57. The lines show that the ignition delay lengthens with increasing water content, resulting in a small peak in the rate of heat release during the premixed combustion phase. The ignition delay from diesel (black line) of approximately 0.5 deg.
CA, increases with DO-EMF25 (blue line) to about 1.5 deg. CA; with DO-EMF40 (red line), it increases to around 2.5 deg. CA. Furthermore, the emulsion directly affects the maximum peak in the rate of heat release: these values decline with increasing water content. Moreover, the effects on after-burning are clearly visible: although the injection duration is delayed to the expansion stroke with the chosen injection strategy, the after-burning is clearly reduced and shows an improved behaviour with a faster energy conversion at this late combustion stage.
0 20 40 60 80 100 120 140 160 180 200
-20 -10 0 10 20 30 40
MPa
deg. CA
Diesel DO-EMF25 DO-EMF40
Figure 57: Rate of heat release and integrated heat release of diesel, DO-EMF25 and DO-EMF40 in ambient air
The average emissions from the rapid compression expansion machine experiments with diesel, DO-EMF25 and DO-EMF40 are listed in Figure 58. With DO-EMF25, the NOx emission is reduced by only 10% compared to diesel. DO-EMF40 reduces the NOx emission by 22% on average. Hence the NOx reduction through emulsion is lower, as found in most of the literature.
The reason for this could be that the rapid compression expansion machine experiments correspond to a low load: the higher the load, the higher the NOx reduction potential from the emulsions [68] [76]. Moreover, the longer ignition delay, with its peak in the rate of heat release, is expected to cause a penalty in NOx reduction. At ignition timing, more fuel is present and ignites instantly; this will cause high temperatures and hence NOx formation. Unfortunately, the unburned hydrocarbons emission increases by applying the emulsions. However, the higher unburned hydrocarbons emissions may be directly related to the injection duration, delayed to the expansion stroke by using emulsions. Nevertheless, carbon monoxide emissions clearly decrease and reflect the tendency from the rate of heat release where after-burning is visibly reduced.
-5 0 5 10 15 20 25
-0.3 0 0.3 0.6 0.9 1.2 1.5
-20 0 20 40 60
KJ kJ / deg.
deg. CA
Diesel DO-EMF25 DO-EMF40
Improved after-burning
Figure 58: Emission results from diesel, DO-EMF25 and DO-EMF40 in ambient air To analyse NOx formation, the flame temperatures are again estimated by applying the two-colour method. The temperature distributions from diesel, DO-EMF25 and DO-EMF40 are illustrated by Figure 59. A clear temperature decrease in the luminous part of the flame can be seen. With DO-EMF25, the hottest spots are eliminated and the flame temperature is reduced.
DO-EMF40 shows a larger temperature reduction than DO-EMF25, as expected. However, the flame becomes thinner and less luminous. The estimated flame temperature decreases (Figure 59) suggest a higher NOx reduction than the measured values (Figure 58); for an extensive discussion about potential reasons for this it is referred to Section 4.8.
0 200 400 600 800
0 10 20 30 40
NOx [ppm]
% water
Diesel DO-EMF25 DO-EMF40
0 10 20 30 40 50
0 10 20 30 40
HC [ppm]
% water
Diesel DO-EMF25 DO-EMF40
0 10 20 30 40 50 60
0 10 20 30 40
CO [ppm]
% water
Diesel DO-EMF25 DO-EMF40
Figure 59: Two-colour method applied on diesel, DO-EMF25 and DO-EMF40 in ambient air
In order to validate the diesel emulsion experiments with the rapid compression expansion machine, experiments are carried out with the optical two-stroke test engine. The experimental conditions are recorded in Table 14.
Table 14: Experimental conditions for diesel emulsions in the optical two-stroke test engine
Fuel Diesel DO-EMF25 DO-EMF40
Compression pressure at injection start 8 MPa
Compression temperature at injection start 870 K
Nozzle diameter 0.2 mm x 4
Injection pressure 150 MPa
Injection start setting [deg. ATDC] -4.0 -4.0 -4.0
Injection end setting [deg. ATDC] 8.0 10.5 12.0
Again, the two-colour method is used to analyse the photographs. Figure 60 shows the results with diesel, DO-EMF25 and DO-EMF40. The estimated flame temperature decreases correspond well with results from the rapid compression expansion machine. Moreover, as a consequence of the longer ignition delay the flame is visible at a later degree crank angle.
Importantly, the results from the optical two-stroke test engine very clearly convey the tendency just discussed with the rapid compression expansion machine: the luminous flame that is visible and can be analysed with the two-colour method becomes thinner and less pixels can be gauged.
As defined by the injection strategy, the total mass of diesel fuel is injected over a longer time period with emulsions, as a consequence of the constant injection flow rate, although this effect cannot be as large as in Figure 59 or Figure 60. The smaller flame core also means a better mixing of fuel with air; however, unlike with the nitrogen-enriched air technique, the emulsions cause an inhomogeneous dilution in the combustion chamber only within the flame; therefore,
Diesel DO-EMF25 DO-EMF40
0 5 10 15 20
deg. ATDC
hot areas with invisible NOx sources outside the visible flame, for example from premixed combustion, are conceivable
Figure 60: Diesel, DO-EMF25 and DO-EMF40 flame temperatures from the optical two-stroke test engine
In order to investigate the emulsion’s potential to improve the after-burning behaviour of a low-grade fuel and to reduce NOx emission at the same time, tests with LCO emulsions are carried out. First, an emulsion consisting of 100% light cycle oil (the properties of which are given in Table 9) and 40% water is examined in the rapid compression expansion machine: the experimental conditions are listed in Table 15.
Table 15: Experimental conditions for light cycle oil and light cycle oil emulsion
Fuel LCO LCO-EMF40
Compression pressure at injection start 10 MPa Compression temperature at injection start 820 K
Nozzle diameter 0.5 mm
Injection pressure 150 MPa
Injection start setting [deg. ATDC] -5.0 -5.0
Injection end setting [deg. ATDC] 15.0 25.5
Light cycle oil has a long ignition delay, resulting in a high peak in the rate of heat release during the premixed combustion phase, as discussed in Section 4.4. Unfortunately, emulsions further
Diesel
DO-EMF25
DO-EMF40
0 5 10 deg. ATDC
increase the ignition delay as shown with the diesel emulsions in this chapter. Looking at Figure 61, it becomes clear that the ignition delay of LCO-EMF40 is unacceptably long. The ignition delays in the experiments vary in the range of 3.5 – 5 deg. CA. Hence, the use of light cycle oil emulsions with such a high percentage of water would require a countermeasure such as a pilot injection. Experiments with light cycle oil emulsions and a pilot injection are carried out and discussed in Section 4.9. Nevertheless, the rate of heat release in Figure 61 shows a considerable improvement in light cycle oil’s after-burning behaviour. Even with the injection delayed to the expansion stroke, LCO-EMF40 shows a much improved after-burning with a faster energy conversion.
Figure 61: Rate of heat release and integrated heat release of light cycle oil and LCO-EMF40 in ambient air
Looking at the emissions, Figure 62 charts improved combustion behaviour: on average, over 20 experiments, the carbon monoxide emissions decrease from 69 to 45 ppm, equal to 35%.
Therefore, the effect from emulsions to the after-burning is even higher in the case of the low-grade fuel light cycle oil than with diesel fuel. Within the light cycle oil experiments, the unburned hydrocarbons emission slightly decreases or barely changes, again showing the potential of the emulsion to improve combustion. With respect to the NOx emission, the following statement can be made: LCO-EMF40 reduces NOx emission in these experiments only by 15% compared to pure light cycle oil. Hence, the effect is even lower than in the case of diesel and DO-EMF-40. It becomes clear that a light cycle oil emulsion with such high water content cannot fulfil IMO Tier III NOx emission requirements.
-5 0 5 10 15 20 25
-0.5 0 0.5 1 1.5 2 2.5
-20 0 20 40 60
kJ kJ/deg.
deg. CA
LCO
LCO-EMF40
Improved after-burning Ignition delay
unacceptable long
Figure 62: Experimental emission results from light cycle oil and LCO-EMF40 in ambient air
Analysis of the flame temperature distribution (calculated by the two-colour method) again shows a clear temperature decrease when using LCO-EMF40. The result is consistent with the diesel emulsions’ cases: the calculated temperature decrease and the low NOx reduction do not match. Again, the luminous flame becomes thinner with LCO-EMF40 and fewer pixels are available for analysis. The results are shown in Figure 62.
0 200 400 600 800
0 10 20 30 40
NOx [ppm]
% water
LCO
LCO-EMF40
40 42 44 46 48 50
0 10 20 30 40
HC [ppm]
% water
LCO
LCO-EMF40
0 20 40 60 80
0 10 20 30 40
CO [ppm]
% water
LCO
LCO-EMF40
Figure 63: Two-colour method applied on light cycle oil and LCO-EMF40 in ambient air In order to visualise the effect of reduced soot formation within the flame through emulsified fuels, the back diffused laser method is applied. The method and its application to the rapid compression expansion machine are explained in Section 3.4. The experimental condition for the back diffused laser experiments with light cycle oil and LCO-EMF25 are recorded in Table 16. A short digression is made on the effect of the injection pressure, as 150 MPa is rather high.
However, with the future trends of increasing injection pressure, 150 MPa seems to be reasonable for such low-grade fuels.
Table 16: Experimental conditions for back diffused laser method for light cycle oil emulsion experiments in the rapid compression expansion machine
Fuel LCO LCO-EMF25
Air conditions
Temperature 418 K
Oxygen concentration 21.0%
Pressure (constant pressure charge) 0.5 MPa Injection conditions
Nozzle hole diameter 0.2 mm
Injection pressure 150 MPa 150 MPa / 100 MPa
Injection start setting [deg. ATDC] -5.0 -5.0
Injection end setting [deg. ATDC] 15.0 20.0 / 22.0
LCO LCO-EMF40
0
5
10
15
20
deg. ATDC
Back diffused laser images from light cycle oil are compared to those from LCO-EMF25 in Figure 64. At 0 and 3 deg. ATDC, the soot cloud from pure light cycle oil (LCO) has a common typical shape. However, LCO-EMF25 forms much less soot within the flame. The flame contour is cognisable. Some soot is formed along the spray path and growths an increase in the crank angle degree.
Figure 64: Back diffused laser images from light cycle oil and LCO-EMF25 experiments with 150 MPa injection pressure
Reducing the injection pressure from 150 MPa to 100 MPa, as shown in Figure 65, shows clearly an effect of the higher injection pressure: a high injection pressure is a countermeasure for soot formation but causes higher NOx emissions [77]. By reducing the pressure to 100 MPa, soot formation within the flame is clearly visible. However, even with the lower injection pressure combined with the emulsion, the soot formation is still much less than that from pure light cycle oil (LCO), with reference to the left series of Figure 64.
LCO LCO-EMF25
0
3
5
10
15
deg. ATDC
Figure 65: Back diffused laser images from LCO-EMF25 experiments with 150 MPa and 100 MPa injection pressure
The potential to reduce soot formation by emulsified fuels is visualised and validated as a successful countermeasure. Hence, one difficulty in applying light cycle oil as a fuel could be solved. In this chapter, it is shown that emulsions have the potential to improve after-burning behaviour and reduce the soot formation of the low-grade fuel light cycle oil. However, the over-long ignition delay requires a countermeasure such as a pilot injection.