Combustion, performance and emissions of a DI-CI engine running on Karanj methyl ester: influence of injection timing

Abstract

This study aims at evaluation of the effect of injection timing on the combustion, performance and emissions of a small power diesel engine, commonly used for agriculture purpose, running on pure bio-diesel, prepared from Karanj (Pongamia pinnata) vegetable oil. The effect of varying injection timing was evaluated in terms of thermal efficiency, specific fuel consumption, power and mean effective pressure, exhaust temperature, cylinder pressure, rate of pressure rise and the heat release rate. Furthermore, the effects on emissions of unburnt hydrocarbons, oxides of nitrogen, carbon monoxide and carbon dioxide and smoke were also studied. It was found that retarding the injection timing by 3° enhances the thermal efficiency by about 8.2% with reduction in emission of oxides of nitrogen.

Keywords:

• injection timing
• Karanj methyl ester
• heat release rate
• emissions

1. Introduction

For diesel engines, a significant research effort has been directed towards using vegetable oils and their derivatives as fuels. Non-edible vegetable oils (in natural or modified form) fall in the category of biofuel. There exist a number of vegetables/plants, which produce oil and hydrocarbon (HC) substances as part of their natural metabolism. These vegetable oils from oil seed crops such as soybean, sunflower, groundnut and mustard, and oil seed from tree origin have got 90–95% energy value of diesel on volume basis and comparable cetane number and can substitute the mineral diesel ranging between 20 and 100%. Bio-diesel is considered a promising alternative fuel for use in diesel engines, boilers and other combustion equipment. The fuel characteristics of bio-diesel are approximately the same as those of fossil diesel fuel and thus may be directly used as a fuel for diesel engines without any modification of the design or equipment. In addition, these are bio-degradable, can be mixed with diesel in any ratio and are free from sulphur.

Although bio-diesel has many advantages over diesel, there are several problems that need to be addressed such as its lower calorific value, higher flashpoint, higher viscosity, poor cold flow properties, poor oxidative stability and sometimes its comparatively higher emission of nitrogen oxides (NO _x ) (Lin and Lin 2006). Bio-diesel obtained from some feed stocks might produce slightly more oxides of nitrogen (1–6%), which is an ozone depressor, than fossil origin fuels, but can be managed by blending bio-diesel and high-speed diesel (Yohaness 2003). It is found that the lower concentrations of bio-diesel blends improve the thermal efficiency. Reduction in emission and brake-specific fuel consumption (BSFC) is also observed while using 10% bio-diesel blend (B10) (Ramadhas et al. 2005, Panwar et al. 2010).

Since the introduction of petroleum fuels, the development of compression ignition (CI) engines has been done keeping the properties of ‘mineral diesel’ in front. The present designs and operating parameters of available engines are standardised for this fuel only. For all other fuels, the operating parameters must be reset in light of the specific fuel properties. Effect of injection parameters (Badami et al. 2002, Payri et al. 2002, Sayin et al. 2008, Jindal et al. 2010b), spray (Watanabe et al. 1998), injection timing and compression ratio (Parlak et al. 2003, Al-Baghdadi 2004, Sayin et al. 2007, Raheman and Ghadge 2008) has been studied in detail at many places. Most of the research studies concluded that in the existing design of engine and parameters at which engines are operating, a blend 20% of bio-diesel with diesel works well (Ramadhas et al. 2005). Many researchers indicated the need of research in the areas of engine modifications so as to suit higher blends without severe drop in performance so that the renewability advantages along with emission reduction can be harnessed to a greater extent (Jindal 2010).

It is commonly accepted that there is some advancement of injection time when bio-diesel is used in place of diesel because of its bulk density. The higher bulk density and viscosity transfers the pressure wave through fuel pipe lines faster and an earlier needle lift will lead to advanced injection. Due to the difference in cetane number, it is often suggested that injection timing be retarded to attain more complete combustion of vegetable oil-based fuels (Randall von, 1999). Late injection of fuel into the combustion chamber helps in reducing the NO _x emission of a diesel engine (Ueki and Miura 1999).

Bio-diesel made from different feed stocks has been tried by many and the effect of feedstocks on engine performance and emissions is well documented (Jindal et al. 2010a, Panwar et al. 2010). One of the major feedstocks researched in India is ‘Pongamia pinnata’, popularly known as ‘Karanj’. Looking at its availability and its bio-diesel potential, this oil is becoming more and more popular in many other countries as well. Evaluation of Karanj esters (Agarwal and Rajamanoharan 2009, Jindal et al. 2010a) indicates its superiority over many other vegetable oils in terms of engine performance, emissions, ease of use and availability.

Karanj (P. pinnata) is an underutilised plant which is grown in many parts of India (Figure 1). Sometimes the oil is contaminated with high free fatty acids (FFAs) depending upon the moisture content in the seed during collection. It is a fast growing leguminous tree with the potential for high oil seed production and the added benefit of an ability to grow on marginal land. The potential of Karanj oil (Karmee and Chadha 2005, Sharma and Singh 2007, Srivastava and Verma 2007, Naik et al. 2008) as a source of fuel for the bio-diesel industry is well recognised.

Figure 1 Plant, fruit and seed of Karanj (P. pinnata).

An effort is made in this study to evaluate the effect of varying the injection timing on the combustion, performance and emission of a 3.5 kW engine fuelled with methyl ester of this oil for establishing the appropriate injection timing. The aim was to establish the modifications required in small, constant speed, direct injection diesel engines used extensively for agricultural applications so that these can be made to run on pure bio-diesel with better performance and at the same time to improve the emission quality.

2. Experiment and procedure

In this study, the selected vegetable oil was transesterified and the major properties were evaluated. Furthermore, the evaluation of methyl ester was done in a CI engine for combustion, performance and emission at different injection timing.

2.1 Transesterification

The transesterification of the oil sample was carried out in the laboratory using standard procedures adopted commonly throughout the world (Van Gerpen et al. 2004). As the Karanj oil contained higher percentage of FFAs, a two-stage transesterification was adopted. The FFA was reduced first by acid-catalysed esterification (using methanol in the presence of sulphuric acid) followed by alkali-catalysed esterification (using methanol in the presence of KOH). After the separation of glycerol, we water washed the ester to remove unreacted methoxide. It was then heated to remove the water traces to obtain clear bio-diesel.

The properties of the so prepared bio-diesel were evaluated in the laboratory using standard test procedures as per ASTM/BIS and are listed in Table 1. The properties evaluated were relative density (by standard RD bottles of 50 ml capacity), calorific value (by adiabatic bomb calorimeter), Kinematic viscosity (by Redwood No. 1 viscometer), flashpoint (by Pensky-Marten closed cup apparatus), cloud and pour points, FFA contents (by chemical titration method) and iodine value (using Wij's solution).

Table 1 Evaluated properties of Karanj oil and its methyl ester.

				Karanj
Property	Unit	Bio-diesel BIS Std	Diesel	Oil	Methyl ester
Relative density	gm/cm^3	0.87–0.90	0.8394	0.9317	0.8749
Calorific value	MJ/kg	–	44.13	39.54	39.19
Kinematic viscosity
at 32°C	cSt	3.5–5.0	3.93	46.03	7.52
at 60°C	cSt		2.77	18.3	3.2
Cloud point	°C	–	< 3	9	2
Pour point	°C	–	< − 5	6	− 2
Flashpoint	°C	>100	56	226	122
FFA	%	< 0.8	–	10.73	0.50
Iodine value		< 115	–	78.44	82.66

2.2 Experimental set-up

The study was carried out in the laboratory on an advanced fully computerised experimental engine test rig comprising a single cylinder, water-cooled, four-stroke diesel engine (3.5 kW), commonly used in agriculture sector for minor irrigation needs, connected to eddy current-type dynamometer for loading. The set-up (Figure 2) includes necessary instruments for online measurement of cylinder pressure, injection pressure and crank angle. One Piezo sensor is mounted on the engine head through a sleeve and another mounted on fuel line near the injector for measurement of pressures. The set-up has transmitters for air and fuel flow measurements, process indicator and engine indicator. Rotameters were provided for cooling water and calorimeter water flow measurement. Provision is also made for online measurement of temperature of exhaust, cooling water and calorimeter water inlet and outlet and load on the engine. These signals are interfaced to a computer through the data acquisition system, and the software displays the P–θ and P–V diagrams. Windows-based engine performance analysis software package ‘Enginesoft LV’ is used for online performance evaluation. The set-up enables the study of engine performance for power, mean effective pressure, thermal efficiency, specific fuel consumption, A/F ratio and heat balance. The specifications of the engine and instrumentation used are given in Table 2.

Figure 2 Engine test set-up.

Table 2 Test engine details.

Item	Make/model/specification
Engine
Make and type	Kirloskar (TV1), single cylinder, DI, four stroke, water cooled
Bore and stroke	87.5 mm × 110 mm
Cubic capacity	0.661 L
Compression ratio	17.5:1
Rated power	3.5 kW at 1500 rpm (load at rated power, 12 kg)
Injector opening pressure	210 bar
Injection timing	23° BTDC static (diesel)
Instrumentation
Dynamometer	Eddy current type, Model AG10 of Saj Test Plant Pvt Ltd
Cylinder pressure sensor	Piezo sensor of PCB Piezotronics, Inc., Model-M111A22; resolution-0.1 psi; sensitivity-1 mV/psi
Fuel pressure sensor	Piezo sensor of PCB Piezotronics, Inc., Model-M108A02; Resolution-0.4 psi; sensitivity-0.5 mV/psi
Load measurement	Load cell – sensortronics make, Model 60001 with digital indicator, range 0–50 kg, supply 230 V AC
Fuel flow measurement	Differential pressure transmitter, make-Yokogawa; Model-EJA110A-DMS5A-92NN
Air flow transmitter	Make-Wika; Model-SL1
Temperature sensor	Type RTD, PT100 and thermocouple, type K
Crank angle sensor	Digital encoder – resolution 1°, speed 5500 rpm with TDC pulse
Engine indicator	Input: Piezo sensor (cylinder pressure and injection pressure), crank angle sensor, no. of channels 2, Communication RS232
Software	‘Enginesoft LV’ Engine performance analysis software (on NI platform)

2.3 Emission measurement

The exhaust gases were sampled from the exhaust line through a specially designed arrangement for diverting the exhaust to the sampling line without increasing the back pressure and were then analysed using a portable multi-gas analyser (make-MRU Airfare, Germany; Model-DELTA 1600 S). It measures carbon monoxide (CO), carbon dioxide (CO₂) and unburnt HC based on a non-dispersive infrared technology. NO _x are measured with the help of electrochemical cells. The range and resolution of the gas analyser is given in Table 3. For the measurement of smoke intensity of the diesel engine's exhaust, we used a diesel smoke meter (model-DSM 2000; make-Manatec Electronics). It is based on the principle of absorption of light, which is an indicative parameter of the level of smoke present in the exhaust sample from a diesel engine. The light passes through a smoke chamber and onto the photodiode which continuously senses the intensity of light incident on it, and converts the light into an electrical signal, which is further processed by signal handling circuits. The output signal is fed finally to a micro controller to give digitised readings. The amount of light falling on the detector will depend upon the opacity of the smoke in the tube, because the smoke tends to obstruct the light more when it is more opaque. The smoke level is displayed in terms of per cent opacity (Hartridge Smoke Unit).

Table 3 Measurement range and resolution of Delta 1600 S gas analyser.

Measured gases	Measured range	Resolution
HC	0–20,000 ppm	1 ppm volume
CO	0–10%	0.01% volume
CO2	0–16%	0.1% volume
NO x	0–5000 ppm	1 ppm

2.4 Experimental procedure

The performance test of the engine was conducted as per IS: 10000 [P:5]:1980. Initially, the engine was run on no-load condition for 20 min for ‘running in’ so that the temperature of the engine coolant reached steady state, after which its speed was adjusted to 1600 ± 10 rpm. The engine was then tested at 5% (minimum load with eddy current dynamometer connected), 25, 50, 75, 100 and 125% loads. For each load condition, the engine was run for at least three minutes after which data were collected. The experiment was replicated three times. Analysis of variance was done to ascertain the significance of cause–effect relationships, and the summary is presented in Table 4. For pressure history, three cycles were recorded and a mean of these was computed for better representation. For all settings, the emission values were recorded thrice and a mean of these was taken for comparison. The performance of the engine at different loads and settings was evaluated in terms of brake thermal efficiency (BTHE), BSFC, indicated power (IP) and brake power (BP), exhaust temperature, indicated mean effective pressure (IMEP), cylinder pressure (P _c), rate of pressure rise (dP/dθ), net heat release rate (dQn/dθ) and emissions of CO, CO₂, unburnt HC and oxides of nitrogen (NO _x ) with exhaust gas opacity. The software enables evaluation of performance from the acquired data using standard relationships. The BTHE is evaluated using the expression BTHE = (BP × 3600 × 100/volumetric fuel flow in 1 h × fuel density × calorific value of fuel). Similarly, BSFC is evaluated on the basis of fuel flow and BP developed by the engine using the expression BSFC = (volumetric fuel flow in 1 h × fuel density/BP). The indicated work done per cylinder per cycle (area of indicator diagram × scale factor × 10⁵) and the IP (indicated work done per cycle × speed/2 × 10^− 3) is computed from the area of indicator diagram.

Table 4 Analysis of variance summary.

Sources of variation	SFC	BTHE	Ex. temp.	IMEP	IP	HC	CO	NO x	CO2	Opacity
Test of between-subject effect
Injection timing	**	**	**	**	**	**	**	**	NS	**
Note: **Significant at 0.01 level; NS, non-significant.

The IMEP is a measure of the indicated work output per unit swept volume, in a form independent of the size and number of cylinders in the engine and engine speed, and is computed as IMEP = indicated work output per cylinder per cycle/swept volume per cylinder. The rate of pressure rise (Δp _c ^*  = Δp _c  × V _i /V _c ) and net heat release rate (dQn/dθ = (γ/γ − 1) ×  p × (dV/dθ)+(1/γ − 1) × V × dp/dθ) are computed using the cylinder pressure history (p–θ) (Stone 1992).

As injection timing plays a crucial role in the start of combustion and quality of combustion, the effects of varying the timing were studied for both advancement and retardation. For changing the injection timing in a jerk-pulse pump, we fitted the pump with different number of shims under the pump body. The standard setting of the engine used for test is with three shims to give standard injection timing of 23° before top dead centre (BTDC) as recommended by the manufacturer. Every single shim, of thickness 0.21 mm, deviates the injection timing by about 3°. The study was done with 3° advancement, normal and 3° and 6° retarded timing.

3. Results and discussion

The effect of transesterification on major properties of the oil is given in Table 1. The relationships between independent variables (load and injection timing) and dependent variables are shown in the figures and the overall effects of injection timing on combustion, engine performance and emissions are discussed in this section.

3.1 Effect on BTHE

At full load, the thermal efficiency is 26.49% with diesel and 24.03% with pure Karanj methyl ester (KME). Lower thermal efficiency (8–10%) is observed in almost all studies with bio-diesel due to lower heating value and higher density (Jindal et al. 2010a). The effect of injection timing on engine performance is significant. It can be seen from Figure 3 that retarding the injection timing by 3° increases the thermal efficiency remarkably. Further retardation is not so beneficial, whereas advancement of injection is not desirable as it leads to drop in thermal efficiency of the engine. With KME as fuel, the thermal efficiency at full load increases from 24.03 to 26.0% on retarding by 3° and to 24.57% on retarding by 6°. On advancing the injection by 3°, the thermal efficiency drops to 23.82%. About 8.2% improvement in thermal efficiency is obtained by retarding the injection timing by 3°. At advanced injection timing, more of the fuel is injected and injection starts early in the cycle leading to earlier pressure rise before the piston reaches TDC position. Greater pressure rise in the compression stroke increases the negative work and consumes the momentum of flywheel. With the reduction in net work output and increased fuel consumption, the thermal efficiency has to drop. Zeng et al. (2006) also reported an increase in thermal efficiency and a corresponding decrease in fuel consumption on retarding the injection timing upto a point after which the trend reversed. They stated that there is an optimum timing at which engine delivers best efficiency which depends on the combustion properties of the fuel.

Figure 3 Effect of injection timing on BTHE.

3.2 Effect on BSFC

At full load, the specific fuel consumption is 0.29 kg/kWh with diesel, and 0.34 kg/kWh with pure KME. The BSFC is also affected by changes in the injection timing corresponding to the changes in thermal efficiency. With the advancement of the injection timing, the specific fuel consumption increases, whereas retarding leads to improvement (Figure 4). With KME as fuel, the BSFC value increases to 0.36 from 0.34 kg/kW h on advancing the injection by 3° and it decreases to 0.32 and 0.33 kg/kW h on retarding the injection by 3° and 6°, respectively. On retarding the injection, the delay period increases but fuel delivery to the cylinder reduces with a higher mean effective pressure in the cycle maintaining the power, thereby reducing the specific fuel consumption. On further delay in injection, unburnt fuel gets exhausted, whereas on advancing the injection, shorter delay with sharp rise in pressure reduces the mean effective pressure. Nwafor (2007) observed increased fuel consumption on advancing the injection timing in a natural gas engine and recommended not to advance the injection under high loading conditions. Parlak et al. (2005) observed reduction in BSFC by 6% on retarding the engine by 4°.

Figure 4 Effect of injection timing on BSFC.

3.3 Effect on exhaust temperature

The temperature of exhaust is higher with diesel than that with KME at standard injection timing. However, the trend line for exhaust temperature with different injection timing indicates an increase in temperature of exhaust gases with retarded injection (Figure 5). As the combustion is delayed and more heat is released in mixing controlled combustion regime, greater amount of heat escapes with exhaust gases. With advanced injection, wall heat transfer is more due to earlier combustion in the cycle leading to lower exhaust temperature.

Figure 5 Effect of injection timing on exhaust temperature.

3.4 Effect on power and mean effective pressure

The effect of injection timing on IMEP, IP and BP is shown in Figure 6. At standard timing, the IMEP is higher with diesel than with KME and hence IP is also high. On advancing the injection, the mean effective pressure in the cycle drops further. This results in lower IP. The IP of the engine increases slightly on retarding the injection by 3°, whereas it decreases on further retarding to 6° or advancing by 3°. When the injection is retarded by 3°, better mean pressure is obtained and the engine runs smoother. The IP and mean effective pressure with KME increases to 5.64 kW and 6.65 bar from 5.19 kW and 6.02 bar, respectively on retarding the injection by 3°. The IP changes to 5.35 and 5.0 kW; and mean effective pressure changes to 6.23 bar and 5.80 bar at 6° retardation and 3° advancement, respectively. Similar increase in brake mean effective pressure on retarding the injection was reported by Zeng et al. (2006) as well. BP of the engine is little affected by the change in injection timing and remains almost the same at all selected timing accept at 6° retardation.

Figure 6 Effect of injection timing on IMEP, IP and BP.

3.5 Effect on cylinder pressure, rate of pressure rise and net heat release rate

The peak cylinder pressure attained with pure KME is higher and the rate of pressure rise is less than that with diesel at standard timing. But the pressure rise is advanced to the extent that it rises even before the piston has reached to TDC causing loss in output power and hence poorer efficiency. Figure 7 represents the effect of injection timing on cylinder pressure, rate of pressure rise and net heat release rate at full load conditions. With changes in injection timing, as expected, the in-cylinder pressure, rate of pressure rise and net rate of heat release also changes. On advancing the injection, the pressure in the cylinder reaches a higher value than the retarded injection scheme. This is mainly due to the fact that, on advancing the injection, larger amount of fuel is injected (injection starting earlier and stopping later). Higher pressure is also found before TDC with advancement due to early start of combustion.

Figure 7 Effect of injection timing on cylinder pressure, rate of pressure rise and net heat release rate at full load.

With the advancement of injection timing, the peak rate of pressure rise increases but it shifts before TDC (358° crank angle) with shorter delay period. On retarding the injection, the rate of pressure rise decreases slightly with a shift away from TDC, and the ignition delay also increases.

Similar effects are seen on the rate of heat release. With the advancement of injection by 3°, the peak rate of heat release is at 358° crank angle and on retarding by 3° and 6°, the peak heat release rate is found at 362° and 363°, respectively against 361° with standard timing. With retardation, larger amount of heat is released in mixing controlled combustion regime resulting in higher mean pressure in the cycle. Others also reported similar results of shifting pressure rise rates and heat release locations (Zeng et al. 2006, Raheman and Ghadge 2008).

3.6 Effect on engine emissions

The effect of injection timing on emissions of HC, CO and CO₂ is shown in Figure 8. HC emissions are significantly affected by change in injection timing of the engine. It is observed that on advancing the engine, HC emissions reduce upto 3° advancement after which the same increases, whereas on retarding the engine, HC emissions increases. Lowest HC emissions are found on 3° advancement. Advanced injection leads to earlier ignition with shorter delay and higher rate of heat release by combustion of HC to a greater extent. Further advancing the injection leads to erroneous running of the engine with higher amounts of HC going to exhaust.

Figure 8 Effect of injection timing on emissions of HC, CO and CO₂ at full load.

On retarding the injection, late burning of fuel due to longer delay and more of mixing controlled combustion leaves some HC unburnt. Retarding the injection by 3° leads to little increase, but further retardation increases the HC to a large extent. With KME, HC increases from 23 to 38 ppm at 3° retarded injection.

CO emissions are highest at standard timing and reduce significantly on advancement. It is also found to be lowest at minimum HC emission timing that is 3° advance, indicating better conversion efficiency. The emissions of CO reduce to 0.06% at 3° advance from 0.10% at standard timing with KME as fuel, whereas it remains unchanged on retarding the injection. The CO₂ emissions are found to be lowest at standard timing which increases with both advancement and retardation. On retarding the injection by 3°, the change is insignificant. Sayin et al. (2008) also studied the effect of injection timing on engine emissions and reported the reduction in HC and CO on advancing the engine injection and the increase in CO₂ emissions on retarding the injection.

Figure 9 represents the effect on emissions of NO _x and smoke. As reported by many studies mentioned in the review, NO _x emission is found to vary with injection timing. The advancement of injection time enhances NO _x emissions, whereas retarding the injection helps reduce the same. Parlak et al. (2005) reported 40% reduction in NO _x with 4° retardation of injection. The higher bulk density and viscosity transfers the pressure wave through the fuel pipe lines faster and an earlier needle lift leads to advanced injection. With additional advancing, the effect is magnified and leads to still higher flame temperature causing an increase in NO _x emission. Retarding the engine compensates for the bulk density effect and hence lower emissions are observed. The results are in line with those obtained by others (Verbiezen et al. 2007, Jindal et al. 2010b).

Figure 9 Effect of injection timing on emissions of NO _x and smoke at full load.

The smoke opacity of engine exhaust is found to decrease in both the cases of advancement or retardation. Minimum opacity is measured for 3° advancement. On further advancing the injection, the opacity starts increasing. At 3° retarded injection, with KME as fuel, smoke opacity decreases from 19.0 to 17.9%. With better mixing controlled combustion, homogeneous combustion in controlled combustion regime reduces the particulate matter in engine exhaust due to fuller combustion of carbon components of the fuel.

4. Conclusion

The fuel properties of bio-diesel are comparable with those of diesel, and lower blends with diesel are found suitable even for long-term uses. Higher blends are still away from acceptance due to poor performance, mainly due to the reason that the present age engines are the result of extensive research keeping only petro-diesel as fuel in mind. Bio-diesel being a fuel of different origin and quality, the engine design needs revision and different settings for optimum performance.

As the combustion advances with pure bio-diesel due to early entry, retarding the injection timing by 3° is found to increase the thermal efficiency by 8.2% and reduce the specific fuel consumption by 9% when KME is used as fuel. Highest exhaust temperature and IP are obtained on 3° retarded injection. By retarding the injection, the fuel delivery is also reduced resulting in slightly lower pressure rise with peak shifting towards outward stroke reducing the negative work. Although lowest HC, CO and smoke emission is found with 3° advancement because of early entry and longer stay, NO_x emission is lower at 3° retarded timings.

Acknowledgements

This study was conducted as a part of a research project sponsored by Petroleum Conservation Research Association, New Delhi. The author fully acknowledges the financial support extended by the agency.