The effect of ethanol–diesel–biodiesel blends on combustion, performance and emissions of a direct injection diesel engine

The effect of ethanol–diesel–biodiesel blends on combustion, performance and emissions of a direct injection diesel engine

Energy Conversion and Management 79 (2014) 698–720 Contents lists available at ScienceDirect Energy Conversion and Management journal homepage: www...

5MB Sizes 0 Downloads 36 Views

Energy Conversion and Management 79 (2014) 698–720

Contents lists available at ScienceDirect

Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman

The effect of ethanol–diesel–biodiesel blends on combustion, performance and emissions of a direct injection diesel engine Gvidonas Labeckas ⇑, Stasys Slavinskas, Marius Mazˇeika Power and Transport Machinery Engineering Institute at Aleksandras Stulginskis University, Student Str. 15, P.O. Box LT-53361, Kaunas Academy, Lithuania

a r t i c l e

i n f o

Article history: Received 25 July 2013 Accepted 30 December 2013 Available online 1 February 2014 Keywords: Diesel engine Ethanol Biodiesel Combustion Heat release Comparative performance Emissions Smoke opacity

a b s t r a c t The article presents the test results of a four-stroke, four-cylinder, naturally aspirated, DI 60 kW diesel engine operating on diesel fuel (DF) and its 5 vol% (E5), 10 vol% (E10), and 15 vol% (E15) blends with anhydrous (99.8%) ethanol (E). An additional ethanol–diesel–biodiesel blend E15B was prepared by adding the 15 vol% of ethanol and 5 vol% of biodiesel (B) to diesel fuel (80 vol%). The purpose of the research was to examine the influence of the ethanol and RME addition to diesel fuel on start of injection, autoignition delay, combustion and maximum heat release rate, engine performance efficiency and emissions of the exhaust when operating over a wide range of loads and speeds. The test results were analysed and compared with a base diesel engine running at the same air–fuel ratios of k = 5.5, 3.0 and 1.5 corresponding to light, medium and high loads. The same air–fuel ratios predict that the energy content delivered per each engine cycle will be almost the same for various ethanol–diesel–biodiesel blends that eliminate some side effects and improve analyses of the test results. A new approach revealed an important role of the fuel bound oxygen, which reflects changes of the autoignition delay more predictably than the cetane number does. The influence of the fuel oxygen on maximum heat release rate, maximum combustion pressure, NOx, CO emissions and smoke opacity of the exhaust is highly dependent on the air–fuel ratio and engine speed. Fuelled with blend E15B the diesel engine develops the brake thermal efficiency of 0.362, i.e. the same as a straight diesel running on slightly richer air–fuel mixture k = 1.5 at rated 2200 rpm speed. Adding of the ethanol to diesel fuel reduces the NOx and the HC emissions for richer combustible mixtures whereas the influence of a higher ethanol mass content on CO emissions and smoke opacity depends on the air–fuel ratio and engine speed. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction One of the biggest challenges of the 21st century is the rapidly growing automotive industry with increasing demand of mineral oil products such as petroleum and commercial diesel fuel. To produce enough diesel fuel technically is feasible but every next barrel of crude oil is getting farther, deeper and harder with a higher extraction, production, and delivery price. The fossil fuel basins have been depleted over the past two industrialisation centuries and enormous amount of heat loss energy and harmful engine emissions were released into the atmosphere. As a number of passenger cars and heavy-duty trucks on the market increases along with power plants and home furnaces, the greater than before fuel consumption and air pollution by the greenhouse carbon dioxide become an ever-increasing problem. High urban population and urbanisation have changed the lifestyle of peoples’, contributed

⇑ Corresponding author. Tel.: +370 37 752 285; fax: +370 37 752 311. E-mail addresses: [email protected] (G. Labeckas), [email protected] lt (S. Slavinskas), [email protected] (M. Mazˇeika). 0196-8904/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.enconman.2013.12.064

to climate changes, quickened ice caps melting, endangered survival of unique sea creatures and wild animals too. The EU Directive 2009/28/EC approved a target of a 20% share of renewable biofuels in overall transport petrol and diesel consumption by 2020 to be introduced in a cost-effective way. The principal reasons for using of biofuels are to lower greenhouse gas emissions, increase farm income, promote rural development and diversification, create a new dynamics in a global agricultural market, increase energy independence and reduce consumer reliance on imported fossil fuels. Alcohol-based fuels have been important energy sources since the 1800s. As early as 1894, France and Germany were using ethanol in internal combustion engines [1]. The first investigations on the use of ethanol in diesel engines were carried out in South Africa in the 1970s and continued in Germany and the USA during the 1980s [2]. There is a commitment of the USA government to increase bioenergy threefold in 10 years, which has added impetus to the search for variable biofuels [3]. Between 2000 and 2010 ethanol production increased from 16.9 to 72.0 billion litres while biodiesel grew from 0.8 to 14.7 billion litres [4]. After 2020, 2nd-generation bioethanol could become a much more significant

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

699

Nomenclature

u ui si k p t

crank angle, degree autoignition delay period, degree autoignition delay time, s air–fuel ratio by mass gas pressure in the cylinder, bar temperature, °C

Abbreviations E ethanol B biodiesel, E15B – RME additive CN cetane number RME rapeseed oil methyl ester E5 5 vol% ethanol/95 vol% diesel fuel E10 10 vol% ethanol/90 vol% diesel fuel E15 15 vol% ethanol/85 vol% diesel fuel E15B 15 vol% ethanol/80 vol% diesel fuel/5 vol% rapeseed oil methyl ester O2 fuel oxygen mass content, wt% TDC top dead center ATDC after top dead center BTDC before top dead center CAD crank angle degree

player in a global biofuels market [5]. Anhydrous ethanol has the advantage over methanol because of having higher miscibility with the diesel fuel and of being of renewable nature [6]. Modern technologies and efficient bio-energy conversion are becoming costwise competitive with fossil fuels, therefore the biofuel economy based on agricultural production will grow rapidly during the 21st century [7]. The experimental studies on the ethanol using in diesel engines have been carried out by Shropshire and Goering [8], Hansen et al. [9–11], Can et al. [12], Zannis et al. [13], Satge de Caro et al. [14] and other researchers [15–18]. Several methods are used, which are known as alcohol fumigation into the intake air ports [19–21], application of dual injection systems [8], using of alcohol-diesel fuel micro-emulsions [22] and preparation of the alcohol-diesel fuel blends [16,19]. The biggest advantages of the blending are the convenience in application and absence of engine modification, however, in case of using hydrous ethanol (190 proof) with the normal diesel fuel at blending ratios of 5/95 and 10/90 by volume, an emulsifying additive ‘‘GE Betz’’ as co-solvent or isopropanol as a heavy alcohol C3H8O is recommended in proportion from 1 vol% to 1.5 vol% to improve miscibility of the ethanol–diesel fuel mixture at lower temperatures, satisfy homogeneity and prevent phase separation [12,23]. Ethanol solubility in the diesel fuel depends on the hydrocarbon composition of the diesel fuel, temperature, content of water and wax in the blend and ambient humidity. To improve miscibility, Asfar and Hamed [22] added a stabiliser (isobutanol) of 3.7 vol%, 7 vol% and 13 vol% to tested 5 vol%, 10 vol% and 20 vol% ethanol– diesel fuel blends. The homogeneous mixture of a diesel, biodiesel and ethanol fuels can also be prepared by mixing it with a magnetic stirrer [24] or a mixer can be mounted inside the fuel tank to prevent phase-separation in straight ethanol–diesel fuel blending [25]. To improve blending of the ethanol with the diesel fuel Abu-Qudais et al. [19] used 0.5 vol% of emulsifier that consisted of a styrene–butadiene copolymer and a polyethyleneoxide–polystyrene copolymer. In that research the optimum percentage for ethanol–diesel fuel blends was 15%, which suggested an increase of 3.6% in brake thermal efficiency, 43.3% in CO emissions, 34% in HC and a reduction of 32% in soot mass concentration.

SOI SOC HRRmax AHRRmax

start of injection, CADs start of combustion, CADs maximum heat release rate, kJ/(m3deg) crank angle corresponding to maximum heat release rate, CADs AI 50 crank angle corresponding to 50% heat release, CADs AI 90 crank angle corresponding to 90% heat release, CADs pmax maximum pressure in the cylinder, MPa Apmax crank angle corresponding to maximum pressure in the cylinder, CADs (dp/du)max maximum pressure gradients in the cylinder, bar/° CA bmep brake mean effective pressure, MPa bsfc brake specific fuel consumption, g/kW h bte brake thermal efficiency NO nitric oxide, ppm NO2 nitrogen dioxide, ppm NOx total nitrogen oxides, ppm CO carbon monoxide, ppm CO2 carbon dioxide, vol% HC unburned hydrocarbons, ppm

In contrast to the above studies, anhydrous ethanol (200 proof) does not need an emulsifying agent to form a transparent solution in diesel fuel, but these solutions can tolerate only up to 0.5% of water [23]. Other researchers also provided tests with various ethanol–diesel fuel blends without using of any additives [26], except in some cases the ignition improver Hicet 3A (0.16% 2-ethylhexyl nitrate) [9] or cetane improver (0.2% iso-amyl nitrite) [18] were used to increase cetane number, improve autoignition and combustion of oxygenated blends. Satge de Caro et al. [14] determined that adding of the 20 vol% of ethanol to diesel fuel decreased cetane number and led to the longer autoignition delay, increased cyclic irregularity and augmented CO and HC emissions. To avoid phase separation between two fractions the rapeseed oil methyl ester from 5 vol% to 10 vol% [27] or soybean oil methyl ester from 2 vol% to 10 vol% [28], as co-solvents, can be added to ethanol–diesel fuel blends. Experiments of a four-cylinder, turbocharged, DI diesel engine fuelled with the 10 vol% and the 15 vol% ethanol–diesel fuel blends revealed that the ethanol addition reduces engine power by 12.5% and 20%, the CO, the SO2 emissions and soot density because of the higher oxygen mass content and lower C/H ratio in ethanol–diesel fuel blends. However, low cetane number and long ignition delay, followed by higher pressures and temperatures inside the cylinder, resulted in slightly higher NOx emission at full load within the speed range of 4500–1500 rpm [12]. Similar changes in the CO emission and higher fuel consumption at adequate loads, concentration of unburned hydrocarbons HC and the total NOx emissions were obtained when running on various diesel–ethanol-1-butanol microemulsions up to 50% by energy content single-cylinder, DI 7.4 kW diesel engine [16]. The reduced cetane number of emulsified fuel blends increased the autoignition delay and caused a higher maximum pressure and combustion temperature in the cylinder that led to higher NOx levels. Experiments with ethanol and rapeseed oil blends showed that the combustion temperature plays an important role in the production of NOx [29,30]. Since the boiling temperature of ethanol is about 2.3 times lower than the start of vaporisation of diesel fuel, mixing of diesel fuel with a lighter ethanol may advance the start of fuel vaporisation that, on the one part, accelerates preparation of the air and fuel

700

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

mixture but, on the other part, a low cetane number of ethanol can increase the NOx emission due to the longer autoignition delay and higher amount of the fuel premixed for rapid combustion. Hansen et al. [3] showed that the ethanol addition of 10 vol% and 15 vol% reduces the NOx by zero to 4–5%, increases the HC and ambiguously affects the CO emissions, however, both HC and CO were below the regulated emissions limit. Authors [10] determined that when operating on 10 vol% dry ethanol, 1 vol% GE Betz additive and 89 vol% low sulphur No. 2 diesel fuel the engine performance over the 500-hour period was not affected apart from an expected 4% decrease in power caused by the lower heating value of the blend. Other tests of turbocharged, DI diesel engine powered with 10 vol% anhydrous bioethanol (99.9%) and diesel fuel blend showed significant reduction of particulate emissions, with no substantial increase in other emissions [26]. Numerical investigations of the effects of oxygenated blends have been conducted on DI and IDI diesel engines operating at constant load [13,24], torque and/or brake mean effective pressure [2,9] and speed [31,32]. Despite a large number of the scientific works conducted on ethanol–diesel–biodiesel blends there is a lack of references dealing with changes in the autoignition delay, combustion peculiarities, maximum heat release rate, performance efficiency, NOx, CO, HC emissions and smoke of the exhaust produced at overall the same air–fuel ratios. It is important to note that the fuel energy delivered per each engine cycle will be almost the same for various ethanol–diesel–biodiesel blends tested at overall identical air–fuel ratios that suggest an intrinsic difference from previous works on biodiesel. A new systematic approach and well-defined preconditions alleviate separation of main factors, which affect the engine power, the brake specific fuel consumption and emissions of the exhaust, and improve analysis and comparison of the test results. Changes in the fuel mass fraction premixed for rapid burning and engine performance will occur because local air–fuel ratios in the fuel spray patterns and in front of the spray tips will not be the same, accessibility of the air-born oxygen and the fuel bound oxygen in heterogeneous mixture will also be dissimilar. The same air–fuel ratios assist in keeping the burning conditions under control that guarantees strict analysis of the effects of the fuel oxygen and the cetane number on the autoignition delay, maximum heat release rate and emissions produced from oxygenated blends. It is important to clarify relationships between the autoignition delay and the maximum heat release rate, on the one part, and the cetane number and the fuel oxygen content, on the other part, identify eventual reasons, reveal important factors, and achieve a new knowledge in this field of investigation. The purpose of the research was to conduct comparative performance of the diesel engine operating on commercial diesel fuel class 2 as a baseline fuel and ethanol–diesel fuel (E5, E10, E15) and ethanol–diesel–biodiesel (E15B) blends. Objectives of the experimental study are as follows: 1. To examine the effects of the ethanol addition to diesel fuel on the start of injection (SOI), the start of combustion (SOC), the autoignition delay period si (ui), combustion peculiarities, the maximum heat release rate HRRmax and maximum pressure pmax in the cylinder when operating on various ethanol–diesel–biodiesel blends. The analysis and comparison of the engine parameters were performed for the same air–fuel ratios representing overall lean k = 5.5, moderate k = 3.0 and slightly richer k = 1.5 combustible mixtures at three ranges of 1400, 1800 and 2200 rpm speed. 2. To study the effects of the fuel oxygen mass (wt%) content on brake mean effective pressure (bmep), brake specific fuel consumption (bsfc) and brake thermal efficiency (bte) when running on various ethanol–diesel–biodiesel blends at specific

air–fuel ratios and the three ranges of speed. The changing tendencies in emissions of nitrogen oxides NO, NO2, NOx, carbon monoxide CO, dioxide CO2, total unburned hydrocarbon THC, residual oxygen O2, vol% content and smoke opacity of the exhaust were analysed and compared with respective parameters of a basic diesel fuel used at almost the same combustion conditions. The research on ethanol–diesel–biodiesel blends conducted for the same air–fuel ratio, which specifies the ratio of ‘‘supply to demand’’ in combustion, makes a major difference from many works on biodiesel, in which the analysis of the measured experimental data was performed mainly for constant torque, load (bmep) and engine speed. Scarcity of the articles dealing with the effects of ethanol–diesel–biodiesel blends on performance of an engine at the same air–fuel ratios encouraged us to perform this experimental study. This method, which was not very often used before, improves theoretical rigour, removes some side effects on the engine performance, and thus allows more attention to be focused on main influencing factors such as the fuel oxygen mass content and the cetane number of the blend. Much of the work was done on biodiesel and these issues are useful for comparison of the test results obtained in light of the findings of other researchers. Through analysis and discussions about the effects of oxygenated blends on the engine performance and emissions proper ethanol–diesel–biodiesel blend can be selected and valuable conclusions extracted from this experimental study. 2. Engine test set up and research methodology The tests have been conducted on a standard, naturally aspirated, four-stroke, four-cylinder, water-cooled, DI diesel engine D-243. Fig. 1 shows installation of the engine test stand and equipment. The experimental test set up consisted of a diesel engine, an engine test bed, the air and fuel mass consumption measuring equipment, the gas analyser and a smoke meter for the exhaust. The specifications of the tested engine and fuel injection system are listed in Table 1. 2.1. Measurement of in-cylinder gas pressure and engine test data A high speed multi-channel indicating system, which consisted of an angle encoder 365C and high performance piezoelectric pressure sensor GU24D coupled to an AVL indicating amplifier IndiModul 622, was introduced for the recording, acquisition and processing of fast crank-angle and time-based gas pressure signals in the first cylinder. The fuel injection line high pressure was measured with the two piezoresistive high-pressure Kistler Inc. sensors 4067A2000S attached to a high-pressure line at the fuel pump and in front of the injector coupled to the Kistler 2-channel charge amplifiermodule 4665 mounted on the signals conditioning platformcompact 2854A with an accuracy of ±0.5% in the pressure range of 0–100 MPa. The start of fuel injection was determined by using the injector’s nozzle-needle-valve lift history recorded with the Wolff Controls Corporation Hall effects position sensor ASMB 470004-1, which was coupled to the Kistler 2-channel charge amplifier-module 5247 mounted on the signals conditioning platform-compact 2854A, with an accuracy of ±0.5% in the needle-valve lift range of 0–0.28 mm. The fuel injection line high pressure and the needle-valve lift signals have been transmitted from the initial conditioning platform 2854A (SCP) to the AVL IndiModul 622 a fast data acquisition and processing system connected to personal computer. Analyzes

701

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

Air Filter

Gas Analyser

Air-Feed Tank

MAF

Bosch Smoke Meter

4 AVL Air Mass Meter

3

2

AC Dynamometer

1 3

AVL Fuel Balance

Kistler microIFEM AVLIndiModul

AVL Software

Fuel Tank

PC

Fig. 1. Schematic arrangement of the engine test stand: (1) AVL crank-angle encoder; (2) piezoelectric in-cylinder gas pressure transducer; (3) two high-pressure transducers mounted on high-pressure injection line at the fuel injection pump and injector; (4) nozzle-needle-valve lift sensor.

MTZ, Minsk (Belarus) In-line, 4 stroke, 4 cylinder, water-cooled 110 mm  125 mm 4.75 dm3 16.0:1 Toroidal in a piston head, direct injection (DI) 60 kW at 2200 rpm (bmep = 0.689 Mpa) In-line, PP4M9P1g-4201 (Czeck Republic) 9.0 mm  8.5 mm DOP122S532-4164 MP 5  0.34 19.0 ± 0.5 Mpa 5.0 mm  7.0 g 25° CAD before top dead center (BTDC)

and acquisition of the experimental data were performed by using the AVL IndiCom Mobile software. The engine torque was measured with a three phase asynchronous 110 kW AC stand dynamometer KS-56-4 with a definition rate of ±1 Nm. The engine speed was measured by using the AVL crank angle encoder 365C mounted at the front-end of the crankshaft that guaranteed an accuracy of less than ±0.2% of measured value. The air mass consumption was measured by using an AVL air mass meter (0–400 kg/h) installed downstream the air cleaning filter before the air tank to reduce pressure pulsations that guaranteed an accuracy of less than ±1% of measured value. The in-line injection pump supplied the fuel at the optimal advance angle of 25° before top dead center (BTDC), which was the same for the normal diesel fuel and ethanol–diesel–biodiesel blends tested at all loads and speeds. The fuel was delivered into a toroidal combustion chamber in a piston head through five-holes injector’s nozzle with the needle-valve lifting pressure adjusted at 19.0 ± 0.5 MPa. The fuel mass consumption was measured by weighting 100 g of fuel on the AVL dynamic fuel balance 733S flex-fuel with an accuracy of ±0.10%. The single-cycle and summarised over 100 engine cycles the incylinder gas pressure versus crank angle was recorded with an accuracy of 0.1° crank angle degree (CAD). The position of TDC was recorded by using the AVL crank angle encoder 365C with an accuracy of ±0.1° CAD. A piezoelectric uncooled transducer GU24D with the measurements range of 0–280 bar mounted into the head of the first

2.2. Measurement of the autoignition delay period The autoignition delay was determined as a period in units of time (si) and/or in crank angle degrees (ui) between start of injection (SOI) and start of combustion (SOC) with an accuracy ±0.1° CAD. As the start of injection was taken crank angle at which the injector-needle-valve lifts up about 5% of its total value. As the start of combustion was taken crank angle at which the curve of the heat release rate crosses the zero line and changes its value from the minus side to plus one. To improve the accuracy of evaluation 100 single-cycle in-cylinder pressure diagrams versus crank angle were in series recorded for every load-speed setting point of an engine. Thus, the heat release rate was calculated by using averaged over 100 combustion cycles in-cylinder pressure-data, instantaneous cylinder volume, and their first order derivative with

90 80 70

0,4 Autoignition delay

0,35 0,3

60

0,25

50

0,2

40 0,15

30

0,1

20 10

SOI

0,05 0

0 -10 -30

Needle valve lift, mm

Engine producer Engine model Cylinder bore  piston stroke Splash volume Compression ratio Combustion chamber Rated power output Injection pump Plunger diameter  stroke Injector’s type Needle valve lifting pressure Needle diameter  mass Static fuel injection timing

cylinder and connected to the microIFEM piezoelectric amplifiersignal conditioning were used to measure gas pressure for every load-speed setting point with an accuracy of ±0.1 bar within the temperature range of 25–200 °C. The gas pressure inside the cylinder, the heat release rate and nozzle-needle-valve lift history versus crank angle were registered as shown in Fig. 2. The accuracies of the measured experimental data of engine performance and exhaust emission parameters and the uncertainties of the calculated test results are summarised in Table 2.

HRR, kJ/(m3 deg); p cyl , bar

Table 1 Engine specifications.

SOC

-20

-10

0

-0,05 10

20

30

Crank angle degrees Fig. 2. Definition of the autoignition delay: the period expressed in crank angle degrees (CADs) between start of injection (SOI) and start of combustion (SOC).

702

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

Table 2 The accuracy of the measured engine performance and emission parameters and the uncertainty of the computed experimental results. Measurements

Accuracy

Torque Speed Cylinder gas pressure Needle-valve lifting Soot opacity NOx CO CO2 HC Computed results Engine power Fuel mass flow rate Air mass flow rate Brake specific fuel consumption Brake thermal efficiency

±1% ±0.1% ±0.1 Mpa ±0.1° (degree) ±0.1% ±5 ppm ±3 ppm ±0.3 vol% ±2 ppm Uncertainty (%) ±1 ±0.1 ±1 ±1.5 ±1.5

respect to crank angle. The data post-processing Software AVL CONCERTOTM advanced version 4.5, 2013 was used that significantly increased productivity and improved accuracy of the results. 2.3. Measurement of engine emissions The emissions of nitric oxide NO (ppm), nitrogen dioxide NO2 (ppm), carbon monoxide CO (ppm), dioxide CO2 (vol%), total unburned hydrocarbons HC (ppm) and residual oxygen O2 (vol%) in the exhaust were measured by using Testo 350 XL flue gas analyser. The total nitrogen oxide NOx emissions were calculated as a sum of both the NO and NO2 components with an accuracy of ±5 ppm. The smoke density D (%) was measured with a ‘‘Bosch’’ RTT 110 opacity-meter, the readings of which are provided as Hartridge units (% opacity) in a scale range of 0–100% with an accuracy of ±0.1%. The temperature of the exhaust gases was measured with a chromel–kopel thermocouple and an indicator N20 that guaranteed an accuracy of ±0.2 °C within the temperature range of 0–650 °C. 2.4. Design and methodology of the research 2.4.1. Preparation of the fuel blends and analysis of basic parameters The experiments started with the investigation of the engine running on low sulphur conventional diesel fuel (DF) to determine the performance characteristics of the engine and exhaust emissions constituting the ‘‘baseline’’ level that was used to compare with the respective values of parameters obtained when running on ethanol–diesel–biodiesel blends. Three ethanol–diesel fuel blends were prepared by pouring anhydrous (99.8% purity) ethanol into the diesel fuel container in the following proportions by volume 5/95, 10/90 and 15/85 and mixing them to keep the fuel blends in homogeneous conditions. This way, stable oxygenated fuel blends E5, E10, E15 were prepared as shown in Table 4. The fuel oxygen mass (wt%) contents, the stoichiometric air–fuel equivalence ratios and net heating values of the tested fuel blends were determined by taking into account the mixing percentage of each component and respective data of anhydrous (200 proof) ethanol, diesel fuel and RME as listed in Table 3. An extra ethanol–diesel–biodiesel blend E15B was prepared by mixing 15 vol% of anhydrous (200 proof) ethanol, 80 vol% of diesel fuel and 5 vol% of rapeseed oil methyl ester as co-solvent to prevent ethanol–diesel phase separation, improve the cetane number and lubricity of the fuel blend. Park et al. [33] added to diesel fuel 10 vol% of soybean oil methyl ester to prevent phase separation between the ULSD and the ethanol fuel. The Ball On-Cylinder

or BOCLE test results showed that biodiesel, or at least soybean and rapeseed oil methyl esters, have superior lubricity compared to commercial low sulphur diesel fuels [34]. The fuel blend E15B differed from simple ethanol–diesel fuel blends as having the highest fuel oxygen mass content of 6.1 wt% and the lowest carbon to hydrogen (C/H) ratio of 6.45, which ensured the stoichometric air–fuel equivalence ratio of 13.55 kg/ kg and net heating value of 40.26 MJ/kg. As in studies [25,26], the fuel blends lubricity, solubility [11], autoignition [3] and cetane number improvers [17,35], ethanol–diesel fuel blends phase stability improving additives [36] and emulsifiers [16] have not been used that excluded modification of the blends and removed the influence of side-agents on the engine performance, smoke opacity and exhaust emission parameters. 2.4.2. Engine test measurement under the condition of same air–fuel ratio The series of load characteristics were taken over 9 seting points when running on each of the above fuels to have engine performance parameters at various loads covering three different ranges of speed: (1) 1400 and 1800 rpm speeds in-between of which maximum torque occurs; (2) 2200 rpm at which the engine develops rated power. The engine parameters, taken as a function of engine load (bmep), were first situated as a function of the air and fuel ratio (k) which traditionally decreases with the increase of the engine load [49]. The chemical formulae of ethanol, RME and general molecular formula of the normal diesel fuel (Table 3) along with the blending ratio of each fuel component were used to calculate the stoichiometric air–fuel ratios of the fuel blends (Table 4). To determine the values of ‘‘lambda’’ respective stoichiometric the air–fuel ratios of each fuel blend and the air and fuel mass consumptions were taken into account. At first, the engine parameters were plotted as a function of the air and fuel ratio for the normal diesel fuel and ethanol–diesel–biodiesel blends E5, E10, E15 and E15B. Then, the autoignition, combustion, engine performance and emission parameters for each of the above fuel blend were defined to account for respective air–fuel ratios of k = 5.5, 3.0 and 1.5 used at three different ranges of speed. The energy contents stored in combustible mixtures prepared from the normal diesel fuel (DF), 0.534, 0.968 and 1.894 MJ/kg and oxygenated fuel blends E5, E10, E15, E15B are almost the same 0.534–0.533, 0.968–0.967 and 1.893–1.888 MJ/kg for respective air–fuel ratios of k = 5.5, 3.0 and 1.5. Determined test conditions alleviate the examination of changes occurred in the effective power and emissions produced by an engine. 3. The test results and analysis Diesel fuel (class 2) was produced at the manufactory ‘‘Orlen Lietuva’’ and its quality parameters satisfied requirements EN 590:2009+A1. Anhydrous ethanol (200 proof) was brought from the producer Ltd. ‘‘Biofuture’’ and its parameters corresponded to standard EN 15376:2009. The purity of ethanol was determined by using the laboratory device Anton Paar density/concentration meter DMA 5000 with an accuracy of ±0.000005 g/cm3 at the temperature of 20 ± 0.001 °C. The RME was brought from the biofuel production company ‘‘Rapsoila’’ and its parameters satisfied main requirements of standard EN 14214:2009. 3.1. Chemical and physical properties of ethanol–diesel–biodiesel blends The test results proved that the miscibility of anhydrous (200 proof) ethanol up to 15 vol% added to diesel fuel is excellent, – so that it matches well with the outcomes of other researchers.

703

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720 Table 3 Properties of ethanol, diesel fuel, and rapeseed oil methyl ester (RME). Property parameters

Test method

Diesel fuel (class 2)

Ethanol 99.8 vol%

RME

Molecular formula Average molecular weight Density at 20 °C, kg/m3 Kinematic viscosity at 40 °C, mm2/s Flash point, open cup, °C Boiling point, °C Cloud point, °C Cold filter plugging point, °C Volatility, min 95% at, °C Auto-ignition temperature, °C Cetane number Iodine number, J2g/100 g Acid value, mg KOH/g Sulphur, mg/kg Phosphor, mg/kg Oxygen, max wt% Carbon to hydrogen ratio (C/H) Stoichiometric air–fuel ratio, kg/kg Net heating value, MJ/kg Carbonates, 10% of residue, mass% Ash content, mass-% Water content, mg/kg Total contamination, mg/kg

– – EN EN EN – EN EN EN – EN EN – EN – – – – EN EN EN EN –

C14H24.2O0.8 6180 830.5 2.07 56 177.8 22 32 340 250 51.5 6 0.06 2.2 Traces 0.4 6.9 14.45 42.95 60.03 60.01 21 2.0

CH3CH2OH 46 790.0 1.40 13 78 626 638 78 365 8 – 60.01 – – 34.8 4.0 9.06 26.95 – – 60.4 vol% –

C19H35.2O2 295 884.7 4.79 178 – 10 16 366 342 53.4 111 0.21 63.0 1.9 10.9 6.5 12.66 37.23 0.29 0.005 250 3.0

ISO 12185:1999 ISO 3104 + AC:2000 ISO 2719:2003 ISO 23015:1999 ISO 116/AC:2002 ISO 3405:2011 ISO 5165:1999 14111:2003 ISO 20846:2004

ISO ISO ISO ISO

8217:2007 10370:1999 6245:2002 12937:2002

Table 4 Specific properties of the tested ethanol–diesel–biodiesel blends. Property parameters

Test method

DF class 2

E5

E10

E15

E15B

Density at 20 °C, kg/m3 Kinematic viscosity at 40 °C, mm2/s Lubricity, corrected wsd, 1.4 lm at 60 °C Cetane number Oxygen content, max wt% Carbon/hydrogen ratio (C/H) Net heating value, MJ/kg Stoichiometric air/fuel ratio, kg/kg

EN EN EN EN – – – –

827.0 2.068 379 51.5 0.4 6.90 42.95 14.45

824.8 1.907 417 49.9 2.1 6.77 42.15 14.18

822.6 1.840 400 46.7 3.9 6.63 41.35 13.91

820.2 1.802 387 44.4 5.6 6.49 40.52 13.64

823.5 1.843 217 45.1 6.1 6.45 40.26 13.55

ISO ISO ISO ISO

12185:1999 3104 + AC:2000 12156-1:2007 5165:1999

Arctic diesel fuel is a better choose for blending with anhydrous ethanol because of the lower wax content and the absence of any biofuel additives. The miscibility of anhydrous (99.8 vol%) ethanol with arctic diesel fuel is good because after 24 h ethanol and diesel fuel blends E5, E10 and E15 were clear and transparent liquids sustaining in one phase at ambient temperature of 20 °C. The molecular weight of ethanol is 3.9 times lower and the density and kinematic viscosity are 4.9% and 1.5 times lower at the temperatures of 20 °C and 40 °C compared to the normal diesel fuel. The carbon to hydrogen ratio and net heating value of ethanol are also 42.0% and 37.3% lower because ethanol contains 3.2 times more oxygen than RME. For this reason, the stoichiometric air–fuel ratio of RME is 12.4% and that of ethanol 37.3% lower compared to conventional diesel fuel that also contributes to production of less NOx, CO and HC emissions due to the lower demand of the air-born oxygen needed for clean combustion of the fuel. On the one part, adding of ethanol to diesel fuel reduces density, viscosity, C/H ratio, flash point and improves volatility and oxygen content in the fuel blend (Table 4) that increases local air–fuel ratios in the very fuel-rich zones, leads to complete combustion at close to stoichiometric conditions and decreases smoke of the exhaust. The lower molecular weight of ethanol and absence of sulphur also contributes to the production of less soot, PM and smoke at the same performance conditions. Adding of 5 vol%, 10 vol% and 15 vol% of anhydrous ethanol to diesel fuel significantly decreased flash and cloud points, making the fuel blends more volatile, which subjected them for a vapour locks in the fuelling system, raised storage and handling requirements.

On the other part, the lower cetane number of ethanol, 1.5 times higher autoignition temperature compared with a basic diesel fuel and nearly threefold higher latent heat of evaporation, which varies in-between 840 kJ/kg [28] and 880 kJ/kg [18], and cooling effect of the fuel sprays may aggravate the autoignition of ethanol–diesel fuel blends and provoke misfiring cycles under unfavourable performance conditions. The latent heat of evaporation of diesel fuel is much suited to the compression ignition engine and varies in-between 250 [26] and 270 kJ/kg [33]. For this reason, when the engine runs under high loads, sharp knocking may emerge for bigger than 15 vol% ethanol additions. The initial boiling points of ethanol and diesel fuel were correspondingly 78 °C and 177.8 °C and a flash point of ethanol was at the temperature 4.3 times lower compared to arctic diesel fuel (Table 3). This means that ethanol begins to evaporate at about 100 °C before diesel fuel does. Both the cloud point (CP) and the cold filter plugging point (CFPP) of ethanol were also lower than respective temperatures of arctic diesel fuel. Adding of a lighter ethanol to diesel fuel decreased the cold filter plugging point (CFPP) of the blend and improved the filtering process therefore lower pressure was needed to filter the same volume of fuel. However, special attention must be paid to avoid ambient humidity and absorption of water because ethanol can deteriorate stability of the fuel blends and accelerate phase separation, cause corrosion of the fuel injection units and aqueous microorganism growth in the fuel tank. Whereas blending diesel fuel with 10 vol% of RME may result in a threefold higher microorganism growth than in pure diesel [37].

704

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

Adding of ethanol to diesel fuel affects certain key properties with particular reference to blend’s stability, viscosity and lubricity, energy content and cetane number. The standard ASTM D975 for conventional DF requires a minimum CN of 40 and EN 590 no less than 51, whereas the standards for biodiesel prescribe a minimum of 47 (ASTM D6751) and 51 according European standard EN14214 [38]. Lubricity is also a potential problem especially for the diesel engine because the fuel plays an important role in lubricating precision parts of the injection pump and injectors. The addition of ethanol to diesel fuel can reduce lubricity of the fuel blend and create potential wear problems in sensitive fuel pump designs. However, with regard to engine durability, blends with up to 15 vol% ethanol in diesel fuel can be considered relatively safe [21]. The wear tests conducted at the oil factory "Orlen Lietuva" by using a High Frequency Reciprocating Ring (HFRR) showed that the wear scar on ball was 6379 lm for the normal diesel fuel according corrected wear scar diameter (wsd), 1.4 lm at the temperature of 60 °C. Adding of 5 vol%, 10 vol% and 15 vol% of ethanol to diesel fuel reduced lubricity of the blends and the wear scar diameter on ball increased by 10.0%, 5.5% and 2.1% compared to the normal diesel fuel at the temperature of 60 °C. Worsened lubricity properties of the blends and a bigger wear scar caused by adding of ethanol to diesel fuel raised reasonable concern about reliability of the fuelling system because other researchers did not notice significant increase in the wear scar caused by the addition of ethanol. Hansen and Zhang showed [10] that the sample taken approximately at the midway of the first 100 hours yielded no abnormal levels of wear, metals, or contaminants. Meanwhile, Torres-Jimenez et al. [36] presented graphical data, which showed that the addition of ethanol to diesel fuel slightly improves lubricity, as the wear scar was lower. The evaporation of ethanol from the lubricating area might play an important role at the test temperature of 60 °C, which is close to the boiling point of ethanol. The reduced rate of the wear scar for higher amounts of ethanol in diesel fuel was uninspected enough, however this probably occurred because the wear testing method EN ISO 12156-1:2007 allows evaporation of some part of ethanol from the sample surface. Whereas in real operating conditions, the ethanol has less chances to escape from the clearances locked in-between the plunger-barrel and the nozzle-valve-needle bodies, – so that the ethanol addition to diesel fuel may have a bigger negative influence on the wear and reliability of the fuelling system. The viscosity of ethanol is much lower compared to diesel fuel therefore lubrication of important fuel injection parts, such as the plunger-barrel and the needle-valve-nozzle bodies, can be in potential danger. To improve lubricity properties, phase stability and expand the use of renewable biofuels the 5 vol% of RME was added to ethanol– diesel fuel blend E15 for the account of diesel fuel. The addition of RME to fuel blend E15 significantly improved lubricating properties of the blend E15B and reduced the wear scar on ball to minimum of 217 lm (42.7%) compared with a basic diesel fuel. In addition to superior lubricity, the RME increased density and viscosity, improved the cetane number, the fuel oxygen mass content and reduced both the calorific value and the stoichiometric air–fuel ratio of the blend E15B (Table 4). 3.2. Injection and autoignition of ethanol–diesel–biodiesel blends The lower density, viscosity and higher bulk modulus of compressibility of ethanol–diesel fuel blends reduced velocity of a highpressure wave propagating within the injection line and affected the injection timing advance. Consequently, the nozzle-needle-valve opening and the start of injection occurred later for ethanol–diesel fuel blends than in case of using the normal diesel fuel. The higher

the percentage of ethanol added to diesel fuel, the more significant delay in the start of injection was registered in CADs. To be precisely, the start of injection took place 1.4°, 0.7° and 0.4° later in the case of running on fuel blend E15 compared to respective values of 12.1°, 12.0° and 12.9° BTDC of ordinary diesel operating on richer air–fuel mixtures k = 1.5 at 1400, 1800 and 2200 rpm speeds. Adding the 5 vol% of RME to ethanol–diesel fuel blend E15, the molecular weight of which is 6.4 times higher than ethanol, improved density and viscosity of the blend E15B (Table 4). As a result, the difference in the start of injection diminished to the minimum of 0.9°, 0.5° and 0.1° CADs compared to normal diesel operation at respective speeds. Replacement of diesel fuel with ethanol–diesel fuel blends is aggravated by a higher latent heat of evaporation, the negative influence of which increases with the increase of the ethanol in the fuel blend. According to the considerable number of experimental test results summarised in Ref. [39], the temperature in the fuel spray core decreased by about 150–200 °C and the autoignition delay was longer for alcohol–diesel fuel blends, especially when operating at light load and low speed. The next disadvantage of ethanol as a potential diesel fuel extender is low cetane number and high autoignition temperature, which significantly affect the ignition delay, combustion, performance efficiency, and emission parameters of the exhaust. The dependencies of the autoignition delay si on the cetane number of ethanol–diesel–biodiesel blends for three specific air– fuel ratios and three ranges of engine speed are plotted in Fig. 3a. It can be seen that the lower the cetane number of the fuel blend, the longer is the autoignition delay si in units of time (ms) and the more period si depends on the composition of the air and fuel mixture and engine speed. The low boiling point of the ethanol seems to have no helpful effect on shortening of the autoignition delay. Researchers Shropshire and Goering [8] noticed that the ethanol causes an increase in the ignition delay in all cases. Moreover, the additional ethanol may lead to misfiring cycles and a rapid fall in engine efficiency because the autoignition delay ui gets longer, especially at low load and speed. The autoignition delay increased from 1.13 ms to 1.52 ms (34.5%) with the decrease of the cetane number from 51.5 (DF) to 44.4 (E15) for lean air–fuel mixture k = 5.5 used at 1400 rpm speed. The higher cetane number 45.1 of the blend E15B did not shorten the autoignition delay and period si further increased to 1.62 ms (43.4%) compared to the normal diesel fuel as show graphs in Fig. 3a. The autoignition delay’s si changing behaviour was very similar within the entire cetane number range of 51.5–44.4 when operating on lean k = 5.5 and moderate k = 3.0 air–fuel mixtures at 1800 and 2200 rpm speeds. The same periods in units of time si mean that the autoignition delays in CADs increase proportionally to the engine speed for all fuel blends tested. The autoignition delay si diminished to the minimum of 0.75 and 0.78 ms for a straight diesel engine operating on slightly richer air–fuel mixtures k = 1.5 used at 1800 and 2200 rpm speeds. The negative influence of low CN on the autoignition delay si slightly diminished for richer combustible mixtures used at 1400 rpm speed due to the later start of injection of oxygenated blends and thus higher temperature of the air and residual gasses inside the cylinder. However, the autoignition delay was still 20.8% and 26.4% longer than regular diesel at 1800 rpm speed and 12.8% and 15.4% longer at rated 2200 rpm speed when running at a fully opened throttle on blends E15 and E15B. The autoignition delay si slightly increased for almost all fuel blends used at richer air–fuel mixtures k = 1.5 due to transition from 1800 to 2200 rpm. As a result, the autoignition delay ui became in the range of 2.1° (E15B) to 2.5° (E5) longer for rated 2200 rpm speed. An unusual variation of the autoignition delay observed at speed of 2200 rpm, which is higher than the delay period (ms) at 1800 rpm speed, might occur because the test results were

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

705

Fig. 3. The autoignition delay si as a function of the cetane number (a) and fuel oxygen mass (wt%) content (b) for various air–fuel equivalence ratios and engine speeds.

analysed and compared for overall the same combustion conditions. The first reason is that the brake mean effective pressure developed by the combustion of the fuel blends was slightly lower at 2200 rpm speed for the same air–fuel ratios of k = 1.5 (Fig. 7) that led to having longer ignition delay. The second reason of a longer ignition delay time might be linked with a higher restriction to free gas flow and lower scavenging degree of the engine cylinder at high 2200 rpm speed. The matter is that the overlap of inlet and exhaust valves is adjusted for the speed range of 1400–1800 rpm to ensure well-organised gas exchange and maximum torque of a naturally aspirated engine. As a result, the cylinder scavenging effect was greater that reduced the amount of residual gases and increased the volumetric efficiency of an engine. The higher pressure at the end of a suction stroke increased the temperature in the cylinder for the start of injection that shortened the ignition delay at 1800 rpm speed. Therefore, probably, the autoignition delay increased a bit faster with the increase of the fuel oxygen content at considered speed. Similar increase of the ignition delay occurred for 2200 rpm speed when running on leaner air–fuel mixtures k = 5.5, whereas the period si actually coincided for moderate k = 3.0 mixtures used over the entire CN range at a higher 1800 and 2200 rpm speed.

Among the most important factors affecting the autoignition of fuel sprays are the cetane number [24], the CN improvers [35], the fuel injection timing advance [28], the pressure and the temperature inside the combustion chamber which depend largely on the engine load [39] and speed [40,41]. Nevertheless, the autoignition delay period si do not always relies on the cetane number only, especially when operating on biofuels with different origin and nature [42]. It should be noted that the autoignition delay si no longer correlated reliably with the cetane number value when the 5 vol% of RME was added to ethanol–diesel fuel blend E15. The fuel blend E15B, the cetane number of which was higher 45.1 than that (44.4) of ethanol–diesel fuel blend E15, suggested the autoignition delay si 6.6% longer for a leaner air–fuel mixture k = 5.5 tested at low 1400 rpm speed. The autoignition delay si gradually decreased for a bigger fuel portion injected (lower k) and higher all of them, – the engine speed and gas pressure and temperature inside the cylinder. Nevertheless, the increase of period si for the blend’s E15B case is unique and not fully reversible, it does not disappear completely and stands out over the entire load-speed range that may affect the overall combustion process and what ends up in the engine exhaust. This finding is strange and uninspected enough because

706

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

adding the 5 vol% of RME, the CN rating of which is the highest one (53.4), to ethanol–diesel blend E15 for the account of diesel fuel should improve the autoignition properties of the blend E15B and shorten ignition delay time. An anomaly like this one notso-often occurs in a normal practice when using conventional diesel fuel, the CN range of which is about 40–60 [43]. The observed changes in the ignition delay si occurred, probably, because the autoignition properties of ethanol–diesel–biodiesel blends were determined by using the cetane engine method EN ISO 5165:1999 which is adapted mainly for testing of the normal diesel fuel. To find out more about not-so-simple phenomenon, the values of autoignition delay si were situated as a function of the fuel oxygen mass content for the tested air–fuel ratios and speeds as shown in Fig. 3b. A new approach revealed an exceptional role of the fuel bound oxygen, which reflects changing behaviour of the autoignition delay si and ui more predictably than the cetane number does. At first, this notice applies to the blend’s E15B case, the components of which are similar in chemical composition but different in origin and nature. The autoignition delay time si increased by 43.4%, 18.9%, 14.0% for overall lean, 21.1%, 22.6%, 22.4% moderate and 14.9%, 21.3%, 15.4% richer air–fuel mixtures responding to the fuel oxygen mass content increased from 0.4 wt% to 6.1 wt% at 1400, 1800 and 2200 rpm speeds. The autoignition delay time si increased about proportionally to the total amount of oxygen in the fuel blend and this notice is true for almost all air–fuel ratios and engine speeds. It would be interesting to find out more about whether the fuel oxygen mass fraction plays a key role in determining duration of the period si in case of using the RME additive only for ethanol–diesel fuel blends treatment or this finding can be applied for other fatty acid methyl esters differing in molecular formula and structure. It is clear that the leaner the air–fuel mixture and the lower the temperature inside the cylinder, the slower the fuel evaporation process and the greater ethanol cooling effect on the measured period si, therefore more chances for unusual events to occur. The measured autoignition delay is actually a sum of both physical and chemical components. The physical delay is the time required for droplets to penetrate into the compressed air environment, break up and evaporate. The chemical delay is largely controlled by the time required for chemical reactions to develop. Taking into account influencing factors, the increase of the ignition delay for the fuel blend E15B, which continues over a wide range of engine loads and speeds, might be caused by the physical delay, i.e. higher density, viscosity, contamination and slower evaporation of RME droplets. The bench tests of a supercharged, four-stoke, four-cylinder, aircooled, DI diesel engine D-144 with a semi-spherical bowl-in-piston combustion chamber showed that the autoignition delay si of a fully loaded engine can be shortened to the minimum of 0.56 ms by speed increased in the range of 1800–2200 rpm. The minimum value of period si depends on the combustion chamber design, compression ratio of an engine, the in-cylinder gas temperature at the start of injection, the fuel cetane number, instantaneous compression rate of the air-charge, which relies on the engine speed and injection timing advance and so called injection criterion K00bgp [40,41,44]. 3.3. Combustion and heat release analysis for ethanol–diesel–biodiesel blends The crank angle AI 50 represents the center of the gravity of differential heat release therefore this factor is important as it affects the fuel energy conversion efficiency of an engine. The center of the gravity of the area usually lies between angles AI 5 and AI 90 for diesel engines and this is a historical value. The shorter is crack

angle ATDC, at which 50% of energy releases in the cylinder, the lower are heat losses during the expansion stroke and the higher the cycle thermal efficiency can be expected. Whereas the crank angle value for 90% mass burned fraction in the integral heat release curve was accepted as the end of combustion. The assumption was made bearing in mind that the remaining 10% of heat release does not greatly affect performance efficiency of an engine. The maximum heat release rate HRRmax in premixed combustion phase relays on the autoignition delay time si, the fuel-spray tip penetration, the spray cone angle and atomisation and evaporation conditions, i.e. on the quantity of fuel premixed for rapid combustion. The maximum heat release rate HRRmax for ethanol–diesel fuel blends in all cases was lower compared to regular diesel operating on lean air–fuel mixtures k = 5.5 in the speed range of 1400– 1800 rpm (Fig. 4a). The lower HRRmax matches well with the test results of other researchers [35], where the beginning of heat release rate retarded for ethanol–diesel fuel blends and HRRmax was lower, except for large load operation. The maximum heat release rate HRRmax increased to a certain extent with the increase of the ethanol oxygen content in the fuel blend for slightly richer combustible mixtures k = 3.0 and 1.5 because of a bigger portion of the fuel was prepared for rapid burning during longer ignition delay. The fuel oxygen has inherent ability to accelerate the rate of combustion, especially when running on richer air–fuel mixtures and low speeds. Therefore the maximum heat release rate increased in the range of 114.7–146.7 kJ/m3 deg (27.9%) and 93.9 (DF) to 111.3 (E15B) kJ/m3 deg (18.5%) for combustible mixtures k = 1.5 used at 1400 and 1800 rpm speeds. The lower HRRmax determined for richer mixture k = 1.5 used at 1800 rpm speed shows that the combustion occurred more likely under premixed-controlled conditions that matches well with a shorter ignition delay time (Fig. 3a), bigger angles corresponding to 50% and 90% of heat release (Fig. 5a) and lower pressure pmax in the cylinder (Fig. 6a). However, the combustion advantages provided by the fuel bound oxygen were largely offset by low cetane number of ethanol–diesel fuel blends, which is frequently observed when running at high 2200 rpm speed. Interaction between influencing factors determined that the maximum heat release rate HRRmax increased to a certain extent when running on less oxygenated blend E5 only and reached the maximum values of 65.1, 96.4 and 138.5 kJ/m3 deg for respective 5.5, 3.0 and 1.5 air–fuel ratios. The biggest (35.6%) influence of the fuel oxygen on maximum heat release rate was achieved for lean air–fuel mixtures k = 5.5, which have slower flame speeds, and the smallest one (15.4%) for slightly richer combustible mixtures k = 1.5, which normally burn with the fastest flame speeds. Whereas, the maximum HRRmax from combustion of moderate mixture k = 3.0 was 24.2% higher for slightly oxygenated blend E5 than conventional diesel fuel. The maximum flame speed for most fuels takes place at an equivalence ratio near 1.2 [43]. The other test results [18] showed that the autoignition delay, the premixed combustion phase, the fraction of heat released, and the maximum pressure inside the cylinder increase, while the diffusive combustion phase and total combustion duration decrease with the increase of the ethanol fraction in the fuel blend. The maximum heat release rate HRRmax from combustion of leaner air–fuel mixture k = 5.5 decreased by 42.4% as the fuel oxygen mass content increased from 0.4 wt% to 3.9 wt% (E10) at 1400 rpm speed. The maximum HRRmax decreased by 38.2% when using more oxygenated blend E15 at a higher 1800 rpm speed too. The cause of such heat release behaviour is that the evaporation of small fuel portions was slow enough when running on lean combustible mixtures and low engine speeds. Therefore, the ethanol addition to diesel fuel has controversial influence on vaporisation process. On the one part, it is good that the front-end volatility of the ethanol is below 78 °C, on the other part, the ethanol provides significant cooling effect

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

707

Fig. 4. The maximum heat release rate HRRmax (a) and its location angle AHRRmax (b) with respect to TDC as a function of fuel oxygen mass (wt%) content for various air–fuel equivalence ratios and engine speeds.

aggravating evaporation of the fuel. For these reasons, the increase of the maximum heat release rate with the increasing engine speed can be caused by a higher gas temperature and in-cylinder gas motion that improve the fuel evaporation and its vapour and the air mixing when running on fuel–lean mixtures. As a cumulative result of all influencing factors, the autoignition delay si increased for a higher fuel bound oxygen content and the maximum heat release corresponding angle AHRRmax moved further from TDC for all air–fuel ratios and engine speeds as shows Fig. 4b. The angle AHRRmax shifted from 0.8° BTDC (DF) to +2.0° ATDC (E15B) CADs responding to the fuel oxygen mass content increased from 0.4 wt% to 6.1 wt% for lean air–fuel mixture k = 5.5 used at low 1400 rpm speed. The heat release angle AHRRmax increased with regard to the normal diesel fuel in the range of 4.9° CADs for an air–fuel ratio of k = 3.0 (E15) to 0.9° CADs for an air–fuel ratio of k = 1.5 (E15B) when running at rated 2200 rpm speed. Fig. 5a, b presents changes in crank angles AI 50 and AI 90 (CADs) as a function of ethanol and RME oxygen mass content in the fuel blend for specific air–fuel ratios and three ranges of speed. Adding ethanol to diesel fuel reduced the cetane number of the fuel blend, increased the fuel oxygen mass content and shifted heat release angles AI 50 and AI 90 further ATDC. The angles AI 50 and AI 90 were dislocated towards a bigger cylinder volume in the expansion stroke because of the low cetane number of oxygenated fuel blends and long autoignition delay. However, the changes

were not so widely scattered and varied within the range of 2–3° CADs for crank angle AI 50 and 1–6° CADs for crank angle AI 90, again depending on the engine speed, the fuel blend used and quality of the air and fuel mixture. The biggest 2.2 and 1.4 times increase in heat release angles AI 50 and AI 90 was registered for mostly oxygenated blend E15B used at lean air–fuel mixture k = 5.5 and low 1400 rpm speed. Demand for the fuel bound oxygen increased when operating on richer combustible mixtures because of the deficiency of the airborn oxygen inside the cylinder. Consequently, the crank angle AI 50, at which the 50% of heat releases in the cylinder, decreased by 9.3% for the fuel blend E15B used at an air–fuel ratio of k = 1.5 and 1400 rpm speed. The closer the combustion process is to constant volume, the higher will be thermal efficiency. This statement is based on the analysis of the brake thermal efficiency results given in Fig. 9. Although the change of crank angle AI 50 is not only reason which determines thermal efficiency and fuel consumption of an engine because the 50% of heat release for blend’s E15B case lasted 19.1% longer at 1800 rpm speed and a bit 1.2% longer at 2200 rpm speed than normal diesel operating on richer air–fuel mixtures. The crank angle AI 90 at which 90% of heat releases in the cylinder was about 17.9% bigger when operating on the fuel blend E15B at 1800 rpm, however, this angle, which with some assumption represents the end of combustion, was less dependent on the

708

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

Fig. 5. Heat release angles AI 50 (a) and AI 90 (b) as a function of fuel oxygen mass (wt%) content for various air–fuel equivalence ratios and engine speeds.

fuel oxygen mass content at 1400 and 2200 rpm speeds. It should be noted that the use of composite blend E15B diminished both heat release angles AI 50 and AI 90 that is important for the cycle thermal efficiency and fuel-efficient engine performance. Using of the fuel blend E15B improved combustion, reduced heat losses, increased the brake mean effective pressure developed from the same amount of energy (Fig. 7) and the brake thermal efficiency (Fig. 9) over the entire speed range 1400–2200 rpm compared to ethanol–diesel fuel blends E5, E10, E15 used at richer air–fuel mixtures k = 1.5. 3.4. Maximum combustion pressure and engine performance smoothness Fig. 6a and b show how the combustion pressure pmax and corresponding angle Apmax change with the increase of the ethanol and RME oxygen content in the fuel blend for various air–fuel ratios and three ranges of engine speed. The maximum in-cylinder pressure pmax decreased for the higher ethanol oxygen mass content and lower cetane number of the fuel blend almost for all loads and speeds, especially when operating on leaner air–fuel mixtures. Contrary to common expectations, more fuel accumulated in the cylinder for the longer period of autoignition delay did not boost up maximum combustion pressure pmax, which for low cetane fuel blend E15 actually decreased by 6.3% compared to regular diesel operating at an air–fuel ratio of k = 5.5 and 1800 rpm speed. The

combustion pressure pmax reduced due to the cooling effect caused by the ethanol, which retarded the combustion process because leaner air–fuel mixtures evaporate much slower at low temperatures in the cylinder. Despite the longer ignition delay time and thus more fuel injected in the combustion chamber, the pressure pmax did not increase with the increase of the ethanol oxygen content in the fuel blend as could be expected. The obtained result shows that the combustion process of overall lean mixtures is largely controlled by the fuel evaporation which is very limited at low temperature in the cylinder. A key point is that the speed of flame propagation, maximum both heat release rate HRRmax and gradients (dp/du)max of combustion pressure rise rely largely on the amount of fuel vaporised for rapid burning, which counts on the temperature created at the end of the compression process, and the engine speeddependent the in-cylinder air swirl turbulence intensity. Therefore the influence of the higher fuel oxygen content on the combustion parameters and most exhaust emissions can be dissimilar for overall lean, moderate and slightly richer air–fuel mixtures used at different speeds, i.e. a reverse in generally well-known tendencies may occur with the increase of the fuel oxygen mass content in the blend. The fuel combustion can be improved to some extent by using three component blend E15B for the engine fuelling. As Fig. 6a shows, the maximum pressure pmax inside the cylinder increased by 2.6% for an air–fuel ratio of k = 3.0 at 1400 rpm speed, and by

709

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

λ 5,5

(b)10 Apmax, CAD ATDC

(a)

λ 5,5

8 6 4 2 0 0

12

3,0

Apmax, CAD ATDC

λ

λ

3

4

5

6

7

2

3

4

5

6

7

2

3

4

5

6

7

3,0

8 6 4 2 0

12

Apmax, CAD ATDC

2

10

0

λ 1,5

1

1

λ 1,5

10 8 6 4 2 0

0

1

Fuel bound oxygen content, wt.% Fig. 6. The maximum combustion pressure pmax (a) and corresponding angle Apmax (b) with respect to TDC as a function of fuel oxygen mass (wt%) content for various air–fuel equivalence ratios and engine speeds.

3.1% and 1.9% for richer air–fuel mixtures k = 1.5 used at 1400 and 2200 rpm speeds. Fuelled with composite fuel blend E15B the fully loaded engine developed the maximum combustion pressure pmax 2.1 and 1.2 bars higher than the values of 68.1 and 64.6 bars of a normal diesel operating at respective speeds. The analysis of the test results revealed that the influence of ethanol oxygen on maximum combustion pressure pmax is ambiguous and depends on many variables which include temperature inside the engine cylinder, turbulence intensity, swirl, air–fuel ratio, presence of fuel oxygen, etc. At low both the engine speed and the in-cylinder gas temperature, the ethanol misses main advantages because the fuel evaporation goes slowly and the flame front propagates in lean air–fuel mixtures with a limited speed. The ethanol as a diesel fuel additive regained an extra impetus when running at high engine loads, speeds and combustion temperatures because the oxidation reactions of rich air–fuel mixtures go faster with a higher flame speed and heat release rate (Fig. 4a). It is known [43] that alcohols have high flame speed, a fast but finite flame speed is desirable in an engine. Fig. 6b shows that the angle Apmax, which corresponds to maximum pressure pmax in the cylinder, moved further ATDC due to increased fuel oxygen content in the blend actually for all air– fuel mixtures and engine speeds. The combustion pressure pmax

sustained at the same or slightly higher level despite dislocation of the angle Apmax further from TDC towards a bigger cylinder volume in the expansion stroke when running on richer combustible mixtures k = 1.5 over a wide range of speeds. The angle Apmax extended by 0.5–2.9° CADs due to the fuel oxygen content increased from 0.4 wt% to 6.1 wt% for lean air–fuel mixtures k = 5.5 used at 1400 and 1800 rpm speeds. The higher heat losses during combustion caused the peak temperature and pressure in the cylinder to be lower from what could be predicted. Changes in the in-cylinder pressure pmax and its location angle Apmax may affect the work produced by the same energy stored in the combustible mixture, brake specific fuel consumption, engine performance efficiency and emissions of the exhaust. Dependencies of the combustion pressure pmax and corresponding angle Apmax on the fuel oxygen mass content show that the engine is able to tolerate anhydrous ethanol up to 15 vol% added to diesel fuel without causing of any larger changes in performance parameters. The angle AHRRmax of maximum heat release rate took place in the range of 2.6° (DF) to 0.2° (E15B) CADs before the maximum combustion pressure Apmax was built up inside the cylinder when running on lean air–fuel mixture k = 5.5 at 1400 rpm speed. The tendency to set the AHRRmax positions to the ones of Apmax with increasing content of oxygen in the fuel blend can also be

710

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

Fig. 7. The brake mean effective pressure (bmep) as a function of fuel oxygen mass (wt%) content for various air–fuel equivalence ratios and engine speeds.

observed when running on slightly richer mixtures k = 3.0 and k = 1.5. The disagreement between both AHRRmax and Apmax angles was reduced to minimum values of 1.1° (E15) and 3.0° (DF) CADs by a higher flame speed in richer combustible mixtures k = 3.0 and 1.5 used at 2200 rpm speed. The peak combustion pressure pmax depends on the engine load, speed, fuel injection parameters, maximum rate of heat release

HRRmax and its location angle AHRRmax with regard to TDC. The behaviour of maximum heat release rate HRRmax differs from that of a single-cylinder DI diesel engine tested by Li et al. [18], who found that the center of heat release curve moved close to the TDC, and the maximum of both heat release rate and a rate of incylinder pressure rise increased for a higher ethanol fraction in the blend. The noted dissimilarity most likely emerged because

350

λ = 1,5

bsfc, g/(kWh)

300 250

1400 rpm 1800 rpm

200

2200 rpm

150 100 0,04

2,1

3,9

5,6

6,1

Fuel oxygen mass fraction, wt% Fig. 8. The brake specific fuel consumption (bsfc) as a function of fuel oxygen mass (wt%) content for various air–fuel equivalence ratios and engine speeds.

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

711

Fig. 9. The brake thermal efficiency (bte) as a function of fuel oxygen mass (wt%) content for various air–fuel equivalence ratios and speeds.

in that study the combustion and the two-phase heat release rate were analysed for the same brake mean effective pressure and engine speed. The maximum gradients (dp/du)max of pressure rise in an engine depend on the amount of fuel premixed for rapid burning, maximum combustion pressure pmax and its location angle Apmax with regard to TDC. The engine operated smoother on composite fuel blend E15B and maximum gradients (dp/du)max of pressure rise on an average were 0.50 bar/° and 0.21 bar/° lower than values of 2.25 bar/° and 3.57 bar/° of a normal diesel running on leaner combustible mixtures k = 5.5 and 3.0 at rated 2200 rpm speed. The maximum pressure gradients (dp/du)max were 1.15 bar/° and 1.04 bar/° lower for blend E15B used with leaner air–fuel mixtures k = 5.5 and 3.0 at 1800 rpm speed. The smoothest 3.99 bar/° (DF) and 3.98 bar/° (E15B) engine performance was also achieved when running on slightly richer mixture k = 1.5 at 1800 rpm speed. Whereas, when fuelling a fully loaded engine with composite blend E15B, the maximum pressure gradients were 1.65 bar/° and 0.61 bar/° higher compared to ordinary diesel operating at 1400 and 2200 rpm speeds, which can be caused by a higher heat release rate HRRmax (Fig. 4a). 3.5. Engine efficiency with ethanol–diesel–biodiesel blends As Fig. 7 shows, the brake mean effective pressure developed by the same energy delivered per each engine cycle depends on the fuel oxygen mass content, the usefulness of which changes with an air–fuel ratio and engine speed, i.e. the utilisation rate of the fuel oxygen relies on the availability of the air-born oxygen in the cylinder. The bmep increased in the range of 0.100 MPa (DF) to 0.108 MPa (E10), i.e. by 8.0% due to the fuel oxygen content increased from 0.4 wt% to 3.9 wt% when running on lean air–fuel mixture k = 5.5 at 1400 rpm speed. More torque and engine power developed by the same energy proved advantages to be realised by the use of slightly oxygenated blend E10 at light loading conditions. The brake mean effective pressure increased by 7.7% for

the same blend E10 at 1800 rpm speed too, whereas the bmep 16.1% higher than easy loaded regular diesel developed the most oxygenated blends E15 and E15B used at rated 2200 rpm speed. Using of plenty oxygenated blends the engine torque and bmep increased because the fuel-side oxygen contributed with an essential help to burn at close to stoichiometric conditions and overreached mixtures at some local spots, which then promoted further reactions until, finally, the combustion reaction occurred. The fuel conserved oxygen plays a very important role in burning of heterogeneous mixtures formed from small fuel portions injected into the combustion chamber at light loads and the effectiveness of fuel-side oxygen depends on the in-cylinder air swirl turbulence intensity which is a function of the engine speed. The in-cylinder air-born oxygen does not burn the fuel completely when incomplete mixing of overall lean mixtures occurs, while the role of fuel-side oxygen in the oxidation reactions increases because of the short time available for completion of each engine cycle at high 2200 rpm speed. The main advantage of fuel bound oxygen is that due to clean combustion of highly heterogeneous mixtures it develops the brake mean effective pressure, torque and engine power a bit higher than an engine consuming the same diesel fuel energy delivered per cycle. The influence of fuel bound oxygen on the work produced by the cylinder’s splash volume slightly decreased for an air–fuel ratio of k = 3.0 because the higher injection pressure improved mixing of the air and fuel, resulting in more fuel particles finding in-cylinder air-born oxygen to react with. Therefore, the bmep 3.0% and 2.1% higher than normal diesel developed an engine running on less oxygenated blend E5 (2.1 wt%) at 1400 and 1800 rpm speeds. The brake mean effective pressure’s behaviour with increasing ethanol oxygen content in the fuel blends changed at rated 2200 rpm speed, – so that the bmep decreased by 5.1% for blend0 s E15 case compared to the normal diesel fuel. Adding the 5 vol% of RME, as co-solvent and cetane improver, to blend E15 for the account of diesel fuel resulted in the brake mean effective pressure 3.3%, 1.4% and 2.1% higher than a normally-fuelled engine develops at

712

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

1400, 1800 and 2200 rpm speeds (Fig. 7). It is interesting to stress that the bmep increased when running on composite fuel blend E15E despite longer ignition delay time si, which seems is not so important if the combustion process goes to completion. Using of low percentage ethanol–diesel fuel blends, such as 10 vol% of ethanol and 90 vol% of diesel fuel, actually did not affect the brake mean effective pressure developed by the combustion of richer air–fuel mixture k = 1.5 at 1400 rpm speed at which maximum torque occurs. However, the bmep decreased by 0.9% for medium 1800 rpm and by 2.6% for rated 2200 rpm speed due to the ethanol oxygen content increased in the range of 0.4 wt% (DF) to 5.6 wt% (E15). The effectiveness of energy conversion reduced because markedly more fuel should be injected per each engine cycle to compensate for the resulting lower net heating value of ethanol–diesel fuel blends. The lower viscosity of ethanol contributed to the delay of a nozzle-needle-valve opening, whereas the lower calorific value of oxygenated blends was compensated by a higher fuel portion that extended the end of injection. The effective work produced per unit of cylinder’s volume decreased because bigger fuel portions had a very small fraction of a second for proper atomisation of the fuel, evaporation, mixing of the fuel vapours with the in-cylinder compressed air-charge, autoignition and complete combustion. The late start of injection caused by the lower density and viscosity of ethanol–diesel fuel blends, low cetane number and long autoignition delay were among main factors leading to incomplete combustion in a short cycle time. For this reason, a fully loaded engine developed less effective power from the same energy stored in ethanol–diesel fuel blends when running at high 2200 rpm speed. Columns of Fig. 7 show that the brake mean effective pressure developed by the combustion of composite fuel blend E15B increased over the entire range of 1400–2200 rpm speed. Adding the 5 vol% of RME to blend E15 most likely reduced the initial spray angle, increased the fuel spray tips’ penetration deeper into the combustion chamber and improved accessibility of the air-born oxygen because of a bigger than before the total surface area of the fuel sprays exposed to high-temperature in-cylinder compressed air-charge. Adding of RME, as a stabilizer of ethanol–diesel fuel blend, suggests an extra advantage because improves the cetane number and lubricity at higher blending ratios, and increases the content of biofuel in the blend, which is one of the objectives determined by the EU directive 2009/28/EC to promote the use of renewable fuels in transport sector. A fully loaded biodiesel suggested the bmep 1.2% and 0.6% higher at 1400 and 1800 rpm speeds, and slightly (1%) lower at 2200 rpm speed than an engine develops consuming the same diesel fuel energy because of improved combustion of the fuel. It can be seen in Fig. 7 that the higher the rotation speed is, the greater are friction losses between moving engine parts and for the gas exchange that lead to the lower bmep and power developed by the same energy at rated 2200 rpm speed. According the test results given in Ref. [8], the ethanol can replace up to 60% of diesel fuel when a low-pressure injection system is used for ethanol spray into the intake manifold of a Case model 188D tractor engine. Adding of ethanol degrades properties of diesel fuel such as the cetane number, viscosity and calorific value therefore up to 20% ethanol could be tolerated, although authors suggest 10% as more effective ratio. Hansen et al. [3] measured a 7–10% decrease in power at rated speed with a 15% dry ethanol, 2.35% PEC additive and 82.65% No. 2 diesel fuel blend run in a Cummins ISB 235 engine. Engine durability tests showed that performance on a blend containing 10% ethanol, 1% GE Betz additive, and 89% diesel fuel was not affected apart from an expected 4% decrease in power caused by lower energy content of the blend [10]. Columns in Fig. 8 show how the brake specific fuel consumption changes with the increase of the fuel oxygen mass content in the

fuel blend when operating at various air–fuel ratios of k = 5.5, 3.0, 1.5 and three ranges of 1400, 1800 and 2200 rpm speed. The bsfc decreased from 446 g/kW h to 409 g/kW h (8.3%) and 470 g/ kW h to 436 g/kW h (7.2%) due to the ethanol oxygen in the fuel blend increased from 0.4 wt% to 3.9 wt% when operating on lean air–fuel mixtures k = 5.5 at 1400 and 1800 rpm speeds. The lower brake specific fuel consumption matches well with a higher effective power developed by an engine running on the fuel blend E10 at lean combustible mixtures and low speeds (Fig. 7). However, the bsfc increased from 632 g/kW h (DF) to 666 g/kW h (E15B), i.e. by 5.4% for lean air–fuel mixture k = 5.5 used at rated 2200 rpm speed. The higher bsfc can be attributed mainly to 6.3% lower net heating value of the blend E15B since the brake thermal efficiency (Fig. 9) did not greatly change with almost any significant response to the increase of the fuel oxygen content and reached 1.5% higher value compared to the normal diesel fuel. The bsfc progressively increased by 5.4%, 6.3% and 15.2% with 1400, 1800 and 2200 rpm speed for composite blend E15B compared with values of 260, 268 and 283 g/kW h of a regular diesel running at an air–fuel ratio of k = 3.0 (Fig. 8). The brake specific fuel consumption was 0.8% lower only for less oxygenated blend E5 used at moderate load and 1400 rpm speed. The bsfc increased with increasing ethanol oxygen content in the fuel blends, when running on richer combustible mixtures k = 1.5 too. However, the higher the engine speed was, the less significant was the bsfc increment rate caused by simultaneous addition of ethanol and RME to diesel fuel. The positive tendencies in the fuel economy with shortening engine cycle can be attributed to a greater than before the role of the fuel oxygen in completing the combustion process since the engine operated on richer air– fuel mixtures and strongly limited time of oxidation reactions. Since there was a lack of the air-born oxygen, the fuel bound oxygen came with essential help to burn rich air–fuel mixtures in local combustion chamber parts saving the fuel energy and reducing the CO and the HC emissions. The bsfc increment rate with regard to the normal diesel fuel decreased 8.2%, 7.2% and 6.5% about proportionally to the increasing 1400, 1800 and 2200 rpm speed since the fuel-rich mixtures prepared from oxygenated blends E15 and E15B burned more efficiently. Experiments with a common-rail, four-cylinder diesel engine showed that the bsfc values were significantly different at low-load conditions, and increased about 6.5% when ethanol–diesel fuel blend was used in the middle-load and high-load conditions [45]. Despite a large quantity of fuel injected, the brake thermal efficiency for ethanol–diesel fuel blends was similar or slightly higher than that of the normal diesel fuel. Whereas, the test results of a single-cylinder, DI diesel engine ZS1100 showed that the bsfc and bte increased for ethanol–diesel fuel blends E10-D and E15D at all loading conditions [46]. Comparison of the results shows that the ethanol effect on the oxidation reactions in the fuel-rich zones, specific fuel consumption and performance efficiency depends on the engine design, loading conditions and a number of influencing factors. The brake specific fuel consumption and net heating value of each fuel component were taken into account to evaluate the brake thermal efficiency of the engine operating on ethanol–diesel–biodiesel blends at specific air–fuel ratios and three ranges of speed. As Fig. 9 shows, the effectiveness of the fuel oxygen on the brake thermal efficiency depends largely on the air and fuel mixture value and engine speed. The brake thermal efficiency increased by 13.3–12.4% and reached the maximum values of 0.213–0.200 for the ethanol fuel oxygen content of 3.9 wt% used at lean air–fuel mixtures k = 5.5 in the speed range of 1400–1800 rpm. A smooth but nevertheless decrease in the performance efficiency shows that an engine does not tolerate higher than the 10 vol% ethanol–diesel fuel blend when operating on lean combustible mixtures and low

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

speeds. The brake thermal efficiency did not vary much with the fuel oxygen mass content increased in the range of 0.4–6.1 wt% for overall lean mixture k = 5.5 used at a higher 2200 rpm speed. Using of the fuel blend E10 (3.9 wt%) suggested the brake thermal efficiency 3.1–1.6% higher within the speed range of 1400– 1800 rpm than normal diesel develops at an air–fuel ratio of k = 3.0 (Fig. 9). However, the brake thermal efficiency decreased by 8.4% and reached the lowest value of 0.271 due to the ethanol oxygen mass content increased in the range of 0.4–5.6 wt% at rated 2200 rpm speed. The energy conversion efficiency decreased with the increase of the ethanol oxygen content in the blend for all engine speeds tested at an air–fuel ratio of k = 1.5 too. Using of the fuel blend E15 suggested the bte 2.1%, 1.3% and 0.5% lower than a fully loaded regular diesel develops at 1400, 1800 and 2200 rpm speeds. It can be seen that the fuel conserved oxygen, which is always on the spot ready to burn the fuel-rich patterns to end, gained extra advantages at a higher speed. As a result, the higher the engine speed was, the lesser decrease of the brake thermal efficiency with the increasing ethanol oxygen content in the blend occurred for overall richer air–fuel mixtures. Using of oxygenated blend E15B improved diesel engine performance on richer combustible mixtures, so that the brake thermal efficiency was only 1.3% and 0.5% lower than respective values of 0.381 and 0.376 of a normal engine running at 1400 and 1800 rpm speeds. Despite late autoignition, the fuel blend E15B suggested the brake thermal efficiency about the same 0.362 as a straight diesel operating on richer air–fuel mixture k = 1.5 at rated 2200 rpm speed. Late autoignition of oxygenated blend should reduce the brake thermal efficiency of the diesel engine as could be expected, but the effect of a longer autoignition delay si was not significant. The analyses of the test results revealed an essential role of fuel bound oxygen, the influence of which on the work produced by the engine, the brake specific fuel consumption, and the brake thermal efficiency depends on the combustion conditions. Therefore the role of fuel bound oxygen in completing the combustion process, ensuring effective performance of an engine and emission parameters can be different with often opposite trends determined for overall lean, moderate and richer air–fuel mixtures. The test results of a heavy-duty, DI Mercedes-Benz diesel engine showed that the brake thermal efficiency of ethanol–diesel fuel blends is slightly higher or equal to the corresponding diesel fuel case [23]. The thermal balance of a constant speed diesel engine running on 5 vol% and 10 vol% ethanol–diesel blends and fumigated ethanol was not different (5%) compared to normal diesel. However, the thermal balance was significantly different compared to diesel fuel when operating on the 15 vol% and 20 vol% ethanol–diesel fuel blends [20]. The test results showed that better performance of an engine on overall lean k = 5.5 and moderate k = 3.0 mixtures ensured less oxygenated ethanol–diesel fuel blends E5 and E10 in the speed range of 1400–1800 rpm. The matter is that to the fuel bound oxygen and the cetane number, as dominant factors, a big group of other players contributed to the improved net work at part-throttle operation. Adding of ethanol to diesel fuel not only increases the fuel-side oxygen, but also improves the fuel injection, atomisation and mixing of the fuel vapours with the in-cylinder compressed air because of a lower density, viscosity, and boiling point of the fuel blend. The test results of an optically-accessible, common-rail diesel engine showed that the spray tip penetration of fuel JP-8 is 16% shorter compared to diesel fuel case at the injection pressure of 30 MPa, and 10% at higher injection pressure of 140 MPa. The decreased spray tip penetration was compensated by 15.9–6.2° wider spray angle of JP-8 under measured fuel injection pressures of diesel fuel [47]. The variation in the spray tip penetration and spray angle is a result of the differences in fuel

713

properties, such as density, boiling point, and viscosity. These properties of JP-8 contribute to higher fuel–air mixing rate and improve atomisation, resulting from shorter spray tip penetration and wider spray angle [48]. One can suppose that adding of a lighter ethanol to diesel fuel leads to similar changes in the fuel injection parameters and improves the air and fuel vapours mixing [33]. 3.6. Exhaust emissions from combustion of ethanol–diesel–biodiesel blends 3.6.1. The effectiveness of fuel oxygen content in NOx reduction The total NOx emission consists of 90–95% of nitric oxide NO formed within and beyond the flame front and the balance percentage belongs to nitrogen dioxide NO2 emissions. The fuel combustion occurs within a narrow (0.1 mm) peak temperature oxidation zone under high pressure and thus time spent by the burned gases in the reaction zone is extremely limited. Therefore the temperature beyond the flame front is almost always higher than within the flame front itself. The increase of temperature occurs because the burned products are further compressed by the combustion process which continues in the cylinder. Consequently, the major part of nitric oxide is formed within a high temperature (P2000 K) zone beyond the flame front and this part of the NO prevails over that produced within the flame front. For this reason, the production of NO emissions depends on the amount of fuel premixed for the autoignition delay, the rapid burning of which takes place in the first combustion phase under increasing pressure and temperature in the cylinder. The mixture of oxygen (O), oxygen gas (O2) and nitrogen gas (N2) accumulated in some combustion chamber parts also affects the NO production [49]. The production of total NOx emissions depends on the engine and combustion chamber design, the fuel composition, the injection system and the air and fuel mixture used for experiments. It can be seen (Fig. 10) that the NOx emissions boosted up in the range of 260 ppm–320 ppm and 203 ppm–266 ppm because of the fuel oxygen content increased from 0.4 wt% to 3.9 wt% for lean air–fuel mixtures k = 5.5 used at 1400 and 1800 rpm speeds. The ethanol oxygen improved combustion of close to stoichiometric and fuel-rich mixtures in some local the combustion chamber areas. Promoted by a higher gas temperature in the cylinder and presence of fuel oxygen, the NOx emissions produced from combustion of fuel blends E5 and E10 increased by 16.5% and 23.1% compared to ordinary diesel running on lean air–fuel mixture at 1400 rpm speed. The NOx emissions increased owing to 10.0% (E5) and 15.4% (E10) higher amounts of the nitric oxide NO compared to standard diesel (241 ppm) since the emissions of nitrogen dioxide NO2 were 28.4% and 18.8% lower than 53.8 ppm produced from the normal diesel fuel. It is interesting to note that the NOx emissions decreased with the increase of the ethanol oxygen mass content in the blend for all air–fuel ratios at rated 2200 rpm speed. Sayin and Canakci [55] tested a single-cylinder, DI diesel engine at constant 15 and 30 Nm torques and 1000–1800 rpm speeds examining the influence of both injection timing and addition from 0 vol% to 5 vol% of ethanol to diesel fuel. Researchers found that the NOx formation depends on the combustion temperature, the biofuel oxygen content and the reaction time. The use of low percentage methanol–diesel and ethanol–diesel fuel blends caused a decrease in the smoke opacity, CO, and HC emissions, however, the NOx emissions were also increased with the use of oxygenated blends at part-throttle operation and low speeds. Whereas, the NOx emission 11.4% higher than the normal diesel fuel produced the combustion of less oxygenated blend E5 only used at an air–fuel ratio of k = 3.0 and 1400 rpm speed. The first reason of such variation is that the production of NOx is unequally sensitive to the cooling effects caused by ethanol–diesel fuel sprays at different air–fuel ratios and engine speeds, most likely therefore

714

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

Fig. 10. The total NOx emissions as a function of fuel oxygen mass (wt%) content for various air–fuel equivalence ratios and engine speeds.

temporal NOx emissions increase ends up at different percentages of ethanol added to diesel fuel. The NOx emissions decreased with the increase of the ethanol oxygen content in the fuel blend when running on richer combustible mixtures k = 3.0 and 1.5 almost for all engine speeds. The NOx emission, produced by the combustion of ethanol–diesel fuel blends, was lower than that emitted from regular diesel in the speed range of 1400–2200 rpm with the reduction being higher, the higher the percentage of ethanol added to diesel fuel. The maximum NOx emissions decreased despite low cetane number of the blends, long autoignition delay (Fig. 3), more fuel premixed for rapid combustion and nearly the same or slightly higher than normal diesel the maximum heat release rate (Fig. 4). The second reason why the NOx emissions were decreased can be associated with the stoichiometric air–fuel equivalence ratio which is 37.3% lower for the ethanol compared to that 14.45 of diesel fuel. As a result, less atmospheric air-burn oxygen participated in the NO creation from overall identical combustible mixtures, whereas the role of ethanol conserved oxygen seems as being not so much significant in NO and NOx production. The third reason is that the combustible mixture of ethanol–diesel fuel blends was less heterogeneous compared to the normal diesel fuel and the fuel spray cone angle was wider, which improved the air and fuel vapours mixing, contributed to more uniform temperature distribution in the cylinder and thus reduction of NOx. Park et al. [28,33] examined the influence of ethanol (99.9%) and diesel fuel blends on combustion and exhaust emissions of a Bosch common-rail, diesel engine run at 1500 rpm and various injection timings. By using high-speed camera Photron, Fastcam-APX RS authors estimated that 10 vol% and 20 vol% ethanol–diesel fuel blends have a shorter spray tip penetration, a larger spray cone angle and smaller droplets compared to pure diesel fuel. An increase in the ethanol blending ratio led to a decrease in the NOx and increase in the CO and HC emissions at the same loads and injection timings. The lower NOx emissions match well with the test results of various ethanol–diesel fuel blends obtained by other researchers [6,32,50]. The lower heating value of ethanol and the higher latent heat of evaporation caused cooling of the fuel sprays and decreased gas temperature at the beginning of combustion. This led to overall

lover temperature levels of the entire engine cycle remaining less reason to increase the NOx emission. Better homogeneity of the air and fuel mixture, fine atomisation of ethanol–diesel fuel droplets and equal distribution of injected portions accompanied by a slight decrease in the spray tip penetration and a slight increase in the spray cone angle [28,33] also contributed to lower NOx emissions because of reduced gas pressure and temperature in the cylinder. On the one part, the lower net heating value and higher evaporation cooling effect of ethanol–diesel fuel blends along with a later start of injection may have lower flame temperatures and burning velocities under unfavourable performance conditions than those of the base diesel fuel, which suppressed NOx formation. On the other part, adding of ethanol not only lowered the cetane number and prolonged ignition delay, but also increased oxygenated fraction of the fuel blends, which contributed to NOx formation. The opposing interaction between influencing factors might not lead to fuel oxygen significant effect on the NOx emission [50]. The test results of a naturally aspirated, DI diesel engine [35] showed that the brake specific fuel consumption increases, the thermal efficiency improves, the autoignition delay gets longer and the total combustion shortens when running on ethanol–diesel fuel blends. As a result, the NOx and smoke emissions decreased, with the CO emissions being higher for low and medium loads. The combustion of ethanol–diesel fuel blend E15, which contained the biggest amount of ethanol oxygen, produced NOx emissions 12.1%, 15.9% and 11.8% lower than a straight diesel operating on slightly richer mixtures k = 1.5 at 1400, 1800 and 2200 rpm speeds. The maximum temperature in the cylinder plays a key role in NOx production therefore adding of the 15 vol% of ethanol to diesel fuel reduced the NOx emissions when operating at the fully opened throttle. According to a big number of experimental test results of various types of biofuels [39], the combustion efficiency of ethanol–diesel fuel blends and the NOx production largely depend on the ignition conditions. At low loads, the cooling effect of ethanol fuel with a higher latent heat of evaporation can be more dominant than the combustion promotion by the oxygen content in the ethanol and diesel fuel blend [33]. A low cetane number of ethanol–diesel fuel blends is the main cause, which aggravated

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

715

Fig. 11. The carbon monoxide CO emissions as a function of fuel oxygen mass (wt%) content for various air–fuel equivalence ratios and engine speeds.

the autoignition of fuel sprays, retarded the start of combustion and its end, reduced the brake thermal efficiency and the flame temperature related NOx emissions. The test results of the International T 444E HT diesel engine fuelled with 5% and 10% ethanol–diesel fuel blends also showed decrease in the NOx emissions by nearly 3%. Therefore, authors assume that ethanol could act as an effective NOx emissions reducing additive [11]. In that research, the blend E10 contained 1% GE Betz additive, 0.11% Lubrizol OS# 147341C cetane improver, and the balance (88.89%) was D2 fuel. Rakopoulos et al. [21] determined that evaporation of the E15-D blend goes faster, the autoignition delay increases, premixed combustion peak goes higher and sharper, emissions of nitric oxide and exhaust soot densities decrease compared to normal diesel operation at both 38% and 75% of full load. Both emissions reduced in spite NO forms within the leaner areas at the jet tip and at the periphery in a high temperature zone whereas soot appears in the fuel-rich areas and between them well known trade-off exists. A comparative study showed a significant decrease in NOx emissions with increasing ethanol concentration in a new DI John Deere 5301 tractor engine and 4039DF008 multi-fuel engine and increase NOx emissions from old Ford 4000 tractor engine run at similar testing conditions [15]. However the CO emissions of Ford and both John Deere engines increased because of the addition of the ethanol. These results show that the combustion chamber design, fuel injection and atomisation quality and engine edging over the years also have influence on the NOx, CO, HC production. The decrease of NOx emissions occurred mainly due to unfavourable performance conditions because adding of the ethanol to diesel fuel delayed actual starts of injection and combustion, lowered gas temperature in the cylinder and crank angles AI 50 and AI 90, corresponding to 50% and 90% of heat release, relocated towards a bigger cylinder volume in the expansion stroke. Other biodiesel tests conducted on the International V-8 diesel fuelled with various biodiesel blends also showed that no correlation exists between the fuel oxygen fraction and the total NOx emissions [17]. A summary of recent results and issues extracted from the

work on various ethanol, petrol and rapeseed oil blends [30] leads to conclusion that the ethanol oxygen and oxygenated additives can contribute to the production of more NOx emissions only if the combustion process goes normally with the same or higher brake thermal efficiency. In contrast to the diesel-alcohol blends case [51], the NOx emissions produced by the combustion of soybean, yellow grease and genetically modified soybean methyl esters [52], rapeseed oil methyl ester [53] and other fatty acids or vegetable oil derived fuels [34] are almost always higher at full load operation. According Ref. [52], the NOx emissions are dependent on the start of fuel injection, combustion timing and the energy released in premixed combustion phase that, in case of using SME, caused higher gas pressure and combustion temperature inside the cylinder and increased the NOx emissions. Wenzel et al. [54] conducted tests on 4.5 L John Deere engine fuelled with standard No. 2 diesel fuel and three 2 vol%, 20 vol% and 100 vol% blends with soybean methyl ester. Researchers showed that using of EGR with rate of 0.5%, 1.5–2% and 5–5.5% is an effective measure for reduction of NOx emissions of an off-road compression ignition engine. Adding of the 5 vol% of RME as co-solvent improved the cetane number, net heating value of the fuel blend and performance efficiency of an engine (Fig. 9), however boosted up the flame temperature related NOx emissions for slightly richer combustible mixtures k = 1.5 over the entire speed 1400–2200 rpm range (Fig. 10). The higher both the engine performance efficiency and the gas temperature in the cylinder supported by more active in the reactions RME oxygen stimulated the growth of NOx emissions rather than the decrease in the cetane number of blend E15B. 3.6.2. The influence of fuel bound oxygen on CO production The CO emissions reduced to minimum values of 265 ppm (25.1%) and 189 ppm (21.9%) when running on the 5 vol% ethanol–diesel fuel blend compared to normal diesel operating on overall lean k = 5.5 and moderate k = 3.0 mixtures at low 1400 rpm speed mainly (Fig. 11). The lower CO emissions correlate well with a higher NOx emissions produced by the combustion of less

716

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

Fig. 12. The total unburned hydrocarbons HC as a function of the cetane number (a) and fuel oxygen mass (wt%) content (b) for various air–fuel equivalence ratios and engine speeds.

oxygenated blends E5 and E10 at respective air–fuel ratios within the speed range of 1400–1800 rpm. The test results of two-cylinder, DI Kubota diesel generator GL7000 showed that using of biodiesel–alcohol blends reduces the NO and PM emissions while increasing both CO and HC emissions at below 70% loads compared to normal diesel [56]. Nevertheless, when the alcohol concentration is low enough (5 vol%), the CO emissions can even decrease because alcohols improve oxygen availability within fuel-rich patterns of overall lean mixture. However, the CO emissions increased at a rate proportional to fuel oxygen content in the blend and reached a nearly 85.6% higher (E15B) level than normal diesel produces running at an air–fuel ratio of k = 5.5 and 2200 rpm speed. It is important to note that the influence of fuel oxygen on the NOx (Fig. 10) and the CO (Fig. 11) emissions produced by the combustion of ethanol–diesel–biodiesel blends for overall lean, moderate and slightly richer air–fuel mixtures depends upon the engine speed. At a higher 2200 rpm speed, harmful CO emissions can arise even from combustion of lean and moderate air–fuel mixtures since the oxidation processes in fuel-rich regions occur at close to stoichiometric conditions. The CO emissions increased significantly with the addition of ethanol to diesel fuel because biofuel transition from the liquid phase to gas phase advances slowly enough in overall lean mixtures and thus there is a lack of the chemical reaction time

for complete burning of the fuel at high 2200 rpm speed. For this reason, the CO emissions from combustion of blend E15B were 15.7%, 51.6% and 2.2 times higher than normal diesel produces when running on an air–fuel ratio of k = 3.0 at 1400, 1800 and 2200 rpm speeds. The CO emissions boosted up since the lower cetane number of the fuel blend increased the autoignition delay time at part-throttle operation. A large excess of CO and HC (Fig. 12) emissions in the exhaust can be caused by low temperature in the cylinder and the flame thick quenching layer, which may occur because of high ethanol vaporisation cooling effect [50]. On the contrary to part-throttle operation, both the NOx and the CO emissions decreased with the increase of the ethanol oxygen content in the fuel blend when running on overall richer mixtures k = 1.5 within the tested speed range of 1400–2200 rpm. The CO emissions produced by the combustion of ethanol–diesel fuel blends were 6.7% (E15), 3.6% (E10) and 5.1% (E15) lower compared to regular diesel operating at respective speeds. The difference between CO emissions produced by the combustion of ethanol–diesel fuel blends was not significant since general CO level decreased substantially for richer air–fuel mixtures. Thus, the combustion of fuel blend E15B produced the CO emission 3.9% lower, then 14.7% and 1.0% higher compared to normal diesel operating on richer air–fuel mixtures k = 1.5 at corresponding speeds. Comparison of Figs. 10 and 11 shows that the NOx

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

717

Fig. 13. Smoke opacity of the exhaust as a function of fuel oxygen mass (wt%) content for various air–fuel equivalence ratios and engine speeds.

and the CO emissions quite differently responded to the increase of the fuel oxygen content stored in overall lean and moderate air– fuel mixtures, but both dangerous pollutants friendly decreased with the increase of the ethanol fuel oxygen content for richer air–fuel mixtures k = 1.5 used at all engine speeds. The only difference in the behaviour of these species was that general level of the NOx emissions almost for all air–fuel ratios, excluding values of k = 1.5, increased with the increase of engine speed, whereas the level of CO emissions decreased for a higher speed of an engine. The opposing trend of carbon monoxide CO variation with the fuel oxygen content at k = 1.5 and high 2200 rpm speed, which is higher than the CO emissions measured at 1800 rpm speed, looks a bit unusual, however this occurred due to changes in the burning conditions (Fig. 11). This result can be attributed to the lower bmep (Fig. 7) developed from the same energy stored in the fuel blends and the higher bsfc rate (Fig. 8) at 2200 rpm speed, which can be regarded as a primary reason, and a greater scavenging degree of the engine cylinder, which occurred at about 1800 rpm speed, as a second-rate cause. The matter is that for the engine D-243 not only the overlap of inlet and exhaust valves is adjusted for the speed region of 1400–1800 rpm, but the form and the length of a suction manifold are also selected for the maximum dynamic aircharge, which pushes the burned gases out of the cylinder, improves the volumetric efficiency of an engine and fuel combustion. 3.6.3. Dependency of THC emissions on cetane number and the fuel oxygen mass content Fig. 12a and b show dependencies of the total unburned hydrocarbons (THC) on cetane number and the fuel oxygen mass content for specific air–fuel equivalence ratios and the three ranges of engine speed. The emission of hydrocarbons HC increased with the decrease of the cetane number in the range of 51.5–46.7 and became 33.6%, 27.9%, 25.7% higher for lean air–fuel mixture k = 5.5, and 33.3%, 21.2%, 16.8% higher for slightly richer mixtures k = 1.5 used at 1400, 1800 and 2200 rpm speeds. The growth of THC emission, which was caused by the cooling effect of ethanol and lower cetane number of the fuel blends, can be regarded as a normal occurrence observed by other researchers [16]. The THC emissions rely on the combustion quality, engine load and speed, remaining within some limits set by chemical and

physical properties of the fuel blend. Rakopoulos et al. [23] tested DI Mercedes-Benz diesel engine running alternately on two 5 vol% and 10 vol% ethanol–diesel fuel blends at 1200 and 1500 rpm speeds and three 20%, 40% and 60% of the full loads. It was stated that low concentration ethanol–diesel blends significantly decrease soot of the exhaust, remain similar or slightly reduce the NOx, the CO emission does not change for two cases and all the others reduce and increase HC emissions compared to neat diesel fuel. It can be seen in Fig. 12a, b that the cetane number of ethanol– diesel–biodiesel blends reflects the THC emissions changing behaviour more predictably than the content of fuel oxygen does, the side effects of which are highly dependent on chemical structure and composition of the fuel blends. An uninspected THC emissions decrease almost always started at the cetane number critical value of about CN  44.4 and the ethanol oxygen mass content of 3.9 wt%, at which the total unburned hydrocarbons decreased to the lowest values for all air–fuel ratios and engine speeds. The HC emissions varied from the highest values emitted from combustion of the fuel blend E10 to the lowest ones measured for ethanol oxygen mass content of 5.6 wt% (E15) and boosted up again for mostly oxygenated blend E15B (6.1 wt%) used at all air–fuel ratios and engine speeds (Fig. 12b). Compared to the normal diesel fuel, the THC emissions reduced in the range of 918 ppm to 332 ppm (2.8 times) for overall lean mixture k = 5.5 when running at rated 2200 rpm speed and 570 ppm to 18 ppm (31.7 times) for slightly richer mixture k = 1.5 at low 1400 rpm speed due to the use of ethanol–diesel fuel blend E15. This end result can be taken as an indicator that an engine would only be able to operate properly on small, up to 10 vol%, ethanol–diesel fuel blends as it would not be longer at full efficiency. Both extremely low cetane number of ethanol and high latent heat of evaporation caused long autoignition delay and slow burning over late phases of the combustion. The HC oxidation later in the expansion stroke, after combustion stopped, resulted in lower THC emission and residual oxygen content in the exhaust accompanied by a higher gas temperature, but late-phase fuel combustion did not improve performance efficiency and effective power developed by the engine. Thus, the fuel oxygen mass content of 3.9 wt% stored in ethanol–diesel fuel blend E10 can be regarded as a critical value, at which the positive effect of ethanol in diesel

718

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

fuel ends up. Adding the 5 vol% of RME to ethanol–diesel fuel blend E15 improved the cetane number (E15B) and engine performance efficiency, slightly increased THC emission, residual oxygen content and decreased temperature of the exhaust. 3.6.4. The influence of fuel bound oxygen on smoke opacity and other emissions The smoke opacity was below 4% level in the speed range of 1400–2200 rpm when operating on blend E15B at an air–fuel ratio of k = 5.5 (Fig. 13). Opacity of the exhaust decreased due to the simultaneous addition of ethanol and RME to diesel fuel, however common smoke level slightly increased for richer combustible mixtures k = 3.0 and 1.5 used at all speeds. The combustion of oxygenated blends E15 and E15B produced smoke opacity 46.2% and 38.5% lower than conventional diesel emits at an air–fuel ratio of k = 3.0 and 1400 rpm speed. However, smoke and other pollutants increased when running on ethanol–diesel fuel blends E5 and E10 at higher 1800 and 2200 rpm speeds since emerging white-coloured unburned fuel vapours reduced transparency of the exhaust. Biodiesel (E15B) produced smoke of the exhaust about the same 4.7% as a straight diesel running at part-throttle and rated 2200 rpm speed because of a greater activity of RME oxygen in the oxidation reactions. One should inject more to compensate for a lower net heating value of the fuel blend when an engine operates under heavy load. This causes the air and fuel mixture to be richer, and produces a higher amount of carbon particles soot. The exhaust smoke produced by a fully loaded normal diesel reached the highest level of 33.7% for 1400 rpm speed at which maximum torque occurs. Using of ethanol–diesel–biodiesel blend E15B improved combustion of slightly richer mixture k = 1.5 and smoke of the exhaust reduced by 26.1% at low 1400 rpm speed. The fuel oxygen suggested benefits in oxidation of some fuel-rich parts at close to stoichiometric conditions and limited the air-swirl turbulence intensity that resulted in less unburned carbon particles exhausted from biodiesel. Since the turbulence intensity and the in-cylinder air swirl increased for a higher 1800 rpm speed, the fuel bound oxygen contributed with a minor degree of improvement. Therefore the formation of visible smoke from combustion of oxygenated blend E15B was 8.4% greater than regular diesel produces at an air–fuel ratio of k = 1.5. The fuel oxygen regained indigenous advantages and the exhaust smoke reduced by 15.6% against normal diesel operating on richer air–fuel mixtures because overall smoke level increased for rated 2200 rpm speed. The studies with a single-cylinder, DI diesel engine revealed ‘‘linearity’’ between engine emissions and oxygen content of the intake air, whereas ‘‘nonlinearity’’ also was evident between soot emissions and oxygen content in oxygenated fuels [13]. Since there was a lack of time for the fuel spray tips development in the combustion chamber, evaporation and homogeneous mixing of the air and fuel vapours at a higher speed, the ethanol and RME oxygen contributed to reducing of soot formation in some the fuel-rich regions of the combustion chamber. Moreover, simultaneous addition of ethanol and RME to diesel fuel naturally reduced the content of sulphur in the fuel blend and thus harmful SO2 emission contributing to lower smoke that can be regarded as environment friendly advantage. The test results of a Cummins-4B diesel running on ethanol (5 vol%), methyl soyate (20 vol%) and diesel fuel blend (75 vol%) showed a significant decrease in PM emissions, but researchers measured 2–14% higher NOx emissions at most tested modes [31]. A common feature of carbon monoxide CO emission (Fig. 11) and smoke opacity (Fig. 13) is their overall lower level produced by the combustion of slightly richer air–fuel mixtures k = 1.5 at 1800 rpm speed compared to that observed at 2200 rpm speed.

This might be caused by the two reasons: (1) the higher scavenging degree of an engine cylinder reduced CO emissions and smoke opacity at medium 1800 rpm speed; (2) the lower brake mean effective pressure (Fig. 7) developed by the same energy and more fuel consumed per unit of effective power (Fig. 8) contributed to higher CO emissions and smoke densities at rated 2200 rpm speed. The effect of fuel oxygen on the engine performance efficiency and smoke opacity depends on testing conditions. Experiments in a steel combustion chamber showed that the increasing percentage of alcohol in alcohol–diesel blend beyond 10 vol% does not improve combustion or reduce harmful pollutants and soot any more [22]. However, the test results of a single-cylinder, diesel engine with a swept volume of 582 cm3 showed that the smoke coefficient decreased by 11–25% for 20% of ethanol added to diesel fuel and by 18.5–33.3% for 15% of ethanol in the blend [19]. In that study, the analysis of the test results performed for different rack settings while keeping the total energy of various ethanol–diesel blends constant for the entire range of 1000–2000 rpm speed. The test results of a HSDI ’’Hydra’’ diesel engine showed that the autoignition delay was higher, both the heat release rate and gas temperature in the cylinder were slightly lower than normal diesel with no big difference in the maximum combustion pressure created by ethanol–diesel fuel blend E15-D. The brake specific fuel consumption and thermal efficiency were slightly higher, the NOx and CO emissions lower, the THC increased and the exhaust soot significantly reduced by the ethanol oxygen compared to normal diesel. In that research [2], the fuel-air equivalence ratios for oxygenated blends were a little lower because the comparison of engine operation was performed for constant speed of 2000 rpm and various bmep = 0, 1.40, 2.57 and 5.37 bars. The CO2 emissions produced by biodiesel strongly depend on the fuel composition, particularly C/H ratio, and the brake specific fuel consumption. Powered by composite blend E15B, a fully loaded engine generated CO2 emission about the same 8.8– 8.9 vol% as a regular diesel produces in the speed range of 1400– 1800 rpm. Biodiesel emitted 1.9% less CO2 emissions at full throttle compared to 9.1 vol% of normal diesel running at rated 2200 rpm speed. Whereas, ethanol–diesel fuel blend E15 suggested the CO2 emission 1.7% and 3.0% higher for 1400 and 1800 rpm speeds and about the same as a straight diesel for 2200 rpm speed. It is thought [7] that using of renewable fuels does not make significant contribution to global warming because biofuels have formidably positive environmental properties and very low sulphur content that mitigates the ‘‘green-hose’’ effect and climate changes. The residual oxygen content O2, vol% for blend’s E15 case was 2.1%, 2.8% and 1.0% lower compared with corresponding vales of 9.0 vol%, 8.9 vol% and 8.8 vol% of normal diesel running on richer mixtures k = 1.5 at 1400, 1800 and 2200 rpm speeds. Whereas the temperature of the exhaust was 4.5–6.0% higher than that, 418–460 °C, of normal diesel tested at respective conditions. The lower residual oxygen content and higher temperature of the exhaust show that the fuel oxidation reactions continued over late phases of the expansion stroke. Using of blend E15B improved combustion and, therefore, the oxygen residue varied in the range from being 0.5% higher to 0.8% lower compared to regular diesel. The exhaust temperature also decreased by 6–8 °C, but it still was 2.9–4.2% higher than conventional diesel. The accomplished study shows that the autoignition delay, combustion, engine performance efficiency and clean stream of the exhaust not only depend on the fuel oxygen mass content, which takes part in the oxidation reactions, but also strongly belong on the chemical structure of the blend. The influence of multicomponent biofuel on combustion and conversion of chemical energy to mechanical energy and heat can be different because – OH group of ethanol is strongly bound and –COOCH3 group of rapeseed oil methyl ester is weakly bound. A key point is that the

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

chemical structure of ethanol CH3CH2OH, in which the hydroxyl functional group (–OH) is attached to a carbon atom by a single (strong) bound, may aggravate complete alcohol oxidation. In particular this carbon center should hardly be saturated, having also single and strong bounds to three other atoms. In contrast to the ethanol case, the chemical structure of rapeseed oil methyl ester CH3(CH2)nCOOCH3 predicts that some part of oxygen is in double bounds with radicals that increases oxidation chances and leads to cleaner combustion and fewer carbon monoxide and hydrocarbon exhaust emissions [1].

4. Conclusions An extended experimental study was conducted with a fourcylinder, naturally aspirated, DI diesel engine to evaluate the effects of using 5 vol% (E5), 10 vol% (E10), 15 vol% (E15) blends of anhydrous (99.8%) ethanol with the normal diesel fuel and ethanol (15 vol%), diesel fuel (80 vol%), biodiesel (5 vol%) blend (E15B). The series of load characteristics were taken running on each of the above fuels to have engine performance parameters for three ranges of 1400, 1800 and 2200 rpm speed. The measured experimental parameters of the engine performance were plotted as a function of the air and fuel ratio for respective speeds. Deviations of the measured combustion, performance and emission parameters from the baseline parameters of the normal diesel fuel (class 2) were determined for the same air–fuel ratios of k = 5.5, 3.0 and 1.5 at three ranges of speed to disclose eventual reasons of changes caused by the addition of ethanol and RME to diesel fuel. The analyses of the test results revealed that the fuel oxygen mass content reflects changes of the autoignition delay time caused by the use of ethanol–diesel–biodiesel blends more predictably than the cetane number does. The autoignition delay time si for mostly oxygenated blend E15B was 15.4% longer than standard diesel (0.78 ms) running on slightly richer air–fuel mixture k = 1.5 at rated 2200 rpm speed. The influence of fuel bound oxygen on maximum heat release rate HRRmax in the first phase of combustion is ambiguous enough. The major positive influence of fuel oxygen on the HRRmax value was registered for overall lean combustible mixtures k = 5.5, which have slower flame speeds, and the smallest one for overall richer mixtures k = 1.5, which normally burn with a faster flame speeds [43]. The heat release history representing crank angles AHRRmax, AI 50 and AI 90 increased due to the addition of 5 vol%, 10 vol% and 15 vol% of ethanol for almost all air–fuel ratios and engine speeds because of the later start of injection, greater cooling effect caused by ethanol fuel evaporation and longer autoignition delay. Adding the 5 vol% of RME to ethanol–diesel fuel blend E15 improved combustion and, in almost all cases, relevant crank angles AHRRmax, AI 50 and AI 90 relocated close to TDC since the cetane number of biodiesel and other properties are similar to the normal diesel fuel. The peak combustion pressure pmax decreased to some extent with the increase of the ethanol oxygen content in the fuel blends almost for all engine performance modes with the biggest 6.3% reduction measured for the fuel blend E15 used at overall lean air–fuel mixture k = 5.5 and 1800 rpm speed. The maximum combustion pressure pmax increased by 2.1 bar (3.1%) and 1.2 bar (1.9%) when running on composite blend E15B against values of 68.1 bar and 64.6 bar of normal diesel operating on richer air–fuel mixtures k = 1.5 at 1400 and 2200 rpm speeds. The respective angles Apmax moved away from TDC due to the increase of fuel bound oxygen in the blend actually for all air–fuel ratios and engine speeds with the biggest dislocation of 2.9° CADs caused by the use of blend E15B at lean combustible mixture k = 5.5 and medium 1800 rpm speed. Biodiesel suggested the same

719

or slightly smoother operation for almost all performance modes, however, the maximum gradients of in-cylinder pressure rise were 1.65 bar/° and 0.61 bar/° higher than normal diesel running on richer air–fuel mixtures k = 1.5 at 1400 and 2200 rpm speeds. Fuelled with blends E15 (5.6 wt%) and E15B (6.1 wt%) the diesel engine developed the brake mean effective pressure 16.1% higher than normal diesel consuming the same diesel fuel energy accumulated by the leaner combustible mixture k = 5.5 at rated 2200 rpm speed. The composite blend E15B suggested the bmep 1.2% and 0.6% higher than a straight diesel develops from richer air–fuel mixtures k = 1.5 at 1400 and 1800 speeds too, however, the bmep was about 1% lower at rated 2200 rpm speed. The brake specific fuel consumption was 8.2%, 7.2% and 6.5% higher for the most oxygenated blends E15 and E15B compared with values of normal diesel operating on richer combustible mixtures k = 1.5 at 1400, 1800 and 2200 rpm speeds. Run on blend E15B, the engine developed the brake thermal efficiency the same of 0.362 as a straight diesel operating on slightly richer air–fuel mixture k = 1.5 at rated 2200 rpm speed. The maximum NOx and the HC emissions generated by the combustion of composite fuel blend E15B were 6.3%, 11.9%, 9.5% and 24.6%, 14.6%, 15.1% lower than normal diesel produces at 1400, 1800, 2200 rpm speeds. The CO emission was 3.9% lower, 14.7% and 1.0% higher, and smoke opacity was 26.1% lower, 8.4% higher, and again 15.6% lower when running on blend E15B at richer air–fuel mixtures k = 1.5 and respective speeds. The lower NOx, HC emissions along with positive changes in the CO emissions and opacities of the exhaust revealed an exclusive role of fuel bound oxygen. The test results show that ethanol (15 vol%), diesel (80 vol%) and biodiesel (5 vol%) blend could be efficiently used for diesel engine powering. The simultaneous addition of anhydrous (200 proof) ethanol and RME to commercial diesel fuel suggests ecological advantages and increases the renewable biofuel concentration in the blend that is one of the targets recommended by the EU Directive 2009/28/EC. Acknowledgements The study authors acknowledge the AVL Company’s Managers Gerald Sommer, Carl Steinmeyer, Bjorn Johansson, Michael Gustafsson, and other highly professional experts for their contribution to the arrangement and update of the diesel engine test equipment at Aleksandras Stulginskis University (ASU, Lithuania). References [1] Minteer S. Alcoholic fuels Saint Louis University Missouri. United States of America, 6000 Broken Sound Parkway NW, Suite 300: CRC Press at Taylor & Francis Group, LLC; 2006. 270 p. [2] Rakopoulos CD, Antonopoulos KA, Rakopoulos DC. Experimental heat release analysis and emissions of a HSDI diesel engine fuelled with ethanol-diesel blends. Energy 2007;32(10):1791–808. [3] Hansen AC, Zhang Q, Lyne P-WL. Ethanol-diesel fuel blends – a review. Bioresour Technol 2005;96(3):277–85. [4] Sorda G, Banse M, Kemfert C. An overview of biofuel policies across the world. Energy Pol 2010;38(11):6977–88. [5] Sims REH, Mabee W, Saddler JN, Taylor M. An overview of second generation biofuel technologies. Bioresour Technol 2010;101(6):1570–80. [6] Rakopoulos DC, Rakopoulos CD, Giakoumis EG, Dimaratos AM, Kyritsis DC. Effect of butanol-diesel fuel blends on the performance and emissions of a high-speed DI diesel engine. Energy Convers Manage 2010;51(10):1989–97. [7] Demirbas A. Biofuels sources, biofuel policy, biofuel economy and global biofuel projections. Energy Convers Manage 2008;49(8):2106–16. [8] Shropshire GJ, Goering CE. Ethanol injection into a Diesel-engine. Trans ASAE 1982;25(3):570–5. [9] Hansen AC, Taylor PW, Lyne L, Meiring P. Heat released in the compressionignition combustion of ethanol. Trans ASAE 1989;32(5):1507–11. [10] Hansen AC, Zhang Q. Engine durability evaluation with E-diesel. Paper Number: 036033. An ASAE meeting presentation, Las Vegas, Nevada, USA; 27–30 July 2003. p. 13.

720

G. Labeckas et al. / Energy Conversion and Management 79 (2014) 698–720

[11] Hansen AC, Gratton MR, Yuan W. Diesel engine performance and NOx emissions from oxygenated biofuels and blends with Diesel fuel. Trans ASABE 2006;49(3):589–95. } Çelikten Ì, Usta N. Effects of ethanol addition on performance and [12] Can O, emissions of a turbocharged indirect injection Diesel engine running at different injection pressures. Energy Convers Manage 2004;45(15– 16):2429–40. [13] Zannis TC, Pariotis EG, Hountalas DT, Rakopoulos DC, Levendis YA. Theoretical study of DI diesel engine performance and pollutant emissions using comparable air-side and fuel-side oxygen addition. Energy Convers Manage 2007;48(11):2962–70. [14] Satge de Caro P, Mouloungui Z, Vaitilingom G, Berge JCh. Interest of combining an additive with diesel-ethanol blends for use in diesel engines. Fuel 2001;80(4):565–74. [15] Jha SK, Fernando S, Columbus E, Willcutt H. A comparative study of exhaust emissions using diesel–biodiesel-ethanol blends in new and used compression ignition engines. Paper Number: 066138. An ASABE meeting presentation, Portland, Oregon, USA; 9–12 July 2006. p. 12. [16] Bhattacharya TK, Chatterjee S, Mishra TN. Performance of a Constant Speed CI Engine on Alcohol-Diesel Microemulsions. Appl Eng Agric, ASAE 2004;20(3):253–7. [17] Yuan W, Gratten MR, Hansen AC. Parametric investigation of NOx emissions from biofuels for compression-ignition engines. Paper number: 056115. An ASAE annual international meeting presentation, paper number: 056115, Tampa, Florida, USA; 17–20 July 2005. p. 21. [18] Li W, Ren Y, Wang X-B, Miao H, Jiang D-M, Huang Z-H. Combustion characteristics of compression ignition engine fuelled with diesel-ethanol blends. Proc ImechE D: J Autom Eng 2008;222:265–74. [19] Abu-Qudais M, Haddad O, Qudaisat M. The effect of alcohol fumigation on diesel engine performance and emissions. Energy Convers Manage 2000;41(4):389–99. [20] Ajav EA, Singh B, Bhattacharya TK. Thermal balance of a single cylinder diesel engine operating on alternative fuels. Energy Convers Manage 2000;41(14):1533–41. [21] Rakopoulos CD, Antonopoulos KA, Hountalas DT. Multi-zone modeling of combustion and emissions formation in DI diesel engine operating on ethanoldiesel fuel blends. Energy Convers Manage 2007;48(4):389–99. [22] Asfar KR, Hamed H. Combustion of fuel blends. Energy Convers Manage 1998;39(10):1081–93. [23] Rakopoulos CD, Rakopoulos DC, Kakaras EC, Giakoumis EG. Effect of ethanoldiesel fuel blends on the performance and exhaust emissions of heavy duty DI diesel engine. Energy Convers Manage 2008;49(11):3155–62. [24] Kwanchareon P, Luengnaruemitchai A, Jai-In S. Solubility of a diesel–biodieselethanol blend, its fuel properties, and its emission characteristics from diesel engine. Fuel 2007;86(7-8):1053–61. [25] Sayin C. Engine performance and exhaust gas emissions of methanol and ethanol-diesel blends. Fuel 2010;89(11):3410–5. [26] Lapuerta M, Armas O, Herreros JM. Emissions from a diesel-bioethanol blend in an automotive diesel engine. Fuel 2008;87(1):25–31. [27] Chen H, Shuai S-J, Wang J-X. Study on combustion characteristics and PM emission of diesel engine using ester-ethanol-diesel blended fuels. Proc Combust Inst 2007;31(2):2981–9. [28] Park SH, Youn IM, Lee CS. Influence of ethanol blends on the combustion performance and exhaust emission characteristics of a four-cylinder diesel engine at various engine loads and injection timings. Fuel 2011;90(2):748–55. [29] Labeckas G, Slavinskas S. Comparative performance of direct injection diesel engine operating on ethanol, petrol and rapeseed oil blends. Energy Convers Manage 2009;50(3):792–801. [30] Labeckas G, Slavinskas S. Study of exhaust emissions of direct injection diesel engine operating on ethanol, petrol and rapeseed oil blends. Energy Convers Manage 2009;50(3):802–12. [31] Shi X, Pang X, Mu Y, He H, Shuai S, Wang J, et al. Emission reduction potential of using ethanol–biodiesel–diesel fuel blend on a heavy-duty diesel engine. Atmos Environ 2006;40(14):2567–74. [32] Sendzikiene E, Makareviciene V, Janulis P. Influence of fuel oxygen content on diesel engine exhaust emissions. Renewable Energy 2006;31(15):2505–12. [33] Park SH, Youn IM, Lee CS. Influence of two-stage injection and exhaust gas recirculation on the emissions reduction in an ethanol-blended diesel-fueled four-cylinder diesel engine. Fuel Process Technol 2010;91(11):1753–60.

[34] Graboski MS, McCormick RL. Combustion of fat and vegetable oil derived fuels in diesel engines. Progress in energy and combustion, vol. 24. Scientific Elsevier Science Ltd; 1998. p. 125–64. [35] Xing-cai L, Jian-guang Y, Wu-gao Z, Zhen H. Effect of cetane number improver on heat release rate and emissions of high speed diesel engine fuelled with ethanol-diesel blend fuel. Fuel 2004;83(14–15):2013–20. [36] Torres-Jimenez E, Jerman MS, Gregorc A, Lisec I, Dorado MP. Physical and chemical properties of ethanol-diesel fuel blends. Fuel 2011;90(2):795–802. [37] Schleicher T, Werkmeister R, Russ W, Meyer-Pittroff R. Microbiological stability of biodiesel-diesel-mixtures. Bioresource Technol 2009;100(2):724–30. [38] Knothe G, Gerpen JV, Krahl J. The biodiesel handbook. Champaign, Illinois, USA: AOCS Press; 2005. p. 312. [39] Lotko W, Lukanin VN, Khachiyan AS. Using of alternative fuels in internal combustion engines. Moscow: MADI [Moscow Automobile and Highway Institute – State Technical University]; 2000. 311 p. [40] Labeckas GS. Some of the relationships between basic parameters of the fuel supply apparatus and the fuel injection and atomisation characteristics. Republican Interdepartmental Scientific Technical Journal ‘‘Internal Combustion Engines, vol. 46. Kharkov: Higher School Publishing; 1987. p. 52–9 [in Russian]. [41] Labeckas GS. The methodology and the results of regression analyses of the relationships between the main parameters of heat release and factors affecting them. Republican interdepartmental scientific technical journal ‘‘internal combustion engines’’, vol. 46. Kharkov: Higher School Publishing; 1987. p. 60–9 [in Russian]. [42] Labeckas G, Slavinskas S. Performance and emission characteristics of a direct injection diesel engine operating on KDV synthetic diesel fuel. Energy Convers Manage 2013;66(2):173–88. [43] Willard W. Pulkrabek (University of Wisconsin -Platteville) Engineering fundamentals of the internal combustion engine, vol. 07458. Upper Saddle River, New Jersey: Prentice Hall; 2009. p. 411. [44] Khachiyan AS, Labeckas GS. The effect of fuel injection and atomisation on heat release and performance of supercharged diesel engine: VII National Congress of the ‘‘Internal Combustion Engines and Combustion’’. India, National Institute of Petroleum, Surathkal; 7–8 of January, 1982. p. 7.01–7.12. [45] Kim H, Choi B. Effect of ethanol-diesel blend fuels on emission and particle size distribution in a common-rail direct injection diesel engine with warm-up catalytic converter. Renewable Energy 2008;33(10):2222–8. [46] Li D, Zhen H, Xingcai L, Wu-gao Z, Jian-guang Y. Physico-chemical properties of ethanol-diesel blend fuel and its effect on performance and emissions of diesel engines. Renewable Energy 2005;30(6):967–76. [47] Lee J, Bae C. Application of JP-8 in a heavy duty diesel engine. Fuel 2011;90(5):1762–70. [48] Myong K, Suzuki H, Senda J, Fujimoto H. Spray inner structure of evaporating multi-component fuel. Fuel 2007;87(2):202–10. [49] Heywood JB. Internal combustion engine fundamentals. Co – Singapore for manufacture and export (International edition); 1988. 930 p. [50] Song CL, Zhou YC, Huang RJ, Wang IQ, Huang QF, Lü G, et al. Influence of ethanol-diesel blended fuels on diesel exhaust emissions and mutagenic and genotoxic activities of particulate extracts. J Hazard Mater 2007;149:255–363. [51] Kumar S, Cho JH, Park J, Moon II. Advances in diesel-alcohol blends and their effects on the performance and emissions of diesel engines. Renew Sustain Energy Rev 2013;22(1):46–72. [52] Yuan W, Hansen AC, Tan Z. Modeling of NOx emissions of biodiesel fuels. Paper number: 056116. An ASAE meeting presentation, Tampa, Florida, USA; 17–20 July 2005. p. 12. [53] Labeckas G, Slavinskas S. The effect of rapeseed oil methyl ester on direct injection Diesel engine performance and exhaust emissions. Energy Convers Manage 2006;47(13–14):1954–67. [54] Wenzel SJ, Hansen AC, Yuan W. Combined impact of biodiesel and exhaust gas recirculation on NOx emissions. Paper number: 056115. An ASAE annual international meeting presentation, paper number: 066136, Portland, Oregon, USA; 9–12 July 2006. p. 12. [55] Sayin C, Canakci M. Effects of injection timing on the engine performance and exhaust emissions of a dual-fuel diesel engine. Energy Convers Manage 2009;50(1):203–13. [56] Yilmaz N, Sanchez TM. Analysis of operating a diesel engine on biodieselethanol and biodiesel-methanol blends. Energy 2012;46(1):126–9.