- Email: [email protected]
Diesel vehicle performance on unaltered waste soybean oil blended with petroleum fuels
Fuel 107 (2013) 757–765
Contents lists available at SciVerse ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Diesel vehicle performance on unaltered waste soybean oil blended with petroleum fuels Eugene P. Wagner ⇑, Patrick D. Lambert, Todd M. Moyle, Maura A. Koehle University of Pittsburgh, Department of Chemistry, Chevron Science Center, Pittsburgh, PA 15260, USA
h i g h l i g h t s " We examine engine performance on unaltered waste soybean oil-petroleum fuel blends. " Fuels containing 15%, 30%, 40% and 50% waste soybean oil were created for analysis. " Power and torque were analyzed on three vehicles on a chassis dynamometer. " Performance for 15%, 30% and 40% WVO blends averaged 1.1% less than pure diesel. " Performance for the 50% WVO blend was 5.4% lower than pure diesel.
a r t i c l e
i n f o
Article history: Received 4 October 2012 Received in revised form 23 January 2013 Accepted 27 January 2013 Available online 13 February 2013 Keywords: Soybean oil Petroleum diesel Engine Horsepower Torque
a b s t r a c t Interest in using unaltered vegetable oil as a fuel in diesel engines has experienced an increase due to uncertainty in the crude oil market supply and the detrimental effects petroleum fuels have on the environment. Unaltered vegetable oil blended with petroleum fuels is less expensive, uses less energy to produce and is more environmentally friendly compared to petroleum diesel or biodiesel. Here we investigate the engine performance of unaltered waste soybean oil blended with petroleum diesel and kerosene for three vehicles. Five biofuel blends ranging from 15% to 50% oil by volume were tested on a 2006 Jeep Liberty CRD, a 1999 Mercedes E300 and a 1984 Mercedes 300TD. A DynoJet 224x chassis dynamometer was used to test vehicle engine performance for horsepower and torque through a range of RPMs. Results for the Jeep showed a modest decrease in horsepower and torque compared to petroleum diesel ranging from 0.9% for the 15% oil blend to 5.0% lower for the 50% oil blend. However, a 30% oil blend showed statistically better performance (P < 0.05) compared to petroleum diesel. For the 1999 Mercedes, horsepower performance was 1.1% lower for the 15% oil blend to 6.4% lower for the 50% oil blend. Engine performance for a 30% blend was statistically the same (P < 0.05) compare to diesel. Finally, horsepower performance was 1.1% lower for the 15% oil blend to 4.7% lower for the 50% oil blend for the 1984 Mercedes. Overall, the performance on these oil blended fuels was excellent and, on average 1.1% lower than petroleum diesel for blends containing 40% or lower waste soybean oil content. The more significant decrease in power between the 40% and 50% oil blends indicates that oil content in these blended fuels should be no more than 40%. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Since the year 2000, gasoline and oil prices have tripled while use and demand has also grown worldwide. Countries with emerging economies, such as China and India, are increasing consumption of gas and oil at record rates. In addition to a potential shortage of fuel to meet the growing demand, there is great concern that the use of fossil fuels for energy is a leading cause of a global warming trend which could lead to undesirable climate ⇑ Corresponding author. Tel.: +1 412 624 2861; fax: +1 412 648 3297. E-mail address: [email protected] (E.P. Wagner). 0016-2361/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2013.01.052
change. Economical, environmentally friendly, and sustainable energy solutions are needed to slow the growth and use of fossil fuels. It is well known that vegetable oil can be used as an energy source in diesel engines. In fact, the first diesel engine developed by Rudolph Diesel was showcased at the 1900 World Fair where he demonstrated the use of peanut oil as a fuel for his engine [1]. It was not until Rudolph’s death that his original engine design was converted to operate on petroleum diesel due to the increasing availability of petroleum products. The concept and practice of operating a diesel vehicle on unaltered vegetable oil has had resurgence in recent years, due to the sharp increase in petroleum costs
758
E.P. Wagner et al. / Fuel 107 (2013) 757–765
and consumption as well as the detrimental effect on the environment. Furthermore, vegetable oil is considered to be nearly a net zero carbon emission fuel because the carbon produced by combustion, typically as carbon dioxide, is absorbed back from the atmosphere by plants [2]. The amount of process energy required to create vegetable oil is minimal, especially when compared to the requirements to produce petroleum diesel or even biodiesel [3]. However, current diesel automotive engine technology does not match well to widespread use of vegetable oil as fuel due to its relatively high viscosity at atmospheric temperatures [4]. One popular solution to the high viscosity is altering the vegetable oil through a chemical synthesis process where the lipid triglyceride (oil) molecule is broken apart at the ester functional group on each fatty acid chain connected to the triglyceride backbone. The end result of this base catalyzed transesterification reaction is the creation of four lower molecular weight molecules, glycerol and three methyl esters. The resulting methyl esters, termed Biodiesel, can then be used directly in a diesel engine or, as more typically done, used as a percent component in petroleum diesel fuel because it has a viscosity and gel point similar to petroleum diesel. An alternative approach is to use chemically unaltered vegetable oil by retrofitting the engine fuel supply system with in-line heat exchangers and a heated fuel tank to lower the oil viscosity. This allows more appropriate flow through the fuel system, fuel injector spray patterns and nebulization of the fuel [5,6]. The complicating issue is that the engine needs to be started and operated on diesel fuel until the temperature of the vegetable oil in a separate fuel tank and supply system is at least 75 °C, which requires the addition of either manual fuel switching mechanisms or an automatic fuel switching system that selects the fuel based on temperature. At this temperature the oil has viscosity similar to that of diesel fuel. Before the engine is stopped, it again needs to be operated on the diesel fuel to purge the fuel system of oil. While this approach has experienced some success [7–14], retrofitting a vehicle with a separate fuel supply system and tank is impractical and cost prohibitive when considered on a large scale. Historically, long term engine wear and coking have also been concerns when operating on 100% unused (neat) or waste vegetable oil (WVO) [15–19]. Most of the issues and concerns reported in these early research efforts are likely due to poor atomization, and insufficient oxygen/air intake levels in the engine cylinder. The diesel engine designs used in these experiments operated at much lower fuel injection pressures than systems found on diesel engines today. As injection pressure increases, nebulization becomes more uniform and the fuel droplet size is smaller [20], which enables a more complete combustion process. Finally, it is not clear in these reports if the fuel was preheated to lower viscosity before injection into the engine. A more practical approach to using unaltered vegetable oil effectively in diesel engines is to lower the viscosity by blending it with other petroleum fuels, such as diesel. There is a growing body of literature and research showing good success with this relatively simple approach. Rakopoulos et al. is one of the more prolific researchers in this area with reports dating back to 1992 indicating the possibility and efficacy of oil–diesel blends [21– 23]. More recent reports by this group continue to show that oil– diesel blended fuels have comparable power performance and emissions to that of petroleum diesel fuel [24–26]. In all these reports the engines were mounted in a laboratory connected to a dynamometer and exhaust gas analysis system. Other investigators, using similar setups, have also reported comparable results for emissions and thermal efficiency of vegetable oil–diesel blends in stationary mounted engines in laboratory settings. While some of these studies use multi-cylinder engines [27–30], many are simple one cylinder engines [31–38]. In general, the thermal efficiency
of oil–diesel blends has been found to be slightly lower, but effectively comparable to pure diesel [27,30,33]. In addition, exhaust gas and engine temperatures are lower with oil–diesel blends, which can be correlated to lower nitrogen oxides (NOx) emissions. Emission levels of unburnt hydrocarbons (HCs), and soot are somewhat mixed among research groups. Most reports do show a lower level of carbon monoxide (CO) when operating on oil–diesel blends. Ultimately, none of the exhaust emission levels are extreme compared to regulation levels. As such, oil–diesel blends can be regarded as a very reasonable alternative fuel that can reduce petroleum fuel use. Certainly, heating a diesel–oil blended fuel prior to injection into the engine to further reduce viscosity is possible and can improve engine performance compared to the same fuel that is unheated [31,32]. Since there appears to be a growing body of positive results for using oil–diesel blends on stationary engines in the laboratory setting, it brings to bear the question of how such fuels will perform in mass production diesel vehicles available to the consumer. The focus of the research efforts presented here was to create blended biofuels with viscosities low enough to allow direct use in mass produced vehicles with little or no modifications and evaluate the power and performance of each vehicle using chassis dynamometer measurements. In addition, on-road performance and fuel economy was assessed. Ultimately, our long range goal is to create practical, economical, and environmentally friendly biofuels by blending waste vegetable oil (WVO) with other petroleum fuels. While using neat vegetable oil makes it easier to control oil physical properties, the cost is much greater for us than obtaining WVO at no cost and processing it to remove particulates and water. Further, using WVO eliminates the debate and concern over using these resources for food versus fuel and provides a means to recycle a waste product.
2. Experimental methods This work consisted of first creating appropriate fuel blends using semi-empirical modeling methods to guide the blend ratios and components and then evaluating performance through chassis dynamometer testing and on-road vehicle performance. Theoretical viscosity modeling of the fuel blends were conducted to determine what blend ratios would be possible in order to lower the viscosity to a specified value while maximizing the energy and oil content. In previous work conducted by the authors, viscosity modeling methods for fuel blends containing any combination of diesel, kerosene, waste vegetable oil and gasoline were developed [39]. Using these semi-empirical methods resulted in many possible blend ratios of diesel, kerosene, oil, ethanol and gasoline that achieve a desired viscosity. Energy content of the fuel needs to be maximized as well. Therefore, our viscosity modeling methods were used, along with energy content values for each component, so that engine performance could be maximized while achieving the identified viscosity limits. By multiplying heats of combustion values for each component with its percent loading in the blend and then summing to obtain the total blend heat of combustion, we were able to model and find specific blend compositions that should produce the highest energy content and presumably the best fuel efficiency and power. The cetane rating of a compression ignition fuel is a very important factor as well and should be monitored in blend formulations. However, it is not a physical property that can be accurately modeled through a summative state function method like heats of combustion. Further, our group does not have access to facilities and equipment to perform cetane measurements as outlined in the ASTM D976 method [40]. Vegetable oils tend to have a slightly lower cetane number compared to petroleum diesel, which can potentially lower engine performance
759
E.P. Wagner et al. / Fuel 107 (2013) 757–765
and may give rise to combustion cyclic irregularity, leading to lower fuel economy and an increase in emitted pollutants. However, Rakopoulos et al. indicate that the diesel engine operates stably even with 100% vegetable oil fuel [41]. While gasoline has a relatively high heat of combustion and a very low viscosity, both of which are positive attributes for blending, the cetane value is very low. Consequently, it was decided that gasoline would not be part of our formulations even though there have been some reports in literature indicating benefits when kept below 10% in a blended diesel fuel [42,43]. Ethanol blended into diesel fuels has also been reported to have some performance and emission benefits [28,44,45]. However, it too significantly lowers the cetane value. Our own research experience with using ethanol as fuel component in our on road vehicle testing resulted in vapor lock when the fuel temperature reached 52 °C. Ethanol could likely be used effectively in colder climates, but in the interest of keeping the array of possible blend components consistent among all created blends for this study, it was not used. Kerosene, also known as #1 diesel, has a lower viscosity than diesel fuel, while still maintaining an acceptable cetane rating. Therefore, it was used as a significant blend component. Based on this information, the guidelines or parameters set for our blended fuels in terms of energy content required them to be created from WVO, diesel and kerosene while maximizing the amount of oil in the fuel. Diesel, WVO, and kerosene all have relatively high cetane values, which minimized the concern for evaluating or knowing the exact cetane value of each blended fuel. Given the seasonal climate change in the region where this work was conducted, it followed that three different blended fuels should be explored and created for this study, each of which would maximize the oil content for a given seasonal climate. Relatively higher temperatures in the summer allow for higher oil content than any other seasonal blends while keeping the fuel viscosity at a low enough value to allow a vehicle to start without pre-heating the fuel. In the autumn and spring blend, an intermediate amount of oil could be used in the fuel, and a winter blend would have the lowest oil content. Our prior research found that the highest viscosity for petroleum diesel fuel before gelling at approximately 15 °C is 15 cSt [39]. Therefore, this was the maximum viscosity value allowed for our blended fuels when starting a cold engine. ASTM standards require a grade 2 diesel fuel to be 6 cSt or lower at a normal engine operating temperature of 40 °C [46]. This was the second viscosity standard set for the blends created here. Using both the energy content and viscosity parameters in our modeling efforts lead to the creation of three blended fuels for this investigation. We have termed these fuels as ‘‘VDiesel’’ or abbreviated VD followed by the percent volume of oil in the blend. The VD15 blend has the lowest percent oil and is useable in winter climate temperatures just like petroleum diesel. The VD30 blend is useable at temperatures of 0 °C and above and the VD 40 blend is useable in summertime climates with temperatures above 15 °C. Two addition trial blends were created for comparative purposes. The VD30-2 was created from a simple binary mix of oils and diesel as a direct comparison to the VD30. Based on our fuel blend
modeling efforts, it was determined that the maximum oil content possible was 50%. Consequently, the VD50 was created to evaluate engine performance at this oil content limit. Table 1 details the blend compositions as well as the lower temperature limit for each of these fuels. Fig. 1 shows the temperature dependent viscosity curve for each of these fuels. Interestingly, the entropic effects of mixing dissimilar molecules lead to a lower gel point for the VD15 blend compared to diesel fuel. Consequently, gelling of the fuels at the lower useable temperature limits was not an issue. Diesel and kerosene were obtained from a commercial fueling station. The waste soybean oil (WSO) was collected from a local restaurant that used a consistent supplier. The oil was stored undisturbed for 1–2 weeks to settle out large particulate material. The oil was then decanted and heated to approximately 80 °C for a time period long enough to drive off water and lower the water content to no more than 300 ppm as measured by the SandyBrea method [47]. The oil was then rough filtered down to 100 lm and subsequently centrifuged down to less than 3 lm using a model OC-20 Dieselcraft oil centrifuge. The WSO free fatty acid content, determined through base titration, ranged from 3% to 6%. The processed oil was volumetrically mixed with the other fuel components to create the VDiesel fuel blends. Evaluations of the VDiesel fuels were conducted through chassis dynamometer testing as well as on-road performance and fuel economy. Three test vehicles were used in this study, a 2006 Jeep Liberty CRD, 1999 Mercedes E300 and a 1984 Mercedes 300TD. All of these vehicles have turbochargers and the pertinent engine specifications are shown in Table 2. Each vehicle was retrofitted with a heat exchanger in the fuel line. The purpose was to control the fuel temperature so that the viscosity would stay within a 4– 6 cSt range during performance testing and for regular driving conditions and daily use. A Kaori K030 brazed plate heat exchanger and temperature probe were installed into the fuel line in the engine bay of each vehicle. Engine coolant was used to heat the oil in the heat exchanger and a Love series TS programmable temperature controller was used to control the fuel temperature through a flow valve on the coolant line. The temperature set point for all experiments was 45 °C and would fluctuate between 40 and 52 °C due to the delay between when the valve was actuated and the measured fuel temperature change. While this may be considered a wide temperature range, it ultimately resulted in only minor viscosity changes and kept the fuel at an appropriately low viscosity during testing. While our eventual goal is to avoid any retrofitting to a vehicle, the cost of this retrofitting was approximately $200 for each vehicle and took a few hours to complete. Compared to installing an entirely separate and second fuel system, as discussed earlier, our retrofitting was quite minimal and guaranteed that our fuel temperature and viscosity would be consistently in an acceptable range. Dynamometer testing was conducted at an automotive performance shop on a DynoJet 224 inertia-type chassis dynamometer. This dynamometer measures to an accuracy of ±0.10 RPM and ±0.010 MPH [48]. In this type of chassis testing, the vehicle was driven onto a rolling resistance drum. An optical pickup to monitor
Table 1 Composition of VDiesel biofuel blends created for study. Each fuel is labeled VD (for VDiesel) followed by the percent oil content. Fuel
VD15 VD30 VD40 VD30-2 VD50
Composition v/v (%) Waste oil
Diesel
Kerosene
Comments
15 30 40 30 50
40 35 28 70 0
45 35 32 0 50
For cold climates (> 20 °C) For temperate climates (>0 °C) For warm climates (>15 °C) Alternate blend of VD30 (for research purposes) Maximized WSO content (for research purposes)
760
E.P. Wagner et al. / Fuel 107 (2013) 757–765
Fig. 1. Predicted temperature dependent viscosity curves for WVO fuel blends. Target operating temperature is achieved with an in-line heat exchanger with the capability of providing temperatures 40° above ambient conditions.
Table 2 Manufacturer’s specification for vehicles and engines and the parameters used for fuel performance analysis. All vehicles had automatic transmissions and turbochargers.
Engine displacement Number of cylinders Compression ratio Bore Stroke Max horsepower Max torque Red line Fuel injection system Injector operation Fuel injection pressure Approximate vehicle mileage at time of testing Gear used for testing Speed range for testing Engine speed range for testing
2006 Jeep
1999 Mercedes
1984 Mercedes
2.8 L 4 17.5:1 94.0 mm 100 mm 160 hp at 3800 rpm 295 lb-ft at 1800 rpm 5000 rpm Common rail direct injection Piezoelectric 1600 bar 42,000 4rd 50–90 mph 2300–4000 rpm
3.0 L 6 22:1 87.0 mm 84.0 mm 174 hp at 4400 rpm 243 lb-ft at 1600 rpm 5000 rpm Simple pre-chamber injection Mechanical (pop valve) 135–145 bar 68,000 3rd 40–75 mph 2750–4950 rpm
3.0 L 5 21.5:1 90.9 mm 92.4 mm 123 hp at 4350 rpm 181 lb-ft at 2400 rpm 4200 rpm Simple pre-chamber injection Mechanical (pop valve) 135–145 bar 180,000 3rd 35–60 mph 2350–3850 rpm
the engine RPM was used so that both horsepower (ft-lbs/s) and torque (lb-ft) could be measured and calculated. Using an inertia type dynamometer requires operating a vehicle in one gear through the power band range. The gear chosen was determined by which one could provide the broadest RPM range and peak power and torque performance (Table 2). Through preliminary testing, the RPM/MPH range was selected on this basis and on the ability to obtain reliable and consistent performance data. A minimum of three dynamometer runs were conducted for each fuel on each vehicle. During a dynamometer test, data was collected every 0.5 MPH. Power and torque values for the 2006 Jeep liberty were collected from 50 to 90 MPH, resulting in a total of 80 data points for each dynamometer run. The same data at the same MPH values were also collected for the diesel fuel. The three runs for each vehicle operating on a blend were averaged and then compared to the average power of the three runs for diesel fuel through percent differences. In addition, statistical comparisons were conducted using paired t-Test with the significance value (P) set at 0.05. This same data collection and analysis procedure was carried out for the other two vehicles with the exception of
the RPM/MPH range. Data from 40 to 70 MPH (2750–4950 RPM) was collected for the 1999 Mercedes and from 35 to 60 MPH (2350–3850 RPM) for the 1984 Mercedes. We were unable to collect VD30-2 data from 50 to 55 MPH on the Jeep and 55 to 60 MPH on the 1984 Mercedes. In these two cases, the corresponding petroleum diesel data was also eliminated in order to conduct appropriate comparative analysis. 3. Results and discussion Figs. 2–4 show the average power and torque performances for VD15, VD30, and VD40 fuels. The VD30-2 and VD50 data were not included on these graphs for clarity purposes. Each of these graphs provides a good overall representation of the power band for each vehicle. Due to scaling and small differences, the performance of each fuel compared to diesel is difficult to observe. Therefore, Figs. 5–7 show the percent power difference of each VDiesel fuel with respect to petroleum diesel. The power performance for all fuel blends, except for the VD50 was very similar to petroleum diesel. Table 3 shows the average engine performance on the blended
E.P. Wagner et al. / Fuel 107 (2013) 757–765
761
Fig. 2. Typical power band for 2006 Jeep Liberty CRD.
Fig. 3. Typical power band for 1999 Mercedes E300.
fuels compared to pure petroleum diesel fuel. The differences in performance were relatively consistent over the power band range for each vehicle. In the case of the Jeep, a noticeably larger change in performance occurred around 2500 RPM. This is where the slope of the power performance is greatest and reaches a maximum. Since the maximum performance for the diesel and blended fuel are at slightly different RPMs, it results in the appearance of a slightly larger difference over this small RPM range compared to
other RPM ranges shown in Fig. 5. All three vehicle performances were statistically lower on the VD15 compared to diesel, with an overall average decrease of 1.0%. The same was true for the VD30, VD40 and VD50 blends with overall average decreases in performance of 1.1%, 1.6% and 5.4% respectively. The results for the VD30-2 were mixed among the vehicles. The 1999 Mercedes on VD30-2 fuel performance was the same as diesel (P = 0.45), the 2006 Jeep performance was significantly better (P < 0.05) and
762
E.P. Wagner et al. / Fuel 107 (2013) 757–765
Fig. 4. Typical power band for 1984 Mercedes 300TD. Data collection stopped short of finding exact horsepower maximum due to 4000 RPM maximum on engine.
Fig. 5. Percent differences from petroleum diesel in horsepower output with respect to speed and RPM for the 2006 Jeep Liberty CRD. Data for VD30-2 was not collected in the 50–55 MPH range.
the 1984 Mercedes performance was significantly lower by 3.2% (P < 0.05). Both the Jeep and 1999 Mercedes performed best on the VD30-2, which does not contain kerosene. However, the 1984 Mercedes performed the best, and almost identically, on the VD15, VD30 and VD40 blends, all of which contained kerosene. Generally speaking, the performance of the Jeep and 1999 Mercedes paralleled each other and both vehicles have a electronic control unit (ECU) for the engine. The 1984 Mercedes is controlled through
older mechanical and pneumatic systems, which may account for why the VD30-2 performance of this vehicle did not mimic that of the other two. The greatest decline in performance was between the VD40 and VD50 blends for all three vehicles indicating that VDiesel blended fuels should contain no more than 40% oil. Each vehicle has been driven over 10,000 miles on various VDiesel blends. From the driver’s perspective, no perceivable driving or power performance difference was found for the VD15, VD30,
E.P. Wagner et al. / Fuel 107 (2013) 757–765
763
Fig. 6. Percent differences from petroleum diesel in horsepower output with respect to speed and RPM for the 1999 Mercedes E300.
Fig. 7. Percent differences from petroleum diesel in horsepower output with respect to speed and RPM for the 1984 Mercedes 300TD. Data for VD30-2 was not collected in the 55–60 MPH range.
VD30-2 and VD40. Only in the case of the VD50 was there a slightly detectable difference in full acceleration situations. Ultimately, the difference was relatively minor and inconsequential in typical driving conditions. There have been no engine or fuel issues or problems while operating on the VDiesel blended fuels. Preliminary on-road fuel economy data for the VD30 blend was collected for the Jeep and 1999 Mercedes vehicles through six complete fuel
tank fills and use. For the Jeep, fuel economy for diesel was 17.8 miles per gallon (MPG) and 19.0 for VD30, an increase of 6.6%. However, fuel economy for the 1999 Mercedes dropped 3.1%, from 23.1 MPG to 22.4 MPG. The exact driving conditions and routes during each test varied slightly, which likely affect the precision and accuracy of the fuel economy data. While more formal fuel economy testing in controlled conditions is necessary, such as on
764
E.P. Wagner et al. / Fuel 107 (2013) 757–765
Table 3 Overall average engine and vehicle performance on biofuel blended fuels compared to pure petroleum diesel fuel. The 1999 Mecedes on VD30-2 fuel performance was the same as diesel (P = 0.45) and the 2006 Jeep performance was significantly better (P < 0.05) on the VD30-2 blend compared to petroleum diesel. All other data indicated a statistically significant decrease in performance (P < 0.05), albeit of minimal practical importance in all cases except for the VD50. Fuel
2006 Jeep (%)
VD15 VD30 VD30-2 VD40 VD50
0.9 ± 0.5 1.5 ± 0.4 +1.1 ± 0.9 1.4 ± 0.5 5.0 ± 1.0
1999 Mercedes (%) 1.1 ± 0.6 0.6 ± 1.1 0.0 ± 1.7 1.4 ± 1.3 6.4 ± 1.1
a dynamometer, these preliminary results do support the efficacy of using blended VDiesel fuels in on-road passenger diesel vehicles. 4. Conclusions The purpose of these experimental studies was to create optimized blended fuels from waste soybean oil, diesel and kerosene using viscosity and energy content models and to evaluate the performance of the fuels in three mass produced diesel passenger vehicles. A total of five different blended fuels with the oil component ranging from 15% to 50% were created. To maintain appropriate viscosity for the fuels, a heat exchanger was installed in the fuel line of each car. The power and torque performance of each vehicle was evaluated using an inertia-type chassis dynamometer. Analysis of the data indicated that performance on the blended fuels was, on average, 1.1% less than diesel for fuels with 15–40% oil content. The minimal decrease in performance for these blends is of little practical significance. More significant decreases in performance, from 4.7% to 6.4%, were recorded for the 50% oil blend. The Jeep and 1999 Mercedes performed best on the VD30-2, which was a simple binary blend of oil and diesel. When comparing this to the VD30, which contains 35% kerosene, it shows how performance is diminished when kerosene is added to the blend. In our opinion, keeping the fuel viscosity as low as possible by adding kerosene is more important than the small (1%) increase in power by eliminating it in the blend. The results found here support findings in other reports and show good correlation between on-road passenger diesel vehicle performance and engine performance in a laboratory setting. In addition, our results provide good supportive evidence for using blended diesel fuels for on road vehicles. There are many other aspects of blended fuels outside the scope of this study that must be investigated and evaluated before these types of fuels become commercially viable. A complete emissions evaluation should be performed. Evaluation of NOx, CO, HC and soot levels have been well studied by other investigators on similar types of blended fuels and the same types of analyses should be conducted for the fuels blends in this work. Chemical analysis of the particulate emissions should also be investigated. In addition, Fuel oxidation, shelf life, corrosive properties and engine wear analysis are all issues that should also be addressed. The importance of vegetable oils as fuels may be strengthened even more if the amount of petroleum products in the blend could be reduced or completely eliminated. More recent research using n-butanol or diethyl ether biofuels as viscosity modifiers for the vegetable oil, instead of diesel and kerosene, has shown good performance and emission results [49]. The efforts and results of this study provide good evidence and support for future research in these suggested areas and continue develop of blended biofuels for mass produced diesel vehicles. References [1] Shrinivasa U. The evolution of diesel engines. Resonance 2012;17(4):365–77. [2] Hossain AK, Davies PA. Plant oils as fuels for compression ignition engines: a technical review and life-cycle analysis. Renew Energy 2010. 351–13.
1984 Mercedes (%) 1.1 ± 0.5 1.3 ± 0.8 3.2 ± 1.0 1.9 ± 0.5 4.7 ± 0.8
Average (%) 1.0 1.1 0.70 1.6 5.4
[3] Esteban B, Baquero G, Puig R, Riba J-R, Rius A. Is it environmentally advantageous to use vegetable oil directly as biofuel instead of converting it to biodiesel? Biomass Bioenergy 2011;35(3):1317–28. [4] Noureddini H, Teoh BC, Clements D. Viscosities of vegetable oils and fatty acids. J Am Oil Chem Soc 2011;69:1189–91. [5] Lefebvre AH. Atomization and sprays. New York: Hemisphere Publishing Corporation; 1989. [6] Rakopoulos CD, Rakopoulos DC, Antonopoulos KA. Multi-zone modeling of diesel engine fuel spray development with vegetable oil, bio-diesel or diesel fuels. Energy Convers Manage 2006;47(11):1550–73. [7] U.S. Department of Energy. Straight vegetable oil as a diesel fuel?. [retrieved 18.08.12]. [8] Barlow K. Undergrad researcher slips past high gas prices. Univ Times 2006;39(2):3. [9] Journey to Forever. Straight vegetable oil as diesel fuel. [retrieved 03.09.12]. [10] Altin R, Cetinkaya S, Yucesu HS. The potential of using vegetable oil fuels as fuel for diesel engines. Energy Convers Manage 2001;42(5):529–38. [11] Bialkowski MT, Pekdemir T, Towers DP, Reuben R, Brautsch M, Elsbett. Effect of fuel temperature and ambient pressure on a common rail rapeseed oil spray. Internal Combust Eng 2004;11(1–2):53–65. [12] Agarwal D, Agarwal AK. Performance and emissions characteristics of Jatropha oil (preheated and blends) in a direct injection compression ignition engine. Appl Therm Eng 2007;27(13):2314–23. [13] de Almeida Silvio CA, Belchior Carlos Rodrigues, Nascimento Marcos VG, Vieira Leonardo dos SR, Guilherme Fleury. Performance of a diesel generator fuelled with palm oil. Fuel 2002;81(16):2097–102. [14] Karaosmanoglu F, Kurt G, Ozaktas T. Long term CI engine test of sunflower oil. Renew Energy 2000;19:219–21. [15] Peterson CL, Auld DL, Korus RA. Winter rape oil fuel for diesel engines: recovery and utilization. JAOCS 1983;60(8):1579–15876. [16] Barsic NJ, Humke AL. Performance and emissions characteristics of a naturally aspirated diesel engine with vegetable oil fuels. SAE 1982:810262. [17] Soltic P, Edenhauser D, Thurnheer T, Schreiber D, Sankowski A. Experimental investigation of mineral diesel fuel, GTL fuel, RME and neat soybean and rapeseed oil combustion in a heavy duty on-road engine with exhaust gas after treatment. Fuel 2009;88(1):1–8. [18] Korus RA, Jo J. A rapid engine test to measure injector fouling in diesel engines using vegetable oil fuels. J Am Oil Chem Soc 1985;62(11):1563–4. [19] Altin R, Cetinkaya S, Yucesu HS. The potential use of vegetable oil fuels as fuel for diesel engines. Energy Convers Manage 2001;41:529–38. [20] Bialkowski MT, Pekdemir T, Towers DP, Reuben R, Brautsch M, Elsbett G. Effects of fuel temperature and ambient pressure on a common rail rapeseed oil spray. J KONES Internal Combust Engine 2004;11(1–2):53–65. [21] Rakopoulos CD. Olive oil as a fuel supplement in DI and IDI diesel engines. Energy 1992;17:787–90. [22] Rakopoulos CD, Antonopoulos KA, Rakopoulos DC, Hountalas DT, Giakoumis EG. Comparative performance and emissions study of a direct injection diesel engine using blends of diesel fuel with vegetable oils or bio-diesels of various origins. Energy Convers Manage 1992;47:3272–87. [23] Rakopoulos CD. Comparative performance and emission studies when using olive oil as a fuel supplement in DI and IDI diesel engines. Renew Energy 1992;2:327–31. [24] Rakopoulos CD, Rakopoulos DC, Hountalas DT, Giakoumis EG, Andritsakis EC. Performance and emissions of bus engine using blends of diesel fuel with biodiesel of sunflower or cottonseed oils derived from Greek feedstock. Fuel 2008;87(2):147–57. [25] Rakopoulos DC, Rakopoulos CD, Giakoumis EG, Dimaratos AM, Founti MA. Comparative environmental behavior of bus engine operating on blends of diesel fuel with four straight vegetable oils of Greek origin: sunflower, cottonseed, corn and olive. Fuel 2011;90(11):3439–46. [26] Rakopoulos DC. Heat release analysis of combustion in heavy-duty turbocharged diesel engine operating on blends of diesel fuel with cottonseed or sunflower oils and their bio-diesel. Fuel 2012;96(6):524–34. [27] Leevijit T, Prateepchaikul G. Comparative performance and emissions of IDIturbo automobile diesel engine operated using degummed, deacidified mixed crude palm oil–diesel blends. Fuel 2011;90(4):1487–91. [28] Labeckas G, Slavinskas S. Comparative performance of direct injection diesel engine operating on ethanol, pertrol and rapeseed oil blends. Energy Convers Manage 2009;50:792–801. [29] Labecki L, Cairns A, Xia J, Megaritis A, Zhao H, Ganippa LC. Combustion and emission of rapeseed oil blends in diesel engine. Appl Energy 2012;95:139–46.
E.P. Wagner et al. / Fuel 107 (2013) 757–765 [30] Bajpai S, Sahoo PK, Das LM. Feasibility of blending Karanja vegetable oil in petro-diesel and utilization in a direct injection diesel engine. Fuel 2009;88(4):705–11. [31] Leenus Jesu Martin M, Edwin Geo V, Kingsly Jeba Singh D, Nagalingam B. A comparative analysis of different methods to improve the performance of cotton seed oil fuelled diesel engine. Fuel 2012;102:372–8. [32] Hazar H, Aydin H. Performance and emission evaluation of a CI engine fueled with preheated raw rapeseed oil (RRO)–diesel blends. Appl Energy 2010;87(3):786–90. [33] Misra RD, Murthy MS. Performance, emission and combustion evaluation of soapnut oil–diesel blends in a compression ignition engine. Fuel 2011;90(7):2514–8. [34] Agarwal AK, Rajamanoharan K. Experimental investigations of the performance and emissions of Karanja oil and its blends in a single cylinder agricultural diesel engine. Appl Energy 2009;86:106–12. [35] Nwafor OMI, Rice G. Performance of rapeseed oil blends in a diesel engine. Appl Energy 1996;54:345–54. [36] Huzayyin AS, Bawady AH, Rady MA, Dawood A. Experimental evaluation of diesel engine performance and emission using blends of jojoba oil and diesel fuel. Energy Convers Manage 2004;45:2093–112. [37] Altun S, Bulut H, Oner C. The comparison of engine performance and exhaust emission characteristics of sesame oil–diesel fuel mixture with diesel fuel in a direct injection diesel engine. Renew Energy 2008;33(8):1791–5. [38] Forson FK, Oduro EK, Hammond-Donkoh E. Performance of jatropha oil blends in a diesel engine. Renew Energy 2004;29:1135–45. [39] Wagner EP, Koehle MA, Moyle TM, Lambert PD. Predicting temperature dependent viscosity for unaltered waste soybean oil blended with petroleum fuels. J Am Oil Chem Soc 2010;87(4):453–9.
765
[40] ASTM D976. Standard test method for calculated cetane index of distillate fuels. West Conshohocken (PA): American Society for Testing and Materials; 2011. [41] Rakopoulos CD, Rakopoulos DC, Giakoumis EG, Dimaratos AM. Investigation of the combustion of neat cottonseed oil or its neat bio-diesel in a HSDI diesel engine by experimental heat release and statistical analyses. Fuel 2010;89:3814–26. [42] Sahin Z. Experimental and theoretical investigation of the effects of gasoline blends on single-cylinder diesel engine performance and exhaust emissions. Energy Fuel 2008;22(5):3201–12. [43] Cha J, Kwon S, Park S. An experimental and modeling study of the combustion and emission characteristics for gasoline-diesel dual-fuel engines. Proc Inst Mech Eng Part D: J Automob Eng 2011;225(6):801–12. [44] Asfar KR, Hamed H. Combustion of fuel blends. Energy Conserv Manage 1998;39(10):1081–93. [45] Bilgin A, Durgun, Sahin Z. The effects of diesel–ethanol blends on diesel engine performance. Energy Sources 2002;24:431–40. [46] ASTM D975-11b. Standard specification for diesel fuel oils. West Conshohocken, PA: American Society for Testing and Materials; 2011. [47] Sandy Brae Laboratories, Inc. Water test kits. [retrieved 03.09.12]. [48] Dynojet Research. Model 224x_dynamometer. [retrieved 03.09.12]. [49] Rakopoulos DC. Combustion and emissions of cottonseed oil and its bio-diesel in blends with either n-butanol or diethyl ether in HSDI diesel engine. Fuel 2012;105:603–13.
Contents lists available at SciVerse ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Diesel vehicle performance on unaltered waste soybean oil blended with petroleum fuels Eugene P. Wagner ⇑, Patrick D. Lambert, Todd M. Moyle, Maura A. Koehle University of Pittsburgh, Department of Chemistry, Chevron Science Center, Pittsburgh, PA 15260, USA
h i g h l i g h t s " We examine engine performance on unaltered waste soybean oil-petroleum fuel blends. " Fuels containing 15%, 30%, 40% and 50% waste soybean oil were created for analysis. " Power and torque were analyzed on three vehicles on a chassis dynamometer. " Performance for 15%, 30% and 40% WVO blends averaged 1.1% less than pure diesel. " Performance for the 50% WVO blend was 5.4% lower than pure diesel.
a r t i c l e
i n f o
Article history: Received 4 October 2012 Received in revised form 23 January 2013 Accepted 27 January 2013 Available online 13 February 2013 Keywords: Soybean oil Petroleum diesel Engine Horsepower Torque
a b s t r a c t Interest in using unaltered vegetable oil as a fuel in diesel engines has experienced an increase due to uncertainty in the crude oil market supply and the detrimental effects petroleum fuels have on the environment. Unaltered vegetable oil blended with petroleum fuels is less expensive, uses less energy to produce and is more environmentally friendly compared to petroleum diesel or biodiesel. Here we investigate the engine performance of unaltered waste soybean oil blended with petroleum diesel and kerosene for three vehicles. Five biofuel blends ranging from 15% to 50% oil by volume were tested on a 2006 Jeep Liberty CRD, a 1999 Mercedes E300 and a 1984 Mercedes 300TD. A DynoJet 224x chassis dynamometer was used to test vehicle engine performance for horsepower and torque through a range of RPMs. Results for the Jeep showed a modest decrease in horsepower and torque compared to petroleum diesel ranging from 0.9% for the 15% oil blend to 5.0% lower for the 50% oil blend. However, a 30% oil blend showed statistically better performance (P < 0.05) compared to petroleum diesel. For the 1999 Mercedes, horsepower performance was 1.1% lower for the 15% oil blend to 6.4% lower for the 50% oil blend. Engine performance for a 30% blend was statistically the same (P < 0.05) compare to diesel. Finally, horsepower performance was 1.1% lower for the 15% oil blend to 4.7% lower for the 50% oil blend for the 1984 Mercedes. Overall, the performance on these oil blended fuels was excellent and, on average 1.1% lower than petroleum diesel for blends containing 40% or lower waste soybean oil content. The more significant decrease in power between the 40% and 50% oil blends indicates that oil content in these blended fuels should be no more than 40%. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Since the year 2000, gasoline and oil prices have tripled while use and demand has also grown worldwide. Countries with emerging economies, such as China and India, are increasing consumption of gas and oil at record rates. In addition to a potential shortage of fuel to meet the growing demand, there is great concern that the use of fossil fuels for energy is a leading cause of a global warming trend which could lead to undesirable climate ⇑ Corresponding author. Tel.: +1 412 624 2861; fax: +1 412 648 3297. E-mail address: [email protected] (E.P. Wagner). 0016-2361/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2013.01.052
change. Economical, environmentally friendly, and sustainable energy solutions are needed to slow the growth and use of fossil fuels. It is well known that vegetable oil can be used as an energy source in diesel engines. In fact, the first diesel engine developed by Rudolph Diesel was showcased at the 1900 World Fair where he demonstrated the use of peanut oil as a fuel for his engine [1]. It was not until Rudolph’s death that his original engine design was converted to operate on petroleum diesel due to the increasing availability of petroleum products. The concept and practice of operating a diesel vehicle on unaltered vegetable oil has had resurgence in recent years, due to the sharp increase in petroleum costs
758
E.P. Wagner et al. / Fuel 107 (2013) 757–765
and consumption as well as the detrimental effect on the environment. Furthermore, vegetable oil is considered to be nearly a net zero carbon emission fuel because the carbon produced by combustion, typically as carbon dioxide, is absorbed back from the atmosphere by plants [2]. The amount of process energy required to create vegetable oil is minimal, especially when compared to the requirements to produce petroleum diesel or even biodiesel [3]. However, current diesel automotive engine technology does not match well to widespread use of vegetable oil as fuel due to its relatively high viscosity at atmospheric temperatures [4]. One popular solution to the high viscosity is altering the vegetable oil through a chemical synthesis process where the lipid triglyceride (oil) molecule is broken apart at the ester functional group on each fatty acid chain connected to the triglyceride backbone. The end result of this base catalyzed transesterification reaction is the creation of four lower molecular weight molecules, glycerol and three methyl esters. The resulting methyl esters, termed Biodiesel, can then be used directly in a diesel engine or, as more typically done, used as a percent component in petroleum diesel fuel because it has a viscosity and gel point similar to petroleum diesel. An alternative approach is to use chemically unaltered vegetable oil by retrofitting the engine fuel supply system with in-line heat exchangers and a heated fuel tank to lower the oil viscosity. This allows more appropriate flow through the fuel system, fuel injector spray patterns and nebulization of the fuel [5,6]. The complicating issue is that the engine needs to be started and operated on diesel fuel until the temperature of the vegetable oil in a separate fuel tank and supply system is at least 75 °C, which requires the addition of either manual fuel switching mechanisms or an automatic fuel switching system that selects the fuel based on temperature. At this temperature the oil has viscosity similar to that of diesel fuel. Before the engine is stopped, it again needs to be operated on the diesel fuel to purge the fuel system of oil. While this approach has experienced some success [7–14], retrofitting a vehicle with a separate fuel supply system and tank is impractical and cost prohibitive when considered on a large scale. Historically, long term engine wear and coking have also been concerns when operating on 100% unused (neat) or waste vegetable oil (WVO) [15–19]. Most of the issues and concerns reported in these early research efforts are likely due to poor atomization, and insufficient oxygen/air intake levels in the engine cylinder. The diesel engine designs used in these experiments operated at much lower fuel injection pressures than systems found on diesel engines today. As injection pressure increases, nebulization becomes more uniform and the fuel droplet size is smaller [20], which enables a more complete combustion process. Finally, it is not clear in these reports if the fuel was preheated to lower viscosity before injection into the engine. A more practical approach to using unaltered vegetable oil effectively in diesel engines is to lower the viscosity by blending it with other petroleum fuels, such as diesel. There is a growing body of literature and research showing good success with this relatively simple approach. Rakopoulos et al. is one of the more prolific researchers in this area with reports dating back to 1992 indicating the possibility and efficacy of oil–diesel blends [21– 23]. More recent reports by this group continue to show that oil– diesel blended fuels have comparable power performance and emissions to that of petroleum diesel fuel [24–26]. In all these reports the engines were mounted in a laboratory connected to a dynamometer and exhaust gas analysis system. Other investigators, using similar setups, have also reported comparable results for emissions and thermal efficiency of vegetable oil–diesel blends in stationary mounted engines in laboratory settings. While some of these studies use multi-cylinder engines [27–30], many are simple one cylinder engines [31–38]. In general, the thermal efficiency
of oil–diesel blends has been found to be slightly lower, but effectively comparable to pure diesel [27,30,33]. In addition, exhaust gas and engine temperatures are lower with oil–diesel blends, which can be correlated to lower nitrogen oxides (NOx) emissions. Emission levels of unburnt hydrocarbons (HCs), and soot are somewhat mixed among research groups. Most reports do show a lower level of carbon monoxide (CO) when operating on oil–diesel blends. Ultimately, none of the exhaust emission levels are extreme compared to regulation levels. As such, oil–diesel blends can be regarded as a very reasonable alternative fuel that can reduce petroleum fuel use. Certainly, heating a diesel–oil blended fuel prior to injection into the engine to further reduce viscosity is possible and can improve engine performance compared to the same fuel that is unheated [31,32]. Since there appears to be a growing body of positive results for using oil–diesel blends on stationary engines in the laboratory setting, it brings to bear the question of how such fuels will perform in mass production diesel vehicles available to the consumer. The focus of the research efforts presented here was to create blended biofuels with viscosities low enough to allow direct use in mass produced vehicles with little or no modifications and evaluate the power and performance of each vehicle using chassis dynamometer measurements. In addition, on-road performance and fuel economy was assessed. Ultimately, our long range goal is to create practical, economical, and environmentally friendly biofuels by blending waste vegetable oil (WVO) with other petroleum fuels. While using neat vegetable oil makes it easier to control oil physical properties, the cost is much greater for us than obtaining WVO at no cost and processing it to remove particulates and water. Further, using WVO eliminates the debate and concern over using these resources for food versus fuel and provides a means to recycle a waste product.
2. Experimental methods This work consisted of first creating appropriate fuel blends using semi-empirical modeling methods to guide the blend ratios and components and then evaluating performance through chassis dynamometer testing and on-road vehicle performance. Theoretical viscosity modeling of the fuel blends were conducted to determine what blend ratios would be possible in order to lower the viscosity to a specified value while maximizing the energy and oil content. In previous work conducted by the authors, viscosity modeling methods for fuel blends containing any combination of diesel, kerosene, waste vegetable oil and gasoline were developed [39]. Using these semi-empirical methods resulted in many possible blend ratios of diesel, kerosene, oil, ethanol and gasoline that achieve a desired viscosity. Energy content of the fuel needs to be maximized as well. Therefore, our viscosity modeling methods were used, along with energy content values for each component, so that engine performance could be maximized while achieving the identified viscosity limits. By multiplying heats of combustion values for each component with its percent loading in the blend and then summing to obtain the total blend heat of combustion, we were able to model and find specific blend compositions that should produce the highest energy content and presumably the best fuel efficiency and power. The cetane rating of a compression ignition fuel is a very important factor as well and should be monitored in blend formulations. However, it is not a physical property that can be accurately modeled through a summative state function method like heats of combustion. Further, our group does not have access to facilities and equipment to perform cetane measurements as outlined in the ASTM D976 method [40]. Vegetable oils tend to have a slightly lower cetane number compared to petroleum diesel, which can potentially lower engine performance
759
E.P. Wagner et al. / Fuel 107 (2013) 757–765
and may give rise to combustion cyclic irregularity, leading to lower fuel economy and an increase in emitted pollutants. However, Rakopoulos et al. indicate that the diesel engine operates stably even with 100% vegetable oil fuel [41]. While gasoline has a relatively high heat of combustion and a very low viscosity, both of which are positive attributes for blending, the cetane value is very low. Consequently, it was decided that gasoline would not be part of our formulations even though there have been some reports in literature indicating benefits when kept below 10% in a blended diesel fuel [42,43]. Ethanol blended into diesel fuels has also been reported to have some performance and emission benefits [28,44,45]. However, it too significantly lowers the cetane value. Our own research experience with using ethanol as fuel component in our on road vehicle testing resulted in vapor lock when the fuel temperature reached 52 °C. Ethanol could likely be used effectively in colder climates, but in the interest of keeping the array of possible blend components consistent among all created blends for this study, it was not used. Kerosene, also known as #1 diesel, has a lower viscosity than diesel fuel, while still maintaining an acceptable cetane rating. Therefore, it was used as a significant blend component. Based on this information, the guidelines or parameters set for our blended fuels in terms of energy content required them to be created from WVO, diesel and kerosene while maximizing the amount of oil in the fuel. Diesel, WVO, and kerosene all have relatively high cetane values, which minimized the concern for evaluating or knowing the exact cetane value of each blended fuel. Given the seasonal climate change in the region where this work was conducted, it followed that three different blended fuels should be explored and created for this study, each of which would maximize the oil content for a given seasonal climate. Relatively higher temperatures in the summer allow for higher oil content than any other seasonal blends while keeping the fuel viscosity at a low enough value to allow a vehicle to start without pre-heating the fuel. In the autumn and spring blend, an intermediate amount of oil could be used in the fuel, and a winter blend would have the lowest oil content. Our prior research found that the highest viscosity for petroleum diesel fuel before gelling at approximately 15 °C is 15 cSt [39]. Therefore, this was the maximum viscosity value allowed for our blended fuels when starting a cold engine. ASTM standards require a grade 2 diesel fuel to be 6 cSt or lower at a normal engine operating temperature of 40 °C [46]. This was the second viscosity standard set for the blends created here. Using both the energy content and viscosity parameters in our modeling efforts lead to the creation of three blended fuels for this investigation. We have termed these fuels as ‘‘VDiesel’’ or abbreviated VD followed by the percent volume of oil in the blend. The VD15 blend has the lowest percent oil and is useable in winter climate temperatures just like petroleum diesel. The VD30 blend is useable at temperatures of 0 °C and above and the VD 40 blend is useable in summertime climates with temperatures above 15 °C. Two addition trial blends were created for comparative purposes. The VD30-2 was created from a simple binary mix of oils and diesel as a direct comparison to the VD30. Based on our fuel blend
modeling efforts, it was determined that the maximum oil content possible was 50%. Consequently, the VD50 was created to evaluate engine performance at this oil content limit. Table 1 details the blend compositions as well as the lower temperature limit for each of these fuels. Fig. 1 shows the temperature dependent viscosity curve for each of these fuels. Interestingly, the entropic effects of mixing dissimilar molecules lead to a lower gel point for the VD15 blend compared to diesel fuel. Consequently, gelling of the fuels at the lower useable temperature limits was not an issue. Diesel and kerosene were obtained from a commercial fueling station. The waste soybean oil (WSO) was collected from a local restaurant that used a consistent supplier. The oil was stored undisturbed for 1–2 weeks to settle out large particulate material. The oil was then decanted and heated to approximately 80 °C for a time period long enough to drive off water and lower the water content to no more than 300 ppm as measured by the SandyBrea method [47]. The oil was then rough filtered down to 100 lm and subsequently centrifuged down to less than 3 lm using a model OC-20 Dieselcraft oil centrifuge. The WSO free fatty acid content, determined through base titration, ranged from 3% to 6%. The processed oil was volumetrically mixed with the other fuel components to create the VDiesel fuel blends. Evaluations of the VDiesel fuels were conducted through chassis dynamometer testing as well as on-road performance and fuel economy. Three test vehicles were used in this study, a 2006 Jeep Liberty CRD, 1999 Mercedes E300 and a 1984 Mercedes 300TD. All of these vehicles have turbochargers and the pertinent engine specifications are shown in Table 2. Each vehicle was retrofitted with a heat exchanger in the fuel line. The purpose was to control the fuel temperature so that the viscosity would stay within a 4– 6 cSt range during performance testing and for regular driving conditions and daily use. A Kaori K030 brazed plate heat exchanger and temperature probe were installed into the fuel line in the engine bay of each vehicle. Engine coolant was used to heat the oil in the heat exchanger and a Love series TS programmable temperature controller was used to control the fuel temperature through a flow valve on the coolant line. The temperature set point for all experiments was 45 °C and would fluctuate between 40 and 52 °C due to the delay between when the valve was actuated and the measured fuel temperature change. While this may be considered a wide temperature range, it ultimately resulted in only minor viscosity changes and kept the fuel at an appropriately low viscosity during testing. While our eventual goal is to avoid any retrofitting to a vehicle, the cost of this retrofitting was approximately $200 for each vehicle and took a few hours to complete. Compared to installing an entirely separate and second fuel system, as discussed earlier, our retrofitting was quite minimal and guaranteed that our fuel temperature and viscosity would be consistently in an acceptable range. Dynamometer testing was conducted at an automotive performance shop on a DynoJet 224 inertia-type chassis dynamometer. This dynamometer measures to an accuracy of ±0.10 RPM and ±0.010 MPH [48]. In this type of chassis testing, the vehicle was driven onto a rolling resistance drum. An optical pickup to monitor
Table 1 Composition of VDiesel biofuel blends created for study. Each fuel is labeled VD (for VDiesel) followed by the percent oil content. Fuel
VD15 VD30 VD40 VD30-2 VD50
Composition v/v (%) Waste oil
Diesel
Kerosene
Comments
15 30 40 30 50
40 35 28 70 0
45 35 32 0 50
For cold climates (> 20 °C) For temperate climates (>0 °C) For warm climates (>15 °C) Alternate blend of VD30 (for research purposes) Maximized WSO content (for research purposes)
760
E.P. Wagner et al. / Fuel 107 (2013) 757–765
Fig. 1. Predicted temperature dependent viscosity curves for WVO fuel blends. Target operating temperature is achieved with an in-line heat exchanger with the capability of providing temperatures 40° above ambient conditions.
Table 2 Manufacturer’s specification for vehicles and engines and the parameters used for fuel performance analysis. All vehicles had automatic transmissions and turbochargers.
Engine displacement Number of cylinders Compression ratio Bore Stroke Max horsepower Max torque Red line Fuel injection system Injector operation Fuel injection pressure Approximate vehicle mileage at time of testing Gear used for testing Speed range for testing Engine speed range for testing
2006 Jeep
1999 Mercedes
1984 Mercedes
2.8 L 4 17.5:1 94.0 mm 100 mm 160 hp at 3800 rpm 295 lb-ft at 1800 rpm 5000 rpm Common rail direct injection Piezoelectric 1600 bar 42,000 4rd 50–90 mph 2300–4000 rpm
3.0 L 6 22:1 87.0 mm 84.0 mm 174 hp at 4400 rpm 243 lb-ft at 1600 rpm 5000 rpm Simple pre-chamber injection Mechanical (pop valve) 135–145 bar 68,000 3rd 40–75 mph 2750–4950 rpm
3.0 L 5 21.5:1 90.9 mm 92.4 mm 123 hp at 4350 rpm 181 lb-ft at 2400 rpm 4200 rpm Simple pre-chamber injection Mechanical (pop valve) 135–145 bar 180,000 3rd 35–60 mph 2350–3850 rpm
the engine RPM was used so that both horsepower (ft-lbs/s) and torque (lb-ft) could be measured and calculated. Using an inertia type dynamometer requires operating a vehicle in one gear through the power band range. The gear chosen was determined by which one could provide the broadest RPM range and peak power and torque performance (Table 2). Through preliminary testing, the RPM/MPH range was selected on this basis and on the ability to obtain reliable and consistent performance data. A minimum of three dynamometer runs were conducted for each fuel on each vehicle. During a dynamometer test, data was collected every 0.5 MPH. Power and torque values for the 2006 Jeep liberty were collected from 50 to 90 MPH, resulting in a total of 80 data points for each dynamometer run. The same data at the same MPH values were also collected for the diesel fuel. The three runs for each vehicle operating on a blend were averaged and then compared to the average power of the three runs for diesel fuel through percent differences. In addition, statistical comparisons were conducted using paired t-Test with the significance value (P) set at 0.05. This same data collection and analysis procedure was carried out for the other two vehicles with the exception of
the RPM/MPH range. Data from 40 to 70 MPH (2750–4950 RPM) was collected for the 1999 Mercedes and from 35 to 60 MPH (2350–3850 RPM) for the 1984 Mercedes. We were unable to collect VD30-2 data from 50 to 55 MPH on the Jeep and 55 to 60 MPH on the 1984 Mercedes. In these two cases, the corresponding petroleum diesel data was also eliminated in order to conduct appropriate comparative analysis. 3. Results and discussion Figs. 2–4 show the average power and torque performances for VD15, VD30, and VD40 fuels. The VD30-2 and VD50 data were not included on these graphs for clarity purposes. Each of these graphs provides a good overall representation of the power band for each vehicle. Due to scaling and small differences, the performance of each fuel compared to diesel is difficult to observe. Therefore, Figs. 5–7 show the percent power difference of each VDiesel fuel with respect to petroleum diesel. The power performance for all fuel blends, except for the VD50 was very similar to petroleum diesel. Table 3 shows the average engine performance on the blended
E.P. Wagner et al. / Fuel 107 (2013) 757–765
761
Fig. 2. Typical power band for 2006 Jeep Liberty CRD.
Fig. 3. Typical power band for 1999 Mercedes E300.
fuels compared to pure petroleum diesel fuel. The differences in performance were relatively consistent over the power band range for each vehicle. In the case of the Jeep, a noticeably larger change in performance occurred around 2500 RPM. This is where the slope of the power performance is greatest and reaches a maximum. Since the maximum performance for the diesel and blended fuel are at slightly different RPMs, it results in the appearance of a slightly larger difference over this small RPM range compared to
other RPM ranges shown in Fig. 5. All three vehicle performances were statistically lower on the VD15 compared to diesel, with an overall average decrease of 1.0%. The same was true for the VD30, VD40 and VD50 blends with overall average decreases in performance of 1.1%, 1.6% and 5.4% respectively. The results for the VD30-2 were mixed among the vehicles. The 1999 Mercedes on VD30-2 fuel performance was the same as diesel (P = 0.45), the 2006 Jeep performance was significantly better (P < 0.05) and
762
E.P. Wagner et al. / Fuel 107 (2013) 757–765
Fig. 4. Typical power band for 1984 Mercedes 300TD. Data collection stopped short of finding exact horsepower maximum due to 4000 RPM maximum on engine.
Fig. 5. Percent differences from petroleum diesel in horsepower output with respect to speed and RPM for the 2006 Jeep Liberty CRD. Data for VD30-2 was not collected in the 50–55 MPH range.
the 1984 Mercedes performance was significantly lower by 3.2% (P < 0.05). Both the Jeep and 1999 Mercedes performed best on the VD30-2, which does not contain kerosene. However, the 1984 Mercedes performed the best, and almost identically, on the VD15, VD30 and VD40 blends, all of which contained kerosene. Generally speaking, the performance of the Jeep and 1999 Mercedes paralleled each other and both vehicles have a electronic control unit (ECU) for the engine. The 1984 Mercedes is controlled through
older mechanical and pneumatic systems, which may account for why the VD30-2 performance of this vehicle did not mimic that of the other two. The greatest decline in performance was between the VD40 and VD50 blends for all three vehicles indicating that VDiesel blended fuels should contain no more than 40% oil. Each vehicle has been driven over 10,000 miles on various VDiesel blends. From the driver’s perspective, no perceivable driving or power performance difference was found for the VD15, VD30,
E.P. Wagner et al. / Fuel 107 (2013) 757–765
763
Fig. 6. Percent differences from petroleum diesel in horsepower output with respect to speed and RPM for the 1999 Mercedes E300.
Fig. 7. Percent differences from petroleum diesel in horsepower output with respect to speed and RPM for the 1984 Mercedes 300TD. Data for VD30-2 was not collected in the 55–60 MPH range.
VD30-2 and VD40. Only in the case of the VD50 was there a slightly detectable difference in full acceleration situations. Ultimately, the difference was relatively minor and inconsequential in typical driving conditions. There have been no engine or fuel issues or problems while operating on the VDiesel blended fuels. Preliminary on-road fuel economy data for the VD30 blend was collected for the Jeep and 1999 Mercedes vehicles through six complete fuel
tank fills and use. For the Jeep, fuel economy for diesel was 17.8 miles per gallon (MPG) and 19.0 for VD30, an increase of 6.6%. However, fuel economy for the 1999 Mercedes dropped 3.1%, from 23.1 MPG to 22.4 MPG. The exact driving conditions and routes during each test varied slightly, which likely affect the precision and accuracy of the fuel economy data. While more formal fuel economy testing in controlled conditions is necessary, such as on
764
E.P. Wagner et al. / Fuel 107 (2013) 757–765
Table 3 Overall average engine and vehicle performance on biofuel blended fuels compared to pure petroleum diesel fuel. The 1999 Mecedes on VD30-2 fuel performance was the same as diesel (P = 0.45) and the 2006 Jeep performance was significantly better (P < 0.05) on the VD30-2 blend compared to petroleum diesel. All other data indicated a statistically significant decrease in performance (P < 0.05), albeit of minimal practical importance in all cases except for the VD50. Fuel
2006 Jeep (%)
VD15 VD30 VD30-2 VD40 VD50
0.9 ± 0.5 1.5 ± 0.4 +1.1 ± 0.9 1.4 ± 0.5 5.0 ± 1.0
1999 Mercedes (%) 1.1 ± 0.6 0.6 ± 1.1 0.0 ± 1.7 1.4 ± 1.3 6.4 ± 1.1
a dynamometer, these preliminary results do support the efficacy of using blended VDiesel fuels in on-road passenger diesel vehicles. 4. Conclusions The purpose of these experimental studies was to create optimized blended fuels from waste soybean oil, diesel and kerosene using viscosity and energy content models and to evaluate the performance of the fuels in three mass produced diesel passenger vehicles. A total of five different blended fuels with the oil component ranging from 15% to 50% were created. To maintain appropriate viscosity for the fuels, a heat exchanger was installed in the fuel line of each car. The power and torque performance of each vehicle was evaluated using an inertia-type chassis dynamometer. Analysis of the data indicated that performance on the blended fuels was, on average, 1.1% less than diesel for fuels with 15–40% oil content. The minimal decrease in performance for these blends is of little practical significance. More significant decreases in performance, from 4.7% to 6.4%, were recorded for the 50% oil blend. The Jeep and 1999 Mercedes performed best on the VD30-2, which was a simple binary blend of oil and diesel. When comparing this to the VD30, which contains 35% kerosene, it shows how performance is diminished when kerosene is added to the blend. In our opinion, keeping the fuel viscosity as low as possible by adding kerosene is more important than the small (1%) increase in power by eliminating it in the blend. The results found here support findings in other reports and show good correlation between on-road passenger diesel vehicle performance and engine performance in a laboratory setting. In addition, our results provide good supportive evidence for using blended diesel fuels for on road vehicles. There are many other aspects of blended fuels outside the scope of this study that must be investigated and evaluated before these types of fuels become commercially viable. A complete emissions evaluation should be performed. Evaluation of NOx, CO, HC and soot levels have been well studied by other investigators on similar types of blended fuels and the same types of analyses should be conducted for the fuels blends in this work. Chemical analysis of the particulate emissions should also be investigated. In addition, Fuel oxidation, shelf life, corrosive properties and engine wear analysis are all issues that should also be addressed. The importance of vegetable oils as fuels may be strengthened even more if the amount of petroleum products in the blend could be reduced or completely eliminated. More recent research using n-butanol or diethyl ether biofuels as viscosity modifiers for the vegetable oil, instead of diesel and kerosene, has shown good performance and emission results [49]. The efforts and results of this study provide good evidence and support for future research in these suggested areas and continue develop of blended biofuels for mass produced diesel vehicles. References [1] Shrinivasa U. The evolution of diesel engines. Resonance 2012;17(4):365–77. [2] Hossain AK, Davies PA. Plant oils as fuels for compression ignition engines: a technical review and life-cycle analysis. Renew Energy 2010. 351–13.
1984 Mercedes (%) 1.1 ± 0.5 1.3 ± 0.8 3.2 ± 1.0 1.9 ± 0.5 4.7 ± 0.8
Average (%) 1.0 1.1 0.70 1.6 5.4
[3] Esteban B, Baquero G, Puig R, Riba J-R, Rius A. Is it environmentally advantageous to use vegetable oil directly as biofuel instead of converting it to biodiesel? Biomass Bioenergy 2011;35(3):1317–28. [4] Noureddini H, Teoh BC, Clements D. Viscosities of vegetable oils and fatty acids. J Am Oil Chem Soc 2011;69:1189–91. [5] Lefebvre AH. Atomization and sprays. New York: Hemisphere Publishing Corporation; 1989. [6] Rakopoulos CD, Rakopoulos DC, Antonopoulos KA. Multi-zone modeling of diesel engine fuel spray development with vegetable oil, bio-diesel or diesel fuels. Energy Convers Manage 2006;47(11):1550–73. [7] U.S. Department of Energy. Straight vegetable oil as a diesel fuel?.
E.P. Wagner et al. / Fuel 107 (2013) 757–765 [30] Bajpai S, Sahoo PK, Das LM. Feasibility of blending Karanja vegetable oil in petro-diesel and utilization in a direct injection diesel engine. Fuel 2009;88(4):705–11. [31] Leenus Jesu Martin M, Edwin Geo V, Kingsly Jeba Singh D, Nagalingam B. A comparative analysis of different methods to improve the performance of cotton seed oil fuelled diesel engine. Fuel 2012;102:372–8. [32] Hazar H, Aydin H. Performance and emission evaluation of a CI engine fueled with preheated raw rapeseed oil (RRO)–diesel blends. Appl Energy 2010;87(3):786–90. [33] Misra RD, Murthy MS. Performance, emission and combustion evaluation of soapnut oil–diesel blends in a compression ignition engine. Fuel 2011;90(7):2514–8. [34] Agarwal AK, Rajamanoharan K. Experimental investigations of the performance and emissions of Karanja oil and its blends in a single cylinder agricultural diesel engine. Appl Energy 2009;86:106–12. [35] Nwafor OMI, Rice G. Performance of rapeseed oil blends in a diesel engine. Appl Energy 1996;54:345–54. [36] Huzayyin AS, Bawady AH, Rady MA, Dawood A. Experimental evaluation of diesel engine performance and emission using blends of jojoba oil and diesel fuel. Energy Convers Manage 2004;45:2093–112. [37] Altun S, Bulut H, Oner C. The comparison of engine performance and exhaust emission characteristics of sesame oil–diesel fuel mixture with diesel fuel in a direct injection diesel engine. Renew Energy 2008;33(8):1791–5. [38] Forson FK, Oduro EK, Hammond-Donkoh E. Performance of jatropha oil blends in a diesel engine. Renew Energy 2004;29:1135–45. [39] Wagner EP, Koehle MA, Moyle TM, Lambert PD. Predicting temperature dependent viscosity for unaltered waste soybean oil blended with petroleum fuels. J Am Oil Chem Soc 2010;87(4):453–9.
765
[40] ASTM D976. Standard test method for calculated cetane index of distillate fuels. West Conshohocken (PA): American Society for Testing and Materials; 2011. [41] Rakopoulos CD, Rakopoulos DC, Giakoumis EG, Dimaratos AM. Investigation of the combustion of neat cottonseed oil or its neat bio-diesel in a HSDI diesel engine by experimental heat release and statistical analyses. Fuel 2010;89:3814–26. [42] Sahin Z. Experimental and theoretical investigation of the effects of gasoline blends on single-cylinder diesel engine performance and exhaust emissions. Energy Fuel 2008;22(5):3201–12. [43] Cha J, Kwon S, Park S. An experimental and modeling study of the combustion and emission characteristics for gasoline-diesel dual-fuel engines. Proc Inst Mech Eng Part D: J Automob Eng 2011;225(6):801–12. [44] Asfar KR, Hamed H. Combustion of fuel blends. Energy Conserv Manage 1998;39(10):1081–93. [45] Bilgin A, Durgun, Sahin Z. The effects of diesel–ethanol blends on diesel engine performance. Energy Sources 2002;24:431–40. [46] ASTM D975-11b. Standard specification for diesel fuel oils. West Conshohocken, PA: American Society for Testing and Materials; 2011. [47] Sandy Brae Laboratories, Inc. Water test kits.