Spray Characteristics of an Airblast Atomizer on Biodiesel Blends

Transcription

1 Spray Characteristics of an Airblast Atomizer on Biodiesel Blends C. R. Krishna and Thomas Butcher Energy Resources Division Brookhaven National Laboratory Building 526, Upton, NY , USA Abstract The cost and supply constraints of petroleum fuels and the concerns of global climate change has renewed interest broadly in finding liquid biofuels as extenders and substitutes for oil. Biodiesel is one such biofuel, which has characteristics very similar to diesel fuel, or more generally, to distillate fuels from petroleum (ASTM # 2 fuel). There is a growing body of literature on the use of blends of biodiesel and distillate fuels in various types of combustion equipment such as diesel engines, and space heating equipment. The work presented here was done in the context of testing such blends in another type of power generation equipment, microturbines. The injector, whose spray characteristics were tested, uses a plain jet type airblast atomizer. The sprays from the atomizer were characterized using a Malvern System 2600 spray analyzer. Several blends of biodiesel with ASTM # 2 oil and also water as a high surface tension liquid were tested at various atomizing air to liquid flow ratios. A major conclusion was that the air to liquid flow ratio was the key parameter determining the spray characteristics as defined by the Sauter mean diameter (SMD) and the mass median diameter (MMD) and these diameters were relatively insensitive to the changes in viscosities and surface tensions of the liquids tested. The commonly used Lefebvre correlation for the SMD was found to represent well the data for the MMD, and it is shown that with a simple multiplier, it could be used to replicate the SMD data as well. The conclusion for the turbine operation is that the overall combustion performance with the biodiesel blends should be similar to that with the petroleum distillate fuel. Corresponding Author 1

2 Background and Introduction The cost and supply constraints of petroleum fuels and the concerns of global climate change has renewed interest broadly in finding liquid biofuels as extenders and substitutes for oil. Most gasoline used in cars nowadays is blended with some levels of ethanol derived in the United States primarily from corn at present. There are projects aimed at producing ethanol from biomass sources, which could be more abundant and economical. Another biofuel that has begun to receive more attention is biodiesel. Biodiesel, as the name implies, has characteristics very similar, to diesel fuel, or more generally, to distillate fuels from petroleum (ASTM # 2 fuel). Hence, it is amenable to blending with diesel and home heating fuels (and obviously with heavier petroleum fuels as well). There is a growing body of literature on the use of blends of biodiesel [1] and distillate fuels in various types of combustion equipment such as diesel engines [2], and space heating equipment [3]. The work being reported here was done in the context of testing such blends in another type of power generation equipment, microturbines. Microturbines are small gas turbines coupled to electrical generators and fired with natural gas or distillate fuels. They are small in the sense that the electric power outputs for individual units are typically much less than a megawatt and are usually sited at the location where the power (and the waste heat) is used, so-called distributed generation [4]. The liquid fired microturbine used in the course of this study is a single-spool, simple cycle gas turbine [4]. It has three fuel injectors in the combustion chamber and each injector is built around a plain jet type air-blast atomizer. It was this typical atomizer that formed the subject of the tests reported here. While the injectors work under varying air pressure and temperature conditions in the turbine, the spray measurements reported here were done while spraying into ambient conditions in the laboratory. The fuel and air pressures entering the injector were varied to change its operating parameters. Hence, the value of the key results are in the differences, if any, seen with the different blends and any such differences could be used to help understand any differences in the combustion performance of the microturbine on the same blends. Experimental The atomizer, which forms the core of the injector, was set up vertically above an enclosure with openings to allow clear access to the laser incident and diffracted beams from a Malvern System 2600 spray analyzer. This enclosure sits above a collecting drum with a small vertical, downward current of air maintained in the enclosure. The set up including part of the Malvern System 2600 spray analyzer is shown in figure 1 below. The fuels are delivered to the atomizer by a small solenoid pump with valves to regulate pressure and hence flow rate. The flow rate is measured by timing the weight change of the reservoir sitting on a scale. The laboratory air supply is connected to the atomizer through a pressure regulator and a rotameter to measure the flow rate of atomizing air. The data is acquired by the Malvern system using their so-called model independent mode. Four fuels were tested. Water was used to establish the protocol and also gave a high surface tension fluid for the tests. This was felt to be important to check the correlation used to represent the mean diameters [5,6], as there was not much difference between the surface tensions of the ASTM # 2 oil and the biodiesel used for the blends. After the water tests, tests were conducted using ASTM # 2 oil, neat biodiesel, designated B 100 and a blend of the two oils, designated B 50. The spray measurements were taken at two fuel flow rates that corresponded to the low and high fuel (ASTM # 2 oil) flow rates for the nozzle during turbine operation. For each fuel flow rate, the air flow rate was varied so that drop sizes could be measured over a range of atomizing air to fuel ratios. The properties of the liquids tested are given in Table 1 below. Results and Discussion Figures 2 and 3 below reproduce digital pictures taken of the # 2 oil and B 100 sprays respectively. They are given here to show the fairly small spray angle of the sprays from the air blast atomizer. Both these pictures were taken at the same low flow rate (see above) and at similar air flow rates. The figures indicate an approximate spray cone angle of about

3 The atomization data will be presented mainly in the form of the mass median diameter, MMD and the Sauter mean diameter, SMD for water and the different fuels. The data from the measurements for the low and high flow rates will be presented separately. These will be compared with the correlation that has been proposed in Lefebvre [7]. Data for the Relative Span Factor [7,8] will be given as well. Figure 4 shows the mass mean diameters (MMD) for water sprays as a function of the ratio of air to liquid flow rates at the two liquid flow rates. The expected behavior of reduced mean diameters as the air to liquid ratios is seen in both cases. Similar behavior is seen for all the fluids tested and figure 5 illustrates this for B 100. Again as expected the Sauter mean diameters (SMD) behave similarly and figures 6 and 7 illustrate this for the same two fluids. The diameters are plotted against the air to liquid flow rate ratio, both because the latter is a key parameter in air atomization and can also be controlled in the equipment operation. This is also borne out emphatically when the mean diameters for all the liquids are compared as shown below in figures 8 and 9 at the high flow rate. Despite the variations in surface tension and viscosity, the key physical parameters, the mean diameters at a particular air to liquid flow ratio do not vary much between the liquids. While it is hard to estimate the error bars for these measurements, one has the feeling that the variations seen, which are not systematic and so not suggesting an effect of these parameters, is within the range of errors. The behavior is similar at the low flow rate too. There have been several correlations published for calculating the mean diameters of air blast atomizers from the physical parameters for the liquids and the air to liquid ratios [5,6,7]. The following correlation, equation (1), from Lefebvre [7], which is given for calculating the SMD, was chosen after trying some of the others as it gave the best reasonable fit, but for MMD instead of SMD as seen below. It is dimensionally correct and this or some modified form is also suggested in Tsai [5] and Mansour [6] although the reproduction in the latter reference has a typographical error. SMD/d 0 = 0.48( /( A U 2 R d 0 )) 0.4 (1+ALR -1 ) (µ 2 L /(( L d 0 )) 0.5 (1+ALR -1 ) (1) Where, SMD Sauter Mean Diameter d 0 diameter of the liquid jet surface tension of the liquid A air density U R relative velocity between the air and liquid flows ALR air to liquid flow rate ratios µ L viscosity of the liquid Figure 10 below compares the correlation with the measured values for the MMD for both the high and low flow rates of the water spray. It can be seen clearly that the correlation is a very good representation of the measured data for the MMD, except at the very low value of the air to liquid ratio. Quite probably, the atomization is not very good at this low value leading to the apparent very large mass mean diameter. Figure 11 shows the same correlation, which is ostensibly for SMD, compared with the measured SMD data. One can see that the agreement is not good, although the trends are parallel, except as before for the lowest air to liquid flow ratio. It would seem that the correlation could me made to fit with a simple multiplier and this is reinforced by comparing the ratios of the MMD to SMD at the different air to liquid flow ratios as shown in figure 12 below. The ratios are fairly constant with the lone point of exception. Figures 13 and 14 give the median diameter ratios for all the liquids tested and one can conclude the same thing despite a certain amount of scatter and no systematic departure. Bossard and Peck [8] have suggested that the droplet size distribution affects the flame from a fuel spray and that it is represented well by the relative span factor as defined by Lefebvre [7]. It is defined as Span= (D 0.9 -D 0.1 )/D 0.5 ) ( 2) Where, D 0.9 the droplet diameter such that 90% of the total liquid volume is in droplets of smaller diameters D 0.1 the droplet diameter such that 10% of the total liquid volume is in droplets of smaller diameters D 0.5 is the mass (volume) median diameter (MMD) The span is compared for all the liquids tested at the two flow rates in figures 15 and 16 below. There is not much difference in the span between the different fuels (biodiesel blends and # 2 oil) suggesting that, broadly speaking, their combustion performance should be similar. 3

4 Conclusions 7. A.H. Lefebvre, Atomization and A plain jet type air blast atomizer, which Sprays, Hemisphere publishing forms the core of an injector in a microturbine was tested. The fuels used were biodiesel and its 8. Corporation, J.A. Brossard and R.E. Peck, Droplet blends in ASTM # 2 oil. Water was also used to size distribution effects in spray provide a high surface tension fluid for comparison. The major conclusion, which is in accord with previous work, is that the air to combustion, Twenty-sixth Symposium (International) on Combustion, The Combustion Institute, , liquid flow ratio is the key parameter controlling the mean drop sizes. One of the most commonly used correlations [7] for the SMD was found to represent the data for the MMD quite well for all the fluids. It could be made to represent the SMD data with a simple multiplying factor. The relative span factor was also similar for all the fluids. All this would suggest that the injector and hence the microturbine combustion performance on the biodiesel blends should be substantially similar to that of the base ASTM # 2 oil. Acknowledgements The results in the paper were developed in the course of a research project funded by the Department of Energy and supported by Capstone Turbine Corporation. The support of Don Geiling, the project manager and Debbie Haught as program manager is very much appreciated. The technical and material support of Craig Smugeresky and Steve Gillette contributed to the accomplishment of the project substantially. Thanks are due George Wei for help with the experiments. References 1. G. Knothe, J. Gerpen, and J. Krahl (Editors), The Biodiesel Handbook, AOCS Press, M.S. Graboski, and R. L. McCormick, Combustion of fat and vegetable oil derived fuels in diesel engines, Prog. Energy Combust. Sci. 24: , C.R. Krishna, Biodiesel blends in Space Heating, BNL-68852, Capstone Turbine Corporation web site: S.C. Tsai and B. Viers, Airblast atomization of viscous Newtonian liquids using twin-fluid jet atomizers of various designs, Journal of Fluids Engineering, 114: , A. Mansour and N. Chigier, Air-blast atomization of non-newtonian liquids, J. Non-Newtonian Fluid Mech., 58: ,

5 Fuel Surface Tension, mn/m Viscosity, mpas Density, Kg/M 3 Water #2 fuel B B Table 1. Properties of fluids tested Figure 1. Spray measurement facility 5

6 Figure 2. Spray of # 2 oil Figure 3. Spray of B 100 6

7 D43, Microns Air to Liquid Ratio Figure 4. MMD for water sprays D43, Microns Air to Fuel Ratio Figure 5. MMD for B 100 sprays 7

8 D32, Microns Air to Liquid Ratio Figure 6. SMD for water sprays D32, Microns Air to Fuel Ratio Figure 7. SMD for B 100 sprays 8

9 D43, Microns Air to Fuel Ratio Water B 50 B 100 Fuel Oil Figure 8. MMD for all the liquids D32, Microns Air to Fuel Ratio Water B 50 B 100 Fuel Oil Figure 9. SMD for all the liquids 9

10 D43, Microns Air to Liquid Ratio LFCorr HFCorr Figure 10. Comparison of the correlation with MMD for the water sprays D32, Microns Air to Liquid Flow Ratio LFCorr HFCorr Figure 11. Comparison of the correlation with SMD for the water sprays 10

11 MMD/SMD Air to Liquid Ratios Figure 12. MMD/SMD for the water sprays MMD/SMD Air to Liquid Flow ratio B 100 B 50 Water # 2 oil Figure 13. MMD/SMD for all the liquids under low flow conditions 11

12 MMD/SMD B 100 B 50 Water # 2 Oil Air to Liquid Flow Ratio Figure 14. MMD/SMD for all liquids under high flow conditions Span Air to Fuel Ratio Water # 2 Fuel B 50 B 100 Figure 15. Relative span factor for all liquids under low flow conditions 12

13 Span Water # 2 Fuel B 50 B Air to Fuel Ratio Figure 16. Relative span factor for all the liquids under high flow conditions 13