A DISSERTATION SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE UNIVERSITY OF MINNESOTA BY. Anil Singh Bika

Transcription

1 Synthesis Gas Use in Internal Combustion Engines A DISSERTATION SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE UNIVERSITY OF MINNESOTA BY Anil Singh Bika IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY Professor David B. Kittelson December 2010

2 Anil Singh Bika 2010

3 ACKNOWLEDGMENTS I would like to begin this section by thanking my advisor, Professor David B. Kittelson, for all of the mentoring, instruction, and support that I received. His door was always open for me. I would also like to acknowledge my committee members Professors Z. Sun, D. Pui, and L. Schmidt, for agreeing to be part of my doctoral process and willing to help me along the way. I would also like to acknowledge all of the help I received from many of the individuals from the 4 th floor engines lab, including Ms. Grace Chan, Mr. Darrick Zarling, and Dr. Win Watts. I would also like to specially thank my fellow graduate students from the lab for making these last 5 years so much fun: Dr. Luke Franklin, Dr. Jacob Swanson, Mr. Andre Olson, Mr. Wei Fang, Mr. Helmer Acevedo, Mr. Lei Tian, Mr. Adam Ragatz, Mr. David Gladis, Mr. Joe Sweeney and Mr. David Bennett. With the group of guys that we had in the lab, work never seemed like work (maybe that s why it took so long). I m not forgetting the rest of my graduate student friends from the department, but there are just so many to list. All of the members that were once a part of GENS (graduate engineers networking society). I would also like to acknowledge the support I received from the machine shop, tool crib, and student machine shop. Most importantly, help from Robin Russell in the machine shop, Mark Ericson in the tool crib, and Mel Chapin and Peter Zimmerman in the student machine shop. I would also like to acknowledge the funding support I received throughout the years from the Minnesota Corngrowers Association and Initiative for Renewable Energy and the Environment (IREE). Lastly, and most importantly, I would like to acknowledge the support from my family in the USA (Minnesota, Chicago, Michigan, and West Virginia) and India (Rajasthan). i

4 DEDICATION This dissertation is dedicated to my grandparents, aunts, uncles, mom, dad, sister, cousins, nieces, and nephews (my family). Most importantly, this dissertation is dedicated to my wife, Soumya. ii

5 Table of Contents ACKNOWLEDGMENTS... i DEDICATION... ii LIST OF TABLES... vi LIST OF FIGURES... vii CHAPTER 1: INTRODUCTION... 1 INTRODUCTION... 2 MOTIVATION... 3 OBJECTIVE... 4 CHAPTER 2: BACKGROUND ON SYNTHESIS GAS AND ENGINES... 5 SYNTHESIS GAS... 6 Background... 6 Production Techniques... 6 Synthesis Gas Properties... 8 ENGINES... 9 Ethanol in CI Engines... 9 Syngas Use in Compression Ignition Engines Syngas Use in Spark Ignition Engines Syngas Use in HCCI Engines CHAPTER 3: THEORETICAL FOUNDATION ENGINES Engine Geometry In-Cylinder Pressure Heat Release Rate Analysis Temperature Calculation Start of Combustion and Combustion Duration IMEP Calculation Fuel to Air Ratio Calculation FUELS Ignition Fundamentals Knock and Octane Number CHAPTER 4: HYDROGEN AND ETHANOL FUELED CI ENGINE OBJECTIVE EXPERIMENTAL SETUP Engine Test Stand Engine Test Procedure RESULTS Ethanol Lubricity H 2 Addition With Ethanol Kinetic Simulation of H 2 Addition DISCUSSION CHAPTER SUMMARY CHAPTER 5: SYNGAS AND DIESEL FUELED CI ENGINE OBJECTIVE iii

6 EXPERIMENTAL SETUP Engine Test Stand Emissions Testing Conditions RESULTS Cycle Efficiency NO x Emissions CO and H 2 Emissions DISCUSSION Cycle Efficiency NO x Emissions CHAPTER SUMMARY CHAPTER 6: SYNGAS FUELED SI ENGINE OBJECTIVE EXPERIMENTAL SETUP Engine Test Stand Engine Test Procedure Knock Quantification RESULTS Knock Limits Constant Equivalence Ratio Tests Constant Compression Ratio Tests CHAPTER SUMMARY CHAPTER 7: HYDROGEN FUELED HCCI ENGINE OBJECTIVE EXPERIMENTAL SETUP Engine Test Stand Testing Methodology RESULTS In-Cylinder Combustion Combustion Characteristics IMEP and Efficiency Heat Transfer to the Walls CHAPTER SUMMARY CHAPTER 8: SYNGAS FUELED HCCI ENGINE OBJECTIVE EXPERIMENTAL SETUP Engine Test Stand Engine Test Procedure RESULTS Effect of Varying Intake Temperature Combustion Characteristics Efficiencies CHAPTER SUMMARY iv

7 CHAPTER 10: CONCLUSIONS HYDROGEN AND ETHANOL FUELED CI ENGINE SYNGAS AND DIESEL FUELED CI ENGINE SYNGAS FUELED SI ENGINE HYDROGEN FUELED HCCI ENGINE SYNGAS FUELED HCCI ENGINE BIBLIOGRAPHY APPENDIX A APPENDIX B v

8 LIST OF TABLES Table 2.1: The properties of hydrogen, carbon monoxide, and methane Table 4.1: Engine test matrix Table 5.1: Engine Specifications Table 5.2: Test matrix Table 6.1: Test Matrix Table 6.2: Summary table of combustion properties for tests at a constant EQR of 0.6 Table 6.3: Summary table of combustion properties for tests at a constant CR of 6:1 Table 7.1: Engine specifications vi

9 LIST OF FIGURES Figure 3.1: Example of an in-cylinder pressure trace Figure 3.2: Example of an engine indicator diagram Figure 3.3: Plot showing IMEP determination graphically from indicator diagram Figure 4.1: Lubricity testing results showing wear scar diameters of various proportions of castor oil and lauric acid in ethanol at 25 C Figure 4.2: Engine operating at a compression ratio of 24:1 and intake temperature of 120 C, with and without 5 slpm of H2- addition with an ethanol injection timing of -13 CAD (ATDC) Figure 4.3: Engine operating at a compression ratio of 24:1 and an intake temperature of 120 C, with and without 5 slpm of H2 addition with an ethanol injection timing of CAD (ATDC) Figure 4.4: Ethanol injection timing and duration of -11 and 23 CAD (ATDC), respectively, at a CR 24:1 and an ethanol only intake temperature of 120 C, compared with the H2 addition condition at 5 slpm and an intake temperature of 80 C Figure 4.5: Ethanol injection timing and duration of -11 and 21 CAD (ATDC), respectively, at a CR 24:1 and an ethanol only intake temperature of 120 C, compared with the H2 addition condition at 5 slpm and an intake temperature of 80 C Figure 4.6: Ethanol injection timing of -13 CAD (ATDC), with straight ethanol at CR 24:1 and intake temperature of 120 C, compared to H2 addition of 10 slpm at CR 19:1 and intake temperature of 80 C Figure 4.7: Ethanol injection timing of -12 CAD (ATDC), with straight ethanol at CR 24:1 and intake temperature of 120 C, compared with H2 addition of 10 slpm at CR 19:1 and intake temperature of 80 C vii

10 Figure 4.8: The ignition delay times of ethanol with hydrogen addition at an equivalence ratio of 0.3 at various temperatures Figure 4.9: The ignition delay times of ethanol with H2 addition at an equivalence ratio of 0.9 at various temperatures Figure 4.10: The ignition delay times of n-heptane with hydrogen addition, at an equivalence ratio of 0.3 at various temperatures Figure 4.11: The ignition delay times of n-heptane with various amounts of H2 addition, at an equivalence ratio of 0.9 Figure 5.1: Syngas and Diesel engine test stand schematic Figure 5.2: Cycle efficiency at 2 bar IMEP Figure 5.3: Cycle efficiency at 4 bar IMEP Figure 5.4: Effect of varying syngas composition on cycle efficiency with engine operating at 2 bar IMEP Figure 5.5: Effect of varying syngas composition on cycle efficiency with engine operating at 4 bar IMEP Figure 5.6: Cycle efficiency reduction with increasing gaseous fuel equivalence ratio 2 bar IMEP Figure 5.7: Cycle efficiency reduction with increasing gaseous fuel equivalence ratio 4 bar IMEP Figure 5.8: NOx emissions at 2 bar IMEP Figure 5.9: NOx Emissions at 4 bar IMEP Figure 5.10: CO Emissions at 2 bar IMEP Figure 5.11: CO Emissions at 4 bar IMEP Figure 5.12: H2 Emissions at 2 bar and 4 bar IMEP Figure 5.13: CO Emissions for 100% H2 condition at 2 bar and 4 bar IMEP Figure 5.14: Cycle efficiency of 100% CO condition at 2 bar IMEP viii

11 Figure 5.15: Cycle efficiency of 100% CO condition at 4 bar IMEP Figure 5.16: Cycle efficiency of 100% H2 condition at 2 bar IMEP Figure 5.17: Cycle efficiency of 100% H2 condition at 4 bar IMEP Figure 5.18: Increasing exhaust emissions of H2 and CO with increasing gaseous fuel equivalence ratio at 2 bar and 4 bar IMEP Figure 5.19: NO2 to NOx ratio at 2 bar IMEP Figure 5.20: NO2 to NOx ratio at 4 bar IMEP Figure 6.1: Showing knocking and knocking in-cylinder pressure traces for: a.) graphical definition of knock index of a knocking cycle with a KI of 0.83 bar, b.) a non-knocking cycle c.) a lightly knocking cycle with a KI of 0.25 bar, and d.) a moderately knocking cycle with a KI of Figure 6.2: Showing the increasing knock resistance of syngas with increasing CO fraction Figure 6.3: Showing how spark timing must be advanced to maintain MBT as the CO fraction in syngas is increased Figure 6.4: Showing the in-cylinder pressure traces and heat release rates with varying H2/CO proportions at MBT spark timing and an equivalence ratio of 0.6 with a compression ratio of: a.) 6:1, b.) 8:1, and c.) 10:1 Figure 6.5: Showing the cumulative heat release rate at MBT spark timings and an equivalence ratio of 0.6 with a compression ratio of: a.) 6:1, b.) 8:1, c. 10:1 Figure 6.6: Showing the rapid burn angle increasing with increasing CO fraction at a compression ratio of 6:1, 8:1, and 10:1. Figure 6.7: Showing the burn duration increasing with increasing CO fraction at a CR of 6:1, 8:1, and 10:1. Figure 6.8: Showing the ignition lag increasing with increasing CO fraction at a CR of 6:1, 8:1, and 10: ix

12 Figure 6.9: Showing how spark timing must be advanced to maintain MBT as the CO fraction in syngas is increased at a constant CR of 6:1. Figure 6.10: Showing the in-cylinder pressure traces and heat release rates with varying H2/CO proportions at MBT spark timing and a compression ratio of 6:1 with an equivalence ratio of: a.) 0.6, b.) 0.7, and c.) 0.8 Figure 6.11: Showing the cumulative heat release rate at MBT spark timings and a compression ratio of 6:1 with an equivalence ratio of: a.) 0.6, b.) 0.7, and c.) 0.8 Figure 6.12: Showing a slight increasing trend in rapid burn angle with increasing CO fraction at a CR of 6:1 and an EQR of 0.6, 0.7, and 0.8. Figure 6.13: Showing the burn duration increasing with increasing CO fraction in the syngas mixture at an EQR of 0.6, 0.7, and 0.8. Figure 6.14: Showing the ignition lag increasing with increasing with increasing CO fraction at an EQR of 0.6, 0.7, and 0.8. Figure 7.1: Showing the H2 HCCI engine test stand schematic. Figure 7.2: In-cylinder pressure traces of the engine operating at various intake temperatures for, (a) lambda = 4.38, (b) lambda 3.64, and (c) lambda 3.16 Figure 7.3: Heat release rate plots of the engine operating at various intake temperatures for, (a) lambda = 4.38, (b) lambda = 3.64, and (c) lambda = 3.16 Figure 7.4: Cumulative heat release plots of the engine operating at various intake temperatures for, (a) lambda = 4.38, (b) lambda = 3.64, and (c) lambda = 3.16 Figure 7.5: Start of combustion for lambda 3.16, 3.64, and 4.38 Figure 7.6: MFB50 for lambda 3.16, 3.64, and 4.38 Figure 7.7: Burn duration for a lambda of 3.16, 3.64, and x

13 Figure 7.8: IMEP varying with intake temperature for a lambda of 3.16, 3.64, and 4.38 Figure 7.9: Cycle efficiency with varying intake temperature for a lambda of 3.16, 3.64, and 4.38 Figure 7.10: Combustion efficiency with varying intake temperatures for a lambda of 3.16, 3.64, and 4.38 Figure 7.11: Cycle efficiency and combustion efficiency of the engine operating at; (a) a lambda of 4.38, (b) lambda of 3.64, and (c) lambda of 3.16 Figure 7.12: Schematic of energy flow from the combustion chamber Figure 7.13: Heat transfer to the walls at various intake temperatures for a lambda of 3.16, 3.64, and 4.38 Figure 8.1: Showing the in-cylinder pressure traces and heat release rates of the engine operating at varying intake temperatures with an: (a) EQR = 0.26 with 100% H2, (b) EQR = 0.3 with 100% H2, (c) EQR = 0.26 with a 50/50 H2/CO ratio, and (d) EQR = 0.30 with a 50/50 H2/CO ratio Figure 8.2: Showing the peak pressure of the best IMEP conditions of the engine operating at an EQR of 0.26 and 0.30 with varying ratios of H2/CO. Figure 8.3: In-cylinder cylinder temperature traces: (a) with an EQR = 0.26 with 100% H2, (b) EQR = 0.3 with 100% H2, (c) EQR = 0.26 with a 50/50 H2/CO ratio, and (d) EQR = 0.30 with a 50/50 H2/CO ratio Figure 8.4: Showing the peak cylinder temperature of the best IMEP conditions of the engine operating at an EQR of 0.26 and 0.30 with varying ratios of H2/CO. Figure 8.5: Showing the intake temperature must be increased for increasing CO fraction in the syngas mixture to maintain best IMEP. The plot shows data points for EQR = 0.26 and EQR = xi

14 Figure 8.6: Showing the crank angle locations where the 10% and 50% cumulative heat release (CA10 and CA50) occurred for: (a) EQR = 0.26 and (b) EQR = Figure 8.7: Showing the rapid burn angle for an EQR = 0.26 and EQR = 0.30, with varying H2/CO ratios. Figure 8.8: Showing the combustion efficiency and cycle efficiency of the engine operating at an EQR= 0.26 and EQR = 0.30 with varying proportions of H2/CO. Figure 8.9: The proportion of combusted fuel energy leaving the system as exhaust, work, and wall heat transfer energies, for an EQR of: (a) 0.26, and (b) Figure A.1: COV IMEP for H 2 only tests for Lambda of 4.38 Figure A.2: COV IMEP for H 2 only tests for Lambda of 3.64 Figure A.3: COV IMEP for H 2 only tests for Lambda of 3.16 Figure B.1: COV IMEP for 50/50 H 2 / CO tests for EQR of 0.26 Figure B.2: COV IMEP for 75/25 H 2 / CO tests for EQR of 0.26 Figure B.3: COV IMEP for 100/0 H 2 / CO tests for EQR of 0.26 Figure B.4: COV IMEP for 50/50 H 2 / CO tests for EQR of 0.30 Figure B.5: COV IMEP for 75/25 H 2 / CO tests for EQR of 0.30 Figure B.6: COV IMEP for 100/0 H 2 / CO tests for EQR of xii

15 CHAPTER 1: INTRODUCTION 1

16 INTRODUCTION The internal combustion engine (ICE) has had a significant impact on the world over the last century. Even though it was invented over 100 years ago, there is still a vast amount of research that is devoted to improving ICE operation. The purpose of an ICE, however, remains the same; to convert chemical fuel energy into useable shaft power. The main mode of this conversion is currently accomplished using either four stroke spark ignition (SI) or compression ignition (CI) engines. The need to reduce fuel consumption and engine emissions has caused significant interest in homogeneous charge compression ignition (HCCI) engines. In HCCI engines, a lean mixture of homogeneously distributed fuel and air is inducted into the cylinder volume during the intake stroke. As the piston approaches top-dead-center (TDC) during the compression stroke, multiple-site autoignition takes place throughout the lean mixture giving rise to a rapid combustion event, with minimal oxides of nitrogen (NO x ) and particulate matter (PM) emissions. These engines combine some positive aspects of both spark ignition and compression ignition engines; however there are difficulties that must be overcome before these engines can be put into production. A wide variety of fuels have been investigated for HCCI use (Onishi, 1979; Noguchi, 1979, Najt and Foster, 1983; Thring, 1989). With gasoline HCCI, the main focus had been on varying the intake temperature and using exhaust gas recirculation (EGR) to control the combustion process. With advances in control technology, researchers began altering the in-cylinder charge mixture via variable valve actuation (VVA) techniques (Zhao, 2007; Milovanovic, 2004). Diesel HCCI utilizes the conventional direct injection (DI) system with electronic control to inject the diesel fuel early in the compression stroke. Another approach to diesel HCCI, called modulated kinetics (MK) mode, has been developed and put into production by the Nissan Motor Company, where the diesel fuel is injected very late and even after top dead center (TDC) giving an extended amount of time for mixing (Kimura, 1999). There has also been HCCI work on alternative fuels 2

17 like natural gas (Yap, 2006; Hosseini, 2008) and DME (Shudo, 2003). The work with synthesis gas as an HCCI fuel, however, is extremely limited. Synthesis gas, also known as syngas, is a mixture of gases containing hydrogen, carbon monoxide, carbon dioxide, and nitrogen along with other gases. Syngas can be produced from a variety of feed stocks ranging from fossil coal to renewable biomass. This allows for syngas production to be tailored to a localized resource. The problem with having a variable feedstock and variable production methods is that the syngas composition can vary significantly. There have been investigations of hydrogen supplemented diesel and spark ignition engines. There is little published work, however, on the utilization of varying blends of synthesis gas in any (SI, CI, or HCCI) combustion regime. MOTIVATION Some specific motivating factors for this dissertation include the scarcity and economic swings of fossil derived fuels, which have increased the awareness and the need for the utilization of alternative fuels in internal combustion engines. Also, increasing CO 2 and greenhouse gas emissions, which are contributing to global climate change, have stressed the importance for efficiency and emissions research in the field. 3

18 OBJECTIVE The objective of this dissertation was to investigate the combustion characteristics of a compression ignition, spark ignition, and homogeneous charge compression ignition engine operating on various blends of synthesis gas. To fully investigate the three ICE operating regimes, experimental investigations were carried out to focus on: 1.) A CI engine operating on ethanol and hydrogen fuel 2.) A CI engine operating on diesel fuel with varying blends of synthesis gas 3.) An SI engine operating on varying blends of synthesis gas 4.) An HCCI engine operating on hydrogen fuel 5.) An HCCI engine operating on varying blends of synthesis gas The three operating modes (CI, SI, and HCCI) were selected because it is unlikely that an engine will be able to operate solely in an HCCI regime throughout the complete load range. The more common CI and SI regimes will likely be necessary for high load engine operation. The results from this doctoral work sheds light into the fundamental aspects of syngas combustion and also provides a foundation for future gasification plant designers and synthesis gas producers, regarding the fuel composition needs of a syngas powered internal combustion engine. The first 3 chapters of this dissertation provide an introduction and background for this doctoral work. The remaining chapters present the results and conclusions. 4

19 CHAPTER 2: BACKGROUND ON SYNTHESIS GAS AND ENGINES 5

20 SYNTHESIS GAS Background A brief introduction to synthesis gas properties and production techniques is provided in this section. A more detailed review can be found in Lieuwen et al. (2010). Although many compounds contain hydrogen; hydrogen is not found in an uncombined form in any significant quantity on earth, like fossil fuels. Various processes exist to obtain the hydrogen from the compounds containing hydrogen. A common form that the hydrogen gas is used in is called synthesis gas, also known as syngas, reformer gas, or producer gas, depending on the production technique. These gases are usually a blend of many gases (H 2, CO, CO 2, N 2, H 2 O, etc) of varying proportions, which have traditionally been used in heating applications and as industrial feedstocks. They have found limited applications as engine fuels, but typically have not been operated cleanly or efficiently. Production Techniques The syngas production process can either be endothermic or exothermic. The most common hydrogen production methods are auto-thermal reforming, steam reforming, and partial oxidation. A variety of hydrocarbons can be used in the steam reforming process. During the steam reforming process, the steam and hydrocarbon fuel source are reacted at high temperatures over a catalyst. Equation 2.1 shows an example of steam methane reforming (SMR) (Lieuwen, 2010), which is currently the main method for hydrogen production in the United States: 3 (2.1) 6

21 The hydrogen produced comes from both the hydrocarbon fuel source and the steam used.. A partial oxidation reaction, shown in equation 2.2, is a reaction that is run at fuel rich conditions, so that some of the hydrocarbon is oxidized with an insufficient amount of air or oxygen, producing the heat needed that is used in the reaction. 2 (2.2) Auto-thermal reforming combines both the partial oxidation and steam reforming reactions. The reactions take place over a catalyst that is chosen based on the selectivity of the gases, and uses the heat generated from the partial oxidation reaction to drive the process. The benefit of auto-thermal reforming is that it takes advantage of the heat generated from the partial oxidation reaction, along with the added H 2 production from the steam in the steam reforming reaction. This process can be beneficial for small applications and mobile transportation systems, by producing the synthesis gas ondemand. Gasification is another method of producing syngas through a combination of partial oxidation and steam reforming. There are a variety of feed stocks used in gasification, from coal to biomass. 7

22 Synthesis Gas Properties The properties of synthesis gas mixtures can vary significantly depending on the feedstock, production technique, and gas mixture proportions. Table 2.1 shows a comparison of H 2, CO, and methane (CH 4 ). Table 2.1: The properties of hydrogen, carbon monoxide, and methane* H 2 CO CH 4 Fuel LHV (MJ/kg) [MJ/N m 3 ] (121) [10.8] (10.2) [12.7] Air-Fuel Ratio (mass) [mole] (34.4) [2.38] (2.46) [2.38] Peak flame Temp 1 atm Flammability Limit φ(lean/rich) 0.01 / / 6.80 Flame Speed at Stoich. (cm/sec) (50.2) [35.8] (17.2) [9.52] / *Sridhar, G., Sridhar, H. V., Dasappa, S., Paul, P. J., Rajan, N. K. S., & Mukunda, H. S. (2005). Development of producer gas engines. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 219(3),

23 ENGINES An excellent review of conventional SI and CI engine operation can be found in various textbooks (Heywood, 1988; Stone, 1999) and will not be given here. A review of HCCI engine operation can also be found elsewhere (Zhao, 2007). This engines background section will provide the reader with a review of the current literature as it pertains to syngas and hydrogen use in internal combustion engines, relevant to the experiments conducted for this doctoral work. The section begins with a literature review on previous experiments conducted on ethanol use in CI engines. This is important for the ethanol and H 2 experiment which is detailed in chapter 4. Ethanol in CI Engines Ethanol, being an oxygenated high octane fuel, has seen much success as an SI engine fuel, with many government regulations requiring 10% blends of ethanol in gasoline and newer flexible-fuel vehicles operating on 85% ethanol. In CI engines, however, ethanol s success has been limited mainly due to lubricity, ignitibility, and miscibility issues (Yilmaz, 2005; Goering, 1998; Nagarajan, 2002; Hardenberg, 1981; King, 1992). Goering et al. (1998) investigated anhydrous and hydrous ethanol in a two-stroke CI engine. To overcome problems with lubricity, a Lubrizol additive was used. It was found that the hydrous ethanol reduced NO x, CO, and HC emissions. The engine power output was also reduced, due to the lower ethanol fuel energy content. After completing the 454 hour test, the only engine related issues were with sticking fuel injectors. Nagarajan et al. (2002) tested the performance of a CI engine running on ethanol with 1% castor oil used as a lubricity additive. Diethyl ether (DEE) was introduced into the intake manifold to serve as an ignition improver, so there was no need for intake air heating. Stable operation was achieved with 3% DEE by ethanol mass share; however the ignition 9

24 delay period of ethanol was lengthened by 8 crank angle degrees (CAD) compared to diesel fuel. Many researchers have investigated the lubricity of alcohols and low-sulfur diesel fuels (Estefan, 1990; Lacey, 1998; Lacey, 2000; Barbour, 2000; Lacey, 1997). A review paper on fuel lubricity can be found by Lacey et al. (1998) which describes the various lubricity testing methods applicable to CI engines. They also describe the details of the different wear mechanisms commonly seen in various diesel fuel injection systems. A range of commercially available additives are also given, however information on ethanol specific lubricity is lacking. Estefan et al. (1990) investigated 106 different compounds in concentrations of 1% by volume in methanol. Although the application was for SI engines, the testing approach utilized the ball on cylinder machine (BOCM), whose data can also be applied to CI engines. Apart from lubricity, another major problem with ethanol use in CI engines is its difficulty to auto-ignite in diesel engine conditions. Alcohol auto-ignition and ignition improvers have been investigated by many researchers (Lee, 1993; Bollentin, 1996; Schaeffer, 1981). Hydrogen, however, as a combustion modifier for an ethanol CI engine has not been investigated. Liu et al. (1995) investigated the ignition delay period of CI dual fuel engines using various gaseous fuels including hydrogen with a diesel pilot fuel. It was found that as the H 2 fuel admission was increased, the length of the diesel ignition delay period also increased. Nielsen et al. (1987) also investigated the ignition delay of a CI dual fuel engine, but they used n-heptane and cetane, along with diesel as pilot fuels. With all three fuels, hydrogen addition into the intake manifold caused an increase in the ignition delay. 10

25 Syngas Use in Compression Ignition Engines There has been significant interest in hydrogen and syngas because of its clean burning properties and the prospect of producing it from various renewable resources. A possible method of using small amounts of hydrogen in transportation applications is through onboard reforming (Li, 2004; Karim, 1966; Kavtaradze, 2005). Another way of hydrogen utilization is with CI dual fuel engines. Dual fuel engines operate on a primary (main) fuel and a secondary (pilot) fuel. The primary fuel source is typically a gaseous fuel such as hydrogen or natural gas, which is injected into the intake manifold. The secondary or pilot fuel is injected into the combustion chamber using the standard fuel injection equipment and is used to ignite the mixture (Lambe, 1992; Bika, 2009; Saravanan, 2009). Sahoo, et al. (2008) presents a critical review of dual fuel diesel engines. They describe how dual fuel engines can be beneficial for controlling both NO x and soot emissions, but have the potential of increasing HC and CO emissions. They explain that the literature shows the brake thermal efficiency to both increase and decrease with gaseous fuel injection, depending on the researcher. Avadhanula et al. (2009) conducted experiments on a stationary diesel engine generator with hydrogen addition. No increase in thermal efficiency was seen with hydrogen addition, with the engine operating at constant load. However, as the hydrogen concentration in the intake air was increased, the NO 2 emissions increased, while the NO emissions decreased at some conditions and increased at others. Lambe et al. (1992) carried out experiments where % of the fuel energy was supplied from hydrogen and the ignition was caused by a pilot quantity of diesel fuel. They noticed a 15% increase in efficiency after modifying the combustion chamber and retarding the diesel fuel injection timing. The NO x emissions were also reduced by up to 70%. Xiao et al. (2009) investigated the performance of an in-direct injection diesel engine fueled with low concentrations of hydrogen in the intake air. For these low 11

26 concentrations it was found that the engine brake power and thermal efficiency were less with hydrogen addition compared to straight diesel operation. The NO x emissions for some conditions were reported as being up to 30 40% lower for dual fuel operation compared to straight diesel operation. Dual fuel engine operation with syngas or a varying gaseous hydrogen content has been investigated by few researchers (Abu-Jrai, 2007; Garnier, 2005; Roy, 2009). Garnier et al. (2005) established a combustion predictive model to determine the ignition delay of a dual fuel syngas-diesel engine. The model also describes the various stages of burning encountered in this type of engine; 1) premixed combustion, 2) gaseous combustion, and 3) diffusion burning. Roy et al. (2009) investigated the performance and emissions of a dual fuel engine, by varying the hydrogen concentration in the syngas from 13.7% to 20%. They found that a broader operating range, in terms of equivalence ratio, could be achieved with the higher hydrogen concentration. The thermal efficiency was also higher for the high hydrogen concentration syngas. The NO x emissions were higher, however, for the higher hydrogen concentration syngas. Syngas Use in Spark Ignition Engines A significant amount of work has been dedicated to engine knock (Gerty, 2006; Baral, 2008; Chun, 1989; Gautam, 2000) and more specifically engine knock with gaseous fuels (Li, 2004; Karim, 1966; Attar, 2003; Rahmouni, 2004). Li et al. (2003) investigated the knock and combustion characteristics of methane, hydrogen, carbon monoxide, and their binary mixtures. They found that the knock limits of dry CO deteriorate significantly in the presence of small amounts of CH 4, H 2, or even H 2 O. These results of carbon monoxide s superior knock resistance over hydrogen and its knock resistance degradation with water was also reported by other researchers (Anzilotti, 1954). Szwaja et al. (2007) investigated combustion knock of hydrogen and also gasoline in SI engines. They used various methods of knock detection ranging from in-cylinder pressure, to piezoelectric accelerometers. It was determined that the knock detection 12

27 techniques used for gasoline engines are also applicable with hydrogen SI engines, with some modifications. Shrestha et al. (2008) investigated the effect of diluents on the knock rating of gaseous fuels. Their knock detection method utilized the third derivative of the in-cylinder pressure with varying blends of methane and hydrogen (Checkel, 1986; Checkel, 1989). With these blends they added either carbon dioxide or nitrogen, both of which increased the knock resistance of the fuels. Of the two diluents, carbon dioxide increased the knock resistance roughly five times over nitrogen. Syngas Use in HCCI Engines The first work with HCCI was performed by Onishi et al. (1979). The name they gave to this combustion was active thermo-atmosphere combustion (ATAC). A two-stroke SI engine with a 7.5:1 compression ratio was used. They were able to shift from SI mode to ATAC mode without issue, and noticed that there was no discernable flame propagation, but instead combustion occurred spontaneously at multiple points. In the same year as Onishi et al., Noguchi et al. (1979) also investigated HCCI, calling it TS (Toyota-Soken) combustion. They used a horizontally opposed two-stroke SI engine. In 1983, Njat and Foster investigated 4-stroke HCCI operation in a CFR engine; giving it the name compression ignited homogeneous charge (CIHC). They used primary reference fuels and looked at varying parameters such as compression ratio, intake air heating, and EGR. They found that the HCCI combustion was controlled by the global hydrocarbon kinetics. They also reported that the low temperature kinetics controlled the ignition of the fuel, while the high temperature kinetics controlled the heat release rate and combustion duration. In 1989, Thring investigated 4-stroke HCCI operation in a single cylinder engine. It was found that significant amounts of EGR (30%) and significant intake air heating (370 C) were required for successful HCCI operation. The operation, however, was restricted to moderate and low loads. Hydrogen HCCI operation, however, has been investigated by few researchers. 13

28 Caton et al. (2009) investigated hydrogen HCCI in a single cylinder indirect injection (IDI) diesel engine. They tested various compression ratios, fueling concentrations, and intake temperatures. They found significant unburned hydrogen in the exhaust (0.5 1%) and the efficiency was lower than engine operation in CI mode. This lower efficiency was attributed to high heat transfer and low combustion efficiency. Gomes Antunes et al. (2008) investigated hydrogen HCCI in a CI engine with a compression ratio of 17:1. The ignition timing was adjusted by heating the intake air. It was found that an operating region from a lambda of 3 to 6 was possible. The maximum brake thermal efficiency was reported as 45%. Stenlaas et al. (2004) investigated the efficiency, combustion phasing and emissions from a hydrogen HCCI engine. Similar to Gomes Antunes et al. (2008), they found that the operating regime was from a lambda of 3 to 6. They found that as the intake temperature was increased, the start of combustion also advanced. They also found that for the loads tested, HCCI engine operation exceeded the thermal efficiency of SI engine operation. Stenlaas et al. (2004) investigated reformed methanol gas as a straight HCCI engine fuel. They compared their findings against an HCCI engine operating on straight hydrogen. They found that they were able to extend the load range beyond H 2 operation alone, however the range was still quite small and only quite lean operation was possible. Syngas HCCI combustion also resulted in a very fast heat release rate in HCCI mode. The SOC control was very difficult and considerable emissions of CO were seen along with some HC emissions which were attributed to the lube oil. Most of the other work related to syngas use in HCCI engines has focused on supplemental use of syngas and have come from three main institutions: 1) University of Alberta, Canada, 2) University of Birmingham, England, and 3) University of Hokkaido, Japan. The focus of these investigations has been on using the syngas mixture as an ignition modifier and combustion controller in a dual fuel HCCI engine. This means that the engine is supplied with a main fuel, such as natural gas or di-methyl ether, and only 14

29 small amounts of syngas is required to either advance or retard the start of combustion (depending on the main fuel). There has also been a focus at these institutions to determine the main cause behind why the syngas affects the combustion in the way it does, by kinetic modeling and investigating the thermal properties of the fuel-air mixture. University of Alberta, Canada The University of Alberta, Canada has done a significant amount of work on using syngas as a combustion modifier with high and low octane fuels. They have also done significant kinetic modeling using both single zone and multi-zone models. Kongsereeparp et al. (2008) began with a Chalmers University Mechanism for pure n- heptane using a genetic algorithm approach to significantly modify the base mechanism. The results showed that the base fuel replacement with a syngas mixture delays the SOC because of reduced HO 2 production, which is the main source of heat release during the cool flame reactions. This caused a lower temperature rise during the primary stage of combustion, which means there will be less radicals available during the cool flame stage, which also delays the main ignition. Kongsereeparp et al. (2008) also focused on looking at how the thermodynamic and chemical properties of syngas mixtures affect ignition and combustion properties. Thermodynamic properties affect temperature during compression while the chemical properties affect chemical reaction rates at any temperature. They used syngas mixtures of 75% H 2 and 25% CO, by volume. They found that the base fuel auto-ignition properties determine whether the syngas acts through a thermodynamic effect or a chemical effect. For high octane fuels that require high compression temperatures, the thermodynamic effect of increasing the mixture specific heat ratio with syngas mixture, hence increasing compression temperatures, causes an advance in SOC. However with low octane fuels, like heptane, which auto-ignite at low temperatures through a progression of cool-flame and other pre-ignition reactions, means that the syngas mixture 15

30 inhibits these cool-flame reactions, and hence slows down the temperature rise due to the chemical affect and retards SOC. For these low octane fuels, it was found that the thermodynamic effect of increased specific heat ratio is not important. Kongsereeparp et al. (2007) used a multi-zone model to predict the effect of syngas blends with natural gas and heptane on combustion. They found that adding syngas to high octane natural gas, advances combustion timing, but adding to heptane retards the timing and lengthens the combustion period. These changes were attributed to the increasing specific heat ratio with increasing syngas fraction, which increases compression temperatures and advances combustion for the high octane fuel conditions. On the other hand, changes in the fuel auto-ignition chemistry, which impacts the heptane combustion, because heptane exhibits a two-stage combustion event, was seen because the H 2 in the syngas blend reduces the radical pool build up in the primary ignition (cool flame) region, which reduces the temperature rise, causing a delay in the main ignition stage. The University of Alberta has also focused on experimental investigations of syngas blends in a CFR octane engine with a variable compression ratio. These tests focused on using natural gas, isooctane, and n-heptane as the main fuel and syngas blends as the secondary fuel to modify the combustion. Hossieni et al. (2009) focused on modeling and experiments by controlling heptane SOC with syngas blends. The blends tried were 75%/25% and 50%/50%, H 2 /CO. They postulated that the low temperature heat release (LTHR) is inhibited by a reduction of intermediate radical mole fractions during the cool flame reactions, and the early stages of the (NTC) delay period. These radicals included H 2 O 2 in which both the productive and destructive reaction rates increased with syngas blend addition, but the net result was a reduction of H 2 O 2 and a destruction of the highly reactive OH radicals. They also showed that the syngas blend fraction did not significantly affect the combustion duration. By replacing the heptane with some syngas, it was thought that the increased 16

31 specific heat ratio would cause an advance in LTHR start, but it did not significantly affect it. It did, however, reduce the overall amount of energy released during this region, which meant that the lost energy from the LTHR required to initiate main combustion stage had to be supplied by the compression process, thus delaying the SOC of the main combustion event. Hosseini et al. (2008) used three base fuels; n-heptane, iso-octane, and Natural gas, to investigate the effects of syngas blending for combustion control. The blends they used were 75%/25% and 50%/50%, H 2 /CO ratio. They found that the syngas blends could provide combustion timing control for all three fuels, even though the fuels had considerably different auto-ignition characteristics. For n-heptane, the syngas blends retarded the SOC and it was found that with increasing H 2 fraction, the SOC is further retarded. This was attributed to H 2 suppressing radical formation and increasing the OH consumption, between the 1 st and 2 nd stage combustion, leading to a delay in SOC. For natural gas, the SOC was advanced with syngas addition and it was reported that the SOC advance was proportional to the amount of H 2 added, rather than the quantity of syngas itself. With iso-octane, the syngas blends retarded ignition and the various blend compositions did not show any difference, meaning the H 2 proportion did not matter. Hosseini et al. (2007) investigated the effect of syngas on low octane fuels, with octane numbers from 0 to 20 in a CFR engine. A blend of 75% H 2 / 25% CO was used. Syngas addition expanded the operating range on the rich side and also retarded SOC timing, by altering the pre-flame chemistry. Also, by adding syngas, the peak pressure, pressure rise rate, and combustion duration were reduced. It was also reported that small amounts of syngas impacted SOC immediately; however, higher syngas mass fractions were needed to effect combustion duration. The smoother pressure rise and retarded combustion were attributed to the chemical kinetic effects diminishing the first stage reactions. Hosseini et al. (2007) investigated the effect of syngas on high octane PRF fuels with octane numbers ranging from 80 to 100. It was found that the syngas fraction (75% 17

32 H 2 /25% CO) retarded the combustion timing, and the HC and CO emissions increased, while the NO x emissions remained relatively constant. An intake temperature of 140 C, compression ratio of 16:1, and less than 30% EGR was used. Their work showed that a condition of higher power demand and later combustion can be achieved with a slight increase in syngas mass fraction, without significantly affecting the power output. University of Birmingham, England Researchers at the University of Birmingham have also investigated syngas mixtures in both dual fuel diesel engines and dual fuel HCCI engines. In the diesel studies, PM emissions were also investigated, along with combustion characteristics. Tsolakis et al. (2008) investigated both B50 (50% biodiesel + 50% diesel) diesel mode operation and ethanol HCCI operation with EGR and simulated reformed fuel EGR (REGR). Two conditions, 10% and 20% by volume of EGR and REGR were inducted into the intake manifold with the REGR made up of 24% of H 2, by volume. They found that the EGR and REGR HCCI conditions retarded the SOC. They attributed this to the displacement of hot trapped residuals within the cylinder by cool EGR. However, it was found that hydrogen addition (REGR case) assisted in the combustion process (even though the SOC was retarded) by reducing the pressure rise rate and allowing more EGR compared to the EGR only case. Yap et al. (2006) demonstrated natural gas HCCI with syngas blending. The syngas was produced from exhaust gas fuel reforming. The exhaust gas and fuel was fed into the reformer and a heater was used to increase the temperature of the reformer inlet to the same temperature as the exhaust valve outlet. It was found that proportions of 10 15% H 2 in the REGR mixture advanced the SOC by 4-6 CAD and also widened the operating region so lower loads could be achieved. 18

33 Wyszynski et al. (2004) investigated natural gas HCCI with simulated REGR, used to modify the combustion. They found that with hydrogen addition there is a decrease in the intake air temperature requirement, and the SOC also advances. Even with residual gas trapping and H 2 addition, some intake heating was still required, unlike gasoline HCCI. They used compression ratios from 12:1 to 14.5:1 and a H 2 proportion of 10 % of the REGR. Hokkaido University, Japan Researchers from Hokkaido University have investigated syngas HCCI with unconventional fuels, like di-methyl ether (DME). They ve performed both experimental tests and numerical simulations to determine how the syngas blends can affect the combustion characteristics of HCCI engine operation. Shudo et al. (2003) investigated the effect of methanol derived syngas addition to DME HCCI combustion. They performed both experimental and kinetic investigations. It was found that adding the hydrogen containing gas retards the auto-ignition of DME considerably. The results showed that the retarded ignition is attributed to the consumption of OH radicals by hydrogen during the low temperature oxidation of DME. They go onto discuss how the OH radical is a chain carrier in the low temperature oxidation, however, when H 2 and CO are introduced into the reaction, they consume the OH radicals through the H 2 + OH H 2 O + H and CO + OH CO 2 + H reactions. Shudo et al. (2006) investigated the effect of CO 2 on the HCCI combustion of DME with hydrogen addition for ignition control. This work focused on experimental data along with Chemkin kinetic modeling. It was found that CO 2 addition retards SOC timing, and also lengthens the combustion duration. The delay in SOC was attributed to the delayed temperature rise during the compression stroke and the low-temperature oxidation processes due to the decreased specific heat ratio, because the in-cylinder gas mean temperature for low and high temperature oxidation were hardly changed. The chemical 19

34 effect of CO 2 was investigated by kinetically modeling an inert CO 2 molecule with the same thermal properties, and there was found to be little difference. Shudo et al. (2004) investigated DME HCCI with methanol derived syngas blending. Their results showed that methanol derived syngas with a higher H 2 fraction had a greater effect on ignition control and the syngas addition retards the start of combustion. The cause of the delay was attributed to the increased CO 2 fraction with a higher specific heat, which lowers the in-cylinder temperature during the compression stroke. The delay was also attributed to the H 2 addition, which consumes the OH radicals formed during the low temperature oxidation period. 20

35 CHAPTER 3: THEORETICAL FOUNDATION 21

36 ENGINES Engine Geometry A critical engine geometrical parameter is the engine compression ratio (CR), defined in equation 3.1. (3.1) In equation 3.1, V d is the displacement volume and V cl is the clearance volume. The compression ratio is crucial, because if everything is held constant and the CR is increased, the thermodynamic cycle efficiency also increases (Heywood, 1988). To perform any heat release and hence combustion analysis, the cylinder volume must be known in crank-angle-space. The crank-slider formula, shown in equation 3.2, is used to calculate cylinder volume based on crankshaft position cos (3.2) In equation 3.2, R is the ratio of the connecting rod length to the crank radius (equivalent to the stroke divided by two). In actual engine testing these measurements are made using an optical encoder mounted on the crankshaft. The encoder is synchronized with the engine cycle using either a camshaft signal or a once-per-cycle signal and the data is acquired on a crank angle basis. Further review of these fundamentals is given in the textbook by Zhao et al. (2001). In-Cylinder Pressure The in-cylinder pressure measurements made from an engine allow the researcher to gather information on combustion duration, pressure rise rate, start of combustion, incylinder temperature, along with many other parameters. Figure 3.1 and 3.2 show 22

37 examples of an engine in-cylinder pressure trace and a pressure-volume diagram, respectively Pressure (bar) CAD (ATDC) Figure 3.1: Example of an in-cylinder pressure trace Pressure (bar) Volume (L) Figure 3.2: Example of an engine indicator diagram 23

38 From the indicator diagram and knowing that the compression process is polytropic in nature, the polytropic constant, n, can be calculated using the following equation (Sonntag and Borgnakke, 2000): (3.3) By plotting an ln (P) vs. ln (V) curve during the compression stroke, a straight line is typically observed. The slope of this line is the polytropic coefficient, n, which is used during the heat release analysis. This value is sometimes interchanged with the ratio of specific heats (gamma). Heat Release Rate Analysis The heat release rate (HRR) is calculated from the in-cylinder pressure data and specifically indicates how much energy is liberated by the fuel. In other words, it shows how much external heat would have to be supplied to the cylinder gas to raise the cylinder pressure to the measured amount. For HCCI engines, it is very reasonable to assume a single zone model, where the temperature increase and reaction zone throughout the cylinder volume is relatively the same. For SI engines, a two zone model that is typically applied. The following derivation can be found in many texts, including (Stone, 1999). To begin with the heat release calculation, the first law of thermodynamics shown earlier is modified by separating the heat transfer term into two components as follows: (3.4) where δq HR is the heat released by combustion and δq HT is the heat transferred to the walls. Substitutions must be made into the above equation to get du and dw in terms of parameters that we are measuring. To do this, the definition of internal energy, shown in equation 3.5, is used. 24

39 (3.5) where c v is the constant volume specific heat. Differentiating the ideal gas law gives: (3.6) Substituting equation 3.6 into equation 3.5 gives: (3.7) Substituting equation 3.7 into our initial first law, and knowing that dw = pdv we have: (3.8) This can be further reduced by realizing that R = c p c v and the specific heat ratio γ=c p /c v, which gives equation 3.9. For this work, a gamma value of 1.3 was assumed. (3.9) If the heat transfer to the walls is not known, and not estimated, then the net heat release rate can be calculated by combining the two heat transfer terms as follows: (3.10) The heat release analysis is now based on quantities that can be measured from the engine such as cylinder volume, pressure, and crankshaft position. 25