Legal forms

Effect of water blending on bioethanol HCCI combustion with forced induction and residual gas trapping

Description
Effect of water blending on bioethanol HCCI combustion with forced induction and residual gas trapping
Categories
Published
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Energy 32 (2007) 2396–2400 Effect of water blending on bioethanol HCCI combustionwith forced induction and residual gas trapping A. Megaritis a,  , D. Yap b , M.L. Wyszynski c a Mechanical Engineering, School of Engineering and Design, Brunel University, West London, Uxbridge UB8 3PH, UK  b Singapore Institute of Manufacturing Technology, 71 Nanyang Drive, Singapore 638075, Singapore c Mechanical and Manufacturing Engineering, School of Engineering, University of Birmingham, Birmingham B15 2TT, UK  Received 27 January 2007 Abstract There is increased interest worldwide in renewable engine fuels as well as in new combustion technologies. Bioethanol is one of thealternative fuels that have been used successfully in spark ignition engines. A combustion technology that currently attracts a lot of interest is the homogeneous charge compression ignition (HCCI) combustion, which has shown potential for low nitrogen oxidesemissions with no particulate matter formation. The authors have shown previously that applying forced induction to bioethanol HCCIwith residual gas trapping results in an extended load range compared to naturally aspirated operation. However, at high boostpressures, high cylinder pressure rise rates develop. Work by other researchers has shown that direct injection of water can be used as acombustion control method. The present work explores water blending as a way that might have an effect on combustion in order tolower the maximum pressure rise rates and further improve emissions. The obtained experimental results show that in contrast tovariable rate direct injection of water, fixed rate water–ethanol blending is counterproductive for the reduction of pressure rise rates athigher loads. In addition, increasing the water content in ethanol results in reduction of the effective load range and increased emissions. r 2007 Elsevier Ltd. All rights reserved. Keywords:  Bioethanol; HCCI; Fuel water blending 1. Introduction Homogeneous charge compression ignition (HCCI) alsoknown as controlled auto-ignition (CAI) combustion isthought to represent one of the next major steps in enginedevelopment for use in transportation vehicles. HCCIengines can have efficiencies close to these of diesel engines,with low levels of emissions of oxides of nitrogen (NO x )and particulate matter (PM). In addition, HCCI engineshave been shown to operate with a range of fuels, e.g.gasoline and natural gas [1,2]. There are still, however,important limitations preventing the full commercializa-tion of HCCI engines such as the limited engine loadoperating range and the challenging HCCI combustioncontrol.One key enabling technology allowing HCCI combus-tion on a moderate compression ratio (CR) engine withoutintake heating is residual gas trapping. It is a viable methodto raise the in-cylinder temperatures to levels required forHCCI operation. As the residual rates increase, the in-cylinder charge temperature also increases, allowing CRstypically found in gasoline engines to be used. In addition,the trapped exhaust gas acts as a diluent [3,4], which isrequired to prevent violent combustion and achievecombustion temperatures low enough so that NO x  forma-tion is dramatically reduced. However, engines employingresidual gas trapping have a limited load range comparedto spark ignition (SI) combustion operation.The requirements for dilution limit the maximum powerdensity of HCCI engines as violent combustion occurswhen the excess air ratio is reduced. As such, the maximumload achieved is dictated by the amount of air or EGR(exhaust gas recirculation) that can be inducted into theengine to provide dilution. Forced induction methods, such ARTICLE IN PRESS www.elsevier.com/locate/energy0360-5442/$-see front matter r 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.energy.2007.05.010  Corresponding author. Tel.: +441895266682; fax: +441895256392. E-mail address:  thanos.megaritis@brunel.ac.uk (A. Megaritis).  as supercharging, have been shown to be effective in raisingthe power density of HCCI engines [5].Bioethanol is considered by many as one of the mostimportant alternatives to gasoline and diesel as it can offersubstantial reductions in consumption of fossil fuels andemission of greenhouse gases. The authors have previouslyshown that it was possible to use bioethanol as a fuel forHCCI operation using a gasoline style engine in conjunc-tion with residual gas trapping (achieved by negative valveoverlap). The naturally aspirated engine operated in HCCImode up to 4.18bar indicated mean effective pressure(IMEP) with moderate intake heating [6] while when forcedinduction was used the IMEP range was extended to7.5bar [7]. However, as the engine load increases, the in-cylinder pressures and the maximum cylinder pressure riserates reach high levels, resulting in excessive combustionnoise.The injection of water, either into the intake manifold orthe cylinder has been shown to reduce NO x  emissions andincrease the HCCI upper boundary load limits. Christensenand Johansson [8] applied water injection into the intakemanifold alongside with different fuels (isooctane, ethanoland natural gas). It was shown that water injection wouldretard the ignition timing and slow down the rate of combustion. The allowable range of water mass flow ratewas no more than 50% of the main fuel mass flow rate forthe naturally aspirated cases and up to 3 times the mainfuel mass flow rate for the supercharged cases. As thecooling effect of the water slowed down the combustionrate, the already high emissions of unburned hydrocarbonsincreased when water injection was applied. The COemissions also increased, indicating that combustionquality was poorer with increasing water injection. OverallNO x  emissions were very low and the use of water injectioncould reduce NO x  to only fractions of a ppm. Taking thisconcept one step further, Iwashiro et al. [9] tested an in-cylinder water injection system in an HCCI engine fuelledwith DME and propane. The concept was to control knockby decreasing the in-cylinder temperature gradientsthrough a directed water vapour layer. They showed thatan increase in IMEP from 4.6 to 7bar could be achievedwith increasing water injection after which, NO x  emissionsrise rapidly. The water injection affects the low tempera-ture heat release, which then affects the spatial position of the main ignition. These researchers found that increasingengine loads require increasing amounts of water injectionalong with advanced injection timings to ensure sufficientevaporation of the water spray.The aim of the work presented here was to exploremethods of reducing the maximum pressure rise rates of bioethanol HCCI combustion with residual gas trappingand forced induction by using water addition. Water wasused in the form of blending with the fuel as it has beenapplied successfully in diesel engine combustion [10]. Waterblending was also considered worth testing because theremoval of water for making neat ethanol requires a largepart of the energy required in the production of ethanolappropriate for fuelling of standard SI and diesel engines.The limitations in the water content of ethanol for SI anddiesel combustion may, however, be more relaxed in thecase of HCCI combustion, which is a very different modeof combustion (e.g. no limitation for the fuel water contentdue to flame propagation requirements as in SI engines). 2. Experimental setup and procedure The experimental work was carried out using a singlecylinder engine with a Rover K series engine head. Thissingle cylinder engine actually forms a quarter of astandard K series 1.8l having been engineered to fit thesingle cylinder crankshaft with custom balance shafts. Forthe HCCI tests presented here, proprietary fixed durationlow lift camshafts were used and the CR was raised to thevalue of 12.5, using a racing style piston. A summary of theengine specifications is given in Table 1.The engine was coupled to a DC dynamometer that wasused to load and motor the engine. The engine torque wasrecorded by a calibrated load cell. In the present work, airfrom a standalone compressor was used for forcedinduction and no external EGR was applied. The engineload was controlled via the boost pressure in steps of 0.4bar and varying fuelling during operation at thatspecific boost pressure. The maximum pressure was limitedto 1.2bar gauge to keep the mechanical stresses below thesafety limit.Negative valve overlap was used to trap residual gasesand the engine throttle was kept wide open throughout thetests. The inlet valve maximum opening point (MOP) wasset at 144 crank angle degrees (CAD) after the exhauststroke top dead centre (TDC). The exhaust valve MOP wasset at 140 CAD before the exhaust stroke TDC. All thetests were conducted at 1500rpm engine speed.A Kistler 6125A pressure transducer (1% measurementaccuracy) was fitted flush with the wall of the combustionchamber, connected via a Kistler 5011 charge amplifier to aNational Instruments data acquisition board. The crank-shaft position was measured using a digital shaft encoderwith an accuracy of   7 1 1 . Software was developed in-house, in the LabVIEW programming environment, torecord the in-cylinder pressure versus crank angle for 100consecutive engine cycles, and to analyse the resulting data.Output from the analysis of consecutive engine cyclesincluded peak engine cylinder pressure, values of IMEP,percentage coefficient of variation (% COV) of IMEP,average values and percentage COV of peak cylinder ARTICLE IN PRESS Table 1Engine specifications summaryEngine type Medusa single cylinder 4-valve engineBore 80mmStroke 88.9mmCompression ratio 12.5Fuelling type Port-injected A. Megaritis et al. / Energy 32 (2007) 2396–2400  2397  pressures, average crank angle for ignition delay, burnduration, and 5 and 95% burn points.A Pierburg (AVL) portable exhaust gas analyser wasused to measure NO x  emissions (chemical method), carbondioxide (nondispersive infrared method, NDIR), carbonmonoxide (NDIR), unburned hydrocarbons (NDIR), aswell as the exhaust gas oxygen content (chemical method).The analyser was calibrated prior to the experimental testswith zero and span gases. The resolution of the analyser forthe NO x  emissions is 1ppm with an accuracy of  7 10ppmwhile for the other emissions the accuracy is 0.01%. Theexcess air ratio  l  (lambda) was also obtained from thePierburg analyser (the analyser provides direct readings of lambda as calculated from the exhaust gas analysismeasurements).Fuelling was via a standard injector located close to theinlet port of the engine. The intake temperature wasmeasured in the intake port approximately 70mm from theintake valve seats and was slightly elevated at 40 1 C, tominimize the effect of the inlet temperature on combustionphasing and to assist homogenization of the charge. Thefuel flow rates were determined by means of a balancemeasuring the amount of fuel consumed by the engine.Anhydrous bioethanol provided by Shell Global Solutions(UK) was used for all the tests. Tests were carried out withpure bioethanol and bioethanol–water blends with watercontents of 10% and 20%. 3. Results and discussion The maximum pressure rise rates versus IMEP for all thetested conditions and fuel-water blends are shown in thescatter plot in Fig. 1. For 10% water content in fuel, thepressure rise rates have a similar range with that of operation with fuel without water addition.For 20% water content in the fuel, the maximum rates of pressure rise for the load range achieved are largely above10bar/CAD. This creates excessive combustion noise andpossibly damage to the engine over prolonged periods of operation. Hence, it appears that increasing the watercontent in fuel is counterproductive for the reduction of pressure rise rates. It can also be noted that with wateraddition the effective load range is much reducedcompared to that of standard fuelling without water.Fig. 2 shows the maximum lambda achieved withincreasing water content in the fuel. There is a relativelysmall change with 10% water content. However, increasingthe water content to 20% decreases the maximum lambdaachieved substantially.It was expected that the water content would be able todecrease the combustion temperatures, thus possiblyslowing down the combustion. Instead it appears thatwater has an adverse effect on the benefits obtained fromresidual gas trapping. Water addition causes cooling downof the in-cylinder gas at inlet valve closure (IVC), resultingin lower in-cylinder temperatures at TDC. This retards thecombustion phasing for a given lambda and creates a needto increase fuelling in order to keep the combustion stable.As a result, lambdas decrease with larger amounts of water.Hence, the dilution levels are lowered resulting in highermaximum cylinder pressure rise rates.The effect of water addition on the excess air ratiorequired for stable combustion results in increased NO x emissions with increasing amounts of water, as shown inFig. 3. This is because the decreased dilution and increasedmaximum cylinder pressure rise rates (which occur withwater addition due to the required increased fuelling asdiscussed above) lead to higher NO x  emissions. Fig. 3 showsthat NO x  emissions increased rapidly with 20% watercontent in the fuel. However, the NO x  emissions were notadversely affected when 10% water in the fuel was used.Although the inlet temperature was kept slightly elevatedat 40 1 C to assist homogenization of the fuel, it wasinsufficient for the complete evaporation of the watercontent of the fuel–water blend. Upon entry of the fuel intothe cylinder and mixing with the hot exhaust gases, thelarge latent heat of vaporization resulted in a decrease of the mixture temperature at IVC. The reason for theobserved tolerance at the lower level of water contentmight be that at that level the evaporation was sufficientand/or that the heat loss was relatively low. ARTICLE IN PRESS 02468101214161804810    M  a  x   i  m  u  m   P  r  e  s  s  u  r  e   R   i  s  e   R  a   t  e   (   b  a  r   /   C   A   D   ) 20% water 10% water 0% water 2IMEP (bar)6 Fig. 1. Maximum cylinder pressure rise rates versus IMEP with differentfuel water contents. 00.511.522.505152025    L  a  m   b   d  a Water Content Fraction (%)10 Fig. 2. Maximum lambda achieved for various fuel water contentfractions at optimal combustion phasing. A. Megaritis et al. / Energy 32 (2007) 2396–2400 2398  It has to be emphasized that the method used here is incontrast to previous work where water has been used as amethod of combustion control using direct injection of varying amounts of water [9]. In the present study, thewater blended is in a fixed ratio to the fuel, which isintroduced in the port. Although this would have been amuch easier method to control combustion by wateraddition, it is clear from the results obtained in this studythat it does not have the desirable effects on bioethanolHCCI combustion control (cylinder pressure rise rates) andexhaust emissions.The present work has, however, demonstrated that withsufficient intake temperatures there is a small tolerance of the combustion process to water contained in the fuel,which can be present due to contamination. Higher intaketemperatures might increase the tolerance to water due toincreased evaporation. Pre-evaporation of the water–fuelblend before reaching the combustion chamber would beexpected to allow high concentrations of water in the fuelbecause in the case of HCCI combustion there are nolimitations related to slow flame propagation resultingfrom the presence of water. This can potentially lead tosubstantial energy savings by elimination of the waterremoval stages in the case of ethanol production specifi-cally for HCCI engines. Clearly the concept of usingethanol–water blends in HCCI engines is worth of furtherinvestigation despite the findings of the present experi-mental work.Finally, it is worth mentioning that although theindicated and brake specific fuel consumption (ISFC andBSFC, respectively) were determined from the acquiredexperimental measurements, they are not presented in thepaper since they are not representative of practical engineoperation and can be misleading. This is because, asmentioned earlier, the engine was boosted by means of anexternal air compressor. Hence, the obtained fuel con-sumption data are not indicative of real energy consump-tion and can lead to erroneous conclusions since the enginewas operated with varying levels of boosting (up to 1.2bargauge). The presented lambda values provide an indicationof the trends of fuel consumption with varying wateraddition to the fuel. 4. Conclusions This paper documents the effects of water blending in thefuel in a bid to reduce the pressure rise rates duringbioethanol HCCI combustion with forced induction andresidual gas trapping.Low concentrations of water in bioethanol appear tohave minimal effects on combustion. However, increasingthe water content to 20% drastically reduces the availableload range and lambda required for combustion.The water contained in the bioethanol–water blendresults in decreased in-cylinder temperatures during com-pression by reducing the gas temperature at IVC.This retards the combustion phasing for a given excessair ratio (lambda). Therefore, in order to maintain stablecombustion, the lambda must be decreased. This inturn lowers the dilution levels and results in substantiallyhigher maximum cylinder pressure rise rates and NO x emissions.Overall it has been shown that fixed rate water–ethanolblending is counterproductive for the reduction of pressurerise rates at higher loads and results in increased NO x emissions and reduced engine load operating range. Thereis, however, a small tolerance of the combustion process towater contained in the fuel, which can be present due tocontamination. Increased intake heating is expected toincrease the tolerance of HCCI combustion to watercontained in ethanol. References [1] Christensen M, Johansson B, Einewall P. Homogeneous chargecompression ignition (HCCI) using iso-octane, ethanol and naturalgas—a comparison with spark ignition operation. SAE TechnicalPaper No. 972824; 1997.[2] Christensen M, Hultqvist A, Johansson B. Demonstrating themultifuel capability of a homogenous charge compression ignitionengine with variable compression ratio. SAE Technical Paper No.1999-01-3679; 1999.[3] Zhao H, Peng Z, Williams J, Ladommatos N. Understanding theeffects of recycled burnt gases on the controlled autoignition (CAI)combustion in four-stroke gasoline engines. SAE Technical PaperNo. 2001-01-3607; 2001.[4] Oakley A, Zhao H, Ma T, Ladommatos N. Dilution effects on thecontrolled auto-ignition (CAI) combustion of hydrocarbon andalcohol fuels. SAE Technical Paper No. 2001-01-3606; 2001.[5] Christensen M, Johansson B, AmnJus P, Mauss F. Superchargedhomogenous charge compression ignition. SAE Technical Paper No.980787; 1998.[6] Yap D, Megaritis A, Wyszynski ML. An investigation intobioethanol homogeneous charge compression ignition (HCCI) engineoperation with residual gas trapping. Energy Fuels 2004;18(5):1315–23.[7] Yap D, Megaritis A. Applying forced induction to bioethanol HCCIoperation with residual gas trapping. Energy Fuels 2005;19(5):1812–21. ARTICLE IN PRESS 012345678905152025    N   O  x   (  g   /   b  r  a   k  e   k   W   h   ) Water Content Fraction (%)10 Fig. 3. NO x  emissions for increasing fuel water content fractions atoptimal combustion phasing. A. Megaritis et al. / Energy 32 (2007) 2396–2400  2399  [8] Christensen M, Johansson B. Homogeneous charge compressionignition with water injection. SAE Technical Paper No. 1999-01-0182;1999.[9] Iwashiro Y, Tsurushim T, Nishijima Y, Asaumi Y, Aoyagi Y. Fuelconsumption improvement and operation range expansion in HCCIby direct water injection. SAE Technical Paper No. 2002-01-0105;2002.[10] Samec N, Kegl B, Dibble RW. Numerical and experimental study of water/oil emulsified fuel combustion in a diesel engine. Fuel2002;81(16):2035–44. ARTICLE IN PRESS A. Megaritis et al. / Energy 32 (2007) 2396–2400 2400

Revisão 7 ano

Apr 16, 2018
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks