energies

Article Spray Formation of a Liquid Dioxide-Water Mixture at Elevated Pressures

Hakduck Kim, Changyeon Kim, Heechang Lim and Juhun Song *

School of Mechanical Engineering, Pusan National University, Busan 46241, Korea; [email protected] (H.K.); [email protected] (C.K.); [email protected] (H.L.) * Correspondence: [email protected]; Tel.: +82-51-510-7330; Fax: +82-51-512-5236

Academic Editor: Mehrdad Massoudi Received: 6 September 2016; Accepted: 8 November 2016; Published: 14 November 2016

Abstract: Liquid -assisted (LCO2-assisted) atomization can be used in coal-water slurry gasification plants to prevent the agglomeration of coal particles. It is essential to understand the atomization behavior of the water-LCO2 mixture leaving the injector nozzle under various conditions, including the CO2 blending ratio, injection pressure, and chamber pressure. In this study, the flash-atomization behavior of a water-LCO2 mixture was evaluated with regard to the spray angle and penetration length during a throttling process. The injector nozzle was mounted downstream of a high-pressure spray-visualization system. Based on the results, the optimal condition for the effective transport of coal particles was proposed.

Keywords: coal gasification; flash atomization; liquid carbon dioxide (LCO2); water-LCO2 mixture; solubility

1. Introduction Pressurized entrained flow gasification is a promising technology for the conversion of low-grade fuels to high-quality synthetic fuels, e.g., and synthetic natural gas. High gasification efficiencies using liquid fuels or slurries can only be achieved through the generation of a fine homogenous spray at a high chamber pressure [1]. During coal-water slurry (CWS) gasification, there is a tendency for the coal particles to agglomerate within individual droplets while the water is evaporating; the agglomerates are then dried out and undergo thermal decomposition in the flame. The resulting coal-particle size distribution (p.s.d.) is determined to a greater degree by the size distribution of the atomized fuel spray than by the initial particle size of the coal. Large particles formed through agglomeration have increased burnout times and produce large fly ash particles [2]. A potential method for obtaining finer spray droplets is CO2-assisted atomization. The effect of disruptive or flash atomization in reducing the p.s.d. of droplets was evaluated. The CO2 is dissolved into the fuel by injection into the line between the pump and the fuel nozzle. During the pressure release in the atomizing nozzle, the dissolved CO2 changes to the gaseous phase and disrupts the coal-water mixture droplets. Pure liquid CO2 (LCO2) has a different impact on the atomization at the injector nozzle when a LCO2 slurry is sprayed [3]. Compared with the conventional spray pattern of H2O, LCO2 produces effective breakup, leading to smaller liquid droplets, because of its lower viscosity and higher momentum flux. This helps to transport coal particles more uniformly inside the gasifier. This feature of better dispersion is attractive because it requires less heat release from partial combustion and thus less supply to the coal gasifiers. However, the effect of the blending ratio of LCO2 in the water on the atomization behavior has not been examined. In a study on the spray formation of pure LCO2 [4], CO2 was expanded through a nozzle from a high pressure to an atmospheric pressure. The flash atomization and breakup mechanisms of LCO2

Energies 2016, 9, 948; doi:10.3390/en9110948 www.mdpi.com/journal/energies Energies 2016, 9, 948 2 of 11 were examined using different liquid contents controlled by the degree of superheat. Under certain conditions, the LCO2 experienced a sudden pressure drop below the saturation pressure. LCO2 with differentEnergies levels 2016, 9,of 948 superheat produced different behaviors of flashing sprays and snow-particle2 of 11 formation. Lee et al. found that sprays in flash-boiling conditions were atomized by bubble growth, were examined using different liquid contents controlled by the degree of superheat. Under certain resulting in a smaller droplet size than that generated by aerodynamic force alone [5]. Liu et al. used conditions, the LCO2 experienced a sudden pressure drop below the saturation pressure. LCO2 with the laser-diffractiondifferent levels of method superheat to determine produced thedifferent p.s.d. behaviors of the spray of flashing [6]. sprays and snow-particle Engelmeierformation. Lee et et al. al. [7 ]found used that a pure sprays carbon in flash-boiling dioxide jet conditions for cutting were applications atomized by where bubble carbon growth, dioxide is compressedresulting in and a smaller expanded droplet via size a sapphirethan that generated nozzle. A by pressure aerodynamic chamber force isalone installed [5]. Liu downstream et al. used of the orificethe laser-diffraction to keep the post-expansion method to determine conditions the p.s.d. above of the thespray triple-point [6]. pressure of carbon dioxide. They demonstratedEngelmeier thatet al. carbon[7] used dioxidea pure carbon has to dioxide be expanded jet for cutting to an applications environment where of increasedcarbon dioxide pressure and thusis compressed increase theand liquid expanded fraction via a tosapphire improve nozzle. theimpact A pressure of the chamber jet. is installed downstream of the orifice to keep the post-expansion conditions above the triple-point pressure of carbon dioxide. Reverchon et al. reported that the solubilization of amount of supercritical CO in liquid solution They demonstrated that carbon dioxide has to be expanded to an environment of increased2 pressure could produce various powders with controllable size and distributions that are useful in several and thus increase the liquid fraction to improve the impact of the jet. fine-particle production processes [8]. Supercritical dissolved-gas atomization is an atomization Reverchon et al. reported that the solubilization of amount of supercritical CO2 in liquid solution processcould in whichproduce carbon various dioxide powders at with a temperature controllable and size pressure and distributions above its that critical are useful point in is several used as the atomizingfine-particle gas [9 ].production It relies on processes dissolved [8]. CO Supercriti2 gas emergingcal dissolved-gas from water atomization to form bubbles. is an atomization Therefore, this techniqueprocess applies in which only carbon to a dioxide limited at range a temperature of liquids and that pressure can hold above significant its critical quantitiespoint is used of as dissolved the gas. Asatomizing a result, gas the [9]. mean It relies droplet on dissolved diameter CO2was gas emerging mainly influenced from water byto form the suddenbubbles. releaseTherefore, of this the CO2 gas dissolvedtechnique inapplies the water. only to a limited range of liquids that can hold significant quantities of dissolved Ingas. general, As a result, the dissolvedthe mean droplet gas at diameter the exit was orifice mainly nucleates influenced to form by the gas sudden bubbles, release as inof effervescentthe CO2 gas dissolved in the water. atomization, and can rapidly evaporate by flashing as well [10]. The result shows that the mean droplet In general, the dissolved gas at the exit orifice nucleates to form gas bubbles, as in effervescent diameter is mainly influenced by the sudden release of the gas dissolved in the liquid. However, the atomization, and can rapidly evaporate by flashing as well [10]. The result shows that the mean behaviordroplet of thediameter water is spray mainly altered influenced by the by atomizationthe sudden release of dissolved of the gas LCO dissolved2 was not in experimentallythe liquid. evaluatedHowever, in our the previous behavior work.of the water That is,spray the al spraytered characteristicsby the atomization of theof dissolved water-LCO LCO2 mixture2 was not must be studied.experimentally evaluated in our previous work. That is, the spray characteristics of the water-LCO2 Inmixture this study,must be the studied. spray patterns of a water-LCO2 mixture were examined during a throttling process. TheIn this injector study, nozzle the spray was patterns mounted of downstream a water-LCO of2 mixture a high-pressure were examined flashing-spray during a throttling system, while the downstreamprocess. The pressureinjector nozzle was controlledwas mounted by downst a back-pressureream of a regulator.high-pressure A high-speedflashing-spray shadowgraph system, while the downstream pressure was controlled by a back-pressure regulator. A high-speed apparatus was used to evaluate the CO2-assisted atomization process. shadowgraph apparatus was used to evaluate the CO2-assisted atomization process. 2. Experimental Setup 2. Experimental Setup Figure1 presents a schematic of the high-pressure spray-visualization system used in this study. Figure 1 presents a schematic of the high-pressure spray-visualization system used in this study. It consistedIt consisted of an of upstreaman upstream chamber chamber (vessel), (vessel), aa middle block, block, and and a downstream a downstream chamber chamber (vessel). (vessel). ThereThere was was an opticalan optical quartz quartz window window on on thethe middlemiddle block block and and the the two two chambers, chambers, enabling enabling the the visualizationvisualization of the of flashing-spray the flashing-spray behavior behavior inside insi thede nozzlethe nozzle and and at the at exitthe exit of the of nozzle.the nozzle. The The window was designedwindow was to withstand designed to pressures withstand up pressures to 100 bar up at to a 100 peak bar temperature at a peak temperature of 80 ◦C, and of 80 a pressure-relief°C, and a valvepressure-relief was installed valve to prevent was installed pressure to prevent rises beyond pressure this rises limit. beyond this limit.

FigureFigure 1. 1.Schematic Schematic of experimental setup. setup.

Energies 2016, 9, 948 3 of 11 Energies 2016, 9, 948 3 of 11

AA saturated saturated mixture mixture of of CO CO22 waswas achieved achieved by by adjusting adjusting the the temperature temperature and and mass mass of of solid solid CO 22 (SCO(SCO22).). Different Different chamber chamber temperatures temperatures were were achieved achieved with with the the use use of of a a constant-temperature constant-temperature bath. bath. AsAs shown inin Figure Figure1, the1, the nozzle nozzle was was installed instal betweenled between the middle the middle block and block the and downstream the downstream chamber. chamber.It was 300 It mmwas long300 mm with long an internalwith an internal diameter diameter of 0.4 mm. of 0.4 Two mm. different Two different materials materials were used were for used the fornozzles. the nozzles. Some wereSome made were ofmade transparent of transparent Pyrex, Pyr enablingex, enabling the visualization the visualization of the of flow the pattern flow pattern inside insidethe nozzle. the nozzle. This was This useful was for useful examining for examining either capillary either actioncapillary or flash-boilingaction or flash-boiling behavior. However, behavior. a However, nozzle a copper was nozzle used for was purposes used for other purposes than internalother than visualization. internal visualization. A data-acquisition A data-acquisition system was systemutilized was to monitor utilized the totemperature monitor the and temperatur pressuree inside and thepressure chambers, inside and the a flowchambers, controller and unit a flow was controllerused to turn unit the was valves used on to andturn off the pneumatically. valves on and off pneumatically. Water–LCOWater–LCO22 mixturesmixtures with with different different percentages percentages LCO LCO22 werewere prepared prepared for for blending. blending. A A water water is is firstfirst filled filled in upper chamber uptoupto thethe desireddesired volume, volume, which which is is determined determined by by water water blending blending ratio ratio on onthe the basis basis of volume.of volume. Then, Then, a solid a solid CO CO2 is2 is poured poured with with desired desired volume volume percent. percent. As As a result,a result, the the CO CO2 in2 inliquid liquid and and vapor vapor phase phase is made is made with waterwith water as it reaches as it reaches ambient ambient temperature temperature at equilibrium. at equilibrium. Before the Beforetest, the the mixture test, isthe agitated mixture by stirringis agitated unit. by Therefore, stirring an unit. efficient, Therefore, continuous an efficient, solubilization continuous of CO2 solubilizationin the liquid solution of CO2 in is the allowed. liquid Thesolution liquid is COallowed.2 exists The as layerliquid above CO2 exists the water as layer layer, above indicating the water low layer,miscibility indicating of CO low2 with miscibility water. This of CO separation2 with water. is clearly This observedseparation in is the clearly inset observed of Figure 1in. Inthe addition, inset of Figureneither 1. droplets In addition, nor neither particles droplets of CO2 norare particles found in of the CO water.2 are found During in the this water. process, During the totalthis process, volume theof thetotal mixture volume remained of the mixture constant. remained The effect constant. of the The blending effect ratioof the was blending investigated ratio was at twoinvestigated different atpressures two different of the pressures upstream of and the downstream upstream and chambers: downstream 85/60 chambers: and 65/40 85/60 bar. and 65/40 gas bar. was Nitrogen used as gasa filler was when used aas pressure a filler when increase a pressure beyond increase the saturation beyond pressure the saturation was necessary. pressure was necessary. FigureFigure 2 illustratesillustrates a aschematic schematic p-T p-T diagram diagram of the of global the global multiphase multiphase behavior behavior of carbon of dioxide- carbon waterdioxide-water between between273 and 310 273 andK as 310can Kbe as determined can be determined from the from literature the literature data [11,12]. data [In11 ,the12]. figure, In the three-phasefigure, three-phase line connects line connects all p, T coordinates all p, T coordinates where tw whereo liquid two phases liquid (a phaseswater-rich (a water-rich liquid L1 and liquid a carbonL1 and dioxide-rich a carbon dioxide-rich liquid L2) liquid coexist L2) with coexist a carbon with adioxide-rich carbon dioxide-rich vapor V. vaporThe pressure V. The pressure at L1L2V at equilibriumL1L2V equilibrium is close to is tw closeo-phase to two-phase line, the liquid line, the and liquid vapor and curve vapor of pure curve CO of2, purebut always CO2, but below always the vaporbelow pressure the vapor of pressure pure CO of2.pure Because CO2 .of Because this similarity of this similarity of the three-phase of the three-phase line and linetwo-phase and two-phase line at testedline at ranges tested rangesof pressure of pressure and temperature and temperature (25 °C), (25 ◦pressureC), pressure release release from from 85 85to to60 60 bar bar has has higher higher presencepresence of of CO CO22-rich-rich liquid(L2) liquid(L2) than than CO CO22-rich-rich vapor(V) vapor(V) since since it it is is located located well well over over the the three-phase three-phase equilibriumequilibrium line. line. In In contrast, contrast, pressure pressure release release from from 65 65 to to 40 40 bar bar experiences experiences a a phase phase transition transition from from thethe CO CO22-rich-rich liquid(L2) liquid(L2) to to the the CO CO22-rich-rich vapor(V) vapor(V) phase. phase. The The detail detaileded properties properties of of the the two two pure pure components,components, water water and and LCO LCO2,2, areare listed listed in in Table Table 11..

Figure 2. Schematic of spray test condition mapping in P–T diagram of CO2-water mixture [12]. Figure 2. Schematic of spray test condition mapping in P–T diagram of CO2-water mixture [12].

Energies 2016, 9, 948 4 of 11

Table 1. Properties of CO2 and water used in the experiment.

Fluids Water CO2 Vapor pressure at 298 K (kPa) 3 6400 Density (mol/L) at 298 K and 64 bar 55.5 16.2 (liquid) Viscosity (mPa·s) at 298 K and 64 bar 0.89 0.057 (liquid)

The constant pressure of the lower chamber was attained using a backpressure regulator mounted downstream of the lower chamber. Without the regulator, the pressure in the lower chamber would have increased as the liquid flow filled the lower chamber, thereby compressing the gas. The dimensions of the high-pressure flashing-spray system are listed in Table2. The visualization region in the lower chamber was provided by an optical window with a diameter of 5 cm. It was desirable to fit most of the spray penetration into this visualization region for all the pressure conditions tested in this study. Thus, wetting of the chamber bottom by spray plumes should be prevented. For this reason, we selected a nozzle with a relatively high ratio of length to diameter, as previously mentioned.

Table 2. Dimensions of high-pressure flashing-spray system.

Parameters Description chamber shape Cylinder chamber diameter/length 5 cm/13.4 cm chamber volume 260 cm3 capillary nozzle inner diameter 0.4 mm capillary nozzle length 300 mm

A shadowgraph technique was employed to examine the flashing-spray patterns of the LCO2 and its mixture with water. During the throttling process across the nozzle, a high-speed camera was used to record consecutive images at a frame rate of 1/40,000 frames per second. Meanwhile, the pressure difference was measured to verify the downstream pressure controlled by the back-pressure regulator.

3. Results and Discussion

3.1. Effect of Water-Blending Ratio in Water-LCO2 Mixture In a previous work [3], spray patterns under three different levels of upstream pressure (pressure difference) were observed using a high-speed camera: 65/40, 75/50, and 85/60 bar. At 65/40 bar, the internal flashing due to bubble formation caused a wider spray angle and shallower penetration length among three sprays. This is sometimes called a “bowl spray” and involves a high level of flash-atomization, which is an effective breakup mechanism of the liquid column, due to a burst of bubbles. A phenomenon similar to the bowl spray was observed with an increasing degree of superheat by Lin et al. [4] and Park et al. [13]. In contrast, a narrower spray angle and deeper penetration were observed at 85/60 bar, where the flash boiling is absent because of the liquid abundance. This is considered a normal jet spray. This difference is caused by the higher velocity of the liquid flow inside the nozzle orifice. At the same pressure of 65/40 bar, the same nozzle length and diameter were used. The spray of H2O was not fully developed compared with that of LCO2. Such weak atomization is expected because of the low velocity and higher surface tension occurring when the H2O flows with a high viscosity inside the nozzle. This breakup mode is similar to the Rayleigh-type breakup mode observed in the experiment of Dechelette et al. [14], which is the main atomization zone of high-viscosity CWSs. However, the behavior of water sprays altered by the addition of LCO2 was not experimentally evaluated in our previous work. Energies 2016, 9, 948 5 of 11

Energies 2016, 9, 948 5 of 11 Figure3 shows the effect of the water fraction blended in the water-LCO 2 mixture on the spray pattern atFigure 65/40 3 bar. shows The the amount effect of of the water water in fraction the mixture blended was in increasedthe water-LCO while2 mixture the total on volumethe spray of the mixturepattern was at maintained. 65/40 bar. The Water amount contents of water of in 0 andthe mixture 100 vol was % correspond increased while to pure the total LCO volume2 and pure of the water, respectively.mixture was The maintained. spray patterns Water under contents both of conditions0 and 100 vol are % correspond similar to pastto pure observations, LCO2 and pure as water, previously mentioned.respectively. As the The water spray content patterns increased under both to conditions 75%, the sprayare similar angle to increasedpast observations, slightly, as from previously 8.5◦ to 9.8◦ and thenmentioned. to 12.3 As◦. Subsequently,the water content it decreasedincreased to to 75%, zero the when spray theangle water increased content slig reachedhtly, from 100%. 8.5° to This9.8° was and then to 12.3°. Subsequently, it decreased to zero when the water content reached 100%. This was because of the increased dissolution of LCO2 when water content in the mixture is high. The high because of the increased dissolution of LCO2 when water content in the mixture is high. The high content of dissolved CO2 can evolve gaseous CO2 and thereby create the flash or disruptive boiling content of dissolved CO2 can evolve gaseous CO2 and thereby create the flash or disruptive boiling duringduring a pressure a pressure release. release.

(a) (b)

(c) (d)

(e)

Figure 3. Effect of water fraction blended in water-LCO2 mixture on spray pattern at 65/40 bar: (a) 0%; ( b) 5%; (c) 25%; (d) 75%; and (e) 100% of water fraction. EnergiesEnergies 20162016,, 99,, 948948 66 ofof 1111

Figure 3. Effect of water fraction blended in water-LCO2 mixture on spray pattern at 65/40 bar: (a) 0%; EnergiesFigure2016, 93., 948Effect of water fraction blended in water-LCO2 mixture on spray pattern at 65/40 bar: (a) 0%;6 of 11 ((bb)) 5%;5%; ((cc)) 25%;25%; ((dd)) 75%;75%; andand ((ee)) 100%100% ofof waterwater fraction.fraction.

2 2 AA similarsimilar trendtrendtrend was waswas observed observedobserved when whenwhen dissolved dissolveddissolved CO COCO2 content2 contentcontent increased increasedincreased in theinin thethe water-CO water-COwater-CO2 mixture2 mixturemixture [2]. 2 [2].Better[2]. BetterBetter atomization atomizationatomization was waswas achieved achievedachieved by byby the thethe increased increasedincreased flash flashflash boiling boilingboiling of ofof CO COCO22 dissolved dissolveddissolved inin the the mixture.mixture. 2 TheThe solubility solubility ofof of COCO CO2 2 ininin waterwater water waswas was fixedfixed fixed onceonce once pressurepressure pressure andand and temperaturetemperature temperature areare aregivengiven given inin thethe in data.data. the data. ForFor ◦ example,Forexample, example, 2.52.5 molmol 2.5 %% mol isis obtainedobtained % is obtained forfor constantconstant for constant solubilitsolubilit solubilityyy atat 6565 barbar at andand 65 bar2525 °C°C and inin 25thethe predictedpredictedC in the predictedresultsresults ofof LarrynresultsLarryn ofW.W. Larryn DiamondDiamond W. Diamond etet al.al. [11].[11]. et WhenWhen al. [11 waterwater]. When makesmakes water upup makes 75%75% ofof up thethe 75% mixture,mixture, of the mixture, absoluteabsolute absolute amountsamounts amounts solublesoluble insolublein thethe waterwater in the becomebecome water higherhigher become comparedcompared higher compared toto thethe casecase to ofof the 25%25% case despitedespite of 25% havinghaving despite thethe having samesame solubility.solubility. the same Therefore, solubility.Therefore, 2 higherTherefore,higher solublesoluble higher COCO2 soluble amountamount CO contributecontribute2 amount toto contribute higherhigher flashflash to sprayspray higher asas flash waterwater spray contentcontent as increases waterincreases content inin thethe increases mixture.mixture. However,inHowever, the mixture. thisthis trendtrend However, isis notnot this consistentconsistent trend is withwith not the consistentthe resuresultslts withofof supercriticalsupercritical the results dissolved-gasdissolved-gas of supercritical atomizationatomization dissolved-gas [9].[9]. TheatomizationThe conecone half-anglehalf-angle [9]. The waswas cone reportedreported half-angle forfor aa was widewide reported rangerange ofof for gas-to-liquidgas-to-liquid a wide range ratiosratios of (GLRs). gas-to-liquid(GLRs). TheThe conecone ratios halfhalf (GLRs). angleangle ◦ ◦ isTheis approximatelyapproximately cone half angle 13°13° is atat approximately aa lowlow GLRGLR andand 13 decreasesdecreasesat a low GLR toto 4°4°and whenwhen decreases thethe GLRGLR to isis 4 increased.increased.when the ThisThis GLR isis is attributedattributed increased. toThisto differencesdifferences is attributed inin thethe to differencesexperiexperimentalmental in theapparatusapparatus experimental andand method.method. apparatus and method. FiguresFigures4 44and andand5 show 55 showshow the changethethe changechange in the inin spray thethe anglesprayspray and angleangle penetration andand penetrationpenetration length, respectively, length,length, respectively,respectively, according accordingtoaccording the water toto fraction. thethe waterwater The fraction.fraction. information TheThe informationinformation was derived waswas from derivedderived the images fromfrom ofthethe Figure imagesimages3. In ofof the FigureFigure spray 3.3. image, InIn thethe spraytwospray lines image,image, are drawn twotwo lineslines and are intersectedare drawndrawn and atand the intersectedintersected nozzle tip. atat The thethe angle nozzlenozzle between tip.tip. TheThe two angleangle lines between isbetween carefully twotwo measured lineslines isis carefullyforcarefully spray angle.measuredmeasured We identifiedforfor sprayspray horizontalangle.angle. WeWe lineidentifiedidentified of spray horizontalhorizontal tip normal lineline to of verticalof sprayspray line tiptip andnormalnormal measured toto verticalvertical distance lineline andfromand measuredmeasured nozzle tip distancedistance to this line. fromfrom This nozzlenozzle is used tiptip toto for thisthis penetration line.line. ThisThis isis length. usedused forfor The penetrationpenetration data at a water length.length. content TheThe datadata of zero atat aa waterarewater excluded contentcontent in ofof these zerozero areare plots. excludedexcluded As the inin water thesethese content plots.plots. AsAs increased, thethe waterwater the contentcontent spray increased,increased, angle and thethe penetration sprayspray angleangle length andand penetrationincreasedpenetration to lengthlength 75 wt %increasedincreased and then toto decreased. 7575 wtwt %% andand At then 75%then waterdecreased.decreased. blend AtAt condition, 75%75% waterwater higher blendblend penetration condition,condition, higher lengthhigher penetrationandpenetration angle were lengthlength observed, andand angleangle indicating werewere observed,observed, a higher indicatingindicating degree of a flasha higherhigher spray. degreedegree In the ofof caseflashflash of spray.spray. higher InIn flash thethe casecase spray, ofof higherdroplethigher flashflash size becomesspray,spray, dropletdroplet smaller sizesize and becomesbecomes more uniform. smallersmaller Therefore,andand moremore dropletuniform.uniform. evaporates Therefore,Therefore, quickly dropletdroplet and evaporatesevaporates particle wouldquicklyquickly agglomerate andand particleparticle less. wouldwould Based agglomerateagglomerate on this observation, less.less. BasedBased a onon water thisthis content observation,observation, of 75 wtaa waterwater % appears contentcontent to of beof 7575 optimal wtwt %% appearsforappears the effective toto bebe optimaloptimal transport forfor the ofthe coal effectiveeffective particles transporttransport by improving ofof coalcoal particlesparticles the spray byby quality. improvingimproving thethe sprayspray quality.quality.

FigureFigure 4.4. ChangeChange inin sprayspray angleangle withwith respectrespect toto waterwater fractionfraction atat 65/4065/4065/40 bar.bar.

Figure 5. Change in penetration length with respect to water fraction at 65/40 bar. Figure 5. Change in penetration length with respectrespect to water fraction at 65/40 bar. bar.

Energies 2016, 9, 948 7 of 11 Energies 2016, 9, 948 7 of 11 Energies 2016, 9, 948 7 of 11 3.2. Effect of Injection Pressure 3.2. Effect of Injection Pressure The spray characteristics were observedobserved forfor aa higherhigher injectioninjection pressurepressure ofof 85/6085/60 bar. Figure 66 The spray characteristics were observed for a higher injection pressure of 85/60 bar. Figure 6 shows the effect of the water-blending ratioratio ofof thethe water-COwater-CO22 mixture on the spray angle at 85/60 bar. bar. shows the effect of the water-blending ratio of the water-CO2 mixture on the spray angle at 85/60 bar. The variation of the spray angle with respect to thethe water fraction as the water content increased was The variation of the spray angle with respect to the water fraction as the water content increased was similar to that at 65/40 bar. bar. However, However, the angle was 15°–21°, 15◦–21◦, which which is far larger than that at 65/40 bar. bar. similar to that at 65/40 bar. However, the angle was 15°–21°, which is far larger than that at 65/40 bar. This can be explainedexplained byby thethe reducedreduced evolutionevolution ofof gaseousgaseous COCO22 inin thethe waterwater droplet.droplet. DissolvedDissolved COCO22 This can be explained by the reduced evolution of gaseous CO2 in the water droplet. Dissolved CO2 has a higher pressure than vapor pressure at the temperature of 25 °C◦C and is likely likely to to remain remain a a liquid. liquid. has a higher pressure than vapor pressure at the temperature of 25 °C and is likely to remain a liquid. Therefore, thethe addition ofof liquid CO22 provides less viscosity for the mixture than for water, where the Therefore, the addition of liquid CO2 provides less viscosity for the mixture than for water, where the velocity andand thus thus the the momentum momentum increase. increase. This This can increasecan increase the spray the spray angle despiteangle despite the lower the amount lower velocity and thus the momentum increase. This can increase the spray angle despite the lower ofamount flash boiling.of flash Theboiling. presence The presence of liquid of in theliquid jet wasin the also jet importantwas also important for the use for of carbonthe use dioxideof carbon in amount of flash boiling. The presence of liquid in the jet was also important for the use of carbon cuttingdioxide applicationsin cutting applications [7]. While [7]. jets While containing jets containing only gas only and gas liquid and showed liquid showed almost noalmost effect no on effect the dioxide in cutting applications [7]. While jets containing only gas and liquid showed almost no effect selectedon the selected samples, samples, the impact the impact increased increased dramatically dramatically with the with presence the presence of liquid. of liquid. on the selected samples, the impact increased dramatically with the presence of liquid.

Figure 6. Effect of water-blending ratio in water-CO2 mixture on spray angle at 85/60 bar. Figure 6. Effect of water-blending ratio in water-CO2 mixture on spray angle at 85/60 bar. Figure 6. Effect of water-blending ratio in water-CO2 mixture on spray angle at 85/60 bar. The images in Figure 7 compare the effects of the water fraction and the injection pressure on TheThe images images in in Figure Figure 77 comparecompare thethe effectseffects ofof thethe waterwater fractionfraction andand thethe injectioninjection pressurepressure onon the spray pattern of the water-CO2 mixture. They confirm that the spray angle increases with the thethe spray spray pattern pattern of of the water-CO2 mixture.mixture. They They confirm confirm that that the the spray spray angle angle increases increases with with the the injection pressure and has little dependence on the water-blending ratio. Compared with the 65/40- injectioninjection pressure andand hashas little little dependence dependence on on the th water-blendinge water-blending ratio. ratio. Compared Compared with with the the 65/40-bar 65/40- bar condition, less flash boiling was apparent at 85/60 bar. barcondition, condition, less less flash flash boiling boiling was was apparent apparent at 85/60 at 85/60 bar. bar.

(a) (b) (a) (b)

Figure 7. Cont.

Energies 2016, 9, 948 8 of 11 Energies 2016, 9, 948 8 of 11 Energies 2016, 9, 948 8 of 11

(c) (d) (c) (d) FigureFigure 7. Comparison ofof effectseffectsof of water water fraction fraction and and injection injection pressure pressure on on spray spray pattern pattern of water-CO of water- Figure 7. Comparison of effects of water fraction and injection pressure on spray pattern of water-2 COmixture:2 mixture: (a) 25% (a) 25% at 65/40 at 65/40 bar; bar; (b) 75%(b) 75% at 65/40 at 65/40 bar; bar; (c) 25%(c) 25% at 85/60at 85/60 bar; bar; and and (d )(d 75%) 75% at 85/60at 85/60 bar. bar CO2 mixture: (a) 25% at 65/40 bar; (b) 75% at 65/40 bar; (c) 25% at 85/60 bar; and (d) 75% at 85/60 bar Figure 8 shows the dependence of the distribution of the liquid-droplet size on the degree of FigureFigure8 8shows shows the the dependence dependence of theof the distribution distribution of the of liquid-dropletthe liquid-droplet size onsize the on degree the degree of flash of flash boiling. A conventional image-analysis technique was employed to obtain the droplet-size boiling.flash boiling. A conventional A conventional image-analysis image-analysis technique tec washnique employed was employed to obtain the to droplet-sizeobtain the distribution.droplet-size distribution. The results show that the mean diameter of the liquid spray at 65/40 bar was 150 μm. Thedistribution. results show The results that the show mean that diameter the mean of thediameter liquid of spray the liquid at 65/40 spray bar at was 65/40 150 barµm. was However, 150 μm. However, the higher pressure of 85/60 bar yielded a larger droplet of 300 μm with a larger deviation theHowever, higher the pressure higher of pressu 85/60re barof 85/60 yielded bar ayielded larger a droplet larger droplet of 300 µ ofm 300 with μm a with larger a larger deviation deviation from from the mean diameter. A similar trend in the droplet-size distribution was observed by Xiao et al. thefrom mean the mean diameter. diameter. A similar A similar trend trend in the in droplet-size the droplet-size distribution distribution was observedwas observed by Xiao by Xiao et al. et [15 al.], [15], who demonstrated that the addition of CO2 in diesel spray produced a far more uniform droplet who[15], demonstratedwho demonstrated that the that addition the addition of CO of2 inCO diesel2 in diesel spray spray produced produced a far a more far more uniform uniform droplet droplet size size because of the flash separation of CO2 from the liquid. The finer and more uniform droplet becausesize because of the of flash the separation flash separation of CO fromof CO the2 from liquid. the The liquid. finer andThe more finer uniform and more droplet uniform confirms droplet the confirms the strong flash atomization2 of the spray at 65/40 bar. This condition is favorable, as small strongconfirms flash the atomization strong flash of atomization the spray at of 65/40 the spray bar. This at 65/40 condition bar. This is favorable, condition as is small favorable, liquid dropletsas small liquid droplets may evaporate quickly and thereby prevent coal agglomeration when coal is present mayliquid evaporate droplets may quickly evaporate and thereby quickly prevent and thereby coal agglomeration prevent coal agglomeration when coal is presentwhen coal [2]. is The present size [2]. The size distribution was measured using the open source software ImageJ. The threshold was distribution[2]. The sizewas distribution measured was using measured the open using source the software open source ImageJ. software The threshold ImageJ. The was threshold appropriately was appropriately chosen to identify the droplet particle from the background. Then, a measuring tool chosenappropriately to identify chosen the to droplet identify particle the droplet from thepart background.icle from the Then,background a measuring. Then, toola measuring was used tool to was used to calculate the size of multiple droplets automatically. It is instructive to note the calculatewas used the to sizecalculate of multiple the size droplets of multiple automatically. dropletsIt automatically. is instructive toIt noteis instructive the limitations to note of thethe limitations of the shadowgraph technique. Smaller droplets scatter away much less light than large shadowgraphlimitations of technique.the shadowgraph Smaller technique. droplets scatter Smaller away droplets much scatter less light away than much large less droplets. light than Therefore, large droplets. Therefore, the smallest droplets in the spray are not detectable with current shadowgraph thedroplets. smallest Therefore, droplets the in thesmallest spray droplets are not detectable in the spray with are current not detectable shadowgraph with current technique. shadowgraph However, technique. However, this effect may not be significant when determining any difference in injection thistechnique. effect may However, not be this significant effect may when not determiningbe significant any when difference determining in injection any difference pressure in since injection it is pressure since it is present at both pressure conditions. presentpressure at since both it pressure is present conditions. at both pressure conditions.

(a) (b) (a) (b) Figure 8. Dependence of atomized liquid droplet distribution on degree of flash boiling at (a) 65/40 Figure 8. Dependence of atomized liquid droplet distribution on degree of flash boiling at (a) 65/40 andFigure (b) 8.85/60Dependence bar. of atomized liquid droplet distribution on degree of flash boiling at (a) 65/40 andand ((bb)) 85/6085/60 bar. bar. 3.3. Effect of Chamber Pressure 3.3. Effect of Chamber Pressure Figure 9 shows a series of spray events of LCO2-water mixture at different chamber pressures. Figure 9 shows a series of spray events of LCO2-water mixture at different chamber pressures. As the chamber pressure increased at the same injection pressure, the flow rate decreased, which As the chamber pressure increased at the same injection pressure, the flow rate decreased, which

Energies 2016, 9, 948 9 of 11

3.3. Effect of Chamber Pressure

EnergiesFigure 2016, 99 , shows948 a series of spray events of LCO 2-water mixture at different chamber pressures.9 of 11 As the chamber pressure increased at the same injection pressure, the flow rate decreased, which reducedreduced thethe sprayspray angleangle asas wellwell asas thethe penetrationpenetration length.length. ThisThis phenomenonphenomenon waswas observedobserved inin thethe visualizationvisualization resultsresults ofof JakobsJakobs etet al.al. [[1]1] andand JohnsonJohnson etet al.al. [[16].16]. TheThe sprayspray angleangle asas aa functionfunction ofof thethe chamberchamber pressurepressure hashas beenbeen measuredmeasured forfor valuesvalues betweenbetween 11 andand 2020 barbar [[1].1]. AnAn increase inin thethe chamber ◦ ◦ pressurepressure fromfrom 11 toto 2020 barbar resultsresults inin aa decreasedecrease inin thethe water-spraywater-spray angleangle fromfrom36 36°to to 2424°.. AnAn increasedincreased chamberchamber pressurepressure appearedappeared toto contractcontract thethe spray,spray, yieldingyielding aa moremore densedense cloudcloud ofof dropletsdroplets comparedcompared withwith thethe correspondingcorresponding atmosphericatmospheric case.case. ThisThis waswas explainedexplained byby thethe smallersmaller volumevolume ofof thethe entrainedentrained airair duedue toto thethe increasedincreased pressure.pressure. ThisThis tendencytendency matchesmatches thethe dieseldiesel sprayspray reportedreported inin [[17].17].

(a) (b)

(c) (d)

Figure 9. Series of spray events of LCO2-water mixture reported at different chamber pressures with Figure 9. Series of spray events of LCO2-water mixture reported at different chamber pressures with samesame injectioninjection pressurepressure (75(75 bar):bar): ((aa)) 7070 bar;bar; ((bb)) 6060 bar;bar; ((cc)) 5050 bar;bar; andand ((dd)) 3030 bar.bar.

Figure 10 shows the temporal variation of two different pressures between the chambers during Figure 10 shows the temporal variation of two different pressures between the chambers during the throttling process. One is the upstream pressure (injection pressure), and the other is the the throttling process. One is the upstream pressure (injection pressure), and the other is the downstream pressure (chamber pressure). The chamber pressure remained constant, indicating the downstream pressure (chamber pressure). The chamber pressure remained constant, indicating proper operation of the back-pressure regulator. The injection pressure in the upper chamber the proper operation of the back-pressure regulator. The injection pressure in the upper chamber experienced a continuous and smooth decay during the throttling process. experienced a continuous and smooth decay during the throttling process.

Energies 2016, 9, 948 10 of 11 Energies 2016, 9, 948 10 of 11 Energies 2016, 9, 948 10 of 11

Figure 10. Temporal variation of differential pressure between chambers during the throttling process Figure 10. TemporalTemporal variation of differential pressure between chambers during the throttling process at 75/70 bar. at 75/70 bar. bar.

Figure 11 plots the spray angle against the pressure difference. When the chamber pressure is Figure 11 plots the spray angle against the pressurepressure difference.difference. When the chamber pressure is high, the pressure difference is low, which indicates a reduced flow rate. The spray angle remains high, the pressure difference is low, whichwhich indicatesindicates aa reducedreduced flowflow rate.rate. The spray angle remains low at 13.5°. However, when the chamber pressure decreases, i.e., the pressure difference increases, low at 13.513.5°.◦. However, However, when the the chamber pressure pressure de decreases,creases, i.e., the pressure difference difference increases, increases, the angle is increased to 20° and afterwards it is not changed substantially. The latter is attributed to the angle is increased to 2020°◦ and afterwards it is not changed substantially. The latter is attributed to choke flow, where vapor is prevalent in the mixture and therefore the flow rate is limited. choke flow,flow, wherewhere vaporvapor isis prevalentprevalent inin thethe mixturemixture andand thereforetherefore thethe flowflow rate rate is is limited. limited.

Figure 11. Variation of spray angle with respect to pressure difference. Figure 11. VariationVariation of spray angle with respect to pressure difference. 4. Conclusions 4. Conclusions 4. Conclusions We investigated the spray characteristics of a mixture of LCO2 and water leaving an injector We investigated the spray characteristics of a mixture of LCO2 and water leaving an injector nozzle.We The investigated optimal amount the spray of CO characteristics2 in the mixture ofa was mixture reported of LCO for the2 and spray water angle leaving and length. an injector This nozzle. The optimal amount of CO2 in the mixture was reported for the spray angle and length. This nozzle.value was The determined optimal amount by the ofdifference CO2 in thein the mixture solubility was reportedof CO2 in for the the water-LCO spray angle2 mixture. and length. The value was determined by the difference in the solubility of CO2 in the water-LCO2 mixture. The Thissmaller value droplets was determined obtained at by65/40 the bar difference indicate in the the strong solubility flash ofatomization CO2 in the of water-LCO the spray compared2 mixture. smaller droplets obtained at 65/40 bar indicate the strong flash atomization of the spray compared Thewith smaller 85/60 bar. droplets Changes obtained in the at chamber 65/40 bar pressure indicate thehave strong no significant flash atomization influence of theon the spray spray compared angle, with 85/60 bar. Changes in the chamber pressure have no significant influence on the spray angle, withexcept 85/60 in the bar. case Changes of a high in chamber the chamber pressure. pressure have no significant influence on the spray angle, except in the case of a high chamber pressure. except in the case of a high chamber pressure. Acknowledgments: The authors thank the Korean government (MKE) for financial support from the manpower Acknowledgments: The authors thank the Korean government (MKE) for financial support from the manpower Acknowledgments:program (No. 20144010200780).The authors This thank work the Koreanwas supported government by a (MKE) National for financialResearch support Foundation from of the Korea manpower grant program (No. 20144010200780). This This work work was was supported by a National Research Foundation of Korea grant (NRF-2016M1A3A1A02005033) funded by the Korean government (MEST). (NRF-2016M1A3A1A02005033) fundedfunded byby thethe KoreanKorean governmentgovernment (MEST).(MEST). Author Contributions: Hakduck Kim and Changyeon Kim performed the spray experiment under the advice Author Contributions:Contributions: HakduckHakduck KimKim andandChangyeon Changyeon Kim Kim performed performed the the spray spray experiment experiment under under the the advice advice of Juhunof Juhun Song. Song. Heechang Heechang Lim Lim reviewed reviewed the the experimental experimental data data and and revised revised a a manuscript. manuscript. JuhunJuhun SongSong supervisedsupervised of Juhun Song. Heechang Lim reviewed the experimental data and revised a manuscript. Juhun Song supervised thethe researchresearch andand wrotewrote thethe manuscript.manuscript. the research and wrote the manuscript. ConflictsConflicts ofof Interest:Interest: TheThe authorsauthors declaredeclare nono conflictconflict ofof interest.interest. Conflicts of Interest: The authors declare no conflict of interest.

References References 1. Jakobs, T.; Djordjevic, N.; Fleck, S.; Mancini, M.; Weber, R.; Kolb, T. Gasification of high viscous slurry R&D 1. Jakobs, T.; Djordjevic, N.; Fleck, S.; Mancini, M.; Weber, R.; Kolb, T. Gasification of high viscous slurry R&D on atomization and numerical simulation. Appl. Energy 2012, 93, 449–456. on atomization and numerical simulation. Appl. Energy 2012, 93, 449–456.

Energies 2016, 9, 948 11 of 11

References

1. Jakobs, T.; Djordjevic, N.; Fleck, S.; Mancini, M.; Weber, R.; Kolb, T. Gasification of high viscous slurry R&D on atomization and numerical simulation. Appl. Energy 2012, 93, 449–456. 2. Yu, T.U.; Kang, S.W.; Beer, J.M.; Teare, J.D.; Sarofim, A.F. Fundamental Aspects of coal-Water Fuel Droplet Combustion and Secondary Atomization of Coal-Water Mixtures, Final Report; Cambridge: London, UK, 1987; Volume II. 3. Kim, K.W.; Kim, H.D.; Kim, C.Y.; Song, J.H. Flash spray characteristics of a coal-liquid carbon dioxide slurry. Korean J. Chem. Eng. 2016, 33, 1612–1619. [CrossRef] 4. Lin, T.C.; Shen, Y.J.; Wang, M.R. Effects of superheat on characteristics of flashing spray and snow particles produced by expanding liquid carbon dioxide. J. Aerosol Sci. 2013, 61, 27–35. [CrossRef] 5. Zeng, Y.; Lee, C.F.F. An atomization model for flash boiling sprays. Combust. Sci. Technol. 2001, 169, 45. [CrossRef] 6. Liu, Y.H.; Calvert, G.; Hare, C.; Ghadiri, M.; Matsusaka, S. Size measurement of particles produced from liquid carbon dioxide. J. Aerosol Sci. 2012, 48, 1. [CrossRef] 7. Engelmeier, L.; Pollak, S.; Kilzer, A.; Weidner, E. Liquid carbon dioxide jets for cutting applications. J. Supercrit. Fluids 2012, 69, 29–33. [CrossRef] 8. Reverchon, E. Supercritical-Assisted atomization to produce micro- and/or nanoparticles of controlled size and distribution. Ind. Eng. Chem. Res. 2002, 41, 2405–2411. [CrossRef]

9. Caputo, G.; Adami, R.; Reverchon, E. Analysis of dissolved-gas atomization: Supercritical CO2 dissolved in water. Ind. Eng. Chem. Res. 2010, 49, 9454–9461. [CrossRef] 10. Sher, E.; Elata, C. Spray formation from pressure cans by flashing. Ind. Eng. Chem. Process Des. Dev. 1977, 16, 237–242. [CrossRef] ◦ 11. Diamond, L.W.; Akinfiev, N.N. Solubility of CO2 in water from −1.5 to 100 C and from 0.1 to 100 MPa: Evaluation of literature data and thermodynamic modeling. Fluid Phase Equilibria 2003, 208, 265–290. [CrossRef] 12. Wendland, M.; Hasse, H.; Maurer, G. Experimental pressure-temperature data on three- and four-phase equilibria of fluid, hydrate, and ice phases in the system carbon dioxide-water. J. Chem. Eng. Data 1999, 44, 901–906. [CrossRef] 13. Park, S.; Lee, S.Y. An experimental investigation of the flash atomization mechanism. At. Sprays 1994, 4, 159–179. [CrossRef] 14. Dechelette, A.; Campanella, O.; Corvalan, C.; Sojka, P.E. An experimental investigation on the breakup of surfactant-laden non-Newtonian jets. Chem. Eng. Sci. 2011, 66, 6367–6374. [CrossRef]

15. Xiao, J.; Qiao, X.; Huang, Z.; Fang, J. Study of droplet size and velocity of fuel containing CO2 spray by means of PDA. Chin. Sci. Bull. 2004, 49, 1195–1199. [CrossRef] 16. Johnson, J.E.; Yoon, S.H.; Naber, J.D.; Lee, S.Y.; Hunter, G.; Truemner, R.; Harcombe, T. Characteristics of 3000 bar diesel spray injection under non-vaporizing and vaporizing conditions. In Proceedings of the 12th Triennial International Conference on Liquid Atomization and Spray Systems (ICLASS), Heidelberg, Germany, 2–6 September 2012. 17. Risberg, M.; Marklund, M. Visualizations of gas-assisted atomization of black liquor and syrup/water mixtures at elevated ambient pressures. At. Sprays 2009, 19, 957–967. [CrossRef]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).