ExperimentalInvestigationonHeatExchangeandSeparationPerformance of an Annular Structured Internal Heat-integrated Distillation Column☆

Lianghua Xu,Dawei Chen,Binghai Yan,Xigang Yuan*

Separation Science and Engineering

ExperimentalInvestigationonHeatExchangeandSeparationPerformance of an Annular Structured Internal Heat-integrated Distillation Column☆

Lianghua Xu,Dawei Chen,Binghai Yan,Xigang Yuan*

State Key Laboratory of Chemical Engineering,School of Chemical Engineering and Technology,Tianjin University,Tianjin 300072,China

A R T I C L EI N F O

Article history:

In this paper heat exchange coeff i cient and separation eff i ciency of an annular structured internal heat-integrated distillation column(HIDiC)were experimentally measured.About 50%heat of the inner column could be transferred to the outer column.The overall heat exchange coeff i cient decreased with an increase in pressure ratio of the inner column and the outer column,but was little affected by the F-factor.The increase of the pressure ratio decreased obviously the separation eff i ciency of the outer column but had little effect on that of the inner column.

?2014TheChemicalIndustry andEngineeringSocietyofChina,andChemicalIndustryPress.Allrightsreserved.

1.Introduction

Distillation is the most widely-used unit operation for separating mixtures.However,the traditional simple conf i guration of the distillation column usually has low energy eff i ciency[1,2].Many new technologies,such as thermally coupled distillation,heat pump distillation and multi-effect distillation,have been proposed to reduce the energy consumption of distillation[3-5].The internally heatintegrated distillation column(HIDiC)which combines the advantages of the low-pressure ratio heat pump distillation with a diabatic operation,isconsideredasthemostenergy-savingandcompactcolumn structure[6].

The typical conf i guration of HIDiC is illustrated in Fig.1.The rectifying and stripping sections of the HIDiC must operate at different pressures to make the temperature of a rectifying section higher than that of the other stripping section.To maintain the pressure difference between the two columns,a compressor and throttling valve must be installed between them.Therefore,the intermediate condensing heat of the rectifying section can be transferred to the stripping section for the intermediate vaporization through the column wall or other special heatexchangers[7].Theheatdutiesofthecondensersandreboilerswill correspondingly be reduced or even removed,thus saving a large amount of energy[8].

The f i rst lab-scale ideal-HIDiC was established in about 1997[9], which was a shell and tube conf i guration where the rectifying section was installed into the stripping section.A bench-scale HIDiC was introduced by Naito et al.in 2000[10],also with a shell and tube structure.An industrial scale HIDiC[11]reported in 2005 consisted of seven double-tube packed column units to increase the heat transfer area,and had an energy saving of more than 76%.

To further increase the heat exchange capacity,a sieve plate HIDiC fi tted with heat transfer panels was developed by Olujic et al.in 2006 [12].Its heat and mass transfer characteristics were investigated by de Rijke et al.[13].

Inrecentyears,manyinvestigations[14,15]abouttheprocessdesign and optimization of the HIDiC are conducted,but the fundamental characteristics of the heat exchange between the two column sections and the effect of heat exchange on the separation performance remain unclear.Owing to a shortage of reliable experimental data,the heat exchange coef fi cient used for designing a HIDiC was usually approximated by those for conventional heat exchangers[14].As a result,the expected energy saving cannot be achieved in practice because of the unrealistic heat exchange coef fi cient used.

In this paper,an annular structured internal heat-integrated distillation column was established.The overall heat exchange coef fi cient(U) between two column sections and the separation ef fi ciency of the distillation column under different operating conditions were measured.The experimental results are interpreted and the performances of heat exchange and separation are discussed.

2.Experimental

2.1.Experimental equipment

Schematic conf i guration of the concentric packed column is shown in Fig.2.The inner and outer columns were f i lled with Dixon packings, whose size is given in Table 1.Both columns were equipped correspondingly with a reboiler and a condenser and could be operatedunder total re fl ux at different pressures[16].The wall of the inner column was inclined to make the cross section area proportional to the vapor fl ow rate.The column wall of the outer column was vertical and surrounded by insulation material.The two column sections' parameters were shown in Table 1.

Fig.1.Schematic conf i guration of the HIDiC.

Fig.2.Schematic conf i guration of the concentric packed distillation column.

2.2.Operation procedure

The binary mixture of ethanol and water was adopted as the working system,and the mass fraction of ethanol was 20%.The two packed columns were pre-f l ooded at total ref l ux to make the packing fully wetted before operated steadily under atmospheric pressure.Then the pressure and temperature at all measuring positions were recorded and the overhead and bottom products were sampled for analysis with gas chromatography.

To transfer heat from the inner column to the outer column,the temperature of the whole inner column must be higher enough than that of the outer column.So the inner column was regulated to a pressure higher than that of the outer column at atmospheric pressure. For the inner column,f i rst the cooling water f l ow rate of the inner column condenser was gradually decreased,at the same time the heat duty of the inner column reboiler was kept constant,so the pressure of the inner column can increase gradually.In this process,for the outer column,the heat duty of the outer column reboiler was gradually decreasedandthe heatdutyof its condenser was kept constanttomake the column at normal pressure state,so that the temperature of the whole inner column was higher than that of the outer column and the heat exchange between the two columns could occur.The cooling water f l ow rate of the inner column and the heat duty of the reboiler in the outer column should be adjusted to perform heat exchange at different pressure ratios between the two columns.

2.3.Calculation

The heat transferred between the two columns can be estimated by the material and heat balances of the inner column assuming no heat loss:

where VDand LDare the mass f l ow rates of the overhead vapor and liquid ref l ux respectively,HVDand HLDrepresent their enthalpies.VBand LBare the mass f l ow rates of the bottom vapor and liquid respectively,and HVBand HLBare their enthalpies.

For total ref l ux operation,we have: The heat duties of the overhead condenser QCand bottom reboiler

QRof the inner column can be obtained by the following heat balances respectively:

Then,the heat transferred between the two columns is

The overall heat exchange coeff i cient(U)is given by

Table 1Parameters of the column

Fig.3.Heat duties of the inner column.

where A was the heat transfer area which was the inside area of the inner column wall and ΔTmwasthe logarithmic temperature difference between the two columns given by

In Eq.(7),TITand TIBare the overhead and bottom temperatures of the inner column respectively,and TOTand TOBrepresent the overhead and bottom temperatures of the outer column,respectively.

The heightequivalentof theoretical plate(HETP,Hetp)indicatingthe separation performance of the distillation column was calculated as

where H is the height of the packing bed and it is 1 m for both columns, and the pressure drops of both columns were negligible.N is the theoretical plate number of each column which was calculated by the McCabe-Thiele graphic method.The vapor-liquid equilibrium data of ethanol and water were calculated by the NRTL equation.

3.Results and Discussion

3.1.Heat exchange

The heat duties of condensers and reboilers of the two columns at different pressure ratios were shown in Figs.3 and 4 respectively, where PIand POwere referred to as the overhead pressures of the inner andouter columns.Theouter columnwasoperated under normal pressure.Therefore,the value of POwas 0.1 MPa,and the overhead pressure of the inner column PIcould be easily got from the value of pressure ratio.

In order to monitor the experimental uncertainty,three parallel experiments were carried out at each pressure ratio.As can be seen in Figs.3 and 4,although the three data points at each pressure ratio scattered a little along both coordinate axes,they were quite consistent in terms of the Q-(or U-)pressure ratio relationship.

Fig.3 shows that when the reboiler heat duty in the inner column was kept unchanged at different pressure ratios,the condenser heat duty was decreased with the increase of pressure ratio,and this is because the heat exchanged between the two columns increases with the increase of the pressure ratios.Similarly,as shown in Fig.4,when the condenser heat duty of the outer column was kept unchanged at different pressure ratios,its reboiler heat duty was decreased with the pressure ratio for the same reason.

The variation of the heat exchange between the two columns QTwith the pressure ratio is shown in Fig.5.It can be seen that,consistent withFigs.3and4,QTincreasedwiththepressureratio,andabout50%of heat of the inner column was transferred to the outer column when the pressureratiowasupto1.6,whichwas1.5timeslargerthantheoneofa similarsizewetted-walldistillationcolumn[17]forthesameseparating system.

Fig.4.Heat duties of the outer column.

Fig.5.Heat exchange load at different pressure ratios.

Fig.6.The change of ΔTmwith pressure ratio.

3.2.Overall heat exchange coeff i cient

AsshowninFig.6,themeantemperaturedifferenceΔTm,whichwas the driving force of heat transfer between the two columns,increased from about 6°C to 13°C with the pressure ratio change.The values of U estimated by Eq.(6),as shown in Fig.7,decreased from about1000 W·°C?1·m?2to about 780 W·°C?1·m?2with the increase of pressure ratios,which means that the heat exchange capacity of the column decreased with the increase of the pressure ratio.According to Fig.7,U can be correlated with the pressure ratio as

Fig.7.The change of U values with pressure ratio.

Fig.9.The separation eff i ciency of both columns at different pressure ratios.

The maximum U value was about 1000 W·°C?1·m?2which was very close to 1100 W·°C?1·m?2of the pilot scale HIDiC reported by Nakanishi et al.[18].Eq.(9),although correlated from our experiment, can be known as simple and representative for the tendency of change of U with respect to the pressure ratios or temperature differences in a HIDiC,which were also reported in the literature[17,19].de Rijke et al.[13]achieved the similar decrease tendency in the research about the heat integrated sieve distillation column,and they accounted for the phenomenon that the f i lm thickness of condensate liquid increased with temperature difference in the rectifying section.With regard to this column,as explained by Nakanishi et al.[18],the liquid condensed on the wall of the inner column can f l ow easily back into the packing,and some dry spots on the surface of the inner column wall could appear when the pressure ratios were increased.This may explain the decrease in U.

So it is inferred that the heat exchange ability of the HIDiC was limited.When the temperature difference between the two sections is very high,the distribution of the liquid and vapor f l ow in the column will be greatly affected,and the performance of heat exchange will deteriorate.So,it is not appropriate to increase the load of heat exchange by simply increasing the pressure ratio.A proper pressure ratio should be selected to achieve a good heat transfer between the inner and outer columns of HIDiC.

Totesttheinf l uenceofthevaporloadontheoverallheattransferfor the inner heat integration,experiments were carried out to measure U with two kinetic energy factors of vapor f l ow(F-factor)which was controlled by the heat duty of the reboiler.The results are shown in Fig.8.The U values at two levels of F-factor were almost the same. Such an indifference of the overall heat exchange coeff i cient to vapor and liquid load in a HIDiC was also reported by Kataoka et al.[20].

Fig.8.U values of two columns at different F-factors(low F-factor:F-factor= 0.57 kg0.5·m?0.5·s?0.5in the inner column,F-factor=0.23 kg0.5·m?0.5·s?0.5in the outer column;high F-factor:F-factor=0.72 kg0.5·m?0.5·s?0.5in the inner column,F-factor=0.34 kg0.5·m?0.5·s?0.5in the outer column).

3.3.Effect of heat integration on separation eff i ciency of distillation column

The experimental Hetpvalues of the two columns under different pressure ratios estimated by Eq.(8)are shown in Fig.9.The pressure ratio value of unity means no heat exchange between two columns. The separation eff i ciency of the inner column is shown to be much higherthantheoutercolumnnomatterwhetherthereisheatexchange or not,and increasesslightly withthepressure ratio.On theother hand, the separation eff i ciency of the outer column obviously decreased with the increase of pressure ratio as shown in Fig.9.The inner heat integration will reduce the irreversibility of the separation process and hence the mass transfer driving force.But as pointed out by Naito et al.[10], the effect of the decrease of driven force on the separation eff i ciency would be weak.As discussed above,the inclined wall of the inner column would make it easier for the condensed liquid on the wall in the inner column to irrigate back to the packing,and the eff i ciency of the inner column would be favored by this effect.The inclined wall could also favor the separation eff i ciency in the outer column because the liquid to be evaporated by the wall should be easier to be collected on the inclined wall.However the outer column suffered from stronger wall effect and decreased the separation eff i ciency of the column, because the former was in an annular structure with two column walls(inside and outside walls).The decreases of the separation eff i ciency in the outer column indicated that the wall effect was overwhelming in such an occasion.

Fig.10 shows the variations of Hetpof the two columns with the pressure ratios for different F-factors.It is seen that the separation eff i ciencies of two columns at high F-factors were slightly lower than those at low F-factors for the outer column,and the inf l uence of F-factor on the Hetpwas even smaller for the inner column.

4.Conclusions

A laboratory scale annular structured internal heat-integrated distillation column was constructed to investigate the heat transfer and separation performance of the internal heat-integrated distillation column.The experimental results were reported and discussed.

Fig.10.Hetpvalues of two columns with heat transfer at different F-factors(○Hetpof the inner column at F-factor=0.57 kg0.5·m?0.5·s?0.5;▽Hetpof the inner column at F-factor=0.72 kg0.5·m?0.5·s?0.5;□Hetpof the outer column at F-factor=0.23 kg0.5·m?0.5·s?0.5;Δ Hetpof the outer column at F-factor=0.34 kg0.5·m?0.5·s?0.5).

The overall heat exchange coeff i cient for the annular structured HIDiC was obtained and correlated with the pressure ratio.The experimental results demonstrated that,although the heat transferred in the internal heat integration increased with the pressure ratio,the overall heat exchange coeff i cient and the separation eff i ciency of the outer column decreased.This indicated that the ability of heat transfer could not be enhanced by increasing the pressure ratio,and a proper pressure ratio should be selected to achieve good heat transfer as well as separation eff i ciency for the HIDiC.The experiments showed that the F-factor had little effect on the overall heat exchange coeff i cient for the HIDiC studied.

[1]D.C.Freshwater,Thermal economy in distillation,Trans.Inst.Chem.Eng.29(1951) 149-160.

[2]D.A.Beneke,S.B.Kim,A.A.Linninger,Pinch point calculations and its implications on robust distillation design,Chin.J.Chem.Eng.19(2011)911-925.

[3]R.S.H.Mah,J.J.Nicholas,R.B.Wodnik,Distillation with secondary ref l ux and vaporization:a comparative evaluation,AICHE J.23(1977)651-657.

[4]Y.M.Wang,P.J.Yao,Simulation and optimization for thermally coupled distillation using artif i cial neural network and genetic algorithm,Chin.J.Chem.Eng.11(3) (2003)307-311.

[5]R.Agrawal,Z.T.Fidkowski,New thermally coupled schemes for ternary distillation, AICHE J.45(3)(1999)485-496.

[6]Z.Olujic,L.Sun,A.de Rijke,P.J.Jansens,Conceptual design of an internally heat integrated propylene-propane splitter,Energy 31(2006)3083-3096.

[7]M.Nakaiwa,K.Huang,A.Endo,T.Ohmori,T.Akiya,T.Takamatsu,Internally heat-integrated distillation columns:a review,Trans.Inst.Chem.Eng.81(2003) 162-177.

[8]A.de Rijke,Development of a Concentric Internally Heat Integrated Distillation Column,(Ph.D.Thesis)Technology University of Delft,Netherlands,2007.

[9]T.Nakanishi,T.Takamatsu,M.Nakaiwa,K.Aso,H.Noda,N.Kuratani,A case study of HIDiC design and energy saving,Comput.Chem.Eng.Symp.Ser.23(1999)855-858. [10]K.Naito,M.Nakaiwa,K.Huang,A.Endo,K.Aso,T.Nakanishi,T.Nakamura,H.Noda, T.Takamatsu,Operation of a bench-scale ideal heat integrated distillation column (HIDiC):an experimental study,Comp.Chem.Eng.24(2000)495-499.

[11]K.Horiuchi,K.Yanagimoto,K.Kataoka,M.Nakaiwa,Energy-saving characteristics of heat integrated distillation column technology applied to multi-component petroleum,IChemE Symp.Ser.152(2006)172-180.

[12]M.Gadalla,L.Jimenez,Z.Olujic,P.J.Jansens,A thermo-hydraulic approach to conceptual design of an internally heat-integrated distillation column(i-HIDiC), Comput.Chem.Eng.31(2007)1346-1354.

[13]A.de Rijke,W.Tesselaer,M.Gadalla,Z.Olujic,P.J.Jansens,Heat and mass transfer characteristics of an annular sieve tray,IChemE Symp.Ser.112(2006) 181-189.

[14]M.Gadalla,Internal heat integrated distillation columns(iHIDiCs)—new systematic design methodology,Chem.Eng.Res.Des.87(2009)1658-1666.

[15]B.Suphanit,Design of internally heat-integrated distillation column(HIDiC): uniform heat transfer area versus uniform heat distribution,Energy 35(2010) 1505-1514.

[16]P.Bai,S.A.Song,M.Sheng,X.F.Li,A dynamic modeling for cyclic total ref l ux batch distillation,Chin.J.Chem.Eng.18(2010)554-561.

[17]V.Lueprasitakul,S.Hasebe,I.Hashimoto,T.Takamatsu,Study of energy eff i ciency of a wetted-wall distillation column with internal heat integration,J.Chem.Eng.Jpn 23 (5)(1990)580-587.

[18]T.Nakanishi,N.Adachi,N.Nishida,K.Iwakabe,K.Horiuchi,M.Nakaiwa,Basic design of a double-pipe unit type internally heat integrated distillation column(HIDiC), Annual AIChE Meeting,Philadelphia,2008,pp.16-21.

[19]H.Noda,T.Mukaida,M.Kaneda,K.Kataoka,M.Nakaiwa,Internal column-tocolumn heat transfer characteristics for energy-saving distillation system,IChemE Symp.Ser.152(2006)737-744.

[20]K.K.Kataoka,H.N.Noda,H.Y.Yamaji,T.Mukaida,M.Kaneda,A compressor-free HIDiC system for recovery of waste solvent mixtures,8th World Congress of Chemical Engineering,Montreal,2009,(paper 512a).

6 January 2013

☆

Supported by the National Key Basic Research Program of China(2012CB720500).

*Corresponding author.

E-mail address:yuanxg@tju.edu.cn(X.Yuan).

http://dx.doi.org/10.1016/j.cjche.2013.05.002

1004-9541/?2014 The Chemical Industry and Engineering Society of China,and Chemical Industry Press.All rights reserved.

Received in revised form 17 April 2013

Accepted 20 May 2013

Available online 4 September 2014

Internal heat-integrated distillation column

Overall heat exchange coeff i cient

Separation eff i ciency