Energy-Efficient Design of an Ethyl Levulinate Reactive Distillation

School of Chemical Engineering, Yeungnam University, Dae-dong 712-749, Republic ... Publication Date (Web): May 4, 2017 ... Two designs, a neat design...
0 downloads 0 Views 664KB Size
Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

Energy Efficient Design of Ethyl Levulinate Reactive Distillation Process via Thermally Coupled with External Heat-Integrated Arrangement Felicia Januarlia Novita, Hao-Yeh Lee, and Moonyong Lee Ind. Eng. Chem. Res., Just Accepted Manuscript • Publication Date (Web): 04 May 2017 Downloaded from http://pubs.acs.org on May 4, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Energy Efficient Design of Ethyl Levulinate Reactive Distillation Process via Thermally Coupled with External Heat-Integrated Arrangement

Felicia Januarlia Novitaa, Hao-Yeh Leeb,* and Moonyong Leea,** a

School of Chemical Engineering, Yeungnam University, Dae-dong 712-749, Republic of

Korea b

Department of Chemical Engineering, National Taiwan University of Science and

Technology, Taipei 10607, Taiwan

Running title: Heat Integrated and Thermally Coupled Reactive Distillation for Ethyl Levulinate Production

Submitted to Industrial & Engineering Chemistry Research

*

Correspondence concerning this article should be addressed to:

Prof. Hao-Yeh Lee, Email: [email protected] Prof. Moonyong Lee, Email: [email protected]

1 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT This paper is leading to propose an energy efficient process design using a reactive distillation (RD) of the esterification reaction for the synthesis of ethyl levulinate (LAEE). Two designs, a neat design and an excess design, were examined to find the best configuration for the process. In the neat design, there are two columns with an equal molar feed to produce LAEE. On the other hand, an additional column is required in the excess design to separate LAEE from the unreacted reactant (Levulinic acid, LA). Later, this unreacted LA will be recycled back into the RD column. Compared to the neat design, the excess design showed superiority for the esterification process in terms of the TAC and energy requirement. Therefore, three energy integration configurations, external heatintegrated (HI), thermally coupled distillation (TCD), and combination of TCD and HI, were investigated in this excess design. A side-reboiler was implemented in the water removal column to overcome the limitation of the temperature difference for heat transferring. The simulation results showed that the configuration of TCD with HI was the best one, which saved 32.9% and 10.5% of energy and TAC, respectively, compared to the conventional RD three columns without HI.

Keywords: ethyl levulinate production, reactive distillation, thermally coupled distillation, process heat integration, energy efficient design

1.

INTRODUCTION Levulinic acid (LA) is a keto acid that can be produced by the acid-catalyzed

dehydration and the hydrolysis of hexose sugars.1 In 2012, BioMetics Inc. established the bio-refine process via dilute acid hydrolysis to produce LA from cellulosic feed-stocks, including paper mill waste, wood waste, and agricultural residues, in 50-70% yield.2-4

2 Environment ACS Paragon Plus

Page 2 of 33

Page 3 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

LA has many derivatives of ester levulinate (EL) which has been proposed for use as a fuel additive.5 Those ELs have their own characteristics.6 As seen in Table S1 (Supporting Information), in among various ELs, ethyl levulinate (LAEE) has a high blending octane number, (RON + MON)/2, of 107.5 and a high oxygen content (33 wt. %), making LAEE better than other ester levulinates for gasoline fuel additives.7 Grove et al.

8

reported the

properties of LAEE and other ester blends with diesel fuel. They found that LAEE could be separated from diesel fuel at temperatures below 0 °C. On the other hand, there is a limit to how much LAEE can be blended into diesel fuel due to its very low cetane number.9 Lake and Burton10 compared test mixtures of ultralow sulfur diesel (ULSD) fuel, biodiesel, and LAEE in a turbocharged engine with those of ULSD and 20 vol% biodiesel (B20). The results revealed that the LAEE blend has the lowest particulate matter (PM) emissions. The LAEE blend also eliminates the small increase in nitrogen oxides (NOx) that have been observed for B20 in many studies.11 LAEE is also used as a flavor additive in some foods and a fragrance additive in cosmetics.12 As a result of these applications, LAEE has attracted considerable attention as a valued chemical that positively impacts the market. As reported by Grand view research13, the demand for LAEE in the market is expected to reach USD 11.8 million by 2022. EL can be produced in two reaction steps. First, LA is dehydrated to Angelica lactone (AL), which is then mixed with an alcohol to form levulinate.14 On the other hand, the two reaction steps makes the process complicated and expensive. Therefore, another method, in which LA reacts directly with alcohol to form levulinate, was considered. Unfortunately, the conventional esterification process has an equilibrium limitation, which leads to high operating costs, due to the separation of unreacted reactants and products. To overcome those problems, reactive distillation (RD), where two important operations; reaction and separation, occur in a single unit, can be used for the esterification process.15 RD can be more

3 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

economical for the bio-based chemical manufacturing of esters than traditional reactorseparator systems because a higher reaction conversion can be achieved with lower energy consumption, which means that the RD column can overcome the equilibrium conversion limitations present in batch or flow reactors.12, 16 Thus far, more than 1000 papers on RD and approximately 236 reaction systems that can be built up with RD configurations have been reported.17 Sharma and Mahajani18 summarized more than 100 important industrial reactions and highlighted the importance of RD in the industry. Tung and Yu19 examined the design of quaternary RD processes based on the relative volatilities of the reactants and products. They concluded six configurations of the RD processes as provided in Table S2 (Supporting Information). From Table S2, Type I is for reactants or products with huge difference in volatility ranking. Type II denotes reactants or products with closer relative volatilities. If the relative volatility ranking is organized in alternating way for the reactants and products, it is classified as Type III. The esterification system in this study included four components and one azeotrope. Table S3 (Supporting Information) lists the boiling point ranking20 of the pure components and the azeotrope. The lightest and heaviest of the system are the reactants, EtOH and LA, respectively. Both products are in the middle ranking of those components. Therefore, the esterification of LA and EtOH process can be classified as Type IR. Sharma and Mahajani highlighted about the importance of RD for the industry in their book chapter18 and also mentioned low conversion of the reaction and high operating costs in the batch or flow reactor. Therefore, this study does not demonstrate the traditional reactorseparator systems for the LAEE production process. From the literature review of LAEE production process, no detailed modeling study of the synthesis of LAEE via the RD configuration has been reported. In addition, mostly the RD studies for heavy esters in the open literature are type IP or IIIP.16 However, the LAEE production process is categorized as

4 Environment ACS Paragon Plus

Page 4 of 33

Page 5 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

type IR. Therefore, the main contributions of this study are to launch the RD design for type IR process and to implement the RD column to product LAEE. Distillation is a popular unit used in separation processes and accounts for 40% to 70% of the costs of chemical plant.21 Therefore, several technologies have been used to reduce the energy requirements. One of the technologies that has attracted a great deal of attention is thermally coupled distillation (TCD). Approximately 30% of energy consumption can be saved using TCD compared to the conventional arrangement.22 Several studies have been carried out to combine a RD column and TCD23-27 to obtain more benefits in term of energy savings. Thermally coupled reactive distillation (TCRD) is more economically promising than the conventional RD process. However, the application of TCRD was often limited due to complexity in the design and control of the configuration. Based on the location of the interconnecting vapor-liquid streams, there are three types of TCRD for ternary separations; (1) TCRD side rectifier, where one reboiler is eliminated and changed with the interconnecting vapor liquid stream and the reactive zone is in the top part of the column, (2) TCRD side stripper, where one condenser is eliminated and changed it with the interconnecting vapor liquid stream and the reactive zone is in the bottom part of the column and (3) TCRD fully coupled, where one condenser and one reboiler are eliminated and changed it with the interconnecting vapor liquid stream in each part and the reactive zone is in the middle of the column. Both TCRD side rectifier and side stripper have a tendency to offer the most thermodynamically efficient designs more compared to the TCRD fully coupled.22,

24

External heat-integration (HI) is also another promising method to improve

energy efficiency, providing significant economic benefits from process intensification (PI). Using the heat of the overhead vapor stream in one column to the side- or bottom- reboiler of the other columns is a simple way to implement HI.24

5 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 33

Another main contribution of this study is implementation of the external HI and TCD into the best design of the LAEE production process to take additional advantages from TCD and HI structures. In this study, a commercial simulator, Aspen Plus, was used to conduct a rigorous simulation. The equilibrium stages were assumed in all column simulations. The total material, component, and energy balances were calculated in each column tray via the RadFrac module in Aspen Plus. For the optimal design of the LAEE production process, the designers’ domain knowledge and engineering trade-off were emphasized.

2.

REACTION KINETIC AND THERMODYNAMIC DATA

2.1. Reaction kinetic The esterification reaction of LA and ethanol (EtOH) in Eq. 1 is a reversible reaction that uses Amberlyst™ 39 as a catalyst. The kinetic data was correlated with the nonidealquasi-homogeneous model to determine the kinetic parameters. k1

C5H8O3 + C2H5OH ⇔ C7H12O3 + H2O

(1)

k2

(A)

(B)

(C)

(D)

where A, B, C, and D are LA, EtOH, LAEE, and water, respectively. The kinetic equation of the above reaction was obtained from Tsai.28 The rate expression of the quasi-homogeneous esterification kinetic model is

 a a  − rA = k1  a A aB − C D  Ka   Therefore, the equilibrium constant ( K a ) of the reversible reaction can be expressed as

6 Environment ACS Paragon Plus

(2)

Page 7 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Ka =

k1 k2

(3)

The unit of the kinetic equation (-rA) is kmol/m3.s, where ai is the liquid activity of each component. The reaction rate constants (k1 and k2) of the esterification reaction are expressed as a function of temperature as follows:

 − E0, f   k1 = A f exp   RT 

(4)

 − E 0 ,r k 2 = Ar exp   RT

(5)

  . 

In Eqs. 4 and 5, Af and Ar are the pre-exponential factors of the forward and backward reactions, respectively. E0,f and E0,r are the activation energies of the forward and backward reactions, respectively. The relationship between the reaction heat (∆hf) and activation energy is ∆ℎ = , − , .

( 6)

The kinetic rate of this esterification process can be obtained using the above equations and the kinetic model parameters in Table S4 (Supporting Information), where R is the universal gas constant (R = 8.314 kJ/kmol.K) and T (Kelvin) is the absolute temperature. Note that the units of the kinetic model parameters in Table S4 need to be converted to fit the requirements of the Aspen Plus simulator. The catalyst density used in this study was assumed to be 770 kg/m3.

2.2. Thermodynamic data

7 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 33

The studied system has four components including two reactants (LA and EtOH) and two products (LAEE and H2O). The non-random two liquid (NRTL) model was used to calculate the liquid activity coefficient (γi) of each component, as written in Eq. 7. nc

ln γ i =

∑τ

ji

G ji x j

j =1 nc

∑G

x

ki k

k =1

nc   τ mj Gmj xm  ∑ nc Gij x j  τ ij − m =1nc ; + ∑ nc  j =1 Gkj xk Gkj xk  ∑ ∑   k =1 k =1 

Gij = exp(− α ijτ ij ) , τ ij = aij +

bij T

+ eij ln T + f ij T ,

α ij = cij + d ij (T − 273.15K ) , τ ii = 0 , Gii = 1

(7)

The NRTL model requires a set of parameters for each binary component system. Since this process has four components, six sets of binary parameters are needed for this quaternary system. The water and ethanol pairing parameters were obtained from the Aspen Plus built-in database, and the remaining parameters refer to the regressed result from Resk’s paper.12 The Hayden-O’Connell (HOC) model

27

was used to consider the association in the

vapor phase. The HOC equation consistently manages the solvation of polar compounds and dimerization of the vapor phase in mixtures containing carboxylic acids.20 Tables S5 and S6 (Supporting Information) list the regressed NRTL parameters and HOC parameters, respectively, used in data regression for the four related components.

3.

PROCESS DESIGN AND OPTIMIZATION There are two important design approaches that can be considered when

implementing RD processes: a neat design and an excess design. A design is called a neat design if reactants (A and B components) have an equal molar feed to satisfy the reaction stoichiometry. In contrast, in the excess design, there is a surplus of one of the reactants (for

8 Environment ACS Paragon Plus

Page 9 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

example: component A) that is fed along with the second fresh feed (component B). As a consequence, an additional column is required in the excess design to separate the unreacted reactant from the products. The amount of excess reactant will be from the amount of recycle. Reactant B is the “limiting reactant” in the RD column and its conversion is high. The conversion of reactant A in the RD column is not high. In the excess design, the excess ratio of the two reactants is an important design variable. The reaction conversion can be improved using a higher excess ratio. This also decreases the number of reactive stages required in the RD column under the same LAEE production rate. On the other hand, the energy requirements will be higher. In contrast, a larger number of reactive stages in the RD column are required for a low excess ratio but the recovery of residual reactant has lower energy consumption. Therefore, there is a tradeoff between those variables. Both neat and excess designs were compared in the following sections.

3.1. Neat design of two columns First, following the approach of Lin29, a system with a low chemical equilibrium constant (Ka) is favored with an excess-reactant design. Therefore, a neat design was implemented because this esterification system has a high Ka value (Ka = 32.18 at 353.5 K) from the reaction kinetics section. Therefore, if the neat design is to be applied into this system, a heavy reactant, LA, is entered into the reflux drum of the RD column and a light reactant, EtOH, is fed into the 2nd stage, which is the optimal feed location for EtOH. The RD column is operated under total reflux. The entire upper section of this column is a reactive section. Rectifying trays are not needed in the RD column because the light reactant, EtOH, goes into the upper section of the RD column, which is the low temperature zone, and the upper section of the column should serve as reactive trays. The reaction takes place in the RD column, and the bottom stream is then fed into the next distillation column to separate two

9 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 33

products. The reaction conversion in the RD column was set to 95.10% so both products (H2O and LAEE) could meet the 95 mol% specification in the next column. Following the design procedure, Figure 1 shows the result of optimal neat design with a stoichiometric feed ratio of EtOH and LA. This neat design was optimized using an iterative optimization procedure for the five main design variables optimized: (1) number of reactive stages in the RD column (Nrxn,RD); (2) number of stripping stages in the RD column (Ns,RD); (3) EtOH feed location; (4) number of stages in C1 (NC1); and (5) C1 feed stage (NF,C1). Figure 2 presents the sensitivity plots of those variables.

TC = 79.8 °C QC = -18.42 Gcal/hr F = 100 kmol/hr X = [0,0,0,1] T = 285.6 °C P = 2 atm F = 100 kmol/hr X = [1,0,0,0] T = 97.1 °C P = 2 atm

F = 1903.08 kmol/hr X= [0.71,0.24,0.03,0.02] T = 79.8 °C P = 1 atm

TC = 90.5 °C QC = -1.73 Gcal/hr

RR = 0.75

F = 200 kmol/hr X= [0.025,0.475,0.475,0.025] T = 100.5 °C P = 1 atm

TR = 100.5 °C QR = 17.17 Gcal/ hr

TR = 205.3 °C QR = 2.42 Gcal/hr

F = 100 kmol/hr X = [0.05,0.95,0,0] T = 90.5 °C P = 1 atm

F = 100 kmol/hr X = [0,0,0.95,0.05] T = 205.3 °C P = 1 atm

Figure 1. Flowsheet of the optimal neat design case, X denotes [XEtOH, XH2O, XLAEE,XLA] in mole fraction.

10 Environment ACS Paragon Plus

Page 11 of 33

5.62

Ns = 12, EtOH feed = 67

Nrx = 81, EtOH feed = 67

5.60

6

TAC (10 $)

6

TAC (10 $)

5.60

5.58

5.58

5.56

5.56

80

81

82

10

83

11

12

13

Number of stripping stages in RDC (-)

Number of reactive stages in RDC (-)

5.382

5.40

Nrx = 81, Ns = 12

NC1 = 11 NC1 = 12 NC1 = 13

6

TAC (10 $)

5.380

6

TAC (10 $)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

NC1 = 14

5.38

5.378

5.36 1

2

3

EtOH feed location

8

9

10

11

12

13

C1 feed stage

Figure 2. Sensitivity plots of the optimization variables for neat design case.

In the design procedure, three assumptions were made to design the RD column: (1) the downcomer occupies 10% tray area; (2) the liquid holdup is set as half-full of catalyst Amberlyst™ 39 in each reactive stage (the catalyst density was assumed to be 770 kg/m3); (3) The weir height is 0.1 m giving a reactive liquid holdup of 0.373 m3. A reflux drum is on the top of the reactive section, which is the maximum kinetic holdup taken to be 20 times of the tray holdup. For column operating pressure, it is well known that at high pressure, a slight envelope in the XY diagram for VLE is not favorable30 because it will increase the cost or duty for separation. Moreover, higher pressure means higher temperature in the column and higher

11 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

utility costs due to hot oil or high pressure steam for heating required. In contrast, operating under vacuum makes the separation easier. However, it requires additional, costly equipment to draw the vacuum and recover material drawn into the vacuum system, which means increases the complexity of the flowsheet.31 A greater cross-sectional area in the column is also needed to accommodate the high volumetric flow of material when a column is operating under vacuum condition. In addition, for a LAEE production process, there is an upper limit on the operating temperature due to the Amberlyst™ 39 that used as a catalyst. The maximum operating temperature of the reactive section in the RD column is 130 °C. Based on these design consideratins, the operating pressure of the columns was chosen at normal pressure (at 1 atm) as the most preferable pressure condition. The temperature of the reaction section in RD column could be kept under 130 °C (See Fig. 8a) and cooling water could be used at normal pressure (at 1 atm). The RD column itself has 94 stages consisting of 81 reactive stages and 12 stripping stages, not including the reboiler. Increasing the reactive stages would increase the reactive holdup, requiring a further decrease in reboiler duty. On the other hand, extra stages require the additional capital cost and catalyst cost. A total of 81 reactive stages were found as a trade-off among the above advantages and disadvantages. The reaction conversion in the RD column was set to 95.1% by adjusting the reboiler duty. Figure 3 shows the liquid composition profiles in the RD column. The composition profile is quite flat until stage 87. This suggests that the temperature change is not significant in this part. Another observation from the composition profile is the remixing effect that the concentration of H2O increases as it approaches the bottom of the column and turns to decrease between stages 93 and 94. The remixing of H2O shows the waste of energy in the column. As a result, a very high reboiler duty (17.17 Gcal/hr) is needed in the RD column to produce LAEE.

12 Environment ACS Paragon Plus

Page 12 of 33

Page 13 of 33

The bottom outlet of the RD column goes into the next column, C1, to separate LAEE and H2O. The separation ability of the C1 column will improve when the total number of stages in C1 increases, which means that the energy requirement will decrease; however, the capital cost increases significantly. From Figure 2, the C1 column has 13 stages and the 11th stage is the optimal feed location. LAEE was obtained at 95 mol% from the bottom of the C1 column and H2O as a byproduct was obtained at 95 mol% from the overhead of the C1 column. The reboiler duty of the C1 column was adjusted to achieve the desired LAEE purity. The reboiler duty of C1 was 2.42 Gcal/hr. As a result, the total duties of the neat design were 19.59 Gcal/hr with the optimal TAC of $ 5.36 x 106.

0 .9 0 .8 0 .7

Mole fraction

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

0 .6 0 .5

E tO H H 2O

0 .4

LAEE LA

0 .3 0 .2 0 .1 0 .0 6

12 18 24 30 36 42 48 54 60 66 72 78 84 90 96

S ta g e

Figure 3. Liquid composition profiles of the RD column for the neat design case.

3.2. Excess design of three columns As reported in the previous section, a high TAC and total duty are required to synthesize LAEE with the neat design, even the purity product is only 95 mol% for LAEE and byproduct, H2O, respectively. This means that the first approach from Lin29 is hardly applied in this esterification system. Another approach from Lin29, which states that the

13 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

system with the lightest reactant, type IR, tends to prefer the excess design, was examined because the esterification process in this study is categorized as type IR. Another reason to implement the excess design is that the composition profile in the RD column is poor using the neat design. As shown in Figure 3, a poor reactant composition distribution in the reactive zone results in a high concentration of EtOH but a low concentration of LA. Recycling the LA reactant (HHK) by adding more LA reactant to the RD column can improve the reactant distribution in the RD column in such a way that the reaction goes towards to the production direction to increase the mole fraction of LA in the RD column. A higher LAEE can be produced due to the higher reaction conversion. From an analysis of the composition profile by Tung and Yu19, if reactant B (HHK) is chosen to be in excess, the reactive zone should be placed in the upper section of the RD column, where the concentration of reactant A (LLK) is higher. Figure 4 shows the optimal flowsheet of the excess design, in which EtOH is considered as a limiting reactant. The overall flowsheet includes a RD column and two other columns. Two fresh feeds (EtOH and LA) are all entered into the reflux drum of the RD column under the total reflux operation. All of the upper section of this column was a reactive section. All products and excess LA leave the bottom of the RD column to the subsequent columns for further separation. Further separation prefers the direct sequence rather than the indirect sequence because of less energy for same separation task. In the indirect sequence, the HHK component will separate from the bottom stream of the first separation column while the upper stream will contain the LK and HK components, which requires considerable energy to achieve this separation task. Moreover, it still needs a great deal of energy to separate the LK and HK components in the last separation column. Negligible EtOH in the feed stream to the C1 column with no azeotrope in the ternary system of water, LAEE, and LA leads to easy separation. The light product (H2O) of 99.5 mol% leaves the first distillation column (C1) as a distillate. The heavy product (LAEE) of 99.5 mol% is obtained from the

14 Environment ACS Paragon Plus

Page 14 of 33

Page 15 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

distillate of the C2 column, whereas the excess reactant (LA) of 24.7 kmol/hr is withdrawn from the bottom of the same column and recycled back to the RD column. The reboiler duties of the C1 and C2 columns were varied to achieve the purity specifications of the products, H2O, and LAEE, while the distillate flowrate of the C1 column and the bottom flowrate of the C2 column were fixed. Total annual cost (TAC) analysis was used to find the optimal design. Eq. 8 expresses the TAC, which consists of the annual operating cost (AOC) and the annualized total capital cost (TCC), with a payback period (n) of 3 years. The calculation is based on Douglas32, ignoring the costs of piping and pumps.

= +

 

(8)

Figure 4. Flowsheet of the optimal excess design, X denotes [XEtOH, XH2O, XLAEE,XLA] in mole fraction.

15 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Seven variables were optimized based on their apparent competing effect on the TAC of the process: (1) number of reactive stages in the RD column (Nrxn,RD), (2) number of stripping stages in the RD column (Ns,RD), (3) number of stages in C1(NF,C1), (4) C1 feed stage (NC1), (5) number of stages in C2 (NF,C2), (6) C2 feed stage (NC2), and (7) recycle flowrate. Their feasible search boundaries were determined from an extensive simulation. In this design flowsheet, all reactants should be fed to the RD column from the top. Three assumptions were made to design the RD column: (1) the downcomer occupies the 10% tray area; (2) the liquid holdup is set as half-full of catalyst Amberlyst™ 39 (assumed density, 770 kg/m3) in each reactive stage; and (3) the weir height was 0.1 m, resulting in a reactive liquid holdup of 0.069 m3. Same as in the neat design case, the top of the reactive section for the excess design was a reflux drum, in which the maximum kinetic holdup is taken to be 20 times the tray holdup.19

The optimal conditions were found using an iterative optimization procedure that are explained as follow: 1.

Fix a number of stripping section (Ns,RD).

2. Guess the stage numbers in the reactive trays (Nrxn,RD). 3. Go back to 2 and change Nrxn,RD until TAC is minimized. 4. Go back to 1 and vary Ns,RD until TAC is minimized. 5. Fix a number of C1 column (NC1). 6. Guess the C1 feed location (NF,C1). 7. Go back to 6 and change NF,C1 until TAC is minimized. 8. Go back to 5 and vary NC1 until TAC is minimized. 9. Fix a number of C2 column (NC2). 10. Guess the C2 feed location (NF,C2).

16 Environment ACS Paragon Plus

Page 16 of 33

Page 17 of 33

11. Go back to 10 and change NF,C2 until TAC is minimized. 12. Go back to 9 and vary NC2 until TAC is minimized. 13. Vary the LA recycle flowrate until TAC is minimized.

2.306

2.310 Nrx = 44

2.304

6

TAC (10 $)

6

TAC (10 $)

Ns = 10

2.302

2.304

2.298

43

44

45

6

46

2.48

2.40

NC1 = 10 NC1 = 11

8

9

NC2 = 17 NC2 = 18

2.38

NC1 = 12

2.44

7

Number of stripping stages in RDC (-)

Number of reactive stages in RDC (-)

NC2 = 19

TAC (10 $)

6

2.40

2.36

6

TAC (10 $)

NC1 = 13

2.36

2.34 2.32

2.32

2.30 2.28 7

8

9

10

11

12

11.0

11.5

2.32

6

2.31

2.30

2.29 24

25

12.0

C2 feed stage

C1 feed stage

TAC (10 $)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

26

27

Recycle flowrate (kmol/hr)

Figure 5. Results of optimization for excess design case.

17 Environment ACS Paragon Plus

12.5

13.0

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5 shows the result of the flowsheet optimization. A total of 44 reactive stages gave the lowest TAC. Increasing this number would add reactive holdup, which would further reduce the reboiler duty required. On the other hand, more stages mean more capital and catalyst cost. A total of 44 reactive stages were found as a trade-off among the above advantages and disadvantages. The separation ability of the columns will improve when the number of total stages of both C1 and C2 increase. This means that the energy requirement will decrease while the capital cost increases. A larger recycle flowrate suggests a higher reactant concentration for the excess reactant and decreases the reactive stages required in the RD column, resulting in a higher recycle cost and energy requirement. In contrast, a smaller recycle flowrate indicates a lower reactant concentration for the excess reactant, which increases the number of reactive stages required in the RD column while it reduces the recycle cost and energy requirement. The optimal flowsheet shown in Figure 4 was finally obtained from TAC analysis. As a result, the reactive section of 44 stages and the stripping section of 7 stages was obtained in the RD column (not including the reboiler stage). The number of stages in the C1 column were 11 stages and the feed stage of the C1 column was located at the 10th stage. The number of stages in the C2 column were 18 stages and the optimal feed stage was found at the 12th stage. The optimal excess ratio was 1.247 with a TAC of $ 2.29x106 and a total reboiler duty of 7.77 Gcal/hr. Figures 6 and 7 show the liquid composition and temperature profiles of the RD, C1, and C2 columns in the optimal excess design, respectively. The composition profile in Figure 7a indicates that the composition of the products improves almost 1.2 and 2 times for H2O and LAEE, respectively, compared to the neat design (from 0.25 to 0.3 mole fraction for H2O, and from 0.05 to 0.1 mole fraction for LAEE in the 1st stage of RD column). The higher product purity with lower energy requirements in the excess design case results mainly from

18 Environment ACS Paragon Plus

Page 18 of 33

Page 19 of 33

the improved reactant distribution in the RD column. Of more importance, the TAC of the excess design was 57.3% lower than that of the neat design. 1.0

0.8

0.6 0.4 0.2 0.0 4

0.8

EtOH H2O

Mole fraction

LAEE LA

Mole fraction

Mole fraction

1.0

1.0 EtOH H2O

0.8

LAEE LA

0.6 0.4 0.2

EtOH H2O

0.4

LAEE LA

0.2 0.0

0.0

8 12 16 20 24 28 32 36 40 44 48 52

0.6

1

2

3

4

5

6

7

Stage

Stage

(a)

(b)

8

9

2

10 11

4

6

8

10

12

14

16

18

Stage

(c)

Figure 6. Liquid composition profiles of the RD (a), C1 (b), and C2 (c) columns in the optimal excess design.

260 Temperature (C)

200

Temperature (C)

100 96 92 88

Temperature (C)

Temperature (C)

180

Temperature (C)

104

Temperature (C)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

160 140 120

240

220

200

100

84 4

8 12 16 20 24 28 32 36 40 44 48 52

Stage

(a)

1

2

3

4

5

6

7

8

9

10 11

Stage

(b)

2

4

6

8

10

12

14

16

18

Stage

(c)

Figure 7. Temperature profiles of the RD (a), C1 (b), and C2 (c) columns in the optimal excess design.

The optimization procedure above can be applied to other processes. Although it does not guarantee the global optimal solution, the main advantage of this procedure is that the designers can analyze the sensitivity of the design variables and the production phenomena based on their knowledge.

3.3.Comparison of the neat design and excess design 19 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table 1 lists the design results of the neat and excess designs. Although there is one additional column in the excess design, the total duty and TAC of the excess design ($ 2.29 x 106) are significantly lower than in the neat design ($ 5.36 x 106). The achievable product purity is also higher using the excess design (99.5 mol% of LAEE and H2O, respectively). Consequently, the excess design was chosen as the best design for this LA and EtOH esterification process. This design was adopted for a further comparison with other process configuration candidates in terms of energy efficiency.

Table 1. Comparison of the design results for the neat and excess designs Neat design Excess design H2 O

0.950

0.995

LAEE

0.950

0.995

RD

17.17

3.68

C1

2.42

2.37

C2

-

1.72

19.59

7.77

RD

$ 2.65 x 106

$ 7.43 x 105

C1

$ 3.76 x 105

$ 2.39 x 105

C2

-

$ 3.86 x 105

RD

$ 2.04 x 106

$ 4.40 x 105

C1

$ 2.92 x 105

$ 2.77 x 105

C2

-

$ 2.04 x 105

$ 5.36 x 106

$ 2.29 x 106

Product purity (mole fraction)

Duty (Gcal/hr)

Total duty (Gcal/hr)

Capital cost

Operating cost

TAC

20 Environment ACS Paragon Plus

Page 20 of 33

Page 21 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

4.

ENERGY INTEGRATION ARRANGEMENT OF THE EXCESS LAEE RD PROCESS

4.1. External HI case One of most common approaches for HI between the adjacent columns is to utilize the latent heat of the overhead vapor stream of one column as the reboiler heat source of the other. For this type of HI, the temperature of the overhead vapor stream in one column should be sufficiently higher than that of the bottom section in another column. A double effect configuration and a heat pump-assisted configuration are two typical options to secure the temperature difference required for this type of HI. On the other hand, these configurations require either an adjustment of the pressure of one of the columns or compressing the overhead stream of one column, which might restrict the economics and implementation.24 In this study, HI using a heat stream and/or a side-reboiler, which requires neither a heat pump nor pressure change, was explored. The minimum approach temperature (∆Tmin) chosen for HI was 10 °C. As shown in Figures 7a and 7c, the reboiler temperature of the RD column is approximately 105.0 °C, whereas the overhead temperature of the C2 column was approximately 199.5 °C. This highlights the potential for HI without pressure up of C2 or heat pump.

21 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 33

Figure 8. Heat integration between the RD and C2 columns.

Figure 8 shows the HI between the RD and C2 columns by implementing a heat stream. In this HI, the latent heat of the overhead stream in the C2 column was transferred to the bottom reboiler in the RD column. As a result, the duty of the bottom reboiler in the RD column could be reduced from 3.68 Gcal/hr to 1.98 Gcal/hr. This HI saved 8.8% and 21.8% of TAC and total energy requirement of the RD and C2 columns, respectively, compared to the conventional excess design without HI. The heat stream here is illustrated as a model of a heat exchanger. Therefore, the area of the heat exchanger can be calculated using the following equation: Q = A.U.∆T

(9)

where Q is the duty (Gcal/hr), U is the overall heat transfer coefficient (kW/m2K), A is the area of the heat exchanger (m2), and ∆T is the temperature difference between the C2 column condenser and the RD reboiler. The duty of this heat stream was 1.71 Gcal/hr (1984.6 kW), the temperature difference was 94.5 K, and U (overall heat transfer coefficient) was taken from Luyben33 as 0.568 kW/m2K. As a result, the required heat exchange area of the heat exchanger was calculated to be 36.98 m2.

22 Environment ACS Paragon Plus

Page 23 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

4.2. Thermally coupled case A thermally coupled distillation (TCD) configuration has gained increasing popularity as an effective way to reduce the energy consumption by reducing the remixing effect in a distillation process22. As indicated in Figure 6a, an apparent remixing effect mostly occurs in the stripping section of the RD column. A partial TCD between the RD and C1 columns was investigated to improve the energy efficiency by reducing the remixing effect in the stripping section of the RD column. A partial TCD between the C1 and C2 columns was not considered due to the slight remixing effect in the C1 column. Figure 9 shows the optimal TCD configuration between the RD and C1 columns. The TCD configuration was implemented in Aspen Plus using an equivalent TCD configuration with a side rectifier.

Figure 9. Optimal flowsheet of the TCD between the RD and C1 columns, X denotes the mole fraction [XEtOH, XH2O, XLAEE, XLA].

To focus on the energy saving effect according to the configuration, the main structures (total number of stages and feed location) of the columns in the TCD were the same as those in the conventional configuration. In this way, all energy savings are credited

23 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

to the thermal-coupling, reflux ratio of RD column, and recycle flowrate. Figure 10 shows the optimization results of those three design variables. The optimal reflux ratio of the RD column was 1.701 and the optimal recycle flowrate was 30 kmol/hr. The reboiler duties of the C1 and C2 columns were 5.31 Gcal/hr and 1.73 Gcal/hr, respectively, under the optimal conditions of the vapor side stream of 261 kmol/hr. As a result, the total duties of the RD and C1 columns reduced by approximately 0.73 Gcal/hr by implementing thermal coupling between the RD and C1 columns.

9.0

9.0

8.8

Recycle Recycle Recycle Recycle

8.6 8.4

flow = flow = flow = flow =

26 28 30 32

kmol/hr kmol/hr kmol/hr kmol/hr

8.2 8.0 7.8 7.6 7.4 7.2 7.0 1.48 1.52 1.56 1.60 1.64 1.68 1.72 1.76

Reflux ratio

Total reboiler duty (Gcal/hr)

Total reboiler duty (Gcal/hr)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 33

8.5

Recycle flow = 26 kmol/hr Recycle flow = 28 kmol/hr Recycle flow = 30 kmol/hr Recycle flow = 32 kmol/hr

8.0

7.5

7.0

6.5 248

252

256

260

264

268

Vapor side stream (kmol/hr)

(a)

(b)

Figure 10. Effect of the reflux ratio of the RD column (a), and vapor side stream (b) for each recycle flowrate on TAC.

Figures 11 and 12 show the liquid composition and temperature profiles of the RD, C1, and C2 columns in the proposed TCD, respectively. The composition at the top section of an RD column profile was similar to the conventional RD design. On the other hand, the remixing effect in the bottom section of the RD column is reduced; 9.39% energy consumption was saved in the TCD configuration with a total energy consumption of 7.04 Gcal/hr. Around 2.0% of TAC could be saved with this configuration compared to the conventional excess design without HI.

24 Environment ACS Paragon Plus

Page 25 of 33

EtOH H2O

0.6

LAEE LA

0.8

0.5 0.4 0.3 0.2

0.8

EtOH H2O

Mole fraction

0.7

Mole fraction

Mole fraction

1.0

1.0

0.8

LAEE LA

0.6 0.4 0.2

EtOH H2O

0.6

LAEE LA

0.4 0.2

0.1

0.0

0.0

0.0 4

1

8 12 16 20 24 28 32 36 40 44 48 52

2

3

4

5

6

7

Stage

Stage

(a)

(b)

8

9

2

10 11

4

6

8

10

12

14

16

18

Stage

(c)

Figure 11. Liquid composition profiles of the RD (a), C1 (b), and C2 (c) columns in the optimal TCD.

255

104 Temperature (C)

Temperature (C)

Temperature (C)

200

96

92

88

Temperature (C)

100

Temperature (C)

Temperature (C)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

180 160 140 120 100

4

8 12 16 20 24 28 32 36 40 44 48 52

Stage

(a)

240

225

210

195

1

2

3

4

5

6

Stage

7

8

9

10 11

(b)

2

4

6

8

10

12

14

16

18

Stage

(c)

Figure 12. Temperature profiles of the RD (a), C1 (b), and C2 (c) columns in the optimal TCD.

4.3. TCD with HI configuration As shown in Figure 9, the latent heat of the overhead of the C2 column was large enough at approximately 1.70 Gcal/hr. From Figures 12b and 12c, although the bottom temperature of the C1 column is approximately 9.1 °C higher than that of the top temperature of C2 column, a sharp decrease in temperature of the rectifying section in the C1 column results a low enough temperature to absorb the latent heat of the overhead of the C2 column through the simple implementation of a side-reboiler. In this configuration, the 10th stage was

25 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

found to be the best location for the side-reboiler implementation to give the minimum energy requirement under the required ∆Tmin constraint. Figure 13 presents the optimal flowsheet of the proposed TCD with HI via a side-reboiler.

Figure 13. Optimal flowsheet of the proposed TCD with HI via a side-reboiler, X denotes the mole fraction [XEtOH, XH2O, XLAEE, XLA].

From the HI method, the bottom reboiler duty of the C1 column could be reduced from 5.31 Gcal/hr to 3.42 Gcal/hr. The overall energy requirement of the proposed TCD with HI case was 5.21 Gcal/hr under the optimal conditions of a vapor side stream of 273 kmol/hr. The net energy saving by the proposed configuration was 2.56 Gcal/hr, which is equivalent to 32.9% energy savings compared to the conventional case. Note that the net saving is higher than the simple sum of those by the TCD configuration (0.73 Gcal/hr) and the HI configuration (1.69 Gcal/hr) due to the difference in recycle flowrates. Each recycle flowrate was found to be the optimal recycled flowrate of each configuration. According to Eq. 9, if a side-reboiler duty is illustrated as a heat exchanger then the required heat exchange area of the heat exchanger is 64.20 m2. Figure 14 presents the optimization variables of the proposed TCD with HI case. The optimal reflux ratio of the RD column was 1.558 and the optimal 26 Environment ACS Paragon Plus

Page 26 of 33

Page 27 of 33

recycle flowrate was 30 kmol/hr. The TAC of this configuration was $ 2.05 x106, which is equivalent to 10.5% saving compared to the conventional excess design without HI.

5.8

Recycle flow = 28 kmol/hr Recycle flow = 30 kmol/hr Recycle flow = 32 kmol/hr

Total reboiler duty (Gcal/hr)

Total reboiler duty (Gcal/hr)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

5.6

5.4

5.6

5.4

5.2

5.2 1.45

Recycle flow = 28 kmol/hr Recycle flow = 30 kmol/hr Recycle flow = 32 kmol/hr

5.8

1.50

1.55

1.60

1.65

1.70

1.75

261

264

267

270

273

Reflux ratio

Vapor side stream (kmol/hr)

(a)

(b)

276

279

Figure 14. Optimization result of the proposed TCD with HI case (Effect of the reflux ratio of RD column (a), and vapor side stream (b) for each recycle flowrate on the TAC).

4.4. Comparison of three configurations Table 2 lists the design results of three configurations. The aim of the proposed combined configuration is to achieve more advantages from both energy integrations. All three configurations saved energy and TAC; however, as shown in Table 2, the TCD with HI case saved energy (32.9%) and TAC (10.5%) mostly. While the TCD alone case saved 9.4% and 2.0% of energy and TAC, respectively, and the HI alone case saved 21.7% and 8.8% of energy and TAC, respectively. The net energy saving of the TCD with HI case was higher than the simple sum of the TCD case and the HI case due to the difference in recycle flowrates. Although, the heat exchange area of the heat exchanger in the TCD with HI case is larger than that in the HI case, the percentage energy saving in the TCD with HI case was still higher. An additional reason for the high energy saving in the TCD with HI case is that thermal coupling can reduce the remixing effect. HI has the most influence here in term of

27 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 33

energy efficiency. The result showed the potential of the TCD with HI configuration for enhancing energy efficiency in the esterification between LA and EtOH system.

Table 2. Comparison of the conventional, TCD, HI, and TCD with HI cases Conventional

TCD

HI

TCD with HI

Total duty (Gcal/hr)

7.77

7.04

6.08

5.21

Recycle flowrate (kmol/hr)

26

30

26

30

Energy saving (%)

-

9.4

21.7

32.9

TAC saving (%)

-

2.0

8.8

10.5

5.

CONCLUSIONS LAEE is an important blending component to biodiesel that causes modifications in

the fuel properties. LAEE is also a good candidate for gasoline fuels additives due to the high blending octane number and high oxygen content. A novel LAEE production process using RD technology was proposed to overcome the equilibrium limitation of esterification. The ultimate goal of this design was to obtain a LAEE with 99.5% purity and maintain the purity of byproduct (H2O) at 99.5%. The effects of several key design variables, including the recycle flowrate, on the economical production of LAEE were investigated. The results showed that the excess design is more economical than the neat design by avoiding much of the energy required to achieve the desired product in the RD column in the neat design. A detailed optimal steady state design was then found through TAC analysis. Finally, an energy integrated arrangement combining a TCD and external HI was proposed to enhance the energy efficiency of the excess RD design configuration. The combined arrangement offered more advantages compared to the TCD and HI alone configurations and achieved significant energy and TAC savings of 32.9% and 10.5%, respectively. Overall, the proposed excess RD

28 Environment ACS Paragon Plus

Page 29 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

design with the combined energy integrations configuration is a promising option for the esterification process between the LA and EtOH system to produce high purity LAEE in an energy efficient manner.

 ASSOCIATED CONTENT The supporting information is available free of charge on the ACS Publications website. Fuel characteristics of each ester levulinate; Process types based on the relative volatility ranking; Boiling point ranking; Kinetic model parameters; Thermodynamic parameters; Total annual costs calculation.

 AUTHOR INFORMATION Corresponding Authors *Tel.: +886-2-2737-6613. E-mail address: [email protected] **Tel.: +82-5-3810-2512. E-mail address: [email protected] (M. Lee)

Notes The authors declare no competing financial interest.

 ACKNOWLEDGMENTS This study was supported by the Basic Science Research Program through the National Research

Foundation

of

Korea

(NRF)

funded

by

the

Ministry

of

Education

(2015R1D1A3A01015621) and by the Priority Research Centers Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2014R1A6A1031189). The authors are also grateful for the support from the Ministry of Science and Technology of Taiwan under grant MOST 104-2221-E-011-149-MY2.

29 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

References (1)

Rackemann, D.W.; Doherty, W.O.S. The conversion of lignocellulosics to levulinic acid.

Biofuels, Bioprod. Bioref. 2011, 5, 198–214.

(2)

Fitzpatrick, S. W., Lignocellulose Degradation to Furfural and Levulinic Acid. U.S.

Patent 4897497; Biofine Incorporated, 1990. (3)

Fitzpatrick, S. W., Production of Levulinic Acid from Carbohydrate-Containing

Materials. U.S. Patent 5608105; Biofine Incorporated, 1997. (4)

Fitzpatrick, S. W. Final Technical Report Commercialization of the Biofine

Technology for LeVulinic Acid Production from Paper Sludge; BioMetics, Inc.; Report No. DOE/CE/41178; U.S. Department of Energy: Washington, DC, 2002. (5)

Joshi, H.; Moser, B. R.; Toler, J.; Smith, W. F.; Walker, T., Ethyl Levulinate: A

Potential Bio-based Diluent for Biodiesel which Improves Cold Flow Properties. Biomass and Bioenergy., 2011, 35, 3262-3266. (6)

Fagan, P. J.; Korovessi, E.; Manzer, L. E.; Mehta, R.; Thomas, S. M. (Dupont)

International Patent No. WO 03/085071 A1, 2003. (7)

Fagan, P. J.; Manzer, L. E. U.S. Patent 7,153,996, 2006.

(8)

Grove, A. P.; Morely, C.; Smith, J.; Stevenson, A. U.S. Patent Application

2005/0144835 A1, 2005. (9)

Murphy, M. M.; Taylor, J. D.; McCormick, R. L. Compendium of Experimental

Cetane Number Data; National Renewable Energy Laboratory (NREL): Golden, CO, NREL/SR-540-36805, Sept2004. (10)

Lake, M. A.; Burton, S. W., U.S. Patent Application 2010/0313467 A1, 2010.

(11)

Yanowitz, J.; McCormick, R. L., Effect of biodiesel blends on North American

heavy-duty diesel engine emissions, Eur. J. Lipid Sci. Technol., 2009, 111 (7), 763–772. (12)

Resk, A. J.; Peereboom, L.; Kolah, A. K.; Miller, D. J.; and Lira, C. T., Phase

30 Environment ACS Paragon Plus

Page 30 of 33

Page 31 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Equilibria in Systems with Levulinic Acid and Ethyl Levulinate. Journal of Chemical and Engineering Data, 2014, 59(4): 1062-1068. (13) 2016.

Grand view research, Ethyl Levulinate Market Worth $11.8 Million by 2022, March Available

from:

http://www.grandviewresearch.com/press-release/global-ethyl-

levulinate-market (accessed 23.11.16). (14)

Huber, G. W.; Iborra, S.; Corma, A., Synthesis of Transportation Fuels from Biomass:

Chemistry, Catalysts, and Engineering. Chem. Rev., 2006, 106, 4044−4098. (15)

Lee, H. Y.; Lee, Y. C.; Chien, I. L.; Huang, H. P., Desidn and Control of a Heat-

Integrated Reactive Distillation System for the Hydrolysis of Methyl Acetate. Ind. Eng. Chem. Res., 2010, 49: 7398-7411. (16)

Chung, Y. H.; Peng, T. H.; Lee, H. Y.; Chen, C. L.; Chien, I. L., Design and Control

of Reactive Distillation System for Esterification of Levulinic Acid and n-Butanol. Ind. Eng. Chem. Res., 2015, 54: 3341-3354. (17)

Luyben, W. L.; Yu, C. C., Reactive Distillation Design and Control; John Wiley &

Sons, Inc.: Hoboken, NJ, 2008. (18)

Sharma, M. M.; Mahajani, S. M., Industrial Applications of Reactive Distillation. In

Reactive Distillation: Status and Future Directions; Sundmacher, K., Kienle, A., Eds.; Wiley−VCH Verlag CmbH & Co. KgaA: Weinheim, Germany, 2002. (19)

Tung, S. T.; Yu., C. C., Effects of relative volatility ranking to the design of reactive

distillation. Aiche Journal, 2007, 53(5): 1278-1297. (20)

Aspen Technology, Aspen distillation synthesis, Version Number V8.6, 2014.

(21)

Humphrey, J. L.; Keller, G. E. Separation Process Technology. McGraw-Hill, 1997.

(22)

Wang, S. J.; Lee, H. Y.; Ho, J. H.; Yu, C. C.; Huang, H. P.; Lee, M. J., Plantwide

Design of Ideal Reactive Distillation Processes with Thermal Coupling. Ind. Eng. Chem. Res, 2010, 49, 3262 – 3274.

31 Environment ACS Paragon Plus

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(23)

Triantafyllou, C.; Smith, R., The Design and Operation of Fully Thermally Coupled

Distillation Columns. Trans. Inst. Chem. Eng., 1992, 70A, 118. (24)

Novita, F. J.; Lee, H. Y.; Lee, M. Y., Self-heat recuperative dividing wall column for

enhancing the energy efficiency of the reactive distillation process in the formic acid production process, Chemical Engineering and Processing: Process Intensification., 2015, 97:144-52. (25)

Mueller, I.; Kloeker, M.; Kenig, E. Y., Rate-based modelling of dividing wall

columns - A new application to reactive systems. In CHISA 2004 16th International Congress of Chemical and Process Engineering, 2004, pp 10297−10317. (26)

Wang, S. J.; Wong, D. S. H.; Yu, S. W. Design and Control of Transesterification

Reactive Distillation with Thermal Coupling., Comput. Chem. Eng., 2008, 32, 3030. (27)

Lee, H. Y.; Lai, I. K.; Huang, H. P.; Chien, I. L. Design and Control of Thermally

Coupled Reactive Distillation for the production of Isopropyl Acetate., Ind. Eng. Chem. Res.,

2012, 51, 11753. (28)

Tsai C. Y., Kinetic Behavior Study on the Synthesis of Ethyl Levulinate over

Heterogeneous Catalyst, Master Thesis, NTUST, Taiwan, 2014. (29)

Lin, L. C., Effects of Relative Volatility Ranking to the Design of Reactive

Distillation: Excess-Reactant Design. M. S. Thesis, National Taiwan University, Taipei, 2007. (30)

Doherty, M. F.; Malone, M. F., Conceptual Design of Distillation Systems; McGraw-

Hill: New York, 2001. (31)

Jobson, M., Energy Considerations in Distillation. In Distillation: Fundamentals and

Principles; Gorak, A., Sorensen, E., Eds.; Academic Press (Elsevier), London, 2014. (32)

Douglas, J. M., Conceptual design of chemical processes, McGraw-Hill: New York,

1998.

32 Environment ACS Paragon Plus

Page 32 of 33

Page 33 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

(33)

Luyben, W., Distillation Design and Control Using Aspen Simulation; John Wiley &

Sons, Inc; New Jersey, 2006.

Table of Contents

33 Environment ACS Paragon Plus