Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation Rupsha Bhattacharyya*, Kalyan Bhanja, Sadhana Mohan Heavy Water Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, Maharashtra, India

article info

abstract

Article history:

Cryogenic distillation is a very widely applied technique for separating the components of a

Received 11 September 2015

mixture of hydrogen isotopes, which may contain up to six isotopic species. This work

Accepted 20 January 2016

presents a simple dynamic model for evaluating the major performance characteristics of a

Available online xxx

cryogenic distillation column for hydrogen isotope separation. Liquid hold up, pressure drop and flooding point have been predicted for the tower using known packing charac-

Keywords:

teristics and fluid properties. The column has been modelled as an equivalent tray tower

Cryogenic distillation

using a known value of the height equivalent to theoretical plate (HETP) and its dynamic

Hydrogen isotopes

behaviour and separation performance under various conditions have been predicted. The

Dynamic simulation

effect of parameters like temperature, reflux ratio, feed point location and feed composition on the final isotopic contents of top and bottom products at steady state has also been studied. Consequences of neglecting the decay heat of tritium in distillation calculations have also been evaluated. This model is intended for preliminary column design work using data from literature, without the need for prior experimental activity. Copyright © 2016, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Introduction Separation of the isotopes of hydrogen has always been an important unit operation in the field of nuclear energy production. For use as the coolant and moderator in pressurized heavy water based reactors (PHWRs) employing natural uranium oxide as the fissile material, the separation and concentration of heavy water from natural sources of light water is a well known and mature process technology, using unit operations like vacuum distillation and catalytic exchange [1]. For the fusion energy field, gaseous hydrogen isotopes like deuterium and tritium are the postulated fuels. The pure form

of each isotope can be separated and recovered in a cryogenic distillation column cascade [2]. For fusion applications, the deuterium-tritium streams are the desired products. Hydrogen isotopes have normal boiling points ranging from 20 to 25 K [3], so a cryogenic refrigeration system based on helium as the working fluid is an auxiliary requirement. Thus cryogenic distillation of hydrogen isotopes is a very energy intensive process. There is a significant body of literature pertaining to the cryogenic distillation of hydrogen isotopes, with a number of papers being available on both theoretical and experimental studies. Among the theoretical investigations, mention can be made of the work of Busigin et al. [4] in which simulation

* Corresponding author. Tel.: þ91 022 2559 2962. E-mail address: [email protected] (R. Bhattacharyya). http://dx.doi.org/10.1016/j.ijhydene.2016.01.106 0360-3199/Copyright © 2016, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

2

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

studies on the detailed design of an isotope separation system for ITER application, for two different kinds of tritium breeding blankets have been presented. Special emphasis has been placed on the cryogenic distillation system (columns and equilibriators promoting isotope exchange reaction), which is at the very heart of hydrogen isotope separation. Optimization of the distillation column cascade to minimize the total tritium inventory handled in it has also been carried out in this work. A simplified semi-rigorous model for predicting transient and steady state behaviour of cryogenic distillation columns for hydrogen isotope separation that allows reduction of computational time has also been developed [5]. This improvement has been brought about by using the concept of a supertray, which is actually a block of trays of a column over which the concentration profiles are very flat. Hence this leads to a reduction in the number of equations to be solved for the column. The model has been well validated by experimental studies. A first principle, theoretical study of the transport phenomena involved in cryogenic distillation of hydrogen isotopes has been carried out by some researchers to illustrate the fact that not all the components of the mixture simultaneously reach equilibrium in the vapour and liquid streams leaving a theoretical stage [6]. Specific simulation packages applicable to hydrogen isotope distillation have also been developed [7] where the non-ideality of the hydrogen isotope solutions has been taken into account and the modified algorithm also ensures faster convergence to the solution. Distillation columns with side streams originating in an equilibriator have also been simulated [8] and location of the side stream was seen to strongly influence the column performance. Decay heat of tritium and difference in the molal latent heats of vaporization of the various species were also taken into account in the model used in this study. To account for the radioactivity of tritiated species, a reactive distillation column model was formulated by a different group of researchers [9] and the code was integrated with a commercial process simulator's distillation module to study column performance. Simulations were validated with experiments performed at Laue-Langevin Institute's tritium extraction plant. Several experimental studies on hydrogen isotope distillation at both laboratory scale and plant scale facilities have been carried out and very important design and operating data have been extracted from them. For example, in a work by Enoeda et al. [10], pressure drop behaviour of the cryogenic distillation column and values of HETP (about 4e6 cm under the conditions used in the study) have been determined experimentally. In another study, a two column arrangement for cryogenic hydrogen isotope distillation has been used to obtain design and operating data for ITER design (e.g. HETP values). A computer code for simulating column performance has been developed. Laser Raman Spectroscopy has been used for determination of the isotopic compositions of the process streams [11]. Test data pertaining to the development and use of two new stainless steel packings for cryogenic hydrogen distillation have been presented recently by Bornea et al. [12]. Experiments on a laboratory scale distillation cryogenic column cascade have been described, parametric studies have been carried out and data for simulation of large scale systems suitable for ITER applications have been collected elsewhere [13]. In another work, full scale columns placed in vacuum

jackets where built and their hydraulic behaviour and separation performance with various packings were experimentally studied. Practical problems during column operation were identified [14]. Practical considerations in the design of cryogenic columns for hydrogen isotope separation have been addressed in yet another work [15]. An experimental facility named TRENTA which is a prototype of the ITER WDS and ISS protium separation column, has been operated at TLK and this work presents design and operating data and information about integration of these two processes so that tritium leakage into the environment can be reduced [16]. Since the hydrogen isotope feed stream is derived from a purge gas stream of helium after passing it through a palladium based membrane permeator, the effect of helium on the column's separation performance, control and condensor performance has been studied experimentally. Removal of helium before feeding the hydrogen isotopes to the column has been found to be necessary if the feed has more than 1% helium [17]. A two column cascade has been proposed as an alternative to the TSTA four column cascade for hydrogen isotope separation as would be required for fusion plants like ITER, with reduction in the number of instruments though there is an increase in the tritium inventory being handled here [18]. In this work, an equilibrium-based dynamic model for simulating the behaviour of a packed cryogenic distillation column for hydrogen isotope separation has been presented. The primary aim of the work is to bring together all the necessary data, correlations, and model equations for the distillation system in one place and to simulate the general behaviour of the distillation column, carry out parametric studies and explore possible design alternatives through simulations, without the need for new experimental work. Important hydrodynamic parameters of column behaviour have also been evaluated using physical property data of the isotopic system and packing characteristics of the column, for which equations from literature have been used. The column configuration has been kept conventional and emphasis is placed on the use of the model for prediction of dynamic composition profiles i.e. the separation behaviour in the column operated under various conditions. The model consists of a set of coupled ordinary differential equations representing material balances along with the vapour liquid equilibrium relationships for the hydrogen isotope system, which have been integrated over time using standard techniques. Generally a cascade of three to four columns is usually employed to recover deuterium and tritium in nearly pure form while removing protium [3]. In this work the focus is on the development of a simulation program, thus results for the first column have been presented in maximum detail and the entire cascade configuration has not been analysed here. The model represents the packed column as a tray tower by using known values of the HETP for the particular type of packing used and gas and liquid flow rates prevailing in it. Model equations such as the ones presented here can be integrated using codes written by the user without depending on expensive commercially available process simulation packages, which very often do not have complete pure component and phase equilibrium data for isotopic systems such as the one considered in this study. Moreover the solution of sets of simultaneous algebraic equations needed in steady state

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

models of multi-component distillation columns often leads to convergence and numerical instability issues. But the solution of sets of first order, non-linear, ordinary differential equations from an initially assumed composition profile using a ‘time-marching’ algorithm is a simpler and a more reliable technique. Further, for design of the control system for the distillation towers, dynamic models are a necessity. Thus the unsteady state model is also more versatile in its application than a steady state one and is an apt alternative to commercially available packages. Energy integration strategies which have been widely addressed for reducing energy requirements in these cryogenic operations [19,20] as well as the design of the refrigeration cycle have not been dealt with in this work.

Thermodynamic properties of the hydrogen isotope system at cryogenic conditions The isotopic hydrogen system consists of 6 species viz. H2, HD, D2, DT, HT and T2. The species HD, DT and HT are formed by isotopic exchange reactions, predominantly in the gas phase, as follows [21]: H2 þ D2 42HD

(2.1)

H2 þ T2 42HT

(2.2)

T2 þ D2 42DT

(2.3)

In this work, all the isotopic species have been considered to be present independently in both the vapour and liquid phases without explicitly considering the reaction equilibrium alongside the vapour liquid equilibrium. The vapour pressures of isotopic hydrogen species were obtained from Equations (2.4)e(2.9) respectively and they were used to calculate the relative volatilities as the ratio of saturation vapour pressure of each component in the mixture to that of a reference component (in this case T2 which is the heaviest component in the mixture) at a particular temperature. PsH2 ¼ 1:085T3  42:23T2 þ 594:5T  2984

(2.4)

PsHD ¼ 1:155T3  53:89T2 þ 905T  5391

(2.5)

PsD2 ¼ 1:246T  66:57T þ 1273T  8591

(2.6)

PsHT ¼ 1:141T  55:74T þ 974:2T  6026

(2.7)

PsDT ¼ 1:251T3  69:7T2 þ 1383T  9654

(2.8)

PsT2 ¼ 1:220T3  69:49T2 þ 1401T  9898

(2.9)

3

3

2

2

The vapour pressure equations were obtained by fitting polynomial equations to tabular data available for hydrogen isotopes in literature [22]. The mixture was assumed to be ideal in both the liquid and vapour phases. This assumption is valid given that the tower typically operates under at a moderate pressure of about 760e1000 mm Hg (absolute) and the liquid phase consists of isotopic species which are chemically very similar and are

3

quite close-boiling components. So the pure component vapour pressure data can provide quite accurate measures of the relative volatilities of the components in the mixture.

Hydrodynamic model for the cryogenic distillation column The cryogenic distillation of hydrogen isotopes is typically carried out in packed towers with countercurrent flow of vapour and liquid at pressures close to ambient conditions. High efficiency corrugated metal wire mesh type of packing or special random packings like Dixon rings have been employed for this separation process. The hydraulic behaviour of the column has been studied for the structured packing in this work. In the present study, all calculations are performed assuming a tray column which provides a separation performance equivalent to the actual packed column, based on available HETP data for the type of packing considered [23]. Typical design and range of operating data for the first hydrogen isotope distillation tower in the cascade considered in this work are presented in Table 1. The tray configuration used to write the unsteady state material balances shown in Section Dynamic simulation model of the cryogenic distillation column is shown in Fig. 1. Many of the correlations and data pertaining to packing geometry used in this work for predicting the hydraulic characteristics of the cryogenic distillation column have actually been empirically derived for other more conventional vapoureliquid systems under different conditions of operating pressure and especially temperature. Thus while the validity of these expressions for a cryogenic system may be in question, in the absence of detailed expressions and data relevant specifically for cryogenic columns, the generally available expressions have been employed in the simulation model.

Calculation of liquid hold up in the column The dynamic liquid hold up in the column was estimated using the following available correlation [25]: !1
2 16 2 3 !101 mL ap sL a2p 8 u ap hLo ¼ 0:93 LS g rL g r2L g

(3.1)

Equation (3.1) was used to determine the total liquid hold up for the entire packed column, which was then divided by the equivalent number of trays to obtain the liquid hold up per tray. The liquid hold ups in the reboiler and condensor drums are generally expected to be more than that on individual trays, so these were taken as twice the tray hold up calculated above. All hold ups were assumed to remain constant during the operation of the column. The physical properties of D2 (liquid) at the column average temperature were used to evaluate the hold up from Equation (3.1) [26].

Pressure drop characteristics of the column The calculation of pressure drop through the packed tower consists of first evaluating the pressure drop in dry condition

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

4

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

Table 1 e Base case design and operating parameters for the first distillation column [3,23,24]. Serial number

Parameter

Value

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Tower diameter Packed height Type of packing HETP Number of equivalent trays Corrugation angle Side dimension of corrugation Specific surface area of packing Packing void fraction Packing factor Cs Cfl Average liquid density Average liquid viscosity Average liquid surface tension Feed flow rate Feed composition

18 19 20

Feed condition Average vapour density Average vapour viscosity

0.15 m 18 m Corrugated metal wire gauze structured packing 0.15 m 120 60 with horizontal 0.009 m 700 m2/m3 0.85 21 m2/m3 2.5 2.25 70 kg m3 95.7*107 Pa s 0.0025 N m1 2.5e5.0 LPH Equilibrium feed at feed temperature starting from an equimolar mixture of isotopically pure H2, D2, T2 Saturated vapour (q ¼ 0) 2.5 kg m3 15*107 Pa s

assuming only the vapour to be flowing, followed by correcting it for the flow of the liquid and the liquid hold up in the irrigated column. Equations (3.2)e(3.5) were used for pressure drop calculation for the structured packing used in the distillation columns [23]. Pressure drop per unit length through dry packing was calculated as rg w2GE DPo ¼f H S

(3.2)

where f ¼ 0:177 þ

88:774 Reg

(3.3)

and wGE ¼

ug 2 sin q

(3.4)

Pressure drop through the irrigated column was calculated from

L

V

Molar liquid hold (M) up on tray # n

xn


5 DP DPo 1 ¼ H H 1  K2 hLo

(3.5)

whereK2 ¼ 0.614 þ 71.35S. The pressure drop at flooding was evaluated from Equation (3.6) [27]. DP=H ðinch water per foot of packingÞ ¼ 0:115F0:7 p

(3.6)

Using the parameters in Table 1, the pressure drop per metre of packing in the distillation tower has been calculated at various values of the reflux ratio (i.e. various possible combinations of gas and liquid flow rates through the column) at a given feed flow rate and presented in Fig. 2. It can be seen that for the typical range of flow rates and the total height of packing considered to be present in the tower in this work, the total pressure drop under irrigated conditions amounts to no more than 0.3 mm Hg. Thus the assumption of constant pressure in the column is very well applicable. By performing a bubble point calculation, it can be shown that the bottom temperature is only about 2e3  C higher than the top temperature. Thus the arithmetic mean of the top and bottom temperatures can be safely assumed as the constant average column operating temperature for the simulation studies since only a small temperature gradient prevails inside the column.

Loading and flooding point characteristics of the column

F L’ V’

Fig. 1 e Schematic diagram of the hypothetical feed tray for the cryogenic column.

Several correlations have been proposed in literature for the calculation of loading and flooding vapour velocities in a packed column [23,28]. Iterative calculations using Equations (3.7)e(3.8) were performed to determine the loading point vapour velocity for a given liquid superficial velocity in the column [23]. Liquid and vapour velocities inside the column were related to the specified feed flow rate, feed quality, the

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

5

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

30

Column pres s ure drop (Pa m -1)

25

20 F = 425 mol/hr 15 F = 325 mol/hr 10 F = 225 mol/hr 5 F = 125 mol/hr 0 50

100

150

200

Reflux ratio

Fig. 2 e Pressure drop through the distillation column as a function of reflux ratio and feed flow rate for given top and bottom liquid compositions (T ¼ 25 K). top and bottom product flow rates calculated on the basis of an assumed component split and the chosen reflux ratio above the minimum required value. Vapour and liquid properties were evaluated at a temperature of 25 K, assuming liquid at bubble point and vapour at dew point.

increased thermal duties, the feed flow rate and the reflux ratio can thus be increased several times over the values being considered in this work, without the danger of the column being flooded.

2 3
0:5
13
16 rffiffiffiffiffi g ε 12m rl l 0:5 4 1 a 5 12ml $uls $u ugl ¼ ls p xs grl grl rg a6

Evaluation of minimum wetting rate for the column (3.7)

p

xs ¼

g !0:4

2 C2s 4uugllsrrgl

ml mg

30:65 qffiffiffi ffi rg 5

(3.8)

rl

Similarly for the calculation of flooding point gas velocity, iterative calculations were done using Equations (3.9)e(3.11). The value of hold up was first estimated from Equation (3.11) which was then used in the calculation of the gas velocity at flooding condition. ugf ¼

1:5 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffirffiffiffiffiffi 0:5     rl ε  hdf 2g hdf ap xfl ε0:5 rg

xfl ¼

2




C2fl 4uugllsrrgl

g !0:2 ml mg

30:39 qffiffiffiffi rg 5

(3.9)

(3.10)

The minimum wetting rate in the cryogenic column was calculated from Equation (3.12) [25]. uSL ¼ 7:7*106


3 29
12 rl sL g ap m4l g

(3.12)

It can be easily shown that the liquid flow rates prevailing in the tower are much above the calculated minimum rate required for wetting the packing, so efficient gaseliquid contact over the surface area provided by the packing can take place in practice.

Dynamic simulation model of the cryogenic distillation column The model equations for the dynamic simulation are based on the following assumptions [29]:

rl

 6  m h3df 3hdf  ε ¼ a2p ε l $uls g rl

(3.11)

Typical results of both loading and flooding point calculations are shown in Fig. 3. The flooding pressure drop for the column is much higher than the calculated pressure drop for the range of gas and liquid flow rates considered here. Provided the condensor and reboiler can adequately handle the

a) Both the vapour and the liquid phases are taken to be ideal solutions. Constant average relative volatilities have been considered. b) The column top temperature has been estimated from the assumed condensor operating pressure. The feed and the reflux are assumed to be at their dew point and bubble point respectively. The reflux condenser is assumed to be a total condenser. c) Constant molal overflow is assumed.

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

6

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

0.24 Loading point gas velocity Flooding point gas velocity

Gas s uperfic ial v eloc ity (m s-1)

0.22

0.2

0.18

0.16

0.14

0.12

0.1 10

15

20

25

30

35

40

45

50

-1

Liquid superficial velocity (m hr ) Fig. 3 e Gas velocity at loading and flooding points as a function of liquid velocity in the distillation column (T ¼ 25 K). d) The stages are assumed to be 100% efficient with respect to mass transfer. e) The pressure drop through the tower has been calculated based on the packing characteristics, which have been assumed to be equivalent to Sulzer™ CY type corrugated metal wire mesh packing. Based on the tower bottom pressure, temperature of the reboiler has been estimated. f) The tower has only the feed, distillate and the bottoms streams and no other side streams. g) The liquid volume or hold up in the reboiler, reflux drum and on each of the hypothetical column plates are wellmixed regions having uniform composition. h) The dynamics of the piping, reboiler and condenser are negligible, thus there are no time lag elements in the system. Vapour phase dynamics are neglected as it is much faster. i) The liquid hold-ups are constant on each tray and in the reboiler and reflux drums. j) The exchange reactions between various isotopic hydrogen species on the trays have been neglected in studying the performance of the column. All the six species have thus been considered independently. k) The column is adiabatic and decay heat release from tritium is neglected. l) Owing to the rather low pressure drop calculated for the column operating conditions as specified in Table 1, a constant column operating pressure has been assumed for the simulation studies. m) The feed composition to the column is obtained by assuming equimolar mixture of isotopically pure hydrogen, deuterium and tritium species which equilibrates at the feed temperature through the exchange reactions, giving rise to the species HD, HT and DT. The objective of the distillation cascade is to recover as much D2 in as pure a form as possible and to obtain a tritium

rich stream of DT-T2. The hydrogen containing streams are treated mainly as reject streams. In the first column of the cascade, the major objective is protium removal through the top stream, while concentrating deuterium and tritium in the column bottoms. The species HT is taken as the light key and D2 is the heavy key. There are no intermediate components in between these keys. An indirect split is attempted wherein the major part of the heavy key and components heavier than the heavy key are separated as the bottoms product in the first column of the cascade itself.

Column model under total reflux conditions For column operation at total reflux condition, the above assumptions were used to write the following material balance equations [29]: For the reboiler or sump (tray number N þ 1): MB

dxNþ1;j ¼ LxN;j  VyNþ1;j dt

(4.1)

For the tray section (for tray number n where n varies from 2 to N): Mt

   dxn;j ¼ L xnþ1;j exn;j  V yn1;j  yn;j dt

(4.2)

For the reflux drum (tray number 1): MD

dx1;j ¼ Vy2;j  Lx1;j dt

(4.3)

For operation under total reflux condition, V is equal to L. The vapoureliquid equilibrium relationship for the multicomponent system, assuming constant average relative volatilities can be written as: aj xi;j yi;j ¼ P aj xi;j

(4.4)

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

7

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

Equations (4.1)e(4.4) were written for each component on each tray and simultaneously integrated to obtain the time dependent composition profiles of each species on each tray. Calculations were continued till steady state composition of each component was attained on each tray. For total reflux operation, a given quantity of the feed as a saturated liquid was assumed to be fed to the sump. The initial feed composition was taken as the initial composition of the liquid on each tray. The numerical integration was carried out with the assumed initial composition profile. The values of the hold up in the sump and on each tray were taken as constants for a particular run. The vapour boil up rate can be varied by varying the thermal energy input to the sump liquid and it is assumed that the required vapour flow rate is instantly established by neglecting heat transfer dynamics in the reboiler. The evolution of the composition profiles on each tray of the tower was then determined and the compositions at steady state were obtained.

The model assumptions were used to write Equations (4.5)e(4.11) for simulating the column behaviour under constant feed introduction and top and bottom product withdrawal [29]. The initial composition on each simulated tray of the column was assumed to be the composition attained from total reflux operation. The equations were then integrated as described in the previous section to obtain the dynamic composition profiles along the column and arrive at the steady state compositions on each tray when the tower is operating continuously. For the reflux drum (Tray 1): dx1;j ¼ Vy2;j  ðL þ DÞx1;j dt

Mt

dxn;j dt

      dxf ;j ¼ L xf 1;j  L0 xf ;j þ V0 yf þ1;jÞ  V yf ;j þ Fxf 0;j dt

(4.12)

Fenske-Underwood-Gilliland (FUG) calculations for hydrogen isotope distillation The FUG calculations provide us with some starting point (e.g. initial assumed top and bottom compositions, initial composition profile or reflux ratios to be considered) for the more detailed tray to tray simulations in case of multi-component distillation [25,30]. These calculations were performed for the hydrogen isotope distillation column for some typical conditions and the results are presented in Table 2. The actual reflux ratio and number of trays in the column are higher than the calculated values, thus establishing the fact that the desired separation is possible.

A single stage flash calculation has been performed in this section to demonstrate the effect of neglecting tritium's decay heat in the analysis of the column's separation performance (Fig. 4). The decay heat from the tritiated species has been included as a feed composition dependent heat addition term in the case of non-adiabatic flashing and for the case where decay heat effect has been neglected, adiabatic flash calculation has been performed, to estimate the flash temperature and the vapour and liquid flow rates and compositions for both cases. Feed has been taken as a saturated liquid in each case and the products leaving the flash stage have been assumed to be in thermodynamic equilibrium. A modification of the Rachford-Rice algorithm for non-

Table 2 e Data and results of FUG calculations for the hydrogen isotope separation column. (4.6)

For the feed tray (n þ 1): Mt

L0 ¼ L þ qF

(4.5)

For trays in the rectification section (2 to n):    ¼ L xn1;j  xn;j þ V ynþ1;j  yn;j

(4.11)

Consequences of neglecting the decay heat of tritiated species in distillation calculations

Column model with continuous feed input and product withdrawal

MD

V0 ¼ V þ Fðq  1Þ

Serial number 1

(4.7)

For trays in the stripping section (n þ 2 to N þ 1): Mt

   dxm;j ¼ L0 xm1;j  xm;j þ V0 ymþ1;j  ym;j dt

2

(4.8)

For the reboiler (tray N þ 2): MB

dxB;j ¼ L' xN;j  WxB;j  V' yB;j dt

(4.9)

Vapour liquid equilibrium relation: aj xi;j yi;j ¼ P aj xi;j

3 4

(4.10)

The relationship between the vapour and liquid flow rates in the enriching and stripping section to the feed quality can be expressed as follows [30]:

5 6

Parameter

Value

xF,1 ¼ 0.1670 xF,2 ¼ 0.2003 xF,3 ¼ 0.1255 xF,4 ¼ 0.1323 xF,5 ¼ 0.2154 xF,6 ¼ 0.1595 Top tray composition assumed xd,1 ¼ 0.998100 xd,2 ¼ 0.001945 xd,3 ¼ 4.522e-10 xd4 ¼ 2.7000e-6 xd,5 ¼ 9.656e-15 xd,6 ¼ 1.572e-20 Average column temperature 20 K Relative volatility (Ps(j)/Ps(6)) 8.0698 4.5698 2.4302 3.3721 1.5581 1.0000 Minimum reflux ratio 1.44 Minimum number of trays 54 Feed composition assumed

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

8

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

isothermal flash calculations has been used to obtain the results in both cases [27]. Stream enthalpies have been evaluated using expressions for each pure isotopic vapour and liquid species available in literature [3]. Equilibrium distribution coefficients (i.e. K values) have been calculated from vapour pressure data for each component. Results are shown in Table 3. It is seen that the temperature after flashing, the vapour to feed ratio and compositions of the products from the flash operation are almost exactly the same, with and without including the effects of decay heat from the tritiated species, even for feed streams having high concentration of the tritiated species (e.g. Case 2 in Table 3). Thus energy balance equations to include the decay heat effect on the composition profiles in the column will not be an absolute necessity in the simulation model, as decay heat has little effect on the separation behaviour and the composition profiles in the column.

Fig. 4 e Schematic of single stage flash vaporization calculation to study the effect of decay heat on separation (Q ¼ 0 when neglecting decay heat, Q ¼ Q (zF) when considering decay heat).

in all subsequent figures. The feed composition (and the starting composition on all trays for the total reflux operation) has been assumed to be that at equilibrium at a given feed temperature, starting from equimolar isotopically pure H2, D2 and T2.

Simulation results and analysis Parametric studies on total reflux operation Some typical results of the tray to tray composition profile calculations for both total reflux and continuous column operation are presented in Figs. 5 and 6. The trays have been differently numbered for the cases of total reflux and continuous distillation, with tray number 1 being the reboiler for the first case and the reflux drum for the second case respectively. Since products are to be withdrawn from only the top and bottom stages of the column, only the top tray and bottom tray liquid compositions have been plotted

The top and bottom tray fluid compositions (i.e. isotopic contents) are of importance since in continuous operation, the two product streams will be withdrawn from these locations only. Based on the calculated composition profiles in terms of mole fractions of the six species, the deuterium and tritium content (in terms of atom ratios) in any stream are computed as in Equations (5.1)e(5.2) and represented in the subsequent figures.

Table 3 e Flash calculation results for mixture of hydrogen isotopes. Serial number

Parameter

1

Feed composition assumed

2 3 4 5

Feed temperature Feed pressure Flash chamber pressure Flash temperature

6

Vapour fraction (j)

7

~B ) Liquid composition (x

8

~D ) Vapour composition (y

Case 1 xF,1 ¼ 0.1670 xF,2 ¼ 0.2003 xF,3 ¼ 0.1255 xF,4 ¼ 0.1323 xF,5 ¼ 0.2154 xF,6 ¼ 0.1595 23.80 K (Bubble point) 1000 Torr 600 Torr With decay heat Without 22.0276 22.0273 With decay heat Without 0.0666 0.0660 With decay heat Without 0.1563 0.1564 0.1972 0.1973 0.1274 0.1274 0.1326 0.1326 0.2213 0.2212 0.1653 0.1652 With decay heat Without 0.3171 0.3172 0.2434 0.2434 0.0991 0.0991 0.1287 0.1287 0.1333 0.1333 0.0784 0.0784

Case 2

decay heat decay heat decay heat

decay heat

xF,1 ¼ 0.0 xF,2 ¼ 0.0 xF,3 ¼ 0.1 xF,4 ¼ 0.0 xF,5 ¼ 0.6 xF,6 ¼ 0.3 25.48 K (Bubble point) 1000 Torr 600 Torr With decay heat Without decay 23.6605 K 23.6604 K With decay heat Without decay 0.0649 0.0640 With decay heat Without decay 0 0 0 0 0.0982 0.0983 0 0 0.5987 0.5988 0.3030 0.3030 With decay heat Without decay 0 0 0 0 0.1253 0.1254 0 0 0.6180 0.6180 0.2567 0.2566

heat heat heat

heat

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

9

Fig. 5 e a: Liquid composition profile evolution in hydrogen isotope distillation at total reflux. (T ¼ 25 K, V ¼ 30 mol h¡1). b: Steady state tray to tray liquid composition profile in hydrogen isotope distillation at total reflux (T ¼ 25 K, V ¼ 30 mol h¡1, Tray # 1 is the reboiler).

D ¼ 0:5xðHDÞ þ 0:5xðDTÞ þ xðD2 Þ ICð1Þ ¼ DþHþT

(5.1)

T ¼ 0:5xðHTÞ þ 0:5xðDTÞ þ xðT2 Þ DþHþT

(5.2)

ICð2Þ ¼

Fig. 7 demonstrates the effect of changing the vapour/ liquid rate at total reflux by varying the heating rate of reboiler liquid on the top and bottom composition profiles. As the vapour rate changes, the liquid rate and hence the liquid superficial velocity change. This causes a change in the liquid hold up on each tray and affects the time required to attain a steady composition profile, though ultimately the same

steady state composition for each component is attained for a given starting composition. To illustrate the difference in the evolution of the composition profiles on changing the boil up rate, the first 10 h of operation have been shown in this figure. Lower the boil up, higher is the time taken to reach steady state in the column. Fig. 8 illustrates the fact that increasing the average column operating temperature (which is directly dependent on the column operating pressure and the pressure drop) leads to a generally poorer separation, with more hydrogen appearing in the bottoms stream and higher tritium content of the top stream as the temperature is raised. This is due to the lower relative volatility at the elevated temperature which leads to

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

10

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

Fig. 6 e a: Liquid composition profile evolution in continuous hydrogen isotope distillation (T ¼ 25 K, F ¼ 225 mol h¡1, R ¼ 100, Feed tray # 59). b: Steady state tray to tray liquid composition profile in continuous hydrogen isotope distillation (T ¼ 25 K, F ¼ 225 mol h¡1, R ¼ 100, Feed tray # 59, Tray # 1 is the reflux drum). poor separation performance in the given tower. It is also evident that the change in composition for every degree change in column temperature is more pronounced for the reboiler liquid as compared with the reflux drum liquid. Figs. 7 and 8 show that the steady top isotopic contents are attained earlier than the steady bottom isotopic content values i.e. there is a difference in the time constants of the mass and energy transfer processes in the top and bottom sections of the column due to the difference in composition of the fluids in these locations.

Parametric studies on continuous column operation As with the operation of the column under total reflux, parametric studies were also carried out for the dynamic

behaviour of the continuous column up to a time equal to first 10 h of column operation. In most cases considered here, steady state compositions are attained within 10 h from start up of the column. Results are presented in terms of the contents of different isotopes of the liquid in the reflux drum and the reboiler in Figs. 9 to 12. It is seen in Fig. 9 that variations in column operating temperature affect the bottom product concentrations more than the top products. Steady state is also attained much earlier in the top trays than in the bottom trays. The dynamic concentration profile evolution is dependent on the average column temperature. The steady state isotopic contents are not very significantly affected by temperature, though the variation in the distribution of the individual species is influenced by it. Increasing the reflux ratio increases the time

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

11

Fig. 7 e Effect of vapour boil up rate on the evolution of composition profiles in batch distillation (Temperature ¼ 25 K, Boil up rates from 36 mol h¡1 to 636 mol h¡1 in increments of 100 mol h¡1).

taken for steady state to be established owing to the difference in the liquid hold ups. There is an improvement in separation performance of the column as well, though only a marginal difference occurs in the isotopic contents of the top and bottom streams with higher reflux ratio, as seen in Fig. 10. Feed flow rate variation influences the dynamic composition profiles but finally the same steady state compositions

are attained, as seen in Fig. 11. From Fig. 12, feed tray location seems to have only negligible influence on the composition profiles in the column and hence the ultimate separation in this case. For all the cases considered in Figs. 9e12, nearly 100% recovery of feed tritium in the bottoms product has been calculated. Further, in the bottoms product, total tritium concentration is about 20% more than that

Fig. 8 e Effect of column average temperature on the evolution of composition profiles in batch distillation (Boil up ¼ 36 mol hr¡1, Temperature from 20 to 25 K in increments of 1 K). Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

12

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

Fig. 9 e Effect of column average temperature on the evolution of composition profiles during continuous distillation (R ¼ 100, F ¼ 225 mol h¡1). in the feed stream. This also indicates that some amount of protium, present as HT is also recovered in the bottoms product and this must be removed in another column in order to obtain pure tritium. For the feed concentration

considered here, this does not represent a very high tritium enrichment factor. Calculations further indicate similar deuterium enrichment in the bottoms product from the first distillation column.

Fig. 10 e Effect of reflux ratio on the evolution of composition profiles during continuous distillation (T ¼ 25 K, F ¼ 225 mol h¡1). Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

13

Fig. 11 e Effect of feed flow rate on the evolution of composition profiles during continuous distillation (T ¼ 25 K, R ¼ 100).

Uncertainties in the simulation model There are several input parameters required to be provided to the simulation model presented here, like the phase equilibrium data, column hydraulic characteristics, HETP values and fluid physical property data. These are mostly experimentally determined quantities, finally presented in the form of best-fit

regression equations or point values in some cases. There is always a measure of experimental error and errors in fitting the experimental data to model equations and thus the numerical values of these input parameters can have significant effect on the column performance predicted by the simulation model. Error in the vapour pressure data of the components is seen to produce negligible effect on the final composition

Fig. 12 e Effect of feed tray location on the evolution of composition profiles during continuous distillation (T ¼ 25 K, R ¼ 100, F ¼ 225 mol h¡1). Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

14

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

profiles, primarily because the ratios of the vapour pressures are used in the calculations. Errors in physical property data are seen to lead to error propagation into the calculated values of liquid hold up, hence a difference in the evolution of the composition profile with temp is noticed, though the final steady state compositions are not affected. Prediction of the HETP is crucial since that governs the number of trays input to the model, for a given total height of packing. Error in HETP value leads to variation in the number of trays and final predicted compositions are affected. 50% error in the HETP value is seen to lead to a maximum variation of about 5% in the isotopic purity calculated from the compositions.

in both phases and comparing the values obtained by using vapour pressure data alone as employed in this work, can be addressed as a separate problem.

Nomenclature ap Cfl Cs D f

Summary and conclusion F The paper presents an unsteady state mathematical model for performance simulation of a cryogenic distillation column for hydrogen isotope separation. Evaluation of some of the most important parameters of column hydraulic behaviour has also been carried out using data from literature. For the dynamic analysis of the column performance, liquid hold up is an important parameter, since hold up affects the time taken to achieve steady state in the column, unlike in the case of steady state simulations where hold up does not appear in the material balance equations. Thus accurate prediction of hold up is significant for dynamic modelling and necessary for calculation of radioactive species inventory (e.g. tritiated species hold up) in the column. Flooding and loading point calculations are pertinent for packed columns since these place practical limits on the maximum throughput that can be handled by a given column. For the typical column throughputs considered in this work, the prevailing vapour velocities are found to be much lower than the flooding velocities for the given liquid rates. Energy effects and chemical isotopic exchange have been ignored in the simulation model without incurring any significant loss of accuracy, as shown by the single stage flash calculation and the vapoureliquid equilibrium data along with material balances have been used to predict the dynamic concentration profiles along the length of the column for both total reflux and continuous column operation. Effect of various operating parameters on the separation performance achievable in the column has been investigated. It is observed that increasing column temperature has an adverse effect on separation achieved but this may be countered by increasing the reflux ratio. Increasing feed flow rate affects the time required to attain the steady state but does not affect the final top and bottom product compositions achieved. The distillation column model presented here can be used as a method of solving the more difficult column design problem through the ‘design by repeated simulation’ philosophy, by varying some of the parameters shown in Table 1. It may further be integrated with models of the cryogenic refrigeration cycle (to which it will provide calculated values of the heating and cooling duty inputs from the reboiler and condensor duties) and it can also be used to provide necessary inputs to energy integration studies to minimize energy requirements for the cryogenic systems. Prediction of separation factors for this distillation, considering an equation of state model to account for non idealities

Fp g hdf hLo H HD HF HW IC(1) IC(2) K2 L L0 MB MD Mt n N P PF PsD2 PsDT PsHD PsHT PsH2 PsT2 q Q R Reg S t

Specific surface area of packing, m2 m3 Characteristic constant for flooding point behaviour of packing, dimensionless Characteristic constant for loading point behaviour of packing, dimensionless Molar flow rate of top product in continuous or flash distillation, mol hr1 Friction factor for flow through corrugated packing, dimensionless Feed flow rate (to continuous column or flash chamber), mol hr1 Packing factor, ft1 Acceleration due to gravity, m s1 Liquid hold up in the packed column at flooding point, m3 liquid m3 packing Liquid hold up in the packed column, m3 liquid m3 packing Height of packing, m Specific enthalpy of stream leaving flash chamber through the top, J mol1 Specific enthalpy of stream entering flash chamber, J mol1 Specific enthalpy of stream leaving flash chamber through the bottom, J mol1 Deuterium isotopic content in any stream, dimensionless Tritium isotopic content in any stream, dimensionless Constant for calculation of wet pressure drop, dimensionless Molar liquid flow rate in rectifying section, mol hr1 Molar liquid flow rate in stripping section, mol hr1 Molar liquid hold up in the reboiler, mol Molar liquid hold up in the reflux drum, mol Molar liquid hold up on each hypothetical tray, mol Tray number index, dimensionless Number of equivalent trays in the column, dimensionless Pressure of streams leaving flash chamber, Torr Pressure of feed before flashing, Torr Vapour pressure of D2, Torr Vapour pressure of DT, Torr Vapour pressure of HD, Torr Vapour pressure of HT, Torr Vapour pressure of H2, Torr Vapour pressure of T2, Torr Fraction liquid in feed, dimensionless Heat input during non-adiabatic flash, J Reflux ratio, dimensionless Packing Reynolds number for gas flow, dimensionless Side dimension of corrugation for corrugated packing, m Time, hr

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

T TF TCW TRS ug ugf ugl uLS uSL V V0 wGE W xo,j xB,j xF,j xi,j x(D2) x(DT) x(HD) x(HT) x(H2) x(T2) fB x yi,j yf D zF ai-j DP DPo ε mg ml rg rL s xfl xs q j

Temperature, K Feed temperature to flash chamber, K Temperature of cooling water in reflux condensor, K Temperature of heating steam in reboiler, K Gas superficial velocity, m s1 Gas superficial velocity at flooding point, m s1 Gas superficial velocity at loading point, m s1 Liquid superficial velocity, m s1 Minimum liquid wetting rate, m3 m2 h1 Molar vapour flow rate in rectifying section, mol h1 Molar vapour flow rate in stripping section, mol h1 Real gas velocity through corrugated packing, m s1 Molar flow rate of bottom product in continuous or flash distillation, mol h1 Mole fraction of jth species in the top liquid product, dimensionless Mole fraction of jth species in the bottom liquid product, dimensionless Mole fraction of jth species in feed stream, dimensionless Mole fraction of jth species in liquid on ith tray, dimensionless Mole fraction of D2 in any liquid stream, dimensionless Mole fraction of DT in any liquid stream, dimensionless Mole fraction of HD in any liquid stream, dimensionless Mole fraction of HT in any liquid stream, dimensionless Mole fraction of H2 in any liquid stream, dimensionless Mole fraction of T2 in any liquid stream, dimensionless Composition vector of the liquid stream from a single stage flash operation, dimensionless Mole fraction of jth species in vapour leaving ith tray, dimensionless Composition vector of the vapour stream from a single stage flash operation, dimensionless Mole fraction of species in feed to flash chamber, dimensionless Relative volatility of ith component with reference to jth component, dimensionless Wet pressure drop, Pa Dry pressure drop, Pa Packing void fraction, dimensionless Gas viscosity, Pa s Liquid viscosity, Pa s Vapour density, kg m3 Liquid density, kg m3 Surface tension, N m1 Coefficient for flooding point behaviour of the column, m s2 Coefficient for loading point behaviour of the column, m s2 Corrugation angle with respect to horizontal, degree Fraction of feed flashing into vapour in a single stage, dimensionless

15

references

[1] Miller AI, Spagnolo DA, DeVore JR. Choice of a process design for simultaneous detritiation and upgrading of heavy water for the advanced neutron source. Nucl Tech 1995;112(2):204e13. [2] Kinoshita M. A cryogenic distillation column cascade for a fusion reactor. Fusion Technol 1984;6(3):629e33. [3] Vasaru G. Tritium isotope separation. 1st ed. Florida: CRC Press, Inc; 1993. [4] Busigin A, Sood SK, Kveton OK, Dinner PJ, Murdoch DK, Leger D. ITER hydrogen isotope separation system conceptual design description. Fus Eng Des 1990;13:77e89. [5] Sarigiannis DA. On order reduction in hydrogen isotope distiliation models. Fus Eng Des 1995;28:406e12. [6] Cristescu I, Ioana Cristescu, Poenariu V, Preda I, Stefan L. Model about the separation of molecular species of hydrogen isotopes by cryogenic distillation using the elementary theory of transport phenomena. Fus Eng Des 2000;49e50:799e804. [7] Kinoshita M. An efficient simulation procedure especially developed for hydrogen isotope distillation columns. Fus Sci Tech 1984;6(3):574e83. [8] Kinoshita M, Naruse Y. Separation characteristics of cryogenic distillation column with a feedback stream for separation of protium tritium systems and tritium. Fus Sci Tech 1983;2(3):410e25. [9] Buvat JC, Latge C, Joulia X, Pautrot GP. Simulation of hydrogen isotope separation by cryogenic distillation. Fus Sci Tech 1992;21(2P2):954e9. [10] Enoeda M, Yamanishi T, Yoshida H, Naruse Y, Fukui H, Muta K. Hydrogen isotope separation characteristics of cryogenic distillation column. Fus Eng Des 1989;10:319e23. [11] Iwai Y, Yamanishi T, O’hira S, Suzuki T, Shu WM, Nishi M. HD-T cryogenic distillation experiments at TPL/JAERI in support of ITER. Fus Eng Des 2002;61e62:553e60. [12] Bornea A, Zamfirache M, Stefan L, Stefanescu I, Preda A. Experimental investigation on hydrogen cryogenic distillation equipped with package made by ICIT. Fus Sci Tech 2015;67(2):266e9. [13] Bornea A, Zamfirache M, Stefanescu I, Preda A, Balteanu O, Stefan I. Investigation related to hydrogen isotopes separation by cryogenic distillation. Fus Sci Tech 2008;54(2):426e9. [14] Trenin VD, Alekseev IA, Baranov IA, Bondarenko SD, Karpov SP, Konoplev KA, et al. Full-scale experimental assembly for hydrogen isotopes separation studies by cryogenic distillation: assembly and results of the studies. Fus Sci Tech 1995;28(3P1):761e6. [15] Sherman RH. Cryogenic hydrogen isotope distillation for the fusion fuel cycle. Fus Sci Tech 1985;8(2P2):2175e83. € rr L, Hellriegel G, Michling R, [16] Cristescu I, Cristescu IR, Do Murdoch D, et al. Experiments on water detritiation and cryogenic distillation at TLK; impact on ITER fuel cycle subsystems interfaces. Fus Sci Tech 2008;54(2):440e5. [17] Kinoshita M. Effects of helium on separation characteristics of cryogenic distillation column cascade for fusion reactor. Fus Sci Tech 1984;6(3):564e73. [18] Kinoshita M. A cryogenic distillation column cascade for a fusion reactor. Fus Sci Tech 1984;6(3):629e33. [19] Jana AK. Heat integrated distillation operation. Appl Energy 2010;87(5):1477e94. [20] Niida K, Shiroko K, Umeda T. A thermodynamic analysis of heat integration in cryogenic separation process systems. J Chem Eng Jpn 1984;17(4):375e81. [21] Yamanishi T, Okuno K. A computer code simulating multistage chemical exchange column under wide range

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106

16

[22]

[23]

[24]

[25]

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e1 6

of operating conditions. In: JAERI Data/Code 96e028. Japan Atomic Energy Research Institute; 1996. Mittelhauser HM, Thodos G. Vapour pressure relationships upto the critical point of hydrogen, deuterium and tritium and their diatomic combinations. Cryogenics 1964;4(6):363e73. Kolev N. Packed bed columns for absorption, desorption, rectification and direct heat transfer. 1st ed. Amsterdam: Elsevier B.V.; 2006. Perry RH, Green DW. Perry's chemical engineers' handbook. 7th ed. New York: The McGraw-Hill Companies, Inc.; 1999. Dutta BK. Principles of mass transfer and separation processes. 1st ed. New Delhi: Prentice Hall of India Pvt. Ltd; 2007.

[26] http://www0.bnl.gov/magnets/staff/gupta/cryogenic-datahandbook/Section4.pdf, last accessed on 1.12.15. [27] Seader JD, Henley EJ, Roper DK. Separation process principles-chemical and biochemical operations. 3rd ed. New Jersey: John Wiley & Sons, Inc; 2011. [28] Stockfleth R, Brunner G. Holdup, pressure drop, and flooding in packed countercurrent columns for the Gas extraction. Ind Eng Chem Res 2001;40:347e56. [29] Ingham J, Dunn IJ, Heinzle E, Prenosil JE, Snape JB. Chemical engineering dynamics. 3rd ed. Weinheim: Wiley-VCH Verlag GmbH & Co. KGaA; 2007. [30] Treybal RE. Mass transfer operations. 3rd ed. Singapore: McGraw Hill Book Company; 1981.

Please cite this article in press as: Bhattacharyya R, et al., Simulation studies of the characteristics of a cryogenic distillation column for hydrogen isotope separation, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.01.106