scieee Science in your language
[en] (orig)

Effect of multiple liquid inlets on mass transfer in rotating packed beds

Abstract

This study deals with optimalization of rotating packed bed design using computational fluid dynamics approach. Comparison of three variants of liquid distributor were performed on 2D geometry. Turbulence was modelled using unsteady RANS approach and volume of fluid technique were used to simulate gas-liquid interphase. Results were compared on basis of liquid holdup evaluation.

Read accessible full text

Effect of multiple liquid inlets on mass transfer in rotating packed beds

Author: Elcner, Jakub; Jícha, Miroslav
Publisher: EDP Sciences
Year: 2020
DOI: 10.1051/matecconf/202032802010
Source: https://dspace.vut.cz/bitstreams/04c69c19-52e3-4cb3-a01a-d91a2d77d86d/download
E ec o Mul iple Liquid Inle s on Mass T ans e
in Ro a ing Packed Beds
Jakub Elcne 1,*, and Mi osla Jicha1
1B no Uni e si y o Technology, Facul y o Mechanical Enginee ing, Depa men o
The modynamics and En i onmen al Enginee ing, Technicka 2986/2, 616 69, B no, Czech Republic
Abs ac . This s udy deals wi h op imaliza ion o o a ing packed bed
design using compu a ional luid dynamics app oach. Compa ison o h ee
a ian s o liquid dis ibu o we e pe o med on 2D geome y. Tu bulence
was modelled using uns eady RANS app oach and olume o luid
echnique we e used o simula e gas-liquid in e phase. Resul s we e
compa ed on basis o liquid holdup e alua ion.
1 In oduc ion
Ca bon dioxide CO2 con ibu es wi h app oxima ely 60 % o global wa ming. I we lea e
aside sus ainable sou ces, hen he echnology CCS (Ca bon Cap u e and S o age) becomes
key s a egy o educ ion o CO2 om ossil uels. In he comple e sys em o CCS
echnology he mos expensi e is he CO2 cap u e ha ep esen s abou 70 % o he o al
cos o equi ed 90 % CO2 educ ion. Resea ch and de elopmen ocuses on in ensi ica ion
o he cap u e p ocess. Acco ding o many s udies he mos ad anced and ma u e
echnology when compa ing wi h o he me hods is chemical abso p ion, which is on
he Technology Readiness Le el 9 (TRL9), and also on he sol en egene a ion side.
Abso p ion is ca ied ou in a ious appa a uses, he mos commonly used a e s a ic
columns, which a e illed wi h di e en ypes o packings. Howe e he s a ic columns
o any sp ay sys ems like sc ubbe s a e in p inciple limi ed by g a i a ion and he p ocess
canno be u he in ensi ied. In addi ion, s a ic columns a e ex emely la ge essels, o en
o e 30 me e s in heigh and la ge in diame e , which leads o high in es men cos and
space.
The e o e, he aim is o in ensi y he abso p ion p ocess. This can be eached by using
a o a y abso be called Ro a ing Packed Bed (RPB). Ro a y abso be can de elop up o
wen y imes he g a i a ional accele a ion and subsequen ly ensu e up o one o de
o magni ude highe ans e o CO2 o he sol en . All his wi h dimensions en imes
smalle han when using s a ic columns o sc ubbe s. The main ask in he de elopmen
o RPB is o cla i y he in luence o geome y, o a ional speed, gas inle , p ima y sol en
dis ibu ion, and ype o packing (wi e mesh o me al oam) on liquid dis ibu ion,
and liquid disin eg a ion and ilm o ma ion, i.e. phase in e ace inside o a ing packing.
The mos sui able way o assess he in luence o a ious pa ame e s and o p opose
an op imal solu ion o RPB is o use compu a ional luid dynamics o simula e he p ocess
* Co esponding au ho : [email p o ec ed]u b .cz
© The Au ho s, published by EDP Sciences. This is an open access a icle dis ibu ed unde he e ms o he C ea i e Commons
A ibu ion License 4.0 (h p://c ea i ecommons.o g/licenses/by/4.0/).
MATEC Web o Con e ences 328, 02010 (2020) h ps://doi.o g/10.1051/ma eccon /202032802010
XXII. AEaNMiFMaE-2020
inside he packing and alida e i by lab scale expe imen s. The e iciency o CO2 cap u e
is e alua ed on he basis o a pa ame e called liquid holdup, i.e. an amoun o sol en ha
is e ained inside he packing and o ms an in e phase be ween sol en and gas.
2 Me hods
2.1 Model
Fo compa ison, a classic a angemen o a o a ing packed bed, based on expe imen s
o [1] and nume ical simula ions o [2], we e selec ed. Th ee a ian s o geome y wi h
iden ical wi e packing and ou le placemen bu a di e en a angemen o liquid
dis ibu o s loca ed in he inne wall o he geome y we e c ea ed. A angemen o inle
dis ibu o s and hei size o each a ian o geome y was designed o achie e an equal
dis ibu ion o liquid in o he wi ed packing and hence main ain he same luid eloci y.
The geome ies can be seen in Figu e 1. Each a ian o he model con ains wi e mesh
packing wi h wi es o ci cula shape. The diame e o he wi es is 0.5 mm and he dis ance
be ween he cen e o he adjacen wi es in he ci cum e en ial di ec ion is 3.5 mm.
The in e nal diame e o he packing is 42 mm and he ou e diame e is 82 mm. Packing
con ains 21 concen ic laye s in o al, and he dis ance be ween wo laye s o wi es in he
adial di ec ion is 1 mm. The oid ac ion o he packing is 0.96. Liquid ou low was
p o ided by en ou le s o he size o 3 mm, e enly dis ibu ed on he ou e wall.
Fig. 1. Va ian s o geome y wi h di e en amoun o liquid dis ibu o s.
A wo-dimensional compu a ional g id wi h uns uc u ed quad ila e al elemen s wi h he
a e age size o elemen s 0.00015 m was gene a ed. A p isma ic laye consis ing o 8 laye s
we e gene a ed on he inne , ou e wall, and wi es o packing o be e ea men o nea -
wall bounda y laye . The a e age size o he g id o all a ian s o geome y was 550 000
o cells. De ails o he compu a ional g id (inle and wi es on he a ian i1) can be seen in
Figu e 2.
Fig. 2. De ail o compu a ional g id.
2.2 Nume ical se ing
Nume ical calcula ions we e pe o med using Simcen e S a -CCM+ sol e in e sion
2019.2 build 14.04.011. Assump ions o calcula ions we e ha he liquid and gas phase
we e bo h New onian and incomp essible wi h no phase change and he whole p ocess is
2
MATEC Web o Con e ences 328, 02010 (2020) h ps://doi.o g/10.1051/ma eccon /202032802010
XXII. AEaNMiFMaE-2020
inside he packing and alida e i by lab scale expe imen s. The e iciency o CO2 cap u e
is e alua ed on he basis o a pa ame e called liquid holdup, i.e. an amoun o sol en ha
is e ained inside he packing and o ms an in e phase be ween sol en and gas.
2 Me hods
2.1 Model
Fo compa ison, a classic a angemen o a o a ing packed bed, based on expe imen s
o [1] and nume ical simula ions o [2], we e selec ed. Th ee a ian s o geome y wi h
iden ical wi e packing and ou le placemen bu a di e en a angemen o liquid
dis ibu o s loca ed in he inne wall o he geome y we e c ea ed. A angemen o inle
dis ibu o s and hei size o each a ian o geome y was designed o achie e an equal
dis ibu ion o liquid in o he wi ed packing and hence main ain he same luid eloci y.
The geome ies can be seen in Figu e 1. Each a ian o he model con ains wi e mesh
packing wi h wi es o ci cula shape. The diame e o he wi es is 0.5 mm and he dis ance
be ween he cen e o he adjacen wi es in he ci cum e en ial di ec ion is 3.5 mm.
The in e nal diame e o he packing is 42 mm and he ou e diame e is 82 mm. Packing
con ains 21 concen ic laye s in o al, and he dis ance be ween wo laye s o wi es in he
adial di ec ion is 1 mm. The oid ac ion o he packing is 0.96. Liquid ou low was
p o ided by en ou le s o he size o 3 mm, e enly dis ibu ed on he ou e wall.
Fig. 1. Va ian s o geome y wi h di e en amoun o liquid dis ibu o s.
A wo-dimensional compu a ional g id wi h uns uc u ed quad ila e al elemen s wi h he
a e age size o elemen s 0.00015 m was gene a ed. A p isma ic laye consis ing o 8 laye s
we e gene a ed on he inne , ou e wall, and wi es o packing o be e ea men o nea -
wall bounda y laye . The a e age size o he g id o all a ian s o geome y was 550 000
o cells. De ails o he compu a ional g id (inle and wi es on he a ian i1) can be seen in
Figu e 2.
Fig. 2. De ail o compu a ional g id.
2.2 Nume ical se ing
Nume ical calcula ions we e pe o med using Simcen e S a -CCM+ sol e in e sion
2019.2 build 14.04.011. Assump ions o calcula ions we e ha he liquid and gas phase
we e bo h New onian and incomp essible wi h no phase change and he whole p ocess is
iso he mal. Su ace ension be ween hese phases is cons an and uni o m and i s alue is
0.0728 N/m. In e ac ion be ween gas (ai : densi y = 1.18415 m3/kg; dynamic iscosi y =
1.85508E-5 Pa/s) and liquid (wa e : densi y = 998.2 m3/kg; dynamic iscosi y = 1.003E-
3 Pa/s) phase was modelled using Volume o Fluid (VOF) mul iphase low model
by acking he gas-liquid in e ace. Because he b eakup o he liquid du ing
he in es iga ed p ocess equi es high esolu ion o he compu a ional g id and hence
demands high compu a ional powe , he RPB had o be modelled on he wo-dimensional
domain. Tu bulence was modelled using he uns eady RANS app oach wi h SST k-ω model
wi h he comp essibili y co ec ion op ion ac i a ed. A ime s ep was se o 1E-5 s wi h
10 inne i e a ions and simula ion was calcula ed o 1.5 s. The con ac angle on he wi es
was se o 150° which co esponds o a hyd ophilic ma e ial, con e sely, he con ac angle
on he walls o RPB was se o 30° (hyd ophobic ma e ial). Mo emen o he wi e packing
was modelled using a mo ing e e ence ame wi h he axis o o a ion in he cen e o he
o a ional domain and he o a ional speed o 1000 pm. 100 % o he liquid phase was se
he inle s o liquid loca ed on he inne wall o he RPB. The eloci y o he luid a he
inle s was 2 m/sec. The iden ical mass low a e in o he domain o all o he h ee a ian s
o RPB was moni o ed using mass low epo and i was con olled by he dimension
o inle nozzle.
2.3 Valida ion
Calcula ions we e alida ed using expe imen al da a om [1], nume ical simula ions o [2],
and wi h he co ela ion p oposed by [3] Values o physical p ope ies we e se acco ding
o expe imen s o [1]. A compa ison o he liquid holdup (su ace a e age o olume
ac ion o wa e ) can be seen in Figu e 3.
Fig. 3. Valida ion o nume ical se ings
The liquid holdup was es ima ed o he whole geome y (s a ic and o a ional domain)
wi h one inle . Ou esul s appea o be o e es ima ed in compa ison wi h he expe imen s
while he calcula ions o [2], unde es ima ed i . The o e all ma ch wi h expe imen s
is sligh ly be e o ou calcula ions.
3 Resul s and discussion
Resul s we e compa ed based on liquid holdup es ima ed o he packing a ea ( o a ional
domain) o neglec he in luence o he liquid agglome a ion on ou e walls o he domain.
Figu e 4 shows he esul o 1.5 s o phenomena.
3
MATEC Web o Con e ences 328, 02010 (2020) h ps://doi.o g/10.1051/ma eccon /202032802010
XXII. AEaNMiFMaE-2020
Fig. 4. Compa ison o liquid holdup o all a ian s o liquid dis ibu ion.
The in luence o he numbe o liquid dis ibu o s on he liquid holdup can be seen om
he esul s. I akes 0.25 s o ini ia e and s abilize he cou se o liquid holdup alues.
A e his ime, a solu ion can be conside ed o be s able wi h small oscilla ions. While he
cases wi h one (i1) o wo (i2) liquid dis ibu o s e inced minimal di e ences in liquid
holdup du ing he in ensi ica ion p ocess, he case wi h ou dis ibu o s (i4) showed be e
pe o mance. Whe eas he same amoun o wa e is deli e ed o he sys em, he di e ences
in liquid holdup mus be caused by he di e en numbe o liquid dis ibu o s. F om he
ac ual layou o wa e in he packing (Figu e 5) can be seen ha he a ea o liquid
dispe sion occu ence g ows wi h he numbe o dis ibu o s, which has a posi i e e ec
on he liquid holdup alues.
Fig. 5. The ac ual olume ac ion o wa e o = 1.5 s.
None heless, his was only an ini ial s udy on he liquid dis ibu o numbe and i s
in luence on he liquid holdup. Since i was pe o med in 2D and i has some simpli ying
assump ion men ioned abo e, i needs o be con i med du ing he nex esea ch.
This wo k was inancially suppo ed by he p ojec “Compu e Simula ions o E ec i e Low-
Emission Ene gy” unded as p ojec No. CZ.02.1.01/0.0/0.0/16_026/0008392 by Ope a ional
P og amme Resea ch, De elopmen and Educa ion, P io i y axis 1: S eng hening capaci y o high-
quali y esea ch. The inancial suppo is g a e ully acknowledged.
Re e ences
1. Y. C. Yang, Y. Xiang, G. W. Chu, H. K. Zou, Y. Luo, M. A owo, J. F. Chen, Chem.
Eng. Sci. 138 (2015)
2. P. Xie, X. Lu, X. Yang, D. Ingham, L. Ma, M. Pou kashanian, Chem. Eng. Sci. 172
(2017)
3. J. R. Bu ns, J. N. Jamil, C. Ramshaw, Chem. Eng. Sci. 55 (2000)
4
MATEC Web o Con e ences 328, 02010 (2020) h ps://doi.o g/10.1051/ma eccon /202032802010
XXII. AEaNMiFMaE-2020