Supe c i ical wa e oxida ion o ene gy p oduc ion by
hyd o he mal lame as in e nal hea sou ce.
Expe imen al esul s and ene ge ic s udy
Pablo Cabeza, Joao Paulo Sil a Quei oz1,, Manuel C iado, C is ina Jimenez,
Ma ia Dolo es Be mejo∗
, Fidel Ma o, Ma ia Jose Coce o
High P essu e P ocess G oup, Dep . o Chemical Enginee ing and En i onmen al
Technology - Uni e sidad de Valladolid - Doc o Me gelina, s/n, 47011, Valladolid - Spain
Abs ac
This wo k p esen s expe imen al and model esul s om a new configu a ion o
a cooled wall eac o wo king wi h wo ou le s: an uppe ou le h ough which
a sal - ee ho effluen (500 - 600 ◦C) is ob ained and a lowe ou le h ough
which an effluen a subc i ical empe a u e dissol ing he p ecipi a ed sal s
is ob ained. Diffe en flow dis ibu ions we e es ed in o de o find he bes
elimina ion condi ions. To al o ganic ca bon emo al o e 99.99% was ob ained
a injec ion empe a u es as low as oom empe a u e, when he ac ion o
p oduc s lea ing he eac o in he uppe effluen is lowe han 70% o he
eed flow. The pe o mance o he eac o was es ed wi h he oxida ion o
a ecalci an compound such as ammonia, using isop opyl alcohol as co- uel.
Remo als highe han 99% o N-NH+
4we e achie ed in bo h effluen s, wo king
wi h empe a u es nea 700 ◦C. Sligh ly be e elimina ions we e ob ained in
he bo om effluen because i s esidence ime in he eac o is longe . The
beha io o he eac o wo king wi h eeds wi h a high concen a ion o sal s
was also es ed. Feeds con aining up o 2.5% w Na2SO4could be injec ed
∗Co esponding au ho . Phone: +34 983423166
Email add esses: [email p o ec ed] (Pablo Cabeza), [email p o ec ed]
(Joao Paulo Sil a Quei oz), [email p o ec ed] (Manuel C iado),
[email p o ec ed] (C is ina Jimenez), [email p o ec ed] (Ma ia Dolo es Be mejo),
[email p o ec ed] (Fidel Ma o), [email p o ec ed] (Ma ia Jose Coce o)
1P esen add ess: Dep . o Chemical Enginee ing - Uni e sidade Fede al de Pe nambuco -
P o . A u de S´a, s/n, - Cidade Uni e si ´a ia - 50740-521, Reci e, PE - B azil
P ep in submi ed o Ene gy Ma ch 2, 2015
in he eac o wi hou plugging p oblems and a o al o ganic ca bon emo al
o 99.7% was achie ed in hese condi ions. Uppe effluen always p esen ed a
concen a ion o sal lowe han 30 ppm. Finally, a heo e ical analysis o he
ene gy eco e y o he eac o wo king wi h wo ou le s was made.
Keywo ds: Supe c i ical wa e oxida ion, Hyd o he mal flames, Renewable
Ene gy, Reac o design
1. In oduc ion1
Since F anck and cowo ke s disco e ed hyd o he mal flames [1] and i could2
be applied o he Supe c i ical Wa e Oxida ion (SCWO), new challenges came3
up o he s udy o SCWO. Fo flammable compounds such as me hane o 4
me hanol, hyd o he mal flame can occu a empe a u es as low as 400 ◦C [2].5
SCWO is he oxida ion o o ganics in wa e unde condi ions abo e i s c i ical6
poin . In p esence o hyd o he mal flames he eac ion imes can be educed o7
he o de o milliseconds [3] wi hou he p oduc ion o sub-p oduc s ypical o 8
con en ional combus ion such as NOx [4] o dioxins [5].9
SCWO wi h a hyd o he mal flame has a numbe o ad an ages o e he10
flameless p ocess. Some o hese ad an ages pe mi o e coming he adi ional11
challenges ha make he success ul and p ofi able comme cializa ion o SCWO12
echnology difficul . The ad an ages include he ollowing [3]:13
•The educed esidence imes (in he o de o milliseconds) allows he con-14
s uc ion o smalle eac o s.15
•I is possible o ca y ou he eac ion wi h eed injec ion empe a u es16
nea o oom empe a u e when using essel eac o s [6, 7]. This a oids17
p oblems such as plugging and co osion in a p ehea ing sys em, ha ing18
an ad an age om he ope a ional and ene gy in eg a ion pe spec i e.19
•Highe ope a ion empe a u es imp o e he ene gy eco e y.20
The fi s eac o p obably wo king wi h a hyd o he mal flame inside was21
he MODAR eac o , wo king in condi ions o concen a ion, empe a u e and22
2
p essu e abo e he igni ion condi ions o me hanol and being able o wo k wi h23
injec ion empe a u es o 25 ◦C and injec ing he ai a 220 ◦C [7]. In he24
ETH o Zu ich, he di ec injec ion o he was e in o a diffusion hyd o he mal25
flame gene a ed inside he eac o was de eloped as a solu ion o a oid he26
ex e nal p ehea ing o he was e up o supe c i ical condi ions [8, 9]. P ´ıkopsk´y27
and cowo ke s in es iga ed he easibili y o injec ing eeds wi h a 3%w o 28
sodium sul a e (Na2SO4) in he anspi ing wall eac o (TWR) wi h a diffusion29
hyd o he mal flame as in e nal hea sou ce [10]. No plugging was obse ed30
du ing he expe imen s, bu sal deposi s we e de ec ed in he uppe ho zone31
o he eac o . In a p e ious in es iga ion o ou esea ch g oup [6], i was32
ound ha using a anspi ing wall eac o , a p emixed hyd o he mal flame33
inside he eac ion chambe could be main ained when injec ing he eed a a34
empe a u e as low as 110 ◦C. Using a simila eac o , eeds wi h up o 4.74% w 35
Na2SO4could be injec ed [11]. The eac o wo ked wi hou plugging, bu he36
eco e y o sal s was only be ween 5% and 50%. Bo h esea ch g oups epo ed37
an inc ease in he empe a u e when sal was injec ed in he eac o [10, 11].38
Zhang e al. [12] s udied he ope a ional pa ame e s o a TWR de eloped o39
gene a e he mal fluids o oil eco e y. They used wa e -me hanol as a ificial40
uel p io o ea ing oil explo a ion was ewa e , and hey ound he limi s o 41
empe a u e o anspi ing flow in o de o a oid he quenching and ex inc ion42
o hyd o he mal flame.43
I has been p o ed ha injec ion o cold eeds o e a hyd o he mal flame is44
only possible when wo king wi h essel eac o s [9, 10, 11] and i is no possible45
when wo king wi h ubula eac o s [13]. This beha io was due o he low46
flame on eloci ies in hyd o he mal flames ha is lowe han 0.1 m/s, in47
compa ison o he highe flame on eloci ies a a mosphe ic condi ions (0.4-348
m/s). This is he eason why flow eloci ies lowe han 0.1 m/s a e necessa y49
o keep a s able hyd o he mal flame whe e cold eagen s can be injec ed [14].50
Ou esea ch g oup has succeeded in keep wo king con inuously a essel eac o 51
injec ing eeds a empe a u es as low as 25 ◦C [15].52
E en hough he mos immedia e applica ion o hyd o he mal flames is in53
3
he SCWO p ocess o was e des uc ion, which is he mos indus ially de el-54
oped hyd o he mal p ocess, i is possible o mo e om he idea o hyd o he mal55
flame as a echnology o he des uc ion o was es o conside i as a echnol-56
ogy o he gene a ion o clean ene gy, which could e en ually subs i u e he57
ac ual echnologies based on a mosphe ic combus ion [16]. Supe c i ical wa e 58
is al eady applied in ene gy fields h ough gasifica ion p ocesses o was e al-59
o iza ion Facchine i e al. [17], R¨onnlund e al. [18]. The efficiency in ene gy60
p oduc ion by SCWO o coal and di ec expansion o he effluen was compa ed61
o he efficiency o o he con en ional powe plan s by Be mejo e al. [19]. I 62
he s eam was p oduced a 650 ◦C and 30 MPa, efficiencies as high as 38% we e63
ob ained by SCWO. Efficiency was as high as 41% i he effluen was ehea ed64
and expanded a second ime. The efficiencies a he same s eam condi ions o 65
pul e ized coal powe plan and p essu ized fluidized bed powe plan we e 3266
and 34% espec i ely. Compa ison is mo e a o able using oxygen en iched ai 67
o e en using pu e oxygen as he oxidan . In his las op ion he cos o he oxi-68
dan mus be assumed. Ne e heless, i is known ha in adi ional combus ion69
powe plan s, oxygen is used o imp o e efficiency. Dona ini e al. [20] simula ed70
a powe plan based on di ec combus ion o pul e ized coal in a SCWO eac o 71
wi h a sys em o CO2cap u e. They epo ed ne efficiencies a ound 27% and72
ound ha he consump ion o he ai sepa a ion uni o oxygen p oduc ion73
s ongly affec s he iabili y o he plan . A simila analysis has been made by74
Ko owicz and Michalski [21], whom ha e p oposed se e al ope a ions in o de 75
o inc ease efficiency o each s ep in a powe plan model: ai sepa a ion uni ,76
boile (SCWO eac o bu ning coal) and s eam u bine.77
A ai e al. [22] p oposed he supe c i ical oxida ion o biomass was es and78
o he sus ainable uels wi h a hyd o he mal flame as a clean ene gy sou ce o 79
eaching a sus ainable socie y wi h a decen alized p oduc ion based on enew-80
able esou ces. Augus ine and Tes e [3] also p opose i s u iliza ion wi h low81
g ade uels. In gene al, his echnology can be applied o he alo iza ion o 82
was e such as was e wa e ea men plan sludge, biomass o plas ic was es o 83
any kind o was e wi h high ene ge ic con en . Basic heo e ical calcula ions84
4
indica e ha eeds wi h an ene gy con en o 930 kJ/kg ( oughly equi alen 85
o an aqueous solu ion wi h 2% ww o hexane) can supply enough ene gy o86
p ehea he eed om oom empe a u e up o 400 ◦C, and o gene a e elec-87
ic powe equi alen o ha consumed by he high p essu e pump and he ai 88
comp esso [23]. A ema kable aspec abou wo king wi h hyd o he mal flames89
is imp o ing ene gy eco e y in SCWO sys em [19]. Hyd o he mal flames allow90
new eac o designs ha no only a e able o injec eeds wi hou p ehea ing91
because o he possibili y o injec ing eac an s a oom empe a u e bu also92
use he hea eleased by he flame o o he pu poses as he ene ge ic in eg a ion93
o he p ocess o o p oduc ion o elec ici y by u bines [24]. Smi h J . e al.94
[25] used exe gy analysis o s udy he pa ial and o al oxida ion o me hane in95
supe c i ical wa e o a hea -in eg a ed supe c i ical wa e eac o and elec i-96
cal ene gy p oduc ion sys em. They assume a di ec expansion o p oduc s (a 97
400 ◦C) in a u bine, ollowed by hea eco e y o he expanded s eam. I was98
ound ha he p ocess could be ene gy sel -sufficien and op imum flow a es99
we e calcula ed in o de o minimize eac o hea equi emen s o maximize ne 100
elec ical wo k. The high empe a u e effluen can also be used as hea sou ce101
in o he hyd o he mal p ocesses, such as lique ac ion o gasifica ion, whe e he102
hea eco e y is a c i ical issue [26, 27]. In he case o was e wi h high con-103
cen a ion o ino ganic subs ances, new eac o designs able o sepa a e hese104
sal s om he effluen mus be de eloped in o de o make i possible o di ec ly105
expand he effluen in an elec ici y p oduc ion u bine.106
The main goal o his wo k is he s udy o he beha io o new cooled wall107
eac o wi h he main pa icula i y o ha ing wo ou le s in o de o y o keep108
he maximum hea eleased by he flame in a clean and high empe a u e flow109
lea ing he eac o om he uppe zone and o he flow a subc i ical condi ions110
wi h he sal s dissol ed going ou o he bo om o he eac o . In his way he111
uppe /lowe effluen ela ion was op imized aking in o accoun he empe a u e112
p ofiles inside he eac o and he o ganic ma e elimina ion in bo h s eams.113
The pe o mance o he eac o wi h ecalci an pollu an s such as ammonia114
was es ed as well as he pe o mance o he eac o wi h eeds con aining sal s.115
5

Figu e 1: Diag am o SCWO acili y wi h wo ou le s.
A CFD model is also used o desc ibe he beha iou o he eac o . Finally, a116
pu ely heo e ical ene gy eco e y s udy o he p ocess wi h he new eac o was117
pe o med, including he possibili y o di ec expansion in hypo he ical de ices.118
2. Expe imen al119
2.1. Expe imen al se up120
All he expe imen s analyzed in his esea ch ha e been ca ied ou in he121
SCWO acili y ins alled in he Uni e si y o Valladolid. I consis s o a con in-122
uous acili y wo king wi h a eed flow o 22.5 L/h, and ai supplied by a ou 123
s age comp esso , wi h a maximum eed a e o 36 kg/h is used as he oxidan .124
The eac o consis o a p essu e essel made o AISI 316 s ainless s eel able o125
s and a maximum p essu e o 30 MPa and a maximum wall empe a u e o 400126
◦C, con aining a eac ion chambe made o Ni-alloy 625 whe e he empe a u e127
be as high as 700 ◦C. Was e wa e eed and ai a e p e iously p essu ized and128
p ehea ed wi h elec ical esis ances o he desi ed empe a u e be o e being129
injec ed by he bo om o he eac o . The eagen s a e conduc ed o he op o 130
he eac o chambe by means o a ubula injec o . A he ou le o he injec-131
6
o he hyd o he mal flame is o med. Cooling wa e , p e iously p essu ized is132
ci cula ing be ween he p essu e essel and he eac ion chambe in oduced by133
he op o he eac o in o de o cool down he essel a a empe a u e lowe 134
han 400 ◦C. This cooling wa e is en e ing in he eac ion chambe h ough i s135
lowe pa and lea ing he eac o by he bo om oge he wi h a ac ion o he136
p oduc s. The es o he p oduc s lea e he eac o by ano he ou le si ua ed137
in he op o he eac o chambe . A e lea ing he eac o , bo h effluen s a e138
cooled down in he in e coole s and dep essu ized. The flow diag am o he a-139
cili y wi h wo ou le s is shown in Figu e 1. Mo e in o ma ion abou he acili y140
can be ound elsewhe e [6, 13]. Figu e 2 shows a scheme o he eac o wi h he141
diffe en posi ion o he mocouples inside he eac ion chambe . The diffe en 142
empe a u e p ofiles a e e e ed a he posi ion o hese ou he mocouples.143
Each effluen ( op and bo om flow) is measu ed wi h a o ame e in o de o144
know he dis ibu ion o he eed flow espec he wo ou le s.
13 mm
Feed
Ai Bo om
P oduc s
Cooling
wa e
945 mm
520 mm
250 mm
120 mm
Top
P oduc s
T1
T2
T3
T4
Figu e 2: Scheme o he eac o wi h he flow dis ibu ion and he posi ions (mm) o he
empe a u e measu emen inside he eac ion chambe .
145
7
2.2. Ma e ials and expe imen al p ocedu e146
The expe imen s analyzed in his esea ch we e pe o med using eeds p e-147
pa ed wi h isop opyl alcohol (IPA, 99% pu i y) and ap wa e wi hou u he 148
pu ifica ion. Fo expe imen s made wi h ammonia i was used ammonia (25%149
in mass). Syn he ic was e con aining sal s we e p epa ed using Na2SO4(pu i y150
>98%).151
P e ious o he beginning o he expe imen he eac o mus be p ehea ed152
elec ically o 400 ◦C. The eac ion is ini ia ed by injec ed ai and was e wa e 153
s eams p ehea ed elec ically up o a empe a u e highe han 400 ◦C. A ew154
minu es a e con inuous injec ing o IPA solu ion and ai s eam he hyd o he -155
mal flame is igni ed. A ha momen a sha p inc ease he empe a u es a he156
op o he eac o (T1 and T2) is egis e ed. Then, he elec ical hea ing o he157
wall o he eac o is u ned off and he cooling wa e flow is connec ed. Fo 158
keeping he maximum empe a u e cons an in alues a ound 600-700 ◦C ill159
he desi ed injec ion empe a u e is eached, IPA concen a ion was inc eased as160
he injec ion empe a u e was dec eased down o he selec ed injec ion empe -161
a u e. A e he a ge injec ion empe a u e is eached ( om 300 ◦C ill oom162
empe a u e, a ound 20-30 ◦C), he uppe flow and bo om flow a e egula ed163
opening o closing he decomp ession al es keeping he ai and he p essu e164
cons an . P essu e mus be s abilized a ound 23 MPa. Se e al s a iona y s a es165
wi h diffe en p epa ed eeds and diffe en flow up/bo om a io a e eached and166
samples o he liquid effluen a e aken.167
To al O ganic Ca bon (TOC) and To al Ni ogen (TN) analysis o he sam-168
ples we e pe o med wi h a TOC 5050 SHIMADZU To al O ganic Ca bon Ana-169
lyze which uses combus ion and IR analysis. The de ec ion limi is 1 ppm. Sal 170
concen a ion is measu ed using a conduc ime e Basic 30 p o ided by C ison.171
Fo doing his, conduc i i ies o solu ions o known Na2SO4concen a ion a e172
measu ed ob aining a linea calib a ion line be ween conduc i i y and Na2SO4
173
concen a ion. Ni a es and ni i es we e cha ac e ized in he liquid effluen by174
ionic ch oma og aphy wi h an IC PAK A column o Wa e s. The de ec ion limi 175
is 1 ppm. NH3and NOxa he gas effluen we e analyzed wi h D ¨age ubes176
8
de ec o s Lab Sa e y Supply CH29401 and CH31001. The NOxde ec ion limi s177
o hese ubes anged om 0.5 o 100 ppm and he NH3de ec ion limi s anged178
om 5 o 70 ppm (s anda d de ia ion o bo h ubes a e be ween 10 and 15%).179
3. Modeling180
A CFD model was pe o med in o de o s udy he in e nal beha io o he181
new eac o . The main elemen s o he eac o ha e been included in he model182
geome y, like he injec o , he eac ion chambe , and he space be ween he183
p essu e shell and he chambe . The eac o is modeled as an axisymme ic 2D184
sys em. The u bulen flow dynamics is modeled by Reynolds-A e aged-Na ie -185
S okes equa ions, using he Realizable k-ϵ u bulence model wi h enhanced wall186
ea men [28]. The densi y o he supe c i ical mix u e is calcula ed by Peng-187
Robinson equa ion o s a e wi h Van de Waals mixing ules, and olume ans-188
la ion (VTPR-EoS) [29]. The olume ansla ion used in densi y calcula ions189
was fi ed o each componen ha cons i u es he sys em (H2O, O2, N2, CO2
190
and IPA), a he ope a ion p essu e o 23 MPa. The olume ansla ion has191
no influence on en halpy calcula ions, hus, specific en halpy (and also cp) is192
gi en by o iginal Peng-Robinson equa ion o s a e (PR-EoS) [30]. The he -193
mal conduc i i y and he molecula iscosi y o he mix u e a e calcula ed as194
a mass- ac ion a e age o he p ope ies o he pu e componen s as unc ion195
o empe a u e. The u bulen diffusion usually o e whelms lamina diffusion,196
and he specifica ion o de ailed lamina diffusion p ope ies in u bulen flows197
is no necessa y. E en so, lamina diffusion coefficien a e es ima ed using he198
me hod o Ma hu and Thodos [31].199
4. Resul s and discussion200
As gene al esul , he new eac o wi h wo ou le s success ully elimina es201
o ganic ma e ial and p o ides a clean s eam wi h high ene gy con en . The202
injec ion a low empe a u es (20 ◦C), a om he c i ical egion, keeps he203
sal s dissol ed inside he injec o , a oiding plugging and co osion. Finally,204
9
!!
"!!
#!!
$!!
%!!
&!!
! !! #!! %!! '!! (!!!
!"#$%
&"''%
$) *+,-. %)'*+,-.
/)(*+,-. 01234567*685935
(a)
!
"
#
$
%
&
% ' ( !!
!"#!$%&'%#())*+
",,&-./#01%'2#3&,0#(4/56+
)*+,-.//.0123304 )*+,/.3123304
(b)
Figu e 9: (a) Tempe a u e p ofiles o diffe en cooling wa e flows. (b) TOC alues in op
and bo om effluen s o diffe en cooling wa e flows.
we e compa ed wi h esul s ob ained wi h mix u es o ammonia and IPA es ed285
in he same eac o wo king wi h only he bo om ou le [4]. Figu e 10 shows286
Ammonia and TOC emo al ep esen ed e sus maximum empe a u e egis-287
e ed inside he eac o . The uppe effluen ac ion was kep cons an a alues288
a ound 50% which means ha he 50% o he eed injec ed (liquid) has been289
aken ou by he uppe ou le . I can be app ecia ed ha empe a u es highe
!
"
#
$
%
&!!
$!! $#! $%! '"! '$! %!!
! "#$%&'()*+,%-./
0%1+2%-34/
()(*#+,-./01/,/-2
()(*#+,-./01/,3-//-4
()(*#+,-51,-./01/
(a)
!"
!#
!$
!%
!&
'""!"
%"" %$" %&" (#" (%" &""
!"#$%&'()*#+,-
#.)/#+0"-
)*+,-./01/,/-2
)*+,-./01/,3-//-4
)*+,-51,-./01/
(b)
Figu e 10: Ammonia emo al (a) and TOC emo al (b) s max empe a u e inside he eac o
o eeds wi h concen a ions be ween 0.5-3% o ammonia and 9-11.5% o IPA wo king wi h
100% bo om flow and wi h 50% op flow.
290
han 700 ◦C a e equi ed o achie e N-NH+
4 emo als o e 99%. These em-291
pe a u es a e highe han hose needed o ob ain he same emo al wi h he292
eac o wo king wi h only one ou le . Table 1 summa izes he a e age esul s293
o emo al o he diffe en expe imen s made wi h mix u es o ammonia and294
IPA. Wo king wi h wo ou le s, i is obse ed ha ammonia emo al is sligh ly295
16

Table 1: Remo al esul s om he expe imen s made wi h diffe en concen a ions o ammonia.
NH+
4o IPAo Tmax TOC TOC N-NH+
4N-NH+
4N-NO−
3N-NO−
3
(%) (%) (◦C) Rem.(%) Rem.(%) Rem.(%) Rem.(%) op bo om
op bo om op bo om (ppm) (ppm)
0.5 11.5 744 99.99 99.99 99.13 99.41 49 38
0.5 10.5 706 99.97 99.96 99.07 99.41 47 21
0.5 10.0 634 99.93 94.71 97.94 91.77 50 14
1.0 10.0 708 99.99 99.99 98.99 99.88 36 74
3.0 9.0 686 99.99 99.97 99.83 99.79 186 78
3.0 9.5 729 99.97 99.98 99.29 99.82 26 27
highe in he bo om effluen han in he op effluen , p obably because he296
esidence ime o he p oduc s comp ising he lowe effluen is longe han he297
one o he op effluen , ha i seems o be oo sho o ha e comple e oxida ion298
o ammonia [4]. Ni a e concen a ion is in gene al highe in he op effluen 299
due o he highe empe a u es. The concen a ions o NOx and NH+
4in he300
gas effluen we e unde he de ec ion limi o 0.5 and 5 ppm espec i ely o all301
he expe imen al condi ions es ed.302
4.7. Beha io o he eac o wo king wi h high sal con en eeds303
The main goal o his new design o he eac o is o ob ain a op effluen 304
a high empe a u e and ee o sal s, becoming his way a ailable o be used305
in sys ems o p oduce ene gy. To achie e ha , sal s con ained in he eed306
mus p ecipi a e and all, lea ing he eac o dissol ed in he bo om effluen 307
while he op effluen is ee o sal s. Fo his pu pose, eeds wi h Na2SO4
308
concen a ions un il 2.5% (25,000 ppm) we e injec ed in he eac o , using IPA309
as uel o ob ain eac ion empe a u es o 700 ◦C, and a eed flow a es o 13-14310
kg/h. In able 2 he main esul s o he expe imen s made wi h eed con aining311
sal s a e summa ized. Equa ion (6) explains how he sal eco e y is calcula ed.312
As can be obse ed in able 2, i is possible o eco e a op effluen almos 313
ee o sal s (wi h conduc i i ies below alues o he ap wa e , equi alen o314
concen a ions o Na2SO4lowe han 30 ppm) and a empe a u es o e 500315
◦C, a ailable o be expanded in a u bine o o he p oduc ion o s eam a high316
17
Table 2: Main esul s o he expe ience made wi h eed con aining 2.5% w o Na2SO4.
F op Fbo om TOC TOC Tmax Tbo om Na2SO4Na2SO4
(kg/h) (kg/h) op bo om (◦C) (◦C) op Reco e y
(ppm) (ppm) (ppm) bo om (%)
7.2 10.2 1.0 209 749 239 24 2.4
7.2 10.2 0.3 352 712 244 23 32.1
7.2 10.2 0.5 599 740 250 24 21.1
7.2 10.2 0.8 23 742 254 23 1.8
7.2 10.2 0.7 69 683 258 23 45.5
7.2 10.2 0.7 16 691 258 26 0.7
A e age 0.7 211 719 251 23.8 17.3
empe a u e ha could be also expanded in a u bine. Paying a en ion o he317
sal eco e y a he bo om flow, i was possible o ob ain an a e age o 17% o 318
sal eco e y. This eco e y is highe han he ob ained wi h he eac o wo king319
wi h only one ou le [32] (a e age o 10%) bu i was no possible o imp o e320
and s abilize he eco e y du ing long imes. This ac could be in e p e ed as321
he possible o ma ion o solid clus e s o sal s swep away by he ou le s eam322
and dissol ed in he cooling sys ems.323
4.8. Ene gy eco e y324
In o de o analyze he possibili y o using he high empe a u e o he325
effluen o SCWO eac o s o p oduce ene gy, an analysis o he op ions o 326
gene a ing ene gy was pe o med.327
4.8.1. Pa ame e calcula ions328
The ollowing equa ions explain how he diffe en pa ame e s o he s udy
we e calcula ed. Fi s ly, he amoun o ene gy eleased by he was e and uel
con ained in he eed is calcula ed as shown in eq. (7).
kW injec ed a he eed = F eedC uel∆Hc, uel
3,600 (7)
whe e ∆Hc, uel is he en halpy o combus ion o IPA (3,750 kJ/kg).
The ene gy consumed is due mainly o he pumping equipmen (pumps and
18
comp esso s). The ac ion o ene gy consumed wi h espec o he ene gy con-
ained in he eed is calcula ed as shown in eq. (8).
Consump ion = kW Consumed
kW injec ed a he eed ·100 (8)
The ene gy p oduc ion is calcula ed using Peng-Robinson Equa ion o S a e
wi h Bos on-Ma hias alpha unc ion conside ing a u bine wi h an isen opic
efficiency o 72 %. The ac ion o ene gy p oduced wi h espec o he ene gy
in oduced in he eed is calcula ed as shown in eq. (9).
Rela i e p oduc ion = kW p oduced
kW injec ed a he eed ·100 (9)
kW p oduced is he ene gy p oduced by di ec expansion o s eam expansion
p oduc ion.
F om he p oduc ion and consump ion is ob ained he pe cen age o he effi-
ciency in ene gy p oduc ion o he sys em o each eac o and kind o oxidan
as shown in eq. (10).
Efficiency = kW p oduced −kW Consumed
kW injec ed a he eed ·100 (10)
Mass and ene gy balances we e sol ed in Aspen Plus so wa e conside ing he -329
mal and olume ic p ope ies calcula ed using Peng-Robinson Equa ion o 330
S a e.331
4.8.2. Ene gy p oduced by s eam expansion332
The mos con en ional me hod o elec ic gene a ion is using he p oduc s333
s eam as hea sou ce o a Rankine cycle. As a guidance o easibili y o he334
p ocess, i has been calcula ed he amoun o s eam which could be gene a ed a 335
diffe en condi ions o h ee small u bines comme cialized by Siemens ( able336
3).337
Inc easing he p essu e o he powe cycle, inc eases he specific wo k p o-338
duced by expanding he s eam. Howe e , he o al amoun o s eam is educed,339
since he hea sou ce is limi ed. Tha can be seen in figu e 11, ha shows he340
composi e cu es o he ho s eam ( eac o p oduc s) and h ee possibili ies o 341
19
Table 3: Cha ac e is ics o comme cial s eam u bines.
Inle P Inle T Powe
(ba ) (◦C) (kWh/(kg-s eam))
SST-040 40 400 0.232
SST-050 101 500 0.272
SST-060 131 530 0.278
cold s eam (wa e -s eam), wi h a diffe ence o 10 ◦C a pinch poin . The hea 342
sou ce is a s eam a 700 ◦C and 23 MPa, wi h he ypical composi ion o SCWO343
effluen using ai as oxidan . Finally, figu e 12 p esen s he amoun o s eam344
ha could be p oduced pe kilog am o ho p oduc s and he ne efficiency o 345
each p essu e le el.
!
"
#
$
%
&
'
(
% ! !%
!"#!$%&'$!()*+,
-.&/%0#1()234/,
)*+,-./0
1123 $
1123 %
1123 &
Figu e 11: P oduc ion o s eam a h ee diffe en p essu e le els.
!"#$# !"#%#
!"#&#
"'#(
")%(
")#(
#
#*)
#*'
#*+
#*$
#*%
#*&
# % )# )%
!!"#"$%#&
'()*+$,-.'()/0123#+
40$**30$5674,8
,-./0
.11232.435
Figu e 12: S eam p oduc ion and efficiency o comme cial u bines.
346
Gi en hese esul s, he ollowing sec ions assume ha he cha ac e is ics347
20
o he s eam ha can be p oduced om he hea con ained in he effluen o 348
he eac o a e hose shown in able 4. The low p essu e s eam p esen ed in
Table 4: Cha ac e is ics o he s eam.
P essu e Tempe a u e
(ba ) (◦C)
High P essu e S eam 46 400
Low P essu e S eam 10 180
349
able 4 could be p oduced wi h eac o effluen s a empe a u es up o 400 ◦C350
( ypical effluen empe a u es o essel eac o s like he anspi ing wall eac o 351
and cooled wall eac o ) and he high p essu e s eam by he effluen s up o 700352
◦C (effluen s o ubula eac o s o he effluen o he new cooled wall eac o 353
desc ibed in his wo k).354
Wi h hese assump ions, he possibili ies o ene gy eco e y o some eac o 355
designs a e compa ed.356
Compa ison o he eco e y ene gy o diffe en eac o s by s eam p oduc ion:357
•Tubula eac o [4, 13]358
This eac o consis ed on a s aigh and emp y ube made o Ni alloy359
C-276 wi h a o al leng h o 5400 mm and a diame e o 1/4” (i.d. 3.86360
mm) gi ing an in e nal olume o 63.2 ml and i was he mally isola ed.361
In his case (figu e 13), he effluen is used fi s ly o p ehea he eed un il362
he injec ion empe a u e (a ound 400 ◦C) and he emaining hea flow is363
used o p oduce s eam.
!"!#$%&%'$()*%
+'',
-%*,!().
/*0&-&.)'$1
2%*,!()3*4
Figu e 13: Sys em eco e y design o a ubula eac o .
364
21

•O iginal cooled wall eac o designed in he Uni e si y o Valladolid (Val-365
ladolid, Spain) [33]366
This eac o is composed by wo concen ic ubes; he inne one is made367
o Inconel 625, and he ou e shell is made o SS 316. Oxida ion eac ion368
akes place inside he inne ube ( eac ion chambe ). In he gap be ween369
bo h ubes, he p essu ized eed s eam is going down and cooling he370
eac ion media a he same ime. In such way, he inne ube does no 371
wi hs and any p essu e a all; ha ing he same p essu e in one side han372
on he o he , and he hickness o he inne ube (Inconel 625) can be373
educed. The effluen o he o iginal CWR (a 400 ◦C) can be used in a374
Rankine cycle as i is shown in figu e 14.
!"
#$$% &'(%)*+,
Figu e 14: Sys em eco e y design o he o iginal CWR.
375
•New cooled wall eac o design [32]376
The new eac o consis s o a e ical Ni-alloy eac ion chambe ha is377
inside o a p essu e essel made o AISI 316 able o s and a maximum378
p essu e o 30 MPa and 400 ◦C. Be ween he walls o he wo essels a379
s eam o cold wa e e ige a es he eac ion essel. The eagen s ( eed380
and oxidan ) a e in oduced in he eac o h ough a ubula injec o up o381
he op o he eac ion chambe . The flame is p oduced ou side o he in-382
jec o , no mally a he op o he eac ion chambe , whe e he maximum383
empe a u e is egis e ed. Reac ion chambe is e ige a ed wi h oom384
empe a u e p essu ized wa e ha flowed be ween he eac ion chambe 385
wall and he inne wall which suppo ed he p essu e, keeping he p essu -386
ized wall a empe a u es lowe han 400 ◦C, and en e ing in he eac ion387
chambe by i s lowe pa mixing wi h he eac ion p oduc s. The p od-388
22
uc s flowed down he eac o lea ing i by i s lowe pa oge he wi h he389
cooling wa e . Following he idea o he o iginal CWR, he p oduc s come
!"#$%&
'#()*+!*
,!!-
./(-)0*1
$((+234
"5*!/
6("#.#1*!57
8/(-)0*2(3
Figu e 15: Sys em eco e y designed o he new CWR wi h he configu a ion wi h one ou le .
390
ou he eac o a empe a u es a ound 325 ◦C and can p oduce s eam o391
be expanded in a Rankine cycle (figu e 15).392
O he configu a ion o his eac o is wo king wi h an ou le a he op o 393
he eac o , hus ha ing wo ou le s: one a high empe a u e and o he 394
a subc i ical empe a u es wi h he sal dissol ed wi h he cooling wa e 395
(figu e 16). Fo pe o ming he compa a i e ene ge ic analysis among he
!"#$%&
'#()*+!*,
-!!.
/(**(0#12(.)3*
$((+456
"7*!2
8("#9#,*!70
12(.)3*4(5
:(1#12(.)3*
;46<#9#,*!70
12(.)3*4(5
Figu e 16: Sys em eco e y designed o he new CWR wi h he configu a ion wi h wo ou le s.
396
diffe en eac o s ypes, ypical ope a ional pa ame e s o each eac o 397
such as uel concen a ion, oxidan excess o e he s oichiome ic amoun 398
(based on he a e age excess used in he majo i y o he expe imen s and399
he accep able oxidan excess o oxidize ni ogen compounds), he pe -400
cen age o cooling wa e and he effluen empe a u e we e fixed. These401
pa ame e s a e shown in able 5. In fi s place, he analysis was pe o med402
conside ing ha he effluen s a e used o gene a e s eam o a Rankine403
cycle. The esul s a e also shown in able 5, a he las wo columns.404
23
As can be obse ed, p oducing elec ici y h ough Rankine cycles p esen
Table 5: Condi ions fixed o he s udy o each eac o , eco e ing ene gy h ough a Rankine
cycle. 5% excess o oxidan is assumed in all cases.
Type o eac o
Hea Cooling Injec ion Effluen Efficiency
flow eed wa e T (◦C) T (◦C)
(kW) (% o eed) Ai O2
O iginal CWR 1,202 0 Room T 400 -16.9% 8.6%
New CWR 1 ou le 1,202 35 Room T 325 -16.8% 4.8%
Tubula eac o 372 0 350-400 700 -21.1% 5.7%
New CWR 2 ou le s 1,208 35 Room T Top 700 -8.0% 19.0%
100% Bo . 300
New CWR 2 ou le s 1,208 35 Room T Top 700 -11.0% 14.0%
70% Bo . 300
405
only posi i e efficiencies ( o be able o co e he ene gy consump ion e-406
qui ed by he pumping equipmen ) when he sys em is using oxygen as407
he oxidan . This is due o he much highe consump ion o ai comp es-408
so s compa ed o liquid oxygen c yogenic pumps. Ac ually, he ene gy409
equi ed (defined in equa ion (8)) when using ai and oxygen ascends o410
28% and 0.2%, espec i ely.411
4.8.3. Ene gy eco e y wi h he new cooled wall eac o 412
Focusing on he new CWR eac o design, a de ailed analysis o he ene ge ic413
eco e y possibili ies is shown abo e. In figu e 17 i is shown ano he possibili y414
o ene gy eco e y o each effluen o he new CWR design, besides scheme415
shown in figu e 16. I was assumed, as obse ed expe imen ally, ha all he416
gases in ol ed in he combus ion lea es he eac o wi h he op effluen : CO2
417
p oduced in he eac ion, N2(when ai is used as oxidan ) and O2 om he418
oxidan mixed wi h he wa e flow, being he bo om effluen conside ed as419
pu e wa e .420
Influence o he dis ibu ion flow h ough diffe en pa ame e s421
To analyze he elec ici y p oduc ion wi h he new CWR eac o , he condi ions422
assumed a e: 5% oxidan excess; 1,208 kW o hea flow eed; flow o cooling423
24
!"#$%&
'#()*+!*,
-!!.
/(**(0#12(.)3*
$((+456
"7*!2
8("#9#,*!70
12(.)3*4(5
:(1#12(.)3*
;42!3*#
!<175,4(5
Figu e 17: Scheme o he di ec expansion o op effluen o he eco e y ene gy wi h he
CWR wi h 2 ou le s.
wa e equi alen o 35% o eed flow; and effluen empe a u es o 700 ◦C and424
300 ◦C, a op and bo om ou le s, espec i ely. The selec ed pe cen age o 425
cooling wa e is based on he op imal ope a ional pa ame e s ob ained wi h he426
new eac o .427
(a) (b)
Figu e 18: Efficiency o he eco e y ene gy o he new eac o wi h (a) di ec expansion and
(b) s eam p oduc ion.
Influence o he kind o ene gy p oduc ion sys em428
Figu e 18 shows he efficiency o he new eac o ob ained by di ec expansion o 429
he flow and s eam p oduc ion wo king wi h ai and wi h oxygen. Diffe en flow430
dis ibu ions a e assumed. As can be obse ed, he ene gy p oduced by di ec 431
expansion o he flow om he eac o is bigge han he ene gy ob ained by he432
p oduc ion o s eam in a Rankine cycle ha could be expanded a e wa ds.433
25