Vinyl Chloride Monomer Process 201 Chemical Process Design: Computer-Aided Case Studies. Alexandre C.

Dimian and Costin Sorin Bildea Copyright © 2008 WILE !"C# "erlag $m%# & Co. '$aA( Weinheim ISB)* +,8!-!.2,!-/00-!0 7 7.1 Basis of Design ,././ Problem Statement "inyl 1hloride monomer 2"C34 is one o5 the leading 1hemi1als 6sed mainly 5or man65a1t6ring poly7inyl 1hloride 28"C4. 9he 8"C :orld:ide prod61tion 1apa1ity in 200. :as o5 a%o6t -. million tons per year( :ith an ann6al gro:th o5 a%o6t -;( pla1ed a5ter polyole5i nes %6t %e5ore styrene polymers. In the /++0s the largest plant in the <SA had a 1apa1ity o5 a%o6t =-. >tons ?/@ ( %6t today there are se7eral plants o7er one million tons. At this s1ale e7en in1remental impro7ements in te1hnology ha7e a signi5i 1ant e1onomi1 impa1t. Comp6ter sim6lation( pro1ess optimiAation and ad7an1ed 1omp6ter ! 1ontrol te1hniB6es play a determinant role. In the past the man65a1t6re o5 "C3 raised 1on1erns regarding haAard( sa5ety and poll6tion. 9here5ore( the "C3 te1hnology :as among the 5i rst to pro5i t 5rom impro7ements s6ggested %y pro1ess sim6lation. As res6lt( the modern "C3 plants are today among the 1leanest and sa5est in the 1hemi1al pro1ess ind6stries. In this proCe1t the ra: materials are ethylene and 1hlorine o5 high p6rity. 9he design re5ers to a 1apa1ity o5 -00 >tonDyear "C3 polymeriAation grade. In5ormation a%o6t te1hnology 6sed in this proCe1t 1an %e 5o6nd in en1y1lopaedi1 :or>s ?/ E -@ ( re7ie: papers ?0 E =@ ( patents( as :ell as on the :e%sites o5 leading prod61ers. 9a%le ,./ sho:s B6ality spe1i5i 1ations( namely regarding >ey imp6rities ?2@ . 9he goal o5 this 1ase st6dy is to ill6strate some generi1 iss6es raised %y the 1on1ept6al design o5 a large ! s1ale pro1ess in7ol7ing se7eral rea1tion and separation se1tions inter1onne1ted in a 1omplex plant :ith re1y1les. 9he st6dy emphasiAes the intri1ate e55e1ts o5 handling the remo7al o5 n6mero6s imp6rities generated in di55erent rea1tors %y a 1ommon separation system( :ith impli1ations on pro1ess dynami1s and plant:ide 1ontrol. 202 7 Vinyl Chloride Monomer Process ,./.2 Health and Safety #ealth and sa5ety reg6lations 2FS#A4 reB6ire the monitoring o5 1on1entrations o5 harm56l spe1ies in all 5a1ilities :here "C3 is prod61ed or 6sed. 9he employees sho6ld not %e exposed to more than / ppm o7er 8 h( or no more than . ppm 5or periods ex1eeding /. min. Chroni1 expos6res to more than /00 ppm may lead to serio6s diseases. Fn the other hand( "C3 is 5l amma%le %y heat( 5l ame and oxidiAing agents. Large 5i res are di55i 16lt to exting6ish. Be1a6se o5 possi%le peroxide 5ormation( the "C3 m6st %e stored or transported 6nder an inert atmosphere. 9he 6se o5 sta%iliAers pre7ents polymeriAation d6ring pro1essing and storage. "C3 is generally transported in railroad tan> 1ars and tan> tr61>s. ,./.conomic !ndices 9a%le ,.2 presents some e1onomi1 indi1es o5 a modern "C3 te1hnology ?=@ ( in :hi1h stoi1hiometri1 7al6es are indi1ated in parentheses. 9he yield 5or 1hlorine may rise o7er +8;( the yield 5or ethylene %eing slightly lo:er %e1a6se o5 losses %y 1om%6stion. 7.2 "eactions and #hermodynamics

In order to in1rease the o7erall yield( the 1hlorine 1ontained in s6%prod61ts( light or hea7ies( 1an %e re1o7ered %y oxidation to #Cl.tility-t VCM O ele1tri1 po:er 2>Wh4 /0.1 86rity spe1i5i 1ations 5or 7inyl 1hloride ?2@ .#2FI /+2.0 E /. O 56el 231al4 ==0 E ..F2JC2#-Cl I 0.2 4( %6t 1hlorine has to %e re1o7ered 5rom the %yprod61t #Cl( as %y means o5 the 1lassi1al Dea1on rea1tion* .2.0 Water 200 Iron( %y :t. E //0 O steam 2t4 0.Cl2 I 0. Me1o7ery o5 #Cl and oxy1hlorination o5 ethylene to EDC* C2#0 I 2#Cl I 0.8 E .=84 O oxygen /-/ E /-+ 2/284 . 2..0.8.. Sin1e the o7erall rea1tion is 7ery exothermal( the "C3 plant sho6ld %e a%le to 1o7er a large part o5 its energy needs.2. 0./>KDmol -.. 9here5ore( employing only dire1t 1hlorination o5 ethylene is more %ene5i 1ial 5rom the en7i! #able 7../ . 9he main 1hemi1al steps are explained %rie5l y %elo:. 7. Dire1t 1hlorination o5 ethylene to /(2 ! ethylenedi1hloride 2EDC4* C2#0 I Cl2JC2#0Cl2 I 2/8 >KDmol 2. !m$%rity Ma&im%m le'el ($$m) A1etylene 2 A1idity( as #Cl %y :t 0. 9hermal 1ra1>ing 2pyrolysis4 o5 EDC to "C3* C2#0Cl2JC2#-Cl I#Cl L . 9he pro1ess %lo1> diagram 1onsists o5 three plants( as sho:n in Nig6re . )ote that the p6ri5i 1ation o5 all EDC streams prod61ed %y synthesis( as :ell as re1y1led( 1an %e done in a 1ommon separation se1tion... By this is meant that all intermediates and %yprod61ts are re1y1led in a :ay that ens6res a tight #able 7.2 E1onomi1 indi1es o5 "C3 pro1esses ?=@ .. Ethylene( propylene 0( 8 Ethylene( di1hloride 2EDC4 /0 )on7olatiles /.. A1etaldehyde 0 Al>alinity( as )ao# %y :t. 0.g-t VCM) O ethylene 0=2 20084 O 1hlorine .B6tadiene = / ! %6tene( 2 ! %6tene -( 0.>KDmol 9here5ore( hal5 o5 the ethylene and the :hole o5 the 1hlorine are 5ed to dire1t 1hlorination( :hile the other hal5 o5 the ethylene goes to oxy1hlorination../ Process Ste$s 3ost o5 the "C3 te1hnologies are %ased on G %alan1ed H pro1esses.20 O 1ooling :ater 2m 4 /.F2JC2#0Cl2 I#2FI 2-8 >KDmol #en1e( an ideal %alan1ed pro1ess 1an %e des1ri%ed %y the o7erall eB6ation* C2#0 I 0.00 Fperators 2menDshi5t4 0 20/ 7 Vinyl Chloride Monomer Process ronmental 7ie:point 2Nig6re . E 0.2.. 3ost o5 the 1hlorinated :aste is prod61ed in the oxy1hlorination step. "a+ material (.2 Reactions and hermodynamics 20* 1los6re o5 the material %alan1e to only "C3 as the 5i nal prod61t( starting 5rom ethylene( 1hlorine and oxygen. /..

. Ethylene and #Cl are gases( 1hlorine and "C3 may %e handled as liB6ids at lo:er temperat6res and adeB6ate press6re( :hile ethylene ! di1hloride is a liB6id..* Chemical 3 "eaction 4nalysis A >ey aspe1t in the "C3 te1hnology is the remo7al o5 n6mero6s %yprod61ts and imp6rities 5ormed in the three rea1tion systems.. By passing thro6gh inter1onne1ted re1y1les some imp6rities prod61ed in one rea1tor may ind61e the 5ormation o5 imp6rities in other rea1tors.02+ Explosion limit in air( . A 1apital 1ost red61tion o5 /.8 L -0./=/ 20. It may %e seen that tra1es o5 a1etylene :ill 5ollo: #Cl in re1y1le.. 1ame thylene Chlorine HCl VCM DC 3ole16lar 5orm6la C 2 # 0 Cl 2 #Cl C 2 # . 2 E -= )one )one 0 E 22 = E /= 7. -20. Large di55eren1es in the physi1al state may %e o%ser7ed..+0= -=. 2C2#= I Cl2 I -D2F2JC2#-Cl I -#2F 9he te1hnology pro5i ts 5rom the extensi7e experien1e gained %y the oxy1hlorination o5 ethylene( %6t is %ased on a 1ompletely di55erent 1atalyst.+= )ormal %oiling point( P C L /0-. Chlorinated :aste 5rom other pro1esses may %e in1inerated to #Cl( and in this :ay re1y1led to the man65a1t6ring o5 "C3. Criti1al 7ol6me( 1m .Cl C 2 # 2 Cl 2 3ole16lar :eight 28.0 . .2 "C3 man65a1t6ring :ith 1hlorine re1o7ery %y #Cl oxidation..0 8-..20. L 8.0= =2.0.. At this stage( it is interesting to examine the properties o5 all 1hemi1al spe1ies( prod61ts( %yprod61ts( intermediates and imp6rities :ith impa1t on the pro1ess design. +8.= .2 L -04 //+.2 Physical Pro$erties 9a%le . 9a%le . 4 at 0 %ar and 2=0 E -20 P C ?+@ ...1 "C3 man65a1t6ring %y a %alan1ed pro1ess 6sing oxy1lorination.. 9he gaseo6s emissions are red61ed pra1ti1ally to Aero. ..displays >ey physi1al properties o5 the main 1omponents. 9his iss6e :ill %e 1learer a5ter presenting the 1hemistry o5 ea1h rea1tion step.0-2 /=.Dmol /2+ /20 8/ /=+ 220 LiB6id density( >gDm ../ L /-../ .-.2.* 8hysi1al properties o5 the main 1omponents.2 P C4 .2#Cl I 0.F2JCl2 I#2F In a ne: pro1ess proposed %y 'ellogg the oxidation o5 #Cl ma>es 6se o5 nitrosyl ! s6l56ri1 a1id 2#)SF .0 0/.0.2 L 8. 9here5ore( the 6se o5 higher press6res and lo:er temperat6res is expe1ted in 7ario6s pro1essing steps.0 gi7es a list o5 normal %oiling points( as :ell as origin and exit points. Nor example( imp6rities generated %y the EDC 205 7 Vinyl Chloride Monomer Process 1ra1>ing %6t not remo7ed 5rom #Cl :ill generate other imp6rities in oxy1hlorination( :hi1h in t6rn( i5 not remo7ed %y EDC p6ri5i 1ation( :ill deteriorate the sele1ti7ity o5 the 1ra1>ing and dire1t 1hlorination steps( a116m6lating e7en more . 9he ne: pro1ess 1laims a 1ost red61tion o5 a%o6t -0. 9he pro1ess is des1ri%ed %y the 5ollo:ing glo%al eB6ation* 0ig%re 7.0.( %e1a6se the ethane pri1e is a%o6t one third that o5 ethylene. 9he amo6nt o5 %yprod61ts diminishes drasti1ally raising the o7erall yield o7er +8.=. #able 7..0 2/=4 Q ! 7ap at n%p( >KDmol /-.0 Criti1al temperat6re( ' 282. Sin1e some 1hlorinated spe1ies ha7e %oiling points 1lose to EDC( its p6ri5i 1ation :ill %e demanding.. 0ig%re 7.. . 9he pro1ess is sa5er %e1a6se the 1onta1t o5 hydro1ar%on mixt6res :ith oxygen is eliminated. may %e estimated 1ompared :ith the oxy1hlorination pro1ess.=0/ -2. 8-.4 +=+ 2 L /04 /2.=/ Criti1al press6re( %ar . Ninally( it is :orthy to mention the ad7ent o5 a ne: generation o5 te1hnologies %ased on ethane oxy1hlorination( as 1ommer1ialiAed %y E"C International ?/0@ .= 02+." Chemical-Reaction Analysis 202 7. 2 L //04 /..

./ to 0. #o:e7er( in order to get 1onsistent material %alan1e it is pre5era%le to 6se stoi1hiometri1 eB6ations and a7oid mass ! yield expressions. Lo: ! temperat6re 1hlorination 2L9C4.a4.8=..0 DC( Fxy Intermediate( to C' 9MI C 2 #Cl .//-.. 9:o methods 1an %e 6sed 5or remo7ing the heat o5 rea1tion( s61h as an internal heat ! trans5er de7i1e( 5or example a 1ooling 1oil( or external heat ex1hanger :ith re1y1ling o5 1old EDC.. C'( Fxy EDC( p6ri5i 1ation lights 9C3 C#Cl .0 C' EDC p6ri5i 1ation( lights DC3 C# 2 Cl 2 00.=. Be1a6se o5 sim6lation 1onstraints( namely in dynami1 mode( the n6m%er o5 1omponents in7ol7ed in the 1hemistry o5 di55erent steps sho6ld %e limited... C' EDC p6ri5i 1ation( lights E9CL C 2 # ...0 C' "C3 separation( sent to sele1ti7e hydrogenation #Cl #Cl E ==. #igh ! temperat6re 1hlorination 2#9C4. Slight ex1ess o5 1hlorine is pre5erred in order to ens6re 1omplete ethylene 1on7ersion. Nor this reason small amo6nts o5 oxygen( o5ten present in the 1hlorine prod61ed %y ele1trolysis( may in1rease the sele1ti7ity to EDC %y inhi%iting the se1ondary rea1tions. C'( Fxy EDC p6ri5i 1ation CLAL C 2 # . 2.0 E . C'( Fxy EDC p6ri5i 1ation( lights DCE R C 2 # 2 Cl 2 0.8 C' "C3 separation( prod61t B<9D C 0 # = E 0. Fxy Separating Fxy 9CE C 2 # . Cl /2. 9he most important %yprod61t is the /(/(2 ! tri1hlorethane 29CE4( as des1ri%ed %y* C2#0 I 2Cl2JC2#-Cl.Cl .+ DC( C'( Fxy EDC p6ri5i 1ation( hea7ies #ea7ies C 0 # 8 Cl 2 /. As a disad7antage :e note that reCe1ting the heat o5 rea1tion to the en7ironment at lo: temperat6re is highly ine55i 1ient. Se1ondary rea1tions ta>e pla1e.. 9he dire1t 1hlorination o5 ethylene may %e 1ond61ted 5ollo:ing t:o te1hniB6es* /... C' "C3 separation( to Fxy 3ECL C# ..DC( C'( Fxy EDC p6ri5i 1ation( lights CL8 C 0 # .=/.0 P C( %elo: the mixt6re %oiling point 2Nig6re ./ Direct Chlorination 9he dire1t 1hlorination o5 ethylene 1onsists o5 rea1tion %et:een dissol7ed gaseo6s rea1tants in the liB6id EDC 5ollo:ing the eB6ation* C2#0 Cl2 C2#0Cl2 2+8 >KDmol I J Q!0 R L2/8 2/4 9he 1atalyst is o5 Le:is ! a1id type( in most 1ases NeCl ./ Chemi1al spe1ies in "C3 pro1ess.Cl .. :t.. Cl -.29CE4I#Cl 2-4 9he 5ormation o5 imp6rities in7ol7es the o116rren1e o5 radi1als..0 C'( Fxy "C3 p6ri5i 1ation "C3 C 2 # . Another maCor dra:%a1> ./ C'( Fxy EDC p6ri5i 1ation DC* dire1t 1hlorinationS C'* 1ra1>ingS Fxy* oxy1hlorination "ery li>ely is the 1hlorination o5 EDC to tri1hloroethane 29ME4* C2#0Cl2 I Cl2JC2#-Cl. Lo:er temperat6re is ad7antageo6s 5or a1hie7ing higher sele1ti7ity( o7er ++. 9he 6se o5 high ! p6rity rea1tants is re1ommended to a7oid the 5ormation o5 a larger spe1tr6m o5 imp6rities that might 1ompli1ate e7en more the EDC p6ri5i 1ation.Cl E /-....-.C'( Fxy EDC p6ri5i 1ation( lights CCl0 CCl 0 . In L9C the rea1tor is a simple gas E liB6id 1onta1t 1ol6mn operating at temperat6res %et:een .I#Cl 224 #able 7. .. C'( Fxy EDC p6ri5i 1ation( lights EDC /(2 ! C 2 # 0 Cl 2 82. Com$onent 0orm%la 1b$6 ( 7 C) 8rigin &it C2#2 C 2 # 2 E 80..( in 1on1entrations o5 0. #en1e( the handling o5 imp6rities is the >ey iss6e in managing the o7erall plant material %alan1e %y "C3 man65a1t6ring.Cl E 20.imp6rities in re1y1les( etc .+.

/( in the a%o7e 6nits.( 8@ . A disad7antage o5 #9C :o6ld %e lo:er sele1ti7ity( %6t some patents 1laim that modi5i ed 1atalysts andDor the 6se o5 inhi%itors ma>e possi%le operation 6p to /. 1an %e a1hie7ed :ith sele1ti7ity in EDC o5 +/ E +=. An amo6nt o5 ethylene is lost %y 1om%6stion at higher temperat6re* . In 5a1t( 9MI and EDC 5orm a lo: ! %oiling point aAeotrope 7ery 1lose to EDC. In parti16lar( 1hloral needs to %e remo7ed immediately a5ter rea1tion %y :ashing %e1a6se o5 its tenden1y to polymeriAation.. 0ig%re 7.* Mea1tion te1hniB6es 5or the dire1t 1hlorination o5 ethylene. 9he 5ormation o5 9MI in the oxy1hlorination rea1tor is d6e to the a1etylene entrained :ith the #Cl %yprod61t 5rom 1ra1>ing. E .....= ... exp2 . Some >ey imp6rities are listed %elo:* /(/(2 ! tri1hloroethane 29CE4( 1hloral 2CCl . .%4.. 9hese data are in good agreement :ith experimental 7al6es o5 /-.! C#F4( tri1hloroethylene 29MI4( /(/ ! and /(2 ! di1hloroethylenes( ethyl 1hloride( 1hloro ! methanes 2methyl ! 1hloride( methylen ! 1hloride( 1hloro5orm4( as :ell as poly1hlorinated high ! %oiling 1omponents. 3ore generally( the 1hemi1al 1on7ersion o5 imp6rities is a po:er56l method 5or diminishing positi7e 5eed%a1> e55e1ts thro6gh re1y1les :ith negati7e in5l 6en1e on operation and the 1atalyst remo7al 5rom EDC %y 1ostly operations and so6r1es o5 poll6tion.E +. D 4m >mol sL/ 2./4 Lr/ R # 2c c # R T L L 2+=2 2 2 D C # Cl2 D2 8 ..8 :ith exp2 . With respe1t to 1hemi1al rate( FreCas ?//@ proposes the 5ollo:ing eB6ations o%tained %y the regression o5 ind6strial data* Lr / R # c c # R L L L -// 2 2 D/ C # Cl2 D/ //0+.0 P C. R I Q!0 R L2+. /0 .-. %ar and +0 to /. 9he 1hemi1al rea1tor may %e integrated as a re%oiler o5 a distillation 1ol6mn( or designed as independent eB6ipment 2Nig6re . 204 9he 1atalyst :idely 6sed is %ased on 1opper2II4 1hloride impregnated on al6mina. Ethylene 1on7ersion o5 +.0 P C. and 2..2 8&ychlorination 9he oxy1hlorination step is des1ri%ed %y the 5ollo:ing glo%al rea1tion* C2#0 #Cl F2 C2#0Cl2 #2F 2+8 >KDmol I 2 I 0. As mentioned( the spe1tr6m and amo6nt o5 imp6rities 5ormed d6ring oxy1hlorination is m61h larger 1ompared :ith dire1t 1hlorination. 9he 5l 6id ! %ed te1hniB6e o55ers more intensi7e heat trans5er( pre7ents the o116rren1e o5 hot spots and allo:s more e55i 1ient 1atalyst regeneration. Se1ondary rea1tions mani5est( as explained next. 9he o7erall pro1ess rate in7ol7es the addition o5 resistan1es 5or mass trans5er and 1hemi1al rea1tion. %ar( in 5i xed ! %ed or 5l 6id ! %ed rea1tors. D 4m mol s 2.282 2/ :ith .24 At =0 P C the rea1tion rate 1onstants are /. 9he se1ond one has %een highlighted %y st6dies in the 5i eld o5 pro1ess dynami1s and plant:ide 1ontrol ?.-+ ?/2@ . 9he highly exothermal rea1tion is 1arried o6t at temperat6res aro6nd 200 P C and press6res o5 /.2 and 2.. to ." Chemical-Reaction Analysis 207 209 7 Vinyl Chloride Monomer Process In #9C( the rea1tion is 1ond61ted at the %oiling point o5 EDC at /.2/. 9he 5ormation o5 9MI is 6ndesired( %e1a6se its remo7al %y distillation is 7ery di55i 16lt. 7. In this manner the heat o5 rea1tion( :hi1h is se7en times the heat o5 EDC 7aporiAation( 1an %e 6sed ad7antageo6sly 5or p6ri5i 1ation. 9:o sol6tions 1an %e adopted* O sele1ti7e 1atalyti1 hydrogenation o5 a1etylene in the #Cl re1y1le streamS O 1hlorination o5 9MI to hea7ies a5ter 1on1entration %y EDC distillation../. Some pro1esses ma>e 6se o5 the 5i rst method.

Cl 2E9CL4 I#Cl 2/04 C2#0 I 2Cl2 R C2#-Cl.DS. All these rea1tions lead 5i nally to a :ide spe1tr6m o5 imp6rities.EDCD2m . At higher temperat6re e7en the C E C %ond in 1hlorinated prod61ts 1an %rea>( 5orming 1hloromethanes.. 2/000 /2 8 =200 RLIRL RL . 2.4 C2#2 I 2Cl2 R C2#Cl.-4 9he rea1tion rate is gi7en in ) m ." Chemical-Reaction Analysis 20: 210 7 Vinyl Chloride Monomer Process ln ln ln #R #R #R /2 . Nor a1ademi1 p6rposes the 5ollo:ing >ineti1 eB6ation may %e applied ?/.C2#0 IF2 R Cl2 I#2F 2..29CE4I#Cl 2. /+-00 20 8. Stoi1hiometri1 eB6ations are gi7en %elo: and may %e 6sed 5or material ! %alan1e p6rposes( altho6gh the tr6e rea1tion me1hanism is m61h more 1omplex* C2#0Cl2 I Cl2 R C2#-Cl.@ * r $#pp $p#p#p R II /2 /22 2// #Cl 2 C# #Cl 2 C#F 2 222D4 2.4 Despite large ! s1ale ind6strial appli1ation( >ineti1 data 5or the 1atalyti1 oxy1hlorination are missing.29EC4 I#Cl 2//4 C2#2 I -Cl2 R 2C#Cl. 9he temperat6re dependen1e o5 the 1onstants is as 5ollo:s* 7.F2 R Cl2 I#2F 2=4 Chlorine is in7ol7ed 56rther in radi1al rea1tions prod61ing many 1hlorinated spe1ies.DS .! 1atalyst h4( the press6re in atm.29MI4I#Cl 284 C2#0 I 2Cl2 IF2 R Cl-CLC#F2CLAL4 I#Cl I#2F 2+4 C2#= I Cl2 R C2#.29C34 2/24 C2#0 I 0Cl2 I F2 R 2CCl0 I 2#2F 2/-4 C2#0 I Cl2 R C2#2Cl2 2DCER4 I2#Cl 2/04 2C2#0Cl2 R C0#8Cl2 2hea7ies4I#Cl 2/.4 Small amo6nts o5 1hlorine appear sin1e the 1opper 1atalyst is a1ti7e 5or oxidation* #Cl I0.

to -0 s. 9he rea1tion me1hanism %y the EDC 1ra1>ing in ind6strial 1onditions is extremely 1omplex. 9he so5t:are S8 MF is a7aila%le 5or the detailed design o5 the rea1tion system( in1l6ding the rea1tion 1oil and the 56rna1e 2 :::.#hermal Crac. . 9he se1ond one emphasiAes the 5ormation o5 a1etylene( 56rther in7ol7ed in the 5ormation o5 other imp6rities( s61h as %enAene and 7inyl ! a1etylene.R C2#2Cl2 2DCE4I#Cl 22-4 C2#-Cl I2Cl2 R C2#Cl. ManAi et al. 9he pa1>age 1an also %e 6sed 5or monitoring the per5orman1e in operation and pre7ent pro%lems( s61h as 5o6ling o5 t6%es %y 1o>e 5ormation.Cl R C2#0 I#Cl 2/+4 C2#0 I C2#2 R C0#= 2B<9D4 2204 C2#2 I C2#-Cl R C0#.4 /(2!C2#0Cl2 R C2#0 ICl2 2/84 C2#. 9he t6%e diameter is sele1ted so as to gi7e a s6per5i 1ial gas 7elo1ity %et:een /0 E 20 mDs. A%o6t 20 E -0 1hemi1al 1omponents ha7e %een identi5i ed %y EDC pyrolysis ?/@ . Chloromethanes( s61h as C#Cl and CCl 0 ( are examples o5 good imp6rities. I5 these are not in present in EDC in s655i 1ient amo6nts( they are added deli%erately.1om 4. 9he se1ond part( the rea1tion Aone( is pla1ed in the radiation 1ham%er.44 and 1hloroprene 2rea1tion 22=44. In the presen1e o5 5ree radi1als they may prod61e a 7ariety o5 higher mole16lar spe1ies( some 6nsat6rated as %6tadiene 2rea1tion 22. Nor this reason( :e :ill 1onsider in this proCe1t the EDC 5rom oxy1hlorination as an external so6r1e entering the p6ri5i 1ation system. )ext( the rea1tions leading to light and hea7y 1hlorinated hydro1ar%ons are :orth mentioning* /(2C2#0Cl2 I Cl2 R At high temperat6re( the EDC de1omposes into "C3 and #Cl %y a 1omplex rea1tion me1hanism dis16ssed 56rther in this se1tion. 9hese spe1ies may a1t as a so6r1e o5 radi1als :ith the e55e1t o5 in1reasing the rea1tion rate and sele1ti7ity at s6%stantially lo:er rea1tion temperat6re./ EDC 1ra1>ing rea1tor. ?/-( /0@ proposed a s1heme in7ol7ing more than 200 elementary rea1tions as :ell as 00 mole16lar and radi1al spe1ies. Some rea1tions :ith impli1ations 5or the rea1tor design are gi7en %elo:* /(2!C2#0Cl2 R C2#-Cl I#Cl 2/=4 C2#-Cl R C2#2 I#Cl 2/. Nor this reason( ad7an1ed p6ri5i 1ation o5 EDC sent to 1ra1>ing is reB6ired.0 4.Cl 2CL84 22/4 9he rea1tions 22/4 and 2224 are parti16larly important.D 9he design o5 a 5l 6id ! %ed rea1tor is di55i 16lt :ith 1on7entional tools. 9he third and the 5o6rth rea1tions explain the 5ormation o5 light 6nsat6rated hydro1ar%ons..29MI4I#Cl 2204 .29CE4I#Cl 2224 C2#-Cl. 9he endothermi1 rea1tion ta>es pla1e at temperat6res %et:een 080 E . 9he rea1tion de7i1e 1onsists o5 a long t6%6lar 1oil pla1ed in a 56rna1e 2Nig6re . 3any originate 5rom spe1ies 1arried :ith the EDC iss6ed 5rom the synthesis steps.spyro. Some imp6rities( s61h as methyl1hloride 2 U =0 ppm4 and %6tadiene 2 U /00 ppm4( are 6ndesired %e1a6se they are di55i 16lt to remo7e d6ring "C3 p6ri5i 1ation..-. Fther 6nsat6rated 1omponents( s61h as 1hloroprene or tri1hloroethylene lead to polymeri1 resid6es that 1an 1a6se the 5o6ling o5 rea1tors.. 9he 5i rst part( hosted in the 1on7e1tion Aone( preheats the rea1tant 6p to the temperat6re :here the pyrolysis rea1tion rate %e1omes signi5i 1ant. 9he 1oil length sho6ld ens6re a spa1e ! time o5 . 9he 5i rst one leads to the main -0 %ar. 9he 1hloroprene( 5ormed %y the addition o5 a1etylene to 7inyl ! 1hloride( is highly 6ndesired. 0ig%re 7..0 P C and press6res 5rom . It is :orth mentioning that the pro1ess itsel5 is 1apa%le o5 prod61ing G good imp6rities H .

presents >ineti1 parameters o5 5i rst ! order rate 1onstants 5or the >ey rea1tions %y EDC pyrolysis %y ass6ming p6rely mole16lar me1hanism ?/0@ ..0 m( 1orresponding to a spatial time o5 a%o6t -0 s.. Fn the other hand( %e1a6se o5 larger a1ti7ation energies the se1ondary rea1tions are promoted %y in1reasing temperat6re.Cl I #Cl 0. /(2 ! C 2 # 0 Cl 2 R C 2 # 0 I Cl 2 /.. /00 mm./4 and 2..T /0 L .Cl R C 2 # 2 I #Cl 0. 2. ...0 T /0 I/.=2 T /0 L . It is >no:n that modern pro1esses operate at m61h lo:er temperat6res and ma>e 6se o5 initiators. 9o 7alidate the >ineti1 data( :e 1onsider an initial 5l o: rate o5 . 9he sim6lation o5 the ethylene 1hlorination 1an %e done as an adia%ati1 8NM :ith a liB6id s6per5i 1ial 7elo1ity %et:een 0. 9his phenomenon is 5a7ored %y higher :all temperat6re( %y the presen1e o5 hea7y 1hlorinated hydro1ar%ons( as :ell as %y some hea7y imp6rities( namely tri1hloroethylene 29MI4. 9a%le .000 >gDh EDC at a rea1tion temperat6re o5 . An important 1on1l6sion is that the mass ! trans5er rate is 5ast 1ompared :ith the 1hemi1al rea1tion./ "eactor Sim%lation .-0 P C and /8 %ar. A 1omparison :ith ind6strial data sho:s that the >ineti1 data 5rom 9a%le ... 9he length is 2. 9he linear 7elo1ity at the rea1tor inletDo6tlet is o5 a%o6t = mDs and /0 mDs( respe1ti7ely( s655i 1ient to ens6re a highly t6r%6lent regime %6t a moderate press6re drop.. 'eeping the rea1tion temperat6re %elo: .2.-+ T /0 L .. Nollo:ing this data the main rea1tion C6st %egins at 080 P C( %6t m61h higher temperat6re is ne1essary to rea1h an e55i 1ient rea1tion rate. 1 ) E (cal-mol) k (/90 7 C) k (220 7 C) /(2 ! C 2 # 0 Cl 2 R C 2 # .8 000 .00 P C pre7ents the 1o>e 5ormation %6t de1reases the rea1tion rate.. 9o o%tain more realisti1 res6lts the pre ! exponential 5a1tor 5or the pyrolysis rea1tion :as modi5i ed to /..2 Pyrolysis of DC 9he pyrolysis rea1tor 1an %e sim6lated in Aspen 8l6s as 8NM :ith po:er ! la: >ineti1s and temperat6re pro5i le or heat d6ty.0./ thylene Chlorination A detailed model 5or a gas E liB6id 1ol6mn :ith external re1ir16lation loop has %een p6%lished %y FreCas ?//@ ..C2#0 I C2#2 R C0#= 2B<9D4 22.2a4 :ith the >ineti1 eB6ations 2.4 2C2#0Cl2 R C0#8Cl. 9he temperat6re sho6ld %e raised to more than . T /0 I/0 =+ 000 0.2 000 /..2hea7ies4 I Cl2 22=4 #ighly 6ndesired is the 5ormation o5 1ar%on deposit./0 T /0 /0 ( :hile the pre ! exponential 5a1tor o5 the a1etylene prod61tion .mDs.. 9he model ta>es into a11o6nt the axial dispersion and mass trans5er 5rom %6%%les. 9he res6lts o%tained %y Aspen 8l6s ?/+@ %y 6sing a s1heme s61h as that sho:n in Nig6re . 7..0..2 'ineti1 parameters 5or EDC pyrolysis 2a5ter ManAi et al.0 P C to a1hie7e 1on7ersions o5 a%o6t . 9he rea1tor 1onsists o5 /= t6%es in parallel :ith an internal diameter o5 #able 7. gi7es some:hat 1onser7ati7e res6lts. Nor these reasons( the pyrolysis temperat6re pro5i le is an optimiAation iss6e.-= T /0 I/0 . 7.T /0 L 2 C 2 # ./ and 0...0. ?/0@ 4." Chemical-Reaction Analysis 211 212 7 Vinyl Chloride Monomer Process ManAi et al. "eaction A (s . As a res6lt( a pse6dohomogeneo6s model 5or liB6id ! phase rea1tion may %e applied 5or design p6rposes. 9his depends on a large n6m%er o5 5a1tors( s61h as 5or example press6re( heat ! load distri%6tion( nat6re and amo6nt o5 imp6rities in EDC( as :ell as the 6se o5 initiators.. 8re7enting the 1o>e 5ormation is a maCor pro%lem in operating the 56rna1e 5or EDC 1ra1>ing. 9here5ore( as already mentioned( it is rational to 6se G initiators H ( s61h as nitromethane( 1hloro5orm or 1ar%on tetra1hloride. ?/-@ demonstrated that the radi1al generation %y initiation is %y 5ar more important than %y 1hlorine elimination or C E C %ond %rea>ing.-0 T /0 L 0 /.0.24 is in good agreement :ith the ind6strial pra1ti1e.0 T /0 L 8 .

)ote that the rea1tion 2 is expressed in term o5 a sele1ti7ity 7aria%le S ( representing the degradation o5 EDC in hea7ies.-+-0 0.-./.. 9he alternati7e 2a4( des1ri%ed in many s1hemes( ma>es 6se o5 5l ash separation :ith liB6id re1y1le and external 1ooling.80+ 0.2%2 20.= 0.0.0. 0.0002 0./.--82 0.8 0.. 0.. %ar.08.000. 9he disad7antage is energy loss %e1a6se o5 the lo: temperat6re 5or heat re1o7ery. 0.0=% L 0./..000/ 0..4 2./0 2-000 /-.+E E 0.000. 9he rea1tions 22-4 to 22..E E 0. )ext( the %eha7ior o5 the rea1tion mixt6re s6%mitted to a simple 5l ash :ill %e in7estigated.000/ 0. 0. +.in1reased to .. 0.-0 P C and /8 %ar is* #able 7.. P C and /...04 S expresses( in a syntheti1 manner( the in5l 6en1e o5 the rea1tion 1onditions on the o7erall sele1ti7ity..-.000/ 0. 2.2+ 0= /0 A5ter pyrolysis( the mixt6re is s6%mitted to rapid 1ooling %y B6en1h. 0. 0.0000 0. #CL 0. #EA"IES 0. Cl 2 0 EDC I Cl 2 J 9CE I #Cl 0.8E E 0. 0.-.000/ ..& Separation System 21* 21/ 7 Vinyl Chloride Monomer Process Com$onent DC VCM HC< C2H2 C<P #C H 4V! S #"! Nlo: >gDh -. 0.2 Alternati7es 5or the B6en1h o5 the rea1tor o6tlet. 9he alternati7es 2%4 and 214 employ a drasti1 1ooling o5 the o6tlet gas %y a liB6id EDC stream 1oming 5rom a do:nstream 6nit( a 5l ash or distillation 1ol6mn. "eaction Con'ersion "eference reactant / C 2 # 2 I "C3 J CL8 0.E E 0.082 0..0=. 9he sele1ti7ity S may %e related :ith the 1on7ersion % o5 EDC as 5ollo:s* S R 0. 9his in5ormation is essential 5or st6dying the 5l exi%ility o5 design. I 0.00 2. 0.0.0002 /.0++ 0.-== 0. . C 2 # 2 2 2EDC J #EA"IES I Cl 2 / ! S EDC . Be1a6se o5 the presen1e o5 polar spe1ies :e sele1t a thermodynami1 0ig%re 7./.-200 0.2-20 0.0. #able 7.. 9his operation pre7ents 56rther de1omposition o5 "C3( %6t also remo7es the 1o>e and other high ! mole16lar imp6rities.000= 0. 9a%le .0000 0.. 0. 0.080 9CE 0.0002 0. 9his rea1tion is responsi%le also 5or 1hlorine prod61tion( 56rther in7ol7ed in other %yprod61ts.2// CL8 0. 7.. )ote that the n6m%er one in the last rea1tion is C6st 1on7ention* :e ass6me that the :hole amo6nt o5 5ree 1hlorine m6st %e 1ons6med in the last rea1tion( 5or :hi1h the real sele1ti7ity is o5 / E 20.0002 0.. 0. 0.9MI =.2 Cl 2 . 0. Se7eral alternati7es are possi%le( as ill6strated in Nig6re ..=.00/0 .2+..7 8hase eB6ili%ri6m o5 the rea1tor ! o6tlet mixt6re at /-.00/.0002 0.0002 0..--=.+8+ I 0.08.0 U % U 0..2 Se$aration System ... Com$onent 0 S" 3 Polar 8S 1"#< 3 "= >?=>?= EDC 0.8.000/ 0.0E E 0.0./ 0irst Se$aration Ste$ 9he synthesis o5 separations %egins %y examining the partial 5l o: rates o5 1omponents at the rea1tor o6tlet.. )ext( the mixt6re is 1ooled and s6%mitted to separation %y a 5l ash.00/. A typi1al mixt6re at .= gi7es details o5 the stoi1hiometri1 rea1tor in Aspen 8l6s.24 R 0.-8.-.-.0. 9he exit gas temperat6re 5rom B6en1h 1an %e s61h that it preser7es a good dri7ing 5or1e 5or the heat ex1hanger #V 6sed 5or energy re1o7ery %y steam generation ?/=@ .5 Mea1tions inp6t in the stoi1hiometri1 rea1tor.-=-8 2.+0 C2#2 0./=-0 0.4 :ere negle1ted( :hile the rea1tions 22=4 to 2-/4 :ere a11o6nted 5or %y a stoi1hiometri1 approa1h. T /0 I/0 . 0.2E E 0.."C3 I Cl 2 J 9CE 0. C 2 # 2 I 2Cl 2 J 9MI I #Cl / Cl 2 7.-/.0002 /0.20.-/0 "C3 0.8 .

. 9his sol6tion ens6res higher B6ality 5or "C3( o%tained as top distillate and 5ree o5 #Cl. Both models gi7e similar res6lts( altho6gh :ith some di55eren1es 5or C 2 # 2 and #Cl( the most 7olatile spe1ies. 9he a%o7e separation seB6en1e( altho6gh the most 6sed in ind6stry( is not 6niB6e. 7.7 Alternati7es 5or the separation o5 the #ClD"C3DEDC mixt6re. 9hese 1ol6mns also ha7e an important role in pro1ess dynami1s and 1ontrol. 9a%le . In a 5i rst approa1h short16t models may %e 6sed to sim6late the distilla! 0ig%re 7. 9he last re%oiled stripper 1ol6mn 2C ! -4 deli7ers EDC 5or 5i nal p6ri5i 1ation. Indeed( a sloppy split in a pre5ra1tionator 2S ! /4 1an ma>e the re%oiler temperat6re 1ompati%le to heat integration %y pro1essDpro1ess ex1hange../ . 9he 1r6de "C3 o%tained at the top o5 2C ! 24 is s6%mitted to 56rther p6ri5i 1ation in order meet the spe1i5i 1ations listed in 9a%le . #en1e( 6sing eB6ation ! o5 ! state models may %e seen as s655i 1iently a116rate 5or the 5i rst split %y 5l ash( %6t in assessing the distillations :e :ill ma>e 6se o5 the )M9LDM' model. 9his s1heme :as a1>no:ledged in an ind6strial paper ?/.& Separation System 212 215 7 Vinyl Chloride Monomer Process are pro7ided to minimiAe the losses in EDC passed in distillate and %ottoms( respe1ti7ely. P C and /. #o:e7er( another interesting sol6tion 1an %e imagined( as in Nig6re . 9he pre5ra1tionator 1o6ld %e simple 5l ash or re%oiled stripping.( the 5ollo:ing G dire1t seB6en1e H 1an %e de7eloped* Split /* top #Cl 2IC 2 # 2 4D%ottoms 2"C3 I EDC4 Split 2* top 1r6de "C3D%ottoms 1r6de EDC Split -* top Lights 2CL84D%ottoms 1r6de EDC Split 0* top p6re EDCD%ottoms 9CE and hea7ies 9he press6res in the distillation 1ol6mns are sele1ted 5or heat ! integration reasons( namely the a7aila%ility o5 a lo: ! temperat6re 6tility 5or #Cl 1ondensation and 1ooling :ater 5or "C3 distillation. 9he p6ri5i 1ation o5 EDC 5rom lights and hea7ies is per5ormed in the to:ers 29 ! /4 and 29 ! 24. In a preliminary sim6lation( the goal is 1losing the material %alan1e :ith re1y1les.. In addition( the 1hemi1al 1on7ersion o5 the nonsat6rated imp6rities to hea7ies may ta>e pla1e %y inCe1ting 1hlorine either in a separate rea1tor or dire1tly in the 1ol6mn 29 ! /4. 9his alternati7e is penaliAed %y ex1essi7e %ottom temperat6re in the 5i rst split at press6res a%o7e ...= presents the 5l o:sheet.. In addition( in the se1ond step an intermediate 1ompression :o6ld %e ne1essary 5or an e55i 1ient separation #ClD"C3. . 7. presents the res6lts at /-.5 Material 3 Balance Sim%lation 9he rea1tor and separation systems des1ri%ed a%o7e 1an %e assem%led in a 5l o:sheet. Applying the he6risti1s in Chapter .. In addition( the s1heme 1an %etter a11ommodate the heat integration %et:een di55erent 1ol6mns( %y 7arying the press6re le7els.. 9he distillation 1ol6mn 2C ! /4 separating #Cl operates at /0 E /2 %ar( 1orresponding to top temperat6res o5 L 2+ to L 20 P C. 9he separation o5 "C3 ta>es pla1e in the 1ol6mn 2C ! 24 at lo:er press6res( 6s6ally 0 to = %ar( 5or :hi1h normal or slightly re5rigerated 1ooling :ater 1an %e 6sed.@ . It 5ollo:s t:o distillation 1ol6mns 5or #Cl and "C3 separations. 9he se1ond 1hoi1e is the )M9LDMedli1h E ':ong model( 5or :hi1h %inary intera1tion parameters are >no:n. Nig6re .2 <i@%id 3 Se$aration System 9he mixt6re is shared approximately in eB6al parts %et:een #Cl( "C3 and EDC. 2le5t ! hand4. It is interesting to note that in pra1ti1e t:o additional small 1ol6mns 0ig%re 7.. Another possi%ility :o6ld 1onsist o5 adopting the G indire1t seB6en1e H .. %ar. 2right ! hand4. In this 1ase( EDC separates in the 5i rst split as %ottoms( 5ollo:ed %y "C3D#Cl distillation( as pi1t6red in Nig6re ..5 9he str61t6re o5 separations %y a G dire1t seB6en1e H .. %ar.option SM ! 8olar EFS.

Nor separation reasons( %oth the n6m%er o5 trays and the LD" tra55i 1 in the 1ol6mn C/0/ sho6ld %e s655i 1iently high. In prin1iple( the remo7al o5 9MI 1o6ld %e enhan1ed %y a side stream. 9he last 1ol6mn C ! 2 s6pplies 1r6de "C3 and EDC to %e re1y1led.' Material-(alance Simulation 217 219 7 Vinyl Chloride Monomer Process 0ig%re 7. 9his approa1h g6arantees the 1onsisten1y o5 the material %alan1e. A5ter 1ooling and 5l ash separation the mixt6re is sent to #Cl separation in the distillation 1ol6mn C20/. 9he B6en1h o5 the o6tlet mixt6re 1onsists o5 1ooling :ith re1y1led EDC.= . Nrom the sim6lation 7ie:point( ma>ing 6se o5 a G design spe1i5i 1ation H is the %est :ay to 5i nely adC6st 7. 9he next 1ol6mn C/02 separates the hea7ies. As a 1onseB6en1e( a s6%stantial amo6nt o5 EDC sho6ld %e entrained in the top distillate.tion 1ol6mns. Similarly( the 1ol6mn C/02 sho6ld %e designed 5or ad7an1ed remo7al o5 9CE and o5 other hea7ies %y allo:ing some EDC to %e dra:n in the %ottom prod61t. 9hese may %e attri%6ted to the 5ollo:ing reasons* /. 1on1entration to pre7ent polymeriAation. As a res6lt( the ex1ess o5 EDC in the top o5 C/0/ and in the %ottom o5 C/02 sho6ld %e re1y1led %y means o5 s6pplementary small 1ol6mns( as sho:n in Nig6re . 9he sim6lation o5 the 1ol6mns C20/ and C202 needs 1are56l analysis o5 spe1i5i 1ations. 9he design o5 #Cl 1ol6mn m6st ens6re %oth high re1o7ery and p6rity o5 #Cl in order to pre7ent a116m6lation in re1y1le. Nresh and re1y1led EDC enters the distillation 1ol6mn o5 lights C/0/. A5ter e7aporation and temperat6re rise the 7apor enters the 1ra1>ing rea1tor 8 MF. It is interesting to note that i5 the design o5 these 1ol6mns is not appropriate( sno:%all e55e1ts o116r %y a116m6lation o5 imp6rities ?+( /0@ .9 8reliminary 5l o:sheet o5 the "C3 plant. $ood 1on7ergen1e is o%tained :ith -. Sampled 7al6es 5or temperat6re( press6re and d6ties are displayed 5or in5ormation. Sno:%all may %e an indi1ation a%o6t the o116rren1e o5 possi%le m6ltiple steady states. We re1all that a sno:%all 1onsists o5 large 7ariations in the re1y1le streams generated %y small 7ariations in the inp6t or o6tp6t streams %e1a6se o5 ex1essi7e sensiti7ity o5 the system. 9his approa1h is eB6i7alent to >eeping 1onstant the inlet rea1tor 5l o:( also a good strategy 5rom the plant:ide 1ontrol 7ie:point 2see Chapter 0 4. theoreti1al stages and a %oil ! 6p ratio o5 0. 1on5l i1ts %et:een spe1i5i 1ation o5 6nits and 1on7ergen1e reB6irements. the distillate rate. We 5o6nd that the most relia%le manner is 5i xing the rate o5 the %ottom prod61t.8 presents the 5l o:sheet. Nirstly( the remo7al o5 9MI in the %ottom o5 C/0/ is hindered %y the aAeotrope that this imp6rity 5orms :ith the EDC. When sim6lating this 5l o:sheet 1on7ergen1e pro%lems appear. thermodynami1 %eha7ior o5 some imp6rities( 2. Another 1onstraint is that the 1on1entration o5 9MI in the p6ri5i ed EDC sho6ld %e >ept %elo: /000 ppm to a7oid 1o>e 5ormation d6ring pyrolysis. 9he next 1ol6mn C202 o55ers the 1hoi1e %et:een spe1i5ying the top distillate and %ottoms 5l o: rate( or the relati7e mode distillateD5eed. Nig6re .. Indeed( the liB6id 1on1entration pro5i le indi1ates a maxim6m o5 9MI near the top( %6t the sim6lation sho:s that this e55e1t is rather limited. 9his means >eeping 1onstant the re1y1le 5l o: rate o5 EDC at a 7al6e 1ompati%le :ith the per5orman1e o5 the 1hemi1al rea1tor. 7. 9he re1o7ery o5 CL8 in the top distillate m6st %e B6antitati7e( %6t :ith less than 8 :t. A se1ond stoi1hiometri1 rea1tor 8 MFS des1ri%es the se1ondary rea1tions.=. 9he distillate 5l o: rate sho6ld ens6re 1omplete #Cl re1o7ery( %6t :ith minim6m losses in "C3. A%o6t -0 stages and re5l 6x 5l o: rate at /000 >gDh are 1on7enient.. Fnly %y allo:ing eno6gh EDC in the top distillate 1an the remo7al o5 %oth 9MI and CL8 %e B6antitati7e.7 . Later( rigoro6s models are inserted 5or 5i nishing the siAing.

9he rea1tor o6tlet has s6%stantial energeti1 potential.0S L . Hot streams P.0S 2/.0S 2.0 J .! W0 /=..0 Separation inlet S0 ! S0A ! S0B /0.+S L 0.2S -.. T 6 d%ty bar6 deg.9 Streams 5or heat integration../S L -.0S /-.. T 6 d%ty bar6 deg.02 Me%oiler C/02 -... C6 MA Nl6e gases 56rna1e /... 9he ! ) 16r7e sho:s that the heat 1ontent o5 the 7apor part is s655i 1ient 5or 1o7ering the d6ty o5 the EDC e7aporator.= J 0-.0 . 9he preheater o5 the gaseo6s EDC is integrated in the 1on7e1tion Aone o5 the 56rna1e.0/-S //00 J 000 +.=S -..nergy !ntegration 9a%le ..28 EDC preheater N0 ! M/ /8. )ote that the de: point is /..0 Mea1tor o6tlet W/ ! W2 ! W.0 J 2/.8 presents the press6res( temperat6res and d6ties o5 the 1andidate streams 5or heat integration..0 J 0-.2/..2S L -.8 J +/.08 Mea1tor inlet Mea1tion mixt6re 22. C6 MA Cold streams P . )ote that the position o5 the heat ex1hanger 1an %e %e5ore or a5ter the B6en1h..0S .. 9he pyrolysis rea1tor is pla1ed in the radiation Aone. 9he se1ond alternati7e o55ers %etter prote1tion against 5o6ling( :hile the temperat6re dri7ing 5or1e remains high.00.8 Me%oiler C/0/ /..0S L 2.8 J /00.2 P C at /. Condenser C202 -S 20.2S +-..8 J /=.S 80.8 J /-.2 Me%oiler C20/ //.0S 0-. J /2-. 9he design o5 the 56rna1e is a 1omplex aspe1t not treated here( in :hi1h heat integration plays an #able 7.+S L 2..0. J +0.0S 00/..0 J -0.S /-0..0EDC 7aporiAer N.0S L /0. 9his sol6tion has the ad7antage o5 5eeding the rea1tor :ith gaseo6s EDC and pre7enting the 1o>e 5ormation. %ar.Condenser C20/ //.0S /20.0S 2/=.S /0-../8 Condenser C/02 -.-0 Condenser C/0/ /.+8 .0 J /-.=S .= J L 28.! N0 /+S /2/. I5 the e7aporation ta>es pla1e inside the 56rna1e an 6psetting e55e1t 1an ta>e pla1e ?/=@ ...

9he optim6m temperat6re is aro6nd -0 P C( :hi1h minimiAes the load o5 the 1ondenser( dri7en %y an expensi7e re5rigeration agent. 9he resid6al heat may %e reCe1ted against air or :ater 1ooling. )ote that the re%oiler o5 C/02 has a relati7ely moderate temperat6re 2/-= P C4( so that 38 steam 1an %e 6sed 5or heating( in t6rn generated %y 6pgrading the L8 steam prod61ed %y dire1t hot 1hlorination( or imported 5rom the other se1tions o5 the plant. 9he remaining stream to C20/ still has a signi5i 1ant potential o5 a%o6t 0. 8la1ing a small heat ex1hanger %e5ore 5l ash separation is re1ommended 5or %etter temperat6re 1ontrol./0 sho:s the s1heme proposed %y L6rgi ?/8@ on its :e%site. J /28.. A5ter this exer1ise( :e may 1on1l6de that %y adeB6ate heat integration the energy reB6irements o5 the pyrolysisDdistillation se1tion 1an %e red61ed 1onsidera%ly.S /28. A part o5 the heat 1an %e 6sed to 1o7er the re%oiler d6ty o5 C20/ %y the 6nit #V/.. Nor ea1h 6nit :ith signi5i 1ant .. 9he 1ondenser o5 C/02 has a d6ty o5 -. P C gi7es a dri7ing 5or1e at the 1old end o5 a%o6t /0 P C. 3W4. 9he a%o7e heat ! integration s1heme is o%7io6sly not 6niB6e.Me%oiler C202 -. )o 6nsta%le %eha7ior ta>e pla1e sin1e the pyrolysis rea1tion is endothermi1.02 3W( high eno6gh to 1o7er the re%oiler d6ty o5 C/0/ o5 -.0. Important sa7ing o5 energy in dri7ing the distillation 1ol6mns 1an %e a1hie7ed.8= 3W( %6t the temperat6re le7el is too lo: 5or heat integration. 9he hot ! o6tlet temperat6re at /-+ P C ens6res a minim6m temperat6re di55eren1e o5 /0 P C :ith respe1t to the re%oiler. 9his operation is sim6lated %y the heat ex1hanger CFFL2. Splitting the liB6id a5ter 5l ash is ne1essary 5or the B6en1h. An interesting energy sa7ing arises %y the integration o5 1ol6mns C/0/ and C/02 %y playing on press6res at /. )o temperat6re indi1ations are gi7en( %6t a5ter this exer1ise it 1an %e easily 6nderstood. 7./8 3W. 9he B6en1her is in 5a1t a m6lti5l ash separator that deli7ers separate gas and liB6id 5eeds 5or the #Cl 1ol6mn.: Nlo:sheet 1on5i g6ration %e5ore heat integration. important role too.%ar( respe1ti7ely.=S 2. S6pplementary sa7ing 1an %e o%tained %y the dire1t integration o5 the 1ol6mns 5or the EDC p6ri5i 1ation. It is interested to o%ser7e that the s1heme ma>es 6se o5 t:o NE#E 6nits to re1o7er the energy o5 the o6tlet gases to preheat the inlet rea1tor mixt6re %oth 5or entering the preheater and the rea1tion Aone to the rea1tion temperat6re.0./ and .7 *nergy +ntegration 21: 220 7 Vinyl Chloride Monomer Process 0ig%re 7. Condensation starts d6ring 1ooling( the 5i nal 7apor 5ra1tion %eing a%o6t 8/.. Me5rigeration is ne1essary 5or lo: ! temperat6re separation o5 #Cl( %6t the amo6nt 1an %e minimiAed %y a116rate 1ol6mn design. 9he enthalpy o5 the rea1tor o6tlet stream 1an %e 6sed 5or ens6ring the re%oiler d6ty o5 the 1ol6mn C202 22. An o6tlet temperat6re o5 /0. 9he res6lting 7apor and liB6id streams are merged again( %6t 1ooled lo: eno6gh to ens6re a 1on7enient thermal 1ondition o5 the 5eed 5or C20/. 9he sele1tion o5 press6res 1an lead to se7eral options that might %e explored %y the reader.9 Dynamic Sim%lation and Plant+ide Control Dynami1 sim6lation is prepared %y siAing the 6nits. #o:e7er( 1are sho6ld %e paid to 1ontrola%ility aspe1ts. 7. Nig6re .7 *nergy +ntegration 221 222 7 Vinyl Chloride Monomer Process 7. 9he hot 6tility 1an %e 1o7ered %y internal means( namely %y the #8 steam raised in the oxy1hlorination stage. Nor example( the heat 1ontent o5 the 5l 6e gases 1an %e 6sed 5or %oth 5eed preheating and steam generation. 9he hot rea1tor e55l 6ent 1an %e 6sed to dri7e the re%oilers o5 the 1ol6mns 5or separating #Cl and "C3.

9he most interesting points are dis16ssed %elo:. At the top o5 the 1ol6mn the 1ooling d6ty is 1onstant( the re5l 6x 5l o: 1ontrols le7el in the 1ondenser dr6m( :hile the 5l o: o5 the 7apor prod61t 1ontrols the press6re.dynami1s( the hold6p is 1al16lated %y ass6ming a residen1e time( typi1ally in the range o5 . Nirstly( :e 5i x the rea1tor ! inlet 5l o: rate and 5eed the 5resh EDC on le7el 1ontrol.4 "C3 1ol6mn 1ondenser. 9here5ore( a 1on1entration ! 1as1ade s1heme :as implemented( :ith 1hloroprene 1on1entration and re%oiler d6ty as 1ontrolled and manip6lated 7aria%les( respe1ti7ely. 2/4 EDC preheater( 224 Need e55l 6ent heat ex1hanger 2NE#E4( 2-4 EDC B6en1her 1ondenser( 204 #Cl 1ol6mn 1ondenser( 2. )ote that this arrangement( :hi1h is reB6ired %e1a6se o5 the small %ottoms stream( 1annot %e implemented i5 a >ettle re%oiler is 6sed. #o:e7er( :hen the plant :as dist6r%ed( a relati7ely large amo6nt o5 "C3 es1aped in the %ottom and ended in the distillate o5 the lights 1ol6mn. 9he tas> o5 the lights 1ol6mn is to remo7e the light 1omponents 5rom the re1y1led EDC( :ith 1hloroprene and tri ! 1hloroethylene %eing the most important imp6rities. A 5i xed 5ra1tion o5 the 5eed is ta>en as %ottom prod61t( in a 5eed5or:ard manner. )ote that 5i xing the re5l 6x and 1ontrolling the le7el %y distillate does not :or>( %e1a6se the distillate rate is 7ery small./2 2a4 sho:s res6lts o5 the dynami1 sim6lation( o%tained :hen a dist6r%an1e . As alternati7e strategy the open ! re1y1le sim6lation is exported to Aspen Dynami1s( :hi1h pro7ides the %asi1 in7entory 1ontrollers. 9he re%oiler d6ty 1ontrols the le7el in the 1ol6mn s6mp. Nig6re . 9he plant:ide 1ontrol str61t6re is presented in Nig6re . 9here5ore( a se1ond 1on1entration E temperat6re 1as1ade 1ontrolling the %ottoms 1omposition :as implemented. Control o5 rea1tor and 1ooling se1tion does not raise any pro%lem. 7. 22/ 7 Vinyl Chloride Monomer Process that #Cl is 1ompletely remo7ed 5rom the liB6id prod61t. 9hen( 1ontrol o5 the distillation 1ol6mns :as implemented. Sin1e the #Cl 1ol6mn operates mainly as a stripper( the temperat6re in the %ottom is 1ontrolled %y manip6lating the steam rate( so as to ens6re 0ig%re 7.2 1o6ld %e reprod61ed :ith small adC6stments o5 the rea1tor ! inlet temperat6re and rea1tor length( 1on7ergen1e o5 the steady ! state sim6lation :ith 1losed re1y1le is 7ery di55i 16lt. A5ter adding temperat6re 1ontrollers aro6nd heat ex1hangers( the re1y1le :as 1losed and sim6lation :as r6n 6ntil a steady state :as o%tained.0. Need ! rate 1hanges 1o6ld %e a11o6nted 5or %y >eeping 1onstant the ratio re%oiler d6tyD5eed thro6gh 5eed5or:ard 1ontrol. Altho6gh the per5orman1e o5 the pyrolysis rea1tor des1ri%ed in Se1tion . )ote that sim6lating a model o5 type G rea1tor ! :ith ! spe1i5i ed ! temperat6re H is not possi%le( %e1a6se o5 a high ! index pro%lem. E /0 min( and 5i nding the 1orresponding 7ol6me. Nor the distillation 1ol6mns tray siAing is ne1essary.. 9he le7el in the 1ondenser dr6m :as 1ontrolled %y the re5l 6x.. Control o5 the hea7ies 1ol6mn is simpler.// . At the "C3 1ol6mn( the p6rity o5 the prod61t stream 21hloroprene less than /0 ppm4( is 1ontrolled 6sing a 1on1entrationDtemperat6re 1as1ade s1heme( :ith re5l 6x as the manip6lated 7aria%le.10 Simpli5i ed 8ND o5 the EDC 1ra1>ingD"C3 separation se1tion :ith heat ! integration s1heme 2a5ter L6rgi Company4.11 Control o5 the 1ra1>ing se1tion o5 the "C3 plant. As a res6lt( %oth the temperat6re o5 the heating medi6m 2%6rn gases4 and the heat ! trans5er 1oe55i 1ient ha7e to %e spe1i5i ed. 9he 1ol6mn is operated at 1onstant re5l 6x( :hile the distillate rate 1ontrols the 1ondenser le7el. 9his approa1h is 7ery approximate as the heat is trans5erred mainly %y radiation. Dynamic Simulation and Plant-ide Control 22* 0ig%re 7. Nast 6nits( s61h as heat ex1hangers( p6mps or mixers 1an %e 1onsidered as instantaneo6s... 9he distillate to 5eed ratio :as >ept 1onstant 6sing 5eed5or:ard 1ontrol. 9his ratio 1an %e 6sed to adC6st the le7el o5 tri ! 1hloroethylene in the %ottom prod61t.

9hese imp6rities m6st %e remo7ed sele1ti7ely 5rom EDC. 9he so6r1e o5 I / is the pyrolysis rea1tor( I 2 appears mainly %y the oxy1hlorination( :hile I .00 >gDh. 9his is the in1enti7e o5 the plant:ide 1ontrol pro%lem.+. As mentioned in Se1tion . P C in the temperat6re o5 the %6rn gases :as introd61ed at time t R .. 9he intera1tions 1an hinder or help >eeping an a11epta%le le7el o5 imp6rities( depending on the %alan1e %et:een positi7e ! and negati7e ! 5eed%a1> e55e1ts. 9his means that( :hen rea1tion 1onditions are 5i xed( prod61tion ! rate 1hanges 1an %e a1hie7ed only at the expense o5 large 7ariations o5 the re1y1le 5l o:.84.+ '4( and 2I . 0ig%re 7.presents a simpli5i ed 5l o:sheet( :hi1h 1on1entrates the essential 5eat6res the G %alan1ed H "C3 te1hnology( as 1on1ept6ally de7eloped in the pre7io6s se1tions( %6t this time :ith the three plants and re1y1les in pla1e* 1hlorination o5 ethylene 2M/4( thermal 1ra1>ing o5 EDC 2M24 and oxy1lorination o5 ethylene 2M-4. Nig6re . 9he in7entory o5 main 1omponents and imp6rities are interdependent( %e1a6se they are 1o6pled thro6gh re1y1les. Fn the 1ontrary( I .. '4( 2I 2 4 tri1hloroethylene 2n%p -.0 '( the top temperat6re( the 7olatility o5 I / ( I 2 ( and I ..+0 and /. Initially( the amo6nts o5 "C3 prod61ed and o5 5resh EDC 5ed in1rease. #o:e7er( these 5l o:s soon de1rease to the initial 7al6es.( 5rom plant:ide 1ontrol three imp6rities are o5 parti16lar interest* 2I / 4 1hloroprene 2n%p --2. As the re1y1le de1reases( more 5resh EDC is ne1essary to >eep 1onstant the rea1tor ! inlet stream.: Plant+ide Control of !m$%rities 9he in7entory o5 imp6rities is a plant:ide 1ontrol pro%lem a1>no:ledged in ind6stry 5or a long time.0 000 >gDh to 8/ . 7.. 7.=.+( 0. I / and I 2 are G %ad H ( sin1e the 5i rst 1an polymeriAe and pl6g the eB6ipment( :hile the se1ond 5a7ors the 1o>e 5ormation %y EDC pyrolysis. I / and I 2 sho6ld %e >ept at lo: ppm le7els( :hile I . 9he dist6r%an1e in1reases the rea1tion rate( :hi1h res6lts in more "C3 and #Cl..has a 1atalyti1 e55e1t on the "C3 5ormation( in some patents %eing introd61ed deli%erately. 9his 1ol6mn re1ei7es 1r6de EDC 5rom three rea1tors. h. #en1e( a systemi1 approa1h %ased on the B6antitati7e e7al6ation o5 the re1y1le e55e1ts is needed./. 9here5ore( the top distillate o5 S2 1an easily remo7e I / and I . 3oreo7er( all the 5l o: rates are 7ery sensiti7e to dist6r%an1es i5 a 1ontrol str61t6re 5i xing the 5l o: rate at the plant inlet is 6sed. 9his pro%lem 1an %e handled systemati1ally %y means o5 dynami1 sim6lation and 1ontrolla%ility analysis.4 tetra1hloromethane 2n%p -0+. In the same time( the le7el o5 imp6rities at the rea1tor inlet de1reases..12 8rod61tion ! rate 1hange o5 the "C3 plant. It is also the pla1e :here three large re1y1le loops 1ross./2 %4( the rea1tor ! inlet 5l o: :as in1reased 5rom . In the se1ond sim6lation 2Nig6re . )ote also that the top distillate o5 S2 1annot 1ontain more than . A5ter pretreatment( the 1r6de EDC is sent to p6ri5i 1ation in the distillation 1ol6mn S2( :hi1h is the >ey 6nit o5 the separation system.( %6t not I 2 .is prod61ed in %oth 1hlorination and oxy1hlorination steps.are sent to :ashingDdrying in the 6nit S0. Cr6de EDC 5rom M/ and M.o5 I2..relati7e to EDC is a%o6t /. 9he implementation o5 1ontrol str61t6res %ased on the 7ie:point o5 standalone 6nits 1an lead to se7ere 1on5l i1ts. 9he top distillate o5 S2 sho6ld remo7e the light imp6rities mentioned a%o7e( :hile the p6ri5i 1ation o5 EDC 5rom hea7ies is 1ontin6ed in the distillation 1ol6mns Sand S. At -. 9he separation o5 imp6rities in S2 is a55e1ted %y 7olatility 1lose to an optimal 7al6e. Small 7ariations in the 5l o: rates o5 lights and hea7ies are o%ser7ed. Dissol7ed gases and 7ery light imp6rities are remo7ed in S/( and 56rther in the distillation 1ol6mn S0( :hi1h is the exit point o5 the light imp6rities. Plant-ide Control o/ +mpurities 222 225 7 Vinyl Chloride Monomer Process 9he plant sim6lation 1onsiders only a red61ed n6m%er o5 6nits( %6t dynami1ally representati7e( as 5ollo:s.

It is :orth mentioning that these 1ontradi1tory reB6irements 1annot %e 56l5i lled %y any standalone design o5 S2. 9he three B6ality spe1i5i 1ations regarding the imp6rities in EDC( a7aila%le %y dire1t 1on1entration meas6rements( s61h as %y IM spe1tros1opy or online 1hromatography( are the outputs o5 the plant:ide 1ontrol pro%lem. 9o pre7ent the a116m6lation o5 I 2 ( a side stream dra:n 5rom S2 is sent to the rea1tor M/( :here 1hlorination to hea7ies ta>es pla1e. 9he e55e1ti7e 1ontrol o5 imp6rities %e1omes possi%le only %y exploiting the positi7e ! 5eed%a1> e55e1ts o5 the re1y1le loops that are %alan1ed %y the negati7e ! 5eed%a1> e55e1ts o5 1hemi1al 1on7ersion and exit streams.introd61ed in the plant. With respe1t to design( S2 is 0ig%re 7.reB6ires only a 5e: stages.( respe1ti7ely. 7.m6st %e >ept aro6nd its optimal 7al6e at 2000 ppm. are sim6lated %y rigoro6s distillation 1ol6mns( as sie7e trays.8. 9h6s( the hea7ies 1reated %y the 1on7ersion o5 lights 1an lea7e the plant thro6gh S. We may also 1onsider manip6lated 7aria%les %elonging to the small 1ol6mn 0ig%re 7. a %ig 1ol6mn :ith a large n6m%er o5 theoreti1al stages operating at high re5l 6x. 9h6s( s6pplementary inp6ts are* D0 E distillate 5l o: rate( and W0 E re%oiler d6ty. 9his ne: rea1tor( designated %y M0( pla1ed %et:een S2 and S0( gi7es the opport6nity to de7elop ne: 5l o:sheet alternati7es %y rero6ting the streams( as depi1ted in Nig6re . Fn the 1ontrary( S/ to S. 9his approa1h is ne1essary %e1a6se yield rea1tors 1an misrepresent the pro1ess./0 . As sho:n %e5ore( the rea1tion net:or> sho6ld %e 5orm6lated so as to 6se a minim6m o5 representati7e 1hemi1al spe1ies( %6t respe1ting the atomi1 %alan1e. A5ter thermal 1ra1>ing the rea1tion mixt6re is B6en1hed and 1ooled 2nonpresented4. Fn the 1ontrary( S0 and S. )ote that all hea7y imp6rities are remo7ed %y the 1ol6mns S.. In the steady state all the rea1tors 1an %e des1ri%ed %y a stoi1hiometri1 approa1h( %6t >ineti1 models are 6se56l 5or M/( M2 and M0 in dynami1 sim6lation ?.and S.( 8@ . A maCor dist6r%an1e o5 the material %alan1e 1an %e sim6lated %y a step 7ariation in an external EDC 5eed. 9he B6ality o5 the EDC sent to pyrolysis m6st 56l5i l stri1t p6rity spe1i5i 1ations( %6t too lo: an imp6rity le7el implies high energy 1ons6mption. may %e 1onsidered %la1> ! %oxes. I / ..2alternati7e C4.( :hi1h :or> in tandem in order to limit the losses in EDC. Nrom a sim6lation 7ie:point 6nits S0( S= and S. 2alternati7e A4( ret6rn to S2 2alternati7e B4 or pass dire1tly to S. A se1ond signi5i 1ant dist6r%an1e is the amo6nt o5 imp6rity I . 9here5ore( it is rational to introd61e a spe1ialiAed rea1tor 5or the 1on7ersion o5 nonsat6rated imp6rities in hea7ies %y liB6id ! phase 1hlorination. Be1a6se o5 the 1onstraint on I / ( the top distillate o5 S2 1arries :ith it a signi5i 1ant amo6nt o5 EDC( :hi1h has to %e re1o7ered and re1y1led to S2 %y the 1ol6mn S0.1* Base ! 1ase 5l o:sheet o5 a %alan1ed "C3 pro1ess 5or handling imp6rities. 9he re1o7ery o5 #Cl and the separation o5 "C3 5rom 6nrea1ted DCE ta>e pla1e in 6nits S= and S. are small 6nits %6t o5 parti16lar importan1e( %eing the only exit o5 light and hea7y imp6rities( respe1ti7ely. Plant-ide Control o/ +mpurities 227 229 7 Vinyl Chloride Monomer Process S0( 1onne1ted to S2 %y a re1y1le( %6t dynami1ally 5aster. 9he 1on1entrations o5 I / and I 2 in the %ottom o5 S2 m6st not ex1eed /00 and =00 ppm( respe1ti7ely( :hile the 1on1entration o5 I . 9he separation in S.1/ Nlo:sheet alternati7es 5or the remo7al o5 imp6rities. By ret6rning the %ottom o5 S0 to the rea1tor M/( some amo6nts o5 imp6rities I / and I 2 1an eliminated( %6t this operation 1ompli1ates the 5l o:sheet. 9he %ase 1ase and three alternati7es :ere e7al6ated %y 1ontrolla%ility analysis . 9he degrees o5 5reedom indi1ate as 5i rst 1hoi1e manip6lated 7aria%les %elonging to the large 1ol6mn S2*D2 E distillate 5l o: rate( SS2 E side ! stream 5l o: rate( and W2 E re%oiler d6ty.

as 8 ! type only. Be1a6se the po:er o5 W2 is m61h larger than SS2( the 1ontroller W2 E I/ is dominating( and the loop SS2 E I 2 is not needed. 9here5ore( the handling o5 imp6rities implies not only their remo7al %y distillation( %6t also the minimiAation at so6r1e( namely %y impro7ing the rea1tion 1onditions. So %oth 1ontrollers are s6pporting ea1h other in reCe1ting the dist6r%an1es. Re/erences 22: 7. #en1e( lea7ing I 2 5ree( %6t ha7ing the g6arantee o5 a %o6nded 7ariation( is a rational 1ompromise that preser7es the ro%6stness o5 the 1ontrol system. )ote that 6sing manip6lated 7aria%les 5rom di55erent 6nits is not a 16rrent 1ontrol pra1ti1e. Chara1teristi1 o5 this pro1ess is the 5ormation o5 a large spe1tr6m o5 1hloro ! hydro1ar%on imp6rities. #o:e7er( the prin1iple o5 proximity is preser7ed( %e1a6se the 1ol6mns S2 and S0 are dynami1ally adCa1ent.?. A dynami1 1ontrolla%ility analysis is enhan1ed %y 1losing the other loops( :hile in alternati7es A and C this e55e1t is hindered.( 8@ ( 5i rstly at steady ! state. S6mming 6p( i5 the in7entory o5 the main 1omponents 1an %e handled %y lo1al 1ontrol loops( the in7entory o5 imp6rities has essentially a plant:ide 1hara1ter.10 Concl%sions 9he "C3 1ase st6dy emphasiAes the 1omplexity o5 designing a large 1hemi1al plant :ith an intri1ate str61t6re d6e to se7eral rea1tors and separation se1tions. #o:e7er( at this point there is not a 1lear distin1tion %et:een the %ase 1ase and alternati7es. In this :ay the p6ri5i 1ation o5 EDC %e1omes an intri1ate pro%lem :ith plant:ide 1hara1ter. 9he 1on1l6sion is that the loops W2 2re%oiler d6ty4 E I / and SS2 2side ! stream 5l o:4 E I 2 are more intera1ti7e than the loop 1ontrolling I :ith D2( D0 or W0. In the %ase 1ase and alternati7e B the e55e1t o5 the 7aria%les %elonging to S0 on I . N6lly dynami1 sim6lation 1on5i rmed the analysis %y implementing the 1ontrollers W2 E I / and D0 E I .( :hi1h o7er a pra1ti1al range o5 5reB6en1y are almost de1o6pled. Some imp6rities are 1ir16lating %et:een the three rea1tors d6e to re1y1le 5l o:s. A %alan1ed pro1ess is designed s61h that only "C3 lea7es the plant( :ith ra: materials e55i 1ien1y 1lose to stoi1hiometry. Controlla%ility st6dy and 1losed ! loop sim6lation indi1ate that the %ase 1ase and alternati7e B ha7e the %est dynami1 properties. #en1e( the plant:ide 1ontrol o%Ce1ti7e 1o6ld %e a1hie7ed :ith only t:o 1ontrol loops( W2 E I / and D0 E I . 9he 5i rst t:o are so6r1es o5 the intermediate EDC( :hi1h is a rea1tant in the third. In parti16lar( the yield o5 pyrolysis 1an %e enhan1ed %y ma>ing 6se o5 initiators( some %eing prod61ed and . 9he 6se o5 D0 o55ers the %est de1o6pling o5 loops.( are s655i 1ient to >eep I 2 %et:een %o6nds( %e1a6se the dist6r%an1es a55e1t the o6tp6ts in the same dire1tion. 9he re%oiler d6ty W2 and the side dra: SS2 also a55e1t the imp6rities I / and I 2 in the same dire1tion. 9he last modi5i 1ation o55ers the shortest path o5 imp6rities and 5aster dynami1s( together :ith %etter prote1tion against pl6gging. 9he re7amp o5 the plant :o6ld 1onsist in adding the small rea1tor M0 5or 1hlorination o5 imp6rities( repiping the 1onne1tion 5rom S0 to S2 and repla1ing the internals in the lo:er part o5 the 1ol6mn S2 :ith 5o6ling ! resistant trays or pa1>ing. 9he rates o5 generation( mainly in 1hemi1al rea1tors( and o5 depletion 2exit streams and 1hemi1al 1on7ersion4( as :ell as the a116m6lation 2liB6id ! phase rea1tors( distillation 1ol6mns and reser7oirs4 1an %e %alan1ed %y the e55e1t o5 re1y1les in order to a1hie7e an a11epta%le eB6ili%ri6m state. Intera1tions thro6gh re1y1les 1an %e exploited to 1reate plant:ide 1ontrol str61t6res that are not possi%le 5rom a standalone 6nit 7ie:point. 9he a%o7e analysis emphasiAes that the most signi5i 1ant impro7ement 1ame 5rom the 1hemi1al 1on7ersion o5 imp6rities %y diminishing the positi7e 5eed%a1> o5 re1y1les. 9hese originate in the three rea1tion systems( 1hlorination( oxy1hlorination and pyrolysis. When the intera1tions %et:een 1ontrollers are ta>en into a11o6nt( only t:o 1ontrollers( W2 E I / and D0 E I .

28+0. *ng. ( Ind6strial Frgani1 Chemistry ( Wiley ! "C# ( 200* Co:5er ( K.D.D.A.K. ( Dimian ( A. *ng.C. Commun.K.re1y1led in the pro1ess itsel5. E -+ ( /++2 . ( Am%ler ( C. ( #ardman ( M. ( Me1o7ering 1hlorine 5rom #Cl ( Chem. 3apan ( 6.3. ( Mo7aglio ( 3. ( "inyl 1hloride monomer . Dire1t 1hlorination and parti16larly oxy1hlorination are important so6r1es o5 energy. Sci. Chem. ( 3ori>a:a ( #. ( Iedema ( 8. Res.W. ( 8rod61ing high p6rity "C3 prod61t ( !ydrocar2on Processing ( /2+ E /-0 ( A6g6st /++2 5 Sae>i ( .. ( 9e1hni1al progresses 5or 8"C prod61tion ( Prog.E -80( /++2 Weissermel ( '. ( 3agistro ( A. Polym. ( Arpel ( #. ( 20.. ( pp. ( 4' ( //+ E /-2 ( 2000 9 Dimian ( A. :hat yo6 sho6ld >no: ( !ydrocar2on Process. ( 3.8. E 88. ( 8ar>s ( K./. ( Nryar ( $. ( $roenendiC> ( A.1thmer *ncyclopedia o/ Chemical echnology ( 2" ( 8=. ( 3odel e7al6ation 5or an ind6strial pro1ess o5 dire1t 1hlorination o5 ethylene in a %6%%le ! 1ol6mn rea1tor ( Chem. ( 0-.K.22 ( 200/ 12 Wa1hi ( S. ( 667 ( /.E.80 E . 9he "C3 pro1ess o55ers good opport6nities 5or energy sa7ing %y heat integration( namely in distillations.K. ( 65 ( /+8= 10 Clegg ( I.3.+0 ( 200/ : Bost:i1> ( L. ( /++8 11 FreCas ( K. E 2/-/ ( 2002 7 $roenendiC> ( A. A=( 2=. *ng. *ng.. "eferences 1 <llmann X s En1y1lopaedia( Chlorinated #ydro1ar%ons( 7ol. Chemi1al 1on7ersion o5 imp6rities a116m6lating in re1y1le pre7ents the o116rren1e o5 sno:%all e55e1ts that other:ise :ill a55e1t the operation o5 rea1tors and separators. *ng.. ( Iedema ( 8. ( 45 ( .( /++2 / 318herson ( M. ( Em6ra ( 9. ( &' ( . ( System approa1h 5or e7al6ating dynami1s and plant:ide 1ontrol o5 1omplex plants ( A+Ch* 3.C. E +2 ( 3ar1h /+.( /+8= 1* ManAi ( E.A. Chem.C. ( Ba>er ( W.. ( Vinyl Chloride0 in $ir# . . 9he handling o5 imp6rities is at the origin o5 1ontrol pro%lems regarding the plant:ide material %alan1e. ( "inyl 1hloride prod61tion pro1ess( <S 8at .+ 2 Wong ( E. ( Dente ( 3. . ( 8yrolysis and 1hlorination o5 small hydro1ar%ons ( Chem. E 00. Sci. .K.E . ( Star>s ( C. ( Me1y1le intera1tion e55e1ts on the 1ontrol o5 imp6rities in a 1omplex plant ( +nd.