CONCEPTUAL DESIGN OF CHEMICAL PROCESSES James M. Douglas nie of Masts(CEPTUAL. DESIGN OF CHEMICAL PROCESSES etry ma ato. Saree Sopeetiton cen ova cogs Sir St ny Meine, het gar rs. Tereca Cyan monn oueeinsaasonser See tt rete et ‘Re prsceon sparc nae ae et toe Ken, Lust Canta Catton Patton Dutt ow ae eres et) ‘orcotn Sega cheia roe. cen a hal ergemeng Stes) barnes ‘When ordering this the use ISBN 0-07-100195-6 Pitan Sgnoe ABOUT THE AUTHOR lames M. Doug PLD. is cane profesor fehl entering she University of Masschusits Prevon he ight atthe Univesity af Rochester sod.arthe Unnersiy of Delaware Bee coeing teaching. beep he yet at ‘ARCO, working on eacor design and contol problems. He has pushed neste in tease reacting eaieeig procs conto cng two Book ‘Sul conceptual proces doen He won the Post Dectoral Felowatip Avert a ARCO, the Faculty Felowip Award at the Univers of Masocbset ad theDeicarsD 10: The loves of my li, My lovely wife, Mary . (Betsy) Douglas, My mother, Carolyn K. sad the memory of my father, Merl H. Douglas, My tno wonderal kids, Lymn and Bob, and to my colleagues, who have taught meso mach about ‘esi and contre, ‘Mike Doherty, Mike Malone, Ka Ne. and Erik Yast, snd to my students, who have sllered so mutCONTENTS, Part 1 Preface A Strategy for Process Synthesis and Analysis. ‘The Nature of Process Synthesis and Analysis Engineering Eeonomics ‘Eematg Copal sd Opening Coste TC sine ol Tot Fe Co ‘Shepig th Eons Anal fe Concept Doi Economic Decision Making: Design of Solvent Recovery System amen ss ConseonsPart I Developing a Conceptual Design and Finding the Best Flowsheet Tnput Information and Batch versus Continuous Input-Output Structare of the Flowshest ‘Dron fr Input ura Secure ‘Belge Warsi, veal Mave! Bic, an Steam Cost Recycle Structure of the Flowshect Fiera ns ews Lents Spree Desi and Cons Tock Eee Ba Separation System ‘Geet Snaefthe Sepeeion Sen Heat- Exchanger Networks Minn Henig and Calin equemens ‘De a anon Emery Hex Exner Netwoes ecg ants Fedeog Be Nomber xchange ‘Se sac am isnt Siig Be wt Part IIL 0 2 ‘Cost Diagrams and the Quick Sereening of Process Alternatives Su Sonne ence Aerie Other Design Tools and Applications Preliminary Process Optimization Con Moder ese Uae penne Orrin Process Retrofit ic Poe Moen Pais Computer-Aided Design Pre (LowiKay Mesias Ge Cee Son ‘Summary ofthe Conceptual Design Procedure land Extensions of the Method ‘ter Spice Asp ote Desi rls Appendixes 2 Shorteut Procedures for Equipment Design ‘Dg of Cat Arie aod Detto Clucomps Ditton Co oben Doe ‘el of eagtin Sys yo Sh ime Dp Ge HDA Case Study Design Data yfroarton Vapor Lig Eauiaie “Feapontre Range ore Mater FLOWTRAN Input forms Shur ‘Cost Data Stmmay Cou Corns Conversion Factors Indexes ‘Author Index Subject Index & BE & o i PREFACE, ‘This aok describes systema procedut forte conepual sip of imited lass of chemical procanee The eal af + coneaal Sigh 10 find the test oes fowshet (2 to select te proses unit andthe nerconecions ang {how unt) and estimate the optimum design conions The problem i ait cle bcaese very many procs allratves coud be coneed. In on expriene indicates that fs than 1% of Mess fr pew desis eve Become To many cscs the processing costs asacited with the vaio press strates der by a order of master moe shat ne ca te shotet ‘ales tosceen the aernatines owen west cri tat wee the eihorhood a te optimum desig condos cachalleratve to prevent ‘Ascarding an lteratve becaeof a poor choice of deh varies, Hence, we the os stuie a am ial screcsing to eliminate ses for dengue ht se prostate Ira proses appear to be probable then we mast some ther ‘tor aclaing say, enronmentl contains control, ee "Weapproath be syathess ed analas olen by tabishin a ieraschy desig decisions With his approach we decompose a vey large and compet ‘roblem ino numberof smaller probiens tha are moc simple to bane By {Seung on the ecsone tat stb ade teach even bea oD ve want 10 nd a solent recovery system), we can Mealy the eng {teholopis that ould be ued ole the problem eg absorption acrpon Smdestion) without recding the posta tht some er echogy (2 ‘membrane proces) might provide a beter suuion. Moreover, by Ising ie Sierate stuoas we can propose foreach destin, we ean senate) enrte sat of proces atertve some cute i pone to wie dig gudtines (re of thumb or heute) 10 make some decom abot the sracure of he Aouatet ander 0argent to eve many of thee beuitn ad we we spe amass i ‘pets un th bwin mayen es Titer making doston 5 Katowing this hoch deson procedure begining designer ean abut be eveTeation anu of exes eaations fr expenence rng ROsNgment af conspral deogn Since shore calcltions ae wel eeseelfe penalty rad in the tine une To sre more aerate is el Ae high OP course asa Senger gum exec, she othe wil be abe 0 NZI ss aeratvs do not ned ote sone for paral type of ‘rece an thereby obi anna in efcncy. Note alto that erpenece ovmalyreauied for nssing the operbiity of desig, and dhtloe a Sew si aye taper eae even ee he Onpaniation ofthe Text ‘Thetentismiean tobe wed ns one-smese, eo lee course in roe sia ‘ckonsatengincriog svete We pee tbe material as 9 etre course. A ‘Mage em sy cated thsughowt the ext to ste the es and te SRS eiementsncide the evaluation a ieratves forthe cena as [dyn wet several oer coe sien The purpone of hese aber ene te ‘Stolp the stodeyt understand the stares ad ferences beeen YS ‘Spevolrosases (ce. singe tents era product Strbution problems cues ‘tere pavsecjoe oss dominate caer bere gid separation costs dominate The cklae between royelng oF semonng hyprodate formed by reversible Feicons te tconomie trades coountred when a gas recyle 36d purge Shen sed te) The som sereeing aeons altogh compat {ded devgn popar cently ued to wei the approximations Part dcuses«suntegy ol yeti apd arly In Chap. is ote that ony aut 15, ese for now digas ver become commerce, hat we eed an ent procdaefor elimiating poor promt. Sia se Asn ‘bien are leas anderen and ween fe pnerat 10" 010% aera Peete ene forage pode! plan, me eed ent way of screening Ricco alerts. Thee decison provide the moviation forthe we of ‘ortot elation Alo, x prose fr decomposing prose Nowshets into Hearts eof simpler problems is preate. ‘Chapter 2 prevents an inedocton to engineering economic, ndoding 3 iscusion of varus menses of oabliy addin, 2 simple enaomie nett is fl for conepual Sips ix developed ‘Chapee 8 preset sry sme desi pron (actualy «aust of “sha cull bea larger despa pote) Thexampe uss how simple 40 Fee procan armen tne nd fr design hers the rip of Sein etc te lnstatons of sgn heer, the inrctons ame prexssing try the ned fora syne impo plnc as unt operations vow and ow shorn design methods can be Sve rv pests the deals ofthe Hsarticl deco proved fo the synthe ant aya emcee design. Chapter # eerie: the information IESG fogs andthe dehion of designing a bah vera a cotinous ESE Saconed,Chnper's prevents the portan decsons fo he input Pepe aractue the erlfcation othe portant design varies theo aes andshrtut procurestocaletatethe steam cots andthe cos of reins owe sewed) Cheyer introduces te addins Becsions ‘Sade te overall ray strctre othe whet, i, the tration of Ee Stemi) with he ems the process Tae ekto os an any REFS Spectr ents ve ead terms ofthe desig variables Ts ‘Ercan i itd to gl rod a ‘A preset the systematic prebmiary deen posure alo ited to sap lid pcre Forth cas f process te strc ofthe spaaton Tyres es tbe geal suc, vapor recovery spt aller andthe Seen is te iid separation sytem end in Chap. 7. Chapter hen see (mess procaine fr the hex encange network Att point Fee eaten existe othe opium Sega conditions ae eve. ‘Ger bane Jsgn sag) eto develop bases dein 3 Tp ss pow ply tn the process aernates a we go sory, to deine ‘eter tere fs semethng aut te pcs that wll ae ll the arenes ‘TBsotahe Provided that our hus cae desi pens be romion Ne we srestnds in Chap 8 to seren the proce lentes Thu, 8s ponte ‘Mem to hey the test pees Howshec Pat if resets soe othe dsp took ad apis Ta the posed reset it Cape though 9, ne sie ease sway exkltions to xine the ‘irs dxpncoditns because ne were continually changing the stutae of weipouatcat ace we have Hdened the fet fone. we can se Tove saan rors Heron ee et “Fe atenimatcapamvation rocdure eps osdeniy he dina evononi {ied for cach ds vat the Gmina! desig rane, and am indi \Soworow ara deg rarble sway frm he opera without knowing the CERNE ore cptmun. Ths approximate optimization analy ao very SER efor ads nd for pteon steadytate conte alton i Chop tlm os the sume echrigusfor proces eos hal we ed © evcop degn or 4 tex plant A systematic prvedore i presented for erceting process incline vomit eplaang the exising pan wih ether, Fame ge er prow aleeatve: The approximate optimization prosedie We hap teny the dominant operating vavables and the equipment ‘da ta preven he operating on fom engined. The, hse on ‘Pee ree aime quent cacy 12 aed unl the incremental imatoed equipment cot falnces the sncemental decrease im operating tn Chap. 12 we dics: the use of computed design progam 10 mpreete fcray ofthe shrtat leans, Chap 1 pees a omen Ti dean proved breoutines of erachical decison proce for sesaod butch proce, and be dscsion of what remains tobe doe er & ‘oneal Jig hes bee come. “The spencer some auilayinfrmaton. The sbortt model for ‘equipment spn we scased In Appt Aad he complete deta a ce ‘Siedy ar gvena Appendix Some supe Jig data ato data ar gen iWAppendies Cand E Acknonledgments {am very appresiative of the ers of A. Esc Anderson (orm wth ARCO), DDuncen Waadeock of Iperal Chena Indust, Edvard €. Haun of UOP Ine, feof Keroy, Univesty of Note neCa King tom daPont FL Sheth fiom Eston, Hoch (ormery wth Halcon Ieroationa, Jobe Sed Students at the Uae of Masschseis and othe stdente fom Imperial (Cherie! Indussn (United Kiogloa). Rohm aed Haas, Monsata, Was Carbide and Cele for many vlusble comments eonceraig the couse ‘Butea In adion, I mun acknotlaige the umcrows contin at allenge Mike Malone made to the tx and wat fo thank my other alge ‘Mike Dobe, Fk Vie and Ka Ng rth eed when he tap he saterial The contbutions of my gradu teas, partly Wayue Fisher, nd al Kirtnod, se not to be acknowledged (Ofcourse Tam epi rac t my lovely wie, etsy, my ee, {yan and eb, and tomy mote, Calye K Doug, forthe support Sig the pepuration ote tat Sir, Pat Levis my asta at, ‘ar Burscherakl who did the (ping provided much need support. James M. Dsl CONCEPTUAL DESIGN OF CHEMICAL. PROCESSESPART I THE STRATEGY AND ANALYSISCHAPTER 1 THE NATURE OF PROCESS SYNTHESIS AND ANALYSIS 1.1_CREATIVE ASPECTS OF PROCESS DESIGN “The pupone of engesng isto create sew material weal. We attempt to ‘compl this goal in chemical eagaetng via th chemical (or Bolopea) {tasormaton anor separation of teas. Proce abd plat design the ‘reative nti whereby we generate eas nthe nate them into ie id proceso producing new mater or fr siaeantspgrding he lot Th any paral compay, we might try o generate now es To produce purchased raw material To convert a mast by produto valable product “Toerete «completly new materi antete ers, fod, oprocsing) ‘To fad a new way of producing an eiig prodect(¢ ew exalt 2 boprocesing ternative) “Toeaplot a new teeology (enc engleting, exper ste) To exploit a ow material of constachon (hhicmpraure- oF high- pressure operation, specially pynes)4 ‘Asan indication fhe tremendous sce of the eapioering elo, we noe that ter 20% ofthe product sod by mache companies were devloped dane Theos decade co. acess Rates eet this exelent record of ices we sould reat ver few ne eas, tit or mprovingexting proces or fr devoping tw roca ea oh Seal In ic he Chan of commercaeton a the svar stg fr 8 roe areclyabout tos athe decopnent sag ty ae sha Oto 3° nate plo punt tage yar abot a0 To CO%E" Ofcourse, me exp that {he cures at for proces moisten wl be hghr than aor oly tte proce bt te economse ears pcated with thse se jes wil ive apcanty lover potent icine srpsng hat 0 ew es in ning cer prove ob uit the same pcm fds for any tp cet Seely Sion epee nates {hat on toma umber of eacer wire meyer ‘nv cf te ot sgn componet of dss method nc proces ‘Suess fey he lection of eqepoen ade inreonnetrs betwen Tat Gripen! wich willackeve a eta gal ella combionon of yuh he Snap cy. er Synthesis and Analyst "eh mse ie i ie a eof {Lib ntaton nied crs oop robs wart ee problem Steven For exams a hemi might rer a new reaction fo make ae ie comme rain. tnd we Tinea the rescton conden tha we tao the hems well ot me formation shou seabed oducts tht we stn rome ‘Sr marsting npnton and ter we ce to opal the ter rman Shure ne todne Sogn oben “Tomy the nicingsuorano, wernt make sumptions bout hal types ef prot ek tuk te od ow howe proce wi toon ‘tc an fas empestne rome ad pron fw ts wl ee ‘Tht he pth Sty. yt sb eer there ae vr rae ‘mes (10 1") mye weigh ear osxompch he eB Ficmedun poems never open endo, Normaly, ne nan fn the proses lene (ot ofthe 10" to 10" cut tat he the lowest ct Put we must alo ese ha The proces wil sty enone coms, easy to stat wpa pea en SES sen we eam ure ae of hum (ori) to elmiate certain proces iSntwet om fares consideration tin may a6 is wees 1 dese et Seamvr and then to compare ther coms. Expeseneed dines ean in ste afealaton bets hey can fon tet Freee Ohs urccalr ant or pup of wie, by analog to antber rece. ower beponing designers normaly mun design and ealate more sere their ta fe the be alert ‘hen cpernced deer consider new ips of problems, whee they Ick capecens un eere thy eanet enti angi ty tse sores (hack ine envelope) desi pore asthe basi for comparing alternatives. These chcotchecmlopeelalations ae ured oly osren alternative. Then if he tcr Spas tobe profable more sgorot dene cleans ate wed to {cipal sin forthe best aerate or the Bs fw erties. Be titundertned an pen sided wear of eign probes and ‘cau athe low soces nte iwch to devlop a ategy fr sone dese Protons: We expect thot he strategy that a begining deiner would wer ‘othess and anaiys would Be diferent from that of a experienced dear. ‘Zee tegen mest cvslante many more process erat, However. by Naethouteat despa proedues wean imi theo required to undertake ‘Gee aoa ealeuation. ngiocering Method Ite rec on the tere af proces phe ada as dead shoe, We recone tat proses desi scoala aL a creative proces Three ‘Tein ty to approach design problems much the same way a8 aie? “vclps a paiing In ctr words, out orginal eign procedures shosld reson to the development pens seth, whee we wan 0 spre all atthe mos sisi Gta of the design, Le we mato over he mos ‘Spemive prs os prces ad the signifi economic trades Anat EX ‘Mums the prehinnary ping a makes odieation, wing ony 0 inte ofthe vate Sina, we matt evaluat fat goo a ese a geerte a uber of proc ternative at mph ead wo mprenecat thisway, we bpefo neat eaonab king rush procs deen Blore Then dhe art nd elo, shading andthe etal of various oct nthe aiming ceases the ess Majer modesto ay beintrodaed the ‘Erma be naroned Ins analogous maret, the ener wes Oe FPOCOMS SSipn ama coating roses or he ot expensive eps Hes, prON SERN atthe apponiate rater and energy balance exclton, a6 ‘Me etain ermeot the sal inexpensive equipment items hate cesar for6 seen cata emer oF ce ns th acs pero bu do note marine onthe tl el es, Ts, mes that both painting 40d proces design proce rough & sere of sucsively mre dead syns and svatation tages Thacher ry {o'asolaionsratgy of the ipe as susesve reece fe cls he fagineerin method Neve that at we make wsttsne inact, we sould “hee malatan focus onthe orl rb. we acept this analogy beeen enprcring design anda, hen we can secogaze some ater intresting atres fthedegPrces Ah at Boe ‘really completes «punting: acemaly the wook ft termined whenever Te ‘Miioat efor reache a pont of dinning sears; sit ie sdded value ‘Somes fom much additonal er, thee ot wart Another etre of am shat theres never ingle sluton toa problem; Ly thr ace a rant of ‘wis of pining reat” Madonta and lil or landscape din pees fpcering ermal deren procing routes can te wid to pone e sae ema for eset the sume cost ‘Stl another analogy betwen eopnerng ‘esen and st that request fo desde how mach den shold be Ince in the various tages of plating orto does pws desi ‘OF course, numereus scenic princes are wd ithe evelopment oa ‘desig, ut the overlay wan af Tafa tt combinaion of wens ad fe erent actviy that hf make pose ge ch # sin Levels of Pngincering Designs Now we se that thete are # numberof eve of engineering designs and cost ‘stmt that me oxpes fo undertake These very fom very mpl ed rap fan mae. Puli nd Diaz cs these den eimatsby te caepn 08 feTabe ich “hey alo gv the elatv cot rete o aban he ena, show in Table? 12 From th tbe we ss bow apy engincrng canteen te inclade mor dea inthe eletone Obvious we want fe avon ltpe design cout rls they eam be economical sti TM Tar hefner of Chal nse Mar Clin Ohi 90, ymin nti a ng Pa a Ch “tie ~25% tates pn dn ves ca) ery heresy stevie ee nn “Type of design eimai 2 Sindy eine ae sina alo bone ree Ferenc sen esjn emn «Buen wimae et cnel ent) bdo no opie a he cnn roche an ona 2a seen us Tk would nr be pons o deep an oreo magiude extn However, an experince Giger can overcome ths cay by deoving aso {es tween the ew poses and other esting proses for Wc Some dat ae {lable the company fer Procedures for devopng onder of mapa {Strats have boo desc in the heratre* but somal rege ome Fora beginning designs, white or no experienc, t wold be url 0 ave asitemaic approach for developing one negte estimates, Wea ‘Seorderafmagstudeagumertsto simpy many othe dig clclston, aod ‘Secan limit out atetion to the major pce of proces qe awe ary ot 2 peciminary prowess den, The pal ofthis tet i To develop » stems Sv once dr Maing hl ap ain” Cag. sata A ae Ys Ps Co “Engiering co to prepare tite (197)5rocedut ofthis type and then o sow how the eats cn bene 0a sty ‘esl evtimates ae considered to be beyond the Scope of his tet Honever ns noed bio the Gunethat new en eve eoties commer ‘penly abou 1 3 that we expect unevake roughly 100 primi designs forenery detailed espn Heme the ethology of conceptual pres desen ‘Mol he mered ie onsderbl deta ‘Other Applications of the Methodology Despite the fact that our primty fc dred othe desig and eatution of sew procesen chef the thology we develop i sel for ther enn Tis incaing bs research nd techn service In siya, we wat to ‘quel most of ou fort stdyngthove nabs that wil have the greatest ‘Sonor pact the ross a cough proses dss wil help oie the Fipvcon parte the procs andthe domiant design teva. Sisal tenia serie acti we look fr wage ofmpronng am exsting proces To ‘complsh ins gah we nerd te ndestond the sigan eonomi aden the process, and it afl to ave precedes avable for obtaining ack ‘Sites the potential payout ove iene Thur the methodology we evloy ‘Stave numerous appiaon i he proces ndastes 124 HIERARCHICAL APPROACH TO, CONCEPTUAL DESIGN “The engineering method or he ati aproac) ines ha we shold solve dege prokiems by fist developing very single solutions and hen ang Sessae ner of det Tosoe how ne cane this approach lor procs eign ‘probens we conser pic flows firs petrchemial proces and then fe Eck Gr vays of stripping aay ies of etal oth we oun the simplest ‘problem ote By sping he pocedre toa mune of dierent ype of [rocese we might able to regal eae aterm tht we can use 2 the {iss forsuesting new proceer, Example: Hyérodealhyation of Toluene (HDA Proce) “The example we conser the hydraeayation of ote to prodcebemsene ‘The easton of iter ee “Toles + Hy + Benzene + CH, a2 wd nee = ipo + Hy 022) (ain Se ie akan nt ee recomny and Dre ric [iar }—f cone ne | we [foe [ne 2 is g estat yale ee eat FRE aes Dl ACHE .3°5801 ‘The homopioous econ tke place inte sng fem 15 low the pete the eacien rate Boo Sw) to 130" (Above ts temperate seer sont of hydrocracking takes place) and as presute of about Bethaaas cee of cron) ea) needed fo prevent coking nd he seer ca ga mst oly enehe 0 TT37F ined to pret SHE {Sie bet each sowing te reac. ‘Onc psie Rowse othe proce shown in Fi 12-1. Th ten and sdropee Tamales a bested at combined i eed tle 1 tpatogn seams ttre they are ed tote react, Te padt ream {Sting heer coma pop methane, tose, les a tea Scipheyl We stung to parte os ofthe hydoges and metane lom seer hy wing purl Conder to conteme te aroma, and then We ‘sk'eway the gt pues We ase he Ig evin he Bh a 1 upp {ern congo tt ho eacor pss (oo shown on he Boned "we woul He tory the yon lenin in the at vapor tu he neha ich enc a input hyrope ed um an ase Pee breton 121, wltscirmaat nthe pas ecye op Hee, ae amie reqonel ro remove bh the ed andthe product methane fom te rcs Noe tht ao er of hob (sgn deine) cab lo estimate the Pecan sincesaton af ean tat shoul be allowed 1 accom nt see oy We docs hi dein vale im mch prea dle. ach drei hve weremave ow fe emaining noun aor eins the aubrey pene her hom contaminating or benzene produc.10 secon 12 miaAneICAL APMROACH TO COMGHETUAL Ouse recycle toluene is seperated trom the unwanted diphenyl Other, aterestive Aowshects can ako be dr, and we diss some of thes a we go hough he nergy Integration “The proos Rowse shown ia Fig 12-1 nt very eat beau ines that the heting and cooing requirements for eey poss ste wl ake pce ‘separate heat exchangers sng extra! vies (eoaig water, ear, hee) Th the lant deca 2 ow denignprowdure has een develop that tani pout fed the minimum basting ad cooling lads oa poses and the es. ‘ichangr nework that eves the “bt energy iueyratoe, Ths rocedue ‘To apy ths new dign proceute, we must know the flow ete and ‘compost ofeach proces steam andthe let end outlet temperatures ofeach process stream. One aterative Bove that res ro hs nergy negation {oasis shown Fig 12-2" Now wee thatthe eat prc set ‘ed to partially reea the ed eneig the eco Then the bot eat ate reused to dive te toene eye column rbot pebeatsome more fed, dive the stabi cola robot, lo spp ct he benzene prod! cls ‘ebotler lod, and Yo preest some move ed ar tbe gues ete the pata Seder Alo the ‘oluene cari is prerze, so tht the condensing fempeature fo toluene higher thn the bie pot the oto steam nthe tenes column Wi this arrangement condensing tole cate wed Spy Sie ofthe benzo reo lon ested wing steam and cling wate om i we compre the erg istepratdfowsbext (Fig 122) wth the fowsbet indicating only te oe or heating td coating (Fig. 12-1), then we wee tat the era iteration analysis makes the dowsbor! more complicated (ie thete ae ‘any more inteonnectiont) Moreover, oaply te ee iteration sali, ‘Terma tow the Now rat and compost feey proses stem, Le all oces het loads incding thos ofthe separation system aswell asl the steam {emperatees Since we not ff amor sh the owes ioe we can exp the ‘cooty ingraton stem en eet ads the genet compton the ross owsteet, we consi he egy nteation abajo atelas step in ou proses eign procure Disaitaton Train Lat es ow coor the train of dntiliton clas shown in Fg 1:21 Sic the ‘waned Sipeng ormed bys veverble ect (eg) me oul aye “— Tie ui12 mens Amun APRON OREN By ct erzee eto ro) Dip ‘he digheyt wih the toluene ade i build ep tam egitim keel Thi ‘emutve mould make it pout eliinate ne of the dation colanes. “hough he Now ate thotgh the reacor would ieee Teese acer he pen at Fp 121 renes, we exget hat the ‘olune diphenyl spt wil be very eat. Therefor, we might be ale 10 we 2 ‘Sdotram solr to sceomplsh a emseeolueedieny spi That we Cull rave he benzene waren temuve the lvee sea sete bel the {Eclant rover the diphenyl ars botom sem (oe Fig 12-3) We cam il {hun wey pure bencee verend we taketh tolvenesiderar of below te {Eat The pry ofthe one renyle wil decreas, however Hit ie econred a3 ‘Sst acomped tan overhead produc Sache no seis fr there folsene the pty might oot be important and desing’ might Ye ‘ovthe Sars we expect hat he mth bese it inte stabizer Gey Then, rooting tencre at 4 sertunm ipa Hy and CH, Fence toluene and diper spiter a pasteurztin column (Fig. 12-) aul aml el Dipherst Mere tain sgh be cheaper than aking the configuration shown i the original fowsbec, ei "Tha heuristic (sgn gene) or separation stems equa knowing ete composion of theta entering the dition train Tas bere me Gnsier the decane scot wath the esi f the dilainsei,me ar ‘Sinica We twat tod te he prow Fey he sie beatexchanee ete Vapor Recovery System Referring gsi to Fig. 121, we conser the vapor Row ein teMsh dam. We ASow tht te eves shots na ah ra and therefore that some of the aromatics wil ave withthe fash vapor Moreover, ome of these aoa Site ot in the pure area. Ofcourse we could recover ths aromatics by iting vapor every Sate either onthe Bash vapor iteam or onthe urge "he vapor scorer) system we ould we one of these: Condensation (igh pee o low temperate bath) ‘Aeorption Assorption To simate whether vapor recovery tem can be economical ji ‘se mus atm he ow rater te vot Tout nthe purge a well ms he ‘roges and methane lw in the purge Hence, Beloe we ee the nes) {nore deen os raporrecvery oat we must spy te remainder fhe weet and ne ns etimate he pros ows. We consider te design of he ‘aor recovery system bere that fo the iu separation sytem eau te ‘Seams from the option for «vapor recone} ftem iste above (Bs # AS sorte) normaly cade» ued eam that set fo the hg epee ‘implied Flowsheet forthe Separation Systems (ur goal oi sway of siligng fowsbeets 1 obvious hat Fi 121i, ‘much simple than Fig 12-2 and heres we deid to do he ener iaepravon {SS Simic we have a Eno the proces low ates to dig vara Tus eovey spies we decided to come these design probes jot belo ‘Seenergyntpation. Thay we ean spy the fowsbet shen in Fig 1211) ‘faving tas shows fa Fig 12-4 The cannons between the vapor and quid ‘covery sytem shown i Fp 12Sare dese I ote deta aes We now ask curves wheter all procener can te reprocnted by the simple fowsect shown nF 12S Sins Howshct contains both gat andViger eooery |_PHE sys cM, Be OU peer [Pose a= ‘aie Lai separnion se | pea Fe pn ym Wr 3 ag AME, 39899) vi sey op, bat oe roto do i ote ay seus component ao yah resto gens (See S71 rae aerate) BoSLEE. ep the fava sil moe mp he vac a ai Bowe arn it anv ton Fi 12-4) Tin, we cot the seme eae general sre oh earn se bre me msi TERR rete he vapor ortega veces tem Recycle Strate of the Flowsbest Now we have bined 1 ey sinpleowsheet forthe proces (Fig 12-6), We an OTM open to estimate be recy fom and ther Beto the mee aad sce ofa gavel compres, fan, Moreover, we can Gu ete Pe Hy che Hy, CHe—ef Reactor Sepution JT Benene Tiuere —e] tem | pipes me Ro FSR 4 a er 1M: AICHE, 33878891 —~ Ka Hy, CHy, co Wenzene HOA Mputasyt ecu 1 Bp ANAS, 9 99H) to understand what dsgn questions are important to obi tis simplified "Spretntton witout worrying aba he atonal completes cuted by the {epurton ssi orth carp negation network For example, we ca sy tfetactors ht determie the nemberctrecylesteams, ea elects inthe reactor, ‘ub bation the rato, tT conan o stip aay levels of ‘ll we see that we wart o wy the rence seule of the Bowes before oniering he deals ofthe separation ste, Ipt-Outpat Stracare of the Plowshest Fiewe 12.6 provides very simple Bonsect, but we cone he posilty of ‘Shining an een slr rpresation, Obvious we aa bor around the Complete poses we wil bee wt the fed and prod seams A st nce (cer Pie 12h thi represeration might em oe os ope Dut it wllab ab ‘inksanding the dg vals that fect he vera meals without inducing ay her oenpcatons Sine rax materi ote normal tl The "ge from 30857 athe prot cost he veal ater beans re eran aor ex Alo, we donot wan to spend ny tne investigating the dag vale inthe ranges ere the rosics and by-products ae worth tem than the raw materia, we coer the npu-utpa strctte of he Flows! andthe deiogs ut ao die siacture oe we onset an yee Pens Limitations 2y snensely simpliping fowsbet, we can devop gee prose fr acting deige protien However, ou opal Dowseet destibed «conti: foun sapor hg procrs that produced single prodot ad ivohed nly Sieh gear or hyrtonc) her a e o cham ng) Prog. ato oe some ve emaresomod cheiprocdate in greater dea However, buch procenes may have a somewhat ‘iter waded stuctre (ne oes carey ent fe operation in» snd ties an cert they ae decid dierent im terms of mathemati ‘rode (ooatly ondary divert or paral diferent equations neal of ‘pst equations or ordinary ailrentl equations) Hen, o fs deeson ‘BEiy should fst dings Reeser bath ad ontinaous process Mieatehy of Decisions ‘we collet thereat dca abov, we can develop systematic approach 10 rose sgn by retusa the desgy problem 10. herarchy of desis ce Fane 12:1 One peat sdantge of hs approach todos htt allows wo {aula eguipent sos and oesimate cons a we pce through tees in the erry Then ithe potent pot bosoms agate a some eve we a look for» pros alerative o emanate the Jsgn projet without Divine 0 tht complete ston othe role “ante advontae ofthe pecre aries fom he fe that as we make cisions abou tesco he Rowse taro eve. we how tha we ‘Sane thes secon, we. wal wecate poser aerate. Thy wih 3 stati desipn poser for senting alters we re much es ike) to ‘Cheick some imprint choices The goal «concept dein Wf fr the Stes alcnative ‘Shortcut Sola Experiences tht itt usualy possible to pata avery rp umber ‘he 1019 ofalteranve bens or ay proses al te poss re imide. Hen it usa to be ble to guctly reduce the nurber of ‘temotes that we eed 10 conser We normally cee hee aerating ‘Sncrotrmgntae argoments 10 simply the process material tlenes he ‘iuipnent design eqattons andthe cov cleitions These shortet aeons ‘Mer are satieenty scart to clininate the 907 oro, ofthe akeoatives ha {f'n corespun! fo poise operation. Then yes and salsa 2 hays sete to spre sletion, we repeat all the clestions more sigroul,besause then ne ca jst the sonal engiesing ef “The ut of shortct solution and the herrea decsion proved sho {ollevelop «proces That i alert chem routes could Reed o make the ‘Sine rode with age miner owes erates foreach route, Hence, “uick estat ofthe range conersio, ala aos ean ete, that fe ‘Soe thecconom optimum for he various rons help the chemist oak ta in the ener here he met proftabe Option might be ohana! end to {Geminteepermsns tat ae ute the ge of pofiabe operation Decomposition Procedres for Existing Processes (0 coune, we ea ao as the approach presented above as 3 decomposition Brosedure or essing process, 10 nin the understanding of he procs 19 Sedesiand the deo nde develop the poco ospstematialy develop ‘tot pres aematiner Thedecomposion procedure we supp ss flows 1. Remon ll dhe Ret exchanges, dems am soge sess 2 Group alle distin clumes (quid separation system Mok) 3 Simpy th eral sete of the separation sytem (Siar to Fit. 12) “4 Lamp (rua wis in single box) the complete separation sytem (sin toFn 28) Lamp the complete roc “Tit dasomposion procedie i rent rm thon that breakdown the fowsbec into dscreterubeysems which alvays retain th Hen, He, to {nial unit oprations To develop process alternatives, we wat to moa the ‘Ssysens, With ou approach me acompih hs tak witha famewore where tte alnay conser the foal plas, tough tbe amouat of deal mtoded at ‘ios tvs changes ierarchial Planing ‘Our stay of wesesive reisements and ou hierarchical doen proce ie ‘Sa tote earch panning tatey Scud inthe ara itligence (Aivitertre Sacer" Hates,‘tcashn spo ve By eomring eu ny erie pcg song en ib imports liveries pray eed pono he ear ae In ourieurchy the ground sate represent the energy tegrated Rowse. and cich lee above i coaisever dic Moreover te pov appeats te unroftale as we proceed through the evel in Tale 121, we lok for Proftabiealaroane or we tetas the projet Before we paced o The ert ve At noted by Sacerdot the herarchy provide an ent proach for veloping des, 13. SUMMARY AND EXERCISES Sommary Process design problem reverend nd only about 1% es for Be ‘sign cver become cormeriaized, Hence, a cent say for developing ‘digi nly to conder only rough, sereeaing ype calculations te. e ‘Siminate poor projects anf por poses akerativer wh miu of et st Coot wae {1 ‘ae [Wore [ease flor} [Fn] Pau Cle 17 ANAS Cine bi) Z a Page sec donne ini rer, i ue is SY“Then i es oti penny aay em roi ad eal the Tea teraat ne he gsc compass ener : on a yh eigen y beting dwn nt a each esa APNE ST Tn teen ease erry cscs sar eres NSC Wearing cn 310 eae he werk techy eng we Tae fr mal 12 Aeclo tthe epring mesh wht woul be he st way to rend» sto ace nnn fn DY py eel 134 Rowen rapes sm nore en {PALHy0 te wed ath fod seam The fect takes plc f Yam aad S72. eaitd to mee Reston ston Tener Restor Seaton Senter _f ent Catena Coie { Catwan “secrion 13 sunny ano asncas 21 Jet tush or th proiton yee Pen nF hye + een Ebene ye + Eitenzee = Datyeene ty foster forthe poe 146 ows er ehenet es shown Fi 135 The prinay ection ae Eine ¢ 10st 2 Ethan =D Eber + 1,0 “herein tks place a SK ad 6 ar an bt 7% conversion th Soh ot cmt co pron {strane 10% yee, hae and 3% mene Show the heey ctl eet Bi Heer sence Ce dh Pr 18 80132. tone fr tenet proton i shown ia Fe 134 [hom BSA sce brcse scant rcs al) 1s Ho. i) The pms wee, vei by-pass (eae and ec eb) a wl a ‘aan tena wat on he Rowse the tle reo and yd nme ctum td hererrl prochaine bm eee aoe Te cit cred ‘ream that 9 cpa compre 1.34 Sobet Bows fom Hyco Proce (he November feo ay ‘eat Devo he chy of owe or be pcs cuaprer 2 ENGINEERING ECONOMICS In Chap. 1 we desided a sytmaic approach that an be wl 10 desl erp desi I saiion we istedtetype of design etme that normaly Se undertaken ove th eo pt The gol of hae etna ofc ‘it dt, ahough the accrac ofthe cleation procedures andthe umoua (ea considered te eleret foreach pe of eta ‘Sine cot estimates ae the deing force foray design sud, we esl fo understand heaton toe to nlade We Serbs precede for feeeaing ‘st state fora coneptual desi his chapter We ein Peseting The ‘esl om a published case tds sn oder to gin. an onl ppt: the "ops of cos dt rege snd hen we ct he det the cot ane Remember tat the est models hat we develop soul be wel ol or screening pocessaleratvs, The cosets tht ae reported Wo manage ‘Shou be prepared ty the appropiate economic specs in the company, esate wl nade contingency als based on exerace od wil ele th cousoftmore tens than me conse. Ths oa on eimater nal il te {eo optimist and hey should be hep cont wl they have ee Yene. 21 COST INFORMATION REQUIRED By considering the result opublake cate sty, we can gt an vv the ‘ind anfrmaton hat we sed to deep aos eximatelora conceal deen Moreover, the femework relating the materi an energy flan stipes see and uty ows cpt and operating cx nd proces petit Soak Tome more appuent The partulr ce stay we cone inoWes the roucton of yloenace bythe yaropcation of benzene” Benzene +34, =*Cyelhexane any Eh eeemtsser ee a7 on mJ Pa, ingen Unay Do any Na. aed by Sh, Wain ‘Our purpose ets nt sess the dt the dese, bat merely tose vba ip eae ae generated ‘lowsheot and Stream Table (ne ofthe most portant tems tat we deen dun «design pret owsbec Gee ig 211). The fowsheet bows the major pees eget. and Coit cash pit otequpnet gen aapeial nomberorname arin 211 ‘Novmal etch steam on he Rows aso tered or umees and a ea {Gol tht contain thse ete uber lien apes a he boo of the owsbect-The steam able conta the Bows each eomponce never) sey Jo's tbe scum lerpeatres so reste in vome eases, etal {esac and ster information for ec stream are added inthe eee te, Operating Costs ‘Once we know the stem ow rates the ream temperate, te ca cate eect ows forte varus wi shown onthe Rowse se Table 21-4. Then {fe Lnom he ui cote wien, we can enlelte te ttl ly costs We ‘mbine thee ter cots wth the rowers costs and other Operaig ‘ayes ooain a summary afte opereting cose Table 212 seminary: Base caseca we wa mn vere Wo BINT Dowie “anorisy aaeiAraa ibnaow ‘Snow a avn ae er = Laan 9 Se meq sen ea) nue es BoeCapital Coss [Aor we have determined the seams flows and sam temperatures, we can Micha dhe euipnent sey ace Table 21. Then we cause on corrltons eee eit Sc 2p tocntmate the delwered equipment cont Next te (Setiwataton fort tate the italledequpment cvs ee Table 2-4), ‘We must ao estimate the working epitalreqsired for the He2 3) i i ; ER ‘ noo ag | i eee : i i 3 [8 [naan ce sen an come of be exp gore i |e conan) =| Proitabiy Estimate ‘We combine the operating nd capital eons long with ome ober ont the the esl etiate the prota the proces ee Table 24-7. The ‘Sharm ov invent ated a rerio potty in theese sty, uta mber sober sess Cn Be ed These re coed nS. 24 Enginceing Boonies Pepsin Foe kina POPE Denn Ma 8 Now that we ca se what types of cot are ince in am ecoaomic na how AON erate tee col data? Fir we conse some ofthe meshods For30 seem 21 coer aronsaron enc seem 21 cr mano semen 31 ee sone BOB ie a a penne = Ft ethome mce rm a =n SS Pet = oie i fe =r = os EE any a ‘aan mae re Lk ra wots oo‘aulatng capita and operating cost then we deci the techniques for ping ‘Shia sod operating conte onthe tame ts net we seus prota ‘eee td ly we preset ape mode has ell or stein proms ‘Terns wh me deep a concept! desi 22_ ESTIMATING CAPITAL AND OPERATING COSTS, tn Table 21 she wy Ku for the various pices oeguiment om the ashe tree tract and in Tate 212 the wy cote were colt. Sia. abe 11) the equipment ss forte owt were ise, and the coats mete ‘Siecle in Tole 21 Ths the fs co ne conser are the peraing od ‘Sit cont stot wth the equipment on the Nowsbet. ‘Operating Cons (Operating cove ate sonal spe to einate.Ooce we know the Bows of intense andthe wy ows ek tear coo water, poner) we ply multiply te Row bythe da value oi that stream. companies hat peat the wily yey es east abd power production, s 8 pra ‘npn the tesco factor are simple to oan I this not the as oweveran anaes ofthe tl se ened fo etmate the cos of seu a ovo pete els For eur peur desi, we assure tata vale “Care must be shen hat the wt vale are gen on 2 thermodynamic consistent bao iy el abd seca should be move expensive than high [route seam, wiih sould be more expensive thas low-pressure sem, ‘Rterratons pes do eccr tine that et apeae hat teresa rt {a burning frocks to make elect orim sing etn fo produce seam. Howerersdeigme bused on unosunl market sors normally pay Dea") ‘Sonom peat ater ew ens One way to Keep uty ess uniform 10 ‘teal ety price (etry, various seam levee andcolig-nate os) 10 herent fava, ee Append E "Te cot cemials can be obtszed trom the marketing department na corp. Fo! sade purpous, caren prc or mort chemicals ca be ond Inthe Chonat Markeing Report rman ofthe wade pbleation. Light ers, for example On Ny CO, cto ae oot hited sn the Chmicel Marketing porter use most ane tldinaiyolong tern contracts The caren pres ovale {le publication at fc iflrest from the pie cba fom the marketing ‘department Benue log term cote strane. Capt Coss ‘As we might expe thee ae a waity of ways of estimating the capital cons of, ripe tat rang rom ery Qk clans wilh ited ncaracy to very hated calnitions that se Nery te-conmuing ul more accurate The mos ‘scram 22 estmaring cxrmn a crenariec comrs | 3 accurate eile is smply to btn a gute rom a vendor; ia bestnchanet Teepe tol you a eat xchange hat has a pects performance apie inf vere on acral dat fr» specie pice. Tt pen 6 shop TUB amacenendors quot wil dpendon how much work 00 hard The MeSdore ques ar wed a he coats of al design, os tovcetul designs we neta faster and simpler approich (ewe ho ot snanttotsto spices proces tad om vendor quotes) Ths ne nomial we cre Prenat: For eramp the capa cost oft est exchange Sa tec terms othe hen nchanger tes, and nt Neary (0 aor reheoates, the mur ete the bf pacing, oF a FE EIS fe desi Sin the ot furnace gen ters of he et Tee Slaten te cow ef «dilation clu iapeced eso te ton hh an dnt Te cost conelations are obtained Wy corel TOREARRET vendors quoter aie the appropri qurment i= varible PURCHASED EQUIPMENT COST CORRELATIONS. A gut extensve sof cee Menntsavaable in PetersandTimmerba® Other creations of his ‘Suni tex plished by Chiton, Happel and Jord, and Guthrie! The roa of Peers and Timer ave aon te on reset although 2 tren more roan upste valle ASPEN. Severe conitions sles FET Scpmcat that ee taken om Guthrie can be ound i Appendix 2. “Se ceaee we te most ntertel n estimating the etal peeing costs er eae ee ble to pedi the ied een contrat han he Fee gupmen cone Te socmpinh ths gok mene! to todace se of isaac ctor. INSTALLED BOUIPMENT COSTS. One othe ers aprons er xiang Dold equpment oss fon the purchased equipment costs as proposed by Tagen hate ttl ntalld equipment cous were approxima eave [2Fins i taal purchased cose, although deen actors oul be wed for aco pocsting plants Ha fousd that more acre etinaes ee oned by esi diferat factors for dierent Kinds of proceing Caen carcrample the purchased cots of dtilation columns, pest segs and intruments toad be mle by 4st exchangrs should Tec by 1: compressors ty 2S: ed enters by 2 and mieeaneous opment by 15 The waco Han tori strated ia Table 21-4 WRC de ie sauce ta Viet Por Set Cat ttm Rr, 31095The iratalaton factors ted in changer, a ne sre th Tnstled Cot~ neta Tostacd Cost = Base CouttlndentiF + Fe) 2) ve pve in Ape sind by, PDOS.36 ser 27 tem cA Ano dream come nee Cx = 1999 exchanger cost; Cy ~ base cot or acarbon steel Noni ead ‘tounge with + 10 sgn presre and between 190 and 1200000 ‘ac wen: Fo-—1 degre coro; Fug ~eteiabret-ranrscton cor ‘the ase coat is In cy~ $202 00150610.4 4 0068140n 4) 25 chonty of cter pcr o equipment Sima, the intalaton torre given {Be oom of equations. Updating Cost Correlations ‘anu the Peters, Timmerhaus. and ASPEN correlations are more reat, Howeve.it {esaboat ther ens o bal chemical plant andso we mast Beate oped {tre cous Cll the cost of almost everything ncreses with ine, and 60 we ‘ost beable to update the sst conelatone Seve methods em be ed fo th Purpose but they al sia that hey vole meng the ase cost in ‘rain ear by the ran oa eet index for some her eat othe cone othe ‘One of themes! pope cnt indice of thi tye inked by Marshal ng nit UGS) and ped mony Chee nen Ae ‘Recordin, ie Ren ery nes, the Chem Enginering lan contre ‘hn index an he materia abo cx index. Some of hese ts ce In ich often experience ferent tation pene to wpdate erage fcr can separate ator ahora mate [By tore Guth coeelaons have the advantage ha he mmerial nd ibosactors at ileret rats o ome id. dea to tecunt or eston INHOUSE COST CORRELATIONS, Many companies have devloped their mn {at coeelatons and tlio latory These re requeny updated By wn Sendorsquctaton and rset somsruction cost Tse company cst core ons thou lays be wed ty are avaiable We te Gui crrlations feos they te seb in he published eat 23 ‘TOTAL CAPITAL INVESTMENT AND TOTAL PRODUCT COSTS. “There ae umetous cone reid 10 tld and operate a cenicalpant other than the operting cone ai th installed equpent eet see Tabs 21 and Ze" Some ef hes costs 28d to the capa nvestmert, whereas others are “peratngcapemes Festus, moto these eons ean eae diy tthe ‘raed cqupment cot through te te of was factor, A very conse Sry of tes tos mes repre ty Peter by Times ands moda ‘enn ef thei fr te tal apa nestmen sb Table 234. The ‘SSrespondingbrenkowe for he ttl product costs sven Table 232 Teiscommon preci the Sevlopment of dein ft acl the size ofa the equpoen and to etna the amount of ies requed. Next. Ute equpment tsar termined, and the uty costs ae eaetted. Then the ‘ascot core ave ded. a ally pofabty aa wndetaken ‘weve fr penny goer den we rk o look or paces iematies ‘room ava design appear ob unproiable Therefore, we wen o develop Sioplfed com me for tal invesinent tll procosig cos and proce Praia. We develop asmple model ef hs type as we dss the nid “otal Capital Invesineat [Acoing to Table 21 the total epi investinent Tot. a the ao the eed capt ivetnent 0d Cap) and the working casa (Work. Cap) “Tol. = ited Cap. + Wer cap oxy[Bento of cal capital neste tao cot Train men a see asa eB ig a SSE Foner epee eat ‘Ppa qu he soo se mt nk ahr cs geo te compe ‘iy sar ice Ft ‘Soar SB ako ey sw ‘secrion 23 YORAL canal ist ate YorAL smawueTcomTs 3% ©) Neg hop erp. asm sup mal TERS nance Sc a, Pao eee mise eons Se “ eto her itp a st Peake re Sep nmi, 42% Fa $8 et nt Fa nie au, cas ge nd won eT SSSI Peres ane ono ommt 2, Sanne me tet 20% FE atom cst option ud miter aa ee ta nat es BP lero (Sm pone tres oa hi pal cn 2 eee Sect sooty. ee posh ion (SOREEIRG Sono retort arcoteebn xan so {fear prton and or a on atc ry tec see $e fg eect eu on nae ae 5, Reems ° Sinem emeien ta Sea Sera 9 Se ae are, Se ee Sa ec eee eer oe eee neater ea naint a oe eit ee eran panne renm nee Samhpeesteoenman eran econ sway pak ste pcm eS Te a te pee sorties coger ‘secon © Gy" nos em ater eee ngs, compre tk wate pe et ety eens on sso9 srcmom 24 yore ‘Crt nig tl pact ote 1 Dest mmc nie er abot 625g hod 4 doceraeurreciats Fee eer Aime ceaseecieams ie Eicon ‘erst (a call SARE gig ninaon toa ogg IPRs conn me se or eyeing or sper mc Saareap Cos Many companies so ice the sar oss as prt of the ait investment ‘ther compans corer the action othe stareap cose that ellen 10 ‘quien madiications as pat othe capital nvesineat eres te fds wed {Sredaionl workove and materials need to srt up te pant re considered ‘peating expenses. The choice among these various posibties depends on the tex station of he company. Howe. for oor parses we inca he trp ‘u(t up) ar prt the nsesinet Hep, E231 becomes Tot tv. Fed Cap. + Work Cap. + Startup 232) From Table 241, lem 1V, wee that Star-up ~014Fved Cap) 3) Working Capital ‘The working aia represents the fds ei oat operate the lant ie {opuy forthe o pay sla, te We tempt to eps the working fol exch oath out of prodact revenues. Nees we met Bake mone [tobe befte ne sommes pets t filth tanks amd tome he a payroll For th esto the working capil cided to be art of ths ftal "breakdown fhe working capitan Table? and resonable Aint ena ef ts cont canbe taken at month spy of tw tater frodics We can grey simpy the nial ivsinet nays hore, we ‘Tove that working capa eae to the nest. Forth eaon, We Work. Cap. ~ 015 Ine) ay Fixed Capital Ivete From Table 23-1 we se tha he red expt inesient i tbe mf be ies ‘The dest cows nce the ote cont (Onsite) o ISBL cons Tins a he fst of OSBL cot (ote baer eis) Disc Cos = Onste + Oe as The omit cost conespond to he installed equipment confor the items showe {nie proses fowsoet Al thee ems se bolt a sea pograpea ars, (Sed the baer hms We can estate he nse costs dey om Guhres “The oie costs or OSBL cout efron lant, cooing towers she tems sted Table2 3 that ae pede forte operation fhe proces st fe bull ina sifercat enrol aren. I common pace to hve cena Tor cooling towers cm generation qupment tc We note or the able ‘he savin Inthe nial fie costs much ng tha tat he (tect in fn thei coats may vay fom a las 40 0 07 fhe hate cont fo an espmson ofan extinct pt 200 #97 of tbe nse Chnster he comtracon oa gaseroote pant (brand new iit tating om ‘Stra or major plan eapanon Thi statin fs aalgous to bung 20 ‘Stitt a ose versus bling new home: nour sdis we conser ony {hat expansions aad we asame tat ite ~ 045 Onsite es “The inv cous desi in he able oe re lumped in two ctor: ()chc owners cost, whith cade te enpnceing, operon, and constrain ‘hpest and (2) congenses ai les (Cming) whieh account both er es42 saevoy 29 TOrAL CAREAL noe ate Tora HOUAC coeTs overlooked inthe pctimiary design an unde t pay th cotracir cor Uingetey llonane oft east 87, should be inde even ne have fe quotes ‘on hand from vendors, Beaute something cat alays go wrong, For out ‘reiminay dep, where we consider only the most exenive picts CQ ‘ent we inclade a contingency ator of 20% Ths, we tte th Indicet Cons ~ Owner Costs + Contin. or (Owne?s Cast = 0050s + Ofc) a9) Coming. ~ ¢20nste + Of) e319) With hese approximations we can write Fixed Cap. ~ Onsite + Ofte + Onna’ Cost + Coting = 12510 + Ome) eu [A Simplified Investment Model ‘The factors we ave sete to we in our atalis should give w easonable satiate ofthe investment forthe type f petocemical proce that we ae oasidring, However, diferent asumtins should be mate for dierent pro: esis and the chose of thse factors ran area here eng npn nected ‘Our gual ito develop x simple meth fr pctiany pres dei so tet Facer should be wed where he ae splice ‘When we combire the expression ove find that ‘Tot In = Faed Cost + Work Cap + Starcup Fined Cap + 015¢Ta Inv) +01(Fed Cap) so that ‘To ln. = L3Fined Cap) eu ‘Then, rom Fa 23:1, “Tot fv. = 1500 254Onete + OMe) o ‘Tou tv. = Leas Onsite + Ofte) es) Forth as fa plant expansion, we substitute Eq 23-70 obtain Tot tn. = L425{Onste + 048Ons)] = 2360856) 3:14) ence once we have etimated th sald equpsiet cst imple mate fo eatinate the foal ivestent although Hi ipo to feet tht ssurpions made in Ege 233,234,237,299, 20d cron 3} era cat en Tra HET css 4 “Total Product Cost, “Table 232 lata breakdown of the tot product cow. Sac the ot proc cost (Tot Prod. Cot) the tam ofthe manlactring cons (Man. Cos and the tener exes (or SARE) we can wae “To. Prod. Cost = Manu, Cort 4 SARE any) ‘The SARE cost often ee about 25% of the sles revenue for chemi Inermednes aug they ray be higher ished product sold ie Co SARE ~ 0025(Revenue) e316) “The maulactring cost che su of he dec production eos the the rg a thea overhead (OVHD): Mans. Cot = Direc Prod. Cos + Fined Charges + Plant OVHD (2417) ‘The direct production cout inci the raw materia, heii, mannan and eeein operating spies (Op Suppl) operating labor, diet sepervision laboratory charges, nd patents ad rsa: Dirt Prod Cost ~ Raw Mail + Ut + Maint + Op Sepniy ‘Pato Supervin FLah + Royaly 23.8) We cas etimate the rawr cote and the utes based on out weimiary dsp elslations From the tbl meee a the mater ‘pairs andthe opwating supplies depend on heed apa vesten nd or Main, ~ O04Fined Cap) en) Supply = G1SMaint) = D1S008%Fied Cap) (25.20) “Te cant fr operating labor, diet supervision and laboratory charge ao can ‘combined iro a snl actor We sume that Labor + Supevis + Lab. = (1 +02-+0154Labo = 13%Labor) 2321) “Te ale indicates that the cost or patents a oye shoud be about 3% of {he otal product ae Royalty = 03(Tot. Prd, Cos) eam When we combi thee relationships, we fd shat Dit Prod. Cos = Raw Mat, 4 Ut + OO44Fie8 Cop) “)Tasitsban + 008TOL Pro’ Cos) @323) “The fined charges (Fined Chg) given in Table 232 inlae local tae, exwBased ‘The intrt chart om Horromed capital depend on the company’s fancing [cyan fr ou rcliniary dene we sume ta intra fr re wed Lo Imre 2326) Rent =0 ean “Teall for deprsition® ay be cleultd in warty of ways and so we Sper ch ny ose rn eb ob och Su, Fo 2421 id hat ry anf 88) “assis ‘Comparing Capital and Operating Coss we want to compare an investment 1, plus annual payments to another faveaunest I, with ierentanbul punts Ry, tose hich she salt, we Sara Spon Was at nopeet neces rary tue cam et dns omen Pyare Rae (2404) Sere conte cae The pans expres br wr pape ei fn be deed in smile way Example 244 Suppose tht we descr 15D mis that we ep ce or 9 ‘lot rn aoa and ah nd hn ne uy oor as eye ‘ithacomentonal ee adel enpee (asin py te tal prc pce Soin, Recording 0 Bg. 2428 csemaqnyasy) er SME (483) 0 + 485 = S805 enc. sete bay the ometon agin stinating Capital Costs [Now we can ue presente calculation o compare aierativs tha hae iret espa and operating css. We can use these ress to aes the ‘roftabil of proces. 25. MEASURES OF PROCESS PROFITABILITY From the dicanan See 2 eal the cot accounting aod wih proces ‘soomicscan income quate complicated Weantpatethe sane ind of sty invatempts to abs the proces prostate OF cour, we per to ae the Sipe pone procedurs for ur preiminary design cacao but we mond Iiethe eats oh as meaning 8 poste Thus, bore we elect procedure fer estialing probaly, we oa to ender te relave advantages and tsadvantages ofthe various eho, Reture on tanestne 1 Tate 21-7 te petty meabre called ws the eur on investment (OD, Wecaleaethsvalueby diving the anna prot bythe oa ivestment 40 muting by 10: eo = Aameal Prt 9g ast) ‘We can bate his etun cn investment on te th prof before ae rhe rai ‘fe tae Soe man cart to seport he aa forthe elation As, is Import to remember ta the woking ast wel ar the orton atest ‘pcs coder asa vest fr tax purposes, shoud be nd the {otal verte ‘The return on jvestcnt a very single esse of te proBtbity, bt cs nar consider the tine value af honey. Moreover. stmt be hse on soe Kind afen verge yeu' operation Scr variable doprecationllances (ches the ACKS method) incensng Weinerance ess eer the poet ie, changin {als valumen ety canna Be soounted for enspt by srerapig. Des et Sborumigs he return on inveniment fen ix oad for presinary Sgn ‘allan Payout Tine ‘ic the imei year takes to coer the fends that we ive Calter fe Fayout period we ae playing poker with someone ees mone), whch i ‘cia stanton, Werecove the weking apa ey monty an therfore we ‘ele the working caitlin the calculation. However, the ection theta op st that coe to be an nvestment boul be adel to the Sed cape Ivete tof theaoun of money ed up the projet. The fds hat we recover tom the projet re the pat afer ae pls te depredation lace ‘hich we cll the ath low 0 that the payout ie Fixed Cap. + Suase-sp Payout tie (x = es16 secon as wena enc moray “This ction abo i vey spew cate, bt i sues fom the sme siatatnge the ROL Hence, we weal ke o contrast these dnp prosedaes et oun enya tat maxon or te ine vlue of mney. nth Teer auls ter undeetanding fhe addon complet required fo obtla Disco ‘Si acter way to jole the dealt of ovsing in 2 new proces Stee maximum snwunt of itte shat we could ford to pay if we Fa, fe evetnen ad he projet would us breakeven Obriowly Dor ante neater ht we could flr to py 120% terest, we kro Seu tar eter tonnes in ts projet athe than na bak Hower, ihe [ern we cold afore to puy was oly 27 we sould abandon he prox When ‘svSStuder mest caution, we rue bat nee often compound a Hence we tonto crate the revenues cost derecnon, anes pai and the lnvestinet om 8 yar bse ba ‘Normally ke sat 9 to bud plan anor thi reason me waar ovement cons tam material an pros pis ies cous et, to eet the erst Dyet the fre rater than athe este’ tine. Rocce, te {Eiulation of be proves potty sould e based onthe neo al costs Seer the dro hos ben made to tart costacton Tn ctber word 0 ALLOCATION OF CAPITAL INVESTMENT. Sic requires about 3116 i ompes proce lan th et ons lB pent ove ths tt] Feriog AL ‘cou we wiv to ay forthe lana ie contr and constetion cc, Ie th oument reper the heal tr preperog the foundations or the ‘upment Ther ne sat isang the equenent as He devered. Thus, the ‘Tree con expends a tae mine abot 10 o 15% ofthe feta. Ding the pers ofboth 2 and 1 ym llen spend 40 or 3875 06 he et cos ‘Gat year and nthe lst yar we normaly spend the remaining 10 t0 15%, However te ovwers costs which are for enptecrng and superesoe, and the oningncies and feraty be spent norm troupout the constuction eid "Toe working capt aed sary cows are invested at ime 270 but erent the working capitis raxoer a the end ofthe projec. Stay ‘MELE else aftneegipment amt eve at he eof te prj. andthe ‘hen aunts to 102 afte phased egupment cost or roughly 39% the Raed (ifalnventment. Ot cover, money tetared fer Ny asa sae ve at Time rio. becuse we oul deposits salt sum in 2 Dank at time 0 and Ieecive he compound intrest on ths fade fo 8 ye, Ags, we ee that the ie ‘Sine llmoney requests to ncsount fo fund a tems af Their presen wae, hich isn the principal, eguiedto acomlate an amount of money S fe ‘Woe The press ele of saris invermeat pies ean be estimated once te {nist rte hs ne specie by wing the lationship ne veloped eae ALLOCATION OF REVENUES AND COSTS. Most aw plats do not rach that al produce eapacty inthe St fen yeas operations often beease a Inaret dace na en fv the product Experience dest tht he production ‘Bice om about 0900 95% during he et coe ab in a, ‘pectiey, of operation, After tat ing, opty te proces operates a a Slay he depreciation allowance wil var each yea, ules tah tie dopeation sede i we Thus, with varie eee anda variable {epecitonalowan he anal profi te income tse andthe et co i ‘Sng tom year to yar We cal the su ofthe snc tet ptt, ich the Froft fe tare put the deprecation allowance, te cask fin, brewne Thi ‘oun! money me actly ue bythe company ech year” OF our the ‘hun flow athe en af the is you, a later years, st be discounted othe present ye again bcos we oul eal he sume mova of money at et {Bey investiga smaer amount of money a tie 0 DISCOUNTED-CASH.FLOW ANALYSIS. Wit the background gen above we {ans up the produ for eles the cout eat ow rat of retry {DCFHOR) Teac yee the re tab he nent (Pine peter aloo the eh ows we cider say general eave where | The alain of the dist conte cam be repreente by pecentaps, such as tre Olsay = G4e1~04 and q = O1 2 Three ct fre tye eb = by = 09 | The ont pdt costs (ren operating expense) ae constant. 44. We eae tgs depresatin so that dy = dy = dy =~ dy '. Weenchde the depredation alowanc rm the total product ex ten we can develop an expression forthe euaity of he present ale ofthe ‘perms aad the income "The doe cot te weer coms, and the contigs ae seat ove the ‘onstruction pti, bt the wovtiog ep andthe sary cos ar Teqted fy at strap Tho the oa ale st ap tie 2 Mfa,Divect Cont + Owners Cont + Comting KI +0 4 Work.Cop.+ Stun} 25)“The present alae fal he ca lows dacounted bck tthe tat -p tne pls Tocoumed ale ofthe working capa andthe salvage vue he eb of te ssc en me a ne pec ‘aprevsons equal to each oer. Unfortunatly there iso simple may to sum the ‘Ses involved that we mst es ta-and-ror procedure oad the eres APPARENTLY UNPROFTTABLE PROCESSES. OF couse, hee is sen in ‘tempting ole the problem by tal and ror the tal as ow ver the i ‘tbe prot plas the Salvage vale ot sda to pay Tr the Seed capi invest! pla the star-up ce In other word, fr te poate we wae tat (No. Year) = Revenue > “Tot Prod Cot (No. Yeats) es In practi, we expect to ccounte thi iitation gute eget many eas for row proceues simply ave not peolabie and the ees of inflaton wl make it Shpen hat we cam never bul plant smlarto one tha lend eis even ‘he market expand. "However, ve do nol want ochminate pos that may become sound investments wien product pes rie because of sappand demand conse lon Thus eq 253 trnitsathed, we fen fqen/0 15 02 in Egy 253 Sn ort fre cont ot cong, ‘Xoking sap carp ny, ol proc cinta eg ac and then we Teeter he seenn we woul ee fo obtain. Nest ema he prod pice {hat covesponds thee evened underke&sipland demand anys {Sacre wrth tee igh xp bn tat pce hie ‘roecon 20 yw might wel put he poet nthe es fr 1 oF; but {etme projection or yr, we mgt continue to work onthe design Again {feent reured 10 make this ection SIMPLIFIED MODEL. As we might expect, te ants comes much sper {te iaveimenty cash ows, and deprecton allowance uf With ‘Svataa tah ows we can cour trs and sony formulas t su the ves S70 8 Means oF mass Pearramie $9 eal! Work Cap. St-up =e Pat Cw see!) 4 (Wk. Cap + Sas: Val X14 0°" tC 11 rok cap + as} = f0S4Revene ~ Tot, Pro. Con) + O48Depree IC + + Work Cap. + Sa Va Of oars me il ed oa 0 ese) ‘aang 264, Cale theDCFROR othe aration iene! and eteCapital Charge Factors "We pete to voi id errr cleans primary process designs, and Jobe would Ite to acount fr the tne val of money nse way i oar {rosy anal Yo ascomplish hs pal we dene capital charge factor evemte Tot. Po Cost = CFT tas) asp sete the oa nvsinent ines the working capital an the star yp coi, RoC‘ccan ase ts sein to tite Rovemle To. Prod. Cont which Inve hourly com own Eq 25.6 a0 thal we obtain ah expen conan [EMO fe carne mon can ssa ot rusiecnroing + camnepecltt +771 "orca. Sh a ass trv soe CCF we obrin cre {EGE} vat eco ca nape a ok Can av Yad esata ind OS In our dcusion a depreciation, we assume that Sale, Val = 003Fxed Cap) @s10) Aso, wean elit the oer quanti sprain in Eq 259 fied api sng the expresione we developed eae Startup ~01(Fxed Cop) ayy ot n= 1MFaed Cap) ea) Work Cap = 0154 ns) a = 015(134Fxed Cap) = 0198FxeH Cap) 25:11) Depree = 0.4Fxe8 Cap) e336) these apeosioatons, the investment terms cacao Ea. 253, and we fd th cer = 0290-4" + 0298 — 029891 + oF 0225+ 0048 aoa + at oor meron 29 sersunes oF rsectas raorneamusry 6 Hence, based om hee simpliyng asain, we see that here isa diet relaionship tetween CCF and the dneoumedcstfow rate of reara Cf Wee FEO tS and The Bod hat CCF 0351; whereas f we er = 03 and Ith we aia CCF 0.06 ly ut pelininay dig eins, fr the sake of spicy, me et [value f= 0.18 she sail value we woul eer comer for a new projet [A Simplied Profiabilty Model Using the defition ofthe CCF Revenue — Tot Prod, Cost = CCFCTOL Ix) 57) ine can subwite Bg. 23-16 for fal favesent and F. 23-3 for toa ‘rodicton cot to ohn ‘CCF(2.% Onsite) = Revenue [L031Raw Mall. + Usp] + 0486 Onst) 4} 002S6(Revens) +213 « 10*Operton, 548) __4rxevenes) — 0436(Ram Mtl Ul) OTEK) +0962» 10" Open Ons * cer mom ss) ‘Hea, fom estimates of the rerensc the rw atrial, the wits, and the {tsa equipment cons (Ons) we cam cleat the CF eorespondig 0 2 we wl aban ny, ienpesne absorbers but wey expensive ‘used on these simple argament wef hat we nan Lo hoot L such rebar os Ofcourse, Lim) ~ 1 iright io be mide of ti ane. However, ve nse Meshape ofthe curves net Lm) = and with gh recoveries we sx that we Iti sain «beter ado of betwen decresng number of aes eed the absorber (cpl cs) and she needing capital and operating costs of he Atmilason column by decreasing L Hote, 8 « Rist es, seems to be emonabe fo chow 1 sich that au 043) ‘ich the common ale of thar. Fractional Recovery in Gas Abmobers For a Sie scles Bow rae, we can always increas the foovey ofthe slate Simply by adiieg tape the ge abo, Hone, there ean ecooomi trade-off Stee a acesing absorber cosas wea Was vests a dectesing value of thesoltc lost Onc thea capital cos (he abvorber and one a Operas ‘the ote os) senator tm 7 ‘COST MODEL. Ii common practise 10 sept operating ins on anu basi Ths. texans te conn trade et alsa pu he cata cos o ‘manned basi A dined ia Set 25, we anal the cpl Cost by ivodcig » capil share ttr (CCT) of tyr where he CCF icles al ‘pial cated expenses (paciation epi abd taesanen, ec) A CCF a {i comesponde to about a 159, ducted eth ow ttc turn (OCPROR), satq 35h Suppose we write ata annual ot (TAC) models TAC(S(90) = (C,S/melNG a 01 NS150 B90) 16S TKN pate) ony ‘OPTIMUM DESIGN. Now ine we ou simpli design ation, Ey 33.12 we Tac = s1s0c,cy.("") + cy (soe -) G44) “The opinuey ection! loo ven by BAC 6 sisacoy, - AC : Fria) 9 MOG nae 6 ose) Uwe consider some typical alice G=S15simat Gy Wmolme | Cy~sesnphey (47) 30) RISTO) ~ 2% meg ich coreponds 10 Fractional recovery = 996% eas) AEE ROR ea omelet tne ‘ohm msn er co putea Cel eeeaa oaSensis “Te inion etre of hs elementary aa isnot the relationship forthe TUGAED eg ta 346 whch eo ect, Bt rater te senitity of he ciara Ea: a From 24-8 wea that we can change ay ofthe tubes eter or the denominator 8) 100% and he cpm actions "it le scsi anager demenstate tat three noincentve foe eting the cs at ue inthe alas. Tis sae behaviors chara ee itue Sorter et design probimsiethe solutions ar fen ey apestie ‘Repti property dat the factions for of the design equation, and desis ae ea sch os heat vane oetients ost data, Ths good enpeerink eerste at we oat some exo terest fe ton bore eee geifcant amount of te pbering acute data or aterptng Tee Vadgncatcutovons That we want to spend as Hie Ge as pense Petnetntmmecr and we moot tha anes o have en enough accuracy tomaks Teton we ae aed th imitations of Roles of Thumb “The rales of thamb we developed see ot based onthe Kremaer equation tnd TREE Tat th operating snd the exis nes are ath For a mem Teepe hee eoottons the misma solvent flow ate evcespons to the sae he ator end ofthe ter when the exiqun composition For & See lant seco) einen withthe entering 8 Go that on ‘etch same reeyseqied she concrtated caf the cols): Fi Se The econmu ade ft dicate tat we want o operate with = 140 A, tnctonal recovery, From Figs 3#3¢ we see that thee resas Bhetpeat soot pinch soue atthe tp (Bue end ofthe aborber. Petecmantrted mite of slate the equ ne might be cue. sows nig 13 Fora ined asonn recover, themininum Bg Bow ates “secrminaty the operating ne econ ans to the ego curve. AS the Na drew ape tht fw egie nove ays tH OM sa prtarcne hens conespndiag change forthe tects, Fig 34-2hand (eke emt enpet tat asog «raven fw sch that [mG somewhat (Genet tnd tempting to sores some es sete than forthe de ‘sew ets coe othe optimum desi cooiitons weer donate eens encounered for adiabatic abseiers. A ‘nsec the stream and ake up bythe slent the te Bes oP SENSE Spvtmin: thisheat tears the temperature the Bgub sea (eer ai approaches the btm end of the tower, Increasing guid Tener nccne the vapor pesrure of the sole, nd om Ba 32-4 wes Ulnar ope fh eu Ue ners Ths the minim Has ow ate O35 Cee « » © Minima tigi Bow when ei With cxcrs when there is pinch Fetes) "rom Fig 344 we sce that «sry sleet in the sent fw rate ove th miners wl alow oto pe om Je 0 a with very ew plates Fhowever the apwar-corving stereo! the equa Une means thatthe ‘Siu sfven ow tate beach large hana corespoding exe where ne {lull mains mother operons Ie seit gud temperate tc, “The init gid lw sate fran adiabatic absorber ony be lO tes pester ha he fule-ofharb ale 1 Yay aed the et bid temperate on “This example iste that the indscrisnate use of rls of thumb my ea 0 An nope esi Te geet, Every ele of huni ame ination oan rom our dtatsion ofthe dxign ofan abate ascbe, we note that me caption af pt al abrir andi oe ae Fata which oot esac station Thu tx common pact to put ‘Saling cis or one or more pump around eoling ops onthe Bottom to of tre tae fo ps ear, to fore tt Behave more as an ote fer he botiom ofthe ter hen Fie = Man (88‘Asin tise to uncetand th inerstion teen prost witsim ode to Aevelop «close 10 an optimum design Siar, sora changes the Aowstee (cag cals in he bottom ays an art!) normal ave ach [ener inact ct proves cronies than crcl oplacaton elena he ‘ptm solvent flow sate foam acaba absorber) We can therefore popose ‘Avoid the ws fadabatic absorber (oles hee oy ‘Sal tepeatne re srs he store). oan 35 SUMMARY, EXERCISES, AND NOMENCLATURE. Summary Aunt of pra ose ae preened cape 1 canbe generate even fr simple proces 1 Wen shortest procedures to slet the bet seraative at we wil des ‘gore, providing that the process proftable (We want set site ine as posi geting an awe (@) Weonlywaattoiecladesficen secray tobe bleto make dedi {@) We abeaye comer the sensty of our elelatione 2 shortcut design proceares fice remonae obs proces lw on 100% ecoveris in separators ano sc egupment digas on 9 5% even a the sreing sae of ses ose Order magitude argument canbe we to simplify design equations 3. Sytem approseh You soul lways consider the oa probe. 1 Changes in he design varias none nit arb) might afet the design 4 Rule of han _beuntor {Ta ruw nner fused as «sven ps abort, consider feiig the ros through the absorber Tei double to rcover more than 997 of valuable components. 1 Cooling waters avaiable st 50°F rom cling owe nd must be rete tothe tower at 12°F or tx. ef MNEs 1 spposh emptor seam cock with caing wt. Eeraet (© Hlbecontenaton rca opera TPF wha pc cl eee to (2) Discuss your ret (6) Dae comic ade-ofinveinthe eg fa confession pet (toh ow tenperte sa gh psu ve are Wo ler © spd scutes ee ohh Scere tc en Fr pp sy sajom [O88 + os a agp ine = opioum pe dase ih Oma fw te (eh p= nay oth n= vocnty Ph K=subsshwh, = 08% My <0 ‘pak, ~ oath X= Bali Many dupont ve (chan arte wos na pes corn ough hy ear aes Swiss por Can oe Ey 3 tdtes rata bomb orp mt ‘hema usr of ayn dale cen ray eae aking {Caparo etn restr ys we ae ss 37% pon) Serle br he bon tr ing aco ZEEE RA tao ttf hi gs a2 cross my ERR MAE ‘SETEE utpen naten pte vemuing to) Reta Mae” we asy Nod ant Lin ronson fe of epeation (CES fined ae crs pyar 10 (p25) ose fn uta Moo a Mota pn een pion tenn “ep epntn rcn te ogand Mocs 3 +86, Comer the ag at ese lene dian atm san = 28) « Fey san scons hee eam 197 at out he fed ara to snes ius cons ao kate he ea on be eats Ste etal et gr ster tap eed ah 3 Agha IT ange mame eh ean ot aes (loners ine F ae wedtr ecto codon nd eal hat SST med fr he ere To ene se nt Be TEicTeinty ogee samaleont ttre D.rtoa 1 Po nd 1 Sy Pon Seas Pee a Fagen iN PATI Moo. ch fn Sp Sr Ne 4.5 secruom 93 summany, enencies ano wowewcrarone 93 LI* we ote ste ane con ft hp te ning mates miu he ae TAC Ay Cate + Cal 8 ass) ‘Wiese ton forth ea bles te nhange a show tha E338 ac cs'S (BEM ait oe Sper ton 6 !Sten—79 ee ste 720 nd mtb pen 0 gh fe in by ‘Seong tat -HN}¢F, "1 Wet hie pponmain show ta che ‘ramen aac fF, aes by sen hah i on ity sae nf naw iar un be pre he Ty Hen wena wee at Wester TT ‘hohe temo seme em oa 337) en oo GM ee re ata i a, Toe ats Calne the ovmds or 4 ce ate Cy S138 F310, Tome F Uo ebb Fh t= 20 Bue A, Ca Shipnationn ye. iy 97 Wath T= DHT Ty MT, aC Suzan ve piel ‘aly eu al nd ener soon Tat fran ae) pt es “ 189)sn hen wee th rel cae eng arp ANC) an a oD gee ovat). ei afvantgrou to ao ander clea wen we ae aang 0 ‘Sauce poion Scud Chup. 7), Malo” agra posite oe {or 8 on the righthand ie of Ey 3.511 amd then ses the vemaing ead Seti ica pa th ne hea ‘Conse aca wet ay = t= Her = Rr O25. ye = 2p: 95% ol the Bae record ocak, adhe and 993% oe ae 89. Inthe opropanati-actone proces single reverse rection) shown in Fit er ese race et ons tee + as) rl an ee yd ot rac? stn) osm) here he en 78 coreonds tthe eye Rw, ME creponde othe of i eye Bow onthe at othe aumento 11 ii eee Prem ee a {Se optimum comerson for fod, sneveie cos? For what pe of ‘Nomeneatere A, Ata of water cooker) Re ota exchanger) C_—Anmuaited con of heat echanger (840-6) Gp Amul os pe plate (Sipe) Cl HeatSapucy [arene “F) oF ost of stm 8/9 ye) Satie value (mel) {Cost cooing wae (860-92) Dutt tow rim ie damier Flow ete oi aE 35:2) Gavvae nts) ‘Scher ate (nl) ‘Slope feline (Fg 324) Nutr of plates Operating presre; mus have mame unt a atm, pi mm) ‘Vapor pesure of sates mot ba he ue uns yt, pi ‘amy Feed aualiy et tad (t/t) Voluneti ow ste (8) Mini eB ratio Temperature of emda tests P) (Overall Reae-aransr cect or the water colat [tee a FI) (perl est trans cox! for he tum exchanger [ihr tr) Contig water ow ate yh) ‘Stem fw ene Ce) Mole ration i ge pe Mol faction of component Mole faction of compost Mole action in ga pase isa Root of Underwood's equation Density = Visco CP) et of aporzaton of team (I) Diteence in vapor ate ConversionPART DEVELOPING A CONCEPTUAL DESIGN AND FINDING THE BEST FLOWSHEETCHAPTER INPUT INFORMATION AND BATCH VERSUS, CONTINUOUS ‘ued No only must we develop Goss used on new reacon shes inveted ‘oxo cman, et so we mi desig al ur compere eco AL INPUT INFORMATION ‘The iafrmaton thats oemaly (cotta mast be gathered) gen the int stages oa dein role ‘The reston formation we nce Yo know lied in Table 41-2 Ofea te posse ta ather much ofthis idormation from the pecet erature. In ‘tthe mig ate{hat mit ake pce Even don) sac mount ol by produc pede as Iaborry experimen, ths by-pofoct may tld ope very lage kvl tne recy op. Hence il the by profcs reduce sunt be Known i order 4 tae, and tnt vay eds toying ngs econo patie MAXIMUM YIELD. Isfomation conceming ow the peodict detbtion anges with conversion ano eaorempertre mela too acta ‘Bfendifiade to obtain, A chemats foc ica wowing diferent ete ‘vempting Wo define econ. looking for may 10 wt 9 oad patent {nim Dunne thee scouting egeton the chemist normaly atempi fd {he exiting data bse wil hve ato the points grouped in = sal cane Sooner clon fo the main yl (OMtan the lags Amun of at wl sendin) "Numerous preceret have been designed to operate at the condition of snaia ved. tis operation fen Gost eestesp To the opr Economic coveson. Asan example, tus conde a simple Hype! ‘eacion sytem As B+ Crubere Behe dese est ar € ha oly oe ‘aloe The concerrations of A Band C verses inet bic eater or ace Tine tabu ecto are shou ig 41-1. When B ake oo temas ‘onentation, pot Pw Fig considerable amount of tndeved termed anda age amount 4iecoesed, Howey we cose pean at enn Qn the dag, sal mount fC foes and ath ee omeret Thus if te peat a poet we wll Iieter reeyee ous se he yeh nh ha yf ae single pected condition da wet check he epi Mote of 8 Produced Motor ofA Converted sd wereerto the conversion ofA € ae atv oA agama htch ‘four definition ost or the HDA procs eS 13). soe in Fp 402 (Note tata wae of eins of mlcty and ld ae sed i he Theatre, so that ralnayerecenay to cock te dhation) Normal raw-material con and eect loses re he dia ate nthe dep ofspetrochernl pret Raw mitra cst ae usualy the ay SAG We aD Bs 90 85 a0 4s 50 wo GF TO TS RD RS OOS KD Calan; © CBIC06: -* CotCa0(o) enor (=m 7 areas ate cme Se etic tonto Jy ae Diet yi- oe Diteat (an sat tange fom 35 10 85% of the total product cmt The optimum economic Conterson i normally Sed by an exnomi trade-off betwen age sety teses and large eactr cost thigh conversions Sauna spi ge eye ‘costs a ow cemersons. Heo the opin econamic conversion han the ‘onvenioncoeesponding Io the maximum ye, shown Ia Fig 413 ‘Otten the seoutng experiments pefoed by chemist wil contain more faformation about the fet of conversion onthe pret dstibution tha the +, 0m 009 2 noo 00.000 20600 oi 02 03 04 03 06 07 08 09 bulk othe data which ae akon a he eighorhoad ofthe mani 8. Agi, ‘ing coe scons with s chemin and rcaing oe her th (ceact about the opium proseig sondns ayia the experiment ‘poztam wil ead vo mre proftate press ‘Unirtanatly mon compensa not orgies fo operate i this mance Ine, he chem appraie hor been completly mana the chest ts en ase 0 another projet before apn egies receive te probe nth stuation the designer should alert o estimate he exonomic mettre or ‘termine the economic ptimur comers, and Heteore Jor doi ome ‘Skonal exprinert ate han st desiring s roses Yo oprate at ma mom pl Catalyst Desetvation Another ce f design dats that lens acking the aly stags of dein The catlet deacon ate The chemist fort re ced on nding 4 more ‘ie or wetv ela at rumerods shocsime rane wih 2 wane ‘Stulitrarecorsiderod Some catats have an operating Me of lof? tele ‘eycnoraon or reptement equtd, 10 obo 4 time coesuning exe Iie equ Bind the dectiation rate Tnlal designe we eet tat hee may he large uncertainties in some of ‘he design dat Thu we eae the senso he ttl produc cat 0 ee Uncrate and we us terest to lp ie the experentl Seon! progr i the dvecton of he highest pote proftay, We ase shoves {ecu for hese intl desig alltions,bcase we recopiz that it wil be eery to repeat the aclstions as more nfrmatonWecotes salableProduction Rat Iw ato dese an plant to mest a eapesing met conto, a it {oes the prodction rate we come th bgt an tat has ever oe ‘Wis hisappreach we oan he genet comy sce (Nora heap pa tn if we uy thom ingen The matin size of aa aly sed by he mata seo on or sore pis af exuipment. Cen thie maximum se ed by relates oe ‘pong he cupoentto the plant te Ta onl aera sec conneey oe ‘We ale conde the pssbityoexceting he maim ize an xising bans Th approach amon stays rege the deeopment a eu eet adthus hashish sa Howeve by mings pe ecoery ef re aap ‘pose oredace he produc petal ana eterna at eg here ut team ak However apogee to btn ee types ef manage prblos might ado srt erence cont ‘recut ne mus abo cries how or tae he mart mg tangef pgella ew lan oaks reat deal ef iterence wither our epee 50% the markt al some of henge esting plat eres chee we hee {Zehr door cig pa oso eee ah cap oe “The podacton rte geil forthe plant igh hang (wally dows) dings dean, Market conditions ae somtanty chasing soe ace "pone o these chang By sng shortcut etbor fer our Peinaey ‘ns He minnie tee reqred to change all the calla Prodeet Parity ‘Th prod party rma ilo fed by marketing omens. fi ‘mish be poset proce aang prodact pari at ilecer oe Ma ‘Remus eet that the product re ver purty peicon ghtehonne eee dsm s eing developed. 1s ao erent to Seen inane eee department about the very high cox that maybe tte wih producog ee iehpcy products ely the development aa new proses, to Oot he Imarkeing eparment doesnt rte ator expecatios fo wicalatic le Raw Materials ‘A chemist normally wes very pure chemical egentsin bortory studs, whereas ‘aturlopuichased raw material always conta seme impuries, Hts we ‘elo gue cme inlormaton fm our marketing gro rbot Taste ‘ce var pty inorder to dese mth to nade arfeation fay aut ofthe design project. Moreover ve mnt work mite chemit to saat ‘he impure inthe raw mates ar lator Wl at thereon Wes secron 4s rervy meommarion 108 sho examin te ele on the separation system, optical, tae amounts of ‘pari can bud up fo large ales nrxyle loop uns he impure ne ‘removed tom te proese, Constr For the ake of sty ne normally at to aod proceting conditions tht are ‘ihn theeslsvelinicofs mintre (There an writen nw fateh ‘ark wllanays el ancaplosive sae snd coun a igution Thre have heey {disturbing le wm o nous pat explosions troapho he hry of ‘he chemi indus, and we want to avid anther creer, Honcess, ‘hale anhydde pln, which onde 3lne wih a, So operate wi hc ‘olive rane although there are many spc my fate ‘Simi we hve to know the roScsing cnitions where Some of oor ‘uate pymerize and foal heneachane surf ct wher thes ight tecome unital od rapely decompose: Numerous atria form coke td nevis soonenee vatuarion x Gy bt i ° — a JH sar wit 4 a oeTan gc o ee (Prom ei Chl Resin, te, My Ye 2 67, RECYCLE FCONOMIC EVALUATION Ou eononi nas te apogee comrl oly the stem cose, prod ps by roduc nau atl cos The wea nt IHDA poco ned he a sc option sober senvenon vats (ne ade mo diag by pen) and whe fhe pps ‘Semin of taoge as a0 (oe bt puted ead ncved pe Tone men ote way B22 Se that se innate rp fow a ane wen he Gnveioieqalvotera snd Bee stows the ine moc ow of eed when he yh pee ‘Sooo oun toro Thus netbeans oko ‘i Taate comnrr sn bh cpl al foot ome eee Pott there pe id bth a pine oven tl on op hut commation ie 71 th ea ete HDR pee {ota star heenuaand cpl and operating cote coapresn iene ws reqoe ana'we not cash ene ean “The ral rte op swe in ig 7 ar ot he eo pina yas bei weanande a parton or ting ad inc en we ape that etic pie nom potent nae oo bethied to fwer eye Bors However, we can tn utp sai 1,500 00.00 a ue Conversion as etwork at sili 68. SUMMARY, EXERCISES, AND. NOMENCLATURE Semmary ‘The decisions ba ed to be made oi the yoy struct ofthe weet are 1. How many sacors are requir? Should some component te separted tween the recor? 2 How many reece sucams ate require? 3 Dowe watt tous a exces of ne reacian at he reactor ink? 4 Isa prsrejte compresnorreguied? How dow i afl the os? 'S Shout th recor be operated aabaticaly, wth det heating coking. 6 Do we wast to shi the equiibeum conversion? How? 7, How do the reactor cnt alfet the ceonomie pena?“Tedesin guidelines we se 1 area lone ak some of thee dso for fs desis 1. rections take place at dfeet temperatures abd press anor they sear dere her pate ce te ree 2. Component yee othe sme sector that ave neighboring being pit Should be eye inthe ame ra A partecilecomprenor rege the recycled components bi at « tener ter tha that of prortene 5 tke eco erpertre pes abr moar ato are caged tos the quiriom conversion. tere us bea opium vale of hse vere {6 Forendothrmi process witha eat ead of ess than 6198 «10° Bf we {av an other renter ih ret beeing, For ler eat eds we en or besten 7, Foreaohermi reactions wee a adiabatic ear the adiabatic temper fare ie fe than 10 Yo 15% ofthe inet tenperatre I the ade Ampere te execs tha lt, ee rect Soong the easter best load es than 610 8» 10° Bae Otherwise, we todace a dient or = heat ser 1 Fer singe easton we choose a corversion of 96 0 0 ft equim 9 The most epee teaciant (othe beast reba sly the Hite 10.1 the euiriam constant of «reese by-product smal eee the reveal byproduct 1, Several deg sideline for ect ae given in Table 6641 12. Themyce ow ofthe intng ace ix iven by F= Fl 3 abe isthe amovat of the ting fase pede for he rection and = the 18, The rele ow of eter corponcats ea be determined ty spying the ‘Bela tsa the etre Exercises (1. Delo the rays wre for he HDA proce with diphenyl reed Gee ESSE) He con ea et x nyse oo Dinh wee seee fai nbd we (ne Ea 1-2 Estithe aodie restr o mpe pl! he esnome pest ers he See re e300 lk ae 90 Bo secon ce seman emcees ano soseverarine U6 {64 Deo the acta ote pu pips ih pots om opera (secke )Amune at Aify = 25H0 Bima ad to he (ioc en bean st wax ectang th Bue fF) The feat ensed fore eeten mopped few # Dover reset Doe {4007 Phe om pel wen tei ep vals Oo Ro alse Bf 2 lg S18 Baa Ioraard reaction rte Soasaa sven by spate) Kg = (1499 1a (1, H9FTE RN rete oer ot eee enn ieee tate cae ere (685, Develop he exc sree for he eyobeane proce ace 47), Ame ee mat aa te Att eatin Toman roth so eS cee ao sos Engi eenrrtei r ath a a Envios Paar ct laa Sree cee ee Seeger amt Maat mee nenes on EE er etemr amare te est setae tae iarben” Wag) somerren Nomenclatare A.BLR.S, Reactive and product components se One a rection for component | * Heatesehange are ip Brat horsepower c Conetetion of componentHw capac aa) Fencton of composition Fopsly of compensa ‘sora ae Feed tein enc of econ regents th) Fra ed i of lu ol) Makeup gt fy (neu How te foment) Recreate flan mol) Horsepower Epltbcam como Noewor went Molureais Mal nd Swi nds tation Production re component ol) Tet an ute presre or compen Prose at sap of er conpanr ‘Tol penetra Yolen on enn) Recor hen oud Gh Grrcomime Rel oo (oo Int a ote empresas compresoe eve) vera et ancien [Bue Conversion en fend ml ano Mat acon sf ompoat Prge compo Me of eacton a acto ompestare ad pes (Bua) Fara cote of component Gil CHAPTER 7 SEPARATION SYSTEM Hee we comer onl the synthesis of separation pte to recover gaseous td Iau components Our dcisoc broken down io thee separate par: (rota aru vapor ecovery sytem, a Ii epuraton term Alo keep fa ming tht we ued wo determi the be weparaton nate fron of he ‘Gaignvanabln te the ange othe dea variable io Fig 67-1 where we obtain Profable operation Th, our previo exoaomic sts tlp Yo simply the 71. GENERAL STRUCTURE OF THE SEPARATION SYSTEM, “To devemin the general stractre of he separation stem, wes determine the ac ofthe reactor effent rum (ee Fig. 1) For napa gud press {het ae only thee poss 1. tthe reactor eluents tigi, we ssume ta we only eed aud sepration ‘sem (ce Fp 71-2) Th ater ight lode tilationcolemesetac on unt, emovzopicdetuation ee, but sormaly thee wil not be aby BS isocbet pas sorption waits eke 2. fe redo luce a tworphe nase, me cae the recor 2 pase ‘plier (or puta fash drum afte he reactor). We snd the but» had ‘Spunaion yc te reactor operating above cing wnt ‘sly col the sear vapor Steam 1K and pases th sam 16Se) 7 i= | ee] recor [ti | tit Sf one oo Tigi eae tr rt ld om J MAIER 5 83 oa resyle 7 Fee Ta ee (6c Fig 71.8 Ihe to temperature ah iad we obtain consis moe ‘scans (nd uo predict components tha fe forme sitesi In onsective rection scheme) then we rey them to thereto (we hare the ‘Suva of etx condenser) However ithe low-temperature fash hai ‘Snail prodocts we se hie stream othe gd recover sytem The Ieper ih aor xy aor ens therettor flue! Uist tos bguid separaticn sate i, cation ta) {empertue) and weap to acheve pe apt (Fis 71-9 orto ‘Sropely conde itr tren. The condensed ged tnt (0 8 gud ‘econery sen, ad he vapor is eto a vapor teaner sen a phe spt snot obliged, we se whether we capes he reactor stem 0 tht pe split wil te obs (ene whether ih fer can be obtain y using only amps on iid eed stream and we le rosea heres doce or afc he rods dst) phase ‘pits ot ohana then we conser the pony of eng both ih Presure and reread pain condenses Ip eae no phase spt can te ‘hain witout ripeation, we ako comer the posit of eeding the Feito een seam dtl oa vapor covery stem. ‘We neo enue tha the ume suas infor te comple range Date spear the cheapest metho of eparcbon andthe assumption Ut ome se we = er | Sete | = vie] fuga i ay Product Tae —| vote |166 scqow cnana smncrne oer sano se Approximate Flash Caeulations To dete he ph fhe et fern sme as, w a we # shy sol approximation procedure to avoid te iaanderersaluion anor ‘SEL ADECOIDENEST sae oqunton cone nano Ove balan: Faved ony Component als Fy Vy bs, 12) viii ek oy When we combine thes expressions, we obit VF vipa) tee) = “WIT bas UK,» Hin Bg 71-6 meee tat tats a6 tH, 1 715, det Lae Pe on ‘Thuan at entinate of he apr und gui Row rate, we ea write VEDs, loracomponests where K,> 10 a L=Sf,orstlenmponeats where k, <0 13) These capresions are equtakat 16 + pret eli, providing there ae 20 componeats ih aK, sue betwen Di and 10. aie "The pe pt eestor he vapor igi quan Howe, vs can sierinpot these equi Tlie ot expen ft ews Thus thetautcomponton is eguitrcm wha vapor coupon! being e140) aay The gui fw of hs component then 5) hat any secron 11 opm emtcrins OF pak anaaarion sare 167 HDA fash or) oman fo eb A Compe Diet 'e Ooomee Now, mec go back and adja the vapor flow fo this os wntobes(t- Pe) auay ‘he remeron eral emp into cy en musetee Geb oy i one, mi) vast A ‘ 1. 134 cmp prin nl a ns i HA a are ea ee ceaeere pas phe etter? SVT ‘otter shoot preter fo i cleltons was publ y King en comer Egh 71 through 713, we ean woe 7 o found » shew ary natn aun,-& ots) Now, ifwe divide Eq 71-18 by a simi expression for component j, we bin Seal #1 1) Gij=T" Ra, cue: Iw speci the fractional recovery fo ne component we ca we F114 to cat he ational ener er Cv er compere yeah han ou en he i specif action rconery af one component normal wilt corpo to Abc emperatire of 100F and the Bath reste. Tay, sme ration might be [Nonidel Mixtures Thete ten sore calcaltion procedures for nnidea mitures, Hower all {CAD mes es FLOWTRAN. PROCESS, DESIGN 0 ASPEN 12. VAPOR RECOVERY SYSTEM ‘Wh we stent ome aor ees sem ed Co mk ro 1. What isthe bet oeaton? 2 What ype f vapor covery sytem i cheapest? ‘Location of Vapor Recovery Sytem ‘het fou choices for the lction ofthe vapor recovery syste 1.7 purge seam 2 The preset A The tac vapor stream 44 None ‘Theres we we to mae his daston ae (ee Fig 721) 8 allows 4: Pc te igor soe eon ug sea fin amos of Value teria are Being tS the pg. The eso forts Peri ‘hat the pare steam nomily haste sas ow rate Vigor tr pis sit 2. Pee fp cnr oe ee eam mae ha {ireious otheyncor operon (etait point) ae present this Site orf tcoying of tome component dae he product sto, Tepe sem soma hese sie fw 2. Pc the apo ecovey syste othe fash spor cam ifboth items I and? eval rth ow re higher, bat me aeempish emo OnE Adjnst the Material Balances? Nott! usin cu sine atrial ane gt SHEE sme natn tune ee eed aga eon Tepussen rt wh gar seni wi gt compre Seton nyc te ened ora or Fos ‘Moetinaes ull poe eed wma We expe fo dy rose see cam tse cet ih al dsp, td there me eo TR Nites cresaor ed be mde so ume 721 HDA rm Bowe et pr ey snr HDX Fetter pues ms ven nae Te pre td ce obo SOTA ESC on mle tpetne. Hee’ we a imate in the BS.St Seen tern in te pre 8 Osea recy On See's avs ty wesectg te ft vale fee compre i REDE Sie yr a au a compare or com ei,Sevclomeat We nigt wel conse this dan are hae dese ‘hse the HDA pes probable, ve dae sued te pe ‘etl mn be sce at ‘xsseyce steam ca ase eat dentvaton Howeee tie bape ‘Sot f team he ah apr sea 25 ee Nee Rowe ate {E1-Ad mon fs eee (ath Seah ite rgd) Wo ‘eo oie ator Te betel he mrt come “ota y+ Reese + He eran Dipl +, ‘Tey ne wou cp hat eo Bsa ent hat eel oe ‘cover wiv cert ghey. Una ec as Seoaee (ie Eran) oot inclaethe ea any feed be trees etn at of eae ot pay Ti iy Sd ‘mn eo he gery steam pre ha eae eke ‘toa a dips oath ah vapor scan operetta tense hg th puesta ay rain) Anetta el te rage kt Sentech cn an ge Upeay W eng eo te cia io a pn cpr nd Sao the hanes te parte eae Oe Srnec tastes ee ete eo ce Te ‘in each of he thee spe ofthe ge ey comp, se feta era’ Nore cl te fas each cpr Hoge to HOF ean ‘ool wa ad tet eines soul un a e toss cere mt Ne cn ha oie ca en ue aman ei a he a teri hn ie mo “4 ‘Pores aiam wil be econ te krtont ums ance owe tives chek tires we moe rows i ne! heb in taf he sh ‘vapor scams, our acmptoce sonrnag the overall and ple mate Salone ae logs ali In partculr,the arc of bene eahing ithe Ash iid seam ot adequate to met the pant oto ae alg he ee emereecoveedin th compres knockout dum wiles the mgatde ‘We oul go back and eis lou materi an caestions. Hower, it wil eae overt again alter ne be specie ua separation ‘tm Since he changes we ince inthe proces Howe ae nf ae, ‘ce fo comic ithe atlas Of cous, ne ar lating to secumte ‘ureniclson We decribesprcedae fr eoeting the mata alaner a “Type of Vapor Reconery System, ‘The most common cokes (with cura! ecotoy) ae 1. Condensaton-highpesure or low temperature, o both 2: Atsorpon 2 Adsorption 4 Membrane separation process ‘5 Renton spam Shortcut design procedures for gas aborbers were dcused ia Chap. 3. The conic taf for the design oft condensation eos afe come in Exercise 3.2 A dig pocedre or aorption prensa bes preset Fair® Neier «design posture! nora sos ovation fr membrane recovery proceses seen to be nvalaie Lathe open Itertute.abough Yendor of from es scan rad M5 eve wih amie Surstexy We design the vapor scovery stem tlre we comsider the gu epuration system because each ofthe vapor tecotery process sal gterates bid ‘team that ste te prt. Forte co ofa gs aber, whee ee eed {owpel asobeat tote abort we in intedare's we eye lop bowen, {he separston rye (ee Fig 72.2) Nermaly we need to stents te se ‘ts ofeach uni to deemiae which the cheapest Ie ne ese ent ay(——_ ne Nepercey F C — "ieee =] a so L_ Die ae mtn ‘Combining the Vapor Recovery System with the i Separation System Itwe we partin ondener and ss dram to pas it the rector flea, tome eth ight gid sompooens wl leave wih th ash vapor tesa ach Str meer ells pect spt) ad theeoe wil not be recovered ia the Bg {ecover sysem Howene tore only asl amount of vapor in he seam [ving the paral condenser andthe fre spt the gid seperation ste ‘oven to be dsiliation, we cold climate he pase spite and eed the ect ‘faetstesm desta. the distil column. “Te sams of lation clue wih the (wo-phas ed will aed to be Inecr handle te ierease poet) tht clare tht fellows a ash ‘rum Hower ths inrened cot may Bele hea tbe costs aoa with ws thor covery spt to fomove the gid component fom the Mash vapor Shea Tere dos ot tem to be a eat valle forming thi eon Sh 0 we pnt fo add ance proces alerative our Bt 7 LIQUID SEPARATION systEM “The decions tat we need to make to spate the Hi sepration syle incl he following 1. How shoul Hight ens removed thy might coma the produc? 2 What shoo be he destin of he ends? momen ro taqune seesaamon seems 173 toi the estore 4 What sparatons an be made by dition What sumnc of column do we ie? {How shold we acormps separation tian ates? ach ofthese decisions ie dicued below. Remember hat we watt make tes (echions a Tunchow ofthe design varies over tbe range a seni pwoftahs operation Light as ‘Some ight nde wil be ced in the ig leaving the phase splits how Figs 113 and 11-4 and orn sme il be dived athe bud steam Icing ibe vapor recovery teem these ight ene might contaminate the fret, they mt Reeve ‘moving ie es ate thse: 1. Drop the pret or incense he temperate fasta and remove the ih ends in phase sie 2 Use paral condemer on the product coun. {Use pasteurization section onthe product column. 4. Uses salir column before the product column “The st thc leanne ar shown in Fig? “The option re nt the der of erasing conta hfe we pele to we te eer entries However, to make «deci frit nde removal Imconay to know the How rate of height ends ad to mate some shore (ielstone a sre CAD re eta te amount eevee 1. Pas cakestions, These ate deus in See 7. 4 Pata condenser. CAD programs bande the problems on some cae the roximate ash clelaons gen in See Tea be eed 1 Pasurzaioncolumas A shorten design ogdure has teen published by Gino and Malone” (ee Appeniis 3) 4. Staines clr This normal dstition clu that removes ip endsAe rt qui i nace, pn te: subir ohn | Note: Recjele vp steam 6 vapor recovery ate f one DESTINATION OF LIGHT ENDS. For the desi ofthe et ends, me ca ‘ot them (pss 0 Bite stem) send bt eds ofl o ye the it nds to the vapor recor stem othe ash drum the Hg ends hae very iu vate weteantto remove hem fom the proses though eat TS veatog ‘uns ae pllaio problems we fo vent therm troup Messe to br {he inding component Ins ofthe it endrarefarumabl wey 0 recover thefue vale: However thei ends ar valuable, ve want etaln tem athe proces If we rey them tothe vapor resorery ssn, we intodoce another reel stem ino the prose, SUMMARY FOW LIGHT ENDS. fight ende wil 20 contaminate the produc, te merely recy them tothe reactor wth reacant-retyde tea ot Temhove ‘hem rom the pros ih y prodt su that set the el spy IF "ght ene wl sortaminste the prod they mathe remove om the process ‘Themethodofremoval adhe dsination of the lphtcns depend on he amount ight ends Hence we met determine the amount of hgh ends a faction of the design variables bere we can maken dein, Aveoropes with Reatans 1 component forms an azsorope wit reactant we hae the choc of eyeing the azcorope or spliting the zeottope and jus eyeing the estat Spins sree }) uu stannox sre TTS te aszarope normally request colunns an therefore isexpenste Hower, ite rey the ascot, we mat ove all he equipmes athe eye 0p tobande the ntementl Row othe extra components A peoeal design heise ‘des ot em ob svaale for making ths cin, and 0 we nly ex ‘vale both alteratives Azeotropic tens ae discussed in more etal ithe Applicability of Dis 1 genera dition isthe feast expense means of separating mites of Ugude Howeve, ifthe clave volts of two component wh aoviog ting pins estan ors, dtilaion boone ery epee is 8 re fen eno regied which soreponds to stare vapor tte eam tlanetr, hag condnsrs ad cbr, an rg steam sod cooing water cst ‘Whenever ne coounter two neiorng components havo ela eat ‘ler thin nmr, me poupUhecoponeetstoger a we eat his ‘Poup ave single Component i the mint In oer words we develop the best ‘Slaton sequence forthe group apd the ober components aadies wesepate the limped components bya oer posedtes (ce Fig 1-2) Colum Sequencing—Simple Colas For sharp spi of thee component mitre with no zeotrpe) we can ter fecover the ight Component sto he Dearie compet Sat thee We {it the remsinng two components (ee Fp 123) Whea the number of ‘omponets ments, the mum alerts increases ery aly Ge Tae ‘i The spits that com be mae i the 14 alerts for «fve-component fniatare ate ated in Table 73-2 1 appeats a twill be sence to met for paras peace, pty se the Bat sequence 4. BO spare pris bec1196 ercrion 23 can strana ear scm 94 tian eran sete 177 Genel hrs for cola sequencing 4)ri} 4 eT © © . ht change sweeten variables Tosimlify hier we might want {o Took for beurstis for calm sequnang, Thee has been » comsiderabe ret Inti {esearch ert in he aes our the past deco 0, nd some of the ress are ea telow ee | [GENERAL HURISTICS There ae see seerl Heute hat ane sed to Sly the selection prosedre fr column boqencs (ee Take 7-3). The fst Auris inti i sed othe fat that the rater ofconsrction of the Calum is much more eapesie than carbon se cores components ate [rset Thos the more ecm that cori component passes Coup he fore epee wl be the dition tei. ‘Rencine component il chang the reparation problem and ths should be coved at oon posble Monemes orto oti eres or he A tunes af actu conditions in onde 0 desea the column ovebeed and £ ‘ovtom temperatures ao thet the rate of polrmeraation gece. Vacum lumos re more cory than preture columns, abd we preter 10 WO The Increased deaning costs ‘We preter to remove products and reyl steams o packed bed reno as 4 dis to vei comin Of the pruct or rye sam wth Bets Tones rast ee, which always accumelte i a process Ii eee To ae Femovea predust rye steam ae bottom eam, ie ena por ete eqns fore prodect wrens fom xr and then condensed apna Atte sme tine n salud puree Cit Cohen? Calms Ga ‘eum say be taken fem te rebolerto prever he ble of contaminants penetra coum" seQUENGIG MELMISTICS TOR SDPLECOLINO A enero =e Ss Soe bara cee senseless wbontop a = | Sate ensy teres pase he pwiaabe 73 see ae be eared ers e Hr te set ee a sae oe v= 2 a a 2 isesteee |78 wenon 74 sq sana ere However, the stand Eh ears i thi a depend on fd compositions, ‘tens tbe second and fourth depend on rate vous Hea expect, ‘hat hve hewn wlan to contedtions teri the moat len componest, Ite Resist thee confit between the St and wcond hei ‘longer it of euratcs has been pls by Tedder and Rods” und some lovestgators have ted 1 order the porno he heute fo rele the enfics” A survey ofthe erature hasbeen presented by Nati Stephanepow Toxand Westenber! and. detae dcanon ofthe intense heres as been paid by Malone a! Some adaonl desea he Bears Bree teow We might ao not that at we chang the conversion na proses. ne pet that the unconverted aca wl oom being the mot lef crane a ery low conversions tothe lest pleat very gh everson. Hem. ‘asics n Tale 7-4 py ta the best clea uoeencs wil change se alee the design varies Smal, note atte ste wae to deve the heuristics were ited to equenn fame colunasbaig gle ed steam that were late fom the rmatnder fhe proses leet es sy beobianed hes we conser he interaction betwen lation in en the ‘nainde ofthe pant INTERACTIONS HETWEEN TIE SEPARATION SYSTEM AND TIE PROCESS. For example suppose we consider the tao flows aleraatne show in Fig ‘73.de and 8. We might conser thse two configurations to be two of the ‘erates ina sequencing problem. However, thre diferent number 9 ‘clay inde Iauieeee lop fr the two astm, ad tere the eye ‘ste wl bein. Hence, the pum conversion wich usally corps {0 trideoff betwen selectivity loses and rryele cit, wl be ier or he ‘0 cates OF couse, we should compare aernativs athe optimum powesiog ond feath sera ae that oo se Mena fst copa ihe two alermaines, ‘rom this Sipe argueot we sce tht he problem of setng the best separation sequence cannot ways elated oe te dei the eae of {e proces ie, the least expesve sequence fr Sued fr tenn contin ‘ght not be the lst expensive sequence (ecauce tbe feed-siteam condition should te chunged to conespod tothe optim Bow) In at, thee might be ‘other eis Sls the sequence hat minimize the umber ofecmas in eee oop Tan. {IN rida G Stepbarpoan an A We Waterers ACA 129-336 2981) ‘MP Mle mos EE Meena Dep AGA, 4, t Bes A tea J +[ccotnt ot] me pm ae a iad @ £ Fix } oa} orice} —[ — wa I | ote sy » See ote ap eye MULTIPLE SEPARATION SEQUENCES. Suppose we conser spuation sy fem tha corespond tothe general owthet enn Fig 11-2 or P32, we ‘pits 0 that some of the most vole “gud components wl eae mt he Aish var and onc they need he recnered and see hid separation fasten However, the Mash id ght conti tre numb of ach enecomponents wel as thos that re tare fom the vapor recover pte Ia ‘Stuns sich thi might better to pt the neering probs ie a arts Tha we would split the fa iid ito ane port containing he ‘ogonens ere fiom the vapor cover system and one poron sence the heave components. Thea we would deign one separtion yt hace single fd siren forthe heavy componente sad one separation pre Natag ruled streams for the componente eure emt vapor fossa AN ALTERNATE APPROACH TO SELECTING COLUMN SEQUENCES The ‘eon for attempting to develo hence tha the Momber of Wenathe Seauetrsinresses very ail tthe number efcomponents resis oe Tae 73.0, However, tht area ge number af plants were foto los alaiog alma te sended to acempsh the separation, Fou simp orto top and one bottow stan ate esd to sorte fie cempovent mitre a fae re seas, ut only for clans are needa! to separate compen oe onporents with reighoring boli points leave ithe seme rea, Te fhe exit seas od sing oly single ores, we Reed To sons the 4 ‘An examination histo neat that 20 can dese ae require to ‘alu al the posites. Teeny column designs requis 4 soviet amount of fort cach of the design x riprow. However, by wing shart ovals Gee Append A) is posable a spicy split aces, ‘ons Using sorta exnigues Gost demonstrated hat Oh evalu he ‘4sequencrs sane instantaerur ona VAX 1-10;¢e he eat sopeed ss s00n a he program was um Kishen ha showe thie 1 soqeenes coe te evalated in only ae seconde onan TBME BC XT, The els of lino nd Kirkwood inst that for modest egeensing problems iti tite: o deve computer codes that evlst he con of eaean erative than itis to se huss Moreover, the nnn tine for tease estes hat the bes Soqence an be determined ay fton ee Complex Colum Rather tho conser ony suences of simple calms (one ovechead td one oom steam) we can comer the use ‘of sierream column acs Stipes and tebe, prtionatr, One st ef eat or cola ot Titi fo compesclum—Tret aod Ro HEB 1.4 share fe EM IESE >. A tee hn he itt tn mae one ada en 9 Soe eure a 7 {ithe 16m and myo onan oo ro recta or gn Dacor get 3 ihe mee ho el soon mae hn 5 ton dc hme dig 2 ince aS mi pe tt 8S etme en td 4 Woe arte sd ones 1 ere dan gm oes nen en 8 oa enn ae epee, fan an ba hee sil ee 2 Dabo nt ih anid fr osing a ital ite odt ranean TR amen sree ep pee pind pf te ‘hi type has been plied by Tedder and Rd (ie Table 75 and Figs 73:5 ‘and 7556) Anoter st has bes presented by Ghioos and Malone" (ce Table 136 Some shorten! dein pocesure hat ate wea ciple oma att jen in Appendix AS ‘COMPLEX COLUMNS IN SEQUENCES. Our goal to compete 4 bavecse sin sap as posible in onder to make a pelinnnry canton of whether 1 Gran BL Mae “Conn Coan mi oD Sy” ae