Polyethylene by the Ziegler process: A preliminary design and cost estimate

POLYETHYLENE BY THE Z IEGLER PROCESS

A PRELIMINARY DESIGN AND COST FSTIMATE by

Roland S.T.B. Driessen

University of Fennsylvania

1957

The author vishes to thank Dr. M. C. Molstad for many suggestions regarding the oomposition of this report. Also, Mesars. O. H. Hariu, E. J. Rollins and D. E. Smi th of the A tlantic Refining Company, Re search and Development Department, were always ready to give valuable advice whieh was sineerely appreciated.

Table of Contente page I Summary ₁ 11 Background 1 111 Introduction

₄

IV _{Process Description}

₁₆

V

Equipment Design

₁₇

VI

Economics

₄₆

VII

Appendix

₅₁

11 Background 1

1- General 1

2. Ziegler Process 3

III

Introduction 4

1. General descr1ption of process 4 2. Discussion of process ₄

3. Economical prospeots 14

IV Process description 16

1- Scope 16

2. Process Flow diagram Opposite 16

3. Description of flow 16

V

Equipment design 17

1- Compressor Section 17

2. Ethylene dryers 20

3. Reactor Section 21

4. Ethane and Ethylene Removal 28

5.

Fla sh Chamber 34

6.

Catalyst Removal Section 37

7. Additional equipment 43

8. Nitric acid recovery section ₄₄

VI

Economics 46

1- Capital Requirements 46

2. Annual Operating Costs

₄₈

3. Rate of return 49

4. Diacussion 50

VII

Appendix 51

I. Smrmary

A prel1m1nary design bas been deve10ped for a plant wbicb vill

he able to produoe 38 m1l1ion __ pounds 'of polyethylene by the Ziegler prooess.

(

Tbe Ziegler process yie1ds a prod~ct of better qua1ity thati that made by tbe conventional high-pressure process, end the manufaoturing co.sts of tbe former are probably lover. A1so, the proeess is extreme1y fiexib1e

, ,

in that it can produce polymers in a vide range of average molecular weights , witb only slight changes in tbe proeess eonditions • To make an

optimam use of this f1exibility, the' polymer is kept from dissolving durilig tbe process, thus eliminating great viacósity variations ~~ need for large solvent circulation.

Tbe eoonomie evaluation of the protiess as deecribed herein is quite inaccurate, but nevertb~lesa c1early indicates tbat tbe process should he investigated. A rate of return

o,t

7C'J1, at loC'Jl, capaoity was

found, and the break-even point lies at 33.5% of max1mum capaoity.

It is believed tbat at this time design and operation of a pilot plant are not feasible, since 'too many basic data are not known. Benob scale experiments sbould first be conducted to obtain the necessary

minimum infonna tion. 11. Background 1)

1. Genernl

Polyetbylene bas been known for a long time (1898) but it was not until 1934 that the first successful steps ware made towards the deve10pment of a commercial process. Taia research led

to

a ~riee of patents issued af ter 1936 to Imperial Chemica1 Induetries, major British

.

chemical concern. Becauae of the Seoond Yorld War, researoh and develop-ment were strongly pushed, and as a result, commercial production started

in tbe early 1940's.

E

_In :> J: u < UI UI <

"

Ö a a ;: 0: a z u z >= z <

..

:E a u

'"

() a al x W o a u ·cö z ~ 111 .~ o o " ,. al 111 '" ui ,. < 3: J: l-a al J: , U I '!': " 0: W

...

UI Z I ~ I 111 :; ë , 111 1 1\1 t'I o Z I \ -o -I I' - - ' 1 - - 0 - I I o

~

I

o L --o

J

10['10

The process, and lAter-deve~oped varjants, is carried out / " under very high preesures

~v~thousand

atmospheres) and elevated

temperaturee (up to 5000C.).

The material bas proved to be highly usetul, 80 that ths pro-duction and consumption rats since its intropro-duction have increased tremendously and continue to do so (Fig. 1).

Polyethylene holds a large share 9f ths thermoplastic resins

~ket and contfnuously 1D.creasss its share. It is expeoted to pass the

bUllon pounds/year mark in 1960, and will he ths f1rst thsrmoplastic - res1n to he produced in such quantlties.

Sin?e 1950, three baslcally nev processes have been announced by the Max Planok Inetitut fur Kohlenforechung, connnonly identified vlth lts present director, Professor Karl Zlegler, tbs Ph'11lips Petroleum Company and the Standard Oll Company of Indjana. All three processes are thoroughly cov~red b,y patente, but procees information is scarce, 1f

existent at all. The former t'Wo processes have been put into commet:0ial operatlon in 1956 and 1957 by leve ral manufaoturers, and more plants are under oonstructlon and in the planning etage. It seems that tbe Standard 011 proeess bas been llcensed, but no plans tor commercial production are known. The nev processee have several features :in common: that appear to have distinct advantages over the oid procees. Most important of these

_.

_.

are relatively low operat1ng temperaturee and pressures and jmproved

2.

product qua1ltiee. Tbe tmproved product qualitiee orlginate in ths higher degree of crystallinity obtained inthe polyetbylene manufactured by tbe new processes. For further detalled information, one refersto the initial study and comparison of ourrent polyethylene processes, whlch initiated underlying study and a simUar one on tbs FhUlips Pr.ocess.

I

.

It should he definitely stated that at this t1me,there is no information available on which a preference for' ei~her of tbe two procesaes could be based.

Af ter the arinounaement of-tbe low-pressure procesaes, a modU'ied

I

version of the high-pressure process was announced, wbicb claimed improved

I

\

próduct qualities over the old process; but gives a product inferior to the low-pressure products.

2. ZiegIer Process 2)-6)

In 1949 Ziegler andco-workers discovered the addition

ot

ethylene to Aluminum alkyl compounds, yielding compounds of the following general structurel

C2H;al

+

02H4 ~ C2H; - CH2 - CH2 -al -~'.'-1 C2H; (

CH

2 - CH2)nal

(al

=

1/3 Al) •

1his discovery opened vide perspect1ves for tbe polymerization of ethylene • However, in this way only polyolefins of relatively 10101

moleoular weight oould be obtained, eince a termination reaction took placer CH,)-(CH2)n-a1

+

_{C2H4 .}

-7

CH,) (CH2)n-2 CH2=CH2

+

C2a;

al Ths frequenoy of this reaotion apparently was too great to allow the

~thesis of solid polymere.

However, it was disoovered that the addition of co~atalysts not only increaaed ~e rate of the addition reaction but also dfminished tbe ratio of termination - addition reactions. Initially, colreidal nickel was used as co-catalyst, but later metal salts were found to be even more effeotive. Outstanding features of tbe nev procesB were tbs lew required

• This mode of express10n eliminates the difficulty of desoribing what

DR'IINC,

ETHYt...éN€.

-

-

_{COf'// PRo ANO}

PUI? 11=. I ' , , . " \ . ~ ,," ...

:-..

lCéAC TOR.

-f' LIGI-I7 END::, lêëHOV/::1t-~ ,cl.. A5 1-1 VqPO.f!.·

..

I Z f·.:rr / ON ~ Cc;TRCYST !Zé .... fnv,!:; t..

.

hL TR.~.;no/V It VVASI-i/N(j Dt<.v IN c, .5TOR.AGé:.

t

Po '-Y E:. 71-/ Y L € rJ

e.

.:50L U€:r./ i -CAIè.R../'t:: /è:., Reel/ze. -UI...A710,v

. its high deeree of crystallinity and low degree of branching. TensUe strength was increased and melting point raiáed considerablYi up to the theoretical1y predict.ed melt:lng tem~ratures of alkanes of 1nf:lnite molecular weight.

Experimental techniques were simple, and' consequently it was expected that the process would find vide commeroial application. At tbia Mme it seems that this expectation was more than justif'ied.

lIl. Introduction

1. General description of process 2)-8)

At present it appears that the process is carried out in the following steps I Ethylene , tree of' non-hydrocarbon contaminants and of

acetylene, is led into a solvent wbich contains the oatalyst and co-catalyst. Solid polyethylene is f'ormed which transforms the liquid

into

a slurry.

Af'ter the reaction is terminated, the solvent is removed

b.1

a combination of f'Utration and distUlation. The catalyst, which remains in the polymer, is decomposed by water and dissolved in dilute acid, which ~s then removed by washing. Finally, the product is dri~d and may then he further processed.

2. Dlscussion of process

The ethylene bas to be entirely free of' its usual inorganic contaminants sucb as H2S, C02, H20 and should further contain no acetylene or sulfur or oxygen compounds. All of these react" with any alumin:um-alkyl

9),

as can be easily illustrated for _H20a

al Alk + H20 ~ al OH

+

Alk H

This thus deactivates the oatalyst completely ~ surf'ioient- of the impuri-ties are present.

s.

The oo-ca.talyst, whioh is TiC14, vUl also.he deoomposed (hydrolysed by

water), but only af ter all the organo-metallio oompound has reaoted. Tbis

. "

is an unfortunate oircumstanoe emoe the Al (Alk») greatly exoeeds the

~iCl4 in prioe. Sinoe the amount Qf Al (Alk») determines tbe average

moleoular weigbt of the polymer, the impurity lev~l should ba 010sely oon-trolled, and beoause tbe prioe is high, impurity levels of lover than parts

( per million

shoul~

aim-e-d a-t-. . -It appears toot any al-alkyl or even any organo-aluminum oom-pound will p8rform satisfaotorily. Hovever, fr.om a manufaoturing .point of view it seems that the lover Al-alkyls (from Al (Eth») on) are preferred, whereas for safety in handling 10) the higher Al-alkyls ap~ar to be more attraotive.

The current prioes of alllIlli.m;m alkyls are quite high, as high as $25!lb., but ths reason for this is not the high produotion oost but the limited availabUity and the pi1ot~plant soale of produotion. Inoreasing application of ths Ziegler prooess, not only for the polymerization of ethylene but also of propylene and in tbe future probably isoprens , ~!.

. .

~-pted to bring ths prices down to belowa dollar a pound. Tbe TiC14 is not expensive, but newest developnents, indioaté that it may he adv~tageou8

. to use TiC1) 11),.12) which seem's to be the actual co-catalyst. If TiCl₄ is used, TiCl) is formed by reduotion witb the aluminum. compound. Oonsequently, smalle.r amounts of both aluminum aJ.kyl and Ti-salt would ba required •

TiCl) is more than ten times as expensive as TiCI4, however.

Fortunately, only small amounts of catalyst and oo-catalyst are needed tor the reaotion. Tbe amount is determined from a ohemioal point of viev by the folloving faotors~

a) The indioated minimum amount is 0.1 gr cata1yst per 1000 gr

( po1yetby1ene. This wou1d presumab1y ba 'tbe amount' toot is

~onsumed

in the reaotion under most idea1 conditions of purity of re~gents.

b) Tbe mo1e ratio of a1uminum alky1 to TiC14 is tbe determining factor for tbe average molecular weight of tbs po1Y!1ler product. Within an approximate ratio range of 1 to

5

it is possible to·obtain average

molecular weigbts from 30.000 to 3.000.000. It may be noted tbat this gives ths process a very great f1exibility witb only minor changes in catalyst quantities, and little or no changes in p~ocess conditions.

c) Within the above mentioned range of cata1yst to co-cata1yst ratio, and above tbe minimum required amount of catalyst ae! mentioned in a}, the reaction rate is directly proportional tó the concentration of co-cata1yst, i f the other conditiona, inc1uding temperature, monomer coneentration and method of cata1yst and eo-oa~alyst preparatiC!n are kapt oonstant.

5),

12)

Summarizing, for a chosen ave ra ge moleeular weight of tbe polymar, eatalyst and eo-catalyst quantities are determined

b.r

desired reaetion rate, .minimum catalyst amounts and ?:,eaction conditions. Sinee it appears toot tbe minimum eata1yst amount is considerably leas than toot usually employed, tbe actual amounts will depend on des1red reaction rata and

opera~ing conditions.

Operating variables that tnrluence the rata of reaction, and

probably to a minor extent also the average molecular weight of the produet~

comprise the following:

a) Preparation of the catalyst -- co-catalyst allJl'ry appears

to

he important. T:i.C1₄ and Al....ukyl are bath liquida, when pure, and are thus re1atively eaay to handle and to a"dd in measurad quantities. However,

7.

the Ti014 transforms into a suspended material when brought in contact with the Al-6.lkyl. '!his is undoubtedly due to its reduction to TiC13 and

possibly lover vallney statee • About the pa.rti~ular techniques to obtain the largest and most effective surface area per unit weight of co-catalyst nothing· ie know; this obviously belongs to the know-how of the manufacturer. Since it is lmown, however, that simple mixing in the presenceof an

organic dlluent will give a satisfactory suspension, if not of opt~ activity, this item certainly would not be an obstacle for aucceseful

.-development of the process.

b) Ziegler claims that the polymerization ean be eondueted at temperatures ranging trom room temperature to 2500C. .Howver, he, states as most advisable the range of 50-l000C .,' whieh is contirmed by' the

available exper:1mental data. Although not quantita1;ively illustrated, it appeare true .that, simllar to other polymerization reaet,ions, an inereaaed reaetion temperature w~ yield lover averàge molecular weights, under

\

oth~rwise constant conditions • Also, inereased temperature will cause an increase in reaction rate, whieh eould tentatively be eolrl'irmed from

available experimental data. However, not enougb data vere available, and

~he eonditions of the experfments vere not suffieiently defined ~o evaluate quantitatively the influence of temperature'on rea,etion rate.

e) Ethylene conc~ntration is an important faotor in tb~ reae~ion

rate. Tbe maximum possible ethyle]le oonc.entration at a certain temperature

. ,

depends on the eomposition of tbe feed gas, the natpre of tbe solvent, and ths pressure. Although only strietly true ,for an ideal gas, it may be

,

said that the ethylene conoentration in "the solvent varies directly wit~

. "

pressure. Although Ziegler definitely states that increased pre~sure speeds

confir.m a definite relationshipbetween pressure and reaction rate.

However, Natta, in his exteneive studies on the polymerization of propylene, found definite proof that tbe reaction rate of t~is polymeriZation was

proportional to tbe conoentration of monomer in tbe solvent.U ) Tbe propylene polymerization employs the same catalysts as does the Ziegler

. .

process, and there is at present no reason to suspect aD1 difference

_-

_..

in reaction mechanism between the two. 'l'hus, it is concluded that the reaotion rate is proportional to the partial pressure of ethylene, under otherwise constant conditions.

Although Ziegler describes a few expe~1ments in which no solvent was employed, the majority was oonduoted wUb solvents, and this is also the procedure recommended by tbe inventor. It should he ~dersto9d tbat the term solvent in this case conoerns only the ethylene, which ie

actually dissolved before it reacts - the polyet~ylene is formed as a solid, 1f not af ter tbe initia1 po1ymerization steps; then very quièk1y afterwards. Tbe polymer , as prepared by this process, has a negligible solubility in any organi~ solvent - produots with an average molecular weight of 30,000-100,000 are not so1uble at all below lOOOC. Small con-centrations of

dissolved~lymer

give extremely viscous solutions. Viscosity of these solutions is strongly dependent onconoentration and ave ra ge moleoular weight. '!'his may explain why it .is advised

te

keep the reaction temperature below lOOOC., avoiding a) operation with highly

viscous so1utions, b) great ohanges in viscosity when changes in moleoular weight of the product are made, 0) resulting wide range of prooess equip-ment requireequip-ments, and d) high solvent-polymer ratio, which in any event

,

would ba uneoonomioal. Consequently, theso-called S~1vent ~SOnlY a.

carrier for tbe polymer that is suspended in it, and no oonsideration w11l be given to the possibllity of actually diseolving the polymere

9.

The range of solvents as recommended b,- Z1egler is extremely vide - any saturated and weIl puTified hydrocarbon or hydrocarbon mixture can be used, from pentane to gas oll. Tbe selection of a suitable solvent is a complex problem which interrelates desired operating temperature,

.

.

pressure and concentration of the monomerJ it is thus closely related to

<

reaction rate and required reactor volume. At a certain temperature and pressure the saturation conaentratión of the ethylene depends on the volatility of the solvent, its mole~r weight and its density. The aforementioned, factors are not solely determ1ning, however - the ahoice of solvent has great influence on an important process step, namely the removal of solvent from the polymer. ,It should be posaible to dètermine an opt:lmum solvent for chosen operating temperature and pressure • Th is determination however is very complex,: and sufficient data laak.

Therefore it is impossible to make a choiee riow which ie absolutely justified. Tentatively it may. be said! that t?e chosen solvent should he

of relative low molecular weight, rela;tive 10w volatility end low price i f considerabIe losses occur and i f à large hold-up proves necessary. Also, i f heat removal by boning is consid,ered, the mol fraction ethylene in the vapor is limited, since at too high en ethylene content no complete conden-sation can be obtained at normal cooling water temperatures, anq a low heat transfer coefficient on the vapori side of the

~ondenser

will occur.

\

All

these reasons led to the still rather arbitrary choice of a

butane-\nJU>~

~

•. }

e.,

~ntane mixture as a solvent.

~

~

Although aliphatiJl impuritios do Il<>t retard or lmpede the

poly-merization reac,tion, they wUI. decreas~ the concentration of ethylene in

i

the solvent since they add to itsvolatflity. Obv10usly, whUe the ethylene reacts and is continuously replenished, the impurity ~thane)

aCCUlllUlates and thus keeps diminishing the ethylene solubility. Thus, in

a continuouB reactor without gae removal, the purity of the ethylene feed bas a definite influenoe on the reaction rate.

Oooling of the reaction ~s important, einoe the heat of poly-merization of polyethylene ie high as compared to other polymeri~a tion

reactions, namely 1800 BTU/lb)'13) At a' scheduled

_.

producti~n

of 4400 lbs/hr.

.

of polymer, the maximwn amount

ot

evolved ~eat would be approximately

L

8 mUlion BTU/hr., vhich might ideally

rep~sent

a credit of $3/hour,

o

already worth an attempt to recover it. Since the temperature at which this heat is 'iven off is rather urifavorably low, it ia believed that heat

recovery should not be considered untll after the optimam-reaction temper-. ature bas been determined, or possibly as a part of this determfnationtemper-.

_.

.

Tbe amount of oiroulated solvent depende onthe allowab.le solida concentration. In the experiments

descr~bed

by Ziegler and others 6) 7), the solid conoentration of tbe slurry at which it could ~o longer he stirred, is frequently indicated and .tal1s then always in the range

ot

10

. . '

to 4(f/, solids (presumably by weight), which is stated by Ziegler in his patente as to Qe the possible range of slurry concentrations as far as tbe possibllity of stirring is qoncerned. PIl.blished values tor the bulk

density of Polyethylene, asprepared by this process, are 200-500 g(114), about 20 to

50%

solids, i.e. in fair agreement with the figures stated

above. !wo faotorsar~ :important in this matter: solid polyethylene swells 'in an organic solvent, and a fibrous ma.t~rial can hold muoh liquide In

hóv far one might speak of ~ solid pbase in. this case is doubtful - it is hard

to

see how a fiber of moleoular thiokness could awell by absorption of solvent. However, it is obvious tbat the packed fihers, trom the moment

11.

they grow on the co-catalyst surface, will hold a large amount' of solvent in their void space, whicb holding will probably he facilitated by tbe normal good solubility of the solvent in the polymer. It would seem from s:1milar cases, that the material, in the form' that it is produced in this

~

.

process, wiIl hold more organic solvent tban it will hold water. When processing a slurry, it should be made ce-rtain that it always ,remains free-flowing, and settling should not be such that blocking of l~s _,. or equipment could occur. The first oondition is controllable by the slurry concentration, and it is belJ.eved that the s6cond one is satisfied:becau5e of tbe low bulk density of the polymer as it is present' in the slurry. This consequently gives only a small difference in density between . apparent solids (solide with solvent in voids) and solvent. ~ additional difficulty is the processing of slurries of thermoplast ic materials like polyethylene because the temperature must not be allowedto exceed the

softening point of the material. If the softening point is p6.ssed, the suspended material is liable to'cover heat transfer surfaces with a very visoous film or aake, thereby greatly reduoing the heat transfer coefficient.

The melting point of the highly oryatalline, high-density material is in the range of 1.35-14000., the softening point is etated to be above

110°0.

Since; as previously stated, impurities in ths etbylene feed (ethane) tend to decrease the concentration of monomer, it is necessary to remove the dissolved impurities from the solvent before this is r~circu­

lated. In tbe particular case of et~e, it will ba advantageous to remove this under elevated pressure so as

to

avoid gre~t overbaad 108s of solvent material, withou~ hàving tó use refrigeration for the overhead condenser.

\

I

f7

Tbe catalyat, which forms the nucleus of the polymer particles,

.-bas to ba decomposed and removed; if this is not done,'the polyethylene discolors and lose,s in mechanical etrength. Also, tbe presence of ' ionfDable matter in the ,polYmer d:iminishes,

its~xcellent\dielectric

properties •. From a competitive stand-point, it is important to obtain a

,

low ash content of the polymer, since the old prooess produces a mate rial

,

.

with ash oontent practically nU. Tbe obvious material for tbe decomposing of the catalyst is water, whieh however bydrolizes the aluminum compound and the titanium salt to waYer-insolubl~ hydroxides; therefore dilute acid has to ba added

t

5)

This aga,in bas to be rE!moved by washing before the product can be' dried. lt

mày

he tha t the poor wettabll1,ty of tbe polymer wUl make efficient wàshing difficult; in this case a volatUe wetting agent (such as an alcohol) might have

t~

b e e <J,

A detailed choice of dryer is not possible at this stage, aince this depends very much on tbe part iele size and partic1e size distribu-tion, about vhich nothing is known. It ia certain that the drying should

_.

ba done under'inert gas, since tbe polymer bas 1ittle resistance against

---~--~ ,

action of oxygen at elevated temperatures. Aa in case with ths slurry, the drying temperature should be kept ba10w the softening point of the polymer. At normal pressure this lies just above the bolling point of water, which would class1fy the operation as v~ry 10w temperature drying, with consequent 10w heat transfer coefficient, low rate of,evaporation, and low tberma1 efficiency. Under tbe given ciroumstances, a cboice betwen tbree dryers seems to he indicatedl

a) ,Tunnel conveyor dryer, which can not he applied i f the material containe duet. This dryer has the advantage of a bigh thermal

operate under vaouum in this dryer, thereby inoreasing speed and efficienoy. In order to use this dryer ,the l' U ter cake bas to he

soored to allov gas flow through the conveyor band.

1J.

b) Indirect heate~ rotary dryer,' which can only ba applied 1f "

, ,

tbe filter cake, wh~ in the drying procsss, breaks up satisfactorily.

, , '

. ",

11' this is tbe case, tbis dryer.might ba tbs best selection, sinoe it appears that vacuum could aasUy ba app11ed to this dryer. This would not be'a standard design, however.

c) Spray dryer, which does outstanding service inthe drying of thermosensitive matt!lrials. The temperature of the' drying gas aan be

considerably h~gher than the highest tempe~ature tbe mataria.l can stand without damage. (Air of, inlet temperature 4000F. was used to dry animal blood.) However, tbe feed to a spray dryer

bas

to he pumpab1e,and it is doubttul whether the polyethylene filter cake fulfills this condition. Bowever, otber filter cakes from rotary filters have been reported dried

in spray dryers • Tbe possibllity of addilig water to make the cake pumpible is of course on first sigbt not attractive. However, the possibUity should ba aonsidered,.

As pr,viously stated, the product, as manufactured by this process, is a fine power vith low bulk density. Tba powder form is very advantageous for mixing operations with other solid particles, like oarbon black 14). Tbis is a oommon operation to :improve on polyethy1ene

weathering qualities, and it is advised to look 'intc the possibility te produce such mixtures as part of the oonsidered operations. Because of the 10w bulk density, ths optimum way of traneportation would be by

raUroad

car~.

Bopper cars should be ueed. Also, it w111 be economical, to keep the minimum amount of product storage. It may ba eco~omica1

to

use hopper car~ for storage , thus hav:ing any-unsold product ready for transportation.

For many uses the powdery po1yethr1ene does not seem to he aatisfactory. For this reason i~might he coneidered to extrude and granulate part of the po1yethylene production, to make sales easier by broaden:ing the market. Since these are purely mechanical operations, not inherent to po1yethy1ene production as such, they have not been considered in this study.

3. Eçonom1ca1 prospects

As a1ready stated, both product ion and conswmption of po1y-ethylene have risen at a great rate in the past years, and expectations are :in general: - first bUlion pound thermop1ast1c material, in 1960 -. Estimates may differ b,y some 100 mil1ion.pounds, however.

Prom a tab1e by R. S. Aries38 ), quote:

In 1958 1t is estimated that rated aapacity for all po1yethylene types wil1 be one bUlion pounds, of which

30%

wil1 be tbe high-density types. Actual production, however, ie expected to ba 760,000,000 pounds, of wh1ch 18% will be the high density types. ~ 1962, production is

,

expected to climb to 1.1 bil1ion pounds, of which 39% wil1 he high-density materialo

The high-density types wil1 be consumed, it is estimated, in app1ications such as molded artic1es, e1ectrical uaes, film, pipe, paper coating, bottles and tubes, and fibers in accordance vith the fo11owing breakdown:

I i

.

1"-' 15 • Mil1ions of pounds/yr. 1962 High Denaity 10w Denslty Tota1

Mo1ded artioles 120 100 320

Electrical _{45 75} 120

Film 100 250 350

Pipe 30 110 140

Paper ooating 35 55

90

BottIes and tubes 25 75 100

Fibers 25 25

. Miscellaneous 60 40 100

-440 705 1145

Unquote. It is obvious that aprediction like the foregoing is subject to much oritio1sm, and it should only be used as an initial, rough gulde. However, it is certain that polyethylene, including its high-density variety, is here to stay and will claim an increasing share of the thermoplastics market.

The raw materials used in the process are in general readlly available. Tbe price of ethylene is rather stable

39~

around 6'/lb. For information on the catalyst, refer to 2.

Lo",-density polyethylene se'lis, according to latest information40 )

for·35~/lb., ",hile tbe bigh-density material selis at 47~/lb. It is belleved that' the latter will gradua11y drop as production oapacity increases. Most people'expect the price of conventional polyethylene to level off at or just belo", tbe 30-cent mark.

Tbe major market for polyetbylene is found in tbe northeast area of the country; it is estimated tbat an area stretching from 100 mnes west and south of Chioago, thraugb New England and to Washington, D. C.

16.

would include better than 75% of the market .40) Al though several poly-ethylene plants on the Gulf Coast clearly indicate thst this is not necessarily a decisive factor, it may be very significant in this case, because of the low bulk density of the product, and the consequently relatively high transportation charges. 55 ) Thus, a plant location in the northeastern part of the U. S. should be considered.

Only one source is known to give information on the licensing costs for the Ziegler process: $1 Million and·3% royalty.43) This first number tends to agree with Ref. 40.

IV. Proceas Description

The discussions, process data and equipment specifications which follow cover the manufacture facilities for

38

Million

Ibs.

of

poly-ethylene per year, manufactured by the Ziegler Process. 2. Process Flow diagram

30. Description of Flow

--

Raw ethyl;ne

~

of

~Purity, t~e

balance

eth~,

is delivered

by pipeline to the compressor. Af ter the first stage it passes an inter.-cooler and knock-out drum and enters the second stage. Af ter he ing cooled and having passed a second knock-out drum, the compressed gas is dried and

.- "...v ..

non-hydrocarbon

im~iti~~

are removed in a silica gel and molecular sievs

?

~

bed. It is then dissolved in the solvent, and the resulting solution is introduced in the lst stage reactors. The liquid catalysts are also

introduced into the reactors, by proportioning pumps. Heat of reaction is carried away by slurry circulation through a heat exchanger. Tbe solvent-polymer slurry flows to the 2nd stage reactors. From here, the reaction _..,

f •

.

:

~'

'I

~.\)

..

17.

mixture is pumped to the light ends fractionating column were dissolved ethane and residual ethylene are removed. Aftar heating in a steam heater, all solvent is flashed off the polymer in a flash vessel. The

I_f_c_o_n_t_am_in_a_te_d_ ~

?

with water, ft can be dr~d. Solids are separated from vapors by a

solvent is condensed and recirculated to the reactors.

-cyclone. Tbe powder is then by gravity and

steamfl~Veyed

to a treating vessel, where dilute inorganic .,., ,catalysts at elevated temperature •

/WV't \ _,

,'- are carried overhead and are condensed and separated. Wet solvent may

"kWf

;.J.

~

end purti'ied from heavier hYdrocar;::s and then reCirCUla:':to

Steam and

~~Of

solvent and acid

lr-

tbe reactors. Solvent make-up is added af ter being dried. The slurry of polyethylene in dilute acid is then filtered and the acid recirculated. The polymer is washed and conveyed to a leaching vessel. From the wash,

~,nitric acid is ~e~~t~and returned to circulation. To rernove

in-organic matter from the polymer, it is leached, filtered and washed in

three subsequent stages. Af ter the last filtration, the product is dried and polyethylene of

~

ash content is obtained. A part of the acid con-tained in the first filtercake ls recovered by distillation.

- - /

v.

Eguipnent Design 1. QQm~ressor Section Compressor Feed: 4400 lbs/hr Ethylene 100 lbs/hr Ethane 14.7 psia 900F.

Required horsepower:

HP

=

N_{s x 0.00436 V1}

PJ.

(n~I) [(~ )\N-~

-

1 ] s

N

s

=

number of stages

V_l = inlet flow, cu.ft/min.

P1

=

inlet presBure

P2

=

outlet pressure

n

=

~

=

1.23 Cv

P2

- =

10, consequent1ya two-stage"compressor is indicated

P1

HP = 190

At tota1 compressor efficiency of

75%:

HP

=

253

Install standard 300 HP drive.

Fuel, steam and e1ectrica1 costs in the plant must be considered before a decislon on the type of drive can be made. As far as siza is concerned, a reciprocating steam engine, an electric motor, a gas or a diesel engine could be c~nsidered.

Since only one compressor is used, a drive of wide flexibility would he preferred from an operating point of view, which would el:1minata the electric motor. An important consideration might he-that because of the

---

.

plantte low steam consumption additional steam consumption might ba

desirabIe (this only in case the plant has to generate lts own steam). It is feIt that no deoision can ba taken on the subjeot, nor bas it to he

taken - estimated investment and operating oosts are practieally the same.

; l ,

lntercool,r: n-1 n

PJ.

L"î

Tl = inlet temperature ,

~

T2 = out1ettemperature T2

=

22eJ

f'

b

Jó'tL

'\

At thls temperature _{there is no danger o~~~t~~}

----=--

_{of t~e}

H

= 50 BTU/lb.

*

Heat to remove'

Coo1ing water in: BooF.

out: 125°F.

A

=

Q

-~--U AT_{e •m•}

A

=

area exchanger ft2

Q = total heat to remove, BTU/hr. U

=

overall heat transfer coefflclent

BTU/hr. sq.ft.

oF.

U gas-liquid

=

10

**

A

=

450 ft2

Install 550 ft2 exchanger

Cooling ~ter requirements; 600 ga1/hr. Aftercooler: ldentica1 to intercooler for same duty. Install 550 ft2 exchanger

Cooling water requirements: 600 ga1/hr

Summary

of egu1pm~reguirements

Two-stage reciprocating compressor, 300 HP 2 Exchangers, E-l and E-2, 550 ft2

*

All enthalples from Ref. 52

**

All values of U are.estimated from values in Ref. 51, 53 and 55.

19.

---2 • Ethylene Dryer~ .31) 17)

Designed for maximum water content of ethylene and 8 hr. cyc1e. Max~ water content in ethylene is unlike1y, but since prooess ls so

,

sensitive for traces of water, no chances should be taken at thls point. Thererore, a primary bed or silica gel is used, rollowed

qy

L1ode's newly introduced molecular sleve. This ",U1 not oo1y re duce tbe water content to <1 ppm, but also remove _{002 ,} H2S and other 0- ~nd S- compounds

to

an extremely 10101 leve1. l7 )

Basis Silica gel bed design (procedure .31), p. 882} Vapor pressure H20 @ 100°F.: 0.95 psia.

Feed: 5000 lbs. °Ethylene/hr., 100% humidlty, 10OOF., 150 psia, visc. 0.0106 cp

Maximum effluent humidlty: 1% Oyc1e: 8 hours

SUica gel: 8-10 mesh, 40 1bs/cu.ft., Sp. heat 0.2 BTU/1b. oF. Superficia1 gas ve10cityz approx.

1/2

ft/sec.

Average moisture content of gel at end of cycle: 30 lbs/100 lbs dry gel

Maximum laad: 20.5 1bs H20/hr.

Resyljï: Vessel diameter 2 t, length 4.5 '

Contente 570 lbs Si1ica ge~

Basis of MobcuJ.ar

"!en

bed d§..§im: In same vessel as sUica gel bed Capacity: 5% of total water 002

+

_{H2S content feedz}

<

5 ppm 002

+

H2S content effiuent: (1 ppm

Break-through capacity: 5 _{lb. H20/lb. dry ads.}

Adsorbent: density 40 1b/cu.rt., ep. heat. 0.3 BTU/lb. OF. Resulting in bed height of 1..3 rt., contain1og 160 1bs. adsorbent.

Re gene rat ion of both bede at 4500F. Required heat per cycle: 100,000 BTU under idea1 conditions. Assuming

50%

thermal efficiency: 200,000 BTU/cyc1e. Since the dryers should remain free of oxygen and oarbon dioxide, the off-gas of the desorption column A-I aould be used

to

advantage ae a purge gas.

21.

Preseure drop through the adeorbent bede amounts to 1.5 in. _H20, and aan thus he neglected.

Summary eguipment reguirements drying seot~

2 Vessels, F-:3 and F-4, 2' diam.,

6'

length 1200 Ibs. Silica gel

320 1bs. Molecular Sieve 200,000 BTU/8 hours :3 • Reactor Sectlon

Basis for design of first stage: Operating temperature 15SOF. Operating pressure 140 psia

. Maximum volume ratio polymer-solvent: 0.15 Feed: 4400 lbs/hr. Ethylene, 107 lbs/hr. Ethane lhlk density polymer I 300 gr/l in air

Raal density poljmer: 950 gr/l Reaotion rate:

R

₌

60 CC,

R

=

Reactlon rate, gr Ethy1ene/l hr C

=

Concentration Ethylene, gr/1 solvent CT

=

Conaentration TiC1₄, cc/1. solvent

Solvent: Choice based on possibility of overhead cooling. Arbitrarily set maximum Ethylene content of equilibrium vapor: 25 mol.%. If 150 gr Ethylene have been absorbed by the solvent, at which point it leaves the first stage

re~ctor,

it will oontain 150 x

~7

=

3.45 gr

Ethane/l. solvent. Using butane and pentane as solvent components, the composition of the solution in the reactor (thus excluding the polymer) is found to be by trial and error:

Ethylene Ethane Butane Pentane Density Molecular weight

Mol. fraction in liqu~d

60.8 0.0327 0.0131 0.672 0.282 526 g/l

Mol. fraction in vapor 0.250

0.071 0.578 0.101

j<

Ethylene content: 7.93 g/l. solution

~

\J'l

~ Solvent compoaition in mol. fractionsl

- -\ri'

I'

:Ik

\~

\:) ,,\

Butane 0.704

r"

"11 Pontane 0.296

'i'

of reaction, using 0.1 cc TiÇ14f1 solvent: \ R

=

47.6 gr Eth/l. solution, hour

~

Solvent circulation at 150 g.Polyethylene/l. solvent:

44

00 x 454 = 3520 gal/hr. 150 x 3.785

Volume

of

aolution in reactors:

Polymerized in lat stage reactor, allowing for 7.93 gr ethylene/l. leaving lat stage reactor:

4170lbs/hr. Thus, at reaction rate 47.6 gril. hr:

Add to this the volume of suspended P91ymer: Total Volume of contents: 1600 Heat balance:

. Heat of reaction 1800 BTU/lb.13 )

Heat contents of catalystshave been neg1ected.

.!Be

4400 lbs/hr. Ethylene, 1000F. 107 Ibs/hr. Ethane, n 10,130 1bs/hr. Butane, It 5,290 Ibs/hr. Pentane, ft Heat of po1ym, 4170 x 1800 Q!ll: (15SOF. ) 235 1bs/hr. Ethylene 107 1bs/hr. Ethane 10,130 1bs/hr. Butane 5,290 Ibs/hr. Pentane 4170 1bs/hr. Po1yethy1ene

Thus, 6,655,000 BTU/hr. has to be removed. Surface of heat exchangers

Coo1ing water: inlet 800 out1st 90°

u

=

125 BTU/br. ft2 oF. Slurry: exit 1000 A

=

1370 ft2 Install 1600 ft2 BTU/lb • 309 328 160 152 246 265 198 188 150 23. BW/hr. 1,360,000 35,000 1,621,000 8~4,000 7₂560₂°00 10,300,000 58,000 28,000 2,020,000 994,000 625,000 3,725,000

Cooling water:

6,655,000

=

80,000

gal/hr.

(10) (8 • .35)

Quantity of circulated slurry

90° I H

=

187 BTU/lb. H

=

142 BTU/lb. Ciroulate 6,655,000

=

{148,000 lbs/hr. 45 Pumps

=-\

Assume a differential head of

€J.-v.'

Total pump - efficiency 0.7

HP

=

148,000 x 100

=

10.7 HP

.3600

x

0.7

x

550

lnstall 15 HP

Alternate for heat removal

Forced circulation of vapor for bë.tter heat transfer coeffioient. Cool va por to 900 F •

By trial and error, when L + V

=

1 "d- '-]C'<: L

=

0.697

Composition of liquid

vapor Mol. fraction Mol. traction

Ethylene 0.1094 09575

Ethane 0.040,3

0.1415

Butane 0.714 0.2626

,

Heat removed per lb. hot vapor:

~.5 ~

Thus, circulate 47,500 lbs. hot vapor/hr.

Blowers for cool vapor

-t~t

Assume maxtmum pressure drop 5 ~~. Use 2-impeller blower.

-w

=

_{V (P2 - Pl )}

Circulate 10,700 lbs/hr. cool vapor, MY

=

37.1 HP

=

5

Condenser

u

=

17, BTU/ft2 hr. oF.

Cooling water 80

4

900F. A

=

1270 ft2

Basis tor deslgn....QL.6Dd staio

Exit concentration ethylene: 1 gril solution Temperature : 5° belou bubble point

Reaction rate:

---.:--- assumed equal to rate at 158°, although in reality _higher R

=

6 g/l. hr.

Consequently, residence time;

t

=

7.93 - 1

=

1.16 bra. 6 Volume of contents: Operating temperature 1.16 x (l.l,) (3520)

=

7.48 630 cu. ft.

Bubble point for solution is found by trial and error to be l7SOF • .---==-.. Since no boiling is desired, the operating temperature wi11

Heat balance

From lat stage reactor Heat' of polym. 206 x 1,800 BTu/hr -3,725,000 374,000 4,099,000

L

S'8°

F-Qlll, 1 TJoF. Ethylene 29 1bs/hr. Ethane 107 lbs/hr. Polyethylene 4.371 lbs/hr. Butane 10,1.30 lbs/hr. Pentane 5,290 lbs/hr.

To remove by oooling: 204,000 BTU/hr. Surface of cooling coU

0001 ing water 80

-7

900 F • U

=

125 BTU/hr. rt2

OF •.

A = 18.6 tt2

Install 25 ft2

Oooling water: 2440 ga1/br. Size Reactor Vessels

BTU/lb. BTU/hr. 25.3 7,400 271 29,000 16.3 712,000 207 2,100,000 198 1,047₂000 .3,895,000

For flexibllity of plant operation, and because of reactor diameter, it is probab1y preferabIe to use two reactor vessels.

For smooth preSBure control, a vapo~ space above the slurry in the lat stage reactor is re qu1re d •

lst and 2nd stage reactors are in the same vessel, separated by

a

horizontal batfle. The same opening in this baffle is used tor slurry tlow and transmisaion of the stirrer shart.

Required volume tor each reactor vessel: lat stage reactor

Vapor space 2nd stage reactor 800 cu ft. .300 cu rt. 315 cu ft. 1415 cu ft.

27.

From this, approximate d±mensions of reactor vessels& Diwmeter 10 ft.

~ngth 18 ft. St irrere. 18)

~"'~

To avoid a vortex, stirrers bave to be installed

ofr-o.nter~he

vessel

\1-\ \

vr

bas to contain baff'les. Flat-blade turbines will he used.

~~~J~. ~\

.V

~ ~

)'

~/Y <~.

V"

Other dimensions:

Diameter turbines: .3 ft. 'standard

~\.:.v

{IJ!

~J

S~t'

off-center __

:G

{:f

peed: 90 RPM

---~

~ol(l~~

0) ~ .

t-

Maximum slurry concentra tion: 150 gr polymer/l. so~ution, resulting in

L;.-JY

1.31 gr polymer

/1.

slurry. ,

.300 gr polymer absorbe 700 mI solvent. Speciflcatlons slurry:

Volume fractlon apparent solide: 0.4.37 Volume fraction ~ solvent: 0.869 Bulk denslty: .38.2 lb/cu. ft.

Bulk viscositYl 9.1 x 10-4 lh/ft. sec.

NRe

.=

n2

Np

~

D

=

d~eter turbine

~e the power number N is independent. of NRe

- Pi

Np~-

pN

₃

n

₅

=

6

P

=

required power, ft. 1b/sec. g :;: gravity constant, 32.2 tt/sec2 , P :;: ,840

, HP

=

10.6

Ona turbine is required for both lat stage and 2nd stage reactor. Instalied HP per st1rrer: 25

Cata1yst hand11n~

A

thorough atudy should ba made ot the hand1ing ot Al (Et)J and TiC1₄• Both are dangerous and poisonous .10) Locations where leakage is 1ia.ble to occur should be we11 vented with a dry inert gaso

Pumps P-3 and P-4

These are proportioning pumps. Approximate maximum capacity 2 gal/hr each. Summarized egpipnent reguirements tor reac~u.illQD

2 Reactor vesse1s

F-5

and

F-6, Diam.

10 tt. Length 18 tt. 2 Pumps P-15 and P-16, each 7.5 HP

240 gal/min. 2 Heat exchangers E-3 and E-4, 800 tt2 each 2 Coo11ng co11s, 13 ft2 aa~h

2 Stirrera, 25 HP each

2 Proportioning pumps, P-3 and P-4, 2 ga1/hr. 4. Etbane and Ej.;Wlene Remova1 52)

Pumps trom reactors to ~owerl

Duty: 20,000 lb/hour, density 38.2 lb/cu ft.

Tota!

~ffiC1en07 !!.4 ~

HP

=

20,000 x.1QQ..

=

2.53

3600 x 0.4 x 550 InstalIed Horse power I 4

Operating specifioations tOMler Pressure: 147 ps1a. Top temperature : 95°F.

Mol. fraction Ethane in bottoms: 0.0001

Assuming tha~ no ethylene reacts af ter leaving 2nd etage reaotor, the composition of the liquid phase of the slurry is:

Ethylene. Ethane Butane Pentane Mol. fraction 0.0041 0.0135 0.693 0.290 29.

From prel1minary rough calculation, approximate composition of bottoms: Butane Pentane Ethane 0.702 0.298 0.0001

Tbe bubble point of this mixture is calcu1ated by trial and error: 199°F. at 147 psia.

Tbe same calculation, starting with overhead composition Ethylene 0.1

Ethane 0.3

~

\

Ethane Butane Pentane 0.526 0.313 0.0014

Minimum theoretical stages to separate Ethane from Butane in desired degree:

LKD I HKD LKW I RKW t X Cl( _: ... SM

=

4.65 If D+W= F log c(LK

Light Key in distillate

Heavy.' !I

..

Light Key in bottoms Heavy 11 ft ft Mol fraction Relative volatility D = 0.0255

F moles

W

=

0.9745 F moles Comp. overhead Ethyl ene 0.160 . Ethane 0.525 Butane 0.315 Pent.ane 0.001

Minimum Refiux Ratio

Detailed method out1ined in Ref. 52, p. 231. It is a modification of the GU1Uand methode

refiux bottoms 0.0314 0.000008

0.14~9 0.0001 0.818 0.702 0.0075 0.298

)

(J

31.

Mol.

%:of feed

vaporized _{(O!D) M;·.}

0.4

_3.9

71.0

_30.9

Thus: _{0.0 3.9}

Column \1111 have reboiler and partial condenser , eaoh counting as one stage •

.

From corre1ation'of theoretioa1 stages with actual ahd minimum ref1ux ratio:

~

3.9

409

5.9 6.9 7.9 9.9 13.9 , Stages

00

10.7 9.2

8.2

7.8 7.5 6.7 Theoretical Plates 00 8.7 7.2

6.2

5.8 5.5 4.7

IJl'"

I Emp10y

oID :::

10

r

U

Add~ca~_J~~~~. ~..!.ty

measure. ).

T~tal :"~oretical

plate"'

~~ ~

P1ate efficiency

Average viscosity on plateel 0.14 cp

.. 1 '

Fluidity :::

VI8ë

=

7.1

Overall plate efficiency 52) :

®

Therefore, use 10 plates.

However, á.ccording to Ref. 20: Since éi( K-1

=

1.33

"

Since a generous allowance already was made by adding 3 theoretical platea, it ia felt that 15 plates will be adequate.

~--Tray spacing above feed platat ltv

Feed tray spacing 3 t

Tray spaeing below feed plate

(~ vapor load) 2i Location of reed plate.

Assume that fraction of actual p1ates abOve the feed wil1 ba the same as that required to effect the same separation of the key components at tatal reflux. For feed plate to top:

SM

=

2.21

ThuB the number of platea above the feed platst

2.21

---- x 17

=

8.1 4.65

Feed ia introduced on p1ate 8, counted from the bottom. Heat ba1ance around tover

BTU/hr

Charge 3,900,000

X 1ba steam @ 17.5 paia, AH=970 970 X

3,900,000

+

970 X

~- 1bs/hr Overhead:

E

thy1 ene , 29

Ethane, 107 Butane, 119 Bottoms: Butane, 10,011 Pentane, 5,290 Polyethy1ene,4,371 H 307 325 30; 225 215 185 9,000 35,000 36,000 2,248,000 1,360,000 809,000

Reflux: Ethylene, Ethane, Buta.ne, Pentane, Heat d u t y S Steam: 1130 1bs/hr. Reboller surface 1bs/hr. 565 . 251 3090 33 U

=

200 BTU/rt2 hr. oF. Steam temperature: 2200 F • A

=

260 ft2

Install 300 ft2 surface area Tpwer diameter AH 85

88

149 156

Vapor lead under top tray

=

ref1ux + overhead Volume: 7.94 ou ft/sec.

Density :0.129 1b/cu ft. Refiux density : 29.0 1b/cu ft.

BTu/hr.

5,000 25,000 460,000

5,000

4,992,000 3,900,000 1,092,000

J

A11owab1e vapor velocity at 1t ft. tray spacing and 1t in. 11quid sea1: VA11 •

=

1.94 ft/seo.

Requ1red tover diameter: 2.3 ft.

Vapor lead under bottam tray

=

Reboller duty

Heat of vap. bottoms Volume: 18.3 cu ft/sec.

Density: 0.130 1b/cu ft. Bottoms densityl 31.3 1b/cu ft.

A11owab1e vapor velocity at 2 ft. tray spaoing and 1 in. 11quid Bea1: VA11 •

=

2.64 ft/sec.

Minimum tover diameter I 2.98 ft.

Use 3 ft. tower diameter. &rtia1 cond~

Duty: 500,000 BTU/hr. U

=

150 BTU/f't2 hr. oF. Coo1ing water: 80 - 85°F. A

=

271 ft2

Install

350

ft2 surface area

Coo1ing vater requirements: 12,000 gal/hr.

34.

Summa.

ry of oguiment requirements for etbane-ethylene remoxa1 gegtlon

1

Distillation column A-I: 15 bubble cap plates

Diameter 3 ft.

Length approximately 35 ft. Reboiler H-l, surface area 300 ft2 Condensor E-7, Burfac~ area 350 ft2 2 Pumps P-l and P-2, each 2 HP

each 33 gal/min.

5. nash Chamber

All solvent bas to be flashed off the polymer. Tbe Fesaure in the vessel should he suoh tbat tho vapors can be totally oondensed at

1000F.

-

(coo1ing water). Thus, this is the pressure at which the bubble point of the mixture is lOOOF..: By trial and error:

P

=

43.5

psia

At this pressure, the dew point of the mixture is found to be by trial and error:

To make sura that all solvent has flashed, the solid polymer should leave the vessel at l25°F.

The average vapor temperature is estimated to he l300Fo

Heat balance, including heater.

BTU/hr.

Ethane tover bottoms 4,417,000

Heater Duty R 4,417,000

+

R Polyethylene, 4371 lbs/hr. Butane, 10,000 Ibs/hr. Pentane, 5,280 lbs/hr. Heater duty Heater Steam,Q 7.5 psia,

2200V,

AH = 970 Required : 1160 lbs/hr.

U

=

200 BTU/ft2 hr. oF. Est. slurry temp. : 200°F.

A

= 281 rt2 Install 350 rt2 H 128 569,000 325 3,250,000 326 1,720,000 5,539,000 4,417,000 1,122,000

A

U-tube exchanger would seem advisable for this type of .duty since it offers the least reeistance and bas the least obstacles.

Condeneer

Flashed vapor

BTu/hr

..!&L

Butane, 10)000 1bs/hr. Pentane, 5,280 1bs/hr. Condenser duty H 160 152

Condenser duty: 2,570,000 BTU/hr.

U

=

150 BTU/rt2 hr. oF.

Coo1ing water: 80 ~ 90o_F.

A = 700 ft2 Install 800 ft2 Burface area Cooling water: 512 gal/min. Vessel 1,600,000 800,000 C 2,400,000 -+C

36 •

Size of vessel hard to predlct. Minimum 101111 be governed by size of cyc1one. Cyclone is not used

b.Y

itse1f only, because

1) Additional pressure vessel is probab1y cheaper than presaur~ resisting cyclone.

2) Residence time of solide is greater, which gives more complete vaporization of the solvent.

Cyclono.

Design inlet velocity : 50 ft/sec. Volume: 9.83 cu ft/sec.

Inlet opening: 0.2 rt2 ;

Basic dimensions or cyclone (Ref. 51, p. 1024) Diameter:

1.3

ft

Length of cylinder and eone: 5.2 ft. From this, estimated size or vesselr

Diameter 2.5 ft. Length 8 ft.

1/

-/200 .5TEAM

"-_.

-F-9 soo STE:AM ... 1766 _HN03 / IG H N03 _O\ERHEAD SISO .'-12.0 COi\JSUMPTlo,',!

.

~_.r: ".!..O.3_. 3 2 7 125. 7 H tV 0.,3 ooE<:=---+---<~----. --2915 H20

---.

/313 hNO.] óe(--'---.. .-- .-.-..

--_.-

... -- ., _{"'-- . '-'----'--"} ---! _{19/7 H 2 0}

..

.1-

- - - -I

I

I

I

I

0 HNO ₃

VI

175 S8 I SO HeO

"'

V-I

1

FIL TE RS

I

I I +37 _{H '" 03} 3933 _{"72 0} V-I .327.5 11 N 0,3 YVAS H. _{8 of i I} Ha.

°

>. --

-I ₃₂₈ h ," 03 404-2 H2 0 ,c-_-10

.

""

.. I

r

-.... ! " i ! --

-HNO ₃ C R..ë -I I

~

SSO ! 874

°

H20 I 1 . \ ' ;" I J.I

U

,\1' TS

i

I ! I : I I i ... 1 0.') _=< ;-11. C.3 392:2! HZ"") ~-_.

__

.. _

-f

629 _NN03 I \ i 39324- H₂ ~J , , !

•

I ,

,

i

! j I

l

... O. I 7 ,'-{ N 0 3 .,

,.

'\ . . 39330 _{H 2 0}

!

+---_._---1

/ V - 2.. FII-TF..q

,

_{la 9. S fiN0 3} 4260 H2 0

.

V-2 ,

t

_I " ". ::.:> t:~ I)

V-3

FIL TER r---.--~

-lt 6.46 4374-V-3 WASH

k---\ _{3.23 HNO,:;} 4-377 H 2 0 p- /2 V ':~4 FIL TER 306 HN0.3 39327 ti,? 0 r---~ \11 0 34- HN03 4-370 H2.0 V-4-WASH 3.9 33

°

Ha

°

<:-_._---I

/H

L D5 /

r

O_17HN~

.

êQJ.yent sur~rum F8

Half hour solvent c1rculat1on capacity. Volume: 215 cu ft.

Dirumeter

5

ft.

Lmgth: 10 ft.

Solvent circulation pump P-5

A pprox:ima te pre Bsure haad: 120 ps i ..., 480 ft. haad Duty: 15,280 Ibs/hr., density 36 Ibs/cu. ft. Efficiency: 0.4

HP

= 9.:3

Install 11 HP

Summary of eguipment reouiraments-f~ilash sec11Qn Vesse1 F-7: Diam. 2.5 ft.

Length 8 ft. Cyc1one: 10 cu ft/sec. Heater H-2 350 ft2 Condenser E-5: 800 ft2

Condensate drum F-8, Diam. 5 ft. Length 10 ft. Pump P-5 11 HP

53 gal/min.

6. Catalyst removal section

Cata1yst is decomposed Qy and disso1ved in a hot 10% nitr1c acid solut1on.

Cata1yst consumption 5.1 1bs/hr.

406

n

For their decomposition is required 14.~lbs/hr _HN03•

Generated are then 16.4 lbs/hr. dissolved nitrates and 3.9 lbs/hr. HOl. Schematic balance of acid and water

in

filtratlQP section

The polymer travels in a ~ lts quantity does not \ change during the operations, and is therefore not shown. Also, the salt

content is not shown, although the prime objective is the removal of

catalyst. The reason is that it is eaaier to picture the dilute acid flow. Tbe ratio solids: HN03 is everywhere equal, except for the fresh supply of HNO;. Since

~

solids leave the system with 109.; lbs HNO;, this ratio prevails in the entire system.

The mass balance is based on the following assumptions: a) Perfect mixing in the mixing ve ssel s •

Vessel F-9

b) Washing on filter reduces HN03 content of filter cake by

50%,

except in the lst wash.

c) The filter cake contains equal amounts of polymer and aqueous solution.

Operating temperature 2200F. Pressure : 17.5 psia

To bring polyethylene and dilute acid up to 2200F., 700 lbs/hr steam are required. In addition, 500 Ibs/hr. steam are used to strip the polymer of remaining solv.ent, thus giving total requirement of 1200 Ibs/hr. steam

( <

20 psia).

Residence time 1 hour, 70% filled Volume: 530 cu ft.

Diameter: 7 ft. Height: 14 ft.

Vessel has to be glass-lined because of HN0

39.

Stirrer: 2 propellers At cormnon N_{Re :}

Np

=

1 (Stirrer shaft off center or vessel baffled)

N = 120 rev

/min.

Diam. propeller: 2 ft. For 2 propellers required 2 HP.

Install 2.5 HP

Heat exchanger E-6

Bacause of highly corrosive nature of slurry and of filtrate, Karbate bas

to ba used.

lat aection: heat exchange

Slurry: 2202

-7

135°F. Filtrate: 100

-1

200°F.

U

=

100 BTU/ft2 hr. oF. Heat duty: 1,690,000 BTU/hr.

A

=

625 ft2 2nd aection: cooler

Slurry: 135

-1

10OOF. Cooling water 80 -~ 900F.

Take U

=

100 BTU/ft2 hr. oF. Heat dutyz 700,000 BTU/hr.

A = 226 ft2

Inatall 1000 ft2 for bath sections. Cooling 'Water: 8400 ga1/hr.

Pump

P-7

Assume head: 100 ft. Laad: 17,0001bs/hr. Efficiency: 0.5

HP = 2

Vessel F-IO

'J" IA ..

iL->

\..-'L

lJ

Operating temperature: 100°F. tÀ~...., 1(---Plastic lined or Raveg.

~

--

-

-Reeidence time: approx. 1 hour Diameter 6 rt.

Height z 12 rt.

Stirrer : identical to stirrer in F-9

Required 2 HP Install 2.5 HP

Vessel F-11

Operating tamperature' 100°F. Plastic lined or Haveg

Res1dence t1me : approx. 1 hour Diameter: 8 rt.

Height: 16 ft.

Stirrer: Diameter propeller 2.5 ft. 2 propellers on shart N = 100 rav./min. Required

HP

=

3.15 Install 4 HP Vsssel F-12 Entirely identical to F-ll Stirrers also identical Filters V-l to~

Nothing is known about the filterability of the product, except the statement 7): _ the material filters extraordinarlly quickly-.

41.

Because resemblance between paper pulp and the

L polyethylene slurry, a low filtration rate of the former is used to estimate the size of the required filters.

Filtration rata: 500 lbs/ft2 day Total polymer: '105,000 lbs/day Required filter area: 200 sq. ft.

Install 200 sq. ft. continuous rotary vacumn filters (Oliver), having

facilities for separate collection of filtrate and wash.

The filters should be covered so the filtration can take place under inert

--gas.

Because of the corrosive nature of filtrate and wash, all metal parts of the filters should he plastic coated.

Pumps P-8 and P-14 Assume he ad 50 ft. Load: 8800 Ibs/hr. Efficiency: 0.4 Required HP: 0.55 Install 0.75 HP

Haveg or plastic coated.

PumPS P-9 to P-13 Assume head 50 ft. Lead: 40,000 Ibs/hr. Efficiency: 0.6 Required HP ,: 1.32 Install 1.5 HP

The length of the conveyo-rs wUI depend on the

is obvious that they should be kept as short as possib1e, since both

investment and operating cost are approximate1y proportional to the length. The conveyors ahould be plastic coated, Which is possible since the wet polyethylene is probably not abrasive (~xy).

Estimate for

5

screw conveyors: Av. length 20 ft. Scope

Capacity 10,000 Ibs/hr. Total required HP: 5

Summary _{of eguipment} reQuirem~nts _{of catalyst} remov~l _section

Vassel F-9 Diam. 7 ft. Height 14 ft. Stirrer 2.5 HP

Exchanger E-6, Karbate, 1000 ft2 Pump P-7, H8.veg, 2 HP, .3.3 gal/min. Vessel F-lO, plastic lined

Diam. 6 ft. He ight 12 ft. Stirrer: 2~5 HP

G1ass-lined

2 Vesse1s F-11 and F-12, plastic 1ined Diam. 8 ft.

Height 16 ft.

4 Rotary vacuum filters, 200 sq. ft. per filter Metal parts plastic:. coated

2 Pumps P-8 and P-14 0.75 HP, 18 gal/min. Haveg or 'coated.

4 Pumps, P-9 to P-13: 1. 5 HP, 80 gal/min. Haveg or coated

7. Addition§..l ~guipne~

~

Ch01ce of dryer 1s not possible at the moment (see introduction). Load: 4400 Ibs H20/hr.

Estimated tbermal efficiency: 50% Heat input: 8,600,000 BTU/hr. If indirect dryer is used witb steam:

Steam 8,600 lbs/hr. Power: estimated 10 HP

Vibrating conveyor

From dryer to storage and loading boppers. For vertical displacement aspiral conveyor.

This could he equipped tor final drying and cooling of the product.

43.

Size depends on outlay of plant and on material bandling properties; none of this information is avallable.

Estimate for required power: 10 HP §:tora ge and loading ljopper§

Because of tbe low bulk density of tbe material, it vlll be advantageous to maintain only a minimum of storage space and sbip tbe material as soon as possible. If tb1s is done, as suggested, by railroad cars, a factor of uncertainty of Car supply is introduced. Therefore, it may be advisable to

have some trucks available for bulk transport.

When produets are made of several specifications, the minimum number of hoppers is three.

The storage space cannot be cut down indefinitely, since the running of the plant at low capacity increases the coat of the product.

The maximum size of a hopper is dictated by product propert1es, which are not known at the present. It certainly would be advisable ~ look into the possibility of using railroad hopper cars for storage.

Ir ~ll product is shipped in bulk by hopper ear, approximately 2t cars of 2,100 cu ft. each are loaded per day.

It is estimated that 7 days storage should ba provided, mostly to buffer fluctuations in sales. This would th en amount to approximately 40,OÓO cu ft storage space.

8. Hitric aç~covery section

No information is available which allows the evaluation of a tower design for this purpose, because of the very great relative volatility of water. However, the most recent information indicates that 2 theoretical stages will result in a practically HN03-free overhead.

Because of the corrosive nature of the fluid (dilute HN03 with Hel, at boiling point), a stoneware flanged packed tower should he used.

Feed : _{8411 1bs/hr. H20} 328 1bs/hr. HN03 100°F. Distillate 5506 IbS/hr. H20 1100 lbS/~ pressure ~

\

Ref1ux : Operating