Document MJbbdKj1wObZQape0Rgo9MwOz
DISTRIBUTION
BrockcvtUo R.JC. Schiataer JtbD nioa (2)
ri-- H. R. Calato* A. 1. HutU|l (2) J. M. Hrslop (2) X. G. Scbwoogorlo
Area Lake G. N. GrtlUafar
ALTC r. x. kium X. A. CoUiaga - L. A. Cbowtlor G. IX Loaimy
Project 7716-75 October 6. 1975
21486001
nTHWF"ftfr SXRV1CX REPORT 1
TRZmi9Q TESTS ON DIAMOND SHAMROCK'S FTC Uflf SIJURIES IN THE 4-INCH DEVE2JOPMXNT COLUMN a R. wsflimiMi
N *P* 4 PllifW IN M A
r
Project 771tjOctober 6, *'y7S
Stripping Tests oa Diamond Shamrock1* PVC Rea in Slurries in the 6-inch Diameter Development Column by C. R. Huddleston and C. J. O'Connor
ctTMMAnv
Diamond Shamrock, a potential licensee of B. F. Goodrich stripping technology, requested tests with three of their resin slurries to establish capability of the method and to provide stripped resin for their evaluation. The three resins were: a low molecular weight homopolymer with 0. 10 porosity, a high molecular weight homopolymer with 0. 22 porosity; and a vinyl acetate copolymer.
These resin slurries were stripped in the 6-inch countercurrent steam stripping column at Bracksvilla under various operating conditions and 5-gallon samples of stripped slurry frosn each run were sent to Dia mond Shamrock, Residual vinyl chloride data were also provided.
The vinyl acetate copolymer handled satisfactorily at 180*F under vacuum but at a lew stripping rate. Under less vacuum at 20(ff, there was significant agglomeration of the resin particles. Successful stripping at 180*1" would require a column with more trays or multipasses through the short column.
The turn humopolymsr resins handled satisfactorily at atmospheric pressure and at peig, hut there was some evidence of resin discoloration. Part of this may be mused by iron contamination from the steel shipping drums since the polyethylene liners were broken in all shipments, but it
! appears that the Dleaeaad Shamrock resins are more heat sensitive than other reelne that have been stripped in the column. Discoloration was avoided hi retests by going to a lower temperature on the feed slurry tank.
21486002
** id
C f
Project 7716 October 6, 1975
j OBJECTIVE
To demonstrate capability of c nutter cur rent column ateam stripping | of vinyl chlorida from Diamond Shanrock'a PVC raaln alurrica in the 6! inch development column to aaeiat -n aale of Goodrich atripping technology.
CONCLUSIONS
Diamond Shamrock1 a low molecular weight hoanopolymer reain can be reduced from 5000 ppm to lesa than 10 ppm in one paaa through the column at 8 peig if the feed la preheated to column operating temperature. When feed waa heated only to i35- 150*F, the vinyl chloride waa reduced to IS ppm in one paaa. At atmoepheric preeeure. the reain waa etripped to 20 ppm in one paaa with the higher feed temperature but only to ISO ppm at the lower feed temperature. Adequate atripping can be attained by ueing a column with more traya.
The high molecular weight homapolymer waa readily etripped to lees than 10 ppm at atmoepheric preeeure even with low preheat :o the feed.
The vinyl acetate copolymer waa handled eatiafactorily without agglomeration under vacuum at 180*1", but the atripping rate waa alow. It ie projected that 33 traya would he required to reduce the vinyl chloride to leas than 10 ppm at 180*F. Stripping rate was doubled at 200*F, but there was significant particle agglomeration and plugging of the downcomera.
Stripping rates of Diamond Shamrock hemopolymare are comperable to those of B. F. Goodrich realms having similar molecular weight and poroeity.
INTRODUCTION
In 1974, B. F. Goodrich Chemical developed atripping techn* logy for
ramoval of vinyl chloride from PVC reein slurries by counter*, urr'nt stem-
in perforated tray colamas. This devalopmant work waa doi.<. n 6-ir a
8-tray column at Brackavilla and in the 30-inch, 17-tray column M
.i
Lake Geoa East. Results from the development were used for design of
21486003
BFG14360
oO-inch, 20-tray colui.uis which art now bing installed at 5 BFG plants in the USA. Goodrich has offered the stripping technology for sale to the industry and several potential purchasers have visited Geon Cast to observe the 30-inch column in operation. Most of them have expressed an interest in having samples of their product stripped in the 6-inch column to demonstrate stripping capability and to provide stripped sam ples of their products for evaluation.
Diamond Shamrock sent SS-gallon drum samples of three of their slurries for testing. This report describes the tests and presents results.
RESULTS AND DISCUSSION
1. Equipment
The 6-inch column Is mads up of stainless steel, in line "bull's eye" sight glass sections obtained from the Ernst Gage Company. Perforated plate trays are installed betareen sections. The trays are stainless steel and are fitted arith a 1 -inch stainless steel tube to serve as a downcomsr and a wier to hold lfc" depth on the tray. This represents a holdup volume about . 1734 gallons psr tray and 1.3172 gallons total. Each tray has 17 holes i" diameter on a ij" equilateral triangular spacing. Open area in the tray is about 1%, About SO pounds per hour of steam is requirsd to support the slurry on tht trays arithoet allowing excessive weeping. With the limited downcoiner capecity, this results la a higher steam to resin requirement than needed for adequate stripping, la production equipment. 0. S pounds steam per pound of resin or less has been adequate.
is added below the bottom tray at a rate controlled by manual setting of the salve and pressure control. Effluent from the bottom tray
Feed slurry Is bentsd in an agitated, closed slurry tank which is fitted with a circulating loop. Pressure in this tank is maintained several pounds higher then the column pressure. Feed to the column is provided by a flush mounted valve in the circulating loop. The vulva la activated by ! a pulse timer to control feed rats. For smoothsst operation, the feed slurry is heated to near the same temperature as the column operating , point. Feed significantly lower in temperature casuses some stoam
a 's rwiwtvo iw o a
21486004
* e
/ t
1
Projc*.i 7 7 16-75 -4- October 6. 1975
condanaatlon and a temporary upaet with each pulae. Aa feed enter tba column, thara ia a flaah above the top tray ao thia tray doaa not receiva tba full vinyl chlorlda load aa maaaurad in tha cold alurry. A mora raaliatic valua la tha vinyl chlorlda laval maaaurad In a hot faad aampla that ia flaahed to tha atmoaphara. Thia technique waa uaad and both hot and cold faad valuaa ara raportad.
Samplaa wara alao taken from traya 3, 6 and 8 and from tha column dlacharge. Sampia tapa ara fluah with tha tray aurfaca.
Aftar rathar aavara diacoloration in tha raain during tha fir at run with tha low molacular waight homopolymar whan faad waa haatad to column opa rating tamparatura, haat on tha faad alurry tank waa reduced to 13S-1S0*F for tha remaining runa. To companaata for ataam condensetion on tha top tray, aomawhat higher ataam rataa wara uaad. Steam rata for calculating ataam/raain ratio a waa cone Ida rad aa that amount which tha coadeneer received.
1L Analytical
A. Vinyl Chloride ldeaeurcmaat
Vinyl chloride content of the aampiea waa determined at Avon Lake Plant Servicea Laboratory by Dot Lewia and John Whitney. Samplaa were aent to thia lab ia tightly aealad, 4-ounce glaaa bottlea with electrical tape wrapped around the cap. Bottle contenta were quickly realurried and filtered ia a Buchner haanal to remove moat of the water. Samplaa of tha wet cake were checked for Nil aolida content and for vinyl chloride con tent by the head apace method gee chromatography unit. Valuaa ware reported aa ppm vinyl chloride ia dry PVC reaia by weight.
B. Reaia Properties
hamrock reported the following data for the three prod-
21486005
>. Mtmnm im u ..
b i cc:>o< liwwli a On*toe*<M C*>*
BFG14362
Project 77 l o- ' -5- October 6, 1^7.
Diamond Shamrock Rssins
Drum Number Resin Type
3 A, B Homopoiymer
Inherent Viscosity Porosity % Retained on 100 mesh % Through 200 mesh Residual Vinyl Chloride
0. 7 0. 1 20 8 3,000-10,000 ppm
4, 4B Homopoiymer
1. 35 0.22 47 2 6,300 ppm
5 Vinyl Acetati
Copolymer
(14% Vinyl Acetate) 0. 59 --
46 trace 3, 000 ppm
11L Column Operation
Before adding feed to the column, it ia allowed to ranch equilibrium with the atmoaphere with steam flow at the desired rata. A portion of the steaum condenses because of heat loss and the condensate discharges from the bottom of the column. The condensate rate is measured. Steam pass ing through the column is condensed and the rate measured. When lead is started, new measurements are taken. The effluent rata less the original condensate is the feed rate. Corrections to this value are needed if either condensation or evaporation results from the entering feed. This correc tion is determined by the difference ia steam condensate from the con denser before and after feed is started.
Residence time in the columa wue estimated tray by tray, assuming that the condensate occurs equally on each tray. Volumes of the feed plus condensate were calculated with consideration of densities of the compo nents. True residence time is somewhat lass than the calculated values because contents on the troys are frothed by the steam. Use of the resi dence times for relative comparisons should be acceptable.
After eeffli leel time for steady state operation to bo established, samples were taken from trays 3, 6 and 8 and from the effluent. Five gal lons of the stripped slurry from each run were collected and sent to Dia mond Shamrock for their evaluation.
21486006
a-...- -
...,
non-rest s > MHifVtfli m o a
- -b--F--G---C--C---'S--'-O--* Dswtos^-t \.tw
Project 7716-'-! -6- October 6, 1975
IV. Operating Data ^nd Analytical Retain
A. Low Molecular Weight Homopolymer
Drum 3B wae selected for testing since it contained more vinyl chlo ride than 3A. The contents of the drum were reslurried and loaded into the feed tank at 33.4% total solids. Slurry was received in a steel drum with thin polyethylene liners which leaked some and were completely destroyed by the agitator during loading to the feed tank. As a result, there was some contamination with iron. The slurry was heated to 220*F in the feed tank and samples showed significant resin discoloration before feeding to the column. Discoloration was not increased by passing through the column, but it did increase with time in the feed tank.
Runs were made at 8 peig, 235*F and two feed rates. Vinyl chloride level was reduced to less than 10 ppm in each run. A third run was made at 2.5 psig. 22i*F and vinyl chloride was reduced to 18 ppm under these conditions. Data from these runs are shown in Table L Figure 1 shows log VCM content plotted against tray number and Figure 1A shows the same values plotted against residence time, h both cases, the effluent values were considered as occurring after one additional tray.
Because of the discoloration* it was decided to make similar runs with drum 3A hut to heat the feed to a leaser temperature, about 135*F. Under these conditions, vinyl chloride was reduced to 16 ppm at 8 psig, 234*F at a normal feed rate hat only to 72 ppm under the same conditions at a very fast feed rate. At l.TS psig* 216*F the vinyl chloride was reduced to 148 ppm at a normal food rata. With feed at 135*F, it is pro jected that 13 trays woald he needed to reach 10 ppm at 8 peig and the fast feed rate and 10 trays at the normal rate. At 1.75 psig, 17 trays would be needed.
The contents of drum 3A also got some cootsmination from the drum. Diacoioeattoa was net evident ia the stripped resin, hat the serum had some illsasleirattna. possibly from coatamiaatioa by other use of the column.
Pi-- foe the three runs are shown la Tahla 1 sad Figures 2 and 2A.
B. ^nleeuinv WeiehS Homopolymer
The contents of drum 4 wore reslurried and loaded late the feed taak at 28.5% solids. Boms coatsmination from the drum occurred hecause of broken liners. When leaded to die slmrry taak, the serum was essentially
T T
21486007
-7-
Project Octobe r >, ! > ,'S
iid nut appear discolored. Color ot tha urum altar atrlpping wn about the same as that of tha warm faad.
Two rune wara mada at 8 psig, 235*F at diffarant feed ratea and one run at 2 paig, 217*F. In all thraa runa, tha vinyl chloride content waa reduced to 5 ppm or laas. Data for tha thraa runa ara ahown in Table 11 and Figuraa 3 and 3A.
Contact with Diamond Shamrock indicated that the serum discolorsj tion waa unexpected so they aant another drum of material labeled 4B for
additional teats. This sample was in a fiber drum with polyethylene liners. Contents of the drum ware loaded to tha faad tank at 30. 2% total solids.
Two runs wore mada at u psig, 235*F at diffarant faad rates and one run at 1. 75 psig, 216*F. In all three runs, the vinyl chloride content was less than 0. 6 ppm. Tha faad contained only 684 ppm compared to 4038 ppm for drum 4, so the lower levels ara as expected. Data for the three runs are shown in Table 11 and Figures 6 and 6A.
There was no discoloration in serum or resin sc it is concluded that the color formation in drum 4 was caused by trace contamination left in the feed tank when used in other service.
C. Vinyl Acetate Copolymer
The contents of drum S wara re slurried and loaded into the feed tank at 37% total solids. Some contamination from the drum occurred. Slurry was heated to 145-150*F and fed to the column which eras under vacuum to keep temperature low. Two runs arere made at a 13-inch vacuum, 180F and different feed rates. Another run was made at a 6-inch vacuum. 200*F. The resin handled with no difficulty at 180*F, hut there was sig nificant agglomeration of particles at 200*F. The agglomerates plugged the downcomers which have fairly narrow clearance from the plates. At 180*F vinyl chloride content was reduced to about 400 ppm at the faster rats and M8 ppm at the slower rate. From the data it is estimated that 30 to 40 trays would be required to reduce the vinyl chloride content to 10 ppm at 180*F. This could be achieved by two passes through a 20-tray column. At 20<FF, ig trays would be required, but this temperature is apparently too severe for the copolymer resin psrticlss.
***"* '-- thres runs srs shown in Tabls 111 and Flguras 4 and 4A.
-* !
i
Projsc' .'i *.-75 October 1975
Dtscussior of Results
From Figures 1, 2, 3, 4 aad 6, aquations of the typ C * Cos**** ware darlvad. C rapraaaata th* pradictad vinyl chlorida cootant la ppm at tray N whaa starting with a faad containing CD ppm. Tha slopa of tha line ia -A, tnd thia valu* caa ba coasldarad a rata of (tripping factor. Similarly, from Figuraa 1A, 2A, 3A, 4A aad (*A, aquations of tha typa
C = CQa* Bt wara darlvad. la thasa aquations -B ia tha slopa of tha line and t ia tha reaideaca time ia minutes. A aad B ara related by the squa-
tion:
B*
A () t (for 8 trays)
By using these equations, the number of trays required to strip slurry from oae level of vinyl chloride to any other level caa be calculated. In using these aquations, CQ should be taken as the vinyl chlorida content actually received by the top tray. When feed enters the column, there la a significant flash of vinyl chloride as the slurry falls to tha top tray. The degree of this flash ia dependent on the temperature of the feed. For these teats the Co or hot feed value tans determined by allowing a sample of the heated feed to flash to tha atmosphere. It became apparent during these tests that ia moat cases the hot feed vinyl chloride content decreased, apparently from leaks ia the feed tank. For later runs, a hot feed sample was taken during the actual nut. This was done for drums 3A, 4B and 5 but not for 3B aad 4. For drums SB aad 4, the hot feed content was esti mated by extrapolation of the tray data.
When two feed rates are meed aader the same operating conditions, data from tha two ruas give a lino with essentially the same slope when plotted against tray somber. This indicates that the number of trays is the controlling factor and not the residence time. Similar data wars obtained during development work with the 30-inch, 17-tray column. Obviously sosns sddmam residence time is required for diffusion of the vinyl chlo ride from the particles. This is demonstrated in the fast feed rats run wMh drum 3A where the slope of the line la significantly different from that at dtt slew feed rats. Tha very fast feed rate apparently did not allow euBtoieet thne for monomer diffusion.
In many of tha plots, the vinyl chloride content as measured for trey 3 falls well below the line. This caa be explained by considering the operating characteristics of the column. Feed is added by a pulse valve which is opea for 1 second, then dosed for 44 or 99 seconds, depending on feed rate desired. The ties of tha pals# depends on pressure differential
p r . .XV'f'Oi
21486009
BFG14366
Project 7716*75 -9- October 6, 1975
! between th columa and tha feed pump discharge. Am m result, th fir at tray receives slug of feed which gradually works it* way down th
' column. Whea tha faad ia aigaificaatly coldar than operating tamparatura 1 thara 1* conatdarabla frothing and holdup on tha top tray and oftan tha | second tray to a leaser extent as the feed heats to operating temperature.
Often the trays barely recover before the next pulse. This results in more residence time on the top two trays than that estimated from the normal level. In addition, the heating of the feed removes appreciably more vinyl chloride than would be achieved if the feed were already at the operating temperature. These factors cause the analytical values at tray 3 to be below the operating line, la production equipment where a steady feed rate is achieved and the feed can be heated ia line to the columa, this problem is eliminated. Ideally a steady feed should be provided for the development column but it is extremely difficult to handle the low volumes of fast settl ing slurry. The pulse feed was chosen as a compromise sad since data from the small column correlate very well with (hose from the 30-inch column, the pulse feed method is considered adequate for development test*.
The pulse effect is dampened out by the first two trays aad flow dis charge from trays further down the column is fairly steady with no indica tion of e pulse ia the offleedl from the bottom tray.
Rate factors from the ten figures are listed for comparison.
N h* ih
00
3
H*
!O
4 r
eu ooooso-t tenet* IS D*v^or't hm*
-10-
Hrc J
lo-75
October b, 1975
Polymer
Stripping
Slurry Rate
Temperature. *F lbs. /hr.
A
-B
Low Molecular Weight Homopolymer Drum 3B
239 235 221
99.5 69.8 76. 1
0. 674 0. 672 0. 377
0. 787 0.566 0. 360
Drum 3A II II
High h^olecular Weight Homopolymer Drum 4
234 234 216
239 235 21?
162.9 74.3 85.4
no. i
100.6 108.7
0.473 0. 598 0. 343
0. 935 0. 592 0. 389
0.872 0.868 0. 589
1.228 0. 978 0.813
Drum 4B** II II
23$
73.5
1. 151 1.093
239
58.0
1. 151 0. 984
216
65.3
0. 895 0.876
Vinyl Acetate Copolymer Drum S
II
ISO 180 200
97.5 85.5 101.8
0. 119 0. 147 0. 266
0. 125 0. 150 0.302
This resin shows much faster rate on a tray basis than Drum 4.
1 1 1
The A rate factors ara plotted against stripping temperstars in Fig ure S. This plot shove the high molecular weight homopolymer with high porosity to bo much faster stripping then the low molecular weight lower porosity resin. To obtnin comparable rates, the low molecular weight resin must be stripped at about 10*F higher temperature. Stripping rate lor the vinyl acetate copolymer is intermediate between the two but at a safe temperature of IUT the rate is very slow, requiring more trays or
isos through tbs shorter column.
rates are established, the number of trays needed to of vinyl chloride from Co to C can be estimated by:
N
21486011
IU rs 'set a ' *" > w *
BFG14368
Project
.5
11- October 0, . < o
VI. Comnarison of Diamond Shamrock Stripping Rates with B. F. Goodrich and Air Products' Rssias
RiU factors derived for sovoral BTC resins from sarllsr data in the 6-inch and 30-inch columns and for Air Products' rssias la tha 6-inch column ars listed for comparison la the following table:
Basin
103EPF76 103EPF76 105EP 105 EP lOZEPn 102EP7S
0. u 0. 12 0.08 0.0s 0.30 0.30
Stripping Tamper-
*r
Bata factors <A) 30-inch
6-inch column column
214-217
0.415*
0.365*
231-237
0.530*
0. 486*
213-220 _ 0.118*
223-237
0.165*
216-210
0.525*
236-237 0.820*
"average of several values
Homopolymer 0.85 L V.
Propylene modified Basin, 0.627 I.V.
Vinyl Acetate Copolymer, 16.4% Vinyl Acetate, 0.344 L V.
Homopolymer 0.7LV.
11 iiiijilinm 1.39 L V.
Vinyl Alettes Copolymer, 14% Vinyl Acetate, 0.39 L V.
Air Products
0.10
220
0.10
236
.093 217-210
.039 .039 .039
100 200 217
niammil 91 216-221
0.1 234-239
0.22 0.22
216-217 239
100 200
0.304 0.602
0.491
0.148 0.285 0.415
0. 360, 0. 389 0.590. 0.673
0.589, 0.895** 0.870, 1.151**
0.147 0.264
-
-
The reaia in Drum 4B stripped much tester than that In Drum 4 when mlunted on a trap basis.
jjv-rees-a s F. t"aMrr-D--- u--S *-
B f OOOWKCH Smetfc *d 0*vio--; (. *
148601Z
1
21486013
TT
BFG14370
7
ft (V O C C ft K *0 O IV I0IO M B * c ihC M
BFG14372
21486015
Ilk
'rrrcjTri ' Li * TlT-iirruu n: H l : r'T :.T : .it: i i i
--14
rreV
i-r f-f
35
jRfr^3W;
' ........................................
r
* T"
*- -
*5
-ir/nm
S T O M frT Z
TTpr !t
i`-g
=4n iirhif?. - SflirirH
f_rr?- f - ;t3 r. .::m
s+-R-
-H-
... |.
| : r" I ! t
ii
i44&
i; ` jjj.
w i4intru_
r; t!ri+i
lI
i
r :t
INCH
BFG14373
S 0 - iftin ICTSOCN 9JIAPH A tlt
M M i>cO O A iT H M C
0V O V U 10 P lV tliQ M l H P
IN C H
CUOCNC
CO.
** w * *
BFG14374
JU1090W Z
r, -r t --
I:
H --
ZA17-
f i*vr
T'TTT^i iTiTvr. ft:t :
a&'*
f
: yb**f*/f **>A>wm
-- L*4pmsj#\h*d Wt :: !:i-j/r'W'rtffy*
r
3X
+T
3^=1 TT xj
5
r- --
ffrtftrH
rrrr 4 -rS4.t4T-J iwt?
ttrt
i&
r+r
:_r.
; TI:
Tit
--i--
*4 ii
--4 M-4
ttn
-I-; -4-rr4LI--J
- J
:tj `I
--r-_.
riT
; B. P.fcOCPHCH
Re**'1"** C*
BFG14375
21486018
;TrTJ r~r
Ft'i&c S
mrrn t : r: tjt* rrrr^Tr . - r rj-rrrr.; r -
___________ WM
4l . I
I :' i ;
x-ifry pfrt-f
km*--
O-rfftii***'
tuaci
l4a-ci*o >*<* omapm At
.tinil
I-t:-
-4"r!+ -i--4-
M
i
: : i t.
4t-
ji <i i j fcH V li f=F
fix: ut-.jrt;:
-:~u
Ui:i
i:; -1: t;
T
r.3:
S*CH Cc-iW
BFG14376
21486019
F,, l^ U-1r9t*
-
3A
^rmrnTr^TinTirf': r:rrr^TT-!TTiTT
;n:
V/'*
Hir i l:
i; W*i
:I
|rH-iri -Hi* JHi; !!
- Miff.
Tj>t**j
rrrrrrr-
-Hr I
::A::~
I- \-
Tr A--
i: : .. -
"t: LI r
3=EB
--t-- '
. 4*--
15:
Jr^~----- H :i
r'r]
-- i
5 I iiu^ ;
T!~ H :j
-urlr -a: rt
i:
Ttrr 7+
-n
i?
ttl:
j.r
[*- r&f::Tri
f-- rH---~ r rf+ I - ' : j 1 ' +-{
! I ri
Hi ft
44
! il i;
Ht -- - -f r-r - *
::t'.:rrr
-4-r+l:
it
~rTr --':'1
MorA'
BFG14377
ZZOVBVTZ
BFG14379
21486023
R19Qr(H
BFG14382
-25-
PtOject 7716
VII. Idjwdai
A. Introduction
Dlaaond Shanrock requested additional taata with chair vinyl acatata copolymer raaln which waa ldantlflad aa Drua 3 la tha original raport. Speci fically they requested that It ba run at a alow faad rata to allow at laaat 15 ainutaa ratantIon tlae la tha coluan. To provlda aora aargln of safaty In pravanting agglomeration and colvam plugging, chaaa taata wara aada at 170*P lnataad of 180*7. Extrapolatlona of tha data wara aada to pradlct raaulta for 180*P baaad on tha factors aatabllahad with tha first drua of notarial.
Soma changas In procadura had baon aada slnca tha aarllar run. A aaall atraaa of watar was added to tha stean faad to ranova suparhaat and pravant drying on tha lowar trays which had baan a savors problaa at low faad ratas. Whan drying or concantratlon occurrad, thara was consldarabla aggloaaratlon and plugging of tha downcoaers. Adding tha watar waa affactiva In ninlaising tha problaa but lnjaction of watar through tha saapla taps was nacassary savaral tiaas at tha vary slow faad rata.
Tha analytical procadura was changad to allow auch longar equilibration tlao in tha head spaca aatbod. Studlas by Vhltnay had shown that tha vinyl acatata eopelynars require auch longar ehan hoaopolyaars to raach equlllbrlua at 90*C. Insufficient tlae resulted In low and erratic values for tha earlier teats. The secoad drua of resin had a auch higher level of vinyl chloride than tha first dna.
Diaaond Shaarock waa interaatad la aothods to obtain 10 ppa in a 20 tray colvan. Projectlooaare shewn oa potaatlal aathods for ranching this goal.
1. Ooaratina Data and Analytical heeulta
Stripping taata wars aada oa levaBbar 13, 1973. Tha slurry, as loadad to tha faad tank, showed a solids esataat of 31.IX aad <360 ppa vinyl chloride.
To obtain tha alower faad rates desired, tha pulse tlaar was sat to opaa tha faad valve far a fraction of oaa secoad. Throa runs, at widely different faad ratas, wara aada without appreciable difficulty. Table 17 shows data for tha throa runs.
/
rg.mi-A *i MhbTVtwuM,
CO ft
8 f vOOOWCH I Drvtljpi.r-1
-26-
Pru <
'14-75
Table IV
Co1 vji Opgrating Data, Dlsnood Shanroch's Vinyl Acetate Copolyeer, Second Drum
Run Nunber Colunn Vacuun Inches Hg
Botton Tonperature, *F Feed, pounds/hr Seconds betweau pulses Resin, pounds/hr Stean, pounds/hr Stsan/Resln Ratio Feed Teaperature, *F RVCM, ppn. Cold Feed RVCM, ppn. Hot Feed Flashed RVCM, ppn, tray 3
" " tray 6 " " tray N " Effluent Residence Tina, aioutea
tray 3 tray 0 tray 0 Effluent
1112A 17 170 24.9 40 7.79 36.4 4.67 114 6360 4326 2302 * 1491 1213 1230
10.35 20.27 26.65 29.81
1112B 17 170 51.8 20 16.21 33.6 2.07 116 6360 3786 2557 2024 1556 1377
4.90 9.69 12.84 14.40
1112C 17 170 137.3 10 42.97 32.8 0.76 116 6360 4076 2769 2431 2003 1770
1.94 3.87 5.15 5.79
The analyst had difficulty and wan not abla to report a specific value. Results ware listed as nore than the values shown.
C. Discussion of tasulta
The data are plotted as log vinyl chloride content against tray nunber In Figure 7 and against residence tins la alautas la Figure 7A. Fran these plots rate factors A and 1 wart derived as discussed earlier in the report. These data show that increasing the residence tlaa by a factor of 5.15 Increased the tray stripping rata by only l.M. (Congare runs k and C) Canparlng runs A and B, Increasing residence tins by a factor of 2.07 Increased the tray rate by 1.2 and rsnparlsg runs B and C, increasing residence tins by a factor of 2.49 Increased trey rata by 1.3B. These data suggest a relationship:
z 0.31; far A -
ta tine
rots rate 2
the three na d solving for x shows: For A - C, 0.23 and for 1 - C, x 0.3d; average x 0.307.
kZ098VfZ
are no-* m mmutu m ut *.
B* tmrd> nS
C*w
BFG14384
./
-27M rata factors froo Figures 7 and 7A ars:
FroJ*'
1 |
Rata. pov>ds/hour
24.9 51.8 137.3
-A Factor .1477
.1231 .0889
-B Factor ".044*
.0763 .1339
Trays Required* 41
49 67
To roduco vinyl ehlorido fro* 4000 ppa to 10 ppa
Fro* tha Japandanca of rata on t*pratura aa ahown in Figure 5, ch*
projactad rata faetora vara aaelaatad for 180*? at tha thraa faad rataa and traya naadad to raduea vinyl chiorIda fro* 4000 to 10 pp* vara estimated:
Slow faad Nadiiai faad Faat faad
k factor at 180*?
57155
0.166
0.120
Trays Needed ------
36 30
No problaa la aatleipatad for thla raala at 180*?, but thara vma appreciabla aggln--ration la tbs aarllar run at 200*? so 180*? is naar tha safa operating taaparatura liait.
a
D. Ooaratloo of a 20 Tray Coin to Obtain Laaa than 10 pp* Hastdual Vinyl Chlorldo
Dlassond reguaatad projections on bow a 20 tray coluan could ba asad to roach
10 pp*. Savoral projections vara aada baaad oa tha davalopaaat colon data.
Thara ara ao diract coaparlaoa data batvaaa production colons and tha dcvalop-
asat colon o* tray rata factor* yat, but comparison batvaaa tha 30 inch colu*n
and tbs inch coltan sbova slallar k factor*. Initial data fro* tha now 60
inch plaat col*aa ara la tha rang* pradlctad by tha 30 inch colu*n ao tha
rata factors |*wnt*d la tb* davalopaaat teats ara sapactad to ba conparabla
to thoaa of tb* prodoctio* colaaa*. Savoral option* salat for reaching 10 pp*
la a 20 tray colaaa. Thao* ara dlscuasod.
M
_ Ml
1. Laaar Tiny]
Cootant la Toad to tb* First Tray
Tbl* ca* b* auaapllabad by aora batch atripplag bofora foodlag to
tb* ala and/or baatlag tb* faad to eon** oparatlag taaparatura la lias to tha *sla*a With tha higher food taaparatura thsre will ba a larger
b foval of vlayl cblorfl* aa tb* slurry oatars tb* coluan. that *ih a cold faad of 6360 pp* tb* level vaa reduced to the raaga ti by 1--^'-g only to Hi*?. The level oould ba reduced further by
fe* U0*? la llaa vltbaut adding algalflcaat beat history to tb* table ? aha** that even vitb tha slow faad rata, a vlayl chloride
at 333 vould ba required to tb* first tray to rea'-S 10 ppa.
ragulrad can ba calculated byt C, 10/a "4B
sg-rstf.* * NiaintMUii
fe-edfv*' rVvf1 p*#-
BFG14385
Projsct 77\> 24-
2. Hake Two Passes Through the Column
Another method to achieve the low vinyl chloride level desired would
be to eeke two peases through the eolian. In this case the equation would
be c m
-A2M and at the slow feed rate a level of 28,640 ppm
could be accepted and reduced to 10 ppa la two paeeee. Soae values are shown la Table V.
3. Recycle Part of the Slurry
In this node of operation, a portion of the slurry would be taken froa the bottom of the coluan aa stripped product while the rasialnder would be returned and aland with freeh fend for another paae. This nethod nay be preferable to making two aeparata paaaea since it would not require extra holding tanka. The equation for predicting feed concentration and recycle fraction is:
10 ppa - [CyX,, 10 (1 - Xj)] a -A>
When Cp la the vinyl chloride concentration In the feed to the top tray and Lrls the fraction of new feed. This equation can be used by setting either the feed concentration or the feed fraction and calculating the other. With a feed containing 4000 ppa the freeh feed fraction Is .132 at tha slow rate and at 1000 ppa the feed fraction.le .33. Soae values are shown in Table V.
14860Z9
1
ea
Project 771 fi -29
Table V
Projection* for a 20 Ttst Colwa at 180*F
KVCM to Plrat Triv 32000 16000 8000 4000 2000
Peat Feed Rate - A Factor 0. 120
V One pace RVCH Two paaaaa
2903 293
1452 132
726 363 66 33
Recycle Nod*
MX recycle KVCM 75X recycle KVCM
SOX recycle KVOI 211 recycle KVCM
59 726
1432 217K
30 364 728 1009
13 8 182 91
363 182 344 271
181 16
3 46 91 136
?
1
2
Rate - A Factor 0 .166
KVOI Oh paaa
1137 378 289 143 72
KVCM IWo pum
41 21 10 5 2.6
Recycle nod*
MX recycle KVCM
23
12
6 3.2 1.5
1 7SX recycle KVOI
289
143
72 36 18
1 SOX recycle KVCM
379
290
143
72
36
2SX recycle KVOI
868
434
217 108
54
Slow Peed Rate - A Factor 0 .199
KVCM Oh paee KVOI IWo paeeee lecycle aedo
MX recycle KVCM 7SX recycle KVCM SOX recycle KVCM 23X recycle MVOI
398 11
12 130 299 4a
299 149 73 37 6 3 1.4 0.7
6 3.2 1.7 0.93
73 37.3
19 9.5
130 73 37 19
224 112 36 28
1000 500
91 45 84
2.7 1.8 23 12 46 23 68 34
36 1.3
1.1 9.3 18.3 27.2
18 0.65
0.7 4.8 9.2 13.6
19 9 .35 0.17
0.6 0.37 4.8 2.5 9.4 4.8
14 7.1
.N MWtlO IK ..<
26 1
liwinfc o*4 0tvtc0c*' vc-f
0E098FTZ
. Jp'jv j i-
a?lti
i*r*
"n /
^. .
~7i //*>> & * Vi *u / /f - ^ ^ ^
f eiibfMtir; Sec^J >*;' -
- ~/br/
/'M"c,`
X- -- NjlZ-TT
1-. i- -
..J...... 4. I
-f--- -------
ir
-t---------r1--
7? - /f#y
12=
3
I :'T
J 4--
.- r- -,1--
!'!: 4-
.4 -
Y I f J;
L:. j.JO^rt? ` f- '* - t?*-** L r* i:-~" TTrt;7..U',... .*t .:. t"7.*--jni--- , *:rf i*.l-tL-"7( " t^bTT
i a*
L.i.
i# ;-. T1 '.
)-/*. />t'jXvTt
~ : .'. i.' '
4-------
r::r.nr. :* "7TT
r
;o"
21486032
BFG14389