Professional Documents
Culture Documents
Bleaching TheorY and Practice
Bleaching TheorY and Practice
-
T.K. ll{ag
Resrrarch Cent.er, (lanadil Packers Ini'., Toronto. C"ntario M6N 1K1' Canada
Introduction
Crrde C)il
In recent years, the process of bleaching hls.inre-a
ieived considerable attention. This has resulted
mr-rch better appreciation of [he fact that the process
affects oils beyond the removal of colored material
and that, in addition. it itifluences the stabilityof ed-
il,te oit products to a decisive extent. Research and rt Extnction Plant
[Vater DegummiBg ( !
li)?
108 T.K. Mog
aor . tt'
f.lr.htdral 4!lt rld
t.t.t-
- {O'-lo{frte-
Ool.iadnl
frrrr i6Pl g3r.;1 1o
l30r-t0lfrf0'
-
ftl.rd3
lO' rtf
t.larltraa Callr. 6*.1
(rrrrlly lyarrtodl- L.r.t ClrF-t
l:l.rl.t.r Clrl..l
ll.l Clrr. o
t
E' probably !e required for red color and chlorophyll
removal. This indicates that for phospholipids, as fbr
Er
b metal removal, process steps other than bleaching
g. \ should be considered alkali-refining, or acid pre-
treatment,pr utilizing -an adsorbent much more effi-
3 \ --\- cient in the removal of these. \
\ The effect of this latter approach is shour in a
a \-r- \ siudy ciue to Pryor et al. (5). Figure g shows the rela-
tivel_v better efficiency of a synthetic silica in phos-
pholipid removai com.pared to an acid-acrivared
2A tI,
Ctry Dorlgc (fl
Wrtr tlcgm.d
Fig.5. Bieached oil lc'ribond red isotherrns fcr scl,bean and
canoia oii (l). .I a r- o'l€o
l,
._ _ o, io6
C.-a.et?a0r-)
^ a.a
ll t a tao
5."
I
I
aa
. toa \'
-Ee,
-*
a---ttr !l
trr
a.'
oq
e a.t
r a a
E
c, a
o
o 0 1 lo ll
a.tti.i ' cray Do'qg. ttl '
.9 (, ta rea)
c
()
.n \'.
_ \ \. _*r".
Fig. 7.
(1).
Bleached oil iron adsorption isotherms for soltrean oil
1\\
D
23 !l, It
ta
Clry Dolre.(t)
l,
la r -- &'tCO
Fig. 6.
Bleached oil chlorophyll adsorption isotherms for ll a\ o__e,!o6
soybean and canola oil (1). ta
\
It
?,,
tl,pical clay dosages used in bleaching these two oils.
grt
rc
\
!l
Chlorophylloid compounds must be removed to the !r
qa
\
above mentioned lorv concentration of 50 micro- Ot
E'
grams. kg or lolver to avoid rapid oxidation of rhe oil Aa-
in rhe presence of lighr (7). a \ '!-
a
Figure 7 deals.,vith iron adsorption. Iron is the ma- I
jor pro-oxidant metal to be removed froni oils in the a
course of refinery processing. It can be seen that il I
a
irc;n is reiatively high., about 0.8 miiiigran.s,'kg. clay ta .'! u
riosages in ihe order of lJi. are required to io.*'ei it to Crrr Dor{. {t}
tl..e Cesired le','el of belou' 0.i ri-,iiiigramsik{. ?his is
n-.uch in e\cess of the cia1.' ciosage needed rii achier,u Fig. 8. l-iitacl'r+rd orl h r;splr r-, ii pi <i aa1 sll-pt i()n rsc t h i, I.'rs tr i r
adequare red ccior and chlorophl,ll removal. evep.
1-.
il;i i.i ).
_sc.r'i*,.ari
Bltruthirtg * Thetrg atd Practk:e lll
TABLE 3
Chlorophyll Hydrolysis in Bleaching of Soybean Oil (0.14% FFA)'
L,oeding
w%xl0 30
No Pretrcat Bleaching
Bleached
Ciay
dosage Temp. Time oil
.C min I-FA ''
20
1.5 r25 3 0.30
.l 0.47
r0
,-) 125 3
i.e 170 3 rl.6'l
.-f rl I ?t! .) (i.6r;
t.5 125 30 0.25
rJ 5r0t5:0 2{i t:;) ;-ii_l
'.).:7
Ctl croph-t'll Conc- PPm l.a) I /t, ,,I,I ( t l).1
') il r :il 3r) (l. ir'-l
Pig. 1ii. Eil!'t'ri si a sl.'nthelic silica prerrealment +rr cilr'ro
l:iri ii is,rtherrns irr :r<:ti:;att'tl t'la-r' hiearhing (i ;. "l-rrpublished daia - T.X. .Hal
112 T.K. Mag
to
t
a
.E 7
!
IE) I
-t
c t
o
a {
!x
o !
a
A
,'
\
t"
G.
t,
l.lt
Cr t6o
t.t o-- Gr106
E
q. trr
o
5i
-: t.
6,!
og
o
o crc
F
GT
e.s
ct
CIay Dosegrc {s}
Fig. 12. Tticof:her,-rl :e i.rlearheri, deoiiori;l.ed sur..bean oil.
effect of differeni cia'; dr-isages ( i ).
Bleaching ThaW and Practirc 113
-
rrlay also be inferred that differences in free acidity These data imply, also, that at the usual bleaching
an.i *olst.,re content of clays have only a minor ef- temperatures, 95-110'C, maximum adsorption is
fect on free fatty acid formation during bleaching' achieved in less than 20 min of contact time.
- Brimberg's data expressed in tl're form of a loga-
tocopherol cbncentrations are usually lowered
somewhat, apparently depending on clay dosage and rithmic equatio4_produces the following function:
oiso .tay einiiency. Figure 12 indicates that for a 1% logC",'Co=-kVt
.ir:'"tug" a lowering of about 10% in total tocopl-tqlgl The constant-k increases with the amount of clay
coricentiation occuis in bleaching of soytrean oil ( 1)' used and the temperature of bleaching. With some
oils k also increases with increasing amounts of water
present in the oil/clay system at t = 0 (11). The same
Kinetics of Adsorption author also found that agitation beyond that re-
The rate at which the various comlrcunds of interest quired to keep the adsorbent (clay) in uniform sus-
in bleaching are adsorbed is, of course, an important pension in the oil, or slurrying instead of adding clay
asnect of tlie process. It especiall-v affects considera- as dry powdq, did not accelerate adsorption.
tions of proceis design and the question ofsuppress- No data on the kinetics of adsorption of oxidation
ins un\a'inted side reactions. Little investigatil-e work products. metals and phosphatides appear to have
"available, but in recent years the subiect has re- been published. Some data are available on the effect
is
ceived more attention in studies by LI. Brimberg (10, of moisture on the removal of trace mteals (i3) and
I I ) and also by IvI. NaC\'( 12). for example. These srud- phosphatides ( 14). These data imply that the rate of
Lr'ru"t" concerned especiall5'with the adsorption of
-hlorophylloid adsorption of all of these compounds is simiiar to
and caiotenoid compou-nds. Figure 13 that determined for caroiene and chloroph3;ll.
summirizes the findings by Brimberg It can be seen As indicated earlier, the tirne required for rnaxi-
that at a bleaching temperature of SffC and usir,g 1% mum adsorption is also important in connection witl'.
of an acid-activated clay, adsorption of chloropLryll side reactions taking piace during bleaching. Side
and carotenoids from rapeseed oil approached equi- reactions of interest are:
librium closely within 5 min of contact time- There Heai-cieavage (breakdorvn) of caro'-enoicis
was a slight improvement in the amount of material Hydroly'sis
adsorbed-up to20 min. Additional contact beyond 20
min. did nol result in signi-ficant further adsorption- Oxidation
Isomerization (including conjugation)
Polymerization
Bleaching Processes
There are many versions of bleaching processes in
use. In most modern plants these installations are
intended either for serni-continuous or continuous
operation. Ferv modern plants op€rate batch pro-
cesses. From the di-ccussicn on bleaching fundamen-
tals, it fcilows that batch processes can hai e sericus
drarvbacks. especiali-v ir-r respect to the time required
to heat the oii anC in resp€ct to the oili'cla1'contact
time. Eoth of these tend to he excessive and hence
Fig. f 3. Adsorption of carotene and chi,:rcphyll n'ith iime allow' unwanied sicie reaciions to take place to a
: . i, ,n5i! Fl-'. E0"(- I r I i) r. greater e:{tep.t.
I l1 T.K. Maq
tall! ?aflA
Fig. 14. Conrinuous bleaching s$item (DE SIIET) (18).
An example of a modern process for continuous ,:^-!.h" oil clay mixture after drying should be about
bleaching is shown in Figure 14. It senes primarily to 0.. lor according .to several authorJ(13. 2l). Excep_
describe in detail modern bleaching praitice and to tions are also cited in the literarure (ll. i8,21) in
point out desirable a^spects as deveiop-ed in the earlier respect to adsorption of phosphatides. metals ind
discussions. chlorophvll, as ment,ioned-befoie. The effect of niois-
Oil to be bleached is heated to bleaching tempera_ ture on.adsorption is probably quite complex. Its ma-
ture, usualll- 95-110"C, the adsorbent is ;dded and nipulation may in some cixes produce significant
rvetted by the oil for dispersal, and the mixture is bleaching economies.
cleaerated, and has moisture removed. Usual adsor- Filters. are_usualiy of the tank r1pe, liorizontal or
bent levels are from 0.2b%-B% of acid-activated vertical. b,ut filter presses are also ifitt'usea. Keeping
bleaching cla1,, or 0.192, to O.bit carbon or silica. T\t,o the adsorbent in uniform suspension in tank filters ii
porvder feedingsystems (not shown) which use more important and is said to be mbre easily accomplished
than one adsorbent, or a filter aid iogether w"ith an in a r-ertical filter tank than in a horizontal filter tank.
adsorbent are often used. Filter aid uiage ma1.be as Average residence time in tank filters is approxi-
filter precoat only, or at rhe rate of abou"r 1091,;lcla.v mately.l 5 min and just a few minutes in lilter presses.
usage. Depending on the size of the main bleaciring section
,u" acid, or."dry," pretreatment is to be gir.en prior
I,f and the tlpe of filter used, average total oil,,6lay con_
tt.r bleaching clay addition, the acid is mereredinto tact time can be more than 10-20 min, u,hich appeu.s
the heated oil stream a-nd uniformll,dispersed by an to be the best compromise betu-een marimum ad_
in-line mixing device. Then contact tirne^is allolved in sorption anci suppression of side reactions.
a stirred reactor vessel as shorvn in Figure 14. Acid A significant portion of the total adsorprion taking
nri-xing methods are available (19, l0)'-r.fricir are ca_ .
place in bieaching occurs in filtration. The oil passinfi
pable olgenerating a large inter:facial surface area fr.rr thrrrlgl 114 filter cake, which is continualll coilecrinfi
:tcid oii coniact. Tilis itlit.-iates the use r_;f a stirred on the filter medium. is in corrtact rvith i r-eq_, higfi
reactor, -q:r., t,s ar:id anci faciiif aies stock cliirn{e-s. ('()Il{:rriiltt-alion r-rf 3d51-r1bent. This acsorbent Stlli h'a.s
Ale'rage residencr. ti;rie in tlie ntain i;ieiichi-ng se,i:-
ti<ili ofthe i)rocess, rviiicii is *iiLtipoeri *ith arr igiru- -.onte frt,e iittir,e site.s. rr'hic'h represeilt significant ca_
paCII.\' ii,r- iur-ther adsorpl i,rit rr!- pi;lii; i.r:i-1-1pt;unii._*
t,.:r'il this e x:lniJ,;ic, i-s frr;nt ab<rui:i-.Jii r..tin. Titi,j'res_
irience tirrie st,-,iirii., tnat-i;e uli,ier \-ar..,,lilfil of ;1i'i-151.t I)i tseitt ii: 'Ie oii pirssi.ng ti-,rrr.rigir iii+ fijrr.,r ..ait Ti;is
arlso.r'p t t, :. is oi t lt e -1,.er.i'li lil in g bi.,i- i3pe a i.: ii i: r,., : r.i
t
t()i't' u'iri. rt()r-ttiiiiial iii-r.ing,.<it- ii m:ir. l-re it: i$.| .\iase: Irlr>nl',' k:r,,ir'ri a:i ihij *il!-i:ss trir:aih i,iie.,i." Initia.ilr.
ir itl-r ilt+-' si,i,oitrl s1.:,rqi' unrier pr.r,ra.ur.e itr sorne pro ii{ i Liti'r'r.ilt1'..ri Ilr.','t,t': iii'ltris*i-iii.lll :iit, silirr:li ,:.rh:ril.r-
('.-'s,i{'s.(.j;1,?1;. a-litr'ri. sir'iint lJ irrie, ti,ri i'i,1 ..u,1 .r,.11
lri. t,r:i ii, ,\:.. iii lulrrrri:rl;;rI.it!-,;,-i,eiii ll;-,.i,.i*- a,i,-ii
Il ltiriirj li-ir'ai-iaI..i(.iii ii_gii:tir;l-:,. -llie ii.,,.risti:i.r-.,(lli(liti tlcitiil :ir'l - :1riir r, {. iii}itr.ii'r: ilrI.i;:r{i-riir.:! ilit i.!tra:ii-;n
tsl.e<u:hing ThatrY and Prat:tk'e 115
-
cvcle.'I'his adsorption occurs under pressure' peroxide value and free fatty acid content are
'
Tire desired bleaching process capabilities can now checked, but less frequently. The anisidilte l'aiue, iron
be sunrmarized as follows: concentration and conjugation values may be moni-
tored in special cases, especially in connection rvith
a) Maximum efficienc-v in removing undesirable problem oils.
compounds, that is,
i). enough mixing to keep the adsorbent uni-
formly'susPended in the oil Disposal of Spent Adsorbents
ii) optimum temperature for adsorption - Increasingly, requirements are that disposal of spent
iii) oltimum moisture content of the oili adsor- adsortrenis-be free of negative effects on the environ-
bent slurr;'
ment. This means, primarily, that the oil retained in
iv) sufficient time for adsorption to take place adsorbents must be reduced to low levels. This is done
b) Suppression of undesirable reactions. that is. by various methods using organic soll'ents, hot waier,
i) limit temPerature oi a combination of these. Combustion is also used. A
ii) limit air access comprehensive oven'iew of disposd- methods \l'as
iii) limit time. presbnted by Fahn and Zschau (22). No regeneration
1b these points the following ma:{ be adCed: and re-use of activated clay seem possible at present,
clay'- addi-
but more effort may well be expended in the future to
c) In acid pretreating p-r-ior to b-leaching
possible investigate '.his further. Also, recent work on syr'.-
tion, ensrrre thai ihe smallesi acid
theric lilica arisorbents (7) suggesls that re-use of
droplet size is achieved for best acid,roi! contact' this t1'pe of adsorbent is feasible. Granular active car-
Thii facilitates phosph'aiide and metal remot'al. bon is routinei-v regenerated and re-used in other
and also facilitates excess acid removal b1'' ti"e industries.
clay.
d) Achieve ail oi the above w'ith rninimum resi-
dence time in the equipment to facilitate stock Summary
changes. Bieaching has assumed increasing import,ance rela-
tive to other refinery processes. It is essentiaily a
Process and Quality Control of Bleaching process of adsorption of dissolved and suspended
iompounds from triglyceride oils onto solid surfaces
Efficient bleaching requires that the key aspects of of ari adsorbent. The compounds to be adsorbed are
the process be monitored and controlled. These as- present in very low concentrations in the order of
pects are listed in Table 4. milligrams/kg. The Freundlich adsorption isotherm
Tire frequency with which measurements are re- desciibes the process best except for the press bleach
quired.,,arles. In process control, all items listed are effect. Acid-activated clays are mostly used, but syn-
,',-,n." o. Iess continuously monitored and adjusted, if thetic silicas and active carbon are also applied.
required, except for moisture in the oil;/clay mixture' There is considerable activity in adsorbent improve-
Molsture detefmination serves to ensure that the op- ments. Investigations of the kinetics of bleaching
timum moisture level is achieved for a particular show that the process reaches equilibrium in min-
bleaching task. utes. Side reacfions are especially important to oil
In quaiity control, red color and, -with some oils, quality and oil losses, making it important to limit
chlorophyli are routinely monitored. In addition, the adsorbent/oil contact time and temperature. Many
process versions exist reflecting past beliefs and
more up-to-date knowledge. Modern processes oper-
TABLE 4 ate in the continuous mode with good control of con-
Process and QualiW Control of Bleaching tact time and temperature. Equipment for acid
pretreating is usuaUy also incorporated. Environ-
Process control mentally sound dispoial of spent adsorbents by using
Temperature them in bui-lding materials or removing the oil before
Pressure dumping in landhll is now much more a requirement
Ailsorbent usage
than in the past. Re-use of activated clays does not
seem feasible, Uut re-use of the more expensive silica
Filter aid usage and carbon adsorbents is being investigated'
N{oisture in oil,'claY mixture
Ciarit-v of f rltrate
Qualit;.'control References
Reri color l. Delaner'. K A.. C.B. Ungertnant-,. Ettiv-tirv' Ari's<.rt';lrit'i'
('li lori;ii h1'l l nical Presenl ai ion ;iiil H:ir-\ l-r a\\'
B I eo,:h i ti g /1. l-t'r:h
'leroxide valtte Filrcli):rrtiiership.,3(r 1 00 L'hagrir Iilr ti.. ('ii'*. ci:rt:ii.
.r'nisi<line valLre )i,io 4* i i-i. L'.S.A.
(
F rr:,.: fait),' aticis 2. lreSirlii. Fi.l-.. in -lfir's-s Irtl;'i.sii't r t'v't'iitiiit;'':' :ii'rl r'<itl '
Irr-n (nickel i "^A.tlsor;:,iiltl anri Ilil, Iirr'|ialt{t'" ir'i' 'rti '-rsj
L'on.!ugation \It'ilr=ri iliii irrt. Iio,ik ('o.. liiri
116
T.K. Llag
117
llti
S.C- loJl
f. _ P!,' ABS.
I\ [5] PRES.
mn#lg
6.0 -
..
Through length),inregration and formula substitu,
[lon, we arrive at tlrt..basic e.quation
sleam requirements (S; in deoriorzriti6n.Iti] for,,_p;r.g
tlr
s= Po ln f vt init,i't ppa t.I ,^,
4.5
Er\. FFA)f toJ
4.0
\t',nnu,
3.5
The equation simply states that the theoretical
slgay. required is equal to the total pressure multi_ 3.0
plied by the moles ot non_volatit" oiftiuiaeO
product of the_vaporization efficiency iimes the b, il; 2.5
por pressure of the volatile matter. This quantity,-a_
multiplied by, rhe narural log of the raiio oi molei is 2.0
votatlte components in the undeodorized oil to of 'r.0
moles of volatile compounds ."*uining in the the 1.5 2
ished product. Therefnre, it can U" ii"E"a that hn-
amount of steam required is directly proportionalthe
PErcEl'fr smIpPI.r3 SiEcrrI
E-0.46
s-1.2% LLast qll*
P-2.rlmmHg
E-0.69
s-1.2r(
P-3.0mmHg
T-2600C
E-1.38
s-0.8%
P-3.&nmHg
T-248oc tlalrq lrry
iy, space does not permit revie$, of ali the makes and
models. Bailey (4) probably introduced the hrst com-
Cootrng lrry
mercial semi-continuous deodorizer, known as the
Girdler. This deodorizer became very popular pri-
marily because of its simple design.
During the development stages, it was discovered Co{cctrng Tny
that an oil level less than 100 mm in an open tray was
.,.ery ineflicient. However, when baffles were placed a
short distance above the surface ofthe oil, the vapor-
ization efficiency increased significantly. It u,as ap-
parent that the deflection created by the oil splashing ,,p- ,j _ ,
with the splash baffle Iocated 60 mm above the oil's Fig. 4. A typical semi-continuous deodorizer.
surface.
Figure 4 symbolizes the semi-continuous concept.
A measured amount (batch) of feedstock is created
in the measuring tank, and is then sequentially heat-
ed, deodorized and cooled by cycling the batches ap-
proximately every 15 to 20 min. This particular dia-
gram illustrates the "thermal siphon" style of heat /5"'
recovery.
Figure 5 illustrates a semi-continuous traywith the
splash or impingement baffles Iocated in the vapor
outiet. The baffles serve to return splash oil created
by the expanding stripping steam as it passes
t.hrough the oil. The shroud showr.l above the baffle
prevents refluxing of condensed fatt1.' acids from
reentering the tra!'and contaminating ti-re freshly de-
oCorked oil.
A iuli.'* continuous deodorizer is illustratetl in Pig-
ure 6. This dia{ram i}lu-strate-c ctiiitiirutiits florv from
lire exiernal deaerator through a vacuurii trim lieat-
er and into the iirst tra;,.at decilorizillg teinperat ures.
'f he t ra;.-s are icientir-ai in srze to the semi r.er-sirin. rr'it It
the addiritin tif veriicai i;ilffies in order ti.i pioriut'e ii
iabl'rintii st.r,le of plug fioiv iF'ig. 7'1. Origl:iaily'. aii
ileating arr.d r:txrlin{ r,\ias a(.c(irnpiis},etri in ihe lra.r,s. t'ig. 5. ,\ s;.lril i'i)nlin:.i,)Lls stiie rlr.r;iiirrl-zt.i trii''
120 SC l1l?
E]
ra|-based
I
I
trl
Bubble Cap
+
Edrrn -l I'i+
u___,
t
Sieve
I
TABI.E 2 TABLE 4
Deodorization Conditions Tocopherol Analysis
percxide values indicaied in Sampte SB L and SB2 can Deodorized Oil 3.0 tr. 35.4
probably be attributed to the Iength of time ( 10 da:si Percent Retained
from the test-runs to the analysis- No additives or in tlii 60.0 G ?4.0
antioxidants were used in the tests. The samples
were sent to an independent taste panel and the re-
sults shown here represent a span of four weeks after ers, fouling rvill not effect lhe operating efficienc5u or
the test runs. All samples were refrigerated and blan- finished oil qualiry.
keted with nitrogen. In summary, the test work and data presented
Table 4 illustrates the retained amounts of toco-
pherols left in the finished oils. Tocopherol is an im- clearly indicates that thin fiIm deodorization is an
portant natural antioxidant and with today's sophis- acceptable alternative to todays practices of tray,
sieve or bubble cap technolos/. Further anall,sis will
ticated consumer, it only seems natural to retain the be compiled regarding the retained oil sample quality
highest ievel of stability without the use of additives.
and published at a later date.
One of our major concerns for using structured
packing, or any packing for that matter, is the ten-
dency to foul when processing highly unsaturated Acknowledgments
oils. After six months of operation, the packing \ras
removed and inspected, and found slightly filmed The refined and bleached oils were pror.ided b.'" CSP Foods,
with polymerized oil. However, none of the channel Canada, and in the Lr.S. by Honeymead Products and the
openings were obstructed. We believe with biannual AE. Staley Company'. A special thanks to POS and their
chemical cleaning associated with liquid oil deodoriz- capable staff.
TABLE 3
Deodorized Oil Analysis
:+a
*Rcpresents samples of commerciali]' tiottie.j ,-;ii-
Ikndarizatiort usry a rrd f>ract.ict' t23
-Tlr