Note: Descriptions are shown in the official language in which they were submitted.
L7~
BACICGROUND O~' TE1~3 INVENT~ON
~ièld of the Invention
Thls invention relates to a p.rocess for sepa:ratin~
solid wax particles from a slurry comprising said wax particles
and a hydroca.rbon oil. More particularly, this process relates
to an improved process for filter1ny solid w~x particles from
a mixture of at least partially dewaxed oi:L and solvent wherein
the impxovement resides in using a needled~felt filter cloth
fabricated from fibers fusible by means of an open flame and
having a singed surface on which the wax is collected~
Descrl~tion of the Priox Art
Waxes axe defined as animal, vegetable or mineral,
depending on their source or origin. In their natural state,
most of .these waxes exist in solution in waxy oil. In order to
separate the wax from the waxy oil the oil is chilled, usually
in the presence of a sol~ent, to precipitate the wax therefrom.
The solvent functions both to reduce the solubility of the
wax in the oil and to reduce the viscosity of the resulting
dewaxed oil thereby greatly facilitating filtration of the
waxy slurry to separate the precipitated wax from the soIution
of dewaxed oil and solvent. Various filtration means may and
have been employed to separate the wax from the oil and solvent
- such as the use of plate and frame presses, shell and leaf
filters, cartrldge filters and xotary-drum filters. Continuous
rotary-drum f~lters are well known and used in the petroleum
industry for wax filtration, particularly for filtering wax
from dewaxed lube oil fra ~ ions~ A typical rotary~drum vacuum
filter comprises a horizontal, cylindrical drum, the lower
~: po.rtion of which is immersed in a trough co~aining the wax
slurry, a cloth filter medium or filter cloth covering the
horizontal surface
- 2 -
1 Of the drum, means Eor applying both vacuum and pressure
2 thereto and means ~or washing and removing wax cake deposited
3 on ~he cloth as the drum continuously rotates around it~
4 horizontal axis. In the~e filters thP ~r~ is divided in~o
compartments or sections, each section being connected to a
6 rotary ~trunnion) valve and ~hen to a discharge head. The
7 wax slurry is fed in~ a filter trough and, as the drum
8 rotates, the face~ of the sections pas~ successively through
g the slurry. In a vacuum drum filter a vacu~m ls applied to
the sections as they pas~ through the slurry, thereby
11 drawing oily filtrate through the filter cloth and deposit-
12 ing wax thereon in ~he form of a cake. As the cake leaves
13 the slurry it con~ains oily filtrate which is removed there-
14 from by the co~.tinued ~pplication of vacuum, al.ong with wash
solvent whîch is evenly distributed or sprayed on th2 suxface
16 o~ the cake. Finally, the washed wax eake is removed from
17 the surfac~ of the ilter cloth by a scraper which i~
18 a~sisted by means o~ blow gas applied to each sectivn o~ ~he
19 drum as it rotates an~ reaches the scraper. In a pressure
~ filter, the dewaxing solvent contains an autorefrigerant~
21 which, by virtue o~ it8 relatively high vapor pressure, is
22 suficient to apply a pressure differential across the filter
23 sur~ace of the drum~ thereby elimina~ing the need vr apply~
ing a vacuum thereto~ `
L~ttle attention ha~ been paid to the ~ypa of
26 ilter cloth employed ~or wax filtration- Fox yParS~ con-
27 ventiona~ woven filter cloth~ have been employed fabricated
28 ~rom natural or sqnthetic yarns. As the cycle of wax deposi
ion and xemoval from the fil~er cloth cGn~inues~ the
throughput 0~ the dew~xed oil and solvent steadily decreases
31 due to elogging o the clGth with wax particles which is
32 cAllecl bl~nding of thP ilter- After the filter has be~n in
~ .
78~
operation Eor a period of time -the blindi.ng continues to a point
where the filter cloth has -to be washed with a solvent such as hot
kerosene to dissolve and wash away -the ~clX par-ticles trapped -there-
i..n which have been causing -the reduction in -the filter rate. In
some applications, it is very possible for this wash cycle to
occupy four hours of a 2a hour day. Thus, not on].y does the blind-
iny of the filter cloth result in a loss of throughput, but it also
requires frequent washings and when the filter cloth is being wash-
ed the filter is down from a production standpoint in that it can-
not be used to filter wax from the slurry.
Therefore, any improvement to wax filtration processes em-
polying a cloth filtration medium or filter cloth which would re-
duce blinding of the fi~ter cloth would be a substantial improve-
ment to the art.
SUMMARY OF THE INVENTION
Thus, the present invention provides a process for separating
particles of solid wax from a slurry _omprising said wax particles
and a hydrocarbon oil by filtering said sl.urry through a cloth
filter med.ium, the improvement which comprises using, as the filter
medium, a needle-felted cloth Eabricated from fibers fusible by
means of a flame and having a singed and fused surface on which
said wax is co1lected, said cloth being further characterized in
havin~ a ~oot mean square surface roughness of said singed and
fused surface in excess of 500 rms microinches and a fouling factor
in excess of aboùt 75%.
The filter cloth can be further characterized in having pre-
ferably a permeability to air in excess of at least about l0, more
preferably 15 and still more preferably at least about 20 cfm
::
: , . . . . . . . .
(cubic :Eeet per minute)/ft. of cloth surface at a di:Eferential
pressure of 0.5 inches of water. The use of such a needle-
felted Eil-ter cloth in wax Eiltra-tion processes such as :Eiltering
wax particles from a slurry comprising a dewaxed
. 10
~ ~ .
.~ ' ' ' ,.
.
: ~ 20
~ - .
.,
~: : : .
;: ~ :
~ 4a -
7~
petroleum oil and a dewaxlng solvent has been found to pro-
vide a higher rate and throughput of dewaxed oil, a lower
blinding rate of the filter cloth and more complete discharye
of the wax from the cloth'then has heretofore been obtainable
using conventivnal fllter cloths woven from natural or syn-
thetic yarns. For example~ ik has been ~ound that with con-
~entional filter fabrics woven from spun yarns, cloth b:inding
can cause a 30% to 40~ reduction in ~eed filter rate after
repeated cloth exposure to wax cake buildup and discharge,
whereas the use of needled-felt filter fabrics heat set and
singed to fuse fibers projectin~ from the surface gave only
a 10% to 15~ reduction in ~eed filter rate. Still further,
with one particular slurry, use of a needled-felt filter
cloth ylelded a 66% increase,in productlvity ~s measured by
total volume of filtrate per unit of time~
The needled-~elt cloth'used in this lnvention is
a non-woven fabrlc manufactured by sandwiching an open mesh
gauze or leno weave ~crim between two batts of loosely packed
and randomly oriented fibers and passing the sandwich through
~20 one or more needle-felting looms ln which it is progressively
compressed and simultaneously subjected to vibratory needling
with~a plurali~y of needles having integral kick-up barbs
which pass through the sandwich and are withdrawn from the
~' fabric. On insertion and withdrawal of the needles the
barb~ push and pull some of the fibers through the scrim as
it is being compressed. After leaving the loom the needle~
felted~fabr~c ls ~urther compressed and at least one surface
thereof is treated by singeing by a flame to melt down and
fuse the ~ibers which project above the surface to ~orm a
~30 di continuous, ret~culated surface in which til~e root: mean
square surface roughness is in GXcess of 500 rms micro-
lnches. Thus, as hereinbefore descrlbed, the fibers :in the
- 5 ~
L.
ba-tts .Erom wh:~ch the n~cl.l~et~-ted ~.r:ic used i.n this inyen~
ti.on are made, must be capable of beiny melted and :Eused.
This elim;nates materials such as cotton wh:i.ch burn when
singed under an open Elame.
sRIEF DESCRIPTION ~F THE DRA~71NGS
__ ___
Fiyure 1 illustrates various types of surEaces of
needle-felted fabrics.
Figure 2 is a plot of feed Eilter rate versus
; number of di~s or cycle~ for a needle-fel-ted, singed filter
cloth compared to a woven cotton cloth obtained from filter-
ing a lube oil wax slurry. Th.ese curves are also reEerred to
as blinding curves~
Figure 3 i`llustrates the method used to measure the
permeability of the cloth.s o:E this invention.
.. .. ... ..
DESCRIPTION OF THE PREFERRED EMBODIMENTS
.
As nereinbefore described t supra, after leavin~
the loom the needle~felte~ fabric useful in this invention has
at least one surface thereo singed with an open flame to melt
and fuse projectiny Eibers to forrn at least one discontinuous,
reticulated surface having a root ~ean square surface roughness
i`n excess of 500 rms microinches. Both sur:Eaces may be singed,
but it is only necessary for that sur.Eace upon which the wax is
to ~e deposi`ted to be singed to form the reticulated or rough
surface~ Needle-felted fabri`cs which are not useful in the
process of this inventi`on i`nclude those fabrics which after
leav~ng the loom are not sinyed and also those -the surface of
whi`ch i`s smoothed by calenderi`ng~ An unsinged and uncalendered
ne:edled~felt surface exhibits poor wax cake discharge because
the wax adheres to the loose surface fibers. A smooth but un-
3n dulatiny surface gives good cake discharge, hut -the cloth fouls
at a rate abo~t equivalent to a conventional woven filter cloth.
: ~6~
L~
A needle-felted fabric use~ul :in the instant invention is one
having a singed surface on which the Eused fibers ~orm globs of
hard polymer several -thousandths oE an inch high on the cloth
surface. This surface gives good cake discharge and significantly
reduces cloth fouling or blinding, resul-!:ing in increased filter
throughput and reduced washing frequency.
The densities of needle-felted ~abrics may range from about
10 to about 30 ounces per sq.yd. with permeabilities to air ranging
- from 2 to about 150 cfm/~t.2 at a pressure drop or differential
across the fabric of 0.5 inches of water. However~ it has been
found that needle-felted fabrics useful in this invention should
have permeabilities in e~cess of 10, more preferably in excess of
15 cfm/ft. and still more preferably in excess of 20 cfm/ft.2.
In some applica-tions, fabrics with a permeability as high as 60
to 120 cfm/ft. have worked in this invention when the waxy slurry
to be filtered is derived from a petroleum lube oil distillate.
The method used to measure the permeability of the cloths is
illustrated in Example 7.
; Textile ~ibers suitable for the preparation and manufacture
of the needle-felted fabrics useful in this invention comprises
both organic and inorganic compositions, a major criterion for
usability in the instant inventlon being fusibility of the fiber
by means of an open flame. Included within this category are
I
glass ~i~ers and suitable therrnoplastic fibers, illustrative but
non-limitlng examples of which include;
1. Isotactic poly a ~ mono-olefins, of which non-limiting
examples include propylene, M.P. 160 to 170C; 3-methyl-butene-1,
M.P. 2~5 to 300C; 4-methyl-pentene-1, M.P. 205 to 235C; 4-methyl-
- 7 -
, - ~ . : . :
:~ . . - - : ,. , , . ., ': ,
. . .. . , , - - : : ., . : ,
:. : - ' .'- .. -. : : . ' . .
hexane-l, M.P. 188C; ancl 4,4-dlmethyl-pen-tene-1, M.P~ 320C.
Preferred is polypropylene.
2. Linear polyamids having -the general ~ormula
.
CNH (CH2) CO-] n made by the polymer1zatlon of lactam.s. Non-
limltlng examples, commonly designated Nylons followed by a
slngle number equal to z-~l, include Nylon 4, M.P. 260C made by
the polycondensation oE pyrrolidon; Nylon 6, M.P. 223C made by
.
the polycondensation of caprolactami Nylon 7,
.
,~ '
.
~.
:~ .
~ I
~ - :
;: :
7a -
~: . : :. :. :
, .. .. ~ . .
. . .
i7YI~
M.P. 233C made by the polycondensation of unanthlactam;
and Nylon 11, M.P. 190 C mad~ by the polycondensatlon o~
amino undecanoic acld. Nylon 6 is preferred.
3. Linear polyamlds and aramids ha~ins the general
~ormula: ~ NH-(CHz)x - NH--CQ-(R)y~CO 3 wharein x is an
integer of from 2 to 10, y is an integer from 1 to 18 and R
is independently selected from the group consisting of
methylene - (CH2) - andphen~lene, made ~y the polycondensation
of diamines and dibasic acids. Non-limiting examples, com-
monly deslgnated Nylon~ followed by two numbers, the first
number indicating the number of carbon atoms ln the diamine
and the second number lndicatlng the number of carbon atoms
in ~he dibasic acid include: Nylon 2-10, M.P. 276C. from
ethylene diamine and sebacic acld; Nylon 66, M.P. 205C.
~rom hexamethylene diamine and adipic acid; Nylon 6-12, M.P.
217C. from hexamethylene diamlne and decanedicarboxylic
acid; Nylon 8-6, M.P. ~50C. from octamethylene diamine and
adlplc acidi and Nylo~ 10-8, M.P~ 217C. from decamethylene
diamine and suberic acid. Usable aramids cvmprlse the poly-
condensatlon products of terephthalic acid and diamines
having 2 to 6 carbon atoms. Preferred is Nylon 66.
4. Linear polyesters ~ree of olefinic unsatura-
tion havlng the general formula: ~O-R-O-CO-C6H4-CQ~n wherein
R ls independently s~lected from the group consisting of
ethylene ~-C~2-CH~-); 1,4-butylene (-CH2-CH2~CH2-CH2-); 1,
~-cyclohexy7ene (~ -7C 7 o~,-xylylene ~-CH2-C6H4-
CH2-~; and 1,4~dlmethylene cyclohexane (-C~I2 -
Preferred is poly(ethylene terephthalate) J M.P. 257-265C~,
depending on the degree o~ c~ystallinity~
5. Nitrile ~ibers of which Acr.ila~*, Creslan*,
Darvan*, Dynel*, Orlon* and Verel* are commercially available.
6. Vinylldene chloride fibers of which Saran
* Trade Mark
.71~
is commerciall~ available.
7. Cellulose triacetate fibers.
The scrim used as an interlayer or sandwich be-
tween the batts may be woven from the same type of fiber
used ~or the batts or may comprisa a different fiber. Yarn
used for the scrim may be a monofilament, multifilament,
or ~pun from staple. It is importan~ that the softening or
melting point of the flber not be exceeded in the filtering
or washing proces3. For example, ln some applications
within the petroleum industry, filtar cloths are washed with
hot kerosene at a temperature of about 200Fo when the cloth
becomes too clogged wlth wax particles. This would preclude
the use of a polypropylene flber based felt in such an oper-
ation and one would have to use a elt made from a fiber
with a higher melting point, such as a DACRON ~ polyester.
As herelnbefore mentioned, supra, the needled-
felt filter cloth used in this invention must have a root
mean square surface roughness in excess of 500 rms micro-
inches, more preferably in excess of about 800 rms micro-
inches, as measured by a screen projection method~ In this
method, the cloth surace is projected onto a ground glass
screen a~d the maximum amplitude of the peaks and valleys
is measured in inches. An rms value in microinches is
; ~ then calculated from this measurement by multiplying this
amplltude by a constant factor o 260,000 derived from com-
puting the root mean square value of a random, saw tooth
wave function, similar in shape to the cloth surfaces.
Another characteristlc o needled-elt cloth suitable for
i~ .
use in this invention ls a foullng factor in excess of 75~
,
Although the process of this invention should be
useful~for filtering wax particles from any wax conta:ining
sLurry, it is pa~ticularly useul for filtering waxy
_ g _
,~/
`::
partlcles ~rom slurrles containing hydrocarbon oils such as
filtering wax precipltated from a mixture of a petrvleum oil
and a dewaxing solvent, especlally when the petroleum oil is
a lube oil fraction.
This lnvention will be more readily understood
by reference to the followlng examples.
EXAMPLE 1
The ~ollowing test procedure was used to evaluate
the blinding characteristlcs of a variety of filter cloths~
A test. slurry was made by mixing one part by volume of a
waxy paraffinic dls~illate having a viscosity of 600 SUS
at 100F. with 3.2 parts by volume of a 45/55 I,V ~ (liquid
volume) mixture of methylethyl ketone ~MEK~ and methyliso-
butyl ketone (MIBK). This solution was heated above the
cloud point of the distillate which was about 130F. a~d
then ch~lled with rapid agitation to about 20F. to crystal-
lize and precipitate the wax thereby forming a cold waxy
slurry. This waxy slurry was then used in the following
procedure.
l. ~ labor~tory size lea~ filter onto which the
filter cloth under test was clamped was dipped into the
chllled slurry for 30 seconds while maintaining suction on
a filter at a pressure of 275 torr and the filtrate was
collected in a flask.
2. The leaf filter was then removed from the
slurry and solvent wash was applied for 30 seconds to wash
the filter cake on the filter cloth.
3. Th~ wax was then discharged from the filter
cloth into a beaker by blowing with alr at a pressure of 2
~30 psig ln the reverse direction through the filter cloth~
4. Steps l, 2 and 3 were repeated for from about
25 to 30 cylesO
-- 10 --
' ~
:
The filter rate was then determined and the fil-
trate, wash and recovered wax from each cycle were combined
and the solvent removed by stripping under vacuum. ~rhe re-
duction in filter rate due to blinding of the ~ilte.r cloth
could then be plotted for each aloth tested. The lower
curve ln Figure 1 is a plot of the filter rate obtained us-
lng a commercial grade of woven cot~on fi.:Lter cloth recom-
mended for wax iltration by manufacturers of rotary vacuum
filters and is typical of the decay in f.ilter rate obtained
with this type of clo~h.
A wide variety of commercially available filter
cloths were tested by this procedure for blinding performance
in dewaxing the wax-containing lube oil slurry, These cloths
comprised a wide range of textlle fibers, yarn constructions
- and wea~es which could be grouped into the following four
main ~a~egories
1~ Con~entlonal filter cloths woven ~rom yarns
spun from a variety of fib~r~.
2. Light-weight fabrics woven from continuous
: 20 multifilament yarnsO
3. Fabrics wove~ from continuous monofilament
yarns. These cloths per~ormed essentially as fine-mesh screens~
4. Needled-felt fabrics manufactured and designed
for filtering du~t particles from gases which were made
~rom polyolefin and polyester s~aple. These fabrics were
singed on one or both sldes in order to melt down and fuse
the fibers protruding above the surface after the needling
operation to yi ld a hard, rough reticulated surface.
As hereinbe~ore mentioned, the filter temperature
30~ was 20~. The cloths used ln the screening test exhibited
. permeabilities ranging rom 1.5 to about lrOOO c~m/ft. of
' ~
`X
.
7~L
filter cloth surEace at a pxe~sure drop of O.S inch~s water
gage. Permeabillties of the needled~felt fabrics ranged
from about 10 to about 130 cfm~f~.2. These cloths exhibited
fouling factors ranging from as low as 26~ for a light fabric
made ~rom a contlnuous polyester filament yarn, about 36
or a woven cloth made from polypropylene, to a fouling
factor oE 87~ obtained ~roln a singed polypropylene need]ed
felt cloth. The conclusions obtalned from this experiment
were:
(a) WoYen cloths having permeabili~ies in the
range of from about 10-100 cfm~ft.2 at a dif~erential pres-
sure of 0O5 inches of water, exhibi~ed blinding rates roughly
equivalent to the traditional woven fabrics currently used
as filter cloths on rotary vacuum filters in dewaxing plants.
(b) Cloths woven ~rom continuous multifilament
~arns possessing low permeability exhlbited extremely high
blinding rates.
tc~ Cloths woven ~rom continuous monofilament
:, .
yarns exhibited only slightly better blinding perfoxmance
~han those c~oths woven from the multifilament yarns and
possessed an additional disadvantage in that the fabxic was
relatively fragile making damage of same likely in commer-
cial rotary wax filters.
(d) Singed! needle-felted fabrics made from poly-
olefin or polyester fibers, having permeabilitles in excess of
20 cfm/ft. and a root mean square surace rou~hness in excess
~;~ of at least S00 rms microinches exhibited superior blinding
; performance with s~stantlal increases in filter rates over ~ -
conventional woven cloths.
EX~MP~ 2
~
The procedure used for the experlment in this
example was essentlally the same as that used ~or Exarnple
.~
,
1, with the exception that four di~.~erent feeds were used
at various dilution levels, ~ilter temperatures ranging
f.rom -10 F. to ~22~ and at varying MEK~MIBK ~olvent ratios.
In this example, a polypropylene needled-felt fabric was
compared to a standard, woven nylon filter cloth used ln
a commercial ketone dewaxing plant~ The results of the
expexlment ~howed tha~ u9e of the needled-elt resulted in
an increase in performance, as measured by fouling factor,
of up to 30~ and a maximum producti~ity increase of as much
as 65.6~. The results for a run uslng a paraffinic, medium
lube oil di.~tillate having a viscoslty of about 500 SUS
at 100F. are plotted in Figure 1 ~or both the polypropylene
ne~dled-felt filter cloth and for th~ commercial woven nylon
~ilter made ~rom a continuous filament warp and a spun
~ilament yarn used in a ketone dewaxing plant. The percent
increase ln producti~ity was obtained by calculating the
ratio of the areas under the respective blinding curves.
It should be noted that a~ well as showing greatly reduced
blinding, the lnltial filter rate of the needled-felt cloth
was frequently substantially highex than that of the commer
cial woven cloth. This factor was taken into consideration
when detexminlng the productivity increaseO
EXAMPLE 3
In this experiment the ~inged needle-felted poly~
propylene filter cloth u~ed in Example 2 was compared with
a woven ilter cloth used in petroleum refinery dewaxlng
:~ plants made from ~ylon. The comparison was made using a
12 inch diameter Door Oliver rokary vacuum filter to dewax
a wax-containing luba oil slurry derived ~rom a paraffinic
lube oil distillate having a viscoslty of 600 SUS at 100F.
; diluted ~ith a mixed ketone dewaxing ~olvent and chilled to
20F. which ~as ~he filtration~temperature. The needled-
- 13 -
felt haA a permeability to alr of 30 cEm/ft.~ at 0.5 inches
of water, a root mean square surace roughness of 1500 rms
mlcroinches as measured by a surface proile projection
method and a laboratory obtained fouling factor of 94.9%.
The commercial nylon clot~ had a permeability to air o
about 15 c~m~ and a foulln~ ~actor of 70% when tested
by the same procedures. The resul~s of this experiment are
llsted in khe tableO
Woven Needle-felted Poly~
~0 ~y~ ro~
Total we~ght of ~eed (lb) 105.3 112.5
Total filter time (mm) 33.0 ~6.3
Average feed filter rate 8.65 11.61
(m3/m2 day)
Dewaxed oil pour point tF.~ ~22 -~ 19~5
No wash oll content t~) 4202 4507
Increase in throughput (%~ - 34
EX~MPLE 4
In this example two sets of experimental runs were
~ 20 made on plant-siz~ rotary vacuum wax filters having stan-
- dardlzed throughputs in whlch the needle-felted cloth used
in Example 2 and 3 was used on one filter and compared~ at
the same:time, w1th the same wax slurry on a secon~ filter
equippad with a standard wo~en rotary vacuum filt~r cloth
made from nylon. In both cases, a 16~ increase in ~eed
filter rate throughput wa~ obtained with the needled~felt
cloth compared with the woven c70tho Further, lt was found
that the hot wash cycle of 6 to ~ hours normally used in
~ : the dewaxing plant could be extended to 40 hours with the
.: .
~eedled~felt cloth on the ~ilters.
EX~MPLE S
In thls experiment a needle-felted fabr:ic made
14 ~
.... ... . . .. .. . . .: . . . ~ , .
78~
from a polyester staple which had been singecl on one side
of the cloth to melt and fuse the projecting fibers which
had a permeability to air of 120 cm~ft~ at a differential
pres~ure of 0.5 inches of water, and a root: mean square
surface roughness o about 1000 rms microinches was tested
by the laboratory leaf-fll~er procedure used in Exampl~s 1
and 2 and found to ha~e a foullng factor of 85~7~. This
compare~ to a fouling factor of about 70~ for a con~entional
filter cloth.
EXAMPLE 6
.
This experiment demons~rates how to calculate
the fouling factor. The ~ouling fclctor is a measure of the
deca~ of the feed filter rate as the filter cloth continues
to blind or become more and more clogged with wax particles
as a function of time or cycles. There are a number of
different ways to expre~ this. For example, a perfect,
nonbllnding fllter cloth would have a fouling factor of
.~
100~ which means that the ~eecl ~ilter rate would be constant
as the dewaxlng operation would continue. Howev0r, this
:20 does not happen in reali~y as shown by the curves in Figure
2. One way to calculate the ~oullng ~actor of a filter
cloth is to determine the area under the blinding curve and
dlvlde it by the rectangular area that would have existed
had the ~ouling factor been 100 (i.e~, a nonblinding cloth).
The area under the curve i9 extremely meaninyful because it
:: represents the total filtrate ~or a gi~en period of time or
number of cycIes. It has been ~ound that the blinding
curves obtained ~rom the ~arious ~loths used in these ex-
amples best fitted an exponential decay model mathemati-
cally expressed as
~ Feed ~llter rate - AD
where D i~ the number of dip~ and A and ~ are fit~ed
~: 15 -
.. ..
7~31
parameters~ There~ore~ ln mathematical terMs r the area under
the bllnding curve ~ay be expressed as follows:
Axea under Blln~ing ~urve ~ ~ ADB~B ~ (
1 B~
The area produced byl~ theoretical ~ilter cloth whlch does
not blind at all ~i~eO~ B ~ O) is:
~D-l)
The fouling factor as determlned by the xatio of the area
under the blinding curve to th~ total are~ obtained wi.th a
nonfoullng or bllnding cloth is then expressed as
B~l DB~l 1 ~ DB~
--~n~nr------ (B~ D-l~
This ~raction, expressed as a percenta~e~ wa~ designated the
cloth fouling factox ln the examples used in this applica-
tion and was used as the main index of filter cloth blind-
lns performance. The ~alue of D was set at 100 dips, or
:.
the rough equivalent of foux hours of rotary ~ilter operation.
: 20 While not wishing to be held to any single theory
~or the improvPd perform~nce of the singed, needle-felted
fabrlcs used in the embodiment of the instant invention for
: wax ~iltration, it is beIieved that the hard, reticulated
; or rough sur~ace created by the ~used filaments of the
fibexs:plays an important part in the improved throughput
rate~obtained with them. It has been observed that poor
wax discharge is obtained with unsinged cloth due to adhesion
; of thb wax to the fiber ends,
he singed, needle-felted cloths useful in the pro-
~cass of this lnvention exhibit ~ xandom distri~ution of varied-
shaped globs.o~ ~used:or melted ~iber ends several thousandths .:
.~ of an inch~in heigh~. The~e globs pro~ide dis- -
6 :-
~,"", ' ' '''', '' ' ~ ., -- .. - - -
~\
continuities i`n l~e wax cake on ~he sur;E~ce o;E the c~oth
which allow the wax cake -to discharge more readily and at
the same ti~le prov~de c~annels for increased rate of filtration.
EXAMPLE 7
This example illus-trates how the permeability of
the cloths were measured in this invention. Referring to
Figure 3, a sample of cloth is clamped :in a ~-inch copper union
mounted in a ~-inch copper pipe. Both the inside diameter of
the pipe and the diameter of the cloth sample exposecl therein
was 11/16 inches. A water manometer was tapped into the pipe
before and aEter the cloth to measure -the pressure oE drop
across the cloth. An air rotameter, calibrated by the dis-
placement of a known volume of water, was connected to one end
of the pipe and the okher end of the pipe was left open to the
a-tmosphere. Air was passed through the calibrated rotameter
into one end of the pipe and through the cloth into the
atmosphere. The pressure drop across the cloth was measured in
inches of water at various air flow rates. The air flow
rates (cfm/ft.2j and corresponding pressure drops tinches of
water) for each sample of clot~ measured were -then plotted on
log-log paper and the flow rate at a ~-inch water pressure drop
was taken as the permeability~
This method of measurin~ the permeabillty has been
found to give results a~out 50~ greater than the results
, ~
obtai`ned usi`ng a Frazier apparatus per ASTM D 737-6~. Thus~
if the method used in this invention gives permeability oE 15
cEm/ft. , the corresponding Frazier permeability will be about
10 cfm~Et. .
,
:
~17
. .
SUPPLEMENTARY DI5CLOS~RE
It has now been found -that the permeabilit~ of the filter
cloth of this invention is most impor-tant: where relatively fine
wax crystals are presen-t i.n the waxy slurry to be filtered. This
is illustrated when the oil -to be dewaxecl is a relatively heavy
oil such as a petroleum lube oil bright stock or deasphalted resid
having an inital boiling point above about 800F. It has been
found that bright stock wax slurries contain fine crystals, some
of which pass through the filter clo-th thereby causing a haze in
the dewaxed oil if the permeability of the cloth is in excess
of about 15 cfm/ft. . Thus, for bright stocks and other heavy
oils containing appreciable amounts of material boiling above
about 1050F, the permeability of the filter cloth will preferably
range between from about 3 to 15 cfm/ft. . When the oil to be
dewaxed is a lighter oil such as a lube oil distillate fraction,
it is not necessary to use a cloth having a permeability as low
as 3 cfm/ft.2.
.
Thus, it has been found that needle-felted fabrlcs useful
in this invention should preferably have permeabilities in excess
of 10 cfmtft.2.
Thus, the invention provides a process for separating
particles of solid wax from a slurry comprising said wax particles
and a hydrocarbon oil by filtering said slurry through a cloth
filter medium, the improvement which comprises using, as the
filter medium, a needle-felted cloth fabricated from fibers
fusible by mean.s of a flame and having a singed and fused surface
on which said wax is collected, said cloth being further charact-
:
- ~ erized in having a root mean square surface roughness of said
.
~ ~ - 18~ - .
. .
. -: .. ~ . -, , , . ., . :
,
.
,
~ ~3~
gL
slnged and Eused surface in excess of 500 rms microinches, a
Eouling factor in excess of about 75~O and a permeability to air
in excess of 3 c:Em/ft. a-t a diE:Eeren-tial pressure of 0.5 inches
of water.
.
!
!
: : :
- 19 -
-
~ . - . - - ::
- : :: : . . . .
., . , . . - . ..
.. ., :. . . ~
- : . ~ - - . . :
