Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
`:
~34Z~2
This invention relates to a process for the re-
moval of solid particles which form a compressible
filter cake from a liquid stream, the stream being
conducted through a continuous filter, such as a rotary
disc or drum filter.
It is known that in the removal of solid particles
from a liquid by filtration the liquid is passed through
a filter in which the solid particles are left behind.
With filtration in its most elementary form the problem
is that when the filter has been in use for some time, a
filter cake develops on it, which results in an
increased resistance of the filter until finally com-
plete blockage occurs. The yield of the filter in terms
of filtered liquid will therefore gradually decrease.
A distinction can be made between two sorts of
filter cake, viz. a compressible and a non-compressible
type. It is known that with non-compressible filter
cakes the yield of filtered liquid in principle
increases according as the~pressure drop across the
filter is increased. With compressible filter cakes, on
the other hand, it is seen that once a threshold pres-
sure drop is reached an increased pressure drop across
the filter causes no perceptible increase in yield of
filtered llquid.
S m ce the;~yield of filtered liquid rapidly decreases
with compresslble filter cakes, ways have been sought to
- 2 - ;
3~Z~32
overcome this problem. This search has produced many
variant.s of the continuou.s filter in which the filter
cake is also removed continuously. Two important examples
of this type of filter are the rotary disc filter and the
rotary drum filter. With both types the pressure drop
across the filter can be created by a positive pressure
on the pressure side and/or reduced pressure on t'ne
discharge side of the filter.
A further problem in the filtration of particles
which form compressible filter cakes from a liquid with
the aid of a continuous filter is that in spite of the
continuous remo~al of the filter cake the yield of
filtered liquid drops relatively rapidly. For instance,
when crystallized~wax particles are removed from a
hydrocarbon oil with a rotary drum filter, it is found
that although the filter cake is scraped off and the `
filter washed in counterflow the drum filter beco~es
blocked within a period of only a few hours to such an
extent that the yield of filtered oil drops below an
acceptable minimum. The filter then has to be taken out
of service and cleaned, which results in a loss of
production. ,-
The present invention aims to provide a method by
which the yield of flltered llquid is lncreased for
continuous filters over the period of time until ef-
fective blocking occurs and/or by which this period is
:
'
'
.
~34Z~
extended. To this end, the pressure drop across the filter is
increased stepwise in the process according to the invention.
Surprisingly, it has been found that - although the yield of
filtered liquid as a function of time is almost independent of
the pressure drop across the filter - a stepwise increase effects
an increase in yield each time. Naturally, the yield gradually
falls again after each increase, but the observed difference in
yield before and aftèr the stepwise increase is found to persist.
Thus, according to the present invention, there is
provided a process for the removal of wax particles, which form
a compressible filter cake, from a hydrocarbon oil, the oil being
conducted to a continuous filter, in which the pressure drop
across the filter is increased stepwise by at least 0.1 bar at
each step.
In this specification the term stepwise increase refers
to an increase which takes place in a jump, i.e. which has the
character of a distinct increase in pressure drop effected within
a short time - hence clearly some time before a new increase.
According to a preferred embodiment of the invention,
the pressure drop across the filter is created by reducing the
pressure on the side of the filter through which the liquid
leaves the filter. The liquid usually flows through continuous
filters from the outside to the inside, so that in the case of
the above-mentioned so-called reduced pressure filter the space
round it can remain at atmospheric pressure and therefore does
not require an exceptionally
~.~L3~
strong construction, whilst inside the filter a reduced
pressure is maintained. In implementing this feature of
the invention, it is advantageous that a relatively
small volume is rapidly brought to the new reduced
pressure.
In order to benefit most from the effect of the
invention, i.e. an increase in yield, it is preferred to
pass the liquid stream through the filter in a filtering
zone and remove the particles left behind on the filter ;
in a cleaning zone by countercurrent washing, scraping
off and/or blowing off with an inert gas in the reverse
direction to that of the flow. In this way the filter
resistance is continuously kept as low as possible. In
addition, the effect of the invention then proves to be
substantial. ;`
The removal of wax from mineral oil products, the
so-called dewaxing, leads to the formation of compres-
sible filter cakes and even with the continuous removal
of the filter cake from a continuous filter, the total
period during which the filter can be kept in operation
is limited to a few hours. Tn view of the quantities of
liquid to be filtered that are usually handled, the
provision accordlng to the invention can result in a `
considerable saving.
In this dewaxing the stepwise increase in the ,
pressure drop across the filter is preferably carried
:"
:
,~
.
~3~28,(3
out in such a way that the pressure drop is increased
each time by at least 0.1 bar. This means a considerable
restriction indeed in the total number of pressure drop
increases possible (the reduced pressure being further
reduced each time) in particular with the continuous
filters whicll already have a reduced pressure inside the
filter such as those which are usually employed for
dewaxing. But it has been found that a few large pres-
sure drop increases have a greater effect than a larger
number of smaller pressure drop increases. It is even
preferred to increase the pressure drop only once every
thirty minutes.
In case the invention is applied to a process in ~-
which the filter is taken out of service as soon as the
quantity of liquid passed through per unit tlme has
dropped below an unacceptable minimum, it is preferred
according to the invention to increase the pressure drop
across the filter stepwise at most three times over the
period between putting the filter into service and
withdrawing it from service. ~
Although differences are~found in individual cases, `
the best result will generally be achieved when the
pressure drop is increased each time at a rate of at
least 1 bar per hour. At too low a rate the result will
be small, at too high a rate the cost of the equipment
that is required for increaslng the pressure drop may
become excessive or the filter may even be damaged. ~;-
-- 6 --
`;
~3~ 2
The invention will be further elucidated herein-
after with reference to an example. The accompanying
figure is a graph showing the yield of filtered liquid
as a function of time for three experiments.
Example
An oil containing 2~.7 %w wax in the form of small
crystals which produced a compressible filter cake was
filtered with the aid of a rotary drum filter in which a
reduced pressure was maintained inside the cylindrical
filter. The filter was continuously scraped.
As it had been found that the rotational speed of
the drum filter has an influence on the oil content of
the filter cake and consequently on the yield of fil-
tered oil, the three experiments were carried out at the
same rotational speed, i.e. 36 revolutions per hour.
The first experiment was performed at a pressure
drop across the filter of 0.8 bar; i.e. the pressure
outside the filter was 1.0 bar and inside the fil~er 0.2
bar. The experiment was started with a clean filter. The
yield of filtered oil was continuously determined with
continuous scraping off and blowing off of wax from the
filter for four hours, after which the experiment was
stopped. The result has been shown in the graph (curve
A), in which the yield of filtered oil in tonnes per day
has been plotted against filtration time in minutes.
The experiment was then repeated with a clean
filter and with a pressure drop across the filter of ;~
only 0.2 bar (see curve B).
- 7 -
~3~
Then, a third experiment was carried out ~see curve
C), which was started with a clean filter again and with
an initial pressure drop across the filter of 0.2 bar.
During this third filtration the pressure drop was
increased stepwise three times according to the inven-
tion, viz. after 70 minutes from 0.2 bar to 0.4 bar,
after 130 minutes from 0.4 bar to 0.6 bar and after 190
minutes from 0.6 to 0.8 bar. These stepwise increases of
the pressure drop were effected each time within lO `
minutes.
The graph shows that at a pressure drop of 0.8 bar
the yield is almost equal to the yield at a pressure -
drop of 0.2 bar (curves A and B); the pressure drop of
0.8 bar initially caused a somewhat higher and subse-
quently (after about 80 minutes) a somewhat lower yield
than the pressure drop of 0.2 bar. ~ `~
From curve C it is seen that the stepwise pressure
drop increases raise the yield to a level that is higher
than that of curves A and B and that the yield is in-
creased in particular during the pressure drop increase ;~
proper, whilst the relative increase persists, at least ~
partly, after each pressure drop increase. -
It will be clear that the surface area under the
curve is a measure of the total yield of filtered liquid
and that this totaI yield is highest for curve C.
'.
