Note: Descriptions are shown in the official language in which they were submitted.
l i 33L~7 ~
BACKGROUND OF THE INVENTION
FIELD OF THE INVENTION
This invention relates to an improvement in the treatment of
whole tailings produced by the hot water extraction process for recovering
bitumen from tar sand. More particularly it relates to a method for
optimizing flocculant addition to whole tailings, said addition be;ng
practiced as part of a procedure for separating the liquid and solid
phases of the tailings.
BACKGROUND
The hot water extraction process, as it is commonly known,
is the process being used commercially to produce bitumen from tar sand.
It is schematically outlined in Figure 4.
In accordance with the process, tar sand is fed into a
rotating conditioning drum. Here it is mixed with hot water, steam, and
a process aid (commonly NaOH). Retention time in the tumbler is typically
kept to less than 10 minutes. In the course of this conditioning step,
the bitumen, which can be likened to a binder, is heated, with the result
that the water, solids and bitumen of the tar sand are less cohesive with
respect to each other.
The product of the conditioning step is a hot, thick slurry
which is screened to remove oversize matter, and then diluted with additional
hot water. The screened, diluted slurry is then advanced through a primary
separation vessel (PSV), where it is retained under quiescent conditions.
Retention time in the PSV is typically in the order of 30 minutes. Be-
cause the components of the slurry are in only loose association, due to the
conditioning step, and have been further diluted, they are now able to
separate in the PS~ under the influence of gravity. Coarse sand contained
in the tar sand, having a gravity of about 2.65, sinks to the bottom of
the PSV and is discharged. This stream is commonly called primary tailings.
- 2 -
1~ 3157'~
The bitumen has a density of around 1~00, that is, close to that of water.
Left to itself~ it therefore tends neither to sink nor to float. However,
air bubbles have been incorporated into the slurry during conditioning in
the tumbler. The bitumen and air bubbles become attached with the result that
bitumen rises to the top in the PSV and is collected and discharged as a
froth. This froth is termed primary froth.
Some bitumen fails to so rise because the size of the air
globules is too small or because of failure to get aerated. This unaffected
bitumen remains in the central region of the PSV and helps make up a
fluid portion known as the middlings. To increase the efficiency of the
process, a middlings stream is continuously withdrawn from the PSV and
advanced to induced air flotation cells where, by vigorous agitation and
the addition of external air, a second yield of bitumen is obtained in the
form of a secondary froth. A watery solids-containing stream termed
secondary tailings, is discharged from the base of each induced air flotation
cell.
The secondary tailings and primary tailings are combined to
form part or all of a waste stream termed herein as "whole tailings".
The secondary froth is highly contaminated with water and
solids in comparison to primary froth. Therefore it is preferred to
upgrade the secondary froth by retaining it in a vessel, termed a froth
cleaner, to allow part of the contaminants to settle out. The underflow
from this vessel may be recycled to the induced air flotation cells.
The primary and secondary froths are then combined and subjected
to dilution centrifuging to separate the bitumen from the remaining water
and solids. More particularly, the combined froths are diluted with a
hydrocarbon diluent, commonly naphtha, and treated in a sequence of scroll
and disc centrifuges. The waste stream from the scroll centrifuges is a
thick stream of coarse solids containing minor amounts of water, fines,
diluent and bitumen. The waste stream from the disc centrifuges is a watery
113157 ~
stream containing minor amounts of solids, diluent and bitumen. The two
waste streams are commonly combined into a stream termed l'centrifuge tailings"
and may be further combined with the primary and secondary tailings to form
whole tailings; alternatively, the centrifuge tailings may be separately
treated to recover contained hydrocarbons. The main product from the
centrifuge circuit is relatively clean bitumen which is in condition to
be upgraded to yield a variety of useful hydrocarbon products.
The hot water process is efficient and has the advantage of
operating under mild conditions. A disadvantage is the production of a
large volume of solids-laden, aqueous tailings. Provision must be made
for storing these tailings and, at least as an intermediate step, they
must, according to present practice, be impounded within dykes that
have to be constructed near the mine area. Such tailings ponds bring
undesirable environmental effects and cover tar sand that is thereby rendered
unavailable to mining.
One extraction operation, producing 120,000 bbl. synthetic
crude per day will, over the 25 year life of the project, create a tailings
pond of around 10 square miles in area. If, as other extraction facilities
are built, the same area of tar sand is covered, the whole deposit is
significantly reduced in size.
In the tailings pond, the solids are supposed to settle to
leave a layer of clear water which can be recycled to the extraction
prucess. Once enough fresh water has been taken on board, an extraction
plant which is self-sufficient in clarified water may be obtained by
this recycle process.
In practice, the coarse solids (i.e. the sand grains) do
settle rapidly, but the fine solids (i.e. -4¢ micron fraction) settle
only slowly over a period of several years.
11315'7~
Tar sand may be viewed as coarse sand grains, each such
grain being covered by an envelope of water, the i~nterstices between
the coated grains being filled with bitumen and fine clay particles.
Generally, as the bitumen content of the tar sand decreases, the clay
content increases. The sand offers little trouble in either the
extraction process for bitumen or the tailings filtration process for
tailings management. However, the clay interferes with both.
It is a characteristic of clay particles that they carry an
electric charge. Hence clay particles suspended in water will settle
only very slowly, if at all, because the effect of gravity on such
minute particles is weak and because the particles are mutually repulsive.
Workers in the field have turned for guidance to the water
clarification art, where large volumes of water are purified for con-
sumption as domestic water. As a result, it has been proposed that coarse
solids be first settled out and the fine matter subsequently flocculated
by conventional flocculating agents. The flocculated solids could then
be removed from the water by centrifuging or filter. Typical examples
of this approach are given in U.S. Patents 3,487,003 and 3,502,575,
issued to Baillie et al and Hepp et al, respectively.
A new process is the subject of patent application
Canadian serial number 311,696 (U.S. serial number 947,996) filed by
J. K. Liu et al., whose assignees are the same as for the instant
invention. The Liu process is based on the discovery that, when flocculant(s)
are added to whole tailings (i.e. with the solids unremoved), the coarse
particles provide nuclei to which the fine particles can adhere. What
is produced is a precipitate of aggregates of coarse and fine solids. In
this state, the mineral matter can be filtered out without the need for
a filter aid. There results a filter cake, small in volume and with little
water, and a filtrate sufficiently clear for immediate recycle to extraction.
The cake is easy to dispose of because it can be compacted to form a base for
reclaimed land. Thus the cake can be disposed of in the mined-out area.
113157;~
The preferred flocculant in the ;nvention of Liu et al is
lime. This is readily available in tar sand regions and does not leave
dissolved residues that could be harmful when recycled to extraction.
The products of decomposed lime are an insoluble carbonate, and one
molecule of water for every reacted molecule of lime. The present
invention, that is now summarized, teaches how to optimize the quantity
of flocculant used.
SUMMARY OF THE INVENTION
The present invention is an improvement on the aforementioned
Liu et al process. In that process, a flocculant, lime is added to the
whole tailings. This leads to co-flocculation or agglomeration, with coarse
sand grains evidently serving as nuclei for the fines to attach to. As a
result, the fines settle with and are distributed in the coarse solids
when the mixture is subjected to filtration with suitable means, preferably
a vacuum filter. In this manner, blinding of the filter cake by a
concentration of fines is avoided and thus effective filtration may be
carried out. This process comprising contacting whole tailings with lime
to effect agglomeration and then filtering the product to separate liquids
and solids, is referred to hereinafter as the "filtration process".
It is the object of the present invention to provide a
technique for optimizing the quantity of flocculant added in the filtration
process. Optimization is achieved for the purposes of this invention when
filtration rate is substantially maximized.
The present invention is based on the discovery that there
is generally linear relationship between the quantity of lime, needed
to give optimum filtration rate in the filtration process, and the fines
content of the tar sand fed to the hot water extraction process.
Surprisingly, the other species of materials present in the whole
tailings - some of which are charged particles - do not affect or
distort the relationship between fines content and the quantity of lime
1~31~7~
needed for maximum filtration rate~
Broadly stated, the invention is an improvement in the hot
water extraction process for recovering b;tumen from tar sand wherein
a coarse and fine solids-laden aqueous whole tailings stream is
generated as a waste product. The improvement comprises monitoring the
fine solids content of the tar sand being fed to the process; treating
the whole tailings with lime to agglomerate fine solids with coarse
solids and to settle the agglomerates, and correlat;ng the amount of lime
used with the fine solids content of the tar sand to optomize subsequent
filtration; and filtering the entire product to separate the water and
solids.
DESCRIPTION OF THE DRAWINGS
Figure 1 is a plot showing how filtration rate and filter cake
moisture vary with different levels of lime addition in the agglomerating
step;
Figure 2 is a plot showing how filtration rate varies with lime
addition in the agglomerating step;
Figure 3 is a plot showing the linear relationship between
optimum lime concentration and fines content of the tar sand - the curves
were developed using tailings obtained from the hot water process fed low
(8 wt. %) fines and high (18 wt. %) fines tar sands feedstocks at different
levels of caustic addition;
Figure 4 is a schematic of the hot water extraction process;
and
Figure 5 is a fanciful drawing showing: (a) typical whole
tailings coarse particles; (b) typical whole tailings fines particles;
and (c) typical agglomerate particles produced by flocculation.
- 1131574
DESCRIPTION OF THE PREFERRED EMBODIMENT
As previously stated, the present invention i`s an improvement
on the known hot water extraction process for recovering bitumen from
tar sand.
As one step of this improvement, one mon;tors the fine
solids content of the tar sand being fed to the extraction process. By
"monitors" is meant that a measure of the fines content of the tar sand
is regularly established. This may be done by one of several ways. For
example, one may analyse the particle size distribution of the solids in the
tar sand using a Microtrac* particle size analyzer or a series of sieves.
"Fines" or "fine solids" are those particles in the -44~ fraction.
Alternatively, one may use the correlation that exists between the fines
content in the tar sand and some of the process streams of the hot water
extraction process. For example, a measure of fines content in the tar
sand may be established by measuring fines content in primary separation
vessel middlings or secondary tailings. Their measure may, for example,
be taken by determining the density of a middlings sample after settling
it for one minute. From the flow rates of the various streams in the plant,
one may then relate the measured fines to the fines content in the tar sand.
In another step, whole tailings from the extraction plant are
treated with lime to agglomerate the greatest part of the fines solids
in said tailings with coarse solids and to settle the agglomerates.
This may be achieved by making up an aqueous slurry containing
about 10% by weight lime in water and mixing it with the whole tailings.
In the laboratory we take a sample of whole tailings into a glass jar or
beaker and weigh the sample by difference. To test the effect of different
* trade mark - Leeds and Northrup, North Wales, Pennsylvania
1 1 3 1 5~
levels of flocculant addit;on, the amount of lime to give final lime dosages
- of from 300 to 700 parts per million is calculated. With the tailings
stirred at 600 revs per minute the desired amount of lime is added either as
a solid or as a lO% by weight slurry. After the flocculant addition,
stirring is continued for a further 2 minutes. The mixture is then
poured into the funnel of a leaf filter test apparatus, described below.
In the plant, one may add ~he slurry to the tailings at a pump box and
pump them together through a line to the vacuum filter. While not critical,
a pumping-mixing time of l-l/2 minutes is desirable. Severe agitation
should be avoided, as the agglomerates are relatively fragile and can break
up, to the detriment of the process.
During this operation, the greatest part of the fines agglomerate
with coarse particles to form agglomerates, a representative sample of which
is shown in Figure 5.
In the last step of the process, the treated product is
subjected to vacuum filtration to separate the liquid and solids. This
may satisfactorily be done with a conventional belt vacuum filter. Good
results have been obtained using a nylon belt having pores of 250 m;cron
diameter. It is desirable to clean the cloth in its passage around the
belt guide rollers to remove adhering bitumen . This may be done by
flushing the underside of the cloth with hot water.
The work done has demonstrated that, by the practise of this
process, the greatest part of the fines do form agglomerates with coarse
particles. These agglomerates settle at a rate comparable to that of
the coarse solids and thus the fines are found distributed fairly uni-
formly throughout the filter cake. The filter cake obtained is relatively
dry, typically containing in the order of about 15% by weight water. The
filtrate is relatively low in solids, typically containing 4% by weight
solids and less than 3% fine solids.
113157~
In accordance with a feature of the invention, the amount of
lime added is correlated with the fine s-olids~ content of the tar sand~
to optimize subsequent filtration.
In a series of laboratory runs, samples of whole tailings
were mixed with varying amounts of lime slurry and filtration efficiency
was tested. More particularly, the mixture samples were each poured into
a funnel lined with 100 mesh filter cloth. The cloth in the funnel
was supported on coarse mesh having openings of about 1/4 inch. The
funnel was drained into a vacuum flask with a side arm at which vacuum
of known force could be applied. The start of liquid being pulled through
the cloth was timed and a further time reading was taken when the surface
of the filter cake was first seen to be dry. The cake was sucked dry
for a further 2 minutes, this being found adequate to remove the free
water. The residual cake was then tested for remaining water (by drying
a known weight); the filtrate was tested for levels of solids (by further
filtration through fine mesh, and drying and weighing resultant solids),
and the filtration rate was known from the time measurements.
A number of samples of the same whole tailings were mixed
with varying amounts of lime and tested as previously described to develop
the curves of Figure 1. These curves show that there is an optimum
lime addition for that tailings composition, at which filtration rate
is a maximum and moisture in the filter cake is minimal. The curves of
Figure 2 show that filtration rate is affected by fines content of the
tar sand, but there is an optimum lime addition for maximum filtration
rate. Figure 3 shows the generally linear relationship which exists
between the amount of lime required for maximum filtration rate and the
fines content of the tar sand. The curve of Figure 3 can conveniently
be used to control lime addition in the filtration step as the type of tar
sand feed varies at the front end of the extraction process.
-- 10 --
