Note: Descriptions are shown in the official language in which they were submitted.
~L'~L~L9 ~8 ~
BACKGROUND OF THE I.NVENTION
This invention relates to an improvement in the treatment
of whole tailings produced by the hot water extraction process for
recoveri.ng bitumen from tar sands
The urgent need for alternative sources of hydrocarbons,
especi.ally for use as fuels, is i:ncreasi'ngly~apparent as reserves of son-
vential crude oi:l becomes us.ed up. An extensi:ve source of hydrocarb.ons. i.s
in the bitumi:nous sands found in vari'ous; parts of the world. Particularly
useful deposits of bitumi.nous sands are in Western Canada, where they are
commonly known as tar s.ands.
~ uch. tar sands are often found near ground surface, thus they
can be mi.ned and transported to an extraction plant for recovery of the
heavy oi.l (bitumen).
The only commercially us.ed process for recovery of bitumen
from mi.ned tar sand is the hot water extracti:on process. In accordance
wi'th this process, the tar sand is -fed i:nto a rotati.ng conditi:oning vessel,known as a tumbler, and mixed wi:tK.hot water, steam, and small amounts of
process ai.d. The most common process: ai.d is sodi:um hydroxi:de; its purpose
i.s to assi.st i.n causi.ng the b.i.tumen to b.e released from the other consti.tuents
of the tar s.and mas.s. The proces is run so that tar sand takes less than
10 mi.nutes to pas.s. th.rough the tumbler. A s.lurry emerges from the tumbler.Thi.s~ s.lurry i.s screened to remove overs;i:ze matter, such as rocks or un-
diges.ted lumps of tar sand, and then i:s di'luted with add;t;onal hot water.
The screened, di.luted slurry i.s then advanced to a quiescent ~one known as.
the primar~ separati.on ves.sel (PS.V~. Because the components o-f the slurry
are in only loose association (as a result of the conditioning in the
tumbler), they are able to separate in the PSV under the influence of
gravity. Hence the s.and~ which has a gravi:ty of a~out 2.65, si.nks to the
bottom of the PSV, and may be pumped out i;n the form of an aqueous mi.xture.
The bitumen has. a dens.i.ty of around 1.00, that i.s, close to that of water.
- 2 - ~
IS
Left to itself, it tends therefore neither to sink nor to float. But
a considerable number of air bubbles have b.een introduced into the slurry
in the tumbler and these attach themselves to the bitumen globules~
In this way the bitumen, i:n an aerated state, rises to the top of the
PSV and may be collected in the form OT a froth. ~ome bitumen fails to so
ri.se because the s.ize of the glohules is too small or because of -failure
to get aerated. This unaffected bi:tumen remains i`n the central region
of the PS~ and helps make up a porti:on known as the "mi:ddlings". To
i.ncrease the effi.ciency of the proces:s~ a middli`ngs stream is continuously
10 withdrawn from the PSV and advanced to induced ai:r flotation cells where,
by vigorous agitati.on and the addition of external air, a second yield
of bitumen i.s. obtai.ned in the form of a secondary froth. The froths
are then combined and the bitumen ;;s separated from the contami:nating water
and mi.neral s.olids. Th.is i.s. don~ by diluti`ng the combined froths with a
15 h~drocarbon diluent and separati.ng the diluted bi:tumen using separatory
means s.uch as centri.fuges. The diluent may then be di:stilled out of the
hydrocarbon product phase to leave pure i:solated bi`tumen that may then
be upgraded by processes known i:n the heavy oil art.
The hot water process. is ef~i:ci:ent and has the advantage of
20 operati.ng under mi.ld condi:tions:. A dis:advantage is the producti.on of alarge volume of soli:ds-laden, aqueous. tai:li:ngs. Provision must be made
ror storing these tai:li.ngs. and, at least as an i.ntermediate step, they
mus.t, accordi.ng to present practice, be i:mpounded within dykes that must
be constructed near the mi:ne area. S.uch` tai`li`ngs ponds bring undesirable
25 environmental effects, and cover tar s:and that is thereby rendered un-
avai.lable to mi`ni.ng.
One extraction operation, producing 120,000 bbl. syntheti:c crude
per day will, over the 25 year life.of the project~ create a tailings pond
of around 10 s.quare mi:les i.n area. I`f, as other extraction facilities
30 are b.ui.lt, the same area of tar s.and i~s; covered7 the whole depos.it is
signi:fi.cantly reduced i.n size.
9~3S
In the tailings pond, the solids are supposed to settlP to
leave a layer of clear water which can be re-cycled to the extraction pro-
cess. Once enough fresh water has: been taken on board, an extraction
plant which is self-suffi:cient in clarified water may be obtained by this
recycle proces;s.
In practice, the coarse S;Ql ids. (i:.e. the sand grains) do
settle rapi:dly, but the fine s.olids (i.e. -44 mi:cron fraction) settle
only slowly over a period of several years. It would be a great advantage
to the tar s.and i.ndus.tr~ f a feasib.le, rapi'd water clarificati:on process
could be devi:sed.
Workers.i:n the field h.ave turned for gui:dance to the water
clarifi.cation art, where large volumes: of water are puri:fied for consumption
as~ domesti:c ~later. As a result9 it has been proposed that coarse soli.ds
be fi.rs.t settled out and the fi:ne matter s;ubsequently flocculated by
conventional flocculati:ng agents. The flocculated solids could then be
removed from the water b~ centri:fugi.ng or filter;:ng. Typical examples
of this approach are given i:n U.S. Patents 3,487,003 and 3,502,575,
i.ss.ued to Bailli:e et al and Hepp et al , respecti:vely.
A new.proces.s is the subject of patent appli:cation Canadi`an
serial numher 311,6.9.6, ~n~O_S.~ o~u~e~ C~ , fi:led by J. K. Liu
et al whos.e assi.gnees. are the same as. for the i:nstant inventi:on. The Li.u
process ;:s. bas.ed on the di:scovery that, when ~locculant(s) are added
to whole tai:li.ngs (i:.e. with the s;oli`ds unremoved), the coarse parti:cles
form nuclei. to whi:ch the fi:ne parti.cles can adh.ere. What is produced i:s
a preci.pitate of aggregates oF coarse and fi.ne soli'ds. In thi.s state, the
mi.neral matter can be fi.ltered out without the need for a filter ai:d.
There results a fi:lter cake, small in vo'lume and wi:th little water, and a
Fi.ltrate suffi:ci:ently clear for immediate. recycle to extraction. The cake
is easy to di:spos.e of because i:t can be compacted to form a base for
reclai:med land. Thus the cake can he di`s;posed of in the mined-out area.
~9~ S
The preferred flocculant i;n th.e inventi.on of Liu et al is
lime. This i.s. readily available i:n tar s.and regions and does not leave
di.ssolved resi:duès that could be harmful when recycled to extraction.
The products of decomposed li:me are an i`nsoluble carbonate~ and one
molecu'le of water for every reacted molecule of li`me. The present
i.nventi.on, that i:s now s.ummari.zed, teaches how to optimi.ze the level of
flocculant needed.
~UMMARY OF THE INVENTI:ON
To bring whole taili:ngs to a state o~ preferred readines.s
for filtrati.on, flocculant i.s added to the ~hole tailings. This leads to
co-flocculati:on, with the large-si:zed grai:ns: evi:dently serving as nuclei.
for the fine soli:d parti.cles to attach to. At optimum flocculant dosage
the effecti.venes.s. of filtrati:on i:s maximized, leadi'ng to rapid filtration
rate, lo~ cake moi.s.ture, and low solids; levels i.n the filtrate. Too little
or too much flocculant prevents the f;:ltration effectivenes:s bei:ng at the
maxi.mum. Further, too much flocculant i`s. wasteful of chem;:cals, and at
the huge ~olumes of tai.lings involved i:n an extraction plant, this could
represent seri.ous. economi.c cost.
Accordi.ng to the i:nventi.on, flocculant dosage is controlled
i.n response to the tai:lings. zeta potential. Before Plocculating, the
taili.ngs. have a negative zeta patenti:al. As flocculant is added, the
zeta potenti.al ri.ses. At the isoelectric point where zeta potential i:s
s.ubstantiall~ zero, it has been found that the taili:ngs are in the most
desi.rable state for ~iltration.
Therefore, in accordance with the i:nvention, the zeta patential
of the tai.lings is. moni:tored as the flocculant i:s added to and mixed
with the whi.le taili:ngs. When the zeta potenti'al is about zero, the
addi.ti.on of flacculant i:s: termi.nated and the flocculated taili.ngs are
treated, as b~ vacuum filtration, to produce a substantially water-free
soli:ds. phase and s.ubstanti.ally clarified aqueous; phase.
9~S
Zeta potential may be monitored by meas.uring the rate at
which a fine particle in suspension travels between two electrodes. This
rate depends on the charge of the particle and it is this charge that is
known as zeta potentia'l. ~e have further found that, although the
tailings are treated i:n the whole state with flocculant, the zeta
potenti:al of fi:ne particles alone i:s i:ndi:cative of the closeness of the
stream to maximum fi:ltration effectiveness. When whole tailings are
treated wi.th li.me flocculant, almost i:mmediatel~ the soli'ds co-flocculate
and visibly begi.n to settle. A s.uhstantially clear supernatant li:quor i.s
left although thi.s liquor does contai.n some finely sus.pended fihes. The
flocculant may thereaFter be added in;stages.; aFter each additi`on a
sample of the supernatant may he withdrawn and subjected to zeta potential
analysis. The zeta potential i:ncreases from a larye negative value to
zero and then increas.es: on the.posi:ti:ve s.ide. At the point of zero
potential of the liquor containi.ng the Fi:nes, the whole tai:li:ngs stream i:s
i.n the preferred conditi.on for filtrati.on. I:t is not necessary therefore
to devise a test to attempt to determine ze.ta potenti;al of large flocs
of co-preci.pitated solids, since the zeta potenti`al of fi.nes alone
indi.cates the zeta potent;al state of the co-precipitated matter.
Broadly stated, the i.nvention i.s. an improvement in the hot
water extracti.on proces.s wherein a soli.ds-laden aqueous whole tailings
stream is generated as. a waste product compri:si:ng: treati:ng said whole
tallings. stream wi.th f'locculati.ng agent and moni.tori.ng its zeta potenti.al;
controlli.ng the amount oF flocculati.ng agent added so as to raise the
zeta potenti.al of the whole tailings~ from an i:niti:al negative value to ahoutzero; and then separati:ng the cons.ti.tuents; of such stream to produce a
substanti.ally water-free soli.ds. phase and a suhstantially clarified
a~ueous phase.
DESCRIPTION OF T!-IE DRA~IINGS
Fi.gure 1 is a plot of four vari.ables aga.inst lime added. The
variables we.re: zeta potential, filtrati:on rate, cake moisture, and
501 ids i:n fi:ltrate.
Figure 2 is a plot of zeta potential versus lime added for
low fines and average fines tar sands.
Fi'gure 3 is a schematic representation comparing the prior
art with the present invention.
DESCRIPTION OF T~E PREFERRED'EM~ODIMENT
The feed for this i`nvention is whole tailings from a tar sand
hot water eYtraction process;.
The preferred flocculant is lime. Lime is readily available
in tar sand reg;ons in the form of limestone~ It i`s a well known pro-
cess to convert this to lime by calcining. I:n use, the lime is mixed
witn water to form a slurry and it is thïs slurry t6at i`s added as the
flocculati~g agent. We have used a slurry containing a60ut lQ wt. % lime.
Since the filtrate is to be recycled to extraction~ the flocculant must
be chosen from substances that will not allow build-up oF species damaging
to extraction. Lime has the advantage that the ultimate reaction products of
any excess of lime with carbon dioxide in the air are calcium carbonate,
which has a low solubility product, and one molecule of water for every
molecule of lime. Thus lime does not create compounds which de'leteriously
affect extraction processing.
The data plotted in Figure 1 was developed using a Zeta-Meter*
on samples of a clarified aqueous phase derived from settled whole tailings
from a tar sand hot water extraction process. The sample was introduced
into an electrophoresis cell and a suitab'le direct current voltage
applied, usually between 100 and 400 volts. The time required for a colloid
particle to kraverse a fixed distance ;n the direction of the anode or
cathode, as viewed under a microscope, was measured. The observed time was
then converted to electrophoretic mobility.
The particular procedure used was as follows:
~* trade mark
9~
1. Approximately 50 mL of sample was required to fill the
cell. If samples appeared di.rty, as was the case with
fine clay slurries, such samples were centrifuged at
15,000 rpm for 20 mi.nutes and the clear liquid decanted
for analysi:s. Care was taken to ensure that the sample
~as at room temperature and the sample temperature was
recorded.
2. The cell holder was posi:ti:oned on the mechanical stage
of the mi.croscope. The cell, fi:lled with the sample, was
placed on the holdèr. The stage was. then adj~sted to
posi.tion the center of th.e cell tube di`rectly beneath the
opti:cal axi:s of the microscope.
3. A thin heam of light ~as focused downward through the
plate glas.s cell holder to îts mi:rror back. This beam was
reflected upward and pas.sed thorugh the cell, causing
reflecting colloids to ~e seèn as tiny rays of li.ght.
4. The microscope was then focused so that the positioning
line of the cell was disti:nct.
5. A dlrect current voltage of lQ0 to 40.0 V was appli:ed to
the cell. If the particles had no charge, they ~ould
remain stati:onar~; negati.ve colloi:ds would mi:grate to-
wards the anode and posi:tively charge colloids woulcl
migrate to the cathode.
6. Di.s.crete particles on or near the counti.ng line were
timed i.n thei.r travers;e of one or more ocular micrometer
divi.sions using the manual timer. Fi:ve or ten discrete
particles were timed i;n thei.r normal direction of travel.
7. The voltage appli.ed across. the cell was recorded, as were
the objective magni.fi:cati.on and the fi:nal temperature
of the sample.
ohjecti.ve magnificati.`on x voltage
-- 8 --
The following precautions. should be noted in carrying out
the test. If the voltage is too hi:gh, the samp'le temperature may rise,
causing particles to describe a s-pi:ral rather than horizontal path.
If this occurs, reduce the applied voltage. If the voltage is too low,
the particles will settle out and be lost From view. For precise work
one should select an opti`mum voltage and track several particles consecu-
tively and then take an average.
Fi:ltration effi.ci.ency was tested by leaf filter tests.
According to this test, ~hole tai:l;ngs w.ere mi:xed with the quantity of li.me
heing i.nvesti.gated and the mixture poured i.nto a funnel li.ned with filter
cloth of the type that would ~e used i`n a cont;'nuous operation. We
use 100 mesh U.~. standard si:eve. The cloth i:n the funnel was. supported
on coarse mesh ~i:th openi:ngs of about ll4 i:nch. The funnel was drai:ned
i.nto a vacuum flas.k with. a si.de arm at wh.ich poi:nt vacuum of known force
could be applied. The start of l;`qu;'d bein~ pulled through the.cloth was
t;.med, and a further time readi.ng was taken w.hen the surface oF the cake
was. f;rst seen to be dry. (Th;s occurred qui:te suddenly and is readily
repeatable.) The cake ~as s.ucked dry for a further 2 m;nutes, thi.s
b.ei.ng found adequate in all cases. to remove free water. The res.i:dual
cake was then tested for remai.ning water ('b.y dryi`ng a known weight); the
fi.ltrate was. tes.ted for 'level of sQli.ds. (b.y further fi:ltration through
f;ne mes.h, and dryi:ng and we;.ghi.ng resultant soli:ds); and the fi'ltration
rate was known From the ti'me meas:urements:.
Turning now to fi:gure 1, ;'t shows that as the zeta potent;al
approaches zero from the negat;.ve s;;:de, cond;.t;.on;ng of the ta;lings For
filtrat;.on reaches max;:mum ef~ect;:veness:. B:eyond th;'s, lime is added
unnecessar;'ly.
In the hot water extra.ct;~on proces.s, as the level of fines
increases in the feed, the dQsage of s:odi:um hydroxi:de (added as; a ~rocess
ai.d) mus.t li:kewi.se be i.ncreas~ed to mai:ntai.n the oil recovery at a maximum.
S
It might be expected that this i.ncrease of sodi.um hydroxide would affect
the quantity of flocculant needed to treat the tail;ngs. It was our
surprising discover~ that, alth.ough.tai:lings hi.gh i:n sodium hydroxide
began at very negati:ve values of zeta potential, li.me addition rapidly
brought that potenti.al to zero. On the other hand, tailings from the
s.ame type of tar sand, but low i`n sodium hydrcxi:de, while starting at a
les.s negati.ve zeta potenti:al, h.ad a slawer res:ponse to lime. That i:s, fortar s.and of the s.ame fi:nes level, ~u~ treated wi:th different amounts of
s.odium hydroxi.de, the slopes of the curves:linking sodium hydroxi:de to
li.me vary such that the effect of s.adium hydroxi.`de is largely eli:mïnated.
Thi;s effect i.s shown i:n Fi:gure 2 whi`ch plots: zeta potential
versus li.me added for a la~. fi:nes. and an average fi`nes tar s.and. For
th.e low fi.n¢s. runs, the lime neede.d to bring th.e zeta potenti:al to zero was.
bet~een 300 and 350 mglkg. For the ave.rage. fines runs the dosage was
800 mg/kg. Different leve1:s. of sodium hydroxi:de added to the extracti.on
process caused the ini;ti:al zeta potenti:al to vary from -23 to -42 mV
but at the i:soelectri:c poi.nt the curves: tended to collect together.
Probably sodi.um hydroxïde i:n extracti:on and li:me in
. flocculation operate by reverse mechani`sms... Extraction process aïds.
deflocculate the clay matter i:n tar s:and and hence encourage releas.e of
the entra.pped bi.tumen. I.t i:s possible that this is an i:ndi`rect mechanlsm
whereby s.urface acti:ve agents~ are fi:rs:t produced between the sodi:um
hydroxi~de and naturally occurri.ng organic aci:ds i`n the tar sand ~i:tumen
and these operate on the clay. The aim of the lime by contrast is to
re-flocculate the clay matter, in our case by coflocculation with the coarse
so1ids. It might be expected therefore that,for the same tar sand,
different levels of sodium hydroxide ~ould require varying amounts of lime.
Our finding that lime requirements are es;sentially independent of sodium
hydroxide content and depend solely on fines content, is an unexpected and
welcome simplification of tar sand tailin~s. m~nagem~n.t.
-- 10 --
No test has been devised to measure the zeta potential of
flocs of co-precipitated tailings solids. It is a fortunate discovery
therefore that the zeta potential of the fines, left in suspension after
the majority of the solids are preci:pi:tated~ may be used to indicate
the zeta potential of the whole tai'li'ngs. Hence, when the zeta potential
of the fines alone is brought to zero, the preci:pi`tated solids are in the
ri.ght condi.tion for vacuum filtrati.on. A convenient point to measure the
potential o~ the fines i:s the fi:ltrate. The small amount of fines
reporting to the filtrate i.s suffi:ci:ent to allow the required zeta
potenti.al measurements to he made. Flocculant slurry may then be
added to the w:hole tailings to give a filtrate potenti:al of zero. In a
conti.nuous pracess batch portions o~ the fi:ltrate can be withdrawn for
testing or a small si.de stream may be di`verted.
Fi.gure 3 i:s a fanciful represèntati.on of tailings management
by the prior art and as taught i.n the pres:ent inventi'on. In the prior
art, whole tailings. are simply allowed to s.ettle under the i.nfluence
of gravity. Although the coarse s.olids settle qu;'ckly the sludge of
f;ne sol;ds and clays settles only ovèr a peri:od of several years
due to the mutual repuls~on oF the parti.cles stabi`lizi:ng the s.ludge.
This slow settling calls for large tai:li`ngs; ponds:.
Accordi:ng to the present i:nventi:on, the whole tailings are
flocculated and may be subjected to s:ome rapi:d separatory means such as
vacuum filtrati.on. Although i:t i:s the condi:ti'on of the co-flocculated
soli.ds that determines optimum fi:ltration, the zeta potential need be
meas.ured only on the fi~nes. i:n the clear aqueous layer. Places where zeta
potenti.al may he canveniently measured are marked "z.p.".
I.t i.s to be understood that the separatory means may i`nclude
mere gravity settl;ng of the flocculated whole tai:li:ngs. If the soli.ds
are to be formed i:nto a heach or layed down by such techniques as cone
ridg;:ng, already taught i'n the mi:ning tai:li:ngs: or dyke building arts,
the inYenti.on may he practi.sed i.n connecti:on with the tai:lings.
