Note: Descriptions are shown in the official language in which they were submitted.
33~0~
FIELD OF THE INVENTION
2 This invention relates to treatment of desanded aqueous tailings
3 produced by a plant using a water-based extraction process to recover bitumen
4 from oil sand. The treatment involves adding calcium sulphate to the desanded
tailings to accelerate the settling rate of fine solids suspended in the tailings water.
6 BACKGROUND OF THE INVENTION
7 Oil sand comprises sand grains that are water wet. Bitumen, a form
8 of oil, is present as a continuous matrix in which the coarse, wet, sand grains are
9 ~ edd~d. Clay particles, referred to as "fines", are entrained in the very thin
water sheaths encapsulating the sand grains. The fines will pass through a 22
11 mesh screen. They include minute particles, termed "ultrafines", which are less
12 than 300 nanometres in diameter.
13 Oil sand is present as an enormous deposit in northern Alberta. This
14 deposit is referred to as the Athabasca deposit.
The bitumen is extracted from the Athabasca oil sands by two large,
16 co",l"e,.iial plants. One such plant is owned by the present assignees and
17 processes about 350,000 tonnes of oil sand per day.
18 Applicant's plant uses what is referred to as the Clark hot water
19 extraction ("CHWE") process, to separate and recover the bitumen from the oil
sand. This process is described in detail in the literature. In broad outline, it
21 involves:
3 ~ ~ ~
Co~ ioni"y the oil sand by mixing it with hot water and a
2 small amount of caustic in a rotating horizontal drum (termed
3 a "tumbler"). Steam is sparged into the slurry to ensure that
4 its exit temperature is about 80C.
In the course of col1diliu"i"9 the viscous bitumen is heated
6 and is separated from the sand grains; it is released into the
7 water phase in the form of minute flecks. At the same time,
8 small air bubbles are entrained in the slurry. Fine bitumen
9 flecks coalesce and form larger globules that contact and coat
air bubbles thereby becoming buoyant;
11 The product slurry leaves the tumbler and is screened, to
12 remove oversize material;
13 The screened slurry is then diluted with additional hot water to
14 produce a slurry containing about 50% solids by mass based
on the original oil sand feed;
16 The diluted slurry is retained in a large thickener-like vessel
17 (called a "PSV") for about 45 minutes. In this vessel the
18 aerated bitumen rises and is recovered as an overflow
19 "primary" froth product. The sand settles and leaves the
2û vessel as an underflow stream C~ plisil~g water and some
21 bitumen. In the mid-section of the PSV there exists a watery
22 mixture cu",,,~risi,~g relatively non-buoyant bitumen and fines -
23 this mixture is referred to as "middlings";
a~
A stream of PSV middlings is mixed with PSV underflow and
2 the mixture is introduced into a deep cone settler, referred to
3 as the tailings oil recovery vessel ("TORV"). In the TORV, the
4 feed mixture is deflected radially as it is fed in and is spread
outwardly and hori~ontally. The out-moving mixture is
6 contacted from below by an upwelling stream of aerated
7 middlings. A secondary yield of froth is produced. The
8 underflow from the TORV, Cu~ ising solids, water and some
9 bitumen is ~ l,aryed as a tailings stream;
10 A stream of middlings is withdrawn from the TORV and is fed
11 to a bank of sub-aerated flotation cells. Here the middlings are
12 subjected to relatively intense aeration and mixing. Bitumen
13 contained in the middlings is recovered in the form of a
14 secondary froth. The underflow streams from the flotation cells
are ~ l,aryed as tailings;
16 The tailings from the TORV and flotation cells are combined to
17 yield a stream referred to as "whole tailings". This stream
18 typically comprises 0.5 wt. % bitumen, 44.9 wt. % water and
19 54.6 wt. % solids;
The whole tailings are, in partm~ dl~ d to a pond, which is
2 enclosed by constructed dykes. More particularly, the whole
3 tailings are discharged onto a sloping l'beach" at one end of
4 the pond. As the whole tailings fan out across the beach, most
of the sand quickly settles out and joins the beach. In the
6 course of this, some water and fines are trapped by the sand
7 forming the beach. The remaining tailings, referred to herein
8 as "clesd,~ded" tailings, join the pond contents. The other
9 portion of whole tailings is ~ ;l,a~yt:d into a remote sand
dump in which the sand settles out. The desanded tailings are
11 then returned to the previously described pond. The desanded
12 tailings typically comprise 0.6 wt. % bitumen, 85.4 wt. % water
13 and 14 wt. % solids. The solids are typically less than 22
14 microns in size. The desanded tailings are fed in at one side
of the pond and clarified water is recycled to the plant from the
16 other end, for use as process water.
17 The desanded tailings settle very slowly to produce clarified tailings
18 which can then be recycled to the process. Typically 1-2 years are required to
19 reach the 50% settled point. Very large tailings settlings basins are required
(5,1'),~ pond has an area of 12km2 and a depth of 45m in the deepest portions).
21 If the rate of settling of the fines in desanded tailings can be
22 siy,li~icd"lly accelerated, then a smaller pond could be used.
3 ~c~
It is known from Canadian patent No. 1,103,184, issued to Liu et al.
2 and United States patent No. 4,282,103, issued to Fuhr and Liu, that the addition
3 of CaO (300-700 ppm) to whole tailings causes the fines and coarse particles to
4 agglomerate, resulting in the fines coming down mixed uniformly with the coarse
particles to produce an agglomerate that can easily be filtered without danger of
6 plugging the filter by fines. The amount of CaO preferably required is the amount
7 needed to reduce the zeta potential of the fines particles to zero. For low fines ores
8 the amount of CaO required is from 300-350 ppm, while for average fines ores the
9 amount required was 800 ppm.
The process is costly, due to the large amount of calcium oxide used
11 and the high capital and operating costs for filtration equipment.
12 It is also known from U.S. Patent No. 4,414,117, issued to Yong et al,
13 that the fines in desanded tailings can be made to settle quickly by removal of
14 carbonate and bicarbonate ions from the system. The patentees assert that agents
such as ion exchange resins, lime or other calcium compounds and mineral acids
16 can be used to remove carbonates and bicarbonates from oil sand plant tailings.
17 They further assert that, in the case of lime, an amount of 800 ppm is required for
18 a tailings sample that contained 10.0 meq/L of carbonate plus bi~dlbulldl~.19 Again, the process is costly due to the large amount of lime required.
For lp~ ' plant, which produces tailings typically having 14.8 meq/L of
21 carbonate plus bicarbonate, the calculated quantity of lime is 1184 ppm.
22 The invention described herein provides a process by which the fines
23 in desanded tailings from oil sand plants can be made to settle quickly by the
24 addition of 100-200 ppm of calcium sulphate.
~ 3 ~
SUMMARY OF THE INVENTION
2 The present invention arose from the discovery that tailings from oil
3 sand plants contain dissolved surface active sodium 11dpl,Il,e~1dL~s and sodium
4 sul~onates at very low conc~ dlions (10-15 ppm). Since surface active materials
are known to be able to adsorb on clay particle surfaces to create high negative6 charges on the particles, it was hy,uuII ,esi~d that the na,c l ~ ndl~s and sulfonates
7 in tailings were causing high negative charges in the fines which in turn were
8 causing slow settling due to the repulsion between the negative charges.
g Both napl,Il,~"dIes and sulfonates are known to form insoluble
calcium compounds. Consequently it was I Iypull ,esi~d that the addition of calcium
11 compounds to tailings should precipitate the na~l,Il,~,1dIt:s and sulfonates from the
12 system and reduce the negative charge on the particles thus permitting fast settling.
13 Since the naphthenates and sulfonates were present at only very low
14 cu"c~"I,dtions in tailings, it was further hypull,~si~d that only very small amounts
of calcium compounds would be required to achieve the desired result. Further, the
16 calcium compound would need to be easily water soluble to be able to rapidly
17 achieve the desired result without the need for thorough mixing. (Insoluble calcium
18 compounds would be expected to react more slowly and require extensive mixing
19 to react with the water soluble naphthenates and sulfonates.) Calcium sulphate and
calcium chloride are the only cheap, readily available, water soluble calcium
21 compounds and calcium sulphate is preferred over calcium chloride due to chloride
22 corrosion problems.
g~
On testing, it was found that the addition of 100-200 ppm of calcium
2 sulphate to desanded tailings resulted in acceleldl~d settling of the fines. There
3 was no need to add calcium sulphate in sufficient quantity to remove all the
4 carbonate and bica,uur,dle as in the Yong et al patent or to add enough calcium
sulphate to reduce the zeta potential (here measured as the equivalent
6 ele~l,upl~ort,lic mobility) to zero as in the Liu et al patent. Nor was there any need
7 to use the coarse sand as a filter media to bind the fine particles as in the Liu et al
8 patent.
9 It was also discovered that when at least 100 ppm of calcium sulphate
was added to desanded CHWE tailings, the tailings became easier to centrifuge to11 a cake suitable for disposal.
12 The invention has been stated in terms of a specified minimum
13 amount of calcium compound added. It needs to be ~lld~laluO.;I that oil sands
14 recovered from different localities and depths of the huge Athabasca oil sand
deposit vary si~u",i~il;d"lly in nature and c~",posilion. We have dt:lt""il~ed that
16 some of the added calcium reacts with nd,ul~ dLes and sulfonates to form
17 insoluble calcium salts. Also, some of the calcium reacts with a portion of the
18 bicdruuridLes present in the pond and river water to also form insoluble calcium
19 salts. The net result is that at least 100 ppm is needed as a minimum to remove
the naphthenates and sulfonates. Usually this addition increases the cu,lc~, Ill~Liol1
21 of calcium in the aqueous phase of the treated tailings by a small amount, typically
22 to 3 - 7 ppm; this calcium collct,,,l,dLiull is a marker or indicator that an effective
23 amount of calcium compound has been added, sufficient to produce a marked
24 improvement in the settling rate of the fines. Due to the varying nature of the oil
3 ~
sand species, the calcium compound addition or dosage may have to be increased
2 to, for example, 2û0 ppm, in order to get effective results. The apu,up,i~Le dosage
3 of calcium compound addition can be de~t:""i"ed by testing the oil sand being
4 p,ucessed in acco,.lal~ce with the examples described below.
DESCRIPTION OF THE DRAWINGS
6 Settling rate data in support of Examples I and ll, are illustrated in
7 Figures 1 - 3, 4 - 7 respectively. More s~ lly:
8 Figure 1 illustrates fine tailings settling rates, over time, for OHWE and9 CHWE p,ucesses,
Figure 2 illustrates the effects of ele1l,upl~ lic mobility on OHWE
11 process settling rate;
12 Figure 3 illustrates the effect of calcium content on OHWE process
13 settling rate;
14 Figure 4 illustrates comparative settling rates with and without the
addition of 130 ppm of CaSO4 to OHWE process tailings derived from 8801(3) oil
1 6 sand;
17 Figure 5 illustrates comparative settling rates with and without the
18 addition of 123 ppm of CaSO4 to CHWE process tailings derived from 8801(3) oil
1 9 sand;
Figure 6 illustrates comparative settling rates with and without the
21 addition of 125 ppm of CaSO4 to OHWE process tailings derived from 8806 oil22 sand; and
~f~ 3~
Figure 7 illustrates comparative settling rates with and without the
2 addition of 169 ppm of CaSO4 to CHWE process tailings derived from 8806 oil3 sand.
4 DESCRIPTION OF THE PRE~t~Eu EMBODIMENT
The invention is supported by the following examples.
6 Example I
7 This example reports on tests carried out on tailings produced from
8 a series of extraction runs performed on a group of different oil sands. Each run
9 was carried out in the same laboratory scale test circuit, described below. Charges
of each of the distinct oil sands were treated with each of the following ~ ,uc~sse~.
11 . the OSLO ("OHWE") process using river water;
12 . the Clark ("CHWE") process using deionized water without use
13 of NaOH process aid; and
14 the Clark ("CHWE") process using tailings pond water from
applicant's plant and 0.02 wt. % NaOH process aid.
16 The river water contained about 60 ppm of calcium ion (see Table 1).
17 The deionized water contained negligible calcium ion. The tailings pond water also
18 contained very little calcium ion (5.~ ppm).
,~ 3 ~d
Asses~l"t:,ll of the results from these tests established the following
2 correlations:
3 that the fines present in the aqueous phase of desanded
4 tailings were slow settling and the tailings were toxic when the
ele~,upl~o,~ mobility (equivalent to zeta potential of the
6 tailings) was high and the calcium cùnc~lllldliull in the
7 aqueous phase was low;
8 that the fines were fast settling when the ele~L,upl~o,~lic
9 mobility was low and the calcium ~ul l~ ldlion was high; and
1û that there was no correlation between total
11 carbonate/bicarbonate COnCt,lllldliull and settling rate.
12 The laboratory testing program produced units of product of about
13 1ûOûmL of desanded tailings. Samples of these tailings were placed in graduated
14 cylinders and the position of the interface which developed as the tailings settled
15 was followed over time. Tailings which were arbitrarily coll~id~ d to be fast settling
16 were observed to reach the half-way point in less than a day and to reach a final
17 settled volume of about 30% in 30-60 days. Tailings which were collsi.lt~ d to be
18 slow settling required longer both to get to the 50% point and to the final 30% point
19 (1 year). The time needed to get to the 50% point was used as a comparative
20 measure of settling rates.
21 Electrophoretic mobility (equivalent to zeta potential) is a property
22 indicative of the electrical charge on the particles. It was measured during the
23 testing using a Rank Brothers Zeta meter' .
24 ' ~trade mark
3~
The tests were carried out on tailings produced in the course of
2 bitumen extraction runs using a batch extraction unit referred to as the "BEU". This
3 test unit was been described in Sanford, E. C. and Seyer, F.A., "Fl~o~ ' 'iCy of
4 Athabasca tar sand using a batch extraction: the role of NaOH", Can. Inst. Mining
and Metall., Bull., 72(803) ppl64-169 (1979).
6 The test unit co",plised a jacketed reaction vessel having a capacity
7 of about 1.5 L and being fitted with a stirrer, a system to introduce dispersed air
8 bubbles near the bottom of the vessel and an outlet at the bottom of the vessel
9 which could be opened to drain the whole tailings from the vessel.
To operate the CHWE process in this vessel, the vessel was charged
11 with 1 50mL of pond water at 80C or Dl water at 80C and the appropriate amount
12 of NaOH. The stirrer and air flow were started with water at 80C being pumped
13 through the jacket of the vessel. Air was supplied at a rate of 150 ml/min. Oil sand
14 (5009) was then added and the resultant slurry stirred for 10 minutes. More water
(pond or Dl) (9OOmL) was then added and the air flow was discontinued. The
16 mixture was stirred for 10 minutes and the bitumen which floated during this time
17 was collected. Air injection was then resumed for a further 5 minutes which caused
18 more bitumen to float. This bitumen was also collected and the whole tailings were
19 drained from the vessel. The whole tailings were allowed to stand for 1 minute
during which time the sand settled and the desanded tailings were then decanted
21 to a graduate cylinder.
12
33g~)
To operate the OHWE process the vessel was charged with 150mL
2 of river water at 80C and the ~ ), u~ le amounts of kerosene and methyl isobutyl
3 carbinol ("MIBC") were added. Oil sand (500g) was then added and the resultant
4 slurry stirred for 4 minutes with air flow (420 ml/min.). The heating bath was then
5 turned off and the river water (9009) at room temperature was added. The mixture
6 was stirred for 10 minutes with a small air flow (rate 30ml/min.) and the bitumen
7 which floated during this time was then collected. The remaining material was
8 stirred for a further 5 minutes with a higher air flow (rate 240 mVmin.) and the
9 bitumen which floated was again collected. The whole tailings were drained from
10 the vessel and allowed to stand for 1 minute to allow the sand to settle. The
11 desanded tailings were then decanted to a graduated cylinder.
12 The relevant details of the extraction runs are described in Tables 2
13 3and4.
14 The whole tailings from each run were desanded by settling them for
15 one minute and then pouring off the supernatant as desanded tailings.
16 Samples of the aqueous desanded tailings were then placed in
17 graduated cylinders and p~liodi~ally observed to determine how long it took to
18 settle to the 50% point.
19 The col1c~"l~liol~ of calcium in the aqueous phase of the tailings was
20 d~r",i"ed. This was done by ultra-filtering the tailings to remove insoluble clays
21 and analysing the filtrate by inductively coupled plasma flame photometry.
13
3 ~ C~
The electrophoretic mobility of the fines in the aqueous phase of the
2 desanded tailings was d. " ;",i"ed according to the method described by L. L.
3 Schramm and R. G. Smith in the article "The Influence of natural surfactants on
4 interfacial charges in the hot water process for recovering bitumen from the
Athabasca oil sands", published in Colloids and Surfaces, 14 (1985) 67-85.
6 The total carbonate (carbonate plus bicarL,olla~) co,~ce"ll~lioll in the
7 aqueous phase of the tailings was measured by ultra-filtering the tailings to remove
8 insoluble clays and analyzing the filtrate by titration with ~ dc~ d hydrochloric
9 acid.
The data from these runs is set forth in Tables 2 - 4 and Figures 1 -
11 3.
12 Settling curves were constructed from the measurements on the
13 position of the interface in the settling cylinders vs. time and a typical set of settling
14 curves is shown in Figure 1. These curves were used to determine the time
required for the tailings to settle to the 50% point and the data from the eAuel i" len ,l~
16 are recorded in Tables 2-4. These tables also contain i,,~u,,,,~liùll about the
17 ~A~Jelilll~llldl conditions for each ~Aueli",e"l and also include electrophoretic
18 mobility data and total carbonate col1ce,,1,dliuns.
19 The data indicated that for the runs in which the oil sands were
~,ucessed with the OHWE process tailings samples characterized by an
Zl ele.il,uullor~ mobility below about 2 had fast settling fines while tailings having
22 a mobility between about -2 and -4 had slow settling fines (see Figure 2). In
23 addition tailings having a calcium content below about 7 ppm had slow settling
24 fines while tailings having a calcium CullCdllll~liol1 greater than about 7 ppm had
14
3~
fast settling fines (see Figure 3). There was no correlation between settling rate
2 and total carbonate cunc~"l,d~i~n.
3 The data from the OHWE runs indicated that the results obtained
4 depended on the oil sand used and the presence of calcium in the river water. If
the produced tailings had calcium Cùl1C~ dliull~ greater than about 7ppm, the
6 tailings were fast settling and of low charge. We hy~.ull ,esi~e that such tailings are
7 low in surface active materials due to the high Cùl~Ctllll,dliùn of calcium. If the
8 produced tailings had calcium conce,,l,dliù,~s less than about 7 ppm, the tailings
9 were slow settling and of high charge. We hy~,ull ,esi~ that such tailings, being low
1û in calcium, contain surface active materials.
11 The tailings from runs using the Clark process using NaOH (which is
12 the process used cu"""eruially) were generally slow settling, highly charged and
13 low in calcium. We hypothesize that these tailings are high in surface active
1~ materials generated by reaction between the NaOH and acid precursors in the oil
1 5 sand.
16 The tailings from the Clark runs using Dl water were su",~ es slow
17 senling, highly charged and low in calcium and sometimes fast settling, of low
18 charge and high in calcium. We hypothesize that some oil sands can liberate19 surface active materials when slurried with Dl water even if no NaOH is used.
Other oil sands when slurried with Dl water do not liberate surface active materials.
~1~33~
Exam~le ll
2 This example supports the following findings:
3 that addition to OHWE tailings of calcium sulphate in an
4 amount of about 100 ppm was effective to convert the tailings
from one in which the fines settled slowly to one in which they
6 settled quickly; and
7 that addition to CHWE tailings of calcium sulphate in an
8 amount of about 175 ppm was effective to convert the tailings
9 from one in which the fines settled slowly to one in which they
settled quickly.
11 The eA~.~,i",e"Ldl procedure used was as follows. Tailings from the
12 BEU were drained into beakers and allowed to stand for one minute to settle the
13 sand. The supernatant was decanted to a beaker and allowed to cool to room
14 temperature (about 2 hours). The mixture was then stirred to mix any settled solids
material and a sample (250 ml) was taken. A portion (50 ml) of this sample was
16 used to determine % solids a portion (100 ml) was c~"l,i~uged and some of the17 centrifugate was used for anion and cation analysis while the rest of the
18 centrifugate was ultra-filtered and used for cation analysis. Calcium sulphate
19 hemihydrate was added as a crystalline solid to the remainder of the tailings and
the mixture was stirred for one hour. A sample (100 ml) was removed and
21 centrifuged with a portion of the centrifugate being used for anion/cation analysis
22 and a portion being ultra-filtered and analyzed for cations. The remaining treated
23 tailings were lld~ d to 1 L graduated cylinders which were sealed with sheets
16
g~g~
of polyfilm and stored at room temperature. An interface developed and the
2 position of this interface was followed with time.
3 The measured data on the position of the interface with time for the
4 experiments was used to plot settling curves and three of these are shown in
Figures 4-7. The time required for the tailings to settle to the 50% point was
6 d~lt""i"edforeach~,~pe,i",~"lfromthesesettlingcurvesandthedataisrecorded
7 in Tables 5 and 6. These Tables also contain information about the ~xpelillltllldl
8 conditions and Table 6 contains total carbonate co,)c~ ldliolls and el~ullulJI,oltlli~
g mobilities.
Figure 4 and Table 5 show that addition of 130 ppm CaSO4 to OHWE
11 process tailings acc~le,dled the time to get to the 50% settled point from about 35
12 days to about 9 days.
13 Treatment of CHWE process tailings derived from the same oil sand
14 with 123 ppm CaSO,~ improved the time to get to the 50% settled point from about
29 days to 9 days. See Figure 5 and Table 6.
16 Treatment of the OHWE tailings from oil sand 88û6 with 125 ppm of
17 calcium sulphate dt,u,t:ased the time required to get to the 50% settled point from
18 143 days to 11 days. See Figure 6 and Table 5.
19 Treatment of the CHWE process tailings from the same oil sand with
169 ppm of calcium sulphate reduced the time to get to the 50% settled point from
21 112 days to 9 days. This is shown in Figure 7 and Table 6.
17
3 3g~
There was a very small reduction in total carbonate Col~C~ dliol- on
2 treatment of CHWE process tailings with calcium sulphate (Table 6) (for example
3 from 15.97 meq/L to 15.43 meq/L for the 8801(3) oil sand). Evidently some of the
4 added calcium sulphate is consumed by reaction with carbonate/bicarbonate.
Clearly it is not necessary to remove the carbonate/bicarbonate to produce fast
6 settling. Calcium sulphate caused a small reduction in electrophoretic mobility (and
7 hence in zeta potential) but it was not necessary to reduce the zeta potential to
8 zero.
9 Exam~le lll
This example dt:llloll~l,dltls that the fines in tailings treated with
11 calcium sulphate are easier to COI15 ~" ' ' with centrifugation than untreated fines.
12 More particularly, samples of treated desanded tailings and untreated
13 desanded tailings were centrifuged in centrifuge tubes, starting at 1000 rpm for 30
14 minutes with measurement of light transmittance after centrifuging was complete.
The tubes and samples were then re-~"l,i~u~ed at 2000, 4000, 6000, 8000 and
16 10,000 rpm with measurement of light tral,:,",illdi,ce after each centrifugation.
17 The data from the tests are set forth in Table 7. The light
18 transmittance data and visual inspection indicated that untreated tailings required
19 about 10000 rpm to form a cake, while the treated tailings formed a cake at about
2000 rpm.
18
33~
TABLE 1
2 WATER ANALYSIS
3 lon RiverWater Pond Water
4 C1-,ppm 14.3 136
5 so42+~ppm 37.5 215
6 HCO3-,meq/L 3.34 14.23
7 Mg2+ppm 17 4.1
8 ca2+ ppm 60.2 5 5
9 pH 7.7 8.6
19
c
~ ~ ) C~ ~ ~ ~ ~'7 co a) co u~ ~ Qo
1~ ~ E C~ 7 ~ c~ N C~i N C~ N N C~ i L
' V N ~ _ _ _ ~ ~ ~ V V V V ~ ~ --
.~
5~ ~ C~ 1~ ~ _ ~ o ~ ~ Q a) C00
3 o co
T ~ E co ,~ ~ ~ co ~ N ~ cO c') CO QO --
_
Q NO ON NO o o NO NO o ON NO ~l æ O NO o ON
L
~ ~ F ~ ~o ~o~ ~~ QOo ~- ~ N ~ N NO CO CO Q ~ N -- N
cn ~ c ~ QO tlO U~ ~) 00 ~ ~ 0~ 00 N ~ Qo ~ ~ _
~ ~ ~, oo a) ~ 0--co co ~ co a~ 1~ ~ ~ a~ ~ a) a) a
o Q CO o CO ~ CO 00 CO aO) æ O ~ t ~ N
N C~ ~ ~ QO ~-- CO 5~ o _ _ C~) ~ QO 1~ ~ 5) CU N CU N N N N
. .
~ 3~a
o ~ ~ 8 ~ o ~ o, _ ~ o a, 2D ~ ~ ~
i N . ~ ~ C~l ~ O O C~i
O ~ _ + + + + + + +
. .
.~ -
_ ` ~ ' -- o ~ ~ '~
~ N ~ i N N i . O O C~
8 o o o g 8 o g o o g o 8 o
Z~C ~ oooooooooooooo
~ '
Om~ ~ 0a~--~0coa~_u,~0
o o ~ O N o 1~ 0--~
C~l ~ ~ LO ~D ~ c0 0 _ ~ N IS~ 0 0 NO N N
33~
CD N _ CD O U~ _
- ~ CO ~ C5 ~7 C" t~ N N
C _ _ _ _ _
~ .~ ~
m ~ Q CC\ C'~ N -- -- O -- ~ N
j~ I L ~ ~, N C'~ N
0 ' ' c U~ a) g CO N ~
~ N ~ N N C'~ -- N
I ~ ~
~ O O O O O O O O
Z ~ ~ o o o o o o o o
a) c
, c .~ ~ ~ ~ ~ _ ~ _ a
N CC 'J 1~ _ ~ ~
-- N c~ ~ u~ C~ 3~ O -- N C') ~ 11~ CO
33g~
E ~ ~ NO N ~ ~ ~ ~ ~ _ ~ ~ N C'~ O a~ r-- N -- -- --
N ~
C ~ ~ o ~0 o ~t~ O NO o N ~ ~ O C~ _ N --
.C
-
u~ 3
~ i~ ~ ON æ æ o~ ON CO ~~ ~OD L~O~ N _ _ _ _ _ _ _ N
~ 8 F o O æ o ~0 NO NO NO NO ~ ~t ~ ~ ~ æ
~ S l N N N N N N -- -- N N -- _ -- ~ N N N N N N N N
O m ~ 0 0 0 a) a~ 0 0
-- o o o o O co ~ a æ o~ 0 0 0 0 0 0 8 8
_ N Cq ~ l/~ tD 1~ a~ 0 0 _ N C~ ~ n CD 1~ a~ 0 0 _ N C') ~ ~ CD 1~ ~D 0
-- -- ~ ~ ~ -- ~ ~ ~ ~ N N N N N N N N N N
3~
N ~ O O O 1~ C'~
0 ~ U'~ q N a~ _ N N N C~ a~ tt~ O~ CO 1
, .
L 5 N N C~ N N N C~ N ~7 N C') N ~g N N N
~ O ~ ~ aN~ a~ "N~ ~ N ~D ~ I~ CO N 1~ ~1
a~ ~ .
.' , ~
~ E
1~1 I
m t~
;~ O o
N ~
o ~ ~ g O N O 1~ 0 CO O ~ O U~ O ~ O ~ O ~ O tD
g~
C~ I
~ ~ ~ O O, O O O, O O O, O O, O. O O, O. O O, O, O
Z C~ g o o o o o o o o o o o o o o o o o o
~ s~ ~
0 0 U~ ~ ~ ~ 0 0 ~ OD 1 a~
O ~ o _ ~ ~ 0 0 ~ ~ _ ~ ~ ~ ~ ~ 0 0 ~ ~ ~ 1
~ ~ ~ ~ ' 8 1~ ,` N N -- ~ 1~ ~ 0 a)
O co a~ CO 0
N 07 ~t U~ CO 1` ~0 0 0 ~ N ~ ~t Ul ~C) ~ CO 0 0 -- N c~ ~ u~ CO
~ ~ -- ~ ~ . ~ ~ -- ~ N N N N N N N
3~
0
tD
., .
,a~
U~
s . ~ U~ 10
O O O O O
~.
.
3~ ~ o O 0 0 0 8
;`
- ææ~2~æ
O a~ ~ ~ co ~ CO
~ . ~
