CA1214421A - Blending tar sands to provide feedstock for hot water process - Google Patents

Abstract

"BLENDING TAR SANDS TO PROVIDE FEEDSTOCKS FOR HOT WATER PROCESS" Two or more tar sand feeds are blended in specified proportions in combination with a specified dosage of NaOH to provide a slurry, having a free surfactant content dissolved in the aqueous phase of the slurry, which yields the maximum possible primary froth production from the hot water extraction process.

Description

? ,~ f~ ~"

2 This invention relates to an improvement of the hot water

3 process for extracting bi-tumen from tar sand. More particularly, it

4 relates to process control, specifically control of process aid addition,whereby primary bitumen froth recovery may be maximized, in spite of 6 the changing nature of the tar sand feed.

7 BACKGROUND OF T~IE INVENTION
_ 8 Tar sand, also known as oil sand and bituminous sand, is 9 now well recognized as a valuable source of hydrocarbons. There are presently two large plants producing synthetic crude from the tar sands 11 of the Athabasca region of Alberta. In these operations, the tar sands 12 are first mined and the bitumen is then extracted from the tar sand by a 13 process called the hot water process. The extracted bitumen is subse-14 quently upgraded by refinery-type processing to produce the synthetic crude.
16 The tar sand is a mixture of sand grains, connate water, 17 fine minerals solids of the particle size of clay minerals, and bitumen.
18 It is commonly believed that the connate water envelopes the grains of 19 sand, the fine solids are distributed in the water sheaths, and the bitumen is trapped in the interstitial spaces between the water-sheathed 21 grains.
22 The hot water process is now well described in the patent 23 and technical literature.
24 In broad summary, this process comprises first conditioning the tar sand, to make it amenable to flotation separation of the bitumen 26 From the solids. Conditioning involves feeding mined tar sand, hot 27 water (180F), an alkaline process aid (usually NaOH), and steam into 28 a rotating horizontal drum wherein the ingredients are agitated together.
29 Typically, the amounts of reagents added are in the follow-ing proportions:

1 tar sand - 3250 tons 2 hot water - 610 tons 3 NaOH - 4 tons (20% NaOH) 4 Enough steam is added to ensure an exit temperature of the mixture from the drum of about 180F. The residence time in the drum is typically 6 about 4 minutes.
7 During conditioning, the mined tar sand (in which the 8 bitumen, connate water and solids are tightly bound together) becomes 9 an aqueous slurry of porridge-like consistency, wherein the components are in loose association.
11 The slurry leaving the drum is screened, to remove oversize12 material, and then flooded or diluted with additional hot water. The 13 product typically comprises 7% by weight bitumen, 43% water and 50%
14 solids. Its temperature is typically 160 - 180F.
The diluted slurry then is transferred to the primary 16 separation operation, where it is introduced into a large separation 17 vessel having a cylindrical upper section and conical lower section.
18 Here the slurry is retained for about 45 minutes in a quiescent con-19 dition. Most of the sand sinks to the bottom and is discharged, to-gether with some fines, water, and bitumen, through an outlet. This 21 discharge is discarded as tailings.
22 The bitumen present in the separation vessel exists in 23 the form of globules, some of which attach themselves to air bubbles 24 entrained in the slurry during conditioning. The aerated bitumen tends to rise through the slurry and is recovered as a froth by a launder 26 extending around the upper lip of the separation vessel. This froth is 27 called primary froth. Typically~ it comprises:
28 66.4% bitumen 29 8.9% solids 24.7% water.

1 Not all of the bitumen becomes sufficiently aerated to rise 2 into the primary froth product~ Much of this bitumen, together with 3 much of the fines, tends to collect in the mid-section of the separation 4 vessel. This aqueous mixture is termed "middlings".
The middlings are withdrawn from the vessel and are fed fi into subaerated flotation cells where secondary separation is 7 practiced. Here the middlings are subjected to vigorous agitation and 8 aeration. Bitumen froth, termed "secondary froth", is produced.
9 Typically, this froth comprises:
23.8% bitumen 1l 17.5% solids 12 58.7% water.
13 It will be noted that the secondary froth is considerably 4 more contaminated with water and solids than the primary froth. One seeks to minimize this contamination, as the froth stream requires 16 downstream treatment to remove solids and water, before it can be fed 17 to the upgrading process.
18 It is desirable to operate the process so that as much 19 of the bitumen as possible reports to the primary froth. The efficiency with which bitumen is collected as primary froth is a measure of the 21 success with which the entire bitumen in the tar sand feed has been 22 brought to a condition amenable for spontaneous flotation. For this 23 reason~ one may consider maximizing primary recovery as optimizing the 2~ entire process.
Now, the tar sand feed to the hot water process is not 26 uniform in nature. Its properties vary in accordance with factors 27 such as bitumen content, fines content, nature of the coarse solids, 28 extent of ageing and weathering after mining, and the chemical nature 29 of the bitumen. This variation in properties of the feedstock requires 1 that the processing conditions be altered -from time to time with a 2 view to maximizing primary froth recovery. Some optimizing techniques, 3 such as regulating middlings density within a preferred range or main-4 taining the temperature within a preferred narrow range, can assist in improving recovery over a narrow variation in the nature of the tar sand 6 feed. But there is a need for identification of a parameter which can be 7 monitored and used to maximize primary froth recovery over a wide range 8 of different tar sand types.
9 At this point~ it is useful to review the role of the "process aid", as it was understood in the past. The originator of the hot water 11 process, Or. Karl Clark, noted that the tar sand was acidic in nature. He 12 taught the need to add an alkaline process aid, such as NaOH, to adjust 13 the pH of the drum slurry to near neutral condition, in order to improve 14 bitumen recovery in the primary separation step. Later investigators taught that it was desirable to maintain a slurry pH in the range of 16 about 8 - 9~ to maximize bitumen recovery.
17 More recently, Dr. Emerson Sanford, co-worker of the 18 present applicants, set forth in Canadian Patents 1,100,074 and 19 1,094,003 that the role of the NaOH was to produce surfactants in the slurry by reaction with carboxylic and sulfonic acid substituents 21 present in the bitumen. He submitted that it was surfactants that 22 were needed to condition the tar sand to free the bitumen from the 23 other tar sand components and render said bitumen amenable to air 24 attachment. He further taught that the level of fines would affect the surfactant requirements. It was believed the fines wou',d adsorb 26 surfactants, thereby diminishing their availability for 'conditioning', In27 summary, he taught that:

1 (1) some process aid was needed for good primary recovery;
2 (2) the process aid functioned by generating surfactants 3 within the slurry, which surfactants were required 4 to maximize bitumen recovery; and (3) different tar sand types, having different fines 6 contents~ would require different quantities of NaOH
7 in order to achieve maximum primary froth production.
8 As mining and geological inspection of the Athabasca g deposit has expanded~ it has become evident that there are oil sand lo types that do not follow the relationships between recovery and process 1l aid addition that one would have anticipated.
12 One such deviation arises from the nature of the clays. It now appears that clays differ in their capacity to adsorb surfactants.
THose that are so placed in the deposit as to be in contact with bitumen can have surfaces thoroughly impregnated with hydrocarbon molecules and 16 may not have sites available for surfactant adsorption. Clays laid 17 down in later erast and forming part of the overburden, may have 18 hydrocarbon-free surfaces,in which case they are strong surfactant 19 adsorbers. If, during mining, some of the overburden gets included in the feed sent to extraction, these non-impregnated clays "poison" the 21 slurry by adsorbing surfactants. When the extraction circuit has been 22 optimized for non-adsorbing clays, the introduction of feed containing 23 overburden clays will lead to reduced recovery. Some oil sands are so 24 rich in surfactant-adsorbing clays that the power of the contained bitumen to contribute surfactants to the slurry is more than offset by 26 the tendency of the clays to adsorb surfactants.
27 A second deviation from ~Inormal~ behaviour is the deteriora-2~3 tion of oil sand after mining. During storage~ feed can age. The 29 mechanism is not understood, but the bitumen surface properties appear to alter, with the result that separation from the solids and attachment 31 of air are made more difficult.

~ ~17 r 1 A third deviation arises in the case of feeds rich in bitumen.
2 Some have been found to produce such a high level of surfactants, ~,Jithout 3 any process aid addition, that the slurry is always in an over-conditioned4 state. Over-conditioning results in bitumen losses from the separa-tion vessel, presumably due to emulsification.
6 Trying to control the process by monitoring some property 7 of the feed is thus liable to failure because the relationships between 8 such property and recovery can be subject to abberrations.
g A safer procedure is to use some property of the slurry, once prepared, rather than of the feed, to determine the needed dosage of 11 process aid.
12 There is thus a need to identify a reliable parameter which can be used to optimize NaOH addition and to determine a strategy for best combining the various types of tar sand to offset their undesirable qualities with respect to surfactant production and consumption.

.
The present invention is based in par-t on the discovery that, 18 for a particular circuit, there is a critical or optimum concentration of19 free surfactant in solution in the aqueous phase of the process slurry (hereinafter ''CO''), which always is required to obtain maximum recovery 21 of bitumen from tar sand in the -form of primary froth.
22 The discovery was arrived at by running a number of batch 23 hot water process extractions on a single tar sand feed, while varying 24 only the amount of process aid used. When primary froth recovery was plotted against concentration of free surfactant in the aqueous phase 26 of the process slurry (hereinafter "Cs"), a peak curve was obtained.
27 When this procedure was repeated for a number of different tar sands, 28 a peak curve was obtained in each case and the peaks were found to occur 29 at substantially the same Cs value. This common peak Cs value is the optimum concentration CO .

l So, for a particular circuit, a single free surfa-tant val~e 2 CO leads to maximum primary froth recovery, regardless of the nature of 3 the tar sand feedstock being processed.
4 Having made this discovery, a general process has been evolved comprising the following steps:
6 (l) determining, for the extraction apparatus or circuit 7 used, a measure of CO for one tar sand feed;
8 (2) then establishing Cs from time to time as different 9 tar sand feeds are processed; and l (3) varying the process aid addition to the process as the ll nature of the tar sand feed changes, to thereby main-l2 tain Cs as close to CO as possible.
3 The general process has now advantageously been applied in 4 connection with feedstocks, for the hot water process, which consist of blends of two or more different tar sand feeds. If a plurality of dif-l6 ferent tar sand feeds, which satisfy criteria set forth below, are 17 selected and combined in certain proportions, which are set forth in a l8 range defined below, and if this blend is combined with process aid in l9 an amount within a range defined below, then, when this mixture is used in the hot water process, two results occur:
2l (l) Cs is found to equate substantially with CO ; and 22 (2) the primary froth recovery obtained is greater than 23 that which would be obtained if one processed each of 24 the tar sand feeds separately.
The use of blended feedstock and process aid as aforesaid 26 in the hot water process is characterized by an important advantage. The 27 invention enables certain tar sand feeds, which could not by themselves 28 be processed a-t CO condition, to be so processed.

1 More particularly, there are certain tar sand feeds whicn, 2 when slurried, initially are consumers of free surfactants. These tar 3 sand feeds usually have a high fines content. The fines tend to adsorb 4 the free surfactants; this is particularly the case when the fines are not well impregnated with bitumen. Thus, when a tar sand of this type 6 is processed, a relatively large quantity of process aid would be required 7 to bring Cs to the CO value. Now, there is a known limit on the amount 8 of process aid which can be used in the hot water process. This is 9 generally taken to be about 0.2 wt. ~, based on the dry tar sand. If this limiting amount of process aicl is exceeded, the hot water process 1l is deleteriously affected by such ef-fects as emulsification of bitumen 12 and poor froth/middlings interface. By blending these free surfactant-3 consuming tar sand feeds with others9 as described below, they can now 4 be processed at CO condition without exceeding the 0.2 wt. % limit.
There are other tar sand feeds, usually high in bitumen 16 content, which, when slurried, produce such a high concentration of 17 free surfactants that they cannot be processed in conventional fashion 18 at CO condition. These tar sand feeds, when slurried, yield a Cs 19 value which is so high that it is on the downslope of the peak curve.
By blending these free surfactant-producing tar sand feeds with free 21 surfactant-consuming feeds, both types of feeds can now be processed 22 at CO condition.
23 Broadly stated, the invention is a process for extracting 24 bitumen from tar sand of varying nature using the hot water process in an extraction circuit, wherein the tar sand is conditioned, by slurrying 26 it with hot water and alkaline process aid with agitation, is diluted 27 with water, and is then retained in a quiescent condition to produce 28 primary bitumen froth. The improvement comprises: selecting a first 29 tar sand feed which, when slurried, is a consumer of free surfactants and a second tar sand feed which, when slurried~is a producer of free 31 surfactants; and blending said first ancl second tar sand feeds with 1 process aid in the conditioning step in amounts selected to yield sub-2 stantially the optimum free surfactant condition, in the aqueous phase 3 of the process slurry for the circuit, required to yield maximum r)rimary 4 froth recovery.
.

DESCRIPTION OF THE DRAI~JINGS
_ _ 6 Figure 1 is a side view of a laboratory hot water process 7 batch extraction unit used to develop the data underlying this invention;
8 Figure 2 is a peak curve plot for various tar sand type 9 samples of primary froth recovery (%) against free surfactant concentration in secondary tailings - the data was obtained by conducting extractions 1l on each feed at varying NaOH additions while holding other conditions 12 constant;
3 Figure 3 is a plot of free surfactant concentration obtained 4 in the aqueous phase of the process slurry (Cs) against dosage of NaOH
process aid used in the process - line A is the surfactant production 16 line obtained by treating samples of surfactant-producing tar sand A
17 with different amounts of NaOH, and line B is the surfactant production 18 line obtained by treating samples of a surfactant-consuming tar sand B -19 the dotted horizontal line corresponds with CO (the optimurn concentration of free surfactant in the aqueous phase of the process slurry for the 21 extraction unit involYed, which concentration corresponds with that 22 associated with maximum primary bitumen froth recoYery);
23 Figure 4 is a plot, similar to that of Figure 3~ for two other 24 tar sands C and D - tar sand C differs from tar sand A in that its intercept with the free surfactant concentration axis is less than CO~
26 whereas A's is greater - tar sand C is a free surfactant-producing tar 27 sand~ and tar sand D is a free surfactant-consuming tar sand~
28 Figure 5 is a plot of maximum primary froth recovery against 29 composition of blend (expressed in terms of % marine ore in the blend) -this plot shows the synergism which arises with respect to primary recovery 31 when certain blends are used; and 1 Figure 6 sets forth titration curves used in the determina ion 2 of free surfactant concentration.

_ _ . _ _ _ _ _ _ _ The invention is a process which has evolved from laboratory experimentation, using bench scale equipment and involving the measure-6 ment of certain parameters and the treatment of the measurement data 7 acquired. It is necessary to go into this experimental background, in 8 order to comprenend the invention.

g The Laboratory Extraction Unit The laboratory batch extraction unit shown in Figure 1 was 11 used to develop the data set forth below. A detailed description o-f this 12 unit and its operation is given in the paper by E. C. Sanford and F. A.
13 Seyer entitled "Processability of Athabasca Tar Sand Using a Batch 14 Extraction Unit: The Role of NaOH" in the Canadian Institute of Mines Bulletin, 72 , (1979), a-t page 16~.
16 Experience has shown that the practise of the hot water 17 extraction process on tar sand in this unit gives results which closely 18 parallel those of the commercial ho-t water extraction process plant operated 19 by the assignees of this application.
The extraction unit of Figure 1 comprises a steel pot 1 21 having a hot water heating ~jacket 2 for temperature control. An agita-tor22 3 for stirring and a sparger 4, -for the introduction of air, extend into 23 the pot.
2~ In the operation of the extraction unit, a charge o-f tar sand, water and NaOH is introduced into the pot. The pot contents are 26 then heated by the jacket. Once at the desired temperature, the charge 27 is agitated and air is sparged into the slurry, to simulate the step oF
28 conditioning in a drum. The air sparging is then s-topped and additional 1 hot water is added, to simulate the step of flooding or dilution.
2 Agitation is continued for a few minutes~ to mix the components. The 3 product is then retained in the pot under quiescent conditions, to 4 simulate primary separation or flotation. Primary bitumen froth is produced during this step. This froth is skimmed off. The residual 6 mixture is then again agitated with vigorous air sparging, to simulate 7 secondary separation. Secondary froth is produced and skimmed off. The 8 material left in the pot is referred to hereafter as 'secondary g tailings'.
lo The determination of Cs values was arrived at using the 1l aqueous phase of this secondary tailings as the feedstock for analysis.
12 The bitumen recoveries were based on the bitumen contents and amounts of the froth produced and skimmed o-ff.

Free Surfactant ~C ) Determination s . ... ...
The invention has necessitated the utilization of a method 16 for measuring Cs ~ the concentration of free surfactants dissolved in 17 the aqueous phase of the process slurry.
18 Before describing the method used, it is useful to explain 19 that, for purposes of this specification, "free surfactants" is a term used to denote the surface active compounds which are in aqueous solution 21 during hot water processing of a tar sand feed. It is believed that, in 22 the main, these compounds are the carboxylate salts which are the reac-tion 23 products o~ NaOH (the process aid) and carboxylic groups associated 24 with the bitumen.
It needs to be noted that not all of the carboxylates in the 26 aforesaid aqueous phase are surface active. Therefore, it is a necessary 27 part of the free surfactant determination method to differentiate between 28 the carboxylates or compounds which are surface active and those that are 29 not. This is further dealt with below.

1 The free surfactant determination procedure which has been 2 utilized takes advantage of the -fact that surfactants collect a~ surfaces.
3 More particularly, a technique known as foam fractionation was used to 4 collect the surface active compounds present in a slurry aqueous phase sample of known volume. This foam fractionation procedure is described 6 in -the paper by C. Bowman entitled "Molecular and Interfacial Properties 7 of Athabasca Tar Sands", published in Proceedings 7th World Petroleum 8 Congress, 3 , (1967), pages 583 - 604.
9 In greater detail, each sample of secondary tailings from the batch extraction unit was centrifuged at 15000 G, to remove suspended 11 solids. A 200 cm3 sample of the remaining aqueous phase was foam fraction-12 ated in a 300 cm3 cylindrica'l vessel equipped with a nitrogen sparger.
3 The introduction of gas was varied as required to yield a separable 4 foam. Fractionation was continued until the sur-face tension of -the residue reached a limiting value, substantially that of pure water, as 16 determined by the known maximum bubble pressure technique.
17 At this point, all of the surface active compounds were 18 isolated in the fractionate and the residue contained only non-surface 19 active compounds. The concentration of non-surface active compounds was assumed to be equal in each of the fractionate and residue fractions.
21 50 cm3 aliquots of each of the fractionate and residue 22 fractions were then titrated against hydroch'loric acid, to establish 23 measures of the concentrations of the surface active and non-surface 24 active compounds.
Typical titration curves are illustrated in Figure 6. It 26 will be noted that the volumes of acid, used to titrate the non-surface 27 active carboxylates in the residue and the non-surface active plus surface28 active carboxylates in the fractionate, can be determined from Figure 6 29 using the end points shown.

1 The data from these titrations were processed using the 2 following mathematical analysis to determine free surfactant concentration 3 (Cs)~ as follows:
4 - let Cnf5 be the concentration of non-surface active carboxylates in the fractionate;
6 - let c5f be the concentration oF surface active carboxylates 7 in the fractionate;
8 - let Cnrs be the concentration of non-surface active 9 carboxylates in the residue;
- let cf be the concentration of total carboxylates in the 11 fractionate;
12 - and let Cr be the concentration of total carboxylates in 3 the residue.

14 Therefore: Cnf5 + c5f = cf (1) and Cnrs = Cr ( 2) 16 At equilibrium, the concen-tration of non-surface active 7 carboxylates is assumed to be the same in both the residue and 18 fractionate.
19 Therefore:
cf = Cr ns ns 21 Thus, combining equations (1) and ~2): Cr + c5f = cf (3) 22 As C and Cr are determined by titration, the value of Cs 23 (the measure of free surfactant concentration) may be determined.
24 Detailed results of a free surfactant determination for one marine ore feed D are now presented, in conjunction with Table I:

2 NaOH Tlgs Vo1ume Vol of Frac'te Vol NHCl Vol Resid Vol 3 wt % vol frac'd frac'te aliquot HCl resid aliquot HC1 F ed cm cm3 - cm3 cm __ ,cm ~ cm cm cm 6 .00 1080 7 .04 1080 '199.5103.7 50.0 .4g .0571 91.3 50.0 .46 8 .08 1080 208.'191.9 50.0 .48 .057'1 115.4 50.0 .43 9 .12 1080 197.8 115.2 50.0 .89 .0571 79.7 50.0 .84 .20 1080 152.2 62.1 61.9 3.34 .0328 89.7 50.0 2.40 11 Total secondary tailings sample volume = 1080 cm3 12 Fractionate carboxylate salt determination:
13 Total volume fractionated = 152 cm3 = V
14 Volume of fractionate = 62 cm = Vfractionate Aliquot volume = 61.7 cm3 = V l;
16 Volume of HCl titrant = 3.34 cm = V
17 Normality of HCl = 0~0328N = N

18 Therefore 19 Concentration of carboxylate salts in fractionate NHCl . V
21 Valiquot 22 = 17.g x 10 4 23 Residue carboxylate salt determination:
24 Residue volume = 90 cm3 Aliquot volume = 50.0 cm3 = V li 26 Volume of HCl titrant = 2.40 cm = V
27 Normality of HCl =0.0328N = N

1 Therefore 2 Concentration of carboxylate salts in residue NHCl . VHC
4 Valiquot = 15.7 x lO 4 6 From the equations 7 Cs = Cf - Cr 8 = 17.8 x 10 4 - 15.7 x 10 4 9 = 2.1 x 10~4 equivalents per litre This is the concentration of free surFactant in the 62.1 cm3 11 fractionate sample. In the original tailings sample there is 12 (2.1 x 10 ) x 62.

13 = 8.6 x 10~5 equivalents per litre.

14 Opt;mum Free Surfactant (C ) Determination As previously stated, a number of hot water process extraction 16 runs were carried out on a particular feed in the batch extraction unit, 17 varying only the amount of NaOH addition. When the Cs values obtained 18 from the runs were plotted, a peak curve, such as one of the curves 19 shown in Figure 1, was obtained. This procedure was repeated for a number of different feeds. It was discovered that the peaks of the 21 curves, corresponding with the maximum primary froth recoveries, fell 22 substantially on a vertical line corresponding with a single Cs value.
23 This particular optimum Cs value is re~erred to as CO .

1 Stated otherwise, maximum primary frotn recovery for the 2 various feeds occurs at only one small range of Cs values. Both below 3 and above that range, which for practical purposes is taken to be a 4 single value CO ~ the primary froth recovery diminishes.
To summarize, for a given circuit or extraction unit, the 6 maximum primary froth recovery for various tar sand feeds always occurs7 at substantially the same free surfactant concentration CO in the aqueous 8 phase of the process slurry.

g The Linear Relationship (Figure 3) _ __ _ ____ In conjunction with determining CO for the extraction unit 1l used, one may determine, for each tar sand feed being considered, the 12 nature of the substantially linear relationship which exists between Cs and Na0H addition.
This may be done by: extracting a plurality of portions of each tar sand feed; using the hot water process at constant conditions 16 except for using different levels of NaOH addition; determining the Cs 17 value for each such extraction; and establishing the nature of the 18 linear relationship by plotting the linear surfactant production line for 19 each feed, which line expresses this relationship.
A typical plot of surfactant production lines, based on 21 experimental runs described below, is set forth in Figure 3. It will be 22 noted that each sur-factant production line is extrapolated to intersect 23 the zero NaOH addition axis and provide an intercept value. Also, a 24 horizontal broken line is provided on the plot, which corresponds with the CO value for the extraction unit used.
26 Line A on Figure 3 is the surfactant production line obtained 27 in connection with extraction of a tar sand feed A, which is relatively 28 high in bitumen content. This feed had the following composition:

2 Oil Water Solids Fines Content 3 Content Content Con'cent (< -44 ) % w/w 4 Tar San Comments % w/w % w/w % w/w solids A (rich) fresh 13.7 1.5 84.8 8.9 6 The extraction procedure used on each tar sand sample, to 7 develop the data for lines A (and line B described below),was as 8 follo~ls:
9 A charge of 500 g of the tar sand, 150 mL of water at 82C, and different amounts of NaOH were introduced into the pot 1.
11 Hot water was circulated through the jacket 2 to bring the 12 charge to 82C and to maintain it there. Once the charge 13 was at temperature, it was agitated with the agitator 3 14 for 10 minutes at 600 rpm while simultaneously introducing air into the charge at 7 mL per second through the sparger 16 4. The air was then switched off and the mixture flooded 17 with 900 mL of hot water (82C). Mixing with the agitator 18 was continued for a further 10 minutes. The agitator was 19 then switched off. The produced primary froth was skimmed off the surface of the mixture and weighed. Samples thereof 21 were analyzed. The residual mixture was then subjected to 22 secondary separation. More particularly, it was agitated 23 at 800 rpm for 5 minutes with air sparging at the rate of 24 4 mL/sec. The secondary froth produced was skimmed off, weighed and analyzed.
26 The extraction results for tar sand feed A were as follows:

2 NaOH Froth Com~osition ~
3 % w/wMass Primary _ (~ w/w) _ Primary 4 tar sandFroth (g) Oi~ ~ Solids Recovery 0.00 53.5 71.1 19.59.4 85.4 6 0.01 28.9 71.2 22.46.4 53.0 7 0,02 32.0 74.8 18.76.5 ~5.0 8 0.03 32.0 75~6 17.37.1 57.4 9 It will be noted from the composition set forth in Table II
that line A is the surfactant production line for a tar sand feed high 11 in bitumen content. As shown in Figure 3, the line obtained has a relatively 12 steep slope, indicating that high concentrations of free surfactant are 13 produced when extractions are carried out.
14 It will further be noted from Figure 3 ~ at zero NaOH
addition, the bitumen-rich tar sand feed A produces such a high con-16 centration of free surfactants that the Cs value is greater than the C
17 value. The data of Table III indicates that the primary froth recoveries 18 from processing this rich tar sand alone fall on the right hand downwardly 19 sloping portion of the feed A curve on Figure 1. Thus, oil recovery by hot water extraction from this tar sand alone is less than the maximum 21 possible, when this feed is processed in the ordinary way with NaOH
22 addition.
23 Line B on Figure 3 is the surfactant production line obtained 24 from extracting a tar sand feed B, known as marine ore, which was low in bitumen content and rich in clay particles. As shown, the slope of this 26 line was less angular than that of line A, indicating that a lower con-27 centration of free surFactants was generated.
28 Tar sand feed B had the following composition:

_ 19 _ 2 Oil W~ter Solids Fines Content 3 Content Content Content (< -44) ,/0 w/w 4 _r Sand Comments % wlw % w/w ~ % w/w _ solids B (marine) aged 60 5.6 10.3 85.0 21.0 days 6 The extraction results for tar sand feed B were as follows:

8 NaOH Froth Composition %
9 % w/wMass Primary (% w/w~ Pri~ary tar sandFroth (g) Oil H2_ ~Solids Recovery 11 0.01 1.01 26.35 64.64 9.01 3.6 12 0.03 1.20 36.45 54.88 8.67 4.3 13 0.06 1.40 42.72 49.-97 7.31 4.1 14 0.12 1.32 38.66 52.94 8.40 4.7 0.16 1.99 47.58 44.98 7.44 7.1 16 ~.20 1.51 40.27 51.52 8.22 5.4 17 0.24 0.84 39.83 52.16 8.01 3.0 18 As can be seen, the primary recovery of bitumen from this 19 feed B was very low when it was processed alone.
Of particular interest is that portion of line B to the left 21 of point 'a' on the plot. Here, for extractions conducted on -feed B
22 at NaOH additions in the range 0.0 to 0.04 wt %, there is a condition 23 of free surfactant consumption. Stated otherwise, in the case of tar 24 sand B and an NaOH addition in said range, the tar sand appears to be adsorbing the free surfactants that it produces when in the extraction 26 slurry.

1 Blending 2 We postulated that, if a first tar sand feed, which is a 3 consumer of free surfactants, were -to be blended with a second tar sand 4 feed, which is a producer of free surfactants, it might be possible to achieve CO conditions. By 'producer' is meant that the tar sand, when 6 processed alone by the hot water process in accordance with conventional 7 conditions, will yield a surfactant production line which, in the context 8 o-f a Figure 3 plot, is always positive (as is the case with line A). By g 'consumer' is meant that the tar sand yields a surfactant production lo line which is partly negative (as is the case with line B).
11 In fact, it has been discovered that there is a synergistic 12 result obtainable, in that certain blends give higher recoveries than 13 would be expected by summing the results of processing each tar sand 14 feed separately.
This synergistic effect is shown in Figure 5. Feeds A and 16 B were blended in various proportions and subjected to bitumen extraction 17 in the previously described extraction unit. The extraction results were 18 as follows:
19 ' A _E VI

Blend Maximum Maximum 21 Composition Primary Recovery Primary 22 (% Mar;ne)(g bitumen/500'g feèd) Recovery (%) 27 , 80 8 22 J~

1 The straight line in Figure 5 is the calculated recovery that 2 was expected from blending alone, taking into account the proportion of 3 each -feed in the blend. The curve depicts the actual recoveries obtained 4 when blends of tar sands A and B where extracted as described above. The upper part of the curve, between the left axis of the plot and the node 6 point~ shows enhanced recoveries due to synergism.
7 The node point on Figure 5 corresponds with point 'a' on ~3 Figure 3.
9 The shaded area on Figure 3 defines the blend proportions which will yield synergism.
1l Figure 4 is a plot showing the surfactant production lines 12 for two tar sand feeds C and D to be blended, where the left hand end 3 of the line for the producer feed is below the CO line when small NaOH
4 additions are used in the extractions. Only when enough NaOH is used, so as to bring the producer line to at least CO , can synergism result. So, 16 for the blend of Figure 4, the shaded area is shifted to the right in 17 comparison to that of Figure 30 18 One may select compatible tar sand feeds to make up a 19 blended feedstock which is amenable to processing at CO condition, by observing the following rules:
21 (1) select a first tar sand feed whose surfactant 22 production line has a negative intercept value when 23 plotted on a plot of the type of Figures 3 and 4 (that 24 is , a feed which will consume free surfactants in the course of hot water extraction);
26 (2) select a second tar sand feed whose surfactant 27 production line has a positive intercept value and 28 which extends above the CO line (that is, a -Feed which 29 will produce free surfactants in the course of hot water extraction in the absence of NaOH);

1 (3) the first feed surfactant production line having to cross 2 the zero free surfactant concentration axis at a NaOH
3 addition value which is less than 0.2 wt. % and greater than (a) the NaOH addition value corresponding with the 6 second feed intercept (in which case the value is 0) 7 or 8 (b) the NaOH addition value corresponding with the point 9 at which the second feed surfactant production line crosses the CO line, 11 whichever is greater. Stated otherwise, the first 12 vertical boundary line, passing through the point where 13 the first feed surfactant production line crosses the 14 zero free surfactant concentration axis, is to the right of or at a greater NaOH value than the second vertical 16 boundary line, which passes through the point where the 17 second feed surfactant production line crosses the CO
18 line or, if it does not so cross, which coincides with 19 the zero NaOH addition axis.

The amount of NaOH used in conjunction with the blended 21 feedstock is selected so as to fall within the range of values between 22 the two vertical boundary lines.
23 Having selected the feeds and NaOH addition to be used, one 24 may determine the proportions of each feed to be used by solving the following equations:
26 C = Xl (Cs 1 + RlP) + X2 (Cs~2 R2 ) 27 1 Xl X2 (5) where:

1 CO is the optimum concentration of free surfactant 2 in the aqueous phase of the process slurry for the 3 extraction unit used 4 Cs 1 and Cs 2 are the intercepts for the selected first and second feeds 6 Rl and R2 are the slopes of the free surfactant production 7 lines for the first and second feeds 8 P is the process aid addition selected g Xl and X2 are the proportions of first and second feeds used to make the blend.
11 There may, of course, be more than two components in the 12 blend, in which case all values of X will give a sum of 1.
3 The improvement in primary froth recovery which can be obtained by selecting free surfactant consuming and producing tar sand feeds as set forth, combining them with a NaOH addition selected from the 16 range between the vertical boundary lines, and combining the tar sands in17 proportions established by solving equations (4) and (5),is demonstrated 18 by the following example involving a blend of feeds A and B.
19 Inspection of Figure 2 shows that CO for the extraction unit of Figure 1 was 1.2 and lQ 4 N ~i.e. equivalents per litre).
21 Turning to Figure 3, Cs a for -feed A at zero NaOH addition was 1.45 x 22 10 4 N. The slope Ra f the free surfactant line for feed A was 44.8 x 23 10 4 . For feed r~, Cs b was -0.45 x 10 4 N and Rb was 11.9 x 10 4 .
24 With this blend, the left hand vertical boundary line coincided with the zero NaOH addition axis, and the right hand vertical boundary 26 line occurred at 0.04 weight % NaOH. That is, the amount of NaOH to be 27 added fell between 0.0 and 0.04 wt. %. With the aid of equations (4) 28 and (5) , one can calculate the amount of NaO~I to be added for any blend29 of these feeds, to give maximum recovery.

- 2~ -~ r~ 1 f {~

1 For example, one can select a blend having a NaOH addition 2 of 0~01%. Solving equations (4) and (5) , one obtains recommended pro-3 portions of feed A and feed B as follows:
4 Xl = 0.69 X2 = 0 31 6 Hence one should blend 7 69% feed A with 8 31% feed B, 9 when one uses a process aid level of 0.01 wt. % NaOH.
In fact, we blended 70% feed A and 30% feed B and perFormed 11 extraction experiments at various NaOH levels from 0.0 to 0.04 wt. %.
12 We then in-terpolated the results to find the bi-tumen in primary froth 13 at the 0.01 wt. % NaOH level. We found the result to be 45 g bitumen.
14 This is very close to the value of 46 g for a 70/30 blend (Figure 5).
It remains to show that an improvement is obtained by pro-16 cessing the blend, compared with processing each feed separately, at 17 the respective CO conditions.
18 Feed A, processed at 0.0 wt. % NaOH, gave 53.5 g bitumen.
19 (0.0 NaOH is the nearest approach to CO for this feed [Figure 3]. The recovery was 85.4% of the maximum attainable [Figure 2]. Feed B was 21 processed at 0.16% NaOH [this being as close as we came to the ideal 22 value of 0.14, in our experimental program] .) It gave 2.0 g bitumen~
23 which, according to Figure 2, was equivalent to 701 % of the maximum 24 attainable.
Hence:
26 Feed Recovery _ _ 27 250 g Feed A
28 + 250 g Feed B
29 processed separately 0.7 x 53.5 + 0.3 x 2.0 = 38 g 250 g Feed A
31 blended with 32 250 g Feed B 45.0 g 1 These points can also be read off Figure 5. As can be seen, t'nere ~,Jas an 2 18% improvement in recovery as a result of blending. It ~las kno~Jn that 3 the value of CO was reached in this case from equations (4) and (5) 4 set forth above.

Claims (3)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A process for extracting bitumen from tar sand of varying nature using the hot water process in an extraction circuit, wherein the tar sand is conditioned, by slurring it with hot water and alkaline process aid with agitation, is diluted with water, and is then retained in a quiescent condition to produce primary bitumen froth, the improvement comprising:
selecting a first tar sand feed which, when slurried, is a consumer of free surfactants and a second tar sand feed which, when slurried, is a producer of free surfactants;
and blending said first and second tar sand feeds and process aid in the conditioning step in amounts selected to yield substantially the optimum free surfactant concentration, in the aqueous phase of the process slurry for the circuit, required to yield maximum primary froth recovery.

2. A process for extracting bitumen from tar sand of varying nature using the hot water process in an extraction circuit, wherein the tar sand is conditioned, by slurrying it with hot water and alkaline process aid with agitation, is diluted with water, and is then retained in a quiescent condition to produce primary bitumen froth, the improvement comprising:
determining, for the circuit involved, a measure of the optimum concentration of free surfactant in the aqueous phase of the process slurry, which substantially corresponds with maximum primary bitumen froth recovery from that circuit;
selecting a first tar sand feed which, when slurried, is a consumer of free surfactants and a second tar sand feed which, when slurried, is a producer of free surfactants;

and blending said first and second tar sand feeds and process aid in the conditioning step in amounts selected to yield substantially the optimum free surfactant concentration, in the aqueous phase of the process slurry for the circuit, required to yield maximum primary froth recovery.

3. In the process of extracting bitumen from tar sand of varying nature using the hot water process in an extraction circuit, wherein the tar sand is slurried in a conditioning drum with hot water and alkaline process aid, agitated, and then retained in a quiescent condition to produce primary bitumen froth, the improvement comprising:
(a) determining, for the circuit involved, a measure of the optimum concentration of free surfactant in the aqueous phase of the process slurry, which substantially corresponds with maximum primary bitumen froth recovery from that circuit;
(b) determining, for each of a plurality of tar sand feeds which are proposed for use to make a blended feedstock, the substantially linear relationship which exists between a measure of the concentrations of free surfactant, present in the aqueous phase of its process slurries, and the amounts of process aid added, when a plurality of hot water pro-cess extractions are practised on each tar sand feed at different levels of process aid addition, all other conditions being maintained substantial-ly constant;
(c) selecting first and second tar sand feeds, from the group tested, such that, if the test data from steps (a) and (b) were plotted on a graph and cross-hatched in accordance with the appropriate figure selected from the plot of free surfactant concentration in the aqueous phase of the process slurry against the amount of process aid added, the first feed would have a surfactant production line whose intercept is above the zero surfactant concentration line and would have values above the optimum surfactant concentration line at a process aid addition value of less than 0.2 wt. %, the second feed would have a surfactant production line whose intercept is below the zero surfactant concentration line and would cross the zero surfactant concentration line at a process aid addition value less than 0.2 wt. %
and greater than the process aid value corresponding with either the first feed intercept or the point where the first feed surfactant pro-duction line crosses the optimum surfactant concentration line, which-ever is greater;
(d) selecting the process aid addition from the range of values between the vertical boundaries of the cross-hatched area, (e) determining the proportions of first and second tar sand feeds to make a blended feedstock so that they satisfy the following equations:
CO = X1 (Cs 1 + R1P) + X2 (Cs 2 + R2P) X1 + X2 where: CO is the value determined in Step (a) Cs 1 and Cs 2 are the intercepts for the selected first and second feeds R1 and R2 are the slopes of the surfactant production lines for said first and second feeds P is the process aid addition selected X1 and X2 are the proportions of first and second feeds used to make the blend;
(f) and supplying to the extraction process said first and second tar sand feeds and process aid in accordance with the values for X1, X2 and P.