CA1272975A

CA1272975A - Separation of fine solids from petroleum oils and the like

Info

Publication number: CA1272975A
Application number: CA000497319A
Authority: CA
Inventors: J. Farnand Redmond
Original assignee: National Research Council of Canada
Current assignee: National Research Council of Canada
Priority date: 1985-12-10
Filing date: 1985-12-10
Publication date: 1990-08-21

Abstract

TITLE

SEPARATION OF FINE SOLIDS FROM PETROLEUM OILS AND THE LIKE

INVENTOR

J. REDMOND FARNAND

ABSTRACT OF THE DISCLOSURE

Dispersed solids, particularly fine peptized hydrophobic solids, present in some petroleum oils, bitumen solutions or concen-trates, heavy oils and the like, may be removed by distributing an additive selected from resorcinol, catechol, formic acid, maleic acid or anhydride, chloral hydrate and asphaltene - precipitating solvents (preferably certain mixtures thereof) throughout the dispersion to cause small agglomerates containing the solids to form, and separat-ing the agglomerates from the residual liquid. This residual liquid is rendered more suitable for various refining processes. Agglomerates containing asphaltenes and mineral solids may be used e.g. in hydrogen generation, for compounding roofing materials and in paving materials.

Description

~2~
Tll;s invention rela~es to the removal o~ fine peptized soli~s ~rom various biLumen solutions or concentrates, heavy peLroleum oils and the ]ike by a specific agglomeration technique. Bitumen solutions e.g. derived from Alberta tar sand processing, and heavy oils e.g.
from Cold l.ake ~lberta or Lloyclminster Saskatchewan are of particular interest as feed material.
Bitumen solutions derived from solvent extracted oil sands usually contain significant quantities o finely dispersed solids (Ignasiak, T.M., Kotlyar, L., Longstaffe, F.J., Strausz, O.P., and Mont&omery, D.S., Separation of Clay from Athabasca Asphaltene, Fuel, 62, 353-362, (19a3). The extent of the solids depends largely on the type of feed material and the method of contact with the extracting solvent (Meadus, F.W., Bassaw, B.P. and Sparks, B.D., Solvent Extraction of Athabasca Oil Sand in a Rotating Mill. Part 2. Solids-Liquid Separation and Bitumen Quality, Fuel Processing Technology, 6, 289-300, (1982j). The lower grade bitumen feeds contain higher percentages of clay resulting in higher amounts of intractable solids in the bitumen solution. Because of the adverse effect of fine solids on the upgrading process, refineries normally apply a specification `! of no more than 1.0% to the ash content of feedstocks.The intractable solids, essentially non-settling, are cha-racterized by fine particle size and a hydrocarbon coating. These solids after ext~nsive extraction with toluene have been shown to contain up to 60% organic carbon. This hydrophobic carbonaceous coating is considerably more soluble in solvents such as methanol and acetone, ~;~ 25 indicating a more polar character compared to bitumen itself.
` Infra-red and NMR spectra of the coated solids indicate a complex ~: .
hydrocarbon structure containing carboxylic acid and sulfonate groups ~, (coating) overlaying a clay (kaolin) matrix. One proposed struc~ure ':' ~ '` : ' ' ' , ' ' ' has fulvic acid-like molecules bonded to the substrate through heavy metal atoms adsorbed on or substituted into the clay lattice in clay tailings (Kessick, M.A., Structure and Properties of Oil Sands Clay Tailings, J. Can. Petrol. Technol., 18~ 49-52~ (1979~ The suspension stability apparently results from fine particle size and strong lnter-action between the particle coating and components of the dissolved bitumen. Clementæ and others also found that montmorillonite and other types of clay absorbed the heavy ends and became hydrophobic (Clementz, D.M., "Interation of Petroleum lleavy Ends with Montmorillo-10 nite" , Clays and Clay Minerals, 24~ 312-319~ (1976)~ Clement:z, D.M., "Alteration of Rock Properties by Adsorption of Petroleum Ueavy Ends: Implications for ~nhanced Oil l~ecovery", SPE/DOE 10683~ (1982)~
Czarnecka, E. and Gillott, J.~., "Formation and Character-ization of Clay Complexes with Bitumen from Athabasca Oil Sand", Clays and ;~ 15 Clay Minerals, 28~ 197 208~ (1980))~ These intractable solids are very di~ficult to remove by filtration and require extensive ccntri-fugation to bring the solids content of the bitumen solution down to acceptable levels.
ln US patent 3~268~07l~ August 23~ 1966 Example X~ the agglome-ration of the oil phase from an aqueous dispersion of tar sands is described (these agglomerates still con~ained 26-28% sand). US Patent 4~057~486 November 8~ 1977 describes the agglomeration of mineral solids from organic solvent or diluent dispersions containing ~ar ;~ ~ sands or oil shale, using about 8-50~o wt. (based on the feed) of aqueous ; 25 agglomerating liquicl. As indicated in Table 1 up to 2~5% Of ; ~ solids (sand) remained with the solvent phase, depending on the amount of settling allowed. It would be desirable to remove increased amounts of these intractable solids.

-2-; ~
, :` ' ~: ' :' .:: : .

gL2~7~i ln the recovery oE bitumen from oil sands by the SESA process (solvent extraction solids agg].omeration) described i.n Canad:ian Patent Applicati.on 486,032 filed ~8 June 1985, the solvent bitumen mi~ture alter being separated from the bulk solids contains 0.4 to greater than 1% intractabl.e solids depending on the feed type and operati.on conditi.on~s~ Expresserl on i:he bitumen component alone, Lhis solids conLent varies from about 1.3 to ~J/~. The amount of solids in t.he bitumen èxtract increases wi.th the ines content (-44~1m) in the original oil sands feed. The hydrophobic nature of these fine solids prohibits the-n from bei.ng ag~lomerated with the main body of the water wetted sol.ids in the SESA process. The present invention in one aspect provides Eor the substanLifll removal oE the i.ntractable solids from biturnen sol.utions produced by t.he SESA procrss.
SIJMMARY OF TIIE INVENTTON
lS In this invention, the fine solids are removed from the bitumen solution, heavy oil or the like by a procedure invo].ving agg].orneration of the solids followed by removal of the agglomerates e.g. by gravity settling/decantati.on, screening etc. Convent:ional wetting agents are usually surface-active agents with a propensity of markedly reducing interfacial tension at low concentrations.
Introduction of such agents into the suspensions treated in this uork usually resulted in the formation of stable emulsions with little : separation of solids.: liowever, it has been found that certain low :
molecular we~ght, water-soluble organic compounds have an affinity for Lhe hydrocarbon-coated solids in oil suspension, aLlowing these particles to be collected by an aqueous solution of the reagent.
: A list of the preferred characteristics found for these agglomera-ting agents is shown below.

-3-.: :

: .
: ' : ' ' ' - . , ' ~L2~2~

(I) ~ligh solubility in water.
~2) Iow mi.scibility with hydrocarbon solvents alone, or in the presence of water.
(3) Contain carboxylic acid and/or hydroxyl groups.

(4) ~lave weakly acidi.c and/or polar character.
_ESCRIPTION 0~ DRAWINCS
Figure I is a graph showing the reduction of ash content of bitumen sol.ution with time for treatment with precipitating so].vent and addi.tive alone, and together.
A few selected additives have been Eound to cause the desired agglomeration i..e. resorcinol, catechol, formi.c acid, maleic acid or anllyclride, chloral hydrate and asphaltene-precipitating solvent.s Mixturt.s of t:ile asphaltene-prccipitating solvenL and one of the othtr acldi.tives (or other polar organic compound having the above characteris-tics) have been tound most effective when asphaltenes are present - in the feed material.
The inven~-ion includes a process of reducing the solids content of petroleum oils, bitumen solutions or concentrates, or heavy oils and the like, containing intractable fine hydrophobic solids,comprising:
~ dispersing an additive throughout the solids-containing feed material, sai.d additive being selected from the group consisting - of resorcinol, catechol, formic acid, maleic acid or anhydride, chloral hydrate and asphaltene-precipitati.ng solvents, and mixtures thereof, in an amount sufficient to cause small agglomerates compr:ising the hydrophobic solids to form; asphaltenes being present when asphaltene-precipitating solvents are selected and water being present when the other additives are selecteci; and - removing the agglomerated solids from the residual liquid.
The agglomerates comprising the fine mineral solids and asphaltenes precipitated on the surfaces of said solids, are believed :; .

to bc novel and arc part of the invention. They are useful as compo-nerlts in rooling maLerials (tar, shingLes, repair compounds) and paving malerials. ~iso ~hey may be uscd in hydrogcn production proces-ses such as the Kashima/Toyo process described in "Chemical and En-gineering Ne~s" Oct:ober 1(), 19~3 page 25; or burned ~or heat value e.g. in a fl~lidi~ed bed combusLor. The5e agglomerates usually are formcd in a size range o[ about 0.2 to about 2 mm. diam. but other sizes are possible by manipul.l~ion Oc variables.
The eff~ct:ive amollnts of additive usually will be within the range about 0.5 to about 3/0 W/W based on the ~eed material (except for the asphaltene-precipiLating solvent which usually is used within the rangc oL about 10 t:o about 40% and rccyclcd).
The asphaltene precipitating solvents are selected from aliphatic solvents having up to about eight carbon atoms. Suitable solvents include petroleum ether, pentane, hexane, octane, and their corresponding isomers.
.
1.ight ends from upgrading plants or refineries could also serve as suitable precipitating solvents, especially the low boiling fraction of Syncrude naphtha or other low boiling paraFfinic cut.
This aspect allows precipitation of the most polar and highest molecular weight portion of asphaltenes, which in turn has been found to act as a bonding agent between the suspended solids with the attendant agglomeration and settling of the solids. The most undesirable portion of the asphaltenes was found to be removed.
This portion also contained the highest amount of metal organic com-pounds. It follows that a portion of the undesirable hetero-atoms (N,S,O) was removed also. Such a treated bitumen product will consti-tute an improved feed for upgrading plants. The quality of the feed

-5-~ .: . . .:
: .

:~ -,: ' : : .

~L2~,~

wil.l i.mprove with the amount of precipitating solvent used.
The rnosL prcferrcd cmbodiment oE the invention is the concur-rent use of both an aspha].tene-prec;.pitating solvent (when asphaltenes are prescnt) and a pol.ar compound characterized by high water solubi.lity and low mi.scibility with hydrocarbon solvents.
It was Eound that the effective concentration requi.red for each cornponent was reduced so significantly that the arnount of ~ch additive needed was well be].ow the eEfective levels required when using either component alone (this is believed to be asynergistic effect). Thus concurrent use of about 5 to 10% solvent plus about 0.2 to 0.5% polar additive (~/W of the feed) was found very effective.
The precipitating solvent method is particularly advantageous for the SESA (solvent extracti.on spherical agglomeration) process liquors not only because of the low levels of water (~ 1%) required but also because economics probably would dictate the use of synthetic naphtha type solvents in any SESA
process commercial operation. These synthetic naphthas have a high content of short chain aliphatic hydrocarbons that contribute apprecia-bly to the subsequent asphaltene precipitation thereby requiring less precipitating solvent to displace a given amount of asphaltenes.
A typi.cal synthetic naphtha contains about 6% aromatics, 25% naphtllenes, and 69% paraffins of which n-pentane and n-hexane make up 22% and 14.5/~ respectively. Because of the presence especially of the latter two par.affins, in a SESA feed liquor, less precipitati.ng so~vent is requi.red to cause thc asphaltenes to agglomerate and precl.pitate the pepti.zed solids in this case.
~s wcll as tar sand sol.vent extraction liquors, this invellt ion is applicable to bitumen concentrates from oil sand hot watcr extracLion processes, various heavy oils including shale o:i.l, and

-6-~``' '` ' :

:: .
~..

. ~
.:
`''' ' .:

oils obtainecl from cnlldnced oil recovery techniques SUC~I as stca injection. A preferrcd feed material is a tar sand solvent extraction spherica] agglomeration process liquor. Even conventional crude oils containing significant amounts of clay soLkls can be treated.
When the additive is resorcinoL, catechol, formic acid, maleic acid or anhydride or chLoral hydrate. some water must be present (or be added) for efEective solids agglomeration. Amounts of water in the range of abouL 0.5 to about 3% w/w of the feed are preferred. Ln some cases it will be desirable to incorporate Lhese additives as a concentrated aqueous solution, prefcrably a saturated solution. When asphaltene-precipitating solvents are used alone, it has been found preferable that the system contain only its natural water content i.e. have 110 additional water addcd.

Many additives were tested before the fully effective ones were idcntified. It was not possible to predict the most efective ones. It; is preEerred to select the most eEfective or economical one for each feed material.
The following examples are illustrative.

; In the experiments with polar additives, water solutions of many different types of chemical additives were added to jars containing 20 to 50 g of bitumen concentrates. The mixtures were then agitated Eairly vigorously in a Spex ~TM) shaker for 3 minutes to promote rapid dispersion of the additives. This was followed by more gentle agitation in a paint shaker for an addltional 6 minutes in order to promote agglomeration of the mineral particles. The mode oE agitation did not appear to be important except that vigorous agita-tion followed by a milder form improved settling rates slight]y.
The suspensions were then al~owed to settle for given time intervals of 15 minutes up to 24 hours depending on the rate of settling achieved.
.~

-7-''-' "',`: '. ~- , : .' '~ .: ' In the case of prec-ipitating solvent adclition, selected solvents were added to the liquors in order to precipitate the asphaltenes on the mineral partic]es. In this case, mild agitation alone appeared to give the best results. Agitation was thus accomplished by rotating the partially-filled jars at about 90% of the critical speed for about 9 minutes (the critical speed is that at which the contents follow the rotating vessel walls thru 360 ). Milder agitation encouraged the loosely formed asphaltene-mineral agglomerates to coalesce and weld into larger si~es with less risk of disintegrating. In the case of mixed polar agents and solvents, the polar agents were added first followed by 3 minutes agitation in a Spex shaker. Subsequently a precipitating solvent was added followed by 6 minutes agitation in a paint shaker. In all cases the treated suspensions were allowed to settle by gravity. Samples were drawn off Erom the supernatent ]5 layer at chosen intervals and ashed at 500 C to determine the solids content. The water content in the bitumen concentrates was determined by the Dean ancl Stark method.
Bitumen concentrates used in these tests were produced by the SESA solvent extraction bench-scale unit at NRC and by the Syncrude ~lot Water flotation plant. The latter product was limited to a single 3 liter sample. The SESA material was produced from diffcerent types of oil sands starting materials which were all high in fines content;
a brief continuous settling step removed the bulk of the coarse suspended material durLng collection. In addition, a composite sample was used.
Ihe identities and compositions of the bit:umen concentrates are given in Table 1.

`: ~, ~' ;. .
" ' ' `' . ~ ':~

~2:~7~;

EXA~PLE 1 SESA I.IQIJORS
-The effect of different types of additives on SESA process bitumcn extract is shown in Table 2A. They represent only a cross section of the many additives tested, and are given here for comparison purposes. In general, screening was carried out with many low molecular weight polar compounds possessing high water solubilities. These additives were composed of alcohols, bases, acids, ketones, aldehydes, phenols, and other types of additives with appropriate functional groups. It was iound that useful additives were soluble to concentrations of at least 20% in water. The most difficult bitumen concentrate - B (Conditioning Drum Oversize) was used in the test results in this Cable 2A. These results may be compared with the effective additives in Table 2B.
Results given in Table 2B compare 6 treating agents found effective for SESA process liquors after a two hour settling period.
Two different concentrations of additives were used in each case.
Formic acid, maleic acid and chloral hydrate in amounts of about 2%
were all effective additives for the four SESA product liquors. Resor-cinol (1~3-dihydroxybenzene) in less than one-fifth the concentration (0.37%) of the other polar type additives was found the most effective.
The better results obtained with resorcinol may have resulted from the acid hydroxyl groups and its extremely high water solubility as well as its insolubility in bitumen components. All the additives in Nos. 3-9 were added as 50% w/w solutions. Catechol (No. 10) was added as a 30% solution because of its lower solubility in water.
Catechol (1,2-dihydroxybenzene) was probably less eEfective than resorci-nol because of its lower water solub~lity and the lower reactivity ''"' ,: .., : .. ..,:.,: : :: -: . : :-: ~- , ~ , .:: , : .

of the hydroxyl groups. The remaining dihydric phenol (1,4-dihydro~y-benzene) has a solubility of only 5.9 g in 100 cc of water, and was therefore not used. It is noteworthy that all polar wetting agents were less efEective on product B (Conditioning Drum Oversize) thar on the other SESA liquors. Because of its history, this material is not only higher in solicls but also contains a proportionately greater arnount of hydrophobic fines compared to other product liquors, therefore making wetting with polar agents more difficult.
Precipitation of asphaltenes with pentane (Table 2B Nos.
11 and 12) was an effective method for reducing solids in SESA product liquors. The water content of these liquors (_ 0.4%) was noL high enough to promote any appreciable amount of emulsion formation with the precipitated asphaltenes. Product B was more amenable to agglomera-tton by the asphaltene-precipitation method than by the use of polar additives. For example 40% pentane (not shown) reduced the solids content to ~0.1% in 15 minutes set;:ling. Therefore, agglomeration of the mineral matter by asphaltenes is probably equally effective for both hydrophobic and hydrophilic minerals. The two hour settling period used in this Table 2B was longer than necessary for some of the examples given (see time dependency in Tables 4A and 4B).

The effect of additive concentration and time on the removal - of solids from SESA process liquor D-l, is shown in Table 3A. This material was low in solids (0.52%), but was chosen because of the availability of a large sample. All additives with the exception of petroleum ether were added in the form of 50% aqueous solutions.
From the control (No. 1), it is shown that 0.22% solids or 42% approxima-tely of the initial solids still remain in suspension after a 24 hour .. . - :.
': ~: ',, . , "
' , : ,:

: : - :' , ~ - . .:

settling period. Formic acid (Nos. 2-5) required concentrations of 1.5 to 2.0% in orcler to be ef~ective in short time periods. Chloral hydrate (Nos. 6-8) was about as effective as formic acid. As noted previously, resorcino] (Nos. 9-13) required much lower concentrations than any of the other polar additives tested. Petroleum ether (P.E.) with the boiling point range of 30-60 C was found to be slightly more effective than pentane for precipitating asphal~enes from SESA process liquor. It was therefore used in the remainder of the asphaltene-precipitation experiments. Nos. 14-20 show the effect of different lo concentrations of P.E. on the removal of solids. Concentrations in the range of 20 to 40% based on the SESA liquor or 60 to 110% based on the bitumen content, produced fairly rapid settling rates.
In No. 16, the D-2 SESA liquor was used. ~This product contai-; ned only 24.1% bitumen. The treatment with only 10% P.E. produced settLing rates at least as fast as 15% P.E. did with the 33% bitumen concentration in the D-l SESA product. The lower amount of~P.E. required ; for the more dilute bitumen solution reflected the higher content of~low molecalar weight aliphatic compounds present in the synthetic naphtha, as well~as;che lower viscosity of the suspending medium.
20~ When~No.~17 possessing only 15% P.E. (insufficient for gravlty settling)wa~s~repeated and~mildly centrifuged, the solids in the supernatent liqui~d wero~reduced~to 0.08% iD about 15~minutes. This~result would nalcate~that the cen~rifugal action produced by hydrocyclones would be~addqu~atd to~eÇEect rapid separatlons wiLh fairly low amounts ol 25~ prec~lplta~ing solvent. The remainder of Table 3A dedls with the joint use~of~polar additives and a precipitating solvent. When these additives ard~usd:d together, ~d synergistic effect occurs permittine~a considerdble reduction in both types of additives. For instance, in No. 21, it ~75 is shown that 5/~ P.E. and k~ss than 0.1% resorcinol reduced the solicls to less than 0.1% in a 6 hour period. This is about as effecLive as 20% P.E. alone (Lactor of four), or 0.375/0 r~sorcinol (factor of four also). Ilowever, doubling both types of additives (No. 22) resulted in more practical settling times. When either formic acid or chloral hydrate (Nos. 24 and 25) was used in conjunction with only 10% P.E., 1/10 as much of these polar additives was required to produce a similar settling rate (compare with Nos. 5 and 8). lJnexpected]y, the water soluble organic acids (acryLic - 26~ acetic - 28, lactic 10 - 29) which had a limited effect when used alone,were almost as effective in combination with P.E. as any of the effective polar additives were in combination with P.E. This would indicate that these acids preferen-tially associate with the asphaltene component of the bitumen. Polymeri-zed acrylic acid, No. 27, had no significant e~fect. Methyl alcohol ~ ~15 (No. 30) which showed no significant improvement when used alone ;~(Table 2A), had a slight favourable effect when used in combination with P.E.
The amount of precipitating solvent can therefore be decreased several fold by~the use of preferably about 0.1 to 0.2% of a suitable water soluble~polar additive. It is possible that as asphaltenes are precipitated from~the bitumen solution, they are preferentially attracted :
to polar additives alrèady adsorbed on the mineral surfaces. A "push-pull~effect" may therefore result. That is, the asphaltenes are being pushed out of solutlon by the aliphc~tic solvent and at the same time 25are being~pulled onto the mineral surfaces by adsorbed polar compounds.
This phenomenon probably results in a greater buildup of precipitated asphaltenes~on the mlneral surfaces.
Results~obtained indicate a~marked dependence on the amount ~: ~ : : . . , ; . : :

~3L2,~97~

of adclitive used, as shown in Table 3A, 1-5 and 9-13 which compares results for formic acid and resorcinol after 30 minutes settling.
~lthough both of these additives are ultimately equally efEective, a significantly greater concentration of formic acid ls required.
This is because, compared to resorcinol, formic acid is considerably more soluble in the bitumen solution, thus resulting in greater diffusion of this additive throughout the continuous phase. Consequently, the concentration of the additive in the dispersed aqueous phase is decrea-sed, causing a reduction in its ability to collect suspended particles, until a miscibility balance is set up.
The mechanism of collection possibly comprises the formation oE a fine dispersion of the aqueous phase under high-shear agitation conditions, followed by contact and adhesion between the suspended particles and these droplets in an analogous manner to emulsion flota-tion. Because of the affinity between the dissolved additive and the polar hydrocarbon coating of the particles, the solids will tend to remain at the hydrocarbon-aqueous interface or be absorbed into the droplets. Reduction in the degree of agitation allows coalescence of the soLids-rich droplets and thus, improved sedimentation rates.
Because the reagents are not surface-active it usually is necessary to have high solution concentrations in order to ensure that an adequate amount of the agent is present at the interface where contact with suspended particles occurs. Thus, when determining reagent requirements the amount of solution added may be changed rather than the concentration of reagent.
The use of anti-solvent agglomeration to remove solids has been shown to be effective in this situation. When an anti-solvent such as pent:ane or hexane, was added to the suspension, the precipitat;on .'. : ~: :. :
: ::: :.::
- -:: ~ :

. .. -: - , ~L2~

oL asphaltenes occurred. ~:ine solids in suspension acted as nuclei for the precipitating asphdlLenes and the surface coating formed provided a boncling medium between particles, resulting in the formation of agglomerates WitlI an increased sedimentation velocity. This efEect has been achieved with naphtha solutioIls of bitumen as shown in Table 3A. For this particular system a 30-60 C petroleum ether has been found to be more efective than pentane. Again the ash levels attainable were Very dependent on the quantity of anti-solvent added. This is illustrated t>y Table 3A Nos. 15-20, which shows ash content, after minutes settling, as a function of added petroleum ether. The amount of anti-solvent used is presented as a weight precentage of the original suspension treated. The quantity of asphaltene precipi-tated increased with the amount of anti-solvent used, and the greater availability of this particle coating agent resulted in larger agglome-rate size and more rapid settling. Addition of an asphaltene solvent,such as toluene, instead of the non-solvent resulted in only a marginal improvement in settling rate, presumably due to viscosity reduction.
Both o the individual treatments proved to be quite capable of removing solids to a satisfactory degree as shown in Tables 2B

and 3A Nos. 1-20. However, the reagent consumption necessary to achieve a rapid rate of sedimentation was quite high. Although an anti-solvent ~would be relatively easy to recover during subsequent bitumen upgrading steps, the polar additives would present more of a problem if recycling became necessary. As a result of these tests it is clear that polar 2~5 additive consumption could be reduced by co-addition with an anti-sol-vent. A typical result is shown in Figure 1 (derived from Table 3A
Nos. 10, 15 and 22) where a combination of resorcinol and petroleum ether was used. It is app~rent that some form of enhanced effect `' -14-. . . . .
: . ~, :

~

:.:~:. . .:: . .: . , ~7~7~

resuLts from the combination of treatments, with both the settling rate and the Einal ash level being signi~icantly improved, coMpared to I:he individual trefltments.
Another unexpected result of the combined treatment was 5 that the system response to a number oE normally weakly active agents was significantly enhanced. Results for several such compounds are shown in Table 3A ~os. 26-30. In the absence of an anti-solvent the same concentrations of these reagents would have given results only marginally better than the control run where no treatment was used.
Similarly, at the 10% level, anti-solvent alone gave only a marginal improvement over ~he control. The cornbination of treatments thus allows a wider selection of reagents to be used? an important considera-I:ion from an economic viewpoint.
EX~MPLE 3 15ll0T WATER PROCESS FROT~I
An attempt was made to remove Lhe solids from a ~roth f]otation concentrate sample which originated at Syncrude (Alberta). This bitumen concentrate had the following composition: bitumen, 38.5%, water, 56.1%;solids, 5.4%. A high proportion of water was occluded and easily 20removed. About 86% of the total water containing about 0.2~/~ solids was removed by agitating this material in a glass jar in a Spex (TM) shaker. rhree successive three minute periods were used with the displaced water being removed after each cycle. This upgraded bitumen concentrate (S.B.-C) now contained 74.4% bitumen and the water and 25mineral components were 18.8% and 6.8% respecLively.
The (S.B.-C) material was then diluted with both toluene and Syncrude naphtha to make the organic phase composition about 32%
bitumen, 10% toluene and 58% Syncrude naphtha. The toluene was added Lo prevent any a~phaltenes from precipitating, thus avoiding emulsifica-''`,:.
-15_ :: :: : ::. . :: : :

-:, . :
: :.. :: .; . , :-~;~}7Z~7~

~ion in Lhe presence ot ~he still high wa~er conte~t. A[ter mixing all the ingredients or lO minutes in a paint shaker, the mixture was al]owed to settle for 3 hours at room temperature. The supernatent bitumen solution (S.B.-D), now partially dewatered and partially demin-eralizecl, was found to contain about 1.1% water and 1.3% ash. AliquoLs of this (S.B.-D) material were subsequently treated with polar agglome-rating agents and a precipitating solvent. Results are shown in Table 3B.
'I'he slow settling rate for untreated (S.B.-D)material is shown in Table 3B No. 31. The addition of 50-60% P.E. produced rapid settling rates. Greater amouncs of P.E. were required than for the SESA product of similar bitumen concentration. This greater amount of precipitating solvent reflects the additional amount required in order to offset the presence of aromatic toluene. Also, emulsified water if present, would tie up clay perhaps in the form of clay films and slow down the settling rate. The polar additives e.g. Nos. 34-36 were all added in high so]ution concentrations (70% additive - 30% water) in order to add little additional water to the 1.1% water contained in the (S.B.-D) material. ~-lere again~ resorcinol was the most effective polar additive but requirements were high for all three polar additives. The higher additive requirements may again be owing to the presence of some emulsi_ fied water. It is possible that the joint use of a suitable demulsifier ancl agglomerating agent mig~t increase the settling rate of the solids in this material. The bulk of the water may also be better removed at elevated temperatures, or by other methods.
~X~MPLE 4 In Table 4, the amount of agglomerate residue, its composition and the asphaltene distribution are given for SESA product liquor D-l treated with P.E. alone and P.E. in combination with selected polar additives. In Table 4 No. l, no addi~ive was used, but the : . :
: :.: ; . , :

liquor was centrifuged severaL hours in order to obtain a well consoli-dflted agglomeraLe residue. By thi.s method, the residue was found to be about 1.7% of the original SESA liquor and containcd only about 2% of the asphaltenes. ITI the remaining five experiments, the residues were obt~;ned by gravity settling and were, therefore, not dircctly comparable to No. 1. In No. 2~ 40% P.E. based on SESA liquor was used as a precipita~i.ng solvent. The precipitating solvent produced a much larger residue (6. l~/o) which contained about 32% aspha].tenes a~d 31.9% o~ the total asphalLencs. The solids con-ent of the res:idue was only about 8.5%. With t:his hi.gh amount of asphaltenes in the residue, it is reasonable to expect that most of this materi.al does not precip:itate di.rectly on the mineral surfaces, but. precipitates randomly in the liquor. On settling, these solid-free asphaltenes may eventually form bridges between the m.ineral particl.es. All results ].5 in Table 4 wei-e based on newly formed agglomerate residues. After Sèparation, the newly formed residue can be sintered and the organic portion reauced considerably without any loss of solids by: 1) allowing it to stand at room temperature for time periods up to 4 hours, 2) allow-ing lt to stand at about 40 to 50C for about ~hour, 3) mild centri-20 fuging for a few minutes. For example, when a SESA residue which had ~ been obtained by treatment with ~0% P.E. (similar to No. 2) was allowed : ~ to stand 2 hours at room temperature, about 39% of its weight was exuded ~ ~ in the form of a mobile liquid. This organic composite material contai-:~ ned onl.y about 0.08% ash.
~: 2~ When 20% P.E. was used as a precipitating solvent (No. 3), ;~ the asphaltene content of the residue ~25.5%) was stiLl high and repre-sentea about 20.5% of the total asphaltenes. When insufficient precipi.-tating solvent was added to bitumen or bitumen solutions to precipitate all the asphaltenes, only the higher molecular weight fractions precipi-.
: -17-,. . ~ . .
- .- ~ , :.-. :. ; ~ . . .;.
:: , :: ~

~2~2~97~

t:ated. I~ has been established that the higher molecular weight frac-tions are preferentially attracted to the clay. This asphaltene com-ponent also contained a high proportion of the inherent ash, i.e.
nickel, vanadium, iron, etc., in the form of chelate structures.
Some of these metals especially nickel and varadium act as catalyst poisons in refining operatio~s. The removal of the high molecular weight fraction o~ the asphaltenes from the bitumen solution will reduce the metals and minerals as well as the hetero-atoms (N,S,O) which are also relatively enriched in this fraction, thus providing an improved feed for the subsequent upgrading operation.
In the resorcinol treated liquor (No. 4), only 4.1% of the original asphaltenes were found in the residue. With the joint addition of both P.E. and appropriate polar additives (Nos. 5-7), only 7 to ~ 9% of the total asphaltenes reported with the residue. Therefore ; 15 the synergistic eftect noted earlier with this combination of precipi-tating so]vent and polar additive is not because of excessive asphaltene precipitation but probably owing to a high proportion of precipitated asphaltenes being attracted to the mineral surfaces as a result of the adsorbed polar additives.

~: ~ : : ~ :: :

:

:

~, -18-~: : ~ : : : :: :

, ~ ' ' . ."''., ' ~ . :

~2~2~

TAsLE_l C~CSI'rION OF BIT~n~ SOUJrIONS

- &~ple Bitunen Water Solids~ Feed No. _ Ori~ iluent Cbnc. Cbntent Cblltent terial ~ ~. wlw %
A SESA ~rsol 27.5 0.1414:65' (2M2~ ~ 3~5 mesh) B SESA guncor N~E~htha 33.00.31 1 66 O~ersLze ~49~ s 325 mesh) + Varsol 0.24(4 003 o(b27~ < 325 mesh) D-l SESA Suncor ~phtha 33.20.37 (1 57) ~(~3do~ < 325 mes~l) D-2 SESA Suncor Na~ltha 24.10.30 0 68 more dllute S.B.* H.W. tas supplied) 3a.5 56.1 (14 03) Flotatiorl Cbnc.

S.B.-D~ H.W. S.B.-Diluted 32.0 1.1l 3 and settled) ~ Syncrude Bitunen (S.B.) ~lotation product from the ~t W~ter (H.W.) process.
$ S.B.-D.= S.B.-Diluted wi~l Syncru~e Naphtha + Tbluene, partiaLly dewatered, an~ partially demineralized.
Soli~ content irl brackets are on solvent/water free ~kasis :: :

,: . ' ,.,, . :': ~ ~ ..

,, : `

~ `
~: ~

T~LE 2A
EFFECT o~ So~lE DIVERSIFIED ADDITIV~S ON SOLIDS
IN B CONCENTR~TX*
No. Additive ~ Additive Solids Content**

Control Nil l.l (3.33) 1 Aeetie Aeid 3% 0.65 11.97) 2 Acrylie Acid 3% 0.50 (l.52) 3 Polyacrylie ~eid 3% 0.76 (2.30) 4 Oxalie Aeid 0.6% 0.64 (l.9~1) ~ialonie Aeid 3% 0.50 (l.52) 6 Glyeollie Aeid 2.0% 0.58 ~1.76) 7 I.aetie ~eid 2.0~ 0.62 (1.88)

8 Tannie Acid 3% 0.78 (2.36)

9 Pyruvie Aeid 3% 0.66 (2.00) ~enzene Sulfonie Aeid2.5% 0.22 [0.67) ll Sulfamie Aei(l 2.0~ 0.63 (1.91) 12 l~y~lro~Luorie Aeid2.0% O.ll (0-33) l3 Amrnonia 3~ 0~62 (l.88) 14 ~lethanolamine 3~ 0.35 (1.06) Sodium Silieate 3~ l.l (3.33) 16 50di~ ~Iydroxide 2~ 0.78 (2.36) 17 Methyl Aleohol 3~ 0.82 (2,48) lB ALlyl Alcohol 3~ 0.64 (l-94) l9 Tetrahydrourfuryl Aleohol 3% 0.74 (2.24) E`urE~ryl ALeohol 3% 0.62 (0.88) 21 Snlfur Dio~i-le 0.5% 0.86 (2.61) 22 Forma1(1~hyde 3~ 0.72 (2.1E3) 23 Urea 3% 0.76 (2.30~
24 Ethyl Urea 3~ 0.71 (2.15) 4-~ly(lroxypyri~3ine 3~ 0.61 ~1.85) 26 lly~roxybutanone 3% 0094 (2.85) 27 Pyridine metllanol 33 0.67 (2.03) 28 ~lethyl ~eetoacetate3% Q.79 (2-3~) 29 Acetonitri.le 3~ 0.68 (2.06) ..
: ~2 llonr settl.in~ perio~
** ~oli~ls content in ~rackets are on solvent/water free basis.

;:
~:,... .
....
,, , , :, T~BLE 2B
EFI~x~r OF SE11X~FED ADDITIVES ON SOLIDS RE~V~L
F~oM SE~ PROlY~CT LIUUORS

No~ Ad~itive ~ P~ditive So1ids C~ntent~ - 2 Elrs. SettLin~

(~.7) (5.2) (4.0)(1.57) 1 Crigina1 t~t Settled 1.3 1.7 1.3 0.52 (3.20) (3.33)(1.9~)(1.08) 2 Cbntrol Na1 0.88 1.1 0.64 0.36 (0.15) (0.5~3)(0.34~(0.15) 3 Fbrmic Acid 2~ 0.04 0.19 0.11 0.05 (0.47) (1.58) (0.36) 4 Fbrmic Acid 1~ 0.13 0.52 -- 0.12 (0-~5) (0.4a~ (0.24) t~leic Acid 2g 0.07 0.16 -- 0.08 (0.~0) (0.67)(0.46) (0.18 6 ~lloral ~lrate 2~ 0.11 0.22 0.15 0.06 (0.~5) 7 alloral H~drate 1~ - -- 0.15 (0.6~)(0.28) (0.18 ~ 8 ~esorcino1 0.37~ -- 0.21 0.09 0.06 : : (0-78) : ~ 9 hesorcinol ~ 0.19~ 0.26 (0.70) (0.36)

10 ~ Cat~chol 0.75~ -- 0.23 -- 0.12 (0.48) (0-65) (0.53) ntane 20~ O.ll 0.1g -- 0.16 (0.84)~ (1.40) (0.60) :12 ~ Pen~ne 10~ ~0~21 0.42 -- 0.18 ~t SbLi(lg (~ontent in bFacke~ .are on solvent/water free bRsis s ~ -21-` , ., , .~ , ' .' ' ! , .
' ~' ' , " ' ' .

TMIE 31~
~cr OF ~DOITIVE c~Nc~rrRATIoNs WIT~I T~E ON RESIDUAL SOLlr~6X IN I.IQIX~R ~l ~e~tin~ Treatin~ I Time in ~burs No. A~ r)t ~1~nt ¦0.25o.5 1 2 4 6 2~ I ~ar~rl<s __ _ _ _ _ _. _ (1.17) (1.05)(1.08)(0.84) (0.66) 1 (bntrol ~lil 0.39 0.35 0.36 0.28 0.22 ~it.=33.2 (1.11) ~0.99)~0.84) (û.7S)(0,63) ~0.2~) 2 Ebrmic Aci~ 0.31 0.370.33 0.28 0.25 0.21 0.08 ~l.02) (0.69) 3 Ebnn~c Acid 0.75 0.34 0.23 (0-~37) (0.30) (0.27) 4 Ebrmic Acid l.5 0.29 0.10 0.09 (0.24) (0.15) Ebrrnic Acid 2.0 0.08 0.0 5 (0.63) (~chnical 6 (~loral ~rate0.75 0.21 Gr~lle) (0.36) 7 Chloral ~rate 1.5 0.12 (0.30) (O.la) 8 Chloral l~rate 2.0 0.10 0.06 (0.81) (0.78)(0.75)(0.75)(0.63)(n?chnicaL
g Resorcin~l 0.09 0.27 0.26 0.25 0.25 0.21 Gra~
(0.96) ~0.~4) (0.7a) (0.69) (0.60) (0.30) 10 E~3sorcinol 0.19 0.32 0.28 0.26 0.23 0.20 0.10 (0.81) (0.6~) (0.21)(0.12)

11 l~esorcinoL 0.37 0.29 0.23 0.07 0.04 (0.5~)(0.30) (0.12) (0.09)

12 ~orcinol 0.75 0.18 O.lO 0.04 0.03 (0.30)(0.2~) (0.12) (0.12)

13 P~sorcinol l.5 O.lO 0.08 0.04 0.04 (0.89) (0.89)(0.76)(0.663~0.44)(l~chnic~l

14 P.E. 30-60 5.0 0.28 0.28 0.24 0.21 0.1~ Gra(le) (1.03) (0.70) (0063) (0.50)(0.36) P.E. 30-6010.0 0.31 0.21 0.19 0.15 O.ll *~ (l.OO) (0.50) (0.23) 16 P.E. 30-60lO.O 0.22 0.11 O.OS Bit.=24.L3 (1.00) (0.S2) (0.52) (O.~la)(0.35) 17 P.E. 30-6015.0 0.2~ 0.15 0.15 0.14 0.10 (0.6S) (0.4~) (0.32)(0.22)~0.1~) 18 P.E. 30-6020.0 0.18 0.13 0.09 0.06 0.04 `` (0.31) (0.2~) 19 P.E. 30-6030.0 0.08 0.06 ., , .: .. .: ,.. .
,-~ , : .

~MI E 3A ~ cont ' d ) Treatin~ Treating ~ Time in Ebur~
No. ~ 0.25 O.S 1 2 4 6 24 I R~nArl~.s (~. 13 ) tO.0~ ) P.E. 30-6040.0 0.03 0.02 (1.0~)(0.95)(0.79)(0.57)~0.2S) 21 P.E. 5~0 ~ 0.34 0 30 0.25 0.18 0.08 sorcinol 0 . 09 ~0.36)(0.26)(0.17~ (0.10) (0.13 22 P.E. 10-0 } o 11 0 08 0 05 0.03 0-04 + ResorcLnol 0.19 (1.13) (0.93) (0.70) t0.46) (0.46) (0.26) (0.10) 23 P.E. , 1 ~ 0.34 0.28 0.21 0.14 0.14 0.08 0.03 ~ ~brmic Acld 0.09 (0.30) (0.10) 24 + FormLc Acid 0 19~ 0-09 0.03 (0.40) (0.2G) (0.13) ioral E~rate 0 19~ 0-12 0.08 0.04 (0.40)(0.20)(0.17) 26 pCE l~ic l~cid 0 19} 0.12 0.06 0.05 (0.83) (0.70) Acry~ic ~ d0 19~ 0~25 0.21 ~M.W. 6000) (0.46)(0.30) (0.20)(0.10) 28 P. E. 10.0 ~ 0.140 09 0. 06 0. 03 Pcetic A~id 0.19 (0.56) (0.33) (0.20~ (0.13) : ~:29 P. E. 10.0 } 0.17 0.10 0. 060. 04 * l~ctic ~cid 0.l9 (0.83)(0,53j (0.17) ~30 P.E.~ 10.0 ~ 0 25 0 16 0.05 ~thyl Alco~lol 0.19 P5 - ~etroleun ether, boilin~ point rat~3e 30 to 60C.
Process liql~or ~2 (24.1~ ~itunen) x ~csil~ual solills i.n br,ic~cets base~ on bltunen content o~ snspeAsiot s.

~: :

: -23-,. :. . ..

, ~ : ~., ,, : :::
.

, ~g~

TA~LE _ EFFECT OF ADDITIVE CO~c~NTRA'rI~S ~TH T~ ~ ON RESI~AL SOLIDS~ IN LI~X~R S.B.-D.
%

Treatin~ Treating ~ - - Time in Ibur~ I
No. ent _~ent ¦0.25 0.5 1 2 4 6 24 ¦ hLn~rks (3 44) ~3.00) (2.~1) 31 (obntrol) Nil l.l 0.96 0.90 32~ Bitllnen (0.25) (0.15) 32 P.E. 60 0.050.03 (0.47) (0.23) 33 P.~. 50 0.10 0.05 (2-53) (l-l9) (0-47) (~dditive as 34 ~brmic Acid 3.0 0.~1 0.38 0.15 70~ Sol.) (2.31) (0.09) (A~ditive as 35 Chloral l~drate 3.0 0.74 0.03 70% Sol.) (1.34) (0.28) (A~ditive as 36 Resorcinol l.S 0.43 0.09 703 ~al.) * S.B.-D. = S~ JDE IIOT ~AT~R Bit~men Cbncentrate, diluted, partially dewatered an(l parti.ally d~nineraliæe~l.
~esidual solids io brack~ts based on bitumen content of suspenslons.

: : :

~:
: ::: ::
:

::

: :~:

:
: : :

:

:~

.:. . ~. :. ::,. .. : :;.: , , . , . ,,:~, . : .

~p I o ~ u~ 7 rl o $ ~ ,~ N ,~

~1 o ~ ro o - ~
___ r~ r~

, ~ ~ ~ r~ ~ ~ O
fi~ 8 5~ g ~ o r N -i E~I ~
O
L~,~ O CO

2 o o c dO ~
3 c~
~1 D~ X~ w~ w~ ~
8 ~ a' &~
~ In ~ X

,. ~.
- ,...

.. . `, . , ~. . ` - .
' ' , ;.' ' .:.. .. .
. .
.
~: : ` ......... :

~ . .. ` ... :, ~`'' '.: ~ ~

Claims

CLAIMS:

1. A process for reducing the hydrophobic solids content of petroleum oils, bitumen solutions or concentrates, or heavy oils and the like, containing fine hydrophobic solids, comprising:
- dispersing an additive throughout the solids-containing feed material, said additive being selected from the group consisting of resorcinol, catechol, formic acid, maleic acid or anhydride, chloral hydrate and asphaltene-precipitating solvents, and mixtures thereof, in an amount sufficient to cause small agglomerates comprising the hydrophobic solids to form; asphaltenes being present when asphaltene-precipitating solvents are selected and water being present when the other additives are selected; and - removing the agglomerated solids from the residual liquid.

2. The process of claim 1 wherein the feed material is selected from a tar sand solvent extraction liquor, a bitumen concentrate from tar sand hot water processes, a heavy oil, and an oil obtained from enhanced oil recovery techniques.

3. The process of claim 1 wherein the feed material contains asphaltenes, the additive used comprises asphaltene-precipitating solvents, and the agglomerated solids include precipitated asphaltenes.

4. The process of claim 3 wherein the feed material is a tar sand solvent extraction-spherical agglomeration process liquor.

5. The process of claim 3 wherein the additive comprises an asphaltene-precipitating solvent and at least one of said other addi-tives listed.

6. The process of claim 3 wherein the additive comprises an asphaltene-precipitating solvent and a polar organic additive having water solubility and low miscibility with hydrocarbons.

CLAIMS (cont.)

7. The process of claims 1, 2 or 3 wherein the amount of the addi-tive used is within the range of about 0.2 to about 3% w/w based on the feed material, except for asphaltene-precipitating solvent which is within the range of about 5 to about 40%.

8. The process of claims, 1, 3 or 4 wherein the asphaltene-precipi-tating solvent is selected from petroleum ether, pentane, hexane and low boiling paraffinic refinery products.

9. The process of claims 3, 5 or 6 wherein the asphaltene-precipi-tating solvent is separated from the residual liquid and recycled.

Agglomerates comprising fine hydrophobic mineral solids and as-phaltenes on the surfaces of said solids, and containing a polar organic additive having water solubility and low miscibility with hydrocarbons.

11. An additive mixture for hydrophobic solids agglomeration com-prising:
(a) an asphaltene-precipitating organic solvent, and (b) an agglomerating-effective amount not more than about 30% w/w based on said solvent, of a water-soluble polar organic compound having low miscibility with hydrocarbons.
12. A process for reducing the content of fine hydrophobic solids of solvent-diluted materials selected from petroleum oils, bitumen solutions or concentrates, shale oil, and heavy crude oils, said materials having fine hydrophobic solids present, comprising:
i) (a) in the case of asphaltene-containing materials dispersing an asphaltene-precipitating solvent and optionally a concentrated aqueous solution of an additive selected from the group consisting of resorcinol, catechol, formic acid, maleic acid or anhydride, chloral hydrate and mix-tures thereof, throughout said materials, CLAIMS (con-t.)

12. (cont.) (b) in the case of such asphaltene-free materials dispersing a con-centrated aqueous solution of an additive selected from the group consisting of resorcinol, catechol, formic acid, maleic acid or anhydride, chloral hy-drate and mixtures thereof throughout said materials, - the amounts dispersed in (a) and (b) and the amounts of water present being sufficient to cause the formation of small agglomerates comprising the hydrophobic solids, ii) agitating to form agglomerates of the hydrophobic solids, and iii) removing the agglomerated solids from the residual material.

13. A process for reducing fine hydrophobic solids content of solvent-diluted materials selected from petroleum oils, bitumen solutions or con-centrates, and heavy crude oils, said materials containing asphaltenes, comprising:
i) dispersing an asphaltene-precipitating solvent and an aqueous solu-tion of a polar additive selected from acrylic acid, acetic acid and lactic acid, throughout said materials the amount of the precipitating solvent, of water and of the additive being sufficient to cause the formation of small agglomerates of the hydrophobic solids, ii) agitating to form agglomerates of the hydrophobic solids, and iii) removing the agglomerated solids from the residual material.