Chemistry of Asphaltenes - ACS Publications - American Chemical


Chemistry of Asphaltenes - ACS Publications - American Chemical...

0 downloads 99 Views 2MB Size

13 Characteristics of Tar Sand Bitumen Downloaded by UNIV OF MISSOURI COLUMBIA on November 27, 2013 | http://pubs.acs.org Publication Date: January 1, 1982 | doi: 10.1021/ba-1981-0195.ch013

Asphaltenes as Studied by Conversion of Bitumen by Hydropyrolysis JAMES W. BUNGER and DONALD E. COGSWELL Department of Mining and Fuels Engineering, University of Utah, Salt Lake City, UT 84112 Chemical characteristics of tar sand bitumen asphaltenes were examined by inference from spectroscopic and hydropyrolysis processing studies of virgin bitumen and deasphaltened bitumen. While average properties differ between asphaltenes and maltenes, major overlap between these fractions is experienced with respect to carbon type, heteroatom type, molecular weight, and volatility. Theoretical considerations suggest that given species are partitioned between the maltene and asphaltene portions further contributing to a lack of chemical distinction between maltene and asphaltene species. When subjected to conversion by hydropyrolysis, both maltenes and virgin bitumen yielded products exhibiting pentane insolubles with maltenes yielding about 11% and asphaltenes yielding (by calculation) 20% insolubles. A strong precursor-product relationship between asphaltenes and pentane insolubles in the product does not exist. No chemical features (e.g., carbon type, heteroatom type, molecular weight, or structure) were found that distinguished species found in the asphaltenes fraction from those in the maltenes fraction. Thus, asphaltenes appear to be chemically nondistinct from maltenes and correlations between asphaltene content and processability, while useful to an engineer, are chemically fortuitous.

A

sphaltenes have been the subject of considerable discussion a n d controversy i n the literature. Controversy a n d a m b i g u i t y arise largely because of the lack of c h e m i c a l definition of asphaltene mixtures for w h i c h composition is dependent u p o n the source material a n d m e t h o d of isolation. W h i l e 0065-2393/81/0195-0219$0.500/0 © 1981 American Chemical Society In Chemistry of Asphaltenes; Bunger, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1982.

Downloaded by UNIV OF MISSOURI COLUMBIA on November 27, 2013 | http://pubs.acs.org Publication Date: January 1, 1982 | doi: 10.1021/ba-1981-0195.ch013

220

CHEMISTRY O F ASPHALTENES

e m p i r i c a l correlations often exist between increased asphaltene content a n d increased difficulty i n processing of feedstocks, the c h e m i c a l basis of these correlations remains to be demonstrated. It is possible that a cause-and-effect relationship between molecules c o m p r i s i n g the asphaltene fraction a n d difficulty i n processing does not exist a n d that these relationships are largely fortuitous i n that the chemistry i n v o l v e d i n asphaltenes isolation is only secondarily related to chemistry of processing. If such is the case, m u c h of the a m b i g u i t y s u r r o u n d i n g asphaltenes derives f r o m the incorrect assumptions about the chemistry of complex h y d r o c a r b o n mixtures i n f e r r e d f r o m e m p i r i c a l correlations. In this work, w e approached the p r o b l e m of chemistry a n d definition of asphaltenes w i t h the f o l l o w i n g hypothesis: (i) asphaltenes are not structurally u n i q u e f r o m nonasphaltenes; (ii) average molecular weights for asphaltenes are not appreciably higher than those f o r nonasphaltenes, a n d strong associa­ tive forces account f o r the inordinately h i g h measured values f o r asphaltene molecular weights; a n d (iii) the in-situ properties of those molecules compris­ i n g the asphaltenes fraction differ considerably f r o m those manifest w h e n isolated i n a n asphaltene fraction. T o test these hypotheses, a tar sand b i t u m e n containing 20 w t % pentane asphaltenes was characterized a n d processed b y hydropyrolysis before a n d after r e m o v a l of asphaltenes. Product yields a n d structure were d e t e r m i n e d a n d the influence of asphaltenes o n results was d e t e r m i n e d b y inferrence. Feedstocks a n d products were characterized a c c o r d i n g to elemental analysis, physical properties, simulated distillation, a n d carbon-type analysis. Infer­ e n c e s m a d e i n this study are discussed i n the context of the reported literature.

Experimental S a m p l e Source a n d P r e p a r a t i o n . The Sunnyside tar sand sample was freshly mined from the old asphalt quarry east of Sunnyside, Utah. Bitumen was extracted with benzene, filtered, and the solvent removed as previously described (J). Deasphaltening was accomplished using 40:1 (v:w) n-pentane-to-sample in which the mixture was stirred and allowed to digest overnight at room temperature. The precipitate was filtered and washed with sparing quantities of chilled n-pentane. The solvent was removed from the fractions as before and the respective fractions from three equalsized deasphaltening runs were combined and used directly for property measure­ ments and as the process feedstock. A third process feedstock was prepared by dissolving a known amount of virgin bitumen in benzene, and combining with this solution 0.7 times the normal amount of asphaltenes. The solvent was removed by rotary evaporation to generate a sample "1.5 normal" in asphaltenes. It was not confirmed that this procedure resulted in completely homogeneous dispersion of asphaltenes in virgin bitumen. E l e m e n t a l A n a l y s i s a n d P h y s i c a l P r o p e r t i e s . Elemental analysis was accomplished by conventional microanalytical techniques i n a commercial testing laboratory. Density, refractive index, average molecular weight (VPO), Conradson carbon residue, and ash content were determined by standard methods. Viscosity was determined by a cone-plate viscometer. Simulated distillation was accomplished using a y " χ 18" column of Anachrome Q, 3% Dexil 300, programmed from - 3 0 to 4

In Chemistry of Asphaltenes; Bunger, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1982.

13.

BUNGER AND COGSWELL

Tar Sand Bitumen

221

Asphaltenes

+ 350°C at 10°/min and an F I D detector, and using as an internal standard an equal volume mixture of C - C n-alkylbenzenes. (See also References 1 and 2). Gas analysis was performed on a %" χ 240" column packed with Chromosorb 102, programmed from —30 to + 2 5 0 ° C at 10°/min and an initial isothermal time of 5 min and using a T C D detector. C a r b o n - 1 3 N M R S p e c t r o s c o p y . High resolution C N M R spectra of maltenes, bitumen, and liquid products taken up in a 1:2 mixture of C D C l were obtained on a Varian XL-100-15 F T spectrometer equipped with a V-4412 probe and a 2.5 megaword disc. A pulse angle of 45° and a repetition period of 60 s were selected to avoid saturation of any nonprotonated carbons. Gated decoupling was also employed to suppress N O E effects in the line intensities. These precautions have been shown for mixtures of model compounds to minimize error in integrated line intensities to less than 10% (3). The combined cross-polarization and magic-angle spinning (CP/MAS) spectra of the asphaltenes were obtained on a single-coil, double-tuned probe similar in design to that of Cross, Hester and Waugh (4) but which has been adapted for an electromagnet. Equipped with a D 0 external lock, this probe accepts either 12-mm liquid sample tubes or the magic-angle spinner (5). Radiofrequency fields of 17 G and 12 G at 25.16 M H z and 100.06 M H z respectively were obtained at 90 W power. Isolation between the C and * H channels is in excess of 60 db. The C P / M A S spectra were taken with the single contact (6) sequence. The 90° pulse for the H spin locking is continuously variable and the amplitude of the C irradiation controlled to within 0.1 db for matching the Hartmann-Hahn (7) condi­ tion. The high-speed magic-angle spinner used has been described in detail elsewhere (5). The particular one we used is constructed of Macor, a machineable glass that gives no background in the C C P / M A S spectra. Additional experimental details and a discussion relating this procedure to solid coals and coal derived liquids have been previously published (8). H y d r o p y r o l y s i s Process. The hydropyrolysis reactor, consisting of a coiled stainless steel tube / i " i d . χ 236" long has been previously described (9,10). Average residence time was calculated from the volumetric throughput and reactor volume. Volumetric measurements made at atmospheric pressure and room temperature were corrected to process conditions assuming ideal gas behavior. (The reaction mixture is typically greater than 97 mol % hydrogen, so residence time may be slightly overestimated by this assumption.) Accumulated coke was determined after one or more runs by burning the coke in air and by measuring the C 0 evolved. Hydrogen consumption (uptake) was calculated for those runs that produced essentially only liquids and gases by first conducting a material balance on the carbon and calculating hydrogen content from the composition of the carbon-containing gases and the elemental analysis of the liquids. Heteroatoms fed but not found in the organic liquids were assumed to exist in their fully hydrogenated form, that is, N H , H S , H 0 , for this calculation. Thus, hydrogen consumption is not underestimated by this procedure. For further description of results of hydropyrolysis and calculation of hydrogen consumption, see References 10-12. 9

1 6

1 3

Downloaded by UNIV OF MISSOURI COLUMBIA on November 27, 2013 | http://pubs.acs.org Publication Date: January 1, 1982 | doi: 10.1021/ba-1981-0195.ch013

3

2

1 3

1 3

l

1 3

3

6

2

3

2

2

Results Feedstock Characteristics.

T h e Sunnyside tar sand sample contains 9.3

wt % b i t u m e n . T h e extracted b i t u m e n was subjected to deasphaltening a n d results of three runs were 20.6, 19.9, a n d 21.4 w t % (20.6% average) of the total b i t u m e n . T h e elemental analysis a n d p h y s i c a l properties of the b i t u m e n ,

In Chemistry of Asphaltenes; Bunger, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1982.

222

CHEMISTRY O F ASPHALTENES

Downloaded by UNIV OF MISSOURI COLUMBIA on November 27, 2013 | http://pubs.acs.org Publication Date: January 1, 1982 | doi: 10.1021/ba-1981-0195.ch013

maltenes, asphaltenes, a n d the asphaltene-enhanced b i t u m e n are g i v e n i n T a b l e I. T h e v i r g i n b i t u m e n is t y p i c a l of U i n t a Basin samples (14), w h i c h are of extremely h i g h molecular weight a n d viscosity, but also of h i g h h y d r o g e n content, l o w sulfur, a n d appreciable nitrogen content c o m p a r e d w i t h m a n y petroleum residues. T h e molecules c o m p r i s i n g maltenes are of a higher average h y d r o g e n content than those i n the v i r g i n b i t u m e n , a n d the asphaltenes contain a higher average heteroatoms content. T h e weighted average of the maltenes a n d the asphaltenes elemental analysis approximates that of the v i r g i n b i t u m e n , l e n d i n g credence to the analysis results. Physical properties show the b i t u m e n a n d subfractions possess h i g h molecular weights a n d l o w A P I gravities. Asphaltene content a n d carbon residue of the v i r g i n b i t u m e n are somewhat higher than other U i n t a Basin bitumens (14). T h e average molecular weight of heavy b i t u m e n fractions, particularly the asphaltenes, is a subject c o m m o n l y debated (14-18). T h e results g i v e n i n T a b l e I are not thought to be representative of the true molecular weights, a n d values as h i g h as 8000 are so erroneous as to be meaningless. T h e results for the asphaltenes show that the intermolecular associations c a n be partially disrupted b y c h a n g i n g the polarity a n d solvating properties of the solvent. T h e apparent higher molecular weight for v i r g i n b i t u m e n exhibited w h e n using p y r i d i n e is presently inexplicable. H o w e v e r , it has previously been observed that solvents of higher polarity c a n lead to a more associated state for nonpolar species ( 14). T h e v i r g i n b i t u m e n a n d fractions were characterized b y C N M R . Results shown i n F i g u r e 1 show that the spectra of the maltenes a n d the v i r g i n b i t u m e n are virtually superimposable. T h e h i g h resolution spectra reveal several resolved peaks i n the saturates region attributable to n - a l k y l species, some of w h i c h are of notably lower intensity i n the maltenes spectrum, but i n general, the vast majority of the saturates carbon (>80%) is f o u n d under the envelope a n d is probably d u e to complex a l i c y c l i c and/or isoparaffin structures. These results are consistent w i t h a n earlier observation (1) that saturated hydrocarbons f r o m U i n t a Basin b i t u m e n are h i g h i n naphthenes a n d l o w i n free paraffins. The C N M R spectra of the asphaltenes obtained i n the solid state is shown i n F i g u r e 2. These results are c o m p a r e d w i t h a spectrum of the maltenes obtained i n solution ( C D C 1 ) . T h e solution spectrum of maltenes has been artificially broadened to produce a p p r o x i m a t e l y the same resolution as the solid-state spectrum of the asphaltenes. It is very apparent f r o m F i g u r e 2 that aside f r o m the relative intensities of the aromatics vs. the saturates portion that the two spectra are essentially identical. T h i s observation has important implications i n that it shows that the gross h y d r o c a r b o n structure of asphaltenes (relative abundance of carbon types) is not significantly different f r o m that of the maltenes. T h e semiquantitative results of the C N M R are g i v e n i n T a b l e II. These results must be considered p r e l i m i n a r y at this t i m e since evidence has not been 1 3

1 3

3

1 3

In Chemistry of Asphaltenes; Bunger, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1982.

13.

BUNGER AND COGSWELL

? ^

? >C S cj »

Downloaded by UNIV OF MISSOURI COLUMBIA on November 27, 2013 | http://pubs.acs.org Publication Date: January 1, 1982 | doi: 10.1021/ba-1981-0195.ch013

« Ι

0) ce

Tar Sand Bitumen

00 Ο 00 Ο -Η ^OOC^COOOLOOLO ςΟΟΟΟ'-Η'-Η'-Ηϊ^· 00 γ-Η

223

Asphaltenes

<

< Ζ

Ζ

Ο 00 r H ρ-Η C0



b© fi •«m

r

ce

a.

^

co

00

l>

CD

LO

^

ζζ

ο LO (Ν



o CD CD TJ< CN| CD LO r^CDc0CDCDO5a> 00

Ή

όο LO 00 Ο

CO

ι-Η

ι-Η

CO

1

fi

i

υ

i~H

00 Γ­

S

^

CD C5

JO E2

SP

α

ϊΊ a

.2

bp © CM ©

Ë c

to «

r j SS Ζ

c £

o ce & 0

Ο Ε ςΛ <

<

^

o

s i ? ce

l i s W22

In Chemistry of Asphaltenes; Bunger, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1982.

Downloaded by UNIV OF MISSOURI COLUMBIA on November 27, 2013 | http://pubs.acs.org Publication Date: January 1, 1982 | doi: 10.1021/ba-1981-0195.ch013

224 CHEMISTRY OF ASPHALTENES

In Chemistry of Asphaltenes; Bunger, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1982.

13.

BUNGER AND COGSWELL

T a b l e I I . S e m i q u a n t i t a t i v e Results of Total Aromatic Carbon Maltenes Virgin bitumen Asphaltene

Downloaded by UNIV OF MISSOURI COLUMBIA on November 27, 2013 | http://pubs.acs.org Publication Date: January 1, 1982 | doi: 10.1021/ba-1981-0195.ch013

225

Tar Sand Bitumen Asphaltenes

17 19 33

1 3

CNMR Total Saturate Carbon 83 81 67

a c q u i r e d to f u l l y ensure that nonprotonated carbons are not saturated b y a too r a p i d repetition period. A n y error due to this effect w o u l d be to reduce the apparent concentration of aromatic carbon; however, this effect w i l l be generally less than 25% of the total aromatic carbon under the conditions r u n . T h e results indicate that for the b i t u m e n a n d maltenes only about one carbon in six is b o u n d i n an aromatic ring, w h i l e for the asphaltenes the proportion is about one i n three. E v e n w i t h the higher aromaticity, the asphaltenes fraction consists p r e d o m i n a n t l y of saturated carbon a n d any theoretical consideration of structure or processability must take this factor into account. It is apparent f r o m the elemental analysis a n d C N M R data that w h i l e the asphaltenes are of a higher molecular weight, aromaticity, a n d heteroatom content than the maltenes, the gross h y d r o c a r b o n structure is not significantly different between these two fractions. 1 3

C o m p o u n d type analysis was not conducted i n this work, but it is instructive to look at some literature results. W i t h respect to heteroatoms, tar sand b i t u m e n a n d petroleum asphaltenes have been variously reported as containing p r e d o m i n a n t l y polar heteroatoms, p r i n c i p a l l y oxygen types (18) or nonpolar heteroatoms, p r i n c i p a l l y nitrogen types (14). T h e difference i n these reported results apparently relates to the m e t h o d of analysis i n w h i c h the f o r m e r is a direct determination, the latter an indirect determination. C e n t r a l to an understanding of the composition of complex h y d r o c a r b o n mixtures is that a l l spectroscopic measurements, elemental analyses, a n d physical properties represent average values a n d that a distribution about the average exists. A s i n m a n y cases, the distribution about an average value is often w i d e c o m p a r e d w i t h the differences i n averages between a maltenes fraction a n d asphaltenes fraction. Thus, there is frequently b r o a d overlap, a n d for any g i v e n value i n the v i c i n i t y of the average there w i l l be species i n both the asphaltenes a n d maltenes possessing essentially identical characteristics. S i m u l a t e d D i s t i l l a t i o n Results. Important insight concerning the molecular-size distribution of asphaltenes vs. maltenes is gained b y the simulated distillation data. T h e b o i l i n g point distribution curves are shown i n F i g u r e 3. These curves were d r a w n as follows. Quantitative simulated distillation data was obtained on the v i r g i n b i t u m e n a n d the maltenes. D i r e c t i n f o r m a t i o n is obtained u p to a n o m i n a l b o i l i n g point of 5 3 5 ° C shown b y the vertical dashed line. T h e area under the curve for the nonvolatile portion is

In Chemistry of Asphaltenes; Bunger, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1982.

Downloaded by UNIV OF MISSOURI COLUMBIA on November 27, 2013 | http://pubs.acs.org Publication Date: January 1, 1982 | doi: 10.1021/ba-1981-0195.ch013

226

CHEMISTRY OF ASPHALTENES

Figure 3. Simulated distillation boiling point curves for virgin bitumen, maltenes, and asphaltenes d e t e r m i n e d d i r e c t l y f r o m the simulated distillation data. T h e e n d point for the nonvolatile portion of the c u r v e was a p p r o x i m a t e d b y k n o w i n g the area under the curve a n d by using molecular weight data to estimate the m e a n b o i l i n g point (this step a d m i t t e d l y requires some judgment). T h e curves were d r a w n i n a smooth fashion rather than some b i m o d a l configuration because, f r o m the standpoint of geochemical maturation, a distribution w i t h a single m a x i m u m is most reasonable. T h e asphaltenes curve is the difference between the other two curves. D e c i m a l values are the weight fraction of v i r g i n b i t u m e n represented under the curve. Attempts to obtain a distillation curve d i r e c t l y o n the asphaltenes fraction resulted i n essentially no F I D response. T h i s result is attributed to the h i g h free energy of association of these molecules. O n c e these species are a l l o w e d to associate, it is difficult to disassociate them. T h e results i n F i g u r e 3 suggest that volatile molecules are contained i n the asphaltene fraction a n d that the overlap i n molecular weight a n d volatility between asphaltenes a n d maltenes is substantial. H y d r o p y r o l y s i s R e s u l t s . T h e maltenes, v i r g i n b i t u m e n , a n d b i t u m e n enhanced to 1.5 times the n o r m a l asphaltene content were processed b y hydropyrolysis at 5 0 0 ° C , 2000 psig h y d r o g e n pressure, a n d 30-s gas residence time. T h e process conditions were selected o n the basis of earlier results (10-12) for tar sand bitumens, w h i c h i n d i c a t e d that gas m a k e a n d the propensity to f o r m coke increased w i t h increasing temperature, a n d that a m i n i m u m h y d r o g e n pressure of 1200-1500 psig was r e q u i r e d to i n h i b i t coking. A moderate temperature of 5 0 0 ° C and, correspondingly, a long residence t i m e was selected i n an attempt to accentuate differences i n results for the three feeds. H y d r o p y r o l y s i s is p a r t i c u l a r l y useful for s t u d y i n g asphaltenes since no coke is f o r m e d a n d the largest, most aromatic molecules are retained i n l i q u i d products where they can be observed a n d characterized. G r a v i m e t r i c results of the process yields are g i v e n i n T a b l e III. T h e percentage of h y d r o c a r b o n gases was calculated to close the material balance on carbon a n d the calculated n o n h y d r o c a r b o n products were d e r i v e d b y

In Chemistry of Asphaltenes; Bunger, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1982.

13.

BUNGER AND COGSWELL

Tar Sand Bitumen

227

Asphaltenes

closing the material balance on the heteroatoms. T h e inclusion of the missing heteroatoms

as N H , H S , a n d H 0 is done only to p r o v i d e a basis for 3

2

2

calculating the m a x i m u m theoretical h y d r o g e n c o n s u m p t i o n a n d is not meant to indicate that these products were actually measured. T h e results i n T a b l e III show that the v i r g i n b i t u m e n that contains the asphaltenes p r o d u c e d relatively m o r e gas a n d n o n h y d r o c a r b o n products than d i d the maltenes. T h i s trend w i t h respect to gases a n d liquids appears to be c o n f i r m e d b y the results of the r u n w i t h the asphaltene-enhanced b i t u m e n ;

Downloaded by UNIV OF MISSOURI COLUMBIA on November 27, 2013 | http://pubs.acs.org Publication Date: January 1, 1982 | doi: 10.1021/ba-1981-0195.ch013

however, appreciable quantities of coke were f o r m e d at the reaction c o n d i tions used a n d good material balances on this r u n were not achieved. W i t h o u t essentially complete reduction of coke f o r m a t i o n b y hydropyrolysis, the significance of results for the asphaltene-enhanced

b i t u m e n are

suspect.

R e m o v a l of carbon i n the f o r m of coke w i l l have an u n k n o w n effect on results that m a y not be attributable to asphaltenes.

These results are i n c l u d e d

p r i n c i p a l l y as negative results to show the d r a m a t i c effects that can result if asphaltenes are not f u l l y dispersed a n d coke f o r m a t i o n is not i n h i b i t e d d u r i n g hydropyrolysis. In an attempt to q u a n t i f y the effect of asphaltenes, a calculation that synthesized results of a hypothetical feed of 100% asphaltenes was m a d e based on the maltene a n d v i r g i n b i t u m e n results. T h i s calculation is useful to break out the relative effect of asphaltenes a n d maltenes to the total results. T h e calculation is based on the general relationship: 0.794 (maltene results)

+

0.206 (asphaltene results) = v i r g i n b i t u m e n results. Results of this calculation show that asphaltenes have a m a r k e d effect o n increasing the gas p r o d u c t i o n , and correspondingly, require appreciable quantities of h y d r o g e n for converTable III. Hydropyrolysis Yields

Maltenes" Hydrocarbon gases Recovered hydrocarbon liquids Calculated nonhydrocarbon products NH HS H0 Coke 3

2

2

c

AsphalteneAsphaltene Enhanced Virgin (Hypothetical) Bitumen Bitumen

ab

0

11

20.1

27.3

NA

54.6

80.3

73.9

42

48.8

0.17 0.13 0.46