Application of the pyrolysis-carbon isotope method for determining the maturity of kerogen in the Bakken shale

Advances in Organic Geochemistry 1985 Org. Geochem. Vol. IO, PP. I 113-l I 17. 1986 Printed in Great Britain. All rights reserved

Copyright

0146-6380/86 $3.00 + 0.00 0 1986 Pergamon Journals Ltd

Application of the pyrolysis-carbonisotope method for determining the maturity of kerogen in the Bakken shale M. E. Department

CONKRIGHT,

W. M.

and K. E.

SACKETT

PETERS*

of Marine Science, University of South Florida, St Petersburg, FL 33701, U.S.A. (Received

19 September

1985; accepted

16 April

1986)

Abstract-A maturity indexing procedure based on the isotopic difference between the total accumulated methane produced by exhaustive pyrolysis and the kerogen (A’)C) and the mole ratio of methane to kerogen carbon (CMR), has been tested by applying a standardized technique, i.e. exhaustive pyrolysis (600°C for 120 hr) of extracted-powdered samples and measurement of the amounts and isotopic composition of the methane and kerogen carbon, on a suite of 15 Bakken shale samples. A linear relation was found between the carbon mole ratio of pyrolysis-derived methane and total organic carbon and the 15“C difference between the pyrolysis-derived methane and total organic carbon (r = -0.79); and between the amount of CH, generatedfrom exhaustivepyrolysisand H/C atomic ratios (r = +0.91). Key words: pyrolysis-carbon isotope method, maturity methane--carbon isotopic composition

based on carbon isotopes, Bakken shale,

The use of this procedure has been supported by data published by Chung and Sackett (1979) who The pyrolysis-carbon isotope method (PCM) for showed that results for a series of 19 coals and 12 determining the maturity of shale and coal samples shales, ranks and maturities estimated by conwas first proposed by Chung and Sackett (1979) and ventional techniques, correlated with A”C and CMR. subsequently refined by Sackett (1984). Maturity The work reported in this paper was undertaken to indices are derived from the isotopic difference be- provide further verification of the pyrolysis-carbon tween the total accumulated methane produced by isotope maturation procedure as outlined above. exhaustive pyrolysis and the kerogen (Ai3C) and the The PCM was applied to a suite of Bakken shale difference in mole ratio of methane (total produced samples provided by the Chevron Oil Field Research by exhaustive pyrolysis) to kerogen carbon (CMR). Company. These samples were of various maturities The use of the PCM is based on the premise that as inferred from atomic H/C values, and data generfollowing post-depositional microbial alteration of ated from Rock-Eva1 pyrolysis such as T,,,,, (temorganics (Fig. l), there is a natural thermally-induced perature of maximum hydrocarbon generation), PI. relative loss of 12C-richmethane which depletes the or production index (S/S, + S, where S, is the methane precursor carbon pool in “C so that during amount of hydrocarbons already present in the rock the subsequent unnatural pyrolysis procedure the and S, the hydrocarbons generated from the pyrolysis methane produced is reduced in amount and 12C of the kerogen), and H.I. (hydrogen index which is content; the degree depending on the extent to which the ratio of S, to organic carbon) (these data are the natural process had proceeded before pyrolysis. presented in Table 1). T,,, and the production index This procedure is attractive for the following reasons. both increase with increasing thermal maturity; H/C 1. It depends on well established physical (i.e. atomic ratios and the hydrogen index decrease with preferential cleavage of 12C-“C bonds over ‘3C-‘2C depth as reactive hydrogen becomes less available. bonds) and chemical (mass balance) laws and objecThe Bakken shale is an organic rich black shale tive chemical and isotopic measurements (Sackett, found as a thin elastic unit in the subsurface of the 1984). Williston Basin which covers most of North Dakota 2. It does not rely on the presence of paly- in the United States (Williams, 1974). It’is of Missisnomorphs, vitrinite or petrographic measurements. sippian age and identified as the source rock of the 3. It can be performed on whole rock samples and Mississippian oils found in this basin (Williams, 1974; does not require kerogen isolation with its potential Zumberge, 1983). It has a wide range of thermal for loss or change in some of the carbon matrix. maturities and contains primarily amorphous kerogen of aquatic origin (Leenheer, 1983). Data for *Chevron Oil Field Research Company, P.O. Box 446, La these samples are presented in Tables 1 and 2 and INTRODUCITON

Habra, CA 90631, U.S.A.

locations

1113

of samples in Fig. 2.

1114

M. E. CONKRIGHT

et al.

-1c

-5

.\” ”

v

a

0

5

I 005

1 01

CH,-CIKEROGEN-C

I

015

(MOLE

RATIO)

Fig. 1. Representation of the shift of methane precursor carbon (MPC) composition with natural maturation and artificial pyrolysis. A”C is the per mil difference between generated 6°C methane and 6°C kerogen, and CMR is the mole ratio of generated methane to parent kerogen (Sackett, 1984). METHODS

The whole rock samples analyzed were ground and extracted with methylene chloride at the Chevron laboratory. After drying, samples were sealed under vacuum in Pyrex tubes while heating to about 150°C and pyrolyzed for 12Ohr at a temperature of 600°C. Methane, the only hydrocarbon present after this time and at this temperature, was separated by fractional distillation at liquid nitrogen temperature and combusted to carbon dioxide. The amount of CO, produced was measured manometrically and its carbon isotopic composition was then measured. A second fraction, referred to as the liquid nitrogen condensable Table

1. Data for Bakken Sample No.

03420-39 36650-l 36653-l 36661-I 38421-4 38422-a 36654-l 38422-21 38422-23 36651-I 36660-I 36659-l 36658-l 36652-l 36662-l 38423-l 5 36649-I ‘Hydrogen *Production

shales from North

Extracted Rock wt (is) 29.0 39.0 1.1 2.4 43.4 31.5 13.0 27.8 35.3 12.5 15.8 5.2 1.5 11.0 2.5 29.0 10.0

Depth (ft) 1545 1562-8 1511 ba95-7 9899 I0002 10006-9 10054 10064 10253-a 10332-50 10638 10796-9 I 1000-26 I I 139-43 I1259 11262-S

Dakota

fraction (LN,CF), was collected as a by-product of the methane separation. This fraction consists of gases which are condensable at liquid nitrogen temperatures. In these exhaustive pyrolysis experiments, the only gas present in the fraction was CO,. b”C of the parent carbon was measured by combusting an aliquot of the whole rock samples using the standardized procedure described by Craig (1953). The total organic carbon percentage (wt%) is based on the amount of CO, generated from the combustion of the shale samples. The data shows our TOC values higher than those reported by Chevron (see Tables 1 and 2). The TOC values used in this study were those determined in our laboratory. (provided

Atomic H/C 0.90 I.2 I.15 0.90 0.83 I .09 1.10 I .03 I .03 0.82 0.16

0.8 0.13 0.81 0.75 0.68 0.10

Index from Rock-Eva1 pyrolysis (mg HC/g TOC). Index. S,/(S, + S,), from Rock-Eva1 pyrolysis.

by Chevron TOC Iwt%l 9.93 9.36 17.85 12.08 10.05 12.12 15.39 19.24 Il.19 12.52 12.25 14.02 13.66 14.23 12.75 II.03 9.55

Oil Field Research

439 426 431 452 457 439 435 432 439 454 453 452 448 445 454 455 450

Company)

HI”

PI*

340 513 543 142 I50

0.12 0.03 0.06 0.15 0.24 0.08 0.11 0.09 0.09 0.26 0.20 0.23 0.22 0.16 0.27 0.42 0.38

810

333 150 120 149 117 95 104 178 88 110 96

Pyrolysis carbon isotope method

1115

Table 2. Experimental data for the change in volume and isotopic composition of methane formed for the pyrolysis, at 600°C (for 120 hr), of North Dakota Bakken shale samples of varying maturities (ranked from hiah to low H/C ratios) Sample No. 36650-I 36653-l 36661-l 38422-8 38422-21 38420-39 38421-4 36651-l 36652-l 36659-l 36660-l 36662-l 36658-l 36649-I 38423-15

TOC W) 1.7 18.8 12.1 12.0 19.6 18.8 II.1 II.0 12.1 12.2 9.1 10.8 13.6 8.7 9.4

“d “C vs the Chicago “H/C ratios provided

6 “C PM’ VW -26.1 -29.0

-28.4 -28.2 -28.8 -28.6 -28.4 -28.4 -26.7 -28.4 -28.2 -28.6 -27.8 -28.4 -26.9

6 “C” CH. w -27.6 -31.6 -30.6 -30.7 -30.1 -30.5 -28.3 -26.6 -25.3 -27.0 -28.2 -21.7 -28.2 -27.0 -24.7

PDB standard. by Chevron Oil Field Research

%d~LNlCF - I.5 -2.6 -2.2 -2.5 -1.3 -1.9 +0.1 + 1.8 +1.4 + 1.4 0.0 +0.9 +0.6 f 1.4 -l-2.2 Company-determined

6°C values were determined using a triple collector Varian Finnigan Mat 250 Isotope Ratio Mass Spectrometer relative to a charcoal working standard (6 ‘“C = - 24.8%0 vs PDB) and reported relative to the Chicago PDB standard. Time series analysis was performed on the least and most mature samples, 36650-I and 38423-15 respectively, with initial H/C ratios of 1.20 and 0.68. These were pyrolyzed for periods of 12-210hr and the same procedure followed as above. RESULTS

AND

DISCUSSION

The premise of the model presented

in Fig. 1

“is that immediately after post-depositional microbial alteration of organics there is a thermally degradable but not renewable pool of kerogenic methane precursor carbon (MPC), such as alkyl side

2.7 7.1 4.1 4.2 14.5 8.8 6.4 2.2 13.4 3.4 1.3 1.8 4.6 2.9 on extracted

CWC,,, (mole ratio) CMR

H/C”

0.222 0.183 0.203 0.213 0.180 0.181 0.107 0.089 0.132 0.152 0.122 0.073 0.080 0.099 0.110

I.20 I.15 I.10 1.09 I .03 0.90 0.83 0.82 0.8 I 0.80 0.76 0.75 0.73 0.70 0.68

shale samples.

chains and linking groups that becomes smaller as methane is produced. The first methane is formed from the most labile and weakly bonded carbon and exhibits the isotopically lightest composition (because the “G”C bond is weaker than the “C -% bond) and, concomitantly, the carbon-carbon bonding in the pool becomes stronger with time, resulting in increasingly smaller isotope fractionations during the methane generation process. Also the pool becomes richer in r3C because of material balance considerations which in itself results in more “C in subsequently formed methane. Referring to Fig. 1, this natural process is indicated by the MPCititi,, pool shifting to the MPC,&,, position. Our laboratory pyrolysis procedure then forces the MPC,, to shift to its final post-pyrolysis position where there is no more methane produced. Thus, the more methane produced relative to the MPC pool and the larger the fractionation that we obtain by pyrolysis, the closer

Saskatchewan

I

Manitoba

1665

3665”;i 30423-15

:I

Fig. 2. Sample locations.

/

1116

M. E. CONKWHT +3.0

+2.0 I +1.0-

0

o.o-

0 2 -4

-l.O-

-2.o-

-‘.OL CH~

0.1

-CIKEROGEN-c

0.2

(MOLE

RATIO)

(CMR)

Fig. 3. Correlation between the Carbon Mole Ratio (CMR) and A”C (6 Y&-6 t3Cprrnlearbon)for the Bakken shale.

is the MPC,,,, to the MPCi,i,i,, pool and the more immature is the kerogen. Conversely, the smaller the amount of methane and isotope fractionation, the closer is the pool to its “final” mature state” (Sackett, 1984).

et al.

PCM. PCM results are compared to the initial H/C as these ratios correlate well with other maturation data generated by Rock-Eva1 pyrolysis (Table 1). These data are T,,,,,,, hydrogen index and production index with correlation coefficients to the H/C ratios of -0.909, +0.833 and -0.896 respectively. The TOC values are those for the samples after extraction with methylene chloride and represent the organic carbon that was actually pyrolyzed. 6 ‘)C values for this TOC ranged from - 26.1 to - 29.OK vs PDB. These light values and high TOC suggest a highly restricted-anoxic depositional environment with a dynamic cycling of CO, between the atmosphere and epeiric sea, fixation by marine algae and bio-oxidation resulting in a 12Cenrichment in dissolved CO2 and a concomitant production of organic matter with 613C values of about -28X. H/C ratios were determined in the Chevron laboratories on the kerogen isolated from the methylene chloride extracted ground whole rock samples. Figures 3, 4 and 5 show plots of CMR vs A’)C, CMR vs H/C and A”C vs H/C, respectively. The correlation coefficients are -0.79, +0.91 and +0.87 respectively. The best correlation is between CMR and H/C and demonstrates the strong linkage between the amounts of methane produced and the hydrogen in the kerogen. Time series experiments were conducted on 36650-I and 38423-15, the samples with the highest and lowest H/C ratios; i.e. the most immature and mature kerogens in the suite. The two sets of data, shown in Fig. 6, are considerably displaced from one another with no overlap in either parameter for the pyrolysis times that were used. For the time seriesexperiment on sample 36650-1, the amount of methane is 2.1-31.7 (a = 12.8) times that of CO, as measured by the LN,CF, and +7.8 to + 16.9% (jz = + 12.4%) lighter. Isotopic exchange equilibrium between CO, and CH, at 600°C would

The model assumesa single pool of MPC, whereas in reality, there are at least two major pools (as a working simplification): aliphatic and aromatic. To demonstrate this dichotomy, the exhaustive pyrolysis procedure was used on two representative pure compounds; n-octadecane (n-C,sH98) representing the aliphatic carbon pool and decacyclene (&H,,,) representing the aromatic carbon pool. Pyrolysis of n-octadecane at 600°C for 48 hr yielded the theoretical methane CMR of 0.53 (based on the hydrogen content). The methane had a A”C of -2.5%. Pyrolysis of decacycleneat 600°C for 48 hr yielded a CMR of only 0.051 (41% of the theoretical amount of methane) with a A13Cof -2.1%. After 96 hr the . 1.20values were 0.052 and - 1.6% respectively, indicating y=3.217x+O.438 that methane generation had reached a limiting value l.lOr=+o.s07 and that some isotopic exchange with the residual carbon may have taken place (A13C shifted by l.OO+0.5%0). These experiments suggest that although the aliphatic-type carbon pool is quantitatively converted to methane (controlled by hydrogen content), the aromatic-type carbon pool is more resistant to thermal degradation and is probably more important, as kerogen is made up in large part of aromatic-type carbon (Tissot and Welte, 1978). Table 2 gives the data for North Dakota Bakken shale samples of depths from 7500 to 11,500 ft in a 0.1 0.2 variety of wells (Fig. 2). These samples are particuCARBON MOLE RATIO larly significant becausethey contain insufficient vitriCH4-C/PARENT C nite and palynomorphs for measurement of vitrinite reflectances or thermal alteration indices, but never- Fig. 4. Correlation between Carbon Mole Ratio (CMR) and H/C ratios. theless can yield maturity information based on

Pyrolysiscarbon isotopemethod

1117

0.70-

-3.0

-2.0

-1.0

0.0

+1.0

+2.0

+3.0,

A13C /

l

(%o)

Fig. 5. Correlation betweenthe A”C

and H/C ratios.

I 0.200

0.100

CH4-WKEROGEN-C

(MOLE

I 0.300

RATIO)

(CMR)

give CO2 compositions about + 11% relative to the CH,, close to that observed in the experiments described here. As the methane generally predominates, its isotopic composition upon formation should be controlling. However, the system is even more complex because of a third component, the carbonaceous residue which also may he exchanging with both the CO1 and the CH4. Mass balance calculations allow us to estimate the A”C of the residue (Table 3) assuming CH,, COz and the organic residue are the only sources of carbon. The calculated values for the residue, for the time series experiment on sample 36650-1, show that the residue becomes lighter with time as the LN$F becomes lighter and the methane heavier. The contradiction between these results and field observations (organic residue becomes heavier) must he resolved by future experiments. Isotopic equilibrium exchange does not negate the use of this method as a maturation indicator. A”C may prove to be less useful than CMR for maturity studies, but there is still a significant correlation between initial H/C ratios and A”C.

The relatively high correlation between CMR, A”C and H/C provide further support for the use of the pyrolysis-carbon isotope method in estimating the maturity of kerogens. The method is particularly attractive because it can he used on whole rock Table 3. Theoretical Sample No. 2.2 2.3 2.4 2.5 2.6 2.7 “(moles CH,)(A”r&.,,)

AT (hr) 24 48 72 96 120 210

A”C

of the organic

Moles CH, 0.185 0.186 0.193 0.214 0.276 0.429

Fig. 6. Changein compositionwith time of Bakkensamples 36650-l and 38423-15with H/C ratios of 1.20 and 0.68 respectively.Numbers by points indicatepyrolysistime in hours. samples and on kerogens which do not contain vitrinite or palynomorphs. researchwas funded by grants from SOHIO and Chevron Oil and ResearchCompany.

Acknowledgements-This

REFERENCES

Chung H. M. and SackettW. M. (1979) Use of stable carbon isotope compositions of pyrolytically derived methaneas maturity indicesfor carbonaceousmaterials. Geochim. Cosmochim. Acta 43, 1979-1988. Craig H. (1953) The geochemistryof the stable carbon isotopes.Geochim. Cosmochim. Acta 3, 53-92. LeenheerM. J. (1984)MississippianBakkenand equivalent formations as source rocks in the Western Canadian Basin.In Advances in Organic Geochemistry 1983 (Edited by SchenckP. A., De Leeuw J. W. and Lijmbach G. W. M.). Org. Geochem. 6, 521-532. PergamonPress, Oxford. Sackett W. M. (1984)Determinationof kerogenmaturity by the pyrolysis-carbonisotope method. In Aduancesin Organic Chemistry 1983 (Edited by SchenckP. A., De Leeuw J. W. and Lijmbach G. W. M.). Org. Geochem. 6, 359-363. PergamonPress,Oxford. Tissot B. P. and Welte D. H. (1978) Petroleum Formarion and Occurrence, 538 pp. Springer,New York. Williams J. A. (1974) Characterizationof oil types in Williston Basin. Am. Assoc. Pet. Geol. Bull. 58, 1243-1252. ZumbergeJ. E. (1983)Tricyclicditerpane distributionsin the correlationof Paleozoiccrude oils from the Williston Basin.In Advances in Organic Geochemistry 1981 (Edited by Bjorey M. et al.), pp. 738-745. Wiley, Chichester.

residue from the time series experiment for sample 36650-I A”C” Idue* A”C LNlCF Moles residue moles LN,CF 0.686 -2.19 0.129 +18.4 0.135 + 17.4 0.679 -2.53 0.672 - 1.98 0.135 +I%6 0.158 + 16.7 0.628 -3.14 0.161 + 19.5 0.563 -4.60 0.181 +l6.1 0.390 -10.44

A”CH 1 -4.1 -3.4 -4.0 -3.1 -2.0 +2.7

+ (moles LNICF)(A”CLNICF)

+ (moles residue)(A’&,,)

= (moles parent C)(A”C,,,c).