Ž . Chemical Geol ogy 149 1998 235–250 Model ling isotope fractionat ion during prima ry cracki ng ofnatural gas: a reaction kinetic approach Bernhard Cramer ),1 , Bernhard M. Krooss, Ralf Littke 2 ( ) Insti tute of Petroleu m and Organic Geoch emist ry ICG-4 ; Resear ch Centre Julic h, 52425 Julich, Germany ¨ ¨ Received 15 April 1997; accepted 3 April 1998 Abstract Ž . A numeri ca l mod el has been devel ope d to comput e sta ble carbon iso top e var iat ion s in nat ura l gas met hane by calculating 13 CH and 12 CH gene ratio n as a set of paralle l first-or der reac tions of prima ry cracki ng. The goal of this work4 4 was to combine the description of isotope fractionation with established kinetic models for gas generation. Stable carbon isotope ratios of methane from sedimentary organic matter are characterized by the initial carbon isotope ratio of methane prec ursor s withi n the organic matter and by a cons tant difference in activatio n energy between 12 C- and 13 C-methane generation from corresponding precursor sites. Methane generation is calculated separately for 12 C- and 13 C-methane. A difference in activation energy automatically implies a temperature dependence of fractionation processes which has not bee n tak en int o consideration in pre vio us wor ks. This new model off ers a the ore tic al exp lanati on and ma themat ica l des cri pti on of the obs erv ed var iab ili ty ofd13 C-values of meth ane durin g open-sys tem pyrol ysis expe rimen ts. Carbo n isotopes of methane within natural gas of thermogenic origin can be simulated for any geological temperature history. The application of the method to two coaly rock samples of the Pokur formation from northern West Siberia results in simulated carbon isotope values of methane which are very similar to those in the natural gas within the reservoirs of the Pokur Ž 13 . formation dC sy42‰ to y54‰ . Th is fi nding supports a ther moge ni c or igin of the ga s at an ea rl y st age of 1 maturation. q 1998 Elsevier Science B.V. All rights reserved. Keywords: Methane; Natural gas; Isotope fractionation; Reaction kinetics; West Siberia; Pokur formation 1. Introduction Stable isotopes of nat ura l gas component s have been used to identify source rocks and to recognize ) Correspondi ng author. 1 Current address: Federal Institute for Geosciences and Natural Ž . Resources BGR , Stilleweg 2, 306 55 Hannover, Ge rmany. E-mail : [email protected]2 Curr ent addr ess: Insti tute of Geol ogy and Geoc hemi stry ofPetroleum and Coal, Aachen University of Technology, Lochner- str. 4-20, 52056 Aachen, Germany. Ž . possib le secondary alterations of the gas. Stahl 1968 Ž . and Sta hl and Ca rey 1975 esta bli shed f irs t empir i- cal relationships between the maturity of source rocks and the stable carbon isotope composition of related gaseous hydrocarbons. Since these early milestones a variety of such empirical relationships has been de- veloped serving as important tools to solve applied Ž geolo gical prob lems Faber, 198 7; Berner and Fab er, . 1988; Shen et al ., 1988; Berner , 1989 . However, the se models have seve re limitations, bec aus e the y are not based on a fundamental unde rst andi ng of 0009-2541r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. Ž . PIIS0009-2541 98 00042-4
16
Embed
Modeling Isotope Fraction During Primary Cracking of Natural Gas
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
8/8/2019 Modeling Isotope Fraction During Primary Cracking of Natural Gas
Modelling isotope fractionation during primary cracking of natural gas: a reaction kinetic approach
Bernhard Cramer ),1, Bernhard M. Krooss, Ralf Littke 2
( ) Institute of Petroleum and Organic Geochemistry ICG-4 ; Research Centre Julich, 52425 Julich, Germany¨ ¨
Received 15 April 1997; accepted 3 April 1998
Abstract
Ž .A numerical model has been developed to compute stable carbon isotope variations in natural gas methane by
calculating13
CH and12
CH generation as a set of parallel first-order reactions of primary cracking. The goal of this work 4 4
was to combine the description of isotope fractionation with established kinetic models for gas generation. Stable carbon
isotope ratios of methane from sedimentary organic matter are characterized by the initial carbon isotope ratio of methane
precursors within the organic matter and by a constant difference in activation energy between12
C- and13
C-methane
generation from corresponding precursor sites. Methane generation is calculated separately for12
C- and13
C-methane. A
difference in activation energy automatically implies a temperature dependence of fractionation processes which has not
been taken into consideration in previous works. This new model offers a theoretical explanation and mathematical
description of the observed variability of d 13
C-values of methane during open-system pyrolysis experiments. Carbon
isotopes of methane within natural gas of thermogenic origin can be simulated for any geological temperature history. The
application of the method to two coaly rock samples of the Pokur formation from northern West Siberia results in simulatedcarbon isotope values of methane which are very similar to those in the natural gas within the reservoirs of the Pokur
Ž 13 .formation d C sy42‰ to y54‰ . This finding supports a thermogenic origin of the gas at an early stage of 1
maturation. q 1998 Elsevier Science B.V. All rights reserved.
13 Ž .Initiald C ‰ y24.1 y23.3OM13 Ž .Final d C ‰ y23.8 y23.1OM
was deposited from Upper Aptian to Cenomanian
times and predominantly consists of sandstones with
coaly organic matter. The two rock samples repre-sent the Cenomanian section of the Pokur formation
Ž .with one sample from the top of the formation Pk1Ž .and the other Pk2 from a section 200 m deeper.
General organic geochemical characteristics of the
organic matter are summarized in Table 1. Accord-Ž . Žing to the concentration of organic carbon TOC 41
.and 63 wt.% the rock samples can be described as
coaly sandstones or sandy coals. Vitrinite reflectance
and T -values clearly indicate that the organic mat-max
Ž .ter of these rocks is immature Table 1 . Hydrogen
Ž .indices HI of 134 and 207 mg HCrgTOC aretypical of humic organic matter in coals at low levels
Ž .of maturation Tissot and Welte, 1984 .
2.2. Pyrolysis experiments
A 200 mg rock sample was pyrolysed using the
open-system pyrolysis apparatus previously de-Ž .scribed by Krooss et al. 1995 . Open-system pyroly-
sis allows the continuous registration of hydrocarbon
formation rates as a function of temperature that
constitute the most common data base for kinetic
modelling of non-isothermal simulation experimentsŽ .Schenk et al., 1997 . Although even under open-sys-
tem conditions a certain overlap of primary gas
generation with secondary cracking of high molecu-
lar weight primary petroleum compounds cannot be
avoided, this method provides probably the closest
approach to primary cracking reactions.
Special tests were performed to control the ovenŽtemperature during the experiments Schaefer et al.,
.1998, in press . Samples were heated from 208C up
to 8008C each at heating rates of 0.1, 0.3 and 2.0 K
miny1. About 40 ml miny1 helium were flushed
through the reactor as carrier gas. Gaseous hydrocar-Ž .bons were separated by gas chromatography GC
Ž .and quantified by a flame ionisation detector FID .
These measurements were carried out every 3 min
during each pyrolysis experiment. More details about
the analytical procedure are found in Schaefer et al.Ž .1998, in press . The evaluation of the experimental
data was performed according to the scheme de-Ž .scribed by Schaefer et al. 1990 and yielded an
activation energy distribution and one common pre-
exponential factor for the entire set of parallel reac-
tions.
For isotope analyses, gas samples were taken
off-line from experiments performed at a heating rateof 2.0 K miny1. During time intervals when no
GC-measurements were made gaseous pyrolysis
products together with helium were flushed throughŽ .a glass sampling container 250 ml equipped with
two stopcocks. After 20 min the flask was closed and
the gas was expanded into an evacuated 70 ml
stainless steel container. For every sample a new
glass flask was installed. For carbon isotope analysis
the gas was injected from the container into a GC
interfaced to an isotope ratio mass spectrometer.
Measurements were carried out using the method of continuous flow.
3. Experimental results
3.1. Methane generation
Characteristics of methane generation during py-
rolysis experiments from the two rock samples were
very similar. As a result, the activation energy distri-butions and pre-exponential factors differ onlyŽ .slightly Fig. 1 . The total methane generation poten-
tial of approximately 29 mgrgTOC determined for
the two samples corresponds to 22% of the total
hydrocarbon generation potential measured by
Rock–Eval pyrolysis for sample Pk1 and to 14% forŽ .sample Pk2 Table 1 . Methane generation is de-
8/8/2019 Modeling Isotope Fraction During Primary Cracking of Natural Gas
( ) B. Cramer et al.r Chemical Geology 149 1998 235–250238
Fig. 1. Reaction kinetic parameters for methane generation from rocks of the Pokur formation evaluated by the gross kinetic model of Ž .Schaefer et al. 1990 .
scribed by 39 parallel reactions for sample Pk1 andŽ .42 reactions for Pk2 Fig. 1 . The highest methane
generation potentials were calculated for activation
energies of 243 kJ moly1
for Pk1 and 247 kJ moly1
for Pk2 at pre-exponential factors of 9.06P1013 sy1
Ž . 13 y1 Ž .Pk1 and 3.20P10 s Pk2 . Both activation
energy distributions show a nearly symmetrical ap-y1 y1 Žpearance from 167 kJ mol to 352 kJ mol Fig.
.1 .
Methane generation during both pyrolysis experi-
ments started at 3008C, reached a maximum genera-Ž .tion at 4908C, and ceased at 8008C Fig. 2 . Mod-
elled methane generation for a geological heating
rate of 2 K Myry1 results in small differences
between both samples. As shown in Fig. 2 modelled
methane generation starts at about 708C. Generally,
the generation curve of sample Pk1 is shifted by 5 K
to higher temperatures in comparison to Pk2. Maxi-
mum methane generation rate from Pk1 occurs at
1878C and from Pk2 at 1818C. Under these condi-
tions, the potential for methane generation of bothŽ .rocks is exhausted at 4008C Fig. 2 .
3.2. Stable carbon isotope Õalues
The stable carbon isotope composition of methane
during pyrolysis experiments is summarized in Table2. Because of the fast volume exchange in the glass
flask with a rate of about 20%rmin the measured
d 13
C-values mainly represent the highest temperatureŽof the respective sampled temperature interval Table
.2 . The comparatively low reproducibility of the13 Žd C-measurements with 1s ranges between 0.5
.and 1.25‰, Table 2 is regarded as a result of low
methane concentrations in the carrier gas collected
during the pyrolysis experiments.
Measured carbon isotope ratios of methane from
both rock samples display similar trends. These mea-
sured d 13
C-values of methane as a function of tem-
perature and extent of methane generation are com-Ž .pared with those of Berner et al. 1995 for methane
pyrolytically generated from xylite and kukersite in
Fig. 3. Interestingly, neither their results nor our
results show a steady increase in carbon isotopeŽ .values over the entire temperature range Fig. 3A .
8/8/2019 Modeling Isotope Fraction During Primary Cracking of Natural Gas
( ) B. Cramer et al.r Chemical Geology 149 1998 235–250 245
Each of the n methane generating reactions was
attributed a generation potential with the same initial
carbon isotope ratio. This carbon isotope ratioŽ .y33.2‰ was derived from the total methane quan-
tity and the isotope ratios of methane measured in
the experiments. The isotope fractionation model
was matched to the measured carbon isotope values
of methane by varying the difference in activation
energy between12
C- and13
C-methane generation
until the square of the difference between model and
measured values had reached its minimum. The ac-
tual computations were performed in a spreadsheet
using a discretized time-temperature history com-
posed of sufficiently small isothermal time steps. To
relate the stable carbon isotope ratio of methane to
the maturation of organic matter an EASY%Ro ele-Ž .ment Sweeney and Burnham, 1990 was integrated
into the program.
5. Results of the modelling of isotope fractionation
The carbon isotope ratios of methane calculated
with the new model are shown in Fig. 5. In principle
all trends of carbon isotope ratios of methane from
pyrolysis experiments, increasing and decreasing
d 13
C-values, are reproduced by the model. Especially
the temperature interval of main methane generation
Ž .between 400 and 6008C Fig. 2 shows good concur-rence of measured and predicted value. Generally,
only the measured minimum and maximum valuesŽ .are not matched Fig. 5 . The modelled trends shown
in Fig. 5 were obtained with almost identical D E -a
values of y103 J moly1 for Pk1 and y102 J moly1
for Pk2. The D E y values are in the same range asa
the differences in zero-point energies between C–C
bonds in isotopically substituted and non-substitutedŽorganic molecules D E between y80 and y250 Jzp
y1 .mol modelled with ab initio quantum chemistryŽ .calculations by Tang and Jenden 1995 . Once again
it should be emphasized that the isotope trends from
pyrolysis experiments were fitted with one common
D E -value for all parallel reactions. The D E con-a a
cept was introduced to account for the differences in
bond strength between12
C–12
C and12
C–13
C bonds
of precursor structures. Although one could argue
that the D E value may not be the same for alla
reactions, a definition and computation of individual
D E values for each reaction is not justified with thea
present database. Such an approach would corre-
spond to a gross over-interpretation of the existing
database.
The modelled D E -values for methane generationa
from rocks of the Pokur formation correspond to
KIE’s between 1.013 and 1.023 for temperatures of Ž .laboratory pyrolysis Fig. 4 and 1.020 and 1.038 for
the temperature range of methane generation at geo-Ž .logical heating rates Fig. 4 .
No realistic results were achieved when applying
the measured carbon isotope ratios of total organicŽ . 13
carbon Table 1 as initial d C-values of methane
precursors. Generally, the slope of the modelledŽ .carbon isotope trend of methane Fig. 5 is deter-
mined by the difference in activation energy. A
parallel shift of the general trend to isotopically
lighter or heavier methane can be realized by chang-ing the initial carbon isotope ratios of the methane
Ž .precursors as shown by Clayton 1991 . The devia-
tions of isotope trends during methane generation
from monotonously increasing isotope ratios must be
seen as a pure result of the dynamics of methane
generation. At the beginning of methane generation
only some reactions of the whole set are involved in
generation processes. At these hypothetical reaction
sites initial carbon isotope fractionation leads to min-
imum carbon isotope ratios of methane. During our
pyrolysis experiments this stage was reached at tem-Ž .peratures of about 3008C Fig. 5 where less than 1%Ž .of total methane was generated Fig. 3 . With in-
creasing temperature the number of participating re-
actions increases. The measured average isotope
composition of methane during this main stage of
generation can be described as a mixture of gener-
ated methane from a maximum number of active
reaction sites. A more or less steady increase of d 13
C
of methane must be expected and actually is ob-
served. In our pyrolysis experiments this stage wasŽ .reached between 400 and 6008C Fig. 5 . At 5008C
the steady isotope trend of both experiments showsŽ .an inflexion Fig. 5 , indicating that the number of
reactions contributing to isotope fractionation de-
creases and methane generation rates have reachedŽ .maximum values Fig. 2 . The maximum values of
carbon isotope ratios at 380 and 6208C also coincide
fairly well with the inflexion of methane generation
8/8/2019 Modeling Isotope Fraction During Primary Cracking of Natural Gas
( ) B. Cramer et al.r Chemical Geology 149 1998 235–250248
Fig. 6. Carbon isotope composition of methane in natural gas from Geologicheskaya field in the Pokur formation as a function of sourceŽ .organic matter maturity compared with modelled trends and empirical relationships. Empirical relationships for type II A and type III
Ž . Ž . Ž . Ž .kerogen C taken from Faber 1987 for type III kerogen B from Shen et al. 1988 .
data. The influence of a release of methane from
deep groundwater on reservoir filling in northernŽ .West Siberia is discussed elsewhere Cramer, 1997 .
7. Conclusion
A new concept is presented to describe stable
carbon isotope fractionation during generation of
natural gas. The model simulates kinetic isotope
effects as differences in activation energy between
reactions involving12
C- and13
C-methane genera-
tion. This approach is based on established and
widely used concepts of hydrocarbon generation ki-
netics and inherently implies a temperature depen-
dence of the kinetic isotope effect. Methane genera-
tion and isotope fractionation during open-system
pyrolysis experiments are calculated from a set of
isotope fractionation as a function of temperature in the CH –4
C H –C H –C H system. Geokhimiya 8, 674–681.2 6 3 8 4 10
Galimov, E.M., Lopatin, N.V., Shabayeva I.Y., The origin of gas
accumulations in Western Siberia. In: Durrance, E.M., Gal-
imov, E.M., Hinkle, M.E., Reimer, G.M., Sugisaki, R., Au-Ž .gustithis, S.S. Eds. , Geochemistry of Gaseous Elements and
Compounds, Theophrastus Publ., Athens, 1990, pp. 401–417.
Grace, J.D., Hart, G.F., Urengoy gas field—USSR West Siberian
Basin, Tyumen District. In: Beaumont, E.A, Foster, N.H.Ž .Eds. , AAPG Treatise of Petroleum Geology, Atlas of Oil and
Gas Fields, Structural Traps: III, Tectonic Fold and Fault
Traps, 1990, A-109, pp. 309–335.¨Hanbaba, P., Juntgen, H., Zur Ubertragbarkeit von Laboratoriums¨—Untersuchungen auf geochemische Prozesse der Gasbildung
aus Steinkohle und uber den Einfluß von Sauerstoff auf die¨Ž .Gasbildung. In: Schenk, P.A., Havenaar, I. Eds. , Adv. Org.