Fuel Vol. 77, No. 112.pp.

23-3 I, 199X
0 1997 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
PII: SOO16-2361(97)00165-S 0016.2361/98 $19.00+0.00
ELSEVIER

Kinetics of petroleum
generation and cracking by
programmed-temperature
closed-system pyrolysis of
Toarcian Shales

V. Dieckmann”, H. J. Schenk, B. Horsfield and D. H. Welte

Forschungszentrum Juelich GmbH, ICG-4, Leo-Brandt-StraQe, D-52428 Juelich, Germany
(Received 14 April 1997; revised 23 June 1997)

Primary kerogen-to-petroleum and secondary oil-to-gas conversion processes in marine source rocks have been
studied contemporaneously by programmed-temperature closed-system (MSSV) pyrolysis of Toarcian Shale
concentrates at heating rates of 0.1, 0.7 and 5.OK min -’ in the temperature range of 300-610°C. All pyrolysates
were analysed by single-step on-line gas chromatography. The cumulative evolution profiles of liquid and gaseous
compounds were deconvoluted into generation curves for oil (C,,), primary gas and secondary gas using
complementary open-system experiments and simple stoichiometric relationships. The subsequent kinetic
analysis resulted in potential versus activation energy distributions which turned out to be comparatively broad for
oil and primary gas and rather narrow for secondary gas, indicating that the forrnerer are generated from more
inhomogeneous precursor materials than the latter. The dominant activation energies were found to increase from
52 (217.9) (oil) to 53 (222) (primary gas) and 55 (230.5) kcal mol-’ (kJ mol-‘) (secondary gas); the best-fit
frequency factors were calculated around 1015 min-’ . By extrapolation to a geological heating rate of 5.3K my-’
(IO-“K min-‘) the onset of oil generation is predicted to occur at 9O”C, the maximum oil formation rate at 140°C
and the onset (peak generation) of primary and secondary gas at 110°C (165°C) and 150°C (18O”C), respectively.
0 1997 Elsevier Science Ltd.

(Keywords: non-isothermal closed-system pyrolysis; compositional kinetics; in-source oil-to-gas cracking; primary
versus secondary gas)

INTRODUCTION Predicting the composition of natural petroleums gener-

ated in source rocks requires kinetic investigations of both
The transformations of sedimentary organic matter upon
the primary generation of individual components from
burial leading from the high-molecular weight kerogen to
kerogen and the secondary transformations among the
oil and gas proceed through a multitude of parallel and
primary products. As a first step towards compositional
consecutive reactions which are unknown in detail, but
modelling, Espitalik et ~1.~ used programmed-temperature
which are recognized to be quasi-irreversible and controlled
open-system pyrolysis to derive kinetic parameters for the
by chemical kinetics’. Kinetic laws are therefore routinely
primary generation of four petroleum fractions from type II
considered as the mathematical link between high tempera-
and III kerogens from the North Sea. Since then, rather
ture-short time and low temperature-long time configura-
complete kinetic schemes of organic matter transformation
tions, allowing to extrapolate from laboratory to natural
including primary kerogen degradation and secondary oil-
heating conditions. However, the extremely complex
to-gas cracking and accompanying reactions have been
structure of the organic components involved* necessitates
elaborated for lacustrine and marine source rocks’ and for
the application of gross kinetic concepts whereby individual
marine and terrestrial source rocks”. These take secondary
product precursors are replaced by bulk potentials which are
as well as primary gas forming reactions into account.
fractions of the total product yield3. Despite these
Secondary reactions are necessarily simulated by heating
simplifications current kinetic models for bulk petroleum
experiments in closed-system configurations under either
generation properly reflect the principal structural features
isothermal’ “‘* or non-isothermal conditions 13-15. These
of different parent kerogens4*5 and as far as marine source
studies generally disconnect secondary from primary
rocks are concerned, they deliver timing predictions which
reactions, because they are mainly concerned with the
are independent of the degree of confinement of the
experimental device6 and which appear to be reliable7. thermal alteration of pooled petroleum and in particular
with in-reservoir oil-to-gas cracking”-13,‘5.‘6. However,
the importance of in-source oil-to-gas conversion has
* Corresponding author. recently been addressed by Pepper and Dodd17.

Fuel 1998 Volume 77 Number l/2 23

Kinetics of petroleum generation and cracking: V. Dieckmann et al.

The present study is concerned with the composition of 3-4 mg kerogen concentrate were sealed in glass capillary
bitumen generated in marine source rocks with emphasis tubes. These tubes were divided into three batches. Each
being placed on a straightforward application of laboratory- batch was placed circulary in a massive cylindrical brass
derived compositional findings to real geological systems. block which serves as sample holder in the furnace for the
In view of this goal bitumen composition is defined in terms non-isothermal heating of the tubes at 0.1, 0.7 and
of only three components which are liquid (C,,) products, 5.OK min-‘. During each heating programme, the sample
primary gas (from kerogen) and secondary gas (from oil). tubes were removed from the furnace at specific tempera-
Their thermal evolution is simulated by programmed- tures spaced by about 5°C. These temperatures were
temperature closed-system (MSSV) pyrolysis of an imma- measured with a thermocouple located in the brass block
ture Toarcian Shale and calculated for geological heating where the temperature errors during the heating pro-
conditions by means of activation energy distributions and grammes were smaller than t O.l”C. Schenk and Horsfield6,
frequency factors individually optimized for each fraction. whose work focused more on primary formation reactions
A sample of the Toarcian Shale was chosen for this study from kerogen, chose final heating temperatures varying
because it represents a typical algal rich kerogen type II between 490 and 530°C. To obtain an accurate database
source rock generating PNA oi118.In addition, bulk kinetic involving enough samples for the primary formation as well
parameters on the basis of open-system pyrolysis of this as for the secondary cracking temperatures, we increased
sample were published”. With the same sample Schenk and the final temperatures in this studie up to 540-610°C. The
Horsfield6 showed that the kinetic analysis of bulk composition of oil and gas formed in each tube was
petroleum formation under programmed-temperature determined by a single-step on-line gas-chromatographic
closed (MSSV) and open system results in similar kinetic analysis*‘. With this technique it is possible to quantify
paramters. These phenomenon is a very important pre- separately the generation profiles of single compounds,
requisite for applicability of kinetic calculation on the basis compound groups or bulk petroleum in milligrams per gram
of MSSV-pyrolysis presented here, in showing that the of kerogen concentrate as a function of temperature27. The
evaluation of these parameters and the resulting temperature role of pressure which is reported to retard the maturation of
predictions for geological heating rates are independent of kerogen28’29and affect the stability of hydrocarbons11,30*31
the analytical configuration. The programmed temperature has recently been comprehensivly discussed3*. It was
closed-system (MSSV) pyrolysis used in the present study proven that the absence of water during maturation processes
ensures a very homogeneous and precise heating of the results in a minor influence of pressure during thermal
sample material during each heating programme because of cracking of organic materials than during generation
the small amounts of kerogen concentrates (3-5 mg) heated processes in hydrous environments.
up to specific temperatures. In addition, liquid and gaseous
hydrocarbons generated at each selected temperatures can
be measured in a single GC run which make it possible to KINETIC MODEL
reach a high data frequency in a comparative short With the MSSV-technique only cumulative amounts of
analytical time*‘. gross hydrocarbons, compound groups or single compounds
can be determined after discrete temperature intervals. After
MATERIALS AND METHODS smoothing by appropriate spline functions, a numerical
differentiation of the evolution profiles of each compound
Sample origin group of interest was carried out. These differentiated
The Toarcian-Shale sample was obtained from corehole curves represent the database for further kinetic calculations
Wenzen which is located in the Hils syncline at the southern according to the mathematical model which has been
rim of the Lower Saxonian Basin (north-west Germany)*‘. described by Schaefer et all9 It is based on the kinetic
Here the organic matter is thought to be thermally analysis of such formation rate (dMld7’) versus temperature
influenced by deep-seated intrusive bodies (Vlotho Massif, (7) curves assuming usually 25 first-order parallel reactions
Uchte Massif, Bramsche Massif) located north-west of the with activation energies Ei regulary spaced between 46
Hils syncline**. The total organic carbon (TOC) content is (192.7) and 70 (293.3) kcal mol-’ (kJ mol-I) and a single
measured as 12.7% with a HI of 660 mg g-’ TOC for this pre-exponential factor A. The model furthermore implies
immature sample (R, = 0.48%), making this a typical the measurement of dMldT versus T profiles at three heating
marine oil prone source rock with a structurally complex rates of r = 0.1,0.7 and 5.OK min-‘. These heating rates are
and compositionally heterogeneous kerogen type II which sufficiently low to ensure correct temperature measure-
was shown to be converted up to 90% into volatile products ments33’34and they are sufficiently different to obtain a
during artificial maturation experiments23. Hydrocarbons reasonable starting value of A from the shift of peak
generated from this algal-rich Toarcian shale during generation temperatures35. A total number of 26 parameters,
maturation processes correspond to PNA-oils18,23,24. namely 25 potentials (partial yields) associated with 25
Because small sample sizes are required for MSSV- activation energies and the pre-exponential factor A are
pyrolysis, a kerogen concentrate was prepared according to optimized by a least-squares iteration method that compares
Durand and Nicaise25. More detailed geochemical work had measured and calculated formation rates at 600 (200 per
been done and published by Rullkotter et al.**, Witte et a1.26 heating rate) temperatures until the corresponding error
and Diippenbecker and Horsfield23. function (sum of squared differences) presents a well-
defined absolute minimum.
Experimental approach
A very useful method for simulating the maturation of RESULTS AND DISCUSSION
organic material by programmed-temperature closed-
system p rolysis is the microscale-sealed-vessel-technique Evolution of liquid and gaseous hydrocarbons in MSSV
(MSSV) yo. The experimental procedure has been described pyrolysates
in detail by Schenk and Horsfield6. Ninety aliquots of Figure r(a) illustrates the cumulative evolution of Cl-s,

24 Fuel 1998 Volume 77 Number I/2

 Kinetics of petroleum generation and cracking: V. Dieckmann et al.

c6-14 and CE+ boiling ranges during programmed- The generation of secondary gas becomes obvious from
temperature closed-system (MSSV) pyrolysis of Toarcian the decrease of C6+ yields [Figure I(b)]. The generation of
Shale kerogen concentrates at heating rates of 0.1, 0.7 and secondary gas at lower temperatures cannot be fully
5.OK rnin- . In accordance with the theory of non- excluded, but multistep-open versus closed Py-GC experi-
isothermal kinetics all evolution profiles are shifted towards ments of the Toarcian kerogen with identical experimental
higher temperatures with increasing rate of heating and configurations provides some evidence that the generation
maximum cumulative yields appear to be independent of the of secondary gas sets in when the formation of primary
heating rate. The apexes of the cumulative C-14 and C,s+ liquids has come to an end. This phenomenon is illustrated
evolution curves mark the temperatures where degradation in Figure 2 where the degradation of the C6+ compounds in
processess overcompensate primary formation. In contrast, the closed-system commences when the liquid hydro-
total gas yields remain constant even at comparatively high carbons generated under open-system condition reach the
temperatures because the observed decrease of pentane, cumulative plateau.The early secondary cracking processes
butane and propane is compensated by an enhanced which are observed between the temperatures of maximum
formation of ethane and especially methane. C is+ and C6+ yields [Figure l(a) and (b)], respectively, take
The thermal degradation of liquid hydrocarbons in a place within the C6+ fraction and do not result in the
closed system does not directly result in the formation of formation of secondary gas. Since there seems to be no
gaseous compounds. The broad spectrum of different liquid significant overlap between primary generation and second-
hydrocarbons with different molecular structures which are ary cracking of Cb+ compounds, the amount of secondary
generated from the kerogen show different cracking gas can be calculated from the decrease of C6+ yields
stabilities and yield different cracking products. Conse- [Figure 1 (b)] assuming that for reasons of hydrogen balance
quently the transition from primary formed liquid hydro- only 70% of cracked C6+ compounds are converted into gas
carbons to secondary gas is marked by a progressive and 30% into pyrobitumen (Coke) according to eqn (1):
decrease of the average molecular weight of detectable
M sec.gas(T)= {Me+max cTk) -M,+,,,(~)}~O.7
hydrocatbons36 and the onset of secondary cracking
reactions in the pyrolysate does not immediately lead to In this equation, M6+max( Tk) represents the maximum of the
the formation of significant amounts of secondary gas. C6+ evolution curve at temperature Tk. M6+res. (7’) is the

I3.1KJmln I
120
100
80
80
40
20 0

20
0
.
0
0.7KJmln (
a 120
z 120
100
E L& 100
80
3
9 60
6 ;;
F ;; B 40

0 F 20
5.OK/mln / 0
120
120
100
100
80
60 80

40 60

20 40

0 20
300 340 380 420 460 500 540 560 620
Temperature(‘C) 300 340 380 420 460 500 540 580
Temperature(“C)

Figure1 (a) Cumulative evolution of Ct_s, C6-14 and C 15+boiling ranges (milligrams per gram of kerogen concentrate) during MSSV
pyrolysis of Posidonia Shale concentrates with increasing temperature (“C) at heating rates of 0.1, 0.7 and 5.OK min-‘. (b) Cumulative
evolution of total gaseous (C 1-s) and liquid (C,,) hydrocarbons (milligrams per gram of kerogen concentrate) during MSSV pyrolysis of
Posidonia Shale concentrates with increasing temperature (“C) at heating rates of 0. I, 0.7 and 5.OK min-’ . Part (a) corresponds to Figure
2 of
Schenk and Horsfield6 with additional data points, especially at higher temperatures

Fuel 1998 Volume 77 Number l/2 25

Kinetics of petroleum generation and cracking: V. Dieckmann et al.

residual amount of liquid hydrocarbons while Msec.gas

corre- Figure 4 illustrates the cumulative yields of primary and
sponds to the resulting amount of secondary gas. The aver- secondary gas calculated from eqns (1) and (5) and the
age conversion factor of 0.7 is assessed from cumulative evolution of liquid hydrocarbons (Ce+) up to
the onset temperatures of oil-to-gas conversion reactions
CH, = pCH, + (1 - x)CH, (2) around 410°C (O.lKmin-‘); 440°C (0.7Kmin-‘) and
where: n = average I-I/C-ratio of Cg+ compounds; m = 470°C (5.OK min-‘).
average H/C-ratio of secondary gas; u = average WC-
ratio of pyrobitumen, which yields Kinetic modelling
In order to use the MSSV-dataset for non-isothermal
x=(n-u)l(m-u) (3) kinetic modelling spline functions were made to fit each of
and a conversion factor of the cumulative evolution profiles given in Figure 4. These
spline functions were differentiated to produce formation
f,=*(12+m)l(12+n) (4) rate versus temperature curves (dotted lines in Figure 5)
Assuming n = 2.2 (average of hexane and triacontane), m = which constituted the input data for kinetic modelling. The
3.2 (average of methane and pentane), u = 0.2”, the con- solid lines in Figure 5 were calculated as best fit from the
version factor is assessed to be fc = 0.7. According to frequency factors and potential versus activation energy
Figure 3 the atomic WC ratio of total (and probably sec- distribution of Figure 6 and TubEe 1. In addition to liquid oil,
ondary) gas in MSSV-pyrolysates increases from 3.2 to 3.7 primary and secondary gas, Table I also contains the kinetic
at the final stage of conversion. However, this increase must parameters for the generation of ‘total’ gas; the latter were
necessarily be compensated (at least partly) by a decrease in determined independently in order to check the reliability of
hydrogen content of the pyrobitumen. Both effects should the kinetic model (see Figure 8). The quality of all kinetic
be balanced and, according to eqns (3) and (4), keep the fits can be judged from Figure 7 which depicts the increase
conversion factor at roughly 0.7. of the error functions with increasing departure of the
The CI_5 fractions in Figure I(a) represent the total gas, frequency factors from the respective best-fit values.
composed of primary gas directly formed from the Figure 6 shows that the hydrocarbon potential versus
macromolecular kerogen and secondary gas sourced from activation energy distribution of the different hydrocarbon
oil-degradation processes. From the calculated yields of classes are shifted to higher energies in going from liquid oil
secondary gas (Msec,& [eqn (l)], the amount of primary gas to primary and secondary gas, whereas the variation of the
(Mprim,gas)can easily be calculated: frequency factors are insignificant. The liquid hydrocarbon
(C,,) potentials are ranged from 48 (201.1) to 54 (226.1)
kcal mol-’ (kJ mol-‘) with a main potential at 52 (217.9)
kcal mol-’ (kJ mol-‘) and a frequency factor of 2.01 X
1OL5min-‘. Similar kinetic parameters for oil generation
-180 ??? ? ? ? ? ?
from marine source rocks are reported by Pepper and
% Corvi37. The main activation energy for primary gas
b 150 ??

“120 ..ooo
00 0
generation is calculated to be 53 (222.1) kcal mol-’
(kJ mol-‘). This is only insignificantly higher than the
e
- 90 0 main energy needed for Cg+ formation while both frequency
0
0 80 Q factors turned out to be nearly identical (see Table I and
4 30
0 .
0 p+WC&~.[ Figure 6). These results are close to the main activation
energies of 51 (213.7)-53 (222.1) kcal mol-’ (kJ mol-‘)
i 0 and the frequency factor of 1.8 X lOI min-’ calculated for
300 350 400 450 500 550 oil and gas formation from marine-elastic source rocks by a
Temperature [“C) model of Braun and Burnham38. In contrast, Pepper and
Corvi37, on the basis of different pyrolysis methods, report
Figure 2 Cumulative evolution of liquid hydrocarbons during much higher kinetic parameters of A = 1.2 X 102’min-’
multistep open- versus closed-system pyrolysis with an identical and E = 67 (280.7) kcal mol-’ (kJ mol-‘) for gas genera-
experimental configuration at a heating rate of 5K min-’ tion. However, the authors pointed out that their gas

3.70 A

a 3.50 4 ??
0 3.30
() 3.10 4 *~**Q+@+*

? 2.90
2.70
2.50-, ““““” t
300 340 380 420 460 500 540 580

Temperature(OC)

Figure 3 Atomic H/C-ratioof total gas (C ,J as a function of temperature(“C)during M&W-pyrolysisof PosidoniaShale concentratesat
a heating rate of 0.7K min-’

26 Fuel 1998 Volume 77 Number l/2

 Kinetics of petroleum generation and cracking: V. Dieckmann et al,

2.0

1.6

1.2
0.8

0.4

1 .o

0.8

0.8

0.4

0.2

0
-

0 a40380 420400600540680

Temperature (‘C)
Figure 4 Cumulative formation of liquid hydrocabons (C,,),
primary gas (C ‘_5prim,)and secondary gas (C 1_5,,.) (milligrams per Figure 5 Formation rates (milligrams per gram of kerogen
gram of kerogen concentrate) as a function of temperature (“C) at concentrate and per degree) of liquid hydrocarbons (C,,), primary
heating rates of 0.1,0.7 and 5.OK min -I. The yields of primary and gas (C ,-s rrim.)and secondary gas (C ,-s ,,,.) during MSSV pyrolysis
secondary gas were calculated on the basis of Figure I(b) and of Posidonia Shale concentrates as a function of temperature (“C)
eqns (1) and (3). Lines represent spline functions smoothing the at heating rates of 0.1, 0.7 and 5.OK mini’. Dotted curves result
experimental datapoints for further kinetic calculations from the differentiation of the splines of Figure 4; full lines were
calulated from the kintic parameters of Figure 6 as a best fit

generation kinetics are based on a poor laboratory data-set. (234.6) kcal mol-’ (kJ mol-‘) with a frequency factor
The potential versus activation energy distribution for the of 6 X lOI min-’ from isothermal oil-to-gas cracking
formation of secondary gas turns out to be extremely reactions in a closed-system and in the presence of the
narrow, indicating a very homogeneous composition of the kerogen residue. These experiments were believed to
corresponding precursor material. Obviously secondary gas simulate oil-to-gas conversion in a source rock environ-
is mainly generated from long chain alkanes which can be ment. The catalysing effect of kerogen may be due to
anticipated to be kinetically homogeneous in so far as they aromatization and condensation within the macromolecular
depict no significant variation of C-C bond energies along framework which release hydrogen for initiation of cracking
the chains39. reactions4’-4”.
Most kinetic investigations of oil-to-gas cracking reac-
tions are based on the thermal degradation of conventional Extrapolation to geological heating conditions
crude oils in the absence of kerogen. Kinetic parameters On the basis of the laboratory-derived kinetic parameters
calculated from these experiments are more or less higher generation curves can be calculated for any geological
than ours which probably reflects the distinctly catalysing heating condition. This extrapolation is illustrated in Figure
role of the kerogen residue. For example, main activation 8 which compares hydrocarbon formation rates for the
energies between 55 (230.4)-60 (251.4) kcal mol-’ slowest artificial heating rate of 0.1 K min-’ and an average
(kJ myi;‘) are reported for isothermal oil-to-gas conver- geological heating rate of lo-“K min-’ (5.3K my-‘). The
sion ’ . Even higher main activation energies of up to 66 corresponding integral evolution profiles are shown in
(276.4) kcal mol- (kJ mol-‘) for the thermal conversion of Figure 9. Bulk petroleum formation from the same source
marine oils into gas have been found by different rock has been predicted to take place in the temperature
authors”.‘3,‘5. More similar to our results, Quigley et ~1.‘~ interval of 90-200°C with T,,, at 150°C for a heating rate
deduced main activation energies of 55 (230.4)-56 of 5.3K my-‘. The dominant compound group in the early

Fuel 1998 Volume 77 Number l/2 27

Kinetics of petroleum generation and cracking: V. Dieckmann et al.

0.50 I 1 stages of petroleum formation are the liquid hydrocarbons.

0.45 -- C6+ (“Oil”) In correspondence to one-dimensional basin modelling46 in
0.40 -- the Hils syncline, the onset of oil generation from Toarcian
0.35
0.30 -- A=2.01 E+15 Shales is extrapolated to occur at 90°C which is similar to
0.25 -- the onset of bulk petroleum formation6. While the peak oil
0.20 -- generation is predicted at 14O”C, the formation of most
of the liquid hydrocarbons is completed at about 150°C.
This temperature interval for the main phase of oil
generation is commonly reported in literature. However,
48 48 50 52 54 58 58 80 82 84 88 88 70 it has to be taken into consideration that different geological
E 0.50 I I heating rates were assumed for extrapolation leading
0 0.45 to somewhat hi her or lower temperature predictions,
0 0.40 respectively8’9.37’F 8,47,48.
a 0.35
5 0.30 The onset of gas generation at 110°C corresponds
3 0.25 exclusively to primary gas whose T,, is predicted at
* 0.20 165°C; at temperatures about 190°C gas formation directly
B 00.;; from kerogen appears to be completed. From 150°C up to
~6 0605
210°C secondary gas resulting from thermal degradation of
liquid oil is predicted to become the dominant gas fraction.
8 48 48 50 52 54 58 58 80 82 84 88 88 70 According to basin modelling results46 these onset tem-
h perature for secondary gas formation were not reached in the
0.50 .
0.45
geological past of the Hils syncline and Gifhom trough, but can
0.40 be expected to be closer to the deep seated-intrusive bodies in
0.35 the north-westerly direction (Vlotbo Masif, Uchte Massif and
0.30 Bramsche Massif) where increasing vitrinite reflectance values
0.25 of up to 4% were measured in the Jurassic strata49’50.
0.20
0.15 Both primary and secondary gas formation rate curves are
0.10 found within the total gas formation rate curves which have
been calculated independently in order to check the self-
Ooo5 consistency of the kinetic model. Figure 10 finally depicts
48 48 50 52 54 58 58 80 82 84 88 88 70 the geological part of Figure 9 including the decrease of
liquid oil yields as a consequence of secondary oil-to-gas
Acttvatlon Energy(kcal/mol) cracking. Decreasing oil yields have been calculated from
the maximum oil yield of about 130 mg g-’ sample and the
increasing yields of secondary gas accordinqito eqn (1).
Figure 6 Fractions of total potentials versus activaton energy With respect to vitrinite reflectance (easy % Ro ) this figure
(kcal mol-‘) and best-fit frequency factor A (min-‘)for liquid
reproduces the principal features of the ‘oil window’
hydrocarbons (C,j+), primary gas (Cl_s prim,)and secondary gas
concept52 with the onset of liquid petroleum generation
(C,_, ,,,,) as derived from MSSV-pyrolysis of Posidonia Shale
concentrates predicted at about 0.6% Ro and the onset of secondary
cracking at about 1.2% Ro. In Figure 10 oil yields do not
decrease to zero because Cg+ products had not completely

Table 1 Kinetic parameters for the formation of liquid hydrocarbons (C,,), total gas (C I_5total),primary gas (C L-5prim,)and secondary gas
(C 1_5,,,,) from MSSV-pyrolysis of Posidonia Shale concentrates at three different heating rates (0.1, 0.7 and 5.OK min-‘)
HC classes Cl-5 G+ C 1-5prim. C I-5 LCC.
Frequency factor (I min-‘) 9.58 x lOI 2.01 x lOI 2.03 X lOI 6.48 X lOI
Total potential (mg g-‘) 111.429 131.715 51.795 52.038

E, [kcal mol-’ (kJ mol-‘)I Fraction of total potential

46 (192.7) 0.00 0.02 0.00 0.00

47 (196.9) 0.00 0.00 0.01 0.00
48 (201.1) 0.01 0.04 0.00 0.00
49 (205.3) 0.01 0.11 0.00 0.00

50 (209.5) 0.00 0.09 0.11 0.00

51 (213.7) 0.10 0.23 0.00 0.00
52 (217.9) 0.01 0.26 0.00 0.00

53 (222.1) 0.00 0.22 0.41 0.00

54 (226.2) 0.39 0.03 0.24 0.00
55 (230.4) 0.19 0.00 0.11 0.77
56 (234.6) 0.11 0.00 0.11 0.19
57 (238.8) 0.18 0.00 0.00 0.00
58 (243.0) 0.00 0.00 0.01 0.00
59 (247.2) 0.00 0.00 0.00 0.04
60 (25 1.4) 0.00 0.00 0.00 0.00

28 Fuel 1998 Volume 77 Number l/2

 Kinetics of petroleum generation and cracking: V. Dieckmann et al.

C6+ Cl -5 Mm.
gm-
8oA
70..

60..

so .'

40 ..
ii
I5 30-.

f 20..
10 ..

0 2 4 6 8, 1\12 14 16 16 20 2 4 6 8 1012141618
og

Cl -5 total

0
0 2 4 6 6 10 12 14 16 18 20 2 4 6 8 10 12 14 16 18 20
logA

Figure 7 Increase of the error functions (S) with increasing departure of the frequency factors, log A, from the best-fit values

Natume j Laboratory m0-m

3.2
e 2.8
*b 2.4
P

- 2.0
d 1.6
1.2
0.8

Figure 8 Formation rate (milligrams per gram of kerogen concentrate and per degree) versus temperature (“C) of liquid hydrocarbons
(G,), total gas CC1-5t,t.), primary gas CC,- 5prim,)
and secondary gas (C 1_5,,,.) for the slowest laboratory heating rate of 0. I K min-’ and an
average geological heating rate of 5.3K my-’

disappeared at the feasible end temperatures of the MSSV the timing and compositional aspects are considered reliable
simulation experiments [Figure l(b)]. in view of fractionations assosiated with expulsion54.
The predicted temperature ranges for secondary oil-to-
gas cracking are by 30°C lower than those derived from
isothermal’ and non-isothermal”.15 pyrolysis of produced
CONCLUSIONS
oils. This is obviously a fundamental difference between
in-source and in-reservoir secondary cracking which may be (1) Programmed-temperature closed-system (MSSV)
due, among others, to the catalytic activities of kerogen pyrolysis at various rates of heating is a suitable approach
residues in source rocks53. Furthermore, it must be for simulating simultaneously primary kerogen-to-
emphasized that the quantitative predictions of Figures 8- petroleum and secondary oil-to-gas transformation
10, being based on MSSV pyrolysis, are only valid for processes in source rocks.
hermetically closed source rocks. More sophistical models (2) The comparison of open and closed-system pyrolysis
should take into account rates of primary migration of oil provides strong evidence that in the case of marine
and gas out of the source rock to predict masses and source rocks the overlap between primary oil generating
volumes of oil and gas generated. However, in view of the and secondary oil cracking reactions is negligible, the
poor quantitative knowledge of such transport processes, latter operating essentially after the former have come
this is very unlikely to be feasible up to now. Nevertheless, to an end.

Fuel 1998 Volume 77 Number l/2 29

Kinetics of petroleum generation and cracking: V. Dieckmann et al.

a00 loo 200 300 400 500 600

Temperature (“C)
Figure 9 Formation (mg g-’ kerogen concentrate) versus temperature (“C) of liquid hydrocarbons (C,,), total gas (C 1-5&, primary gas
(C,_5 rrim,)and secondary gas (Cl_5 ,,.) for the slowest laboratory heating rate of O.lK min-’ and an average geological heating rate of
5.3K my-’

SCHEMATIC Oil-WINDOW window with the onset of oil generation predicted to

(Posidonia Shale) occur at about 0.6% Ro and the onset of secondary
oil-to-gas cracking at about 1.2% Ro for a geological
cumulatlva yield (mglg sample)
heating rate of 5.3K my-r.
Easy % Ro 50 100 150
%Ro rc

REFERENCES
0.35 - fdl
Y-l 1 Tissot, B., Rev. Inst. Fr. P&t., 1969, 24,470.
2 Rullkotter, J. and Michaelis, W., Advances in Organic Geo-
0.65 - 100 chemistry 1989 Part II: Molecular Geochemistry, 1990,
829.
3 Tissot, B. and Espitalie, J., Rev. Inst. Fr. Pit., 1975,30,743.
1.18 - 150 4 Ungerer, P. and Pelet, R., Nature, 1987, 327, 52.
5 Schenk H. J., Horsfield B., Kross B., Schaefer R. G. and
Schwochau K., Petroleum and Basin Evolution 1997, ed. D.
2.32 - 200 H. Welte, B. Horsfield and D. Baker. Springer, Heidelberg,
1997, p. 231.
Schenk, H. J. and Horsfield, B., Geochimica Cosmochimica
3.78 - 250 Acta, 1993, 57, 623-630.
Schenk, H. J. and Horsfield, B., Zbl. Geol. Paliiont. 1995,
H 1 l/12.
4.63 - 300 Espitalie, J., Ungerer, P. M., Irvin, H. and Marquis, F.,
Advances in organic geochemistry 1987. Organic
Geochemistry, 1988,13,839.
350 9 Bumham, A. K. and Braun, R. L., Advances in organic
geochemistry 1989. Organic Geochemistry, 1990,16,27.
Figure 10 Cumulative evolution of liquid and gaseous hydro- 10 Bihar, F., Kressmann, S. and Rudkiewicz, J. L., Geochem-
carbons including formation and degradation (mg g-’ kerogen istry 1991. Organic Geochemistry, 1992, 19, 173.
concentrate) as a function of temperature (“C) and vitrinite 11 Ungerer, P., Bihar, F., Villalba, M., Heum, 0. R. and
reflectance (easy % Ro5’) at a geological heating rate of Audibert, A., Advances in organic geochemistry 1987.
5.3K my-’ Organic Geochemistry, 1988, 13, 857.
12 Bihar, F., Ungerer, P., Kressmann, S. and Rudkiewicz, J. L.,
Rev. Inst. Fr. P&t., 1991, 46, 151.
(3) The cumulative MSSV evolution profiles of liquid and
gaseous components can be deconvoluted into genera- 13 Horsfield, B., Schenk, H. J., Mills, N. and Welte, D. H.,
Advances in organic geochemistry 1991. Organic
tion curves for oil (C,,), primary gas and secondary gas
Geochemistry, 1992,19, 191.
using simple stoichiometric relationships.
14 Schenk H. J. and Horsfield B., Poster Sessions from the 17th
(4) Kinetic frequency factors and potential versus activation International Meeting on Organic Geochemistry, San
energy distributions can be calculated individually for Sebastian, Spain, ed. J. 0. Grimalt and C. Dorronsoro,
these three compound classes. 1995, p. 88.
(5) By extrapolation to geological heating conditions the I5 Schenk, H. J., Di Primio, R. and Horsfield, B., Organic
model reproduces the principal features of the oil Geochemistry 1997 (in press).

30 Fuel 1998 Volume 77 Number l/2

 Kinetics of petroleum generation and cracking: V. Dieckmann et al.

16 Quigley T. M., Mackenzie A. S. and Gray J. R., Migration of 36 Phillips, C. D., Haidar, N. I. and Poon, Y. C., Fuel, 1985,64,
Hydrocarbons in Sedimentary Basisns, ed. B. Doligez, 678.
1987, p. 649. 37 Pepper, A. S. and Corvi, P. J., Marine and Petroleum
17 Pepper, A. S. and Dodd, T. A., Marine and Petroleum Geology, 1995, 12, 291.
Geology, 1995, 12, 321. 38 Braun, R. L. and Bumham. A. K., Organic Geochemistry,
18 Horsfield, B., Review of Palaeobotany and Palynology, 1992. 19, 161.
1990, 65, 357. 39 Claxton M. J., Patience R. L. and Park P. J. D., Poster
19 Schaefer, R. G.. Schenk, H. J., Hardelauf, H. and Harms, R., Sessions from the 16th International Meeting on Organic
Advances in organic geochemistry 1989. Organic Geochemistry, Stavanger, Norway, ed. K. Bygard. Falch
Geochemistry, 1990, 16, 115. Hurtikrykk. Oslo, 1993, p. 198.
20 Horsfield, B., Disko, U. and Leistner, F.. Geologische 40 Kuo. L. -C. and Michael, G. E., Organic Geochemistry,
Rundschau, 1989, 78, 361. 1994, 21, 911.
21 Mann, U., Facies, 1987, 17, 181. 41 Brooks, J. D. and Smith, J. W.. Geochimica Cosmochimica
22 Rullkotter, J., Leythaeuser, D., Horsfield, B., Littke, R.. Acta. 1969. 41, 1223.
Mann, U., Miiller, P. J., Radke, M., Schaefer, R. G., 42 Evans, E. J., Batts, B. D. and Smith, J. W.. APEA Journal,
Schenk, H. -J., Schwochau, K., Witte, E. G. and Welte, D. 1984,62, 1584.
H., Advances in organic geochemistry 1987. Organic 43 Herschkowitz, F., Olmstead, W. N., Rhodes, R. P. and Rose,
Geochemistry, 1988, 13, 847. K. D., American Chemistry Society Symposium, 1983. 15,
23 Diippenbecker, S. and Horsfield, B., Advances in organic 301.
geochemistry 1989. Organic Geochemistry, 1990,16, 259. 44 Smith, J. W., Gilbert, T. D. and Batts. B. D., APEA Journal,
24 Latter, S. R. and Senftle, J. T., Nature, 1985, 318, 277. 1987, 27, 98.
25 Durand B. and Nicaise G., Kerogen. Insoluble Organic 45 Serio, M. A.. Hamblen, D. G., Markam, J. R. and Solomon,
Matter from Sedimentan: Rocks, ed. B. Durand. Technip, P. R., Energy and Fuels, 1987, 1, 138.
Paris, 1980, p. 35. 46 Dueppenbecker S. J. and Welte D. H., Proceedings of the
26 Witte. E. G., Schenk, H. J., Miiller, P. J. and Schwochau, K., 13th World Petroleum Congress, 1991, vol. 3, p. 1.
Advances in organic geochemistry 1987. Organic 47 Quigley, T. M. and Mackenzie, A. S., Nature, 1988, 333,
Geochemistry, 1988, 13, 1039. 549.
27 Horsfield, B. and Dueppenbecker, S. J., Journal of Analyti- 48 Reynholds. J. G. and Burnham. A. K., Organic
cal and Applied Pyrolysis, 199 1, 20, 107. Geochemistry, 1995,23, 11.
28 Goffe, B. and Villey, M., Bulletin Minerale, 1984, 107, 8 1. 49 Bartenstein, H., Teichmiiller, M. and Teichmuller, R.,
29 Chiaramonte, M. A. and Novelli, L., Advances in organic Fortschr. Geol. Rheinl. u. Westf, 1971, 18, 501.
geochemistry 1985. Organic Geochemistry, 1986,10, 28 1. 50 Koch, J. and Amemann, H.. Geology Journal, 1975. A29,
30 Fabuss, B. M., Smith, J. 0. and Satterfield, C. N., Advances 45.
in Petroleum and Chemical References, 1964, 3, 157. 51 Sweeney, J. J. and Bumham, A. K., AAPG, 1990,10, 1559.
31 Domine, F., Organic Geochemistry, 1991, 17, 619. 52 Tissot, B. P. and Welte D. H., Petroleum Formation and
32 Michels, R., Landais. P.. Torkelson, B. E. and Philp. R. T., Occurence, 2nd edn. Springer, Berlin, 1984.
Geochimica Cosmochimica Acta, 1995.59, 1589. 53 Horsfield, B. and Schenk, H. J., Kinetics of gas generation in
33 Gregg, M. L., Campell, J. H. and Taylor, J. R., Fuel, 1981, overmature sediments: source versus reservoir. Nature
60, 179. 1997 (in press).
34 Burnham, A. K., Braun, R. L. and Samoun, A. M., Advances 54 Leythaeuser, D., Littke, R., Radke. M. and Schaefer, R. G.,
in organic geochemistry 1987. Organic Geochemist?, Advances in organic geochemistry 1988. Organic
1988, 13, 839. Geochemistry. 1988, 13, 489.
35 Van Heek, K. H. and Jiintgen, H., Ber. Bunsenges. Phys.
Chem. 3 1968 .172 1223.

Fuel 1998 Volume 77 Number l/2 31