Genetic Classification of Petroleum Systems!
Gerard Demaison? and Bradley J. Huizinga
ABSTRACT
Our genetic classification of petroleum systems is
founded on a simple working nomenclature that con-
sists of combining qualifiers from each of the following
three categories: (1) charge factor (supercharged, nor
mally charged, or undercharged), (2) migration
drainage style (vertically drained or laterally drained),
and (3) entrapment style (high impedance or low
impedance)
The charge factor is estimated on the basis of the
richness and volumetrics of mature source rocks. The
source potential index (SPD, which combines source-
rock richness and thickness into a single parameter, is a
convenient shortcut for comparing the petroleum
potential of diverse source rocks containing dissimilar
kerogen types and for rapidly estimating regional,
charging capacity. In extensively explored basins, a
positive correlation exists between the magnitude of
SPL and basin-wide petroleum reserves,
The migration drainage style is determined from the
structural and stratigraphic framework of a basin. Verti-
cal-migration drainage, which occurs mainly through
faults and fracture systems breaching a seal, is charac-
teristic of petroleum systems contained within rift
basins, deltaic sequences, salt-dome provinces, wrench
basins, and fold-and-thrust belts, In contrast, lateral-
migration drainage is dominant wherever stratigraphi-
cally continuous seal-reservoir “doublets” extend over a
Copyright 1981. The Amercan Assocation of Petroleum Geclogisl. All
rots reserved
“Manuscript received March 19, 1990; revised manuscript received
Feoruary 5, 1991; tal acceptance May 17, 199%
*Goclogy Deparment, Stanoré Unveriy, Sianfor, Calforia 84305, or
P.0..Bor 1877, Capi, Calorie 9500,
Schavion Overseas Petroleum in, 6001 Bolinger Canyon Read, San
Ramon, Cafoma 9583,
‘Ths pope’ was presented at the AAPG Research Conference on te
Petroleum Potential of Seomentary Basins, Leesburg, Virgina. Api 26,
1988, and atthe 1969 and 1991 AAPG Annual Conventions. The waters
Woh to ackrowiodge W. Dow and &. Perodon forthe nial concept oF
petroleum system, B. Tissot Yor frat introducing the idea of integrating
Ihckness and veld into a single numeral expression, G. Ulrishek for his
‘ary erica revow of our maruserpl, and. Schnesdermann whose suppor
and encouragement resuted in tie pape’ being witlen and presented. We
are especialy ngebied fo L. Magoon. whe has retined the concept ot
petcleum systom over recent years and has aso spent considerable efotn
Improving aur onginal manuseript In adalton, the euthors gratefully
‘acknowledge the consicive views of Chevron’ management, namely
MW Boyoe, EL Couch, and RE Kropachot
very large area in a tectonically stable province (e.g
commonly foreland or intracratonic platform basins),
Recognition of the dominant migration style helps to
predict the location of zones of petroleum occurrence
in relation to the “hydrocarbon kitche
The entrapment style, which is also dependent on.
the structural framework and the presence and effec-
tiveness of seals, describes the degree of resistance
Ge., impedance) working against dispersion of the
petroleum charge.
Application of these working concepts should help
to significantly reduce geologic risk, particularly in new
ventures-type exploration
INTRODUCTION
During the last 40 years, several tectonic classi
fications of sedimentary basins have been developed
by petroleum geologists to provide a common frame-
work of reference for the Earth's sedimentary basins
(e.g., Weeks, 1952; Perrodon, 1972; Klemme, 1975,
1980, 1983; Bally and Snelson, 1980; Kingston et al.,
1983, Helwig, 1985). These classifications have bk
developed with the purpose of finding relations
between tectonic style and petroleum occurrence. The
most useful outcome of these studies has been to
demonstrate the statistical predominance of giant fields,
and major petroleum reserves in cratonic sags and fore-
deep basins, in contrast to lesser reserves in pull-apart
and thrust-belt settings (Klemme, 1975, 1980, 1983;
Bois et al., 1982; Carmalt and St. John, 1986). Also.
field-size distributions in petroleum basins generally
show some relation to basin classification and basin
size (Klemme, 1983). Furthermore, recognition of tec
tonic style helps to predict regional retention efficiency
(eg., in the evaluation of thrust belts; Vann, 1989) or
abnormally high heat-flow regimes affecting source
rock maturation and oil preservation (e.g., in proto-
‘oceanic rifts such as the Red Sea or Afar Triangle),
Nevertheless, any degree of petroleum richness,
ranging from near barren to highly prolific, can be
observed among specific examples of each basin type.
Bally and Snelson (1980) recognized this point by con-
cluding that “the classification of basins does little to
improve our hydrocarbon volume forecasting ability.
This sobering opinion was echoed by Ulmishek (1986):
1626“Tectonic basin classifications are limited in their effec-
tiveness of rating the richness of basins.” Grunau,
(1987) went even further in stating: “A classification
system which is based on measurable relevant criteria,
and which is tuned to the needs of the explorationist,
has not yet been developed.” The reason the analog
approach does little to improve our hydrocarbon vol-
lume forecasting ability is that source-rock volumes and,
richness are unrelated to tectonic style. In other words,
tectonically “look-alike” basins of similar size do not
necessarily contain identical volumes of mature source
rock and, thus, do not possess analogous regional
charging potential. Furthermore, Bois et al, (1982) and
Ulmishek and Klemme (1990) have documented that
paleolatitudinal and paleoclimatic considerations are
more important than tectonic style in explaining the
world’s geographic distribution of prolific source rocks
and major petroleum reserves. For instance, these
authors show that highly favorable paleogeographic
factors led to the inordinate development of prolific
Upper Jurassic and middle to Upper Cretaceous source
rocks in the greater Tethyan realm, which is the Earth's
hest endowed region in terms of petroleum reserves
(75% of the oil and 61% of the gas). The importance of
source-rock distribution, rather than tectonie sty!
also illustrated by the following examples.
(1) In peri-Andean foreland basins, middle to Upper
Cretaceous source rocks show an exponential decrease
in petroleum potential from north to south, The tecton-
ie style, which can be classified as a foreland ramp, and
regional seals are similar; however, source rock rich:
ness and thickness, as well as discovered Cretaceous-
sourced petroleum reserves, tend to wane from north
to south,
(2) Coastal basins along the Brazilian Atlantic margin
show very unequal petroleum richness, despite exten-
sive exploration, The Campos basin alone contains 65%
of all Brazilian reserves (Bacoccoli et al., 1989; Mohriak
et al., 1990), whereas the neighboring Espirito Santo
and Santos basins hold only minor reserves, The
unequal occurrence of petroleum in Brazilian coastal
basins is dependent on variations in the thickness and
regional distribution of Lower Cretaceous lacustrine
source rocks deposited in early rift settings. Otherwise,
the tectonic style, the effectiveness of regional seals,
and source-rock maturity levels are similar in all the
basins along this South Atlantic margin,
(3) Carbonate platform systems, whether associated
with passive margins or foreland basins, also show
extreme variations in petroleum richness, ranging from
prolific (e.g., Campeche-Reforma basin in Mexico, Cen-
tral Arabian basin) to marginal (e.g., South Florida
basin, Adriatic basin). The differences in petroleum
richness between the Campeche-Reforma basin and the
South Florida basin are caused not so much by differ-
ences in tectonic styles as by inequalities in source-
rock volumes and basin geothermal parameters, In the
‘Campeche-Reforma basin, the prolific Upper Jurassic
Demaison and Huizinga 1627
source-rock sequence is stratigraphically continuous,
thick, and widespread (Gonzalez and Holguin, 1991).
This source rock was deposited in a deep-water anoxic
environment during a major marine transgression. The
geothermal gradients are average in the Campeche-
Reforma basin, In contrast, the South Florida basin con-
tains rather thin Lower Cretaceous source beds that are
erratically distributed, both vertically and laterally. They
were deposited in ponded shallow-water, hypersaline,
anoxic “basins,” which occurred inside the carbonate
platform during a period of overall low sea level. Fur-
thermore, the geothermal gradients tend to be lower in
the South Florida basin
Because source-rock volume and richness are unre-
lated to tectonic style, we fell that a genetic classic
tion scheme based on the processes of petroleum for-
mation, migration, and entrapment had to be devised:
to supplement conventional tectonic classifications.
The geologic framework of a basin provides only the
natural setting for these physico-chemical processes to
interact in a manner that can lead to either concentra
tion or dispersion of petroleum. The geologic history of
a basin is the “program” that sequentially activates
these physico-chemical processes during the passage
cof geologic time (Perrodon, 1980).
‘Three important geologic factors control the accu-
mulation of petroleum in the subsurface and, thus, are
essential 10 the existence of viable “petroleum sy
tems": (1) volumetrically adequate petroleum gener
tion, occurring during or after the time of trap forma-
tion; Q) favorable migration-drainage geometry
leading to the focused movement of hydrocarbons into
traps rather than to dispersion and loss of hydrocar-
bons in either subsurface migration “waste zones” or
up to the surface; and (3) the existence of volumetrical-
ly adequate traps, capable of retaining their petroleum,
charge from the earliest time of trap filling t0 the pre-
sent day.
Our plan is to genetically classify the main types of
petroleum systems and to qualify and quantify, when-
ever possible, the key factors that control the occur-
rence, abundance, and habitat of oil and gas.
‘THE PETROLEUM SYSTEM
A “petroleum system” is a dynamic, petroleum gen-
erating and concentrating physico-chemical system,
functioning in a geologic space and time scale. A
petroleum system requires the timely convergence of
certain geologic elements and events essential to the
formation of petroleum deposits (¢.g., mature source
rock, expulsion, secondary migration, accumulation,
and retention; Nijhuis and Baak, 1990). This definition
is quite similar to the earlier concepts of “source-reser-
voir oil system” (Dow, 1974), “Petroleum System" (Per
rodon, 1980, 1983; Perrodion and Masse, 1984; Magoon,
1987, 1988), “hydrocarbon machine” (Meissner et al.,1628
PETROLEUM
‘SYSTEM
MIGRATION.
GENERATIVE
JENTRAPMENT|
‘SUBSYSTEM easerereea
| fo
® cance — |) @ wararion U@ewrrapwenr
FACTOR DRANG SME
|-Supercharged Vertically Drained || -High Impedance
-ormaty Charged
Feaeetacea || Lateraty Drained || Low impedance
a
[A PETROLEUM SYSTEM IS CLASSIFIED
"ACCORDING TO THE QUALIFIERS
‘SELECTED FROM THESE
‘THREE CATEGORIES
Figure 1—Flow diagram for the genetic classification of
a petroleum system.
1984), and “Independent Petroliferous System”
(Ulmishek, 1986). A “petroleum basin” (or province) i
a geologic entity containing at least one or more
petroleum systems
‘The concept of the “petroleum system” is not syn-
onymous with “play,” which is defined by Bois (1975)
as “a continuous portion of sedimentary volume which
contains pools showing the following characteristics
(1) reservoirs within the same productive sequence
occur throughout the zone, (2) hydrocarbons are of
similar chemical composition, and (3) traps are of the
same type.
A petroleum system is composed of two subsystems
(Figure 1)
G) A generative subsystem provides a certain supply
of petroleum during a given time span. Generative sub-
systems are essentially controlled by chemical process-
5, consisting of biochemical transformation of dead
organisms into kerogen during the source depositional
stage (Demaison et al., 1984) and thermo-chemical
kinetics, which control the transformation of kerogen,
into petroleum (Tissot et al., 1987).
(2) A migration-entrapment subsystem gathers
petroleum from the mature source rocks and distributes
it in a manner that may lead to either concentration of
petroleum into economic accumulations or loss of
petroleum due to dispersion and destruction. Migra.
tion-entrapment subsystems are predominantly con-
Genetic Classification of Petroleum Systems
trolled by physical processes, including the buoyant
rise of petroleum in water, fluid flow and capillary
pressures in porous media (llling, 1939; Showalter,
1979), and pressure-temperature-composition relation”
ships affecting phase behavior before and during
petroleum entrapment, The understanding of petro-
Jeum migration processes has made great advances
since the mid-1970s (Durand, 1988)
‘The purpose of the proposed genetic classification of
petroleum systems is t0 describe and predict (1) the rel-
ative charging potential of petroleum systems and, thus,
segments of petroleum basins and (2) the geographic
location of zones of petroleum occurrence or plays in
basins. As a part of this genetic classification, the source
potential index (SPD is introduced as a useful tool for
approximating regional charging potential. Furthermore,
patterns of regional petroleum occurrence can be pre-
dicted by delineating generative basins (i.c., *hydrocar-
bon kitchens") and recognizing the migration style and
distances (Demaison, 1984). The final result of integrat-
ing hydrocarbon-kitchen mapping, migration style, and
SPI ranking is a more effective evaluation of regional
petroleum potential than that achieved by using the tec-
tonic analog approach alone, A prerequisite to the
genetic approach, however, is that the tectonic frame-
work, sequence stratigraphy, geologic history, thermal
history, and a reasonably adequate geochemical data-
base must be fully integrated.
‘The themes reviewed in this paper do not address
the prediction, before drilling, of the volume and com-
position of petroleum trapped in individual prospects.
Advanced computer-aided methodologies for quantita-
tive prospect evaluation have already been described
by Nederlof (1979, 1981), Sluijk and Nederlof (1984),
and Nijhuis and Baak (1990). Instead, the present
genetic classification scheme has been devised to con
duct rapid, semiquantitative, regional evaluations.
EVALUATION OF GENERATIVE SUBSYSTEMS
Regional Charging
In every petroleum system, the principal constraint
to petroleum richness is the adequacy of the generative
subsystem, which must be powerful enough to provide
sufficient petroleum charge to the migration-entrap-
‘ment subsystem. The overriding importance of the gen-
erative subsystem is obvious: if there is no petroleum
generation in the subsurface, then all of the other nec
essary requirements of a petroleum system (e.g., struc-
ture, reservoir, seal) lose relevance.
Charge has been defined as the “hydrocarbon vol-
umes available for entrapment” (Sluijk and Nederlof,
1984). Charge volume equals the volume of petroleum,
generated in the drainage area of a trap minus the vol-
ume lost through migration processes. Overall migra~
tion losses result from the summation of primary-migration losses (expulsion from the source rocks into the
carriers) and secondary-migration losses (in the carri-
ers, between the source beds and the trap).
The term “regional charge” is used to represent the
total amount of petroleum available for entrapment in a
regional hydrocarbon kitchen. (Regional charge
quantity of petroleum generated in a regional hydrocar-
bon kitchen minus expulsion and migration losses.)
The regional charge is dependent on the original
source-rock richness and volumetrics in the generative
parts of the basin, Source-rock richness is most accu-
rately expressed in terms of the hydrocarbon genetic
potential (i.e., S; + aasured by Rock-
Eval analyses on well cuttings, core, and outerop sam-
ples. Hydrocarbon kitchens are delineated on source
rocks by projecting source-rock maturity data from key
wells onto seismic depth maps and/or by employing
kinetic-modeling methods. Following these delin-
ations, the volumes of mature source rock can be esti:
‘mated with the assistance of geochemical, stratigraphic,
and seismic information. Finally, a gross approximation
of the amount of petroleum generated in the kitchen
can be obtained using source-rock genetic potentials,
(in kg HC/t rock), mature source-tock volumes, source-
rock densities, and maturity conversion factors (e.g,
modeled transformation ratios). ‘The resulting amount
of petroleum (in kilograms) can be converted to any
unit of volume required by the evaluator.
In an effort to extend beyond this point, deterministic
methods using uncalibrated volumetric calculations have
been proposed to predict the amount of petroleum pre-
sent in an undrilled prospect or sector of a basin, These
schemes are unworkable because of the great uncertain:
ty associated with some, or all, of the input parameters.
For example, expulsion efficiency and secondary-migra
tion losses are not directly observable and, thus, defy
accurate measurement, The calcukated amounts of
petroleum generated from the drainage areas are invari-
ably several orders of magnitude higher than what has
already been found or can be reasonably expected in
associated traps. To bring the calculated petroleum
quantities to plausible propostions, the evaluator has to
apply discount factors to compensate for expulsion and
secondary-migration losses. These discounts are either
assumed” by the evaluator, usually in response 10 per-
sonal biases, or are sometimes obtained from numerical
modeling. Although the former approach kicks any sci-
entific merit, we believe that the modeling of petroleum
losses also leaves much to be desired in terms of geolog
ic realism because it is oversimplified. We agree with
Schowalter (1989) that: “Migration losses and trap effi-
Giency will vary with dip angle, oil/water density, inter-
facial tension, wettability, and rock heterogeneity along
the migration path, This ever expanding level of com:
plexity in a realistic subsurface situation suggests tha
quantitative geologic modeling of this problem is com-
plex to the point of being insolvable.” For the same fun-
damental reasons, mathematical modeling of expulsion
response), 2s m
Demaison and Huizinga 1629
efficiency from large source-rock volumes of varying
thickness, sedimentologic fabric, mineral matrix, kero:
gen richness, maturity, and pressure regime is also
fraught with difficulties. Beyond the deduction that the
hydrocarbon losses in petroleum systems are large, itis
doubtful that operationally reliable charge predictions
will ever be derived from uncalibrated deterministic
models, Some calibrated methods exist for predicting the
petroleum charge volumes available 10 undrilled traps;
however, these methods require large data bases and
complex, intensive statistical treatments (Sluijs and Ned:
erlof, 1984)
To sidestep the operational difficulties of
regional charge, we have turned 1 a
cal method, which we refer to as th
index (SPD)
stat
leulating
implified statisti-
source potential
Definition and Calculation of
the Source Potential Index (SPI)
The source potential index (SPD, or “cumulative
hydrocarbon potential” (Tissot et al., 1980), is defined
as the maximum quantity of hydrocarbons (in metric
tons) that can be generated within a column of source
rock under 1m? of surface area (Figure 2), Because SPI
niSizS2)p
1000
'SP1= The maximum quentity of hydrocarbons
that can be generated within a column
‘of source rock under 1m? of
face area (In matric tons of
hydrocarbons/square meter).
SPI
‘= source rock thickness in meters.
average genetic potential in kilograms
hydrocarbons/metric ton rock.
source-rock density in metric
tons/cuble meter.
S82
Figure 2—Definition and calculation of the source po-
tential index (SPD.1630
is a measure of cumulative petroleum potential. it is
important to bear in mind that (1) this parameter does,
not distinguish between oil-gencrating vs. gas-generat
ing capacity, and (2) the total quantity of hydrocarbons
will not be fully realized unless the source-rock
sequence is completely matured during burial. The SPI,
which effectively combines source-rock thickness and
richness into a single parameter, is calculated according
to the following equation:
HFS
1000
SPI =
SPI = source potential index in metric tons hydrocar-
bons/square meter
b= source-rock thickness in meters. The source-
rock thickness excludes intervening intervals that lack
significant source potential; therefore, multiple source
beds are combined to yield a cumulative thickness at
the given locality. In practice, we have generally con-
fined each source-rock sequence to include the cumu-
lative thickness of only those fine-grained rocks show-
ing genetic potentials in excess of 2 kg HC/t rock. (in
certain cases where biomarker analyses unambiguously
correlate oils to lean source rocks, this lower limit has
been reduced to 1 kg HC/t rock.) Also, the source-rock
thickness is corrected for well deviation, the dip of the
source-rock sequence, or other structural complexities.
(3) + 5p) = average genetic potential in kilograms
hydrocarbons/metric ton of rock. The genetic potential,
which is a semiquantitative measure of source-rock
richness, is defined as the S, + S; yield from Rock-Eval
pyrolysis (Espitalie et al., 1977). The S, represents the
kilograms of hydrocarbons that are thermally distilled
from 1 t of rock and the S, represents the kilograms of
hydrocarbons that are generated by pyrolytic degrada:
tion of the kerogen in 1 t of rock (Espitalie et al., 19775,
Peters, 1986). Based on the systematic Rock-Eval pyrol-
ysis of closely spaced samples (approximately every 10
‘m) in a given well penetration or measured outcrop
section, the average genetic potential for a source rock
is determined by summing a sequence of rectangular
areas, constructed by cross-plotting sample depths vs.
the genetic potentials of individual whole-rock
ples, and dividing the resulting sum by the total uncor
rected source rock thickness,
= source-rock density in metric tons/cubic meter.
Although specific source-rock densities should be used
in the SPI determinations, we have simplified our cal
culations to date by arbitrarily assigning a density of 2.5
Um’ to all source rocks.
For source rocks that are immature across an entire
basin, the acquisition of SPIs fulfills no useful purpose
because hydrocarbon generation has never taken
place in the subsurface. SPI calculations for a given
source rock only have relevance wherever thermally
mature hydrocarbon kitchens have been identified
Genetic Classification of Petroleum Systems
The well penetrations or measured outcrop sections
selected for SPI determination should be located with-
in or immediately adjacent to the generative basin (ie.
hydrocarbon kitchen). Since SPI is a measure of
petroleum potential, the magnitude of SPI at a given
location is optimal when the source rock is immature
to early mature. A source rock that has attained a
‘middle oil-window" rank or higher maturity shows
significantly reduced SPI because the average genetic
potential is depleted due to petroleum expulsion. In
certain cases, theoretical estimates of SPI prior to mat-
uration can be calculated using the average genetic
potential from a similar organic-facies equivalent locat
ed in a less mature sector of the basin. At a given loca-
tion, the difference between the theoretical SPI and
present-day residual SPI can provide a crude estimate
of the amount of petroleum expelled from the column
of source rock under 1 m? (in metric tons hydroca
bons/square meter).
Each SPI is usually obtained from an individual
exploration well or measured outcrop section; there-
fore, the resulting value is representative of the source
Unit at a specific location in the basin of interest. Some
source rocks show nearly uniform richness and thick:
ress over long distances; therefore, an SPI measured in,
‘an immature sector is representative of the original SPI
of a more mature, lateral equivalent of the same
source-rock unit. Although many other source rocks
show significant lateral variations in thickness and
organic facies, a general knowledge of paleogeogra-
phy, source-tock depositional models (Demaison et al,
1984; Jones, 1987), and possible thickness changes
(c.g,, derived from seismic data) allow an interpreter to
determine whether the source rock in the undrilled sec-
tor of the generative basin was likely to have shown a
higher or lower SPI than those obtained from nearby
wells or measured outcrop sections. Therefore, SPI cal-
culations often serve to consirain the possible lateral
changes in quantitative source-rock potential.
The concept and the calculations necessary to for-
mulate cumulative hydrocarbon potential, whi
have renamed SPI for practical reasons, were first
applied by Tissot et al. (1980) to show the variability in
hydrocarbon source potential of Cretaceous black
shales in Atlantic basins. Identical calculations were
later expanded to a global geochemical data base by
Demaison (1988) and Demaison and Huizinga (1989),
‘The original purpose for developing an SPI data base
was to create the operational recognition that sourc
rock volume, which is in part a function of thicknes
as important as source-rock richness,
Our concept of SPI is different from the “Source
Potential Rating Index” proposed by Dembicki and
Pirkle (1985). Their parameter is calculated by multiply-
ing the average total organic carbon content (96TOC) of
the source rock, the thickness of mature source rock,
and certain maturity scaling factors. Although the
method of Dembicki and Pirkle (1985) can be usefulfor mapping regional trends from a given source rock
in a well-known basin, it is not suited for making valid
comparisons of different source-rock sections from
basins around the world because it does not take the
variability of kerogen types into account. In contrast,
our SPI ranking scale has been developed for the pri-
mary purpose of making these global comparisons. We
have made ranking of very different source rocks possi-
ble by using hydrocarbon potential, rather than relative
Units of effectively realized generation, and by express-
ing source-rock richness in terms of the genetic poten-
tial (.e., Rock-Eval S; + Sp yield), rather than 9%6TOC.
Our use of the genetic potential, instead of %TOC,
allows for a fundamentally viable ranking of the
petroleum potential of diverse source rocks containing
issimilar kerogen types. Thus, lean source rocks con-
taining Type Ill kerogen but having substantial thick-
nesses (e.g, Tertiary sequence of the Niger delta, Juras-
sic source rocks of the Barrow-Dampier basin) can be
‘compared on the same SPI ranking scale to high-yield
source rocks containing Type I or Il kerogen and hav-
ing lower stratigraphic thicknesses (¢.g., Upper Jurassic
source rocks of the North Sea or West Siberia basin).
SPI can also be used for mapping regional variations in
a given source-rock unit, as discussed elsewhere in our
study.
Applications of SPI to Exploration
A relative source-rock ranking system has been devel-
‘oped by compiling the average SPls of individual source
rocks from various basins (Figure 3; Table 1). For each
source rock, the result listed in Table 1 has been deter-
mined by calculating SPI at as many locations as possi-
ble, within or adjacent to the hydrocarbon kitchens, and
averaging these values. Although these averaged results,
are tabulated in numerical order from highest to lowest,
only substantial relative differences are considered
important in the SPI comparison of two or more source
rocks, Some basins contain multiple source rocks, which
are listed separately in Table 1 (e.g., the Miocene and
Eocene-Oligocene source rocks of the San Joaquin
basin) of may not all appear due to lack of data
In the appraisal of a petroleum system, the source
rocks have to be evaluated in the context of the over-
all structural and stratigraphic framework, which deter-
mines the migration style. We have developed a pre-
liminary SPI classification (Figure 4), which has been
empirically derived from worldwide measurements of
SPI and a general knowledge of the discovered hydro-
carbon reserves that are likely to be derived from
these specific sources. In vertically drained petroleum
systems, SPIs are classified as low (SPI < 5), moderate
GS SPI < 15), and high (15 £ SPD. In contrast, lateral-
ly drained petroleum systems are significantly less
demanding on the magnitude of SPI than are vertically
drained systems. Compared to vertically drained
Demaison and Huizinga 1631
petroleum systems, laterally drained systems common.
ly contain broader areas of mature source rock, due to
more laterally extensive source-rock development
(eg., marine source rocks in laterally drained foreland
basins vs. lacustrine source rocks in vertically drained
early rift basins) and larger hydrocarbon-kitchen size
(eg., in laterally drained foreland basins vs. vertically
drained fold-and-thrust belts). Traps in laterally
drained systems tend to fetch petroleum from larger
areas of source-rock drainage, relative to those in ver
tically drained systems. Because areal extent is also
critical in determining source-rock volume, we tenta-
tively propose lower limits to define the SPI categories
in laterally drained systems (low: SPI < 2; moderate: 2
SPI <7; high: 7 € SPI). So far, our database includes
SPls from only a limited number of laterally drained
basins (e.g., Central Arabia, Maturin, West Siberia,
Paris, Illinois, Oriente, Williston); therefore, the SPI
categories for laterally drained systems are subject to
future refinement.
In extensively explored basins, a positive correlation
exists between the magnitude of SPI and basin-wide
petroleum reserves. Source rocks with high SPI, as
defined in Figure 4, are statistically associated with
large petroleum reserves and production from giant oil
and gas fields (Figure 3; Table 1). In contrast, basins
containing source rocks with low SPI, as defined in Fig-
ure 4, are associated with either small basin-wide
petroleum reserves and production from subgiant oil
and gas fields or are nonproductive (Figure 3; Table 1).
Although the use of SPI categories that shift with
respect to the size of drainage areas (Figure 4) shows,
great promise as an operationally viable shortcut for
estimating regional charging capacity, we will continue
to evaluate and refine this sliding SPI scale as more
data become available.
In the genetic classification of a petroleum system,
the generative subsystem is described by the charge fac=
tor (Figure 1), which semiquantitatively characterizes
the total amount of hydrocarbons available for entrap-
‘ment (.e., regional charge). The three terms used to
describe the charge factor (ie., supercharged, normally
charged, or undercharged) are provisionally assigned.
according to the magnitude of SPI (i.e., high, moderate,
or low SPI, respectively, as defined in Figure 4),
After a source rock showing a favorable SPI rating is
identified, it is economically important to delineate the
specific sectors of a basin that have the highest charg-
ing capacity. This is primarily achieved by the integra-
tion of SPI maps and maturity maps. The evaluation of
regional trends in hydrocarbon charging capacity is
also enhanced with the aid of paleogeographic maps, a
knowledge of source-rock depositional models
(Demaison et al., 1984), and a recognition of migration
styles. This information will help in interpolating the
charging capacity for areas occurring between locations
of known SPI and possibly assist in extrapolating
trends away from measured SPI data points,