Evolution and Origin of Deep Reservoir Water at the Activo Luna Oil Field, Gulf of Mexico, Mexico

Abstract

Petroleum wells of the Activo de Producción Luna oil field at the Mexican Gulf Coast are partially invaded by formation water at a production depth between 5000 and 6000 m. Measured 14C activities between less than 0.9 and 13.7% modern carbon reflect a homogenous, late Pleistocene-early Holocene age (40-10 ka) for the regional infiltration of meteoric and marine water into the reservoir. Before infiltration, both components were partially affected by atmospheric evaporation, which explains the hypersaline composition of some formation waters. Very positive δ18O values (up to + 12.5 ‰) of the formation waters are caused by strong secondary water-rock interaction processes and reflect close to equilibrium conditions between the carbonate host rock and the fluids. The formation of biogenic and/or thermocatalytic methane in some parts of the petroleum reservoir is indicated by δ13C values up to +20.4‰. Southwest-northeast-directed hydraulic migration of the deep aquifer between camps Sen and Escuintle-Pijije-Caparroso is indicated by interference tests and pressure drawdown characteristics, whereas northwest-southeast-trending thrust faults restrict communication toward the Luna and Tizón camps in the most northeastern part of the oil field. On a local scale, vertical zonation trends of the fluids with decreasing salinity toward upper parts of the common aquifer are related to separation processes by gravity and/or by the rising of condensed vapor. The migration of the fluids is mainly related to southwest-northeast-trending fractures and microfractures, whereas northwest-southeast- and northeast-south-west-trending reverse and normal faults, respectively, behave irregularly as barrier or as flow conduits. Recently, the extraction of petroleum caused an increased mobilization of the hydrodynamic aquifer system.

Full text

AAPG Bulletin, v. 86, no. 3 (March 2002), pp. 457–484 457
Evolution and origin of deep
reservoir water at the
Activo Luna oil field,
Gulf of Mexico, Mexico
P. Birkle, J. J. Rosillo Arago´n,E.Portugal,
and J. L. Fong Aguilar
ABSTRACT
Petroleum wells of the Activo de Produccio´n Luna oil field at the
Mexican Gulf Coast are partially invaded by formation water at a
production depth between 5000 and 6000 m. Measured
14
Cactiv-
ities between less than 0.9 and 13.7% modern carbon reflect a ho-
mogenous, late Pleistocene–early Holocene age (40–10 ka) for the
regional infiltration of meteoric and marine water into the reservoir.
Before infiltration, both components were partially affected by at-
mospheric evaporation, which explains the hypersaline composi-
tion of some formation waters. Very positive d
18
Ovalues (up to
Ⳮ12.5‰) of the formation waters are caused by strong secondary
water-rock interaction processes and reflect close to equilibrium
conditions between the carbonate host rock and the fluids. The
formation of biogenic and/or thermocatalytic methane in some
parts of the petroleum reservoir is indicated by d
13
Cvalues up to
Ⳮ20.4‰. Southwest-northeast–directed hydraulic migration of the
deep aquifer between camps Sen and Escuintle-Pijije-Caparroso is
indicated by interference tests and pressure drawdown character-
istics, whereas northwest-southeast–trending thrust faults restrict
communication toward the Luna and Tizo´n camps in the most
northeastern part of the oil field. On a local scale, vertical zonation
trends of the fluids with decreasing salinity toward upper parts of
the common aquifer are related to separation processes by gravity
and/or by the rising of condensed vapor. The migration of the fluids
is mainly related to southwest-northeast–trending fractures and mi-
crofractures, whereas northwest-southeast– and northeast-south-
west–trending reverse and normal faults, respectively, behave irreg-
ularly as barrier or as flow conduits. Recently, the extraction of
petroleum caused an increased mobilization of the hydrodynamic
aquifer system.
Copyright 䉷2002. The American Association of Petroleum Geologists. All rights reserved.
Manuscript received February 15, 2000; revised manuscript received April 5, 2001; final acceptance June
20, 2001.
AUTHORS
P. Birkle Instituto de Investigaciones
Ele´ctricas, Unidad Geotermia, A.P.1-475,
Cuernavaca, Morelos, 62001 Mexico;
birkle@iie.org.mx
Peter Birkle received his master’s degree in
geology/petrology from the Eberhard-Karls
University in Tu¨bingen (Germany) in 1992,
followed by a Ph.D. in hydrogeology /
hydrochemistry from the Technical University
of Freiberg, Saxony (Germany) in 1998. In
1993, he joined the research team of the
Geothermal Department at the Institute for
Electrical Research (Instituto de
Investigaciones Ele´ctricas) in Cuernavaca,
Mexico, as a specialist in deep groundwater
systems. Birkle’s current research includes
hydrogeological modeling of geothermal and
petroleum reservoirs, assessment studies of
reservoir extraction, environmental impacts by
geothermal exploitation, and identifying the
origin and pathways of deep groundwater
with isotopic and hydrochemical methods.
J. J. Rosillo Arago´n PEMEX—Exploracio´n
yProduccio´n, Activo de Produccio´n Luna,
Disen˜o y Evaluacio´n de Explotacio´n,
Comalcalco, Tabasco, 86388 Mexico;
jrrosiar@sur.pep.pemex.com
Jorge Rosillo Arago´n received his bachelor’s
degree in engineering geology from the
Autonomous University of Mexico (U.N.A.M.)
in 1975. Until 1991, he worked at the Mexican
Institute of Petroleum (IMP) and as a teaching
professor at the U.N.A.M. and the
Polytechnical Institute. From 1992 to 1993, he
was the executive manager of the Department
for Reservoir Characterization. Recently,
Rosillo Arago´n has been senior geologist and
coordinator of the Reservoir Characterization
Group at the Activo Luna oil field (PEMEX-
PEP), with major research interests in
structural effects on petroleum exploitation.
E. Portugal Instituto de Investigaciones
Ele´ctricas, Unidad Geotermia, A.P.1-475,
Cuernavaca, Morelos, 62001 Mexico;
portugal@iie.org.mx
Enrique Portugal joined the Institute for
Electrical Research in Cuernavaca, Mexico, as
aresearcher in 1984. Presently, he is in
charge of the isotope laboratory at the

458 Activo Luna Oil Field (Gulf of Mexico, Mexico)
INTRODUCTION
Waters from deeply buried geological formations are of special hy-
drogeological interest because they reflect the hydrodynamic or hy-
drostatic conditions of recent and/or past times. The origin of saline
waters in many sedimentary basins is still controversial. Based on
chemical analysis, early studies postulated that most of the deep
waters represented modified original seawater, locked in during the
time of sediment deposition (Rubey, 1951; Chave, 1960; Degens
et al., 1964; White, 1965). The chemical evolution of connate wa-
ter was explained by water-rock interaction processes (White,
1965). In contrast to these earlier chemical studies, the develop-
ment of isotope techniques in the 1960s revealed the initial mete-
oric origin for major parts of the formation water. The original con-
nate water of the formation was flushed out of the sediments by
compaction processes and replaced subsequently by younger me-
teoric water (Clayton et al., 1966). The first model postulates the
existence of impermeable reservoir conditions with hydrostatic
connate water, whereas the later one reflects an active, hydrody-
namic system with circulating fluids. A mixture of both models
seems to be the most probable approach for most petroleum res-
ervoirs. Carpenter (1978) and Knauth and Beeunas (1986) sug-
gested just a partial migration of the connate fluids, whereas the
subsequent invasion of meteoric or marine water causes mixing
processes.
Apart from its scientific interest, formation water can affect the
economy of producing oil fields. Cases of brine water invasion in
petroleum reservoirs, especially in sedimentary basins, are known
from a variety of global oil fields, such as the Western Canada sed-
imentary basin (Hitchon and Friedman, 1969) and the central Mis-
sissippi Salt Dome basin (Kharaka et al., 1987). The majority of oil
wells, especially in the more mature North American fields, pro-
duce more water than they do oil (Peachey et al., 1998). In the
special case of the oil field Activo de Produccio´n Luna in Mexico,
increasing volumes of invading water into the petroleum wells were
detected in the past few years. A hydrochemical and isotopic study
on the origin of the Activo Luna reservoir water was carried out
from the Instituto de Investigaciones Ele´ctricas for the national oil
company PEMEX (P. Birkle et al., 1999, unpublished data).
The present study reveals data on the hydrochemical and iso-
topic composition of formation water from 15 wells of the Activo
de Produccio´n Luna oil field, as well as reference samples from
rivers, lagoons, shallow wells, seawater from the Gulf of Mexico,
and precipitation water. The residence time and/or the age of the
reservoir water was determined by measuring the
14
Cactivities and
the incorporation of the values into black-box models. The stable
isotopes
13
C,
18
O, and D, as well as the major and trace element
composition, are used to reconstruct secondary effects, such as
water-rock interaction or methane formation, that modify the pri-
mary composition of the fluids during their evolution. Based on the
geological and tectonic history of the region and the chemical-
ACKNOWLEDGEMENTS
We thank the staff from Activo Luna for pro-
viding information and technical support. We
appreciate the helpful comments from the re-
viewers L. M. Walter and B. Hitchon.
Geothermal Department. He received his
master’s degree in chemical engineering at
the Autonomous University of the State of
Morelos (UAEM), Mexico, in 1992. His major
interests are stable isotope studies of
geothermal fluids.
J. L. Fong Aguilar PEMEX—Exploracio´n y
Produccio´n, Activo de Produccio´n Luna,
Disen˜o y Evaluacio´n de Explotacio´n,
Comalcalco, Tabasco, 86388 Mexico;
jsfongag@sur.pep.pemex.com
Jose´Luis Fong Aguilar received his bachelor’s
and master’s degrees in petroleum
engineering from the National Autonomous
University of Mexico (UNAM), Faculty of
Engineering, in 1983 and 1991, respectively.
In 1983, he began to work for the National
Oil Company PEMEX (Petro´leos Mexicanos),
mainly in the realization and evaluation of
reservoir simulations, as well as exploitation
studies of oil fields in the southern production
zone of Mexico. Simultaneously, he
collaborated as teaching professor at the
Faculty of Engineering of the UNAM. He is the
former coordinator of the Design and
Exploitation Department and the actual
administrator of the Activo Luna oilfield
(PEMEX-PEP).

Birkle et al. 459
isotopic composition of the fluids, a hydrogeological
model is presented for the Activo de Produccio´nLuna
oil field.
TOPOGRAPHICAL LOCATION AND
PRODUCTION DATA
The Activo Luna oil field is located in the coastal
swamps of the Gulf of Mexico between the urban cen-
ters of Frontera, Comalcalco, and Nacajuca and is nat-
urally limited to the north by the Gulf Coast and to
the east by the Grijalva River (Figure 1). It is located
70 km north of Villahermosa, the capital of the state
of Tabasco and comprises an area of approximately
1500 km
2
.The field has been operating since 1982,
with a 1998 production of 85,000 b/day of petroleum
and 270 MMCFGD from an average reservoir depth
of 5000 m at five separate production camps: Sen-
Cardo, Pijije-Caparroso-Escuintle, and Escarbado ex-
tract volatile gas, whereas Luna-Palapa and Tizo´n pro-
duce gas and condensate. With 19 production wells,
Sen represents the most important field, followed by
Luna-Palapa (18), Pijije-Caparroso-Escuintle (17),
Tizo´n (2), Escarbado (2), and Cardo (1) (Rosillo,
1998).
REGIONAL FRAMEWORK
The Activo Luna oil field forms part of the northwest-
southeast–trending Villahermosa uplift horst struc-
ture, which is separated from the Comalcalco basin to
the northwest, and from the Macuspana basin to the
southeast by the Comalcalco and Frontera faults, re-
spectively (Salvador, 1991). These structures were
formed during the Laramide orogeny, which modified
the paleogeography of the western flank of the Gulf of
Mexico basin. The resulting formation of the Sierra
Madre Oriental was accompanied by a series of de-
pressions, known as the Tertiary basins of eastern Mex-
ico: the Burgos, Tampico-Misantla, and Veracruz ba-
sins, and the so-called Southeastern basins (Isthmus
Saline, Comalcalco, and Macuspana basins). The Villa-
hermosa uplift comprises between 5000 and 10000 m
of Tertiary and Mesozoic sediment layers in the west-
ern and eastern part, respectively (Salvador, 1991).
Upper Middle Jurassic salt deposits as part of hy-
persaline basins formed at the same time in the north-
ern (East Texas basin, North Louisiana salt basin, Mis-
sissippi Salt Basin, Jasper arch, Adams County high)
and southern part (Bay of Campeche) of the Gulf of
Mexico. Their thickness decreases toward the south-
eastern part of Mexico, and they are very thin or absent
in the Tabasco region (Martin and Foote, 1981; Sal-
vador, 1991). The main ascent into younger formations
occurred during the Cenozoic.
LOCAL FRAMEWORK
The petroleum wells from the Sen and Luna camps
derive their production from Upper Jurassic (Kim-
meridgian), Cretaceous, and Tertiary rock units,
whereas the other camps are related exclusively to
Cretaceous formations. The Cretaceous and upper
parts of the Jurassic Tithonian comprise limestone
with mudstone to wackestone texture, and planktonic
foraminifers, deposited in a deep-water marine envi-
ronment. The lower Tithonian and the Kimmeridgian
Jurassic layers consist of micro- to mesocrystalline do-
lomite, originally deposited within an environment of
platform carbonates (Rosillo, 1998). The overlying
Cenozoic units are formed by gray, fossiliferous ma-
rine shales with intercalations of quartz sandstone,
bioclastic limestone, conglomerates, bentonic mate-
rial, and volcanic-ash beds (Ricoy, 1989). The Pleis-
tocene section includes fluvial-deltaic systems that
were probably deposited by the ancestors of the Co-
atzacoalcos, Uzpanapa, Grijalva, and Usumacinta riv-
ers (Salvador, 1991).
The primary porosity of the Cretaceous strata is
practically nonexistent (1–2%), and the matrix of the
reservoir rocks is practically impermeable (0.004–1.7
md). The hydrocarbon production comes from open
macro- and microfractures generated very recently
(Miocene–Pliocene?). Temperatures above 160C
(Table 1) at the Cretaceous reservoir could imply its
recent generation. Average porosity and permeability
values of 2–9% and 25 md, respectively, were mea-
sured from 99 core samples from the Caparroso-
Escuintle area. Furthermore, maximum open spaces
of 14 mm diameter and permeability values of more
than 10 d were detected by Stonley wave anomalies
at the Caparroso-197 well (Rosillo, 2000). The con-
ductive fractures are oriented in northeast-southwest
and southeast-northwest directions, with openings
between 0.00001 and more than 0.2 cm. Between
the Caparroso-81 and Escuintle-201 wells, the mega-
fractures of normal faults are located at a regular dis-
tance of 300 m at the Upper Cretaceous level (Ros-
illo, 2000). A correlation between the lithological

460 Activo Luna Oil Field (Gulf of Mexico, Mexico)
composition of the reservoir rocks and the distribu-
tion of fractures, however, cannot be observed.
Oligocene compressive tectonic events formed
northwest-southeast–oriented thrust faults, as well as
the folding of the formations, as part of horst and gra-
ben structures and allowed the formation and migra-
tion of the hydrocarbons starting in the Miocene (Ros-
illo, 1998). Figure 2 shows the altitude (meters below
sea level [mbsl]) of the top of the Upper Cretaceous
units, as well as the distribution of the northwest-
southeast– and northeast-southwest–oriented thrust
and normal faults, respectively. Figure 3 shows a geo-
logical cross section through the Activo Luna oil field
from southwest to northeast, incorporating the Sen,
Figure 1. Location of the indi-
vidual camps of the Activo Luna
oil field, located close to the
coast of the Gulf of Mexico. Ro-
man numerals ⳱surface water
samples; Arabic numerals ⳱
studied formation water sam-
ples (symbol description in Ta-
ble 1).

Birkle et al. 461
Table 1. Chemical (Concentrations in mg/L), Stable (
2
H,
18
O,
13
C), and Radioactive (
14
C) Isotopic Composition of the Formation Waters from the Activo Luna Reservoir*
Well pH
T
surface
(C)
T
bottom
(C)
Conduc-
tivity
(lS/cm)
Salinity
(mg/L) Cl Na Mg K Ca HCO
3
SO
4
FB Li Si Mn Fe
Sen-65 (S-65) 6.1 58.4 153.8 132,700 115,200 69,500 35,000 481 1,640 7,770 210 116 6.95 163.1 21.8 60.3 2.33 97.2
Sen-121 (S-121) 7.0 31.0 154.0 52,000 26,000 16,100 7,220 138 262 2,090 n.d. 63.3 2.6 38.5 6.94 15.2 1.23 80.0
Escuintle-1 (E-1) 5.6 56.0 168.0 486,000 277,500 173,000 78,500 1,130 4,650 18,800 150 82.1 16.2 397.0 45.1 39.5 6.54 188
Escuintle-2 (E-2) 5.8 38.8 160.0 131,600 103,600 70,800 22,400 480 2,610 6,590 230 259 11.0 210.3 25.5 71.4 1.23 51.2
Caparroso-35 (C-35) 5.3 52.2 163.8 172,600 179,100 108,500 52,200 902 3,540 12,900 151 111 7.1 279.2 32.3 56.9 4.89 156
Caparroso-85 (C-85) 5.4 35.0 162.0 480,000 284,200 168,500 79,000 1,830 11,100 22,200 151 115 10.1 373.3 35.3 33.4 5.96 222
Caparosso-192 (C-192) 6.2 44.2 165.0 149,800 147,900 90,500 41,400 822 1,780 11,200 200 203 6.75 228.2 25.8 52.6 3.59 112
Pijije-1A (P-1A) 5.4 37.2 n.m. 190,000 199,700 124,900 56,100 999 2,910 13,800 170 12.2 4.55 202.7 51.7 56.1 5.76 139
Pijije-12 (P-12) 6.0 33.1 165.5 3,780 3,660 2,680 854 3 13 65 225 51 2.45 52.6 0.959 3.84 0.293 9.28
Pijije-21 (P-21) 6.3 59.4 166.0 2,590 510 330 72 2 5 104 n.d. n.d. 1.7 21.5 0.176 2.0 1.91 29.3
Pijije-41 (P-41) 5.1 51.6 166.8 192,000 216,900 134,700 61,500 1,020 2,000 16,500 155 38.9 12.4 253.7 40.8 50.8 6.17 183
Luna-2 (L-2) 6.5 49.8 166.0 1,110 70 57 8 0.4 12372 5.89 0.7 24.3 0.1 2.0 0.013 0.3
Luna-3B (L-3B) 5.2 45.5 163.0 530,000 292,600 180,000 82,000 1,020 5,040 23,000 175 n.d. 9.45 222.5 51.6 33.9 29.6 223
Tizo´n-1 (T-1) 6.6 39.2 175.4 16,300 6,190 4,550 1,390 14 37 169 1,205 28 2.8 18.7 1.08 8.44 0.407 1.37
Tizo´n-3DL (T-3D) 6.8 38.4 173.0 33,400 19,600 11,800 6,990 5 510 52 1,275 197 14.3 154.6 8.81 88.1 7.4 0.46
Well Ge As Se Br Rb Sr Te I Cs Ba Tl
Balance
(%)
14
C-
Technique**
d
13
C
(‰)
14
C
mod
(pMC)
2r
(pMC)
d
18
O
(‰)
dD
(‰)
Sen-65 (S-65) 0.03 0.121 1.23 658.5 4.39 1,158 0.111 34.2 1.25 37.9 0.12 0.7 CON 3.4 9.16 1.00 10.4 ⳮ22
Sen-121 (S-121) 0.01 0.088 0.305 136.6 0.573 334.9 0.03 4.43 0.159 6.54 0.003 ⳮ2.1 AMS 0.4 13.73 0.16 5.2 ⳮ33
Escuintle-1 (E-1) 0.16 0.384 2.6 1,318 9.22 1,986 0.332 41.2 3.07 77.9 0.229 ⳮ3.4 CON 1.5 3.59 1.53 12.5 ⳮ23
Escuintle-2 (E-2) 0.44 0.061 0.971 508.4 4.42 647.4 0.053 22 1.28 20.5 0.033 ⳮ15.5 CON 2.5 1.22 0.89 12.5 ⳮ22
Pijije-1A (P-1A) 0.14 0.297 1.92 1,022 6.53 1,300 0.165 43.4 2.92 94.7 0.505 ⳮ3.5 AMS 9.8 1.86 0.16 9.2 ⳮ25
Pijije-12 (P-12) 0.01 0.005 0.02 4.69 0.033 4.42 0.02 1.16 0.004 0.396 0.001 ⳮ31.1 CON 2.7 9.27 0.90 8.0 ⳮ24
Pijije-21 (P-21) 0.001 0.003 0.02 0.339 0.014 2.34 0.02 0.082 0.0002 1.58 0.007 ⳮ3.9 AMS ⳮ3.7 4.59 0.15 6.2 ⳮ32
Pijije-41 (P-41) 0.09 0.269 1.99 1,026 3.43 1,690 0.197 38 1.49 63.9 0.395 ⳮ2.2 CON 1.3 2.01 1.01 11.9 ⳮ28
Caparroso-35 (C-35) 0.11 0.223 1.63 844.6 6.8 1,078 0.104 23 2.19 47.6 0.412 0.3 CON 4.3 1.43 0.86 11.8 ⳮ23
Caparroso-85 (C-85) 0.07 0.482 2.76 1,430 13.5 1,671 0.253 22.9 2.53 90.8 0.408 2.3 AMS 20.4 2.24 0.16 11.9 ⳮ22
Caparosso-192 (C-192) 0.12 0.129 3.74 2,034 3.79 1,373 0.138 30 1.6 23.9 0.039 ⳮ1.7 AMS 8.9 3.35 0.16 11.3 ⳮ20
Luna-2 (L-2) 0.001 0.003 0.02 0.332 0.001 0.968 0.02 0.02 0.0002 2.31 0.002 ⳮ88.9 AMS 6 1.06 0.16 9.9 2
Luna-3B (L-3B) 0.01 0.548 3.03 1,492 12.4 2,069 0.429 43.2 3.69 177 0.199 ⳮ1.5 CON 4.6 2.96 1.05 11.2 ⳮ15
Tizo´n-1 (T-1) 0.001 0.003 0.02 4.88 0.058 34.8 0.02 0.486 0.008 2.42 0.0006 ⳮ34.1 CON 6.5 8.47 1.17 10.2 5
Tizo´n-3DL (T-3D) 0.26 0.003 0.102 58.1 0.648 29.8 0.02 14.7 0.16 13.8 0.0005 ⳮ5.4 CON 7.2 0.96 – 10.8 ⳮ19
*n.d ⳱not detected at that lower limit; n.m. ⳱not measured.
**CON ⳱conventional method; AMS ⳱acceleration mass spectrometry.

462 Activo Luna Oil Field (Gulf of Mexico, Mexico)
Figure 2. Detail map of the individual camps of the Activo Luna oil field, including altitude isolines (mbsl) of the top of the Upper
Cretaceous units (in mbsl), distribution of northwest-southeast– and northeast-southwest–oriented thrust and normal faults, and the
interpreted flow direction of the deep aquifer.
Escarbado, Caparroso, Pijije, Luna, and Tizo´n camps.
Besides the tectonic structures, salt intrusions caused
discontinuities of the geological formations and prob-
ably in the hydraulic connection of deep aquifers and
their fluids. Lithological evidence of salt cores from
some isolated wells, such as below the Upper Creta-
ceous units of the Caparroso-45 and Caparroso-85
wells and at the bottom of the Escuintle-201 well, does
not allow definition of the diapiric or bedded type of
salt intrusion structure. Seismic profiles, however, such
as the 5 km long section between the Manea camp
(close to well Manea-1) and the Luna camp (close to
well Luna-11A) (Figure 4; section AA⬘in Figure 2)
indicate a horizontal sheet structure of the Jurassic salt
intrusions, as observed for the eastern part of the Loui-
siana slope (Fairchild and Nelson, 1989) and for the
Mississippi Fan fold belt (Sarkar et al., 1995).
In the initial production phase of the Activo Luna
field, the primary piezometric level of the water-
hydrocarbon contact was Ⳳ5600 mbsl for the Sen and

Birkle et al. 463
Figure 3. Geological cross section from southwest to northeast of the Activo Luna oil field (modified after Rosillo, 2000). UC
Men
⳱
Upper Cretaceous (Mendez Formation); UC
Sf
⳱Upper Cretaceous (San Felipe Formation); UC
San
⳱Upper Cretaceous (Agua Nueva
Formation); MC ⳱middle Cretaceous; LC ⳱Lower Cretaceous; UJ
Tit
⳱Upper Jurassic (Tithonian Formation); UJ
Kim
⳱Upper
Jurassic (Kimmeridgian Formation).
Caparroso camps, whereas 5900 and 6100 mbsl were
measured in the Luna and Tizo´n camps, respectively
(Figure 3).
STRUCTURAL EVOLUTION
The deposition of Jurassic and Cretaceous carbonate
sediments was interrupted by the generation of local
discordances during the middle Cretaceous. From the
Late Cretaceous to the late Eocene, the deposition of
carbonate sediments continued with an increasing
component of terrigenous sediments. The maximum
folding of structures and the formation of thrust faults
occurred during the Oligocene. The subsidence of the
Reforma-Akal basin, with normal faulting and terrig-
enous sedimentation, occurred mainly in the Miocene,
when the hydrocarbon generation window reached
Upper Jurassic Tithonian units. From the Pliocene to
the Holocene, subsidence continued, and normal
faults were formed down to the Tertiary. The seismic
profile section (Figure 4) between the Manea camp
(close to well Manea-1) and the Luna camp (close to
well Luna-11A) shows the dominance of reverse
faults within the Mesozoic formations. Probably, hy-
drocarbons are still being generated and expelled at
the Activo Luna area (Rosillo, 1998). In a similar case,
large volumes of gases, generated by the Antrim Shale
in the Michigan Basin, are likely to be only on the
order of a few tens of thousands of years old (Martini
et al., 1998).
METHODS
Oil field water was sampled at 15 production wells
from the Sen, Pijije-Caparroso-Escuintle, Luna, and
Tizo´n camps. The sampled waters are from (1) Upper

464 Activo Luna Oil Field (Gulf of Mexico, Mexico)
Figure 4. Seismic profile sec-
tion between the Manea camp
and the Luna camp (Section
AA⬘in Figure 2).
Cretaceous units: Sen-121, Sen-65, Pijije-21 (Mendez
Formation); Caparroso-192 (San Felipe Formation);
Escuintle-1, Pijije-12, Caparroso-85 (Agua Nueva For-
mation); (2) middle Cretaceous units: Escuintle-2,
Pijije-41, Pijije-1A, Tizo´n-1, Tizo´n-3DL; (3) Lower
Cretaceous units: Caparroso-35; and (4) Upper Juras-
sic units: Luna-2, Luna-3B (Kimmeridgian). The water
fraction was separated from the petroleum phase di-
rectly in the field by accumulation of the water fraction
at a dead-end extraction line, facilitated by differences
in density. Nine reference samples (baseline) were
taken from rivers, lagoons, and shallow wells, as well
as from ocean water from the Gulf of Mexico and pre-
cipitation water. The sample localities and site descrip-
tions are shown in Figure 1 and Tables 1 and 2,
respectively.
Depending on the carbon content of the sample
and the available amount of sample material, the con-
ventional technique, as well as the accelerated mass
spectrometry (AMS) technique, was applied to mea-
sure the
14
Ccontent; measuring was carried out by
GEOCHRON Laboratories, Cambridge, Massachu-
setts. Nine samples were analyzed by the conventional
14
Cage method. The samples were made alkaline (pH
9.0) by adding NaOH. The bicarbonates were pre-
cipitated in the form of BaCO
3
by adding BaCl
2
.Six
samples of 1 L each were taken directly at the well
head and placed in sealed Nalgene bottles for their
AMS measurement. The homogeneous range of
14
C
concentrations from both methods reflects the lack of
contamination (by atmospheric CO
2
)ofthe conven-
tional samples during sampling and preparation.
The water samples for the analysis of major
and trace elements were prefiltered using 0.45 lm
Millepore filters, acidified with HNO
3
-Suprapur, and
analyzed by ACTLABS, Ontario, Canada, using the
inductively coupled plasma–mass spectrometry (ICP-
MS) technique. The anions were measured using a
colorimeter and atomic absorption spectroscopy
(AAS) at the Facultad de Ciencias Quı´micas e Ingen-
ierı´a of the Universidad Auto´noma del Estado de Mo-
relos, Cuernavaca, Mexico. The d
18
Oand dDisotope
values were analyzed using the mass spectrometry
(MS) technique at the Unidad de Geotermia of the
Instituto de Investigaciones Ele´ctricas, Cuernavaca,
Mexico. The dDvalues were determined using the

Birkle et al. 465
Table 2. Chemical and Isotopic Composition of Surface Water Samples from the Activo Luna Area*
Sample Water Type Locality pH
Salinity
(mg/L)
Cond.
(lS/cm)
14
C-
Technique**
d
13
C
(‰)
14
C
mod
(pMC) 2r
d
18
O
(‰)
dD
(‰)
IPrecipitation Adjacent to well
Pijije-1A
6.1 18 5 AMS ⳮ16.1 85.51 0.79 ⳮ2.9 ⳮ15.0
II Lagoon Lagoon Mecoaca´n 7.1 1,397 2,270 CON ⳮ11.5 119.03 2.39 ⳮ4.5 ⳮ30.0
III Seawater:
Gulf of Mexico
Miramar Beach 8.2 32,020 49,400 CON ⳮ1.3 110.67 1.40 0.9 ⳮ1.0
IV River River Gonza´lez 7.4 547 937 CON ⳮ10.9 103.38 1.99 ⳮ5.2 ⳮ35.0
VShallow well Base Camp Pijije 7.5 1,317 2,190 CON ⳮ19.3 1.37 0.47 ⳮ3.1 ⳮ19.0
VI River River de la Pigua 7.5 171 260 CON ⳮ10.3 98.01 2.02 ⳮ5.6 ⳮ37.0
VII Shallow well Base Camp Luna 7.6 988 3,290 n.a. n.a. n.a. n.a. ⳮ2.6 ⳮ13.0
VIII River Victoria Channel 7.7 173 389 n.a. n.a. n.a. n.a. ⳮ5.6 ⳮ38.0
IX Lagoon Lagoon Santa Anita 7.0 423 853 n.a. n.a. n.a. n.a. ⳮ3.7 ⳮ28.0
*n.a. ⳱not analyzed.
**CON ⳱conventional method; AMS ⳱acceleration mass spectrometry.
zinc-reduction method described by Tanweer (1993)
for brine samples. Oxygen isotope ratios were mea-
sured using the CO
2
equilibration method (Epstein
and Mayeda, 1953). The equation of Sofer and Gat
(1972) was used to correct the
18
Oactivity data be-
cause of salinity effects. Isotopic analyses are given as
standard drelative to standard mean ocean water
(SMOW), with a precision of Ⳳ2and Ⳳ0.15‰ for
dDand d
18
O, respectively.
RESULTS
Hydrochemical Composition
Table 1 shows the temperature of the formation water
at the surface during sampling, the measured reservoir
temperature, conductivity, pH, and the concentration
of major elements and selected trace elements. The
water samples from the oil wells are characterized by
avery heterogeneous chemical composition, with sa-
linity values from 0.07 (Luna-2) to 292.6 g/L (Luna-
3B). Samples such as those from Luna-2 have a low-
mineralized drinking water composition. The Sen-121
sample is close to seawater salinity, whereas the
Caparroso-85 sample exceeds by eight times the salt
concentration of ocean water. Hitchon (1996) ex-
plains large differences in salinity of two aquifers due
to the separation by strong aquitards. According to the
classification proposed by Davis (1964), four types of
Activo Luna reservoir waters can be distinguished:
•Freshwater: Luna-2, Pijije-21 (total dissolved solids
[TDS] 1.0 g/L)
•Brackish water: Pijije-12, Tizo´n-1 (TDS: 1.0–10
g/L)
•Saline water: Tizo´n-3DL, Sen-121 (TDS: 10–100
g/L)
•Brine: Escuintle-2, Sen-65, Caparroso-192, Capar-
roso-35, Pijije-1A, Pijije-41, Escuintle-1, Caparroso-
85, Luna-3B (TDS 100 g/L)
On a Piper plot (Figure 5), the Activo Luna formation
waters are typical chloride-type waters with the excep-
tion of the sodium bicarbonate composition of the
Luna-2 sample (L-2). Based on the cationic composi-
tion, the samples can be further distinguished into an
alkaline-sodium type of water without (Tizo´n-1,
Tizo´n-3DL, Pijije-12) or with little calc-alkaline influ-
ence (other wells in Figure 5), as well as calc-alkaline
calcium water (Pijije-21).
For each sample, a linear correlation between the
abundance of major elements and the concentration of
metals and nonmetals was observed. The very saline
Luna-3B, Caparroso-85, Escuintle-1, Pijije-1A, Capar-
roso-192, and Escuintle-2 brines are especially en-
riched in trace elements (maximum concentrations in
parentheses) (Table 1): Sr (2069 mg/L), Br (2034 mg/
L), B (397 mg/L), Fe (223 mg/L), Ba (177 mg /L), Mn
(29.6 mg/L), Li (51.7 mg/L), I (43.4 mg/L), F (16.2
mg/L), Rb (13.5 mg /L), Se (3.74 mg/L), Cs (3.69 mg/
L), As (0.55 mg/L), Tl (0.51 mg/L), Ge (0.44 mg/L),
and Te (0.43 mg/L).

466 Activo Luna Oil Field (Gulf of Mexico, Mexico)
Figure 5. Piper plot: chemical
classification of the Activo Luna
formation waters.
Primary Origin of the Oil Field Waters: Residence Time
Because of their low carbon contents and /or littleavail-
able sample volume, the water samples from Sen-121,
Pijije-21, Pijije-1A, Luna-2, Caparroso-192, and
Caparroso-85, as well as one precipitation sample, were
analyzed by the AMS technique. The conventional pre-
cipitation technique was applied to the remaining sam-
ples (Table 1). The
14
Cvalues of the formation waters
range between less than 0.96 and 13.73 Ⳳ0.16% mod-
ern carbon (pMC), whereas values between 85.51 Ⳳ
0.79 and 119.03 Ⳳ2.39 pMC from the rivers, lagoon,
and the Gulf of Mexico seawater reflect the typical at-
mospheric composition of surface water.
Several uncertainties, such as the concentration of
the initial atmospheric input of
14
Cand the type of
infiltration event, make it difficult to calculate the res-
idence time of deep formation waters.
Initial Atmospheric Input
In general, a standardized initial
14
Cconcentration of
85 Ⳳ5pMC, which is independent of the climatic
conditions, is applied to calculate the
14
Cresidence
time of groundwater (Vogel, 1970). As a result, the
residence time of the Activo Luna deep waters ranges
between 16.49 ka (Sen-121) and more than 38.63 ka
(Tizo´n-3DL) for model 1 (Table 3).
The complete reaction of soil CO
2
with carbonate
rock, however, causes an isotopic equilibrium between
both components and a dilution of 50% of the initial
14
Cconcentration:
14
CO (soil) ⳭHOⳭCaCO
22 3
2
Ⳮⳮ
14
ⳮ
rCa ⳭHCO ⳭHCO
33
In the case of complete isotopic equilibrium, the
14
C
content in the new /CO
2
product is 50% of
ⳮ
HCO
3
the initial biogenetic CO
2
(Geyh, 1980). For the Ac-
tivo Luna reservoir, the dead carbon content of the
Jurassic–Cretaceous reservoir limestones could cause
the mentioned dilution of the atmospheric
14
Cinput
and a minimum residence time of the fluids (model 2).
Age variations between 10.76 and more than 32.9 ka
for model 2 indicate that maximum interaction pro-
cesses between brine and dead carbon limestone cause
aslight decrease of the residence time of the reservoir
fluids (Table 3).
The fractionation factor of
14
Cis2.3 times higher
than the corresponding
13
Cfactor, thus the calculated
ages of both models were corrected by the following
factor (Saliege and Fontes, 1984):
13
2.3 (dCⳭ25)‰
sample

Birkle et al. 467
Table 3. Residence Time of the Activo Luna Reservoir Water Since its Infiltration (in years)*
Sample
Model 1:
a
o
⳱85%, q⳱1.0
Model 1:
d
13
C-Corrected
Model 2:
a
o
⳱85%, q⳱0.5
Model 2:
d
13
C-Corrected
Tizo´n-3DL 37,990 38,630 32,260 32,900
Tizo´n-1 19,730 20,350 14,000 14,620
Sen-121 15,990 16,490 10,260 10,760
Sen-65 19,340 19,890 13,610 14,170
Pijij-41 31,880 32,400 26,150 26,670
Pijije-21 25,050 25,470 19,320 19,740
Pijije-12 19,240 19,790 13,510 14,050
Pijije-1A 32,520 33,210 26,790 27,480
Luna-3B 28,680 29,260 22,950 23,530
Luna-2 37,170 37,780 31,440 32,050
Esc-2 36,010 36,550 30,280 30,820
Esc-1 27,090 27,610 21,350 21,870
Cap-192 27,660 28,330 21,930 22,600
Cap-85 30,990 31,900 25,250 26,170
Cap-35 34,700 35,270 28,970 29,540
*Considering an initial
14
Cvalue of 85 pMC (model 1) and the effect of dead carbon by water-rock interaction processes (model 2); a
o
⳱initial activity; q⳱dilution
factor.
Type of Infiltration Processes
The piston-flow model represents a hydrodynamic
model, with the hypothesis that the initial tracer con-
centration is not altered by dispersion, diffusion, or any
other interaction effect during subterranean migration.
Asingle defined infiltration event and the homoge-
neous velocity of the water particles over the flow path
are assumed. The initial concentration of the tracer, c
in
,
decreases exclusively by the radioactive decay of the
isotope (Moser and Rauert, 1980),
ⳮ
ks
c(t)⳱c(tⳮs)e
out in
where c
out
represents the measured
14
Cconcentration
or activity at time t, sis the residence time, and kis
the decay constant. Applying the piston-flow model,
the age of the Activo Luna reservoir waters ranges be-
tween 15 (Sen-121) and more than 37 ka (Tizo´n-
3DL). This is not very unusual, as Martini et al. (1998)
describe a common glacial age for naturally occurring
influx of freshwater into gas and oil reservoirs. In many
mid-continent basins, remnant glacial waters have
been identified isotopically, such as in the Illinois basin
(Stueber and Walter, 1994), in western Ontario
(Weaver et al., 1995), and in Great Plains aquifers (Sie-
gel and Mandle, 1984).
The exponential model assumes continuing infil-
tration of meteoric water over different time periods,
with individual fractions diminished as their residence
time increases within the aquifer. Because of the mix-
ing of waters of varying ages, the equation is expressed
by a folded integral:

1
ⳮ
ks
ⳮ
(s/¯s)
c(t)⳱c(tⳮs)e•eds
out in
冮
0
s
The model results in an average residence time of 49–
55 k.y. for the water particles within the Activo Luna
reservoir.
An inverse trend exists between the measured
14
C
activities and d
18
Ovalues (Figure 6). The youngest wa-
ter samples (group A) show more negative d
18
Oval-
ues, whereas older waters (with low
14
Cactivities) are
enriched in d
18
O(group B). This indicates the time
dependence intensity of water-rock interaction pro-
cesses: a relative short residence within the reservoir
(20 k.y.), corresponding to a shorter duration for
chemical-thermal processes, causes minor water-rock
interaction processes (group A). In contrast, an ex-
tended residence time (20 k.y.) allows more time for
interaction with the limestone.
Secondary Effects
Avariety of chemical, thermal, and physical processes
can affect the primary chemical and isotopic compo-
sition of groundwater.

468 Activo Luna Oil Field (Gulf of Mexico, Mexico)
Figure 6. Correlation of the
14
Cand d
18
Ocomposition of the
Activo Luna formation waters.
Figure 7. Log Cl vs. log Br for the Activo Luna formation
waters: evolution from an initial seawater composition along the
seawater evaporation trajectory and deviations by halite disso-
lution and mixing processes.
Chemical Indications
Evaporation Effect
The chemical evolution of brines produced by evapo-
ration of seawater has been studied by Zherebtsova and
Volkova (1966), Herrmann and Knake (1973), and
Carpenter (1978). The ratio of the inert constituent
bromine to other constituents is an especially sensitive
indicator of the evaporation process and the formation
of evaporite minerals. During the evaporation of sea-
water, the Cl/Br ratio shows a linear, positive trend
until the saturation of halite is reached (Figure 7). At
this point, halite starts precipitating, whereas bromine
becomes concentrated in the residual fluid, causing a
horizontal trend of the evaporation trajectory.
The evolution of the Activo Luna reservoir water
was simulated using G
EOCHEMIST
’
S
W
ORKBENCH
soft-
ware (Bethke, 1994). The composition of the Playa
Miramar sample from the Gulf of Mexico was used as
the initial seawater composition. The trajectory line for
the evaporation of seawater was very similar to those
documented by Carpenter (1978) and Knauth (1988).
The samples Caparroso-85, Escuintle-1, and Luna-3B
are located on the horizontal section of the evaporation
line, which can be interpreted as an extreme advanced
evaporation process of seawater with ongoing precipi-
tation of halite. Samples below the trajectory line suf-
fered evaporation of marine water at the surface and
are additionally affected by mixing processes in the
form of subsequent dilution with fresh water or marine
water (Carpenter, 1978; Knauth, 1988; Connolly et
al., 1990). The numbers (fine horizontal lines) in Fig-
ure 7 give the percentage of dilution. Pijije-41, Pijije-
1A, and Caparroso-35 were mixed with seawater (20–
40%), whereas Escuintle-2 and Sen-65 show seawater
dilutions above 60%. The Sen-121 and Tizo´n-3DL
samples seems to reflect an original marine composi-
tion but can be also related to the addition of fresh-
water to the brine. This trend is reflected by a line
parallel to the evaporation trajectory (Figure 7) (Car-
penter, 1978). The low mineralization of the Tizo´n-1,
Pijije-21, Pijije-12, and Luna-2 samples (not shown in
Figure 7), however, implies a meteoric origin for those
samples. None of the samples is affected by halite dis-
solution, which would be reflected in a Cl/Br ratio
above the evaporation trajectory.
The classification diagram by Rittenhouse (1967)
is based on the TDS and bromine concentration of the
samples (Figure 8). Similar to the results of the Cl/Br
diagram, none of the samples seems to be affected by
halite dissolution (group III in Figure 8). The compo-
sition of the Sen-121 and Tizo´n-3DL samples is very
similar to marine water from the Gulf of Mexico
(“Ocean”), whereas most samples are affected by the
excessive evaporation of seawater (group V). Luna-2,
Pijije-21, Pijije-12, and Tizo´n-1 composition can be ex-

Birkle et al. 469
Figure 8. Log total dissolved
solids (TDS) vs. log Br: classifi-
cation of the Activo Luna for-
mation waters according to
Rittenhouse (1967).
Figure 9. Diagram of Cl /Br and Na/Br ratios of the Activo
Luna formation waters. Most are formed by the evaporation of
seawater (SW
ET
).
plained by a pure meteoric origin, as are the reference
samples from the Lagoon Mecoaca´n (“Lag-Mec”) (Ta-
ble 2), or by dilution of marine water with a meteoric
component (group IV).
The Na-Cl-Br systematics, graphically presented as
Cl/Br and Na /Br ratios, distinguish two formation
trends of formation water (Walter et al., 1990; Kesler
et al., 1996). Br
ⳮ
behaves conservatively during evap-
oration of seawater as it is sparingly incorporated into
halite (Stoessel and Carpenter, 1986), whereas halite
dissolution increases both Na
Ⳮ
and Cl
ⳮ
by equimolar
amounts relative to Br
ⳮ
concentration. Therefore,
samples below the seawater composition are formed
by evaporation of seawater; values above the seawater
composition are formed by halite dissolution (Figure
9). The molar ratios of Cl/Br to Na/Br for most Activo
Luna samples plot below the seawater value, indicating
that residual evaporated seawater is left in the basin
(Figure 9). In contrast, the diluted Piije-12, Tizo´n-1,
and Pijije-21 fossil meteoric waters are not influenced
by evaporation of seawater nor by dissolution of halite.
Membrane Effect
Shales may act as membranes to concentrate chloride,
along with other ions such as calcium (Bredehoeft et
al., 1963). This concept is not supported by data from
the Gulf Coast (Land and Macpherson, 1992), where
natural pressure gradients probably are insufficient to
accomplish filtration (Manheim and Horn, 1968). Us-
ing the Rb-Li-Ca-Na elemental distribution, the Rb /Li
ratio should increase drastically along with the Ca/Na
ratio because of filtration (Kharaka and Smalley,
1976). Except for the diluted Pijije-21, Pijije-12, and
Tizo´n-3DL waters, all Activo Luna formation waters

Figure 10. Log Na/Ca vs. log Li/Br for the Activo Luna for-
mation waters, showing constant Na/Ca ratios and variable
Li/Rb ratios for circled samples.
have a very constant Na/Ca ratio (circled area in Figure
10), providing one argument to reject membrane fil-
tration as a significant process in controlling the salinity
and composition of the Activo Luna waters.
Stable Isotope Indications (dD/d
18
O): Water-Rock Interaction
and Evaporation
In general, the 15 formation water samples are char-
acterized by very positive d
18
Ovalues, whereas the 9
surface samples are located close to the global meteoric
water line (Tables 1, 2; Figure 11). In general, evapo-
rating brines in coastal regions have a maximal d
18
O
value of Ⳮ6‰ because of back exchange of the brine
with atmospheric water vapor (Lloyd, 1966), also me-
teoric waters in evaporating pools of extremely arid
regions can exceed this value (Fontes and Gonfiantini,
1970). Humid climatic conditions of the Gulf region
during the Pleistocene–Holocene exclude evaporation
as the principal process for enriched maximal d
18
Oval-
ues. Especially values up to Ⳮ12.5‰ indicate the ex-
istence of strong interaction processes between the flu-
ids and the host rocks (Figure 11). Three groups can
be distinguished based on the evolution of the individ-
ual reservoir fluids.
(1) The positive trend of both isotope ratios for
the wells Tizo´n-1 and Luna-2, together with a very low
salinity, indicates the evaporation of meteoric water
before its infiltration into the reservoir. Secondary fil-
tering of the evaporated water by shales, which act as
membrane ultrafilters (Hanshaw and Zen, 1965;
White, 1965), could explain the extremely low salinity
of both samples. Shales occur in the Tertiary rocks
overlying Cretaceous–Jurassic units from the Activo
Luna reservoir (e.g., J. J. Rosillo, 1999, personal com-
munication). Continuous membranes, however, prob-
ably could not exist in the extensively faulted Gulf
Coast section where physical compaction, chemical
compaction (for example redistribution of CaCO
3
),
and salt flowage are active processes (Land and Mac-
pherson, 1992). Fresh water, however, infiltrated di-
Figure 11. Relation of dDand d
18
O: various alteration processes cause a heterogeneous evolution of the formation waters.

Birkle et al. 471
rectly into the subsoil of the wells Pijije-21 and Pijije-
12, and these areas experienced moderate (Pijije-21)
to strong (Pijije-12) water-rock interaction.
(2) The chemistry of the saline samples from
Tizo´n-3DL and Sen-121 seems to reflect a marine or-
igin; however, the combination of depletion in dDand
enrichment in d
18
Osuggests the evolution of primary
meteoric water by secondary water-rock interaction
processes.
(3) The extreme evaporation of seawater—
described as superevaporation (Gonfiantini, 1965)—
before its infiltration into the subsurface causes a typ-
ical regressive tendency line in the evolution of the
oxygen and hydrogen isotope composition. Evapora-
tion of about 80% of the total seawater volume results
in decreasing d
18
Oand dDvalues. The reversal occurs
as a result of the decreased activity of water, which
reduces the humidity contrast between the boundary
layer and the open atmosphere. As a second option,
clay-water reaction effects could also cause a dDde-
pletion relative to seawater (Yeh, 1980), but this does
not explain the salinity of the brines. Superevaporation
applies for the hypersaline brines from Luna-3B,
Caparroso-85, and Escuintle-1, and partially for Pijije-
41, Pijije-1A, Caparroso-35, Caparroso-192, Sen-65,
and Escuintle-2. The Br/Cl ratio of the water samples
(Figure 7) indicates the absence of halite dissolution as
asecondary process. Additionally, the elevated Br/Cl
ratio of the Pijije-41, Pijije-1A, Caparroso-35, Capar-
roso-192, Sen-65, and Escuintle-2 brines indicates the
dilution of the evaporated brines with meteoric water,
which can be due to secondary influx of freshwater into
the isolated lagoon.
The elevated d
18
Ovalues indicate the presence of
considerable secondary water-rock processes in the Ac-
tivo Luna reservoir. A positive oxygen shift from the
initial meteoric water composition toward the final for-
mation water composition, correlated with a constant
hydrogen isotope value, indicates the effect of water-
rock interaction processes. Very positive d
18
Ovalues
reflect strong water-rock interaction processes with full
equilibrium conditions between the fluid and solid me-
dium (samples Sen-65, Caparroso-192, Caparoso-35,
Escuintle-2), whereas wells Sen-121 and Pijije-21
show moderate interaction processes. The most prob-
able process for the enrichment of the formation wa-
ters in
18
Oisthe exchange with
18
O-rich solids in the
sedimentary pile, especially with calcite. In general, the
origin of heavy formation waters (d
18
O⳱Ⳮ10 to
Ⳮ11‰) in the Illinois, Michigan, and Alberta basins
and the Gulf Coast is attributed to the exchange with
calcites enriched in
18
O(d
18
O⳱Ⳮ25 to Ⳮ32‰)
(Clayton et al., 1966).
The calculation of the initial isotopic composition
of the infiltrating fossil meteoric water component is
difficult to determine because of the homogenization
of the
18
Oisotopic composition of the Activo Luna
formation waters during secondary water-rock inter-
action processes. The horizontal trend lines in Figure
11 originate from a recent meteoric water composi-
tion, whereas in reality, colder climatic conditions
during glacial periods should imply more negative
isotopic values for the initial fossil meteoric water
component.
Stable Isotope Indications (d
13
C): Water-Rock Interaction and
Biogenetic Influence
The stable isotope
13
Cisanexcellent tracer for the
reconstruction of the evolution of carbonates in aqui-
fers because of its considerable natural variations in dif-
ferent carbon reservoirs. Figure 12 shows typical d
13
C
ranges in the natural environment: Atmospheric CO
2
is characterized by average d
13
Cvalues of ⳮ7‰,
whereas biogenic processes, such as respiration and the
cultivation of compost from plants in the unsaturated
soil zone, cause the formation of soil CO
2
,with typical
soil values of ⳮ20 to ⳮ30‰ in humid areas and ⳮ25
to ⳮ7‰ in semiarid to arid regions (Pearson and Han-
shaw, 1970; Mann, 1983). Analyzed d
13
Cvalues be-
tween ⳮ10.3 (River de la Pigua) and ⳮ19.3‰ (shal-
low Pijije aquifer) for rivers and lagoons in the Activo
Luna area confirm the atmospheric and biogenetic
character of the surface samples (Table 2). Carbonates
of sedimentary basins show typical values of Ⳳ0‰,
and crude oil, similar to coal, ranges from ⳮ34 to
ⳮ18‰ (Stahl, 1979). d
13
Cvalues of magmatic CO
2
gases range between ⳮ5and ⳮ8‰ (Taylor, 1986).
The generation of methane by bacteria results in a vari-
ation of d
13
Cbetween ⳮ50 and ⳮ93‰ for biogenic
methane (Schoell, 1980; Grossman et al., 1989; Ar-
avena et al., 1995), whereas magmatic methane ranges
between ⳮ52 and ⳮ12‰ (Fritz et al., 1987; Sher-
wood Lollar et al., 1993), with values generally above
ⳮ40‰ (Clark and Fritz, 1997). Thermocatalytic
methane ranges between ⳮ25 and ⳮ38‰ (Barker and
Fritz, 1981).
With d
13
Cvalues from ⳮ3.7 to Ⳮ20.4‰, the Ac-
tivo Luna formation waters are characterized by a wide
range of isotopic composition (Figure 12). Water sam-
ples with d
13
Cvalues between ⳮ3.7 and Ⳮ3.0‰ are
probably formed by secondary interaction processes
between meteoric water and the Jurassic–Cretaceous

472 Activo Luna Oil Field (Gulf of Mexico, Mexico)
limestone, which causes an isotopic homogenization of
both phases. In this case, d
13
Cvalues of infiltrating me-
teoric water become diluted by the dissolved inorganic
carbon from the source carbonate minerals (Clark and
Fritz, 1997). The elevated positive values (between
Ⳮ3.0 and Ⳮ20.4‰) of wells Sen-65, Caparroso-35,
Luna-3B, Luna-2, Tizo´n-1, Tizo´n-3DL, Caparroso-
192, Pijije-1A, and Caparosso-85 (in increasing order)
are, however, very uncommon in natural environ-
ments. These elevated values indicate the progressive
conversion of CO
2
to methane (as part of a closed or
CO
2
-limited system), as shown in the Upper Devonian
Antrim shales, Michigan basin (Martini et al., 1996,
1998), where the d
13
Cofresidual CO
2
becomes in-
creasingly more positive. Another documented expla-
nation for the formation of extremely positive d
13
C
values is the bacterial reduction of CO
2
to CH
4
in
brackish groundwater of the St. Lawrence Lowlands
dolomites (Clark and Fritz, 1997), which causes an en-
richment of
13
Cinthe waters. Chloride concentrations
of 3 molal or greater inhibit many microbial commu-
nities, whereas those greater than 4 molal practically
extinguish methanogens (Zinder, 1993). Therefore,
the occurrence of the microbial methanogenesis pro-
cesses is probable for the low-chloride (3molal) sam-
ples Luna-2, Tizo´n-1, Tizo´n-3DL, Sen-65, and Capar-
roso-192, restricted for Caparroso-35 and Pijije-1A
(between 3 and 4 molal), and inhibited for Caparroso-
85 and Luna-3B (4molal). The positive d
13
Cvalues
for the latter sites must be explained by thermo-
catalytic methanogenesis.
Regional or Local Aquifer Systems?
Hydrological Communication
Variations in the chemical and isotopic composition of
groundwater samples within a defined area can be re-
lated (1) to the geochemical evolution of the water
during flow within a single aquifer, or (2) to their af-
filiation to different aquifer systems with variations in
their permeability and porosity characteristics.
Figure 13 shows the correlation between the chlo-
ride concentration of formation waters and their
depths. Waters from the same geological unit but from
different depths, such as the samples Pijije-12, Pijije-
21, Sen-121, Sen-65, Caparroso-192, Caparroso-85,
and Escuintle-1 from the Upper Cretaceous forma-
tions, range significantly in their chemical composition.
The chloride concentration ranges from 330 (well
Pijije-21) to 168,500 mg/L (well Caparroso-85). The
range in salinity between the individual samples indi-
cates the independence of the flow regime from the
host rock and the preferential migration of the fluids
by open fractures and fault systems. The homogeneous
lithological composition of the reservoir—Cretaceous
wackestones and packstones and Upper Jurassic dolo-
mites—does not cause variations in the chemical com-
position of the formation waters.
Figure 12. Comparison of
typical d
13
Cranges in the envi-
ronment of the Activo Luna for-
mation water.

Birkle et al. 473
Figure 13. Correlation be-
tween chloride concentration,
depth, and the host rock of the
Activo Luna formation waters.
Lateral Communication
Isotopic and Chemical Variety of the Formation Waters
In the theoretical case of the existence of local aquifers
with lack of communication between the individual
wells, a lateral discontinuity of the chemical and iso-
topic composition should be observed. The distribu-
tion map of the d
18
Ovalues shows minimum values in
camp Sen in the southwest part of the oil field (5.2‰),
increasing toward the northeast with a homogeneous
composition (11.8 to 12.5‰) in the Escuintle and Ca-
parroso fields (Figure 14). Toward the north-north-
west, a natural barrier seems to separate the adjacent
Pijije camp from Escuintle and Caparroso, showing val-
ues from 11.9 to 6.2‰. In the northeast part of the
Activo Luna, the Luna and Tizo´n camps are very ho-
mogeneous in their stable oxygen isotope composition
(10.2 to 10.8‰).
Asimilar tendency was observed for the chloride
values: two maximum peaks within an irregular distri-
bution map are located in the Escuintle-Caparroso
(maximum value ⳱173 g/L) and Luna fields (maxi-
mum value ⳱180 g/L). It can be concluded that no
clear trend can be observed for the lateral distribution
of a regional aquifer at the Activo Luna oil field.
Interference Pressure Test
As an experiment to study the hydraulic connection
between the production wells, Activo Luna oil field
engineers performed an interference pressure test
at the Caparroso camp between wells Caparroso-5,
Caparroso-15, Caparroso-15/RM, and Caparroso-35.
As a result, hydraulic communication was detected in
asouthwest-northeast direction, whereas no contact
was observed from northwest to southeast (L. A. Pa-
vo´n, 1998, unpublished data).
Pressure Drawdown Data
The bottom pressure of the production wells has been
measured since the beginning in the Activo Luna oil
field to predict the pressure behavior for future petro-
leum exploitation. Figure 15 shows pressure data from
the individual Sen, Escarbado, Escuintle-Caparroso-
Pijije, Luna, and Tizo´n camps from 1988 to 2000. To
compare the single camps, the measured pressure
value is corrected to a reference depth of 5200 mbsl.

474 Activo Luna Oil Field (Gulf of Mexico, Mexico)
Figure 14. Lateral distribution
of d
18
Ovalues in the Activo
Luna oil field.
Figure 15. Pressure-drawdown curves from individual camps of the Activo Luna oil field, related to petroleum extraction.

Birkle et al. 475
Between 1983 and 1998, the pressure values of the
Sen, Escarbado, and Escuintle-Caparroso-Pijije camps
declined simultaneously from 820 to 580 kg/cm
2
,
indicating hydraulic communication among these
camps. Separated pressure-drawdown curves for the
Luna and Tizo´n camps, however, reflect the local
hydraulic character of each camp, and the lack of com-
munication between each of the reservoirs of Sen-
Escarbado-Escuintle-Caparroso-Pijije, Tizo´n, and Luna,
respectively.
Both the composition of the formation waters and
pressure field data of the reservoir indicate the exis-
tence of a southwest-northeast–directed flow regime of
the deep aquifer between the Sen, Escarbado, Escuin-
tle, Caparroso, and Pijije camps. Pressure data, how-
ever, indicates the lack of hydraulic communication
between the three blocks Sen-Escarbado-Escuintle-
Caparroso-Pijije, Luna, and Tizo´n, which is probably
related to a northwest-southeast–directed barrier of
thrust faults, as shown in Figure 2.
Vertical Communication
Avertical tendency in the chemical composition of
deep waters indicates the occurrence of one or more
of the following processes and aquifer characteristics:
1. Existence of vertical communication between dif-
ferent levels of one single aquifer
2. Gravity-influenced separation of fluids with varia-
tions in salinity (and therefore density)
3. Occurrence of recent or fossil infiltration processes
4. Temperature-related increase in velocity of the
thermodynamic reactions
In the case of the Activo Luna wells, no linear corre-
lation was found between chloride and depth on a re-
gional scale (Figure 13). Some of the shallowest wells,
such as Caparroso-85 (4722–4731 m) and Escuintle-1
(4885–4915 m) have the highest Cl
ⳮ
concentrations,
which contradicts the recent influence of infiltrating
meteoric water from the surface. At the same time, the
deepest wells (Tizo´n-1: 5978–5991 m; Tizo´n-3DL:
6114–6128 m) have very low Cl
ⳮ
concentrations. A
linear correlation, however, can be observed on a local
scale: the wells of each individual camp, such as Sen,
Tizo´n, Luna, and Pijije, show increasing salt contents
with depth. For example, wells Pijije-21 (5300–5321
m), Pijije-12 (5316–5347 m), Pijije-41 (5365–5390
m), and Pijije-1A (5449–5465 m) show an increasing
chloride content from 330 (Pijije-21) and 2680 mg /L
(Pijije-12) toward 124,900 (Pijije-1A) and 134,700
mg/L (Pijije-41), which indicates (1) the existence of
vertical flow pathways and local communication
among adjacent wells or (2) a vertical zonation of the
fluids by differences in density. The level of the division
line between lower (Pijije-12, Pijije-21) and higher sa-
line (Pijije-1A, Pijije-41, Caparroso-35) formation wa-
ter is shown in the profile in Figure 16. The same effect
can be observed for the vertical distribution of the
d
18
Ovalues.
The concentration of dissolved silica in discharging
water from deep wells has been used as a criterion to
identify physical processes in production zones of res-
ervoirs (Land and Macpherson, 1992; Truesdell et al.,
1995). This application is based on the temperature-
related solubility of silica and the speed of reequilibra-
tion during dilution or ebullition processes in the res-
ervoir. Samples with identical reservoir (T
reservoir
)and
silica (T
SiO2
)temperatures, reflected by the straight
line in Figure 17, are under equilibrium conditions.
Comparing the measured reservoir temperatures from
the Activo Luna reservoir with the temperature of dis-
solved silica (quartz) (Fournier, 1991) shows that most
formation waters are under equilibrium conditions,
whereas E-1, L-3B, and C-85 are about 40Cbelow
the equilibrium line; however, the strong deviation of
P-21, P-12, L-2, and S-121 formation waters from the
equilibrium line indicates dilution by condensed vapor
at the reservoir (Figure 17). Therefore, vertical rising
of condensed vapor could explain the formation of
some low-saline waters in the Activo Luna reservoir.
Overpressure conditions at structural highs and
depressed structural lows can provide dynamic flow
processes, that is, between sand and shale layers, as
shown in the offshore Gulf of Mexico (Stump and
Flemings, 2000). Stratigraphic layering focuses fluid
flow toward structural highs. An early Laramide flex-
ural bulge caused hydrofracturing and subsequent kar-
stification at the Veracruz basin (Ferket et al., 2000),
which is adjacent to the Villahermosa uplift. The latter
process can be explained by the infiltration of meteoric
water. Pleistocene recharge processes in northern
Michigan are explained by repeated advances and re-
treats of ice sheets causing the opening of preexisting
fractures within the Antrim Shale (Martini et al.,
1998). Overpressure conditions during the Laramide
orogeny caused horizontal fluid flow toward the fore-
lands (Ferket et al., 2000). Similar horizontal expulsion
processes during the Laramide orogeny with subse-
quent vertical infiltration of meteoric water to a depth
of more than 2000 m is documented for Devonian car-
bonate aquifers in the Western Canada sedimentary

476 Activo Luna Oil Field (Gulf of Mexico, Mexico)
Figure 16. Conceptual geological-structural model of the Pijije-Escuintle and Caparroso camps, including information on related formation waters (% SW
ET
⳱mixing percentage
of evaporated seawater component; % MW ⳱mixing percentage of meteoric water component). The migration of the deep waters along individual faults and tectonics is further
discussed in the text (location of section BB⬘is shown in Figure 2; stratigraphic abbreviations are explained in Figure 3 caption).

Birkle et al. 477
Figure 17. Comparison of the measured reservoir tempera-
tures (T
reservoir
)from the Activo Luna reservoir with tempera-
tures of dissolved silica (quartz) (T
SiO2
).
basin (Machel et al., 2000). In the case of the Canadian
shield, infiltration of dense evaporitic brine is explained
by a density gradient associated with brine production
(seawater evaporation at the surface), which may have
enabled brine to infiltrate to depths of several kilo-
meters (Spencer, 1987). Sedimentary column experi-
ments showed the upward displacement of consolida-
tional fluid flow by compaction (Bitzer et al., 2000).
Another driving mechanism of geopressured zones is
the vertical migration of heated brines from hot sub-
stances by fractures into carbonate host rocks (Fowler,
1994). Also, alpine deformation processes with over-
pressured systems provided conduits for upflowing flu-
ids in Middle Jurassic sandstones of Egypt’s Western
Desert during the latest Eocene (Rossi et al., 2000). At
the Gulf of Mexico continental slope, recent migration
of gaseous and liquid hydrocarbon from deep subsur-
face (2000–3000 m) reservoirs into near-surface sedi-
ments and to the seawater surface has been described
by Kennicutt et al. (1988). The expulsion of the pa-
leofluids and the infiltration of meteoric water at the
Activo Luna reservoir can be explained by similar
processes.
Water Types and Recharge Processes
The correlation of radioactive isotope data with con-
servative tracers is useful for the determination of mix-
ing processes during different periods of water evolu-
tion. The oldest formation waters of the Activo Luna
reservoir (20 ka;
14
C4pMC) are characterized by
avery heterogeneous distribution of the conservative
tracer chloride (group B in Figure 18). As shown in the
Br vs. Cl, and Br vs. TDS distribution (Figures 7, 8),
this group is characterized by the mixture of two end
members: extremely evaporated seawater (C-35, L-
3B, E-1) and infiltrated, fossil meteoric water (Luna-2,
Tizo´n-3DL), which explains the wide chloride range.
Intermediate compositions, such as in Escuintle-2, Ca-
parroso-192, Caparroso-35, Pijije-1A, and Pijije-41,
reflect mixing processes between both components.
Subsequently, formation waters with
14
Cactivities
between 8 and 14 pMC (group A) indicate a second
infiltration event of low-saline, meteoric water be-
tween 15 and 10 ka in the Tizo´n (Tizo´n-1) and Pijije
areas (Pijije-12) (according to age calculations with
model 2, as shown in a previous section). At the same
time, the infiltration of primary (Sen-121) and slightly
evaporated (Sen-65) marine water is proposed for both
sampled wells in the Sen area, reflected by Cl
ⳮ
con-
centrations of 16,100 and 69,500 mg/L, respectively.
The influence of halite dissolution on the chloride en-
richment can be excluded (see also the previous section
“Evaporation Effect”). The formation water from well
Pijije-21 (P-21) is the only sampled site with a
14
Cage
between that of group A and that of group B.
Similar to chloride, the trace elements lithium, ru-
bidium, manganese, molybdenum, strontium (Figure
19), and boron (Figure 20) are unaffected by solubility
controls to upper limit concentrations and can be used
as mixing process indicators. As shown for chloride,
the linear trend of the formation waters in Sr and B vs.
14
C, respectively, indicate mixing processes between
two late Pleistocene components (mixing line in Fig-
ures 19, 20)—evaporated seawater (SW
ET
)and me-
teoric water (MW)—represented by the two end mem-
bers Luna-3B–Escuintle-1–Caparosso-85 and Luna-2,
respectively. The depletion of the younger group A
(15 ka) in strontium and boron, however, reflects the
infiltration of exclusively meteoric water and marine
water between 15 and 10 ka.
Local Hydrogeological Model
Camp Escuintle-Caparroso-Pijije
Compressive thrust sheets in a northwest-southeast
direction represent the dominating tectonic structure
of the Escuintle-Caparroso-Pijije zone (Figure 2).
Younger, extensive systems in the form of normal

478 Activo Luna Oil Field (Gulf of Mexico, Mexico)
Figure 18. Chloride vs.
14
C
concentrations: a younger
(10–15 ka), less-saline water
type (group A) can be distin-
guished from an older
(20 ka) and chemically vari-
able water type (group B).
faults, and horst and graben structures divide the
northwest-southeast–trending anticlines in isolated
blocks, partially caused by updoming salt structures, as
shown in the profile from Figure 16. In the northern
part of the saline structure, the wells Pijije-1A (P-1A),
Caparroso-35 (C-35), and Pijije-41 (P-41) are charac-
terized by a mixing water type between evaporated
seawater (60–80%) and meteoric water (20–40%),
with
14
Cactivities below 4 pMC (group B from Figure
18), whereas Pijije-12 and Pijije-21 represent younger
(
14
C4pMC), low-saline meteoric water (group A
from Figure 18). The abundance of heavy group B
brines in the deeper part of the horst structures (Lower
and middle Cretaceous), and meteoric water in the up-
per part (Upper Cretaceous) (see division line in Figure
16) can be explained by vertical, density-driven gravity
processes. Besides, the normal fault (no. 1 in Figure 16)
facilitates the hydraulic continuity between the fault
blocks of C-35 and P-41, whereas the reverse fault be-
tween P-12 and P-21 (no. 2 in Figure 16) assures the
migration of meteoric water between the mentioned
wells.
On the southern side of the salt structure, the sam-
pled Caparroso and Escuintle formation waters are all
group B brines (
14
C4pMC). The normal faults (no.
3and no. 4) seem to form a barrier between high-saline
(Caparroso-85, Escuintle-1) and moderate-saline for-
mation waters (Caparroso-192). The continuation of
an indicated normal fault structure between Escuintle-
1and Escuintle-2 is not known.
Camp Luna
The formation water from the adjacent wells Luna-2
and Luna-3B were sampled from the Upper Jurassic
reservoir at a depth between 5229 and 5335 m and
5405 and 5426 m, respectively. Both show the same

Birkle et al. 479
Figure 19. Sr vs.
14
Cconcentrations: the illustrated linear
trend indicates mixing processes between two late Pleistocene
components—evaporated seawater (SW
ET
)and meteoric water
(MW)—represented by the two end members Luna-3B–
Escuintle-1–Caparosso-85 and Luna-2, respectively.
Figure 20. Bvs.
14
Cconcentrations: mixing processes similar
to those shown in Figure 19.
14
Ccharacteristics (group B,
14
C4pMC), but the
chemical composition ranges from freshwater (Luna-
2) to hypersaline composition (Luna-3B). No indica-
tions exist for the existence of tectonic structures be-
tween both wells, thus the chemical heterogeneity is
due to vertical separation processes by differences in
density and/or by the formation of condensed water in
the upper part of the aquifer.
Regional Hydrogeological Model
Based on the correlation of geological, structural, and
geophysical data with the recent hydrochemical and
isotopic information, a hydrogeological model for the
evolution of the Activo Luna reservoir can be con-
structed as follows (Figure 21).
(1) During the Late Jurassic and Cretaceous pe-
riods, the cover of the Mexican Gulf Coast with a pre-
Atlantic column caused the deposition of platform and
deep-sea carbonates as part of a marine environment.
Probably, the deposition of limestone in this period in-
cluded the enclosure of marine water in the form of
pore inclusions (Figure 21a).
(2) Later, probably from the Eocene to Miocene
periods, the primary formation water was expelled
from the carbonates by the following processes (Figure
21b):
•The compaction of the stratigraphic column by the
deposition of overlying Tertiary sediments caused a
space reduction of the sediment pores.
•Compressive folding during the Oligocene produced
the required tectonic stress to initiate the flow
migration.
•Normal and thrust faulting during the Oligocene and
Miocene, respectively, resulted in the formation of
pathways for the vertical circulation, which facili-
tated the migration of the primary fluids by fractures
and microfractures.
(3) As part of two infiltration periods during gla-
ciation from the late Pleistocene to early Holocene
(40–20 ka and 15–10 ka), meteoric and marine water
infiltrated from the surface into the petroleum reser-
voir to a depth of 5000 to 6000 m. During this period,
the hydrodynamic system of a regional aquifer covered
the Mexican Gulf Coast (Figure 21c). Remnants of the
fossil formation water were probably replaced by sec-
ondary water intrusions. Depending on the site, surface
water experienced different physical processes prior to
infiltration. The brines, encountered in wells Luna-3B,
Caparroso-85, and Escuintle-1, are of original marine
type, which experienced strong evaporation processes
at the surface. This is probably related to their location
in coastal, shallow-water areas or as part of isolated
saline lagoons. This process caused the evolution of
hypersaline brines before their infiltration into the

480 Activo Luna Oil Field (Gulf of Mexico, Mexico)
Figure 21. Hydrogeological model for the evolution of the
Activo Luna reservoir. (a) Deposition of carbonate sediments
with trapped marine water. (b) Migration of formation water.
(c) Infiltration of surface water. (d) Chemical and physical pro-
cesses: 1 ⳱water-rock interaction; 2 ⳱methane genesis; 3 ⳱
vertical zonation.
subsoil. Sen-121 is probably of primary marine origin.
Wells with low-salinity waters, such as Tizo´n-3DL,
Tizo´n-1, Pijije-12, Pijije-21, and Luna-2, were formed
by the direct infiltration of meteoric water into the res-
ervoir, although the water samples from Tizo´n-1 and
Luna-2 seem to be slightly affected by evaporation.
The formation waters from Sen-65, Escuintle-2,
Caparroso-192, Caparroso-35, Pijije-1A, and Pijije-41
represent mixed waters between the meteoric and hy-
persaline component.
(4) Since the early (?) Holocene, secondary pro-
cesses, such as the interaction of the fluids with the
limestone units (all samples), the formation of meth-
ane (Caparroso, Luna, and Tizo´ncamps), and the grav-
itative separation of the fluids by differences in density
and/or the rise of condensed vapor into upper parts of
the aquifer caused modifications in the chemical and
isotopic composition of the infiltrating surface waters.
These secondary processes, as well as the ongoing for-
mation of natural barriers by faulting from the Oligo-
cene to the Holocene, explain the partial evolution of
local aquifers, separated by tectonic structures. Halite
dissolution from salt domes/layers and the membrane
effects by shales were not observed.
(5) During the last 15 years, the exploitation of
the Activo Luna petroleum reservoir caused a pertur-
bation of the local aquifers, including the drawdown of
the hydrostatic pressure of the reservoir, the rise of the
piezometric water level of the deep aquifers, and the
invasion of water into the petroleum production wells
by coning processes. Some tectonic structures and
faults are natural barriers to flow, whereas others ap-
pear to be conduits for the aquifer migration.
COMPARISON WITH OTHER
SEDIMENTARY BASINS OF THE GULF OF
MEXICO
In contrast to the Activo Luna formation waters, the
chemistry of the bromine- and rubidium-rich saline
formation waters from the northern Gulf of Mexico
reflects interaction with the Jurassic Louann evaporites
(Land et al., 1995), which is probably due to major
thicknesses of the salts in the described section. For-
mation water from salt mines in southern Louisiana
were incorporated during diapiric rise of the salt at a
depth of 3–4 km and have been trapped within the salt
for 10–13 m.y. (Knauth et al., 1980). Most waters from
the Cenozoic oil and gas reservoir in the northern Gulf
of Mexico sedimentary basins are derived from de-
watering of the clastic sediments themselves or enter
from the underlying Mesozoic strata (Land and Mac-
pherson, 1992). Three types of water can be distin-
guished, partially on a very local scale: dilute sodium
acetate water (typical of shale-rich sections, with origi-
nal seawater composition modified by microbial reac-
tions), sodium chloride–rich water (formed by disso-
lution of halite)—the most abundant type—and
calcium-rich water (albitization of plagioclase and by
injection from underlying Mesozoic strata) (Land and
Macpherson, 1992). This heterogeneity is most easily
explained by active fluid flow of an allochthonous sys-
tem and mixing. The latter process characterizes al-
most every field studied to date (Land et al., 1988).
Stratigraphic-tectonic and chemical arguments ex-

Birkle et al. 481
clude the influence of shale membrane filtration (re-
verse osmosis) on the composition and salinity of Gulf
Coast brines (Land and Macpherson, 1992). Changes
in in-situ density provide an important mechanism for
density-driven water circulation (Ranganathan and
Hanor, 1989). Deeper samples at the northern Gulf of
Mexico are characterized by positive d
18
Ovalues con-
sistent with extensive mineral-water reactions (Su-
checki and Land, 1983). The most probable process
for the enrichment of the formation waters with
18
O
is the exchange with
18
O-rich solids in the sedimentary
pile, especially with calcite. The origin of formation
waters in the Illinois, Michigan, and Alberta basins and
in the Gulf Coast is attributed to the exchange with
18
O-rich solids in the sedimentary pile, especially with
calcite (Clayton et al., 1966).
Recharge processes during the Pleistocene are con-
firmed for different sedimentary basins: a significant
influx of late Pleistocene meteoric water is shown for
the eastern flank of the Williston basin, which mixed
with formation water (Grasby and Betcher, 2000).
Both the Activo Luna formation waters and the
northern Gulf of Mexico deep waters show a strong
variation in their chemical composition and are af-
fected by density-driven separation processes, whereas
membrane filtration is not proven. Both regions are
characterized by strong exchange processes of the for-
mation waters with
18
O-enriched calcite.
DISCUSSION
The extreme discrepancies in the chemical and isotopic
composition of the Activo Luna reservoir waters seem
to indicate different origins for the individual samples.
At first glance, hypersaline brines seem to have been
fossil-water formed during the deposition of the host
formation, whereas low-salinity waters are similar to
surface water, such as the fluids used for drilling. In
fact, secondary processes, such as the formation of
methane, camouflage the homogenous primary origin
of the Activo Luna waters, which infiltrated into the
Jurassic and Cretaceous host rocks during the late
Pleistocene–early Holocene. Possibly, the fossil re-
charge event was related to major periods of increased
humidity, such as observed for northern Mexico at the
end of the late Pleistocene (18–11 ka) and early Ho-
locene (11–8.9 ka) (Ortega-Ramı´rez et al., 1998). Both
the marine transgression of the Mexican Gulf and the
consequent cover of the coastal area, as well as in-
creased precipitation rates, could accelerate the infil-
tration rate of surface water. Further isotopic studies
in adjacent petroleum fields of the Mexican Gulf Coast
could reconfirm the indicated large-scale climatic and
hydrologic changes during the late Pleistocene–early
Holocene.
Activo Luna faults seem to be ambiguous in their
hydraulic behavior: Regional, northwest-southeast–
trending thrust sheets (reverse faults) and fractures
communicate the Sen reservoir with the Escuintle-
Caparroso-Pijije camps (see interpreted flow direction
in Figure 2); however, the same type of faults restricts
flow migration from Escuintle-Caparroso-Pijije to
Luna and Tizo´n.Similar, younger normal faults, which
cut the reverse system in a southwest-northeast direc-
tion, have either blocking or conduit functions. On a
regional scale, the lateral communication between the
aquifers of the individual fields of the Activo Luna oil
field is probably inhibited by compressive tectonic
structures, such as thrust faults, which cause the for-
mation of isolated, trapped aquifers. On a local scale,
vertical normal faults and fractures facilitate the de-
scent of concentrated brine water. In general, fractures,
parallel and perpendicular to the reverse faults with
opening fractures between 0.0001 and more than 0.2
cm (Rosillo, 1998), are probably the most effective
transport medium for flow migration. Vertical migra-
tion of the fluids, observed as a coning effect around
production wells and the abundance of hypersaline
brines in the deeper parts of the reservoir seems to
predominate over migration by horizontal conduits.
Because of the overall distribution of formation
water at the Activo Luna field, future decisions about
new drilling sites should be concerned with the local
pattern of vertical fault systems, especially the struc-
tural correlation with adjacent, water-invaded wells.
Because the Activo Luna aquifer system is separated
into isolated local systems, the evaluation of new ex-
ploitation sites should be carried out on a local
scale.
CONCLUSIONS
The primary origin of the deep waters of the Activo
Luna petroleum reservoir, located at a depth of 5000–
6000 m in the Mexican Gulf Coast, is related to the
regional infiltration of meteoric and marine water dur-
ing the late Pleistocene and/or early Holocene (40–10
ka) as part of a hydrodynamic flow system. Before its
infiltration, major parts of the surface waters were af-
fected by atmospheric evaporation processes. Fossil

482 Activo Luna Oil Field (Gulf of Mexico, Mexico)
waters, related to the deposition of the Jurassic–
Cretaceous host rocks, were probably replaced through
younger recharge events. Secondary alteration pro-
cesses, such as the interaction between the fluids and
carbonate host rocks, the vertical zonation within the
aquifer by the descent of concentrated brines, as well
as the formation of methane, caused individual evo-
lution of the chemical and isotopic groundwater com-
position in each camp. In general, a southwest-north-
east–directed migration of the deep aquifer is
postulated for some local sections of the oil field. Re-
cently, the exploitation of petroleum caused a remo-
bilization of the deep groundwater flows. Especially on
alocal scale, adjacent wells are likely to become con-
nected by vertical extensional structures.
REFERENCES CITED
Aravena, R., L. I. Wassenaar, and L. N. Plummer, 1995, Estimating
14
Cgroundwater ages in a methanogenic aquifer: Water Re-
sources Research, v. 31, p. 2307–2317.
Barker, J. F., and P. Fritz, 1981, The occurrence and origin of meth-
ane in some groundwater flow systems: Canadian Journal of
Earth Sciences, v. 18, p. 1802–1816.
Bethke, C., 1994, The geochemist’s workbench: a users guide toRxn,
Act2, Tact and Gtplot: Urbana-Champaign, University of Illi-
nois Hydrogeology Program, 212 p.
Bitzer, K., J. Salas, and C. Ayora, 2000, Fluid pressure, flow velocities
and transport processes in a consolidating sedimentary column
with transient hydraulic properties: Journal of GeochemicalEx-
ploration, v. 69–70, p. 127–131.
Bredehoeft, J. D., C. R. Blyth, W. A. White, and G. B. Maxey,1963,
Possible mechanism for concentration of brines in subsurface
formations: AAPG Bulletin, v. 47, p. 257–269.
Carpenter, A. B., 1978, Origin and chemical evolution of brines in
sedimentary basins: Oklahoma Geological Circular, v. 79,
p. 60–77.
Chave, K. E., 1960, Evidence on history of sea water from chemistry
of deeper surface waters of ancient basins: AAPG Bulletin,
v. 44, p. 357–370.
Clark, I., and P. Fritz, 1997, Environmental isotopes in hydrogeology:
Boca Raton, Florida, Lewis Publishers, 348 p.
Clayton, N. R., I. Friedman, D. L. Graf, T. K. Mayeda, W. F. Meents,
and M. F. Shimp, 1966, The origin of formation waters—I.iso-
topic composition: Journal of Geophysical Research, v. 71,
p. 3869–3882.
Connolly, C. A., L. M. Walter, H. Baadsgaard, and F. J. Longstaffe,
1990, Origin and evolution of formation waters, Alberta basins:
Western Canada sedimentary basin: I. chemistry: Applied Geo-
chemistry, v. 5, p. 375–395.
Davis, S. N., 1964, The chemistry of saline waters by R. A. Krieger—
discussion: Ground Water, v. 2, no. 1, p. 51.
Degens, E. T., J. M. Hunt, J. H. Reuter, and W. E. Reed, 1964,Data
on the distribution of amino acids and oxygen isotopes in pe-
troleum brine waters of various geological ages: Sedimentology,
v. 3, p. 199–255.
Epstein, S., and T. Mayeda, 1953, Variation of O-18 content of wa-
ters from natural sources: Geochimica et Cosmochimica Acta,
v. 4, p. 89–103.
Fairchild, L. H., and T. H. Nelson, 1989, Emplacement and evolu-
tion of salt sills (abs.): Houston Geological Society Bulletin,
v. 32, no. 1, p. 6–7.
Ferket, H., F. Roure, R. Swennen, and S. Ortun˜o, 2000, Fluid mi-
gration placed into the deformation history of fold-and-thrust
belts: an example from the Veracruz basin (Mexico): Journal
of Geochemical Exploration, v. 69–70, p. 275–279.
Fontes, J. C., and R. Gonfiantini, 1970, Comportement isotopique
au cours de l’e´vaporation de deux bassins Sahariens: Earth and
Planetary Science Letters, v. 3, p. 258–266.
Fournier O. R., 1991, Water geothermometer applied to geothermal
energy, in F. D’Amore, ed., Applications of geochemistry in
geothermal reservoir development: Rome, Italy, UNITAR
(United Nations Institute for Training and Research)/UNDP
Centre on Small Energy Resources, p. 37–69.
Fowler, A. D., 1994, The role of geopressure zones in the formation
of hydrothermal Pb-Zn Mississippi Valley type mineralization
in sedimentary basins, in J. Parnell, ed., Geofluids: origin, mi-
gration and evolution of fluids in sedimentary basins: Geological
Society Special Publication 78, p. 293–300.
Fritz, P., S. K. Frape, and M. Miles, 1987, Methane in the crystalline
rocks of the Canadian shield, in P. Fritz and S. K. Frape, eds.,
Saline water and gases in crystalline rocks: Geological Associ-
ation of Canada Special Paper 33, p. 211–223.
Geyh, M. A., 1980, Isotope als Tracer und Strahlungsquellen fu¨r
hydrologische Untersuchungen, in H. Moser and W. Rauert,
eds., Isotopenmethoden in der Hydrogeologie: Stuttgart, Ger-
many, Verlag Gebru¨der Borntraeger, p. 43–55.
Gonfiantini, R., 1965, Effeti isotopici nell’evaporazione di acque sal-
ate: Atti Soc. Toscana Sci. Nat. Pisa, Ser A, v. 72, p. 550–569.
Grasby, S. E., and R. Betcher, 2000, Pleistocene recharge and flow
reversal in the Williston basin, central North America: Journal
of Geochemical Exploration, v. 69–70, p. 403–407.
Grossman, E. L., B. K. Coffman, S. J. Fritz, and H. Wada, 1989,
Bacterial production of methane and its influence on ground-
water chemistry in east-central Texas aquifers: Geology, v. 17,
p. 495–499.
Hanshaw, B. B., and E. Zen, 1965, Osmotic equilibrium and over-
thrust faulting: Geological Society of America Bulletin, v. 76,
p. 1379–1386.
Herrmann, A. G., and D. Knake, 1973, Geochemistry of modern sea
water and brines from salt pan: main components and bromine
distribution: Contributions to Mineralogy and Petrology, v. 40,
p. 1–24.
Hitchon, B., 1996, Rapid evaluation of the hydrochemistry of a sed-
imentary basin using only ‘standard’ formation water analyses:
example from the Canadian portion of the Williston basin: Ap-
plied Geochemistry, v. 11, p. 789–795.
Hitchon, B., and I. Friedman, 1969, Geochemistry and origin offor-
mation waters in the Western Canada sedimentary basin—I.
stable isotopes of hydrogen and oxygen: Geochimica et Cos-
mochimica Acta, v. 33, p. 1321–1349.
Kennicutt II, M. G., J. M. Brooks, and G. J. Denoux, 1988, Leakage
of deep, reservoired petroleum to the near surface on the Gulf
of Mexico continental slope: Marine Chemistry, v. 24, p. 39–
59.
Kesler, S. E., A. M. Martini, M. S. Appold, L. M. Walter, T. J. Hus-
ton, and F. C. Furman, 1996, Na-Cl-Br systematics of fluid in-
clusions from Mississippi Valley-type deposits, Appalachianba-
sin: constraints on solute origin and migration paths:
Geochimica et Cosmochimica Acta, v. 33, p. 1321–1349.
Kharaka, Y. K., and W. C. Smalley, 1976, Flow of water and solutes
through compacted clays: AAPG Bulletin, v. 60, p. 973–980.
Kharaka, Y. K., A. S. Maest, W. W. Carothers, L. M. Law, P. J.
Lamothe, and T. L. Fries, 1987, Geochemistry of metal-rich
brines from central Mississippi Salt Dome basin, U.S.A.: Ap-
plied Geochemistry, v. 2, p. 543–561.

Birkle et al. 483
Knauth, P. L., 1988, Origin and mixing history of brines, of Palo
Duro Basin, Texas, USA: Applied Geochemistry, v. 3, p. 455–
474.
Knauth, P. L., and M. A. Beeunas, 1986, Isotope geochemistry of
fluid inclusions in Permian halite with implications for the iso-
topic history of ocean water and the origin of saline formation
waters: Geochimica et Cosmochimica Acta, v. 50, p. 419–434.
Knauth, P. L., M. B. Kumar, and J. D. Martinez, 1980, Isotope geo-
chemistry of water in Gulf Coast salt domes: Journal of Geo-
physical Research, v. 85, p. 4863–4871.
Land, L. S., and G. L. Macpherson, 1992, Origin of saline formation
waters, Cenozoic section, Gulf of Mexico sedimentary basin:
AAPG Bulletin, v. 76, p. 1344–1362.
Land L. S., G. L. Macpherson, and L. E. Mack, 1988, The geochem-
istry of saline formation waters, Miocene, offshore, Louisiana:
Gulf Coast Association of Geological Societies Transactions,
v. 38, p. 503–511.
Land, L. S., R. A. Eustice, L. E. Mack, and J. Horita, 1995,Reactivity
of evaporites during burial: an example from the Jurassic of
Alabama: Geochimica et Cosmochimica Acta, v. 59, p. 3765–
3778.
Lloyd, R. M., 1966, Oxygen isotope enrichment of sea water by
evaporation: Geochimica et Cosmochimica Acta, v. 30, p. 801–
814.
Machel, H. G., P. A. Cavell, B. E. Buschkuehle, and K. Michael,
2000, Tectonically induced fluid flow in Devonian carbonate
aquifers of the Western Canada sedimentary basin: Journal of
Geochemical Exploration, v. 69–70, p. 213–217.
Manheim, F. T., and M. K. Horn, 1968, Composition of deepersub-
surface waters along the Atlantic continental margin: South-
western Geology, v. 9, p. 215–236.
Mann, W. B., 1983, An international reference material for radio-
carbon dating in M. Stuiver and R. Kra, eds., Proceedings of the
11th International Radiocarbon Dating Conference: v. 25,
no. 2, p. 519–528.
Martin, R. G., and R. Q. Foote, 1981, Geology and geophysics of
the Maritime Boundary assessment areas, in R. B. Powers, ed.,
Geologic framework, petroleum potential, petroleum resource
estimates, mineral and geothermal resources, geologic hazards,
and deep-water drilling technology of the Maritime Boundary
region in the Gulf of Mexico: U.S. Geological Survey OpenFile
Report 81-265, p. 30–67.
Martini, A. M., J. M. Budai, L. M. Walter, and M. Schoell, 1996,
Microbial generation of economic accumulations of methane
within a shallow organic-rich shale: Nature, v. 383, p. 155–157.
Martini, A. M., L. M. Walter, J. M. Budai, T. C. W. Ku., C. J. Kaiser,
and M. Schoell, 1998, Genetic and temporal relations between
formation waters and biogenic methane: Upper Devonian An-
trim Shale, Michigan basin, USA: Geochimica et Cosmochim-
ica Acta, v. 62, p. 1699–1720.
Moser, H., and W. Rauert, 1980, Isotopenmethoden in der Hydro-
geologie: Berlin, Gebru¨der Borntra¨ger, 388 p.
Ortega-Ramı´rez, J. R., A. Valiente-Banuet, J. Urrutia-Fucugauchi,
C. A. Mortera-Gutie´rrez, and G. Alvarado-Valdez, 1998, Pa-
leoclimatic changes during the late Pleistocene–Holocene in La-
guna Babı´cora, near the ChihuahuanDesert, Me´xico: Canadian
Journal of Earth Sciences, v. 35, p. 1168–1179.
Peachey, B. R., S. C. Solanki, T. A. Zahacy, and K. Piers, 1998,
Downhole oil/water separation moves into high gear: Journal
of Canadian Petroleum Geology, v. 37, p. 34–41.
Pearson Jr., F. J., and B. B. Hanshaw, 1970, Sources of dissolved
carbonate species in groundwater and their effects on carbon-
14 dating, in Isotope hydrology: Proceedings Symposium, Vi-
enna, Austria, International Atomic Energy Agency, p. 271–
286.
Ranganathan, V., and J. S. Hanor, 1989, Perched brine plumesabove
salt domes and dewatering of geopressured sediments: Journal
of Hydrology, v. 110, p. 63–86.
Ricoy, J. M., 1989, Tertiary terrigenous depositional systems of the
Mexican Isthmus basin: Ph.D. thesis, University of Texas at
Austin, Austin, Texas, 145 p.
Rittenhouse, G., 1967, Bromine in oil-field waters and its use in de-
termining possibilities of origin of these waters: AAPGBulletin,
v. 51, p. 2430–2440.
Rosillo J., 1998, Caracterizacio´n de yacimientos en el Activo de
Produccio´n Luna: Gaceta de la Asociacio´n de Ingenierı´a Petro-
lera de Me´xico (AIPM), 6 p.
Rosillo, J., 2000, Caracterizacio´ndel fracturamiento natural, estudio
piloto en el Activo de Produccio´n Luna: XXXVIII Congreso
Nacional de la Asociacio´n de Ingenieros Petroleros de Me´xico
(AIPM), Memorias, 16 p.
Rossi, C., R. H. Goldstein, and R. Marfil, 2000, Pore fluid evolution
and quartz diagenesis in the Khatatba Formation, WesternDes-
ert, Egypt: Journal of Geochemical Exploration, v. 69–70,
p. 91–96.
Rubey, W. W., 1951, Geological history of sea water: Geological
Society of America Bulletin, v. 62, p. 1111–1148.
Saliege, J. F., and J. Ch. Fontes, 1984, Essai de de´termination ex-
pe´rimentale du fractionnement des isotopes
13
Cet
14
Cducar-
bone au cours de processus naturels: International Journal of
Applied Radiation and Isotopes, v. 35, p. 55–62.
Salvador, A., ed. 1991, The Gulf of Mexico Basin: Boulder, Colo-
rado, Geological Society of America, The Geology of North
America, v. J, 568 p.
Sarkar, A., J. A. Nunn, and J. S. Hanor, 1995, Free thermohaline
convection beneath allochthonous salt sheets: an agent for salt
dissolution and fluid flow in Gulf Coast sediments: Journal of
Geophysical Research, v. 100, p. 18085–18092.
Schoell, M., 1980, The hydrogen and carbon isotopic composition
of methane from natural gas of various origins: Geochimica et
Cosmochimica Acta, v. 44, p. 649–662.
Sherwood Lollar, B., S. K. Frape, S. M. Weise, P. Fritz, S. A. Macko,
and J. A. Welhan, 1993, Abiogenic methanogenesis in crystal-
line rocks: Geochimica et Cosmochimica Acta, v. 57, p. 5087–
5097.
Siegel, D. I., and R. J. Mandle, 1984, Isotopic evidence for glacial
meltwater recharge to the Cambrian–Ordovician aquifer,
north-central United States: Quaternary Research, v. 22,
p. 328–335.
Sofer, Z., and J. R. Gat, 1972, Activities and concentrations of
oxygen-18 in concentrated aqueous salt solutions: analytical
and geophysical implications: Earth and Planetary ScienceLet-
ters, v. 15, p. 232–238.
Spencer, R. J., 1987, Origin of Ca-Cl brines in Devonian formations,
Western Canada sedimentary basin: Applied Geochemistry,
v. 2, p. 373–384.
Stahl, W. J., 1979, Carbon isotopes in petroleum geochemistry, in
E. Ja¨ger and J. C. Hunziker, eds., Lectures in isotope geology:
Berlin, Springer-Verlag, p. 274–282.
Stoessel, R. K., and A. B. Carpenter, 1986, Stoichiometricsaturation
tests of NaCl
1
ⳮ
x
Br
x
and KCl
1
ⳮ
x
Br
x
:Geochimica et Cosmo-
chimica Acta, v. 50, p. 1465–1474.
Stueber, A. M., and L. M. Walter, 1994, Glacial recharge andpaleo-
hydrologic flow systems in the Illinois basin: evidence from
chemistry of Ordovician carbonate (Galena) formation waters:
Geological Society of America Bulletin, v. 106, p. 1430–1439.
Stump, B. B., and P. B. Flemings, 2000, Overpressure and fluid flow
in dipping structures of the offshore Gulf of Mexico (E.I. 330
field): Journal of Geochemical Exploration, v. 69–70, p. 23–
38.
Suchecki, R. K., and L. S. Land, 1983, Isotopic geochemistry of
burial-metamorphosed volcanogenic sediments, Great Valley

484 Activo Luna Oil Field (Gulf of Mexico, Mexico)
sequence, northern California: Geochimica et Cosmochimica
Acta, v. 47, p. 1487–1499.
Tanweer, A., 1993, Determination of deuterium in brines and
hypersalines aqueous solution by mass spectrometry using zinc
as agent reducing: Analyst, v. 118, p. 835–838.
Taylor, B., 1986, Magmatic volatiles: isotopic variation of C, H, O,
and S, in W. J. Valley, P. H. Taylor, and R. J. O’Neil, eds.,
Stable isotopes in high temperature geological processes: Re-
views in Mineralogy, v. 16, p. 185–225.
Truesdell, A. H., M. J. Lippmann, J. L. Quijano, and F. D’Amore,
1995, Chemical and physical indicators of reservoir processes
in exploited high-temperature liquid-dominated geothermal
fields: Proceedings of the World Geothermal Congress,
p. 1933–1938.
Vogel, J. C., 1970, Carbon-14 dating of groundwater, in Isotope hy-
drology: Proceedings Symposium, Vienna, Austria, Interna-
tional Atomic Energy Agency, p. 225–239.
Walter, L. M., A. M. Stueber, and T. J. Huston, 1990, Br-Cl-Na
systematics in the Illinois basin fluids: constraints on fluidorigin
and evolution: Geology, v. 18, p. 315–318.
Weaver, T. R., S. K. Frape, and J. A. Cherry, 1995, Recent cross-
formational fluid flow and mixing in the shallow Michigan ba-
sin: Geological Society of America Bulletin, v. 107, p. 697–
707.
White, D. E., 1965, Saline waters of sedimentary rocks, in A. Young
and G. E. Galley, eds., Fluids in subsurface environments:
AAPG Memoir 4, p. 342–366.
Yeh, H. W., 1980, D/H ratios and late-stage dehydration of shales
during burial: Geochimica et Cosmochimica Acta, v. 47,
p. 1487–1499.
Zherebtsova, I. K., and N. N. Volkova, 1966, Experimental study of
behavior of trace elements in the process of natural solar evap-
oration of Black Sea water and Sasyk-Sivash brine: Geochem-
istry International, v. 3, p. 656–670.
Zinder, S. H., 1993, Physiological ecology of methanogens, in J.
Ferry, ed., Methanogenesis ecology, physiology, biochemistry,
and genetics: New York, Chapman and Hall, p. 128–206.