E. Díaz-Martínez e I. Rábano (Eds.

), 4th European Meeting on the Palaeontology and Stratigraphy of Latin America

Cuadernos del Museo Geominero, nº 8. Instituto Geológico y Minero de España, Madrid, 2007. ISBN 978-84-7840-707-1
© Instituto Geológico y Minero de España

UPPER CRETACEOUS PALEOGEOGRAPHY AND RESERVOIR DISTRIBUTION IN

THE UPPER MAGDALENA VALLEY OF COLOMBIA

G. Veloza-Fajardo, M. de Freitas and M. Mantilla

Hocol SA, Colombia,

[email protected]

Keywords: Campanian-Maastrichtian, paleogeography, reservoir distribution.

INTRODUCTION

Upper Magdalena Valley (UMV) is an intermontane basin bounded by the Chusma Fault system in the
W and by the Garzón-Suaza Fault system in the E flank; the northern and southern limits are the Girardot
Fold belt and the Altamira Fault respectively. It is comprised by two sub-basins named Girardot (GSB, north-
ern portion) and Neiva (NSB, south) separated by the Natagaima and Pata highs, which expose the pre-
Cretaceous basement (Fig. 1). The UMV is the third most important hydrocarbon province in Colombia. Five
decades of exploration resulted in the discovery of 35 oil fields, which produce from three main reservoirs.
The Campanian-Maastrichtian interval contains the best reservoir rocks in the basin.
Most of the previous attempts to establish the paleogeography and reservoir potential assessment of
the Campanian-Maastrichtian section in the UMV were limited to localized data and centered on specific
topics, being poorly representative of the entire basin and not properly addressing the regional tectonic
implications.

REGIONAL SETTING

The Upper Magdalena Valley (UMV) is an NE-SW elongated intermontane basin separating the Central
and Eastern Cordilleras of Colombia (Fig. 1). The UMV extends for nearly 400 km, with an average width
of 50 km. Some key references for the understanding of the structure, stratigraphy and hydrocarbon poten-
tial of the basin include the following: Beltrán & Gallo (1979), Schamel (1991), Van der Wiel (1991), Gómez
(2003).

STRATIGRAPHIC MODEL

Controversy about the stratigraphic nomenclature of the Upper Cretaceous of the UMV has occurred
in the past years because of the interpretation of the stratigraphy presented in each area, where regional

399
 G. Veloza-Fajardo, M. de Freitas and M. Mantilla

concepts were applied to local studies. The stratigraphy defined in the GSB differs of the one defined for
the NSB (Table 1), where just one sandstone level is present above the Lower Lidita Fm., and no formal
stratigraphic nomenclature has been defined for this sub-basin.
Eleven informal units or cycles, named from base to top Kml1-5, Upper Lidita Unit and units Kmu1-5
were defined for this time span (from Campanian–Middle Maastrichtian). Units Kml1-5 are time equiva-
lent deposits with the Arenisca Dura Fm. of the Cordillera Oriental area and rest conformably over the lime-
stones and shales of the Villeta Gr.; and deposits of the cycles Kmu1-5 are equivalent with the deposits of
the Arenisca Labor-Tierna of the same area. The equivalent units between GSB and NSB are presented in
Table 1. Five principal areas were defined in this study on the basis of similar depositional environments
and thicknesses. These areas are named Fusa, Ortega, North Neiva, South Neiva and Altamira, each with
its own stratigraphic relationships, depending of the depositional environment and how the tectonic evo-
lution affected the preservation of these deposits (Fig. 1).

Figure 1. Generalized geologic map of the Upper Magdalena Valley showing major fault systems and the study area.

400
 UPPER CRETACEOUS PALEOGEOGRAPHY AND RESERVOIR DISTRIBUTION IN THE UPPER MAGDALENA VALLEY OF COLOMBIA

Table 1. Historical development of the stratigraphic nomenclature of the Upper Cretaceous succession in the UMV.

Database and methodology

Almost 100 wells and 40 surface sections were analyzed and correlated with the purpose of a better
understanding of the depositional environments and of the depositional sequences. All the wells were
adjusted to a unique set of logs, then those wells were corrected for dip (when dipmeter data were avail-
able) for obtaining the true stratigraphic thickness (TST). Two reference levels, Lower and Upper Lidita Fms.
were used as regional maximum flooding surfaces (Guerrero et al., 2000). We assume those units to be
approximately isochronal across the basin. The facies association for each interval were defined on the
basis of cores’ descriptions and observations of the outcrops (Veloza et al., 2007). More than 200 petro-
graphic and biostratigraphic samples were tied to each defined interval.
Parasequences represent relatively conformable successions of genetically related bedsets, bounded by
marine flooding surfaces and their correlative surfaces. The concept of facies dislocation is used to deter-
mine the nature of the top of the sequence (Monserrate Fm top), where rocks of a shallower facies rests
directly on rocks of a significant deeper facies (Emery & Myers, 1996). A facies dislocation implies a fall in
relative sea level and the development of a subaerial unconformity. It thus marks a sequence boundary or
its correlative conformity.

Age and stratigraphic relations

Integration of biostratigraphic information from many wells, and surface samples from published
reports allow us to develop a chronostratigraphic framework for this study. Almost all the published bios-
tratigraphic information of the UMV has been obtained for the northern portion of the basin, GSB.
In the Ortega area biostratigraphic reports from the Toldado and Rio Saldaña wells (Ecopetrol-ICP,
2000) and surface samples (Guerrero et al., 2000) suggest an Santonian age for the Lower Lidita level,
Lower Campanian for the Kml interval, Middle–Upper Campanian for the Upper Lidita and Upper
Campanian –Middle Maastrichtian for the Kmu sequence. This is the same for Fusa area, where samples
form Atadero-1 well were analyzed. However, for this area a subtle paraconformity was identified on the
top of the Guadalupe Gr. (Bayona et al., 2003), where the Middle Maastrichtian time span is absent.

401
 G. Veloza-Fajardo, M. de Freitas and M. Mantilla

In the North Neiva area, private reports from the Cebu-5 well (Robertson Research, 1984) indicate a
Santonian age for the interval identified as Lower Lidita level, and indicate an undefined Campanian-
Maastrichtian? age for the overlying layers. The same conclusion was achieved by palynological analysis
made on the Yaguará area, in Los Mangos-60 well (Duque-Caro, 1998), where the Upper Lidita level
appear on top of the analyzed interval. In the southernmost portion of the UMV, in the Iskana-1 well, paly-
nological samples indicate a lack of biozones on the top of the Cretaceous sequence, where the interval
defined as Caliza del Tobo represents the Campanian-Early Maastrichtian and the overlying deposits of the
Guaduala Fm. yield palynomorphs that suggest a late Maastrichtian age (Tepma, 2001).
All these observations are in agreement with the unconformity observed in northern Ecuador, where
the Cretaceous marine strata of the Napo Fm. (Albian-Campanian) is overlain by the Maastrichtian-
Paleocene Tena Fm. (Balkwill et al., 1995; Barragán et al., 2005), and south Colombia, in the Putumayo
area between the Rumiyaco and Las Iglesias Fms. (Mora et al., 1998).

Litostratigraphy, petrography and sedimentary environments

From north to south, in the Fusa area, the deposits of the Kml units represent environments from upper
offshore (Kml1) to middle shoreface (Kml3) and eventually mouth bar deposits are present. These intervals
are represented by three progradational cycles, from the Kml1 to Kml3. The Kml4, Kml5 and Upper Lidita
intervals represent the transgression of the sea level, with deposits ranging from middle shoreface to off-
shore, where the last cycle was deposited. The average thickness of this interval in the Fusa area is 160
feet. The Kmu deposits in this area range in environments from middle shoreface to foreshore. Sequences
Kmu1-Kmu3 are principally progradational and Kmu4 and Kmu5 are agradational, representing environ-
ments where the accommodation space created was compensated with the sediment influx. The average
thickness of this sequence is 70 feet. Petrographic samples from the Atadero-1, Revancha-2ST and surface
samples (Ecopetrol-ICP, 2000) shows that since the Kml unit, lithic and feldspar fragments are present,
being classified as subarkoses and sublithoarenites. The samples from the Kmu unit exhibit principally
quartz rich sandstones with just few feldspar fragments. Variations in sandstone composition from base to
top (from immature to mature sandstones) may be due to the rework of sediments, deposited in environ-
ments near to the shore line.
In the Ortega area, the depositional environments in the Kml cycles range between offshore and the
transition to lower shoreface. These deposits present an agradational path, and minimal deflections were
observed in the well logs. Such response of the curves obey to the high silt content of the rock, typical of
those environments. The average thickness of this interval is 90 feet. The Kmu sequence is principally agra-
dational from the Kmu1 to Kmu3 intervals, and becomes progradational in the Kmu4 and Kmu5
sequences. The Kmu5 represents an abrupt facies shift, from marine to alluvial environments, which records
the uplifting of the Central Cordillera and that was preserved in this area when the accommodation space
exceeded the rate of uplift and sediment supply. This interval has an average thickness of 80 feet and
increases northward reaching 200 feet. Compositionally these sediments are almost all of them subarkos-
es and sublithoarenites (Guerrero et al., 2000).
In the North Neiva area the observed facies presents a progradational path in all the cycles of the Kml
interval.The environments vary from lower shoreface to middle shoreface, and eventually mouth bar deposits
appear. Of the Kmu unit, only sequences 1 and 2 were preserved in the northern portion of the area. In the
Cebu-5 well the upper Kmu2 interval was eroded or non deposited. The average thickness is 70 and 60 feet
for Kml and Kmu, respectively. Sandstones composition from this area comes principally from wells of the
Cebu Field, where sandstone composition varies from base to top from subarkoses to quartzarenites.

402
 UPPER CRETACEOUS PALEOGEOGRAPHY AND RESERVOIR DISTRIBUTION IN THE UPPER MAGDALENA VALLEY OF COLOMBIA

Figure 2. Paleogeographic map, with palinspastic restoration, of the UMV area. (A) Early Campanian (+/- 80Ma) corresponds
approximately to Kml2-3; (B) Early Maastrichtian (+/- 70 Ma) corresponds approximately to Kmu 4- 5. The present position of the
cities and present faults are included for reference.

In the South Neiva area just a few wells have all the Kml interval complete. In the northern portion of
this area the uppermost level found is the Upper Lidita, which is in unconformable contact with the
Guaduala Group. The depositional environments range from lower shoreface to upper offshore, and even-
tually mouth and mouth bars deposits are found. The Kml cycles in this area present principally a progra-
dational path. The interval’ thickness ranges from 40-60 feet. Sandstones composition of Kml unit in this
area is principally subarkosic and exhibits great amounts of feldspar content. Some calcareous quartzaren-
ites are present too.
In the Altamira area, few cycles are preserved, namely the Kml1 to 3. These cycles represent a progra-
dational path, as found in all of the study area. These deposits have always been interpreted as offshore
deposits. In our interpretation, based mostly on petrography of surface samples, they represent high ener-
gy, near shore environments (middle to upper shoreface), which occurred far from clastic sediment input,
allowing the development of biogenic habitats. The average thickness of the sedimentary sequence in this
area is 50 feet. Compositionally the Kml unit in the Altamira area is principally of calcareous affinity and
some calcareous sandstone is present. Calcareous rocks varies from sparite dominated bioclastic grain-
stones to micrite dominated packstones showing the proximity to the shoreline because of the energy of
the deposit (Osorio & Rodriguez, 2000; Ecopetrol-ICP, 2000).
In a general sense, provenance analysis does not indicate a clear source area for the sediments of the
analyzed interval in the UMV. The petrographic samples analyzed fall in the upper corner of the classifica-
tion Qz-F-L ternary diagram (90:6:4). The source rock for these deposits are the Guyana Shield (granite and
granodiorite), ancestral Central Cordillera (andesite) and the Cretaceous substrate (intraclasts). The pres-

403
 G. Veloza-Fajardo, M. de Freitas and M. Mantilla

ence of well rounded quartz grains in compositionally immature sandstones suggests that these grains
were recycled from quartzose underlying units.

DISCUSSION

In the UMV, the Campanian-Lower Maastrichtian sequence varies considerably in thickness, from 130
feet in the south to 1300 feet in the north (Figs 3A-B & 4). This ten-fold thickness variation reflects major
environmental changes in the basin. The Santonian-Lower Campanian deposits (Arenisca Dura Fm’s time
equivalents) shifted facies from offshore to middle shoreface. Also, the absence of the Labor-Tierna Fms.
time equivalents reflects the erosion or non-deposition during the Late Campanian-Middle Maastrichtian
time span, at least in the NSB. We suggest that a progressive, south to north uplift of the ancestral Central
Cordillera, from the Campanian through to Maastrichtian produced a paraconformity between the strata
of the Guaduala (or Seca) Fm and the Monserrate Fm in this area, shifting depositional environments from
alluvial plain to lower shoreface, respectively.
Figure 2 represents our paleogeographic reconstruction for the Early Campanian and Early
Maastrichtian times. The Kmu cycles are only fully developed in the GSB and the northernmost NSB, being
absent in the southern NSB, where the Guaduala rest paraconformably over the Kml cycles. Coarse clas-
tics derived from the Central Cordillera developed in the western flank of the GSB during the uppermost
cycle Kmu5, directly overlying shales of the Nivel de Lutitas Fm. These deposits, assigned to the La Tabla
and Cimarrona Fms. and Kmu5 cycle of this study, form important reservoirs in this part of the basin and
are genetically related with the overlying Seca Fm and are in a paraconformable relation with the under-
lying Kmu4 or 3 (Buscavida Fm.) cycles.

Reservoir distribution

The Campanian–Maastrichtian section contains important hydrocarbon reservoirs in the UMV. A series
of oil and gas fields produce from sandstones of this interval, including Dina, Tello and Guando. Reservoir
characteristics and distribution in the basin are controlled primarily by the original depositional environ-
ments and the burial history, which is strongly marked by the tectonic pulses that affected the UMV and
adjacent areas since the Campanian. The overall reservoir distribution and quality of the
Campanian–Maastrichtian interval may be summarized as follows:
• The Kml cycles are regionally the most important reservoirs in the basin. They are particularly well
developed in the eastern margin of the GSB and the northern NSB (Fig. 3C), corresponding to the Mon-
serrate, and Dura sandstones; distal facies corresponding to the Nivel de Lutitas y Arenas form non
reservoirs in the Ortega area, western portion of the GSB (Fig. 3C). Along the eastern and southern
margin of the basin, carbonate facies associated with the El Tobo limestones form poor reservoir rocks
(Fig. 3C);
• In most of the central area of the GSB, the best reservoir facies (attaining more than 20% porosities)
have been removed during the “early Andean” event in the Oligocene (De Freitas et al., 2006).The out-
line of subcrop of the upper Cretaceous in the basin is indicated (Fig. 3);
• The development of good coarse sand reservoirs in the Ortega area, along the western margin of the
GSB, belonging to the uppermost cycle Kmu5 reflects the uplift of the Central Cordillera (Fig. 3E).These
westerly derived deposits immediately overlie fine grained deposits which are non reservoirs (Fig. 3D).

404
 UPPER CRETACEOUS PALEOGEOGRAPHY AND RESERVOIR DISTRIBUTION IN THE UPPER MAGDALENA VALLEY OF COLOMBIA

Figure 3. Gross thickness map for the Kml (A) and Kmu (B) intervals. Average porosities map of the Kmll-5 sequences (C), Kmul-4
sequences (D) and Kmu5 stratigraphic cycle (E). Maps without palinspastic restoration.

CONCLUSIONS

Eleven cycles were recognized in the Campanian to early Maastrichtian sequence in the Upper
Magdalena Valley of Colombia. A ten-fold thickness variation is observed from south to north (130 ft to
1300 ft, respectively). Together with facies changes and stratal relationships, this thickness variation is
interpreted to reflect a progressive south to north uplift of the Central Cordillera from the Campanian
onwards, producing a paraconformity developed between the Maastrichtian-Paleocene Guaduala Fm and
the Campanian-early Maastrichtian cycles studied.
Provenance analysis shows more than one source area for the sediments of the analyzed interval.
Guyana Shield, Ancestral central Cordillera and Cretaceous sediments were the principal areas that sourced
this units. Although those were probably the principal source areas, recent studies indicate that there were
some tectonic events (+/- 100Ma, Van der Wiel, 1991; Jaimes & De Freitas, 2006; Veloza et al., 2007) that
uplift areas like the Garzón Massif or the Macarena Range that could serve as source area too.
Petrographically, almost all the samples were classified as quartzarenites, subarkoses and sublithoarenites.
The Campanian-Mastrichtian sequence contains important reservoirs in the basin. The distribution of
Campanian-Maastrichtian reservoirs in the UMV are controlled primarily by their depositional environ-

405
406
G. Veloza-Fajardo, M. de Freitas and M. Mantilla

Figure 4. Stratigraphic correlation of the Santonian - Maastrichtian sequence across the Upper Magdalena Valley.
 UPPER CRETACEOUS PALEOGEOGRAPHY AND RESERVOIR DISTRIBUTION IN THE UPPER MAGDALENA VALLEY OF COLOMBIA

ments and subsequent burial history. The “early Andean” compressional event (Oligocene) removed vast
amounts of prime reservoirs from this interval. Carbonate non reservoir rocks developed along the south-
eastern margin of the Neiva subbasin during the Kml cycles. Good coarse sand reservoirs developed in the
western Girardot subbasin, reflecting the uplift of the Central Cordillera during the uppermost cycle kmu5
(Early Maastrichtian).
Apart from the evident stratal patterns shown in the well correlation and the gross thickness of the
studied interval, evidences that support this hypothesis include: (1) absence of biozones of the middle
Campanian- lower Maastrichtian in the Fusagasuga (Bayona et al, 2003) and in the Altamira areas (Tepma,
2001); (2) abrupt change of facies between the Guaduala Fm and Monserrate Fm in the NSB and in the
southern portion of GSB; (3) velocity anomalies on top of Monserrate Fm (De Freitas, in prep.) that sug-
gest subaerial exposure of the sediments; (4) presence of feldspars (mostly microcline), lithic fragments in
low amounts and intraclasts, indicating a far source area, reworked sediments or both.

Acknowledgements

We are thankful to Hocol for permissions to publish these ideas. Discussion with many geologists
improved this document.

REFERENCES

Balkwill, H. R., Rodríguez, G., Paredes, F.I. and Almeida, J.P. (1995). Northern part of Oriente basin, Ecuador: reflection
seismic expression of structures. En: A.J. Tankard, R. Suárez Soruco and H.J. Welsink (Eds.), Petroleum basins of
South America. AAPG Memoir, 62: 559-571.
Barragán, R., Toro-Alava, J., Jaillard, E., White, H., Toulkeridis, T., Montenegro, J. and Medina, G. (2005). Lower
Maastrichtian syntectonic sedimentation along the Subandean Zone and its relationship with an accretionary event
of an oceanic terrane registered in the Cordillera Occidental of Ecuador. Proceedings VI ISAG, Barcelona.
Barrio, C. and Coffield, D. (1992). Late Cretaceous stratigraphy of the Upper Magdalena Basin in the
Payande–Chaparral Segment (Western Girardot Sub-Basin), Colombia. Journal of South American Earth Sciences,
5(2).
Bayona, G., Cortes, M., Jaramillo, C. and Llinas, R. (2003). The Tertiary Fusagasuga Succession: a record of the complex
Latest Cretaceous-pre-Miocene deformation in an area between the Magdalena Valley and Sabana de Bogotá.
Actas VII Simposio Bolivariano de Exploración Petrolera en las Cuencas Subandinas.
Beltran, N. and Gallo, J. (1979). The geology of the Neiva Sub-Basin, Upper Magdalena Basin, southern portion. En:
Geological field trips Colombia, 1959-1978, ACGGP, Bogota: 253-275.
De Freitas, M., Vidal, G. and Mantilla, M. (2006). Structural evolution and hydrocarbon entrapment in the Balcon Field
area, upper Magdalena Valley, Colombia. Actas IX Simposio Bolivariano de Exploración en Cuencas Subandinas.
Duque-Caro, H. (1998). Micropaleontology and biostratigraphy of the well Los Mangos 60, Yaguara Area. Braspetro
internal report.
Ecopetrol-ICP (2001). Evaluación integrada del tren de producción de la cuenca del Valle Superior del Magdalena,
Departamento del Tolima. Internal report.
Emery, D. and Myers, K. (1996). Sequence Stratigraphy. Blackwell Science Ed., 297 pp.
Gómez, E., Jordan, T., Allmendiger, R., Hegarty, K., Kelley, S. and Heizler, M. (2003). Controls on architecture of the Late
Cretaceous to Cenozoic southern Middle Magdalena Valley Basin, Colombia. GSA Bulletin, 115(2).

407
 G. Veloza-Fajardo, M. de Freitas and M. Mantilla

Guerrero, J., Sarmiento, G. and Navarrete, R. (2000). The Stratigraphy of the W side of the Cretaceous Colombian Basin
in the Upper Magdalena Valley. Reevaluation of selected areas and type localities including Aipe, Guaduas, Ortega
and Piedras. Geología Colombiana, 25.
Jaimes, E. and De Freitas, M. (in press). An Albian-Cenomanian unconformity in the northern Andes: Evidence and tec-
tonic significance. Journal of South American Earth Sciences.
McCourt, W.J., Aspden, J.A. and Brook, M. (1984). New geological and geochronological data from the Colombian
Andes; continental growth by multiple accretion. Journal of the Geological Society (London), 141: 831–845.
Mojica, J. and Franco, R. (1990). Estructura y Evolución Tectónica del Valle Medio y Superior del Magdalena, Colombia.
Geología Colombiana, 17: 41-64.
Mora, A., Venegas, D. and Vergara, L. (1998). Estratigrafía del Cretácico Superior y Terciario Inferior en el sector norte
de la Cuenca de Putumayo, Departamento de Caqueta, Colombia. Geología Colombiana, 23.
Osorio, F. and Rodríguez, A. (2000). Análisis facial para la interpretación paleoambiental de la Formación El Tobo
(Carbonatos de la Formación Monserrate) del Cretácico Superior en el Valle Superior del Magdalena, Departamento
del Huila, Colombia. B.Sc. Thesis, Manizales, Universidad de Caldas.
Robertson Research (1984). Biostratigraphy of a core simples from Cebu-5 well, Hocol. Internal report.
Schamel, S. (1991). Middle and Upper Magdalena Basins, Colombia. En: K.T. Biddle (ed.) Active margin basins. AAPG
Memoir, 52: 283-301.
Tepma (2001). Biostratigraphical study of the Cretaceous Section-South Gaitanas Block (Upper Magdalena Valley
Basin-Colombia). Internal report.
Van der Wiel, A.M. (1991). Uplift and volcanism of the SE Colombian Andes in relation to Neogene sedimentation in
the Upper Magdalena Valley. Ph. Thesis, University of Wageningen, Amsterdam, 208 pp.
Veloza, G., Mora, A., De Freitas., M. and Mantilla M. (2007). Dislocación de facies en el tope de la secuencia cretáci-
ca de la Subcuenca de Neiva, Valle Superior del Magdalena y sus implicaciones en el modelo estratigráfico secuen-
cial colombiano. Actas del XI Congreso Geológico Colombiano.

408