1. INTRODUCTION
The Lower Cretaceous Caballos Formation exposed in the Upper Magdalena Valley, has been related either to the Cretaceous extensional basins that characterizes the eastern sector of the Colombian and Ecuadorian Andes [1], or to a foreland basin associated to compressive event of this age that which has been mostly recognized in the Cordillera Real of Ecuador [2,3].
Seismic interpretation, stratigraphic and petrographic observations [4,5], suggest that the accumulation of the Caballos Formation in the Upper Magdalena Valley was limited to the east and west by structural highs that may correlate with the Garzón paleo massif and proto-Central Cordillera.
Given the oil-producing nature of this formation in the Upper Magdalena Valley and Putumayo basins, several studies that have deal with its reservoir qualities have suggested the existence of lateral environmental variations, which are expressed by the strong lateral variations on facies type and thickness [5], or to the presence of calcareous allochemicals and the variation in the heat flow near the Lower Cretaceous fault areas, which would control the dissolution and the generation of secondary porosity [6].
This paper presents petrographic, heavy mineral analysis, detrital zircon U-Pb Geochronology, and basic petrophysical analysis from two stratigraphic sections of the Caballos Formation exposed in the Neiva and Girardot subbasins within the Upper Magdalena Valley Figure 1.
The new results make it possible to characterize the provenance of the Caballos Formation and to review the source are configuration during basin filling in the Aptian- Albian. Additionally, the analysis of the relationship between petrophysical attributes and petrographic characteristics suggests that the quartz and feldspar contents, the formation of pseudo-matrix and dissolution processes affect the quality of the siliciclastic facies in the Caballos Formation.
2. THEORETICAL FRAME
GEOLOGICAL FRAMEWORK
The geological record between the Jurassic and the Early Cretaceous in the Central and Eastern Colombian Andes is related to the subduction of the Farallón plate, and several extensional events associated with the fragmentation of Pangea and the formation of the Gulf of Mexico [1], [7]-[10]. This complex tectonic configuration is recorded by the growth of different magmatic arcs during a large part of the Jurassic, and the generation of a series of extensional basins within the continent [11]. During the Lower Cretaceous, the continued formation of the proto-Caribbean and the opening of the Atlantic [12], as well as several collisional events have shape the margin [2].
The Central Cordillera of the Colombian Andes exhibits a Devonian, Permo-Triassic and Jurassic plutonic and metamorphic substrate [2],[13]-[16]. Different Lower Cretaceous clastic sequences are discontinuously deposited over this basement rocks, and are characterized by a transgressive record that extends from the Berrasian to the Albian-Aptian, and ends with the development of oceanic crust and the subsequent construction of a magmatic arc that likely extends to the Upper Cretaceous [10],[17]-[18].
The Eastern Cordillera including its continuation to the south in the Garzón Massif includes a Proterozoic metamorphic rocks with radiometric ages between 980 Ma-1180 Ma [19,20] and Early to Middle Jurassic granitoids [21], as well as (meta) sedimentary rocks from the Lower to Upper Paleozoic [22]-[24]. An extensive Jurassic to Cretaceous clastic and chemical record with minor Cretaceous gabbroic intrusions overly this Precambrian to Paleozoic rocks [25,26]. The Lower Cretaceous record is characterized by a transgressive nature, associated with conditions of high subsidence within an extensional basin.
In the Middle Magdalena Valley, the clastic record has been similarly associated with an extensional history since the Berrasian-Hauterivian [27,28], which was likely deposited over the volcano-sedimentary and plutonic sequences of Triassic to Jurassic in age. This record is probably characterized by the development of high-energy fluvial environments, followed by a deepening of the basin towards transitional environments in the Aptian [1],[29]-[31]. In the Upper Magdalena Valley, the Putumayo basin and the eastern Ecuadorian basin, sedimentary accumulation began in the Barremian and extends to the Albian-Aptian, and is also characterized by extensive subsidence and a transgressive nature, deposited in a continental to transitional environments [32,1], which has been interpreted as associated with an extensional [1] or collisional record [2,3].
Later, the stratigraphic record present in the Eastern Cordillera and the Middle and Upper Magdalena Valley is characterized by stable tectonic conditions, at least between the Turonian and the Campanian [33]Figure 2.
LOWER CRETACEOUS RECORD IN THE UPPER MAGDALENA VALLEY
The Upper Magdalena Valley is a depression bound by the Central and Eastern Cordillera characterized by 400 km in length, and a width that can reach 65 km.
Based on the lateral variations of facies, it has been suggested that the basin where the Caballos Formation was accumulated was characterized by a block related paleotopography [34,5], bounded by identifiable structural highs in the seismic lines, which include an extensive paleo-high to the east identified as Arco del Vaupés, that may include segments of the Garzón Massif and Serranía de la Macarena, and to the west another topographic high that could be related to the proto-Central Cordillera.
The Mesozoic sedimentary fill of the Upper Magdalena Valley comprises units that span from the Triassic to the Cretaceous (Figure 2). The Triassic-Jurassic includes continental clastic sediments, a calcareous record, effusive and pyroclastic volcanism, as well as plutonic rocks of acid to intermediate composition, which have been grouped in the Luisa, Payandé and Saldaña formations respectively [25]. Different intermediate to felsic plutonic bodies with an extensive geological history between 190 Ma and 139 Ma intrude some of these sedimentary units and the older metamorphic basement [35].
In the Lower Cretaceous, sedimentation possibly begins during the Barremian with the deposition of the Yaví Formation [5], which is characterized by continental siliciclastic facies and pyroclastic volcanic material. These rocks are in angular unconformity on the pre-Cretaceous basement and in gradational contact with the overlying Albian-Aptian Caballos Formation [36].
The Caballos Formation is characterized by mix clastic and calcareous sedimentation accumulated in transitional and continental environments, and reach a maximum thickness of 450m [32,37], which can be divided into three members [38]-[41].
The lower member is characterized by intercalations of mudstones and sandstones of up to 250m of thickness [37]. Towards the base it exhibits thin beds of coarse-grained to massive conglomeratic sandstones, with low-angle lamination or trough cross stratification [42,6]. This member of Aptian in age [43], has been related to fluvial and estuarine environments [40,41,6].
The middle member includes muddy and calcareous intercalations with thin packets of sandstones. This succession has a thicknesses between 80m and 100m [44,32,42,6], and has been related to coastal environments [41,34], having been accumulated during the Upper Aptian [43].
The upper member of the Caballos Formation exhibits layers of medium to fine grain sandstones with interspersed mudstones, whose maximum thickness can reach 60m [45]. It has been related to fluvio-deltaic environments and coastal areas, which likely accumulated during the Middle Albian [43].
Conformable with the Caballos Formation are intercalations of shales, black chert, and micritic and esparitic limestones from the Villeta Group, that were accumulated in a shallow platform environment during the Campanian and Cenomanian [43].
EXPERIMENTAL DEVELOPMENT
PETROGRAPHY AND IMAGE ANALYSIS
Twenty-one sandstone samples from the Caballos Formation collected in the Ocal Creek section (Huila) and in the Cobre Creek (Tolima) in the Upper Magdalena Valley were petrographically analyzed.
Roundness and sphericity was determined in 100 grains [46], and 300 grains were counted in accordance with the Gazzi-Dickinson and Indiana methods [47]. Petrographic discrimination of quartz types was carried out on 100 grains [48].
The porous space was measured in 14 samples by analyzing the images obtained in an Epson V500 scanner, with two different software packages: IMAGEJ-JPOR® (49) and JMICROVISION® [50].
HEAVY MINERALS
Heavy minerals were analyzed in nine sandstone samples from the lower and upper members of the Caballos Formation.
The sandstones were disintegrated using a metallic mortar and sieved using 400um disposable mesh. Subsequently, sodium polyungstane was used to separate the fraction with a density > 2.89 g / cm3. The grains were immersed in Meltmount® with a refractive index of 1.539. Up to 300 translucent grains were counted using the line method [51].
DETRITAL ZIRCON U-PB GEOCHRONOLOGY
U-Pb detrital zircons from four-samples were analyzed with the U-Pb LA-ICP-MS (Laser Ablation Inductively Coupled Plasma Mass Spectrometry) technique. Preparation was carried out in the Zirchron LLC laboratories and the analyses were conducted at the Washington State University in Pullman, United States [52]. The Plesovice standard with an age of 337.1 ± 0.4 Ma was used for this study [53]. Analytical results are present in Appendix 1.
The analyses were performed on the zircon cores in order to avoid zircons with complex histories [54]. Individual ages for each sample were processed and plotted using Probability Density Plots (PDP) with the Isoplot 3.62® program [55]. Individual analyses that exceeded 10% analytical error and 20% unconformity were discarded for interpretations and discussion. However, unconformity was not evaluated for grains less than 500 Ma, given that the abundance of 207Pb cannot be measured with sufficient precision to evaluate the unconformity between the ages 206Pb/238U and 206Pb/207Pb [56]. In addition, Appendix 2 shows the Tera- Wasserburg concordia diagrams prepared for each of the analyzed samples.
PETROPHYSICAL ANALYSIS
Porosity was calculated by means of Boyle porosimetry in 10 rock core samples, and through the gravimetric method in an irregularly shaped samples [57]. Samples were taken in outcrops from both study sections (Figures 3 and 4) with a portable drill. Permeability tests were performed on 7 samples following the procedure provided by (58), using a confining pressure of 2000 psi similar to those found in the fields relating to the Caballos Formation in the Magdalena Valley.
SCANNING ELECTRON MICROSCOPY (SEM) AND X-RAY DIFFRACTION.
In order to analyze the clayey fraction, a total of four samples were analyzed in the JEOL JSM-5910LV Scanning Electron Microscope, with a BSE detector and an EDX spectrometer at Universidad Nacional de Colombia (Medellín). The clayey fraction less than <2um was analyzed after removing organic matter and carbonates with Morgan's solution and 35% hydrogen peroxide. The standard conditions were Ka1 = 1.540598, 45 kV; 40 mA with a 2 theta between 3° and 72° (with a step of 300 seconds in total and an angle of 0.0334°). The semi-quantitative estimate followed the protocol provided by [59].
4. RESULTS
Seven samples from the lower and upper members of the Caballos Formation were sampled in the Ocal section (41) (Figure 3). Whereas for the Cobre section (60) we have select nine samples also from the Lower and Upper members (Figure 4).
PETROGRAPHY
EL OCAL CREEK SECTION:
Sandstones from the lower member are of medium to fine grain, subangular and sub-rounded, with both low and high sphericity and poor to moderate selection. The contacts between the grains are mainly concave-convex and sutured. The matrix content varies between 4% and 18%, and include pseudomatrix associated with the disintegration of mudstone lithics. XRD analysis suggest that the matrix is dominated by the presence of kaolinite and interstratified illite-smectite ( Figure 5, Table 1 ).
The sandstones are quartz arenites, subarkoses and sublitharenites [61]; see Table 1, Figure 6. Quartz grain include monocrystalline quartz with straight extinction and sedimentary and diffuse polycrystalline quartz with more than three grains. Feldspar is less than <7%, and include mainly K-feldspar with a pertitic texture and alteration to clays. Lithics include felsic volcanic and plutonic fragments with quartz and feldspar (Figure 6), reddish brown to yellow mudstones and chert, and muscovite schists. Flexed muscovite was also seen in proportions lower than 5%, whereas zircon, epidote, tourmaline, hornblende, calcite and cordierite are present in proportion less than 1%. Siliceous, ferruginous and calcareous cements were identified (3% -14%). It was also common to find organic matter and oil partially covering pores and generally accompanied by epidote, tourmaline and pyrite.
Sandstones of the upper member are of medium size, subangular to sub-rounded, with low to high sphericity and poor to moderate selection. Contacts between grains are mainly tangential, with minor concave-convex and suture. Matrix is mainly fine silt and mud and is present with proportions between 11% and 12% (Figure 5). XRD analyses indicate that the most abundant clay minerals are interstratified illite-smectite and kaolinite.
The sandstones are calcareous quartz arenites [61]. Monocrystalline quartz with straight extinction predominates, while sedimentary and diffuse polycrystalline quartz are also present in smaller quantities. Polycrystalline quartz grains are found with 2 and three grain and polygonal to diffuse contacts. Feldspar is less than 7%, and include K-feldspar (microcline) with perthitic texture, whereas the plagioclase is altered to clay minerals. Lithics include mudstone, chert and quatz-feldspar sandstones that are affected by ductile deformation (Figure 6). Glauconite is presented in a proportion between 5% and 15%, whereas zircon, epidote, tourmaline, garnet, monazite and diopside are present in proportions below 2%. Calcareous and ferruginous cements were identified (7% -11%).
EL COBRE CREEK SECTION:
Sandstones from the lower member are fine to coarse grain, rounded to sub-rounded grains, low to high sphericity and poor selection. Contacts are tangential, with minor concave-convex and sutured relations. Matrix contents varies between 9% to 18%, and is dominated by clay minerals (6%) (Figure 5), that based on the XRD analysis they are mainly illite-smectite and kaolinite (Table 1).
Compositionally, the sandstones show appreciable amounts of lithics, and are classified as litharenites, sublitharenites and subarkoses [61], see Table 1, Figure 6). Monocrystalline quartz with a straight and undulatory extinction is the most abundant quartz type, whereas polycrystalline quartz is mostly sedimentary and diffuse with more than three grains, altough minor polygonal quartz was also identified. The feldspar contents in the samples varies between 7% - 10%, and it relates to K-feldspar that may include perthitic texture and is altered to sericite or carbonates. The volcanic lithics are felsic and contain quartz and feldspar, the plutonic ones are also felsic by include muscovite (Figure 6), the sedimentary ones are laminated mudstones, quartz-rich and k-feldspar rich sandstones, and the metamorphic lithics are mainly muscovite schists (Figure 6). Flexed muscovites and biotites were observed in proportions smaller than 5%. Other minerals such as zircon, epidote, chlorite, glauconite and diopside are present in a proportion below 1%. Cements are ferruginous and siliceous (1% -5%) and it is also common to find organic matter and small sheets of coal partially covering pores.
Sandstones from the upper member varies from coarse to very fine grain, with sub rounded to rounded grains, and high and low sphericity, with moderate to good selection. Contacts between the grains are predominantly concave-convex and sutured, with minor angential contacts. The matrix varies between 4% and 11%, and is dominated by clay in sheet arrangements (Figure 5) that corresponds to kaolinite (Table 1).
Sandstones are very quartz rich (> 85%) and are classified as sublitharenite and subarkose [61]; Figure 6). Monocrystalline quartz with undulatory extinction is the most abundant quartz type, followed by sedimentary and diffuse polycrystalline quartz with more than three grains. Feldspar varies between 2% -15%, and nclude mainly K-feldspar with perthitic texture and plagioclase, Lithic fragments include felsic volcanics microcrystailine quartz: mudstone, chert, quartz-rich sandstones and sandstones rich in potassium feldspar. Metamorphic lithics include muscovite schists. Flexed muscovite, biotite, hornblende and zoisite have values below 5%. Other minerals such as zircon, epidote, chlorite, glauconite, titanite and diopside are present in a proportion less than 1%.
Siliceous and calcareous cements were identified (1%-6%). It was common to find organic matter and small remnants of crude oil that partially covering pores, and is usually accompanied by heavy minerals such as epidote, tourmaline, anatase and pyrite.
HEAVY MINERALS
EL OCAL SECTION:
The most abundant ultrastable mineral in the lower member is zircon, followed by rutile and tourmaline (Figure 5). Moderately stable and stable minerals include garnets (colorless to reddish) monazites, apatite and titanite. Unstable to highly unstable minerals nclude zoisite, clinozoisite, epidote, hornblende, diopside, augite, olivine, sphalerite and enstatite in very low proportions (Table 2) Opaque minerals vary between 18-24%.
In the upper member, the most abundant mineral is also zircon followed by rutile and tourmaline (Figure 5). Moderately stable and stable minerals include garnet, apatite and, to a lesser extent titanite and sillimanite. Ustable minerals and highly unstable minerals include zoisite, clinozoisite, epidote and muscovites (Table 2). Opaque minerals are 17% in abundance.
EL COBRE SECTION:
The most abundant mineral in the lower member is zircon, followed by rutile and tourmaline (Figure 5). Moderately stable minerals nclude garnet, apatite, muscovite and sillimanite (Figure 5), while unstable minerals include zoisite, clinozoisite and epidote. The opaques vary between 36-42%. In the upper member the most significant mineral is zircon, followed by rutile and tourmaline (Figure 5). Moderately stable and stable minerals include garnets, muscovite and apatite (Table 2).
Section | Ocal | Cobre | |||||||||
Member | Lower | Upper | Lower | Upper | |||||||
Sample | EJD009 | EJD010 | EJD012a | EJD001 | EJD007 | EJD014 | EJD014a | EJD019 | EJD020 | ||
Zircon | 52.53 | 31.64 | 41.51 | 10.44 | 11.89 | 20.00 | 11.89 | 21.32 | 17.30 | ||
Ultrastable (%) | Rutile | 5.07 | 4.02 | 4.58 | 0.78 | 3.24 | 1.89 | 1.62 | 5.79 | 1.62 | |
Tourmaline | 0.80 | 1.88 | 0.81 | 3.39 | 2.43 | 0.81 | 0.81 | 2.89 | 2.43 | ||
Anatase | 0 | 0 | 0 | 0 | 0 | 0.2695 | 0 | 0.7895 | 0 | ||
Apatite | 1.33 | 0.54 | 0.54 | 8.88 | 1.89 | 0.54 | 2.16 | 2.89 | 0.00 | ||
Stable (%) | Garnet | 0.27 | 0.00 | 0.00 | 2.87 | 1.62 | 1.35 | 0.00 | 3.42 | 0.00 | |
Monazite | 0.00 | 8.85 | 7.55 | 0.00 | 0.00 | 0.00 | 0.00 | 0.79 | 0.00 | ||
Moderately Stable (%) | Sillimanite | 0.53 | 0.00 | 0.00 | 0.52 | 0.00 | 0.81 | 0.00 | 0.00 | 0.00 | |
Spinel | 0.00 | 0.00 | 0.00 | 1.04 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | ||
Sphene | 0.27 | 0.00 | 0.00 | 0.52 | 0.27 | 0.00 | 0.00 | 0.00 | 0.00 | ||
Epidote | 5.07 | 6.97 | 3.50 | 0.52 | 1.62 | 8.38 | 11.62 | 5.00 | 0.54 | ||
Zoisite | 4.00 | 5.09 | 7.28 | 4.18 | 15.14 | 9.73 | 9.73 | 16.58 | 1.35 | ||
Clinozoisite | 2.13 | 4.56 | 4.04 | 11.75 | 4.86 | 2.16 | 3.24 | 5.00 | 0.54 | ||
Biotite | 0.00 | 0.00 | 0.00 | 3.92 | 6.22 | 0.00 | 0.00 | 0.26 | 27.57 | ||
Muscovite | 3.47 | 5.63 | 5.93 | 6.01 | 10.54 | 1.62 | 5.68 | 7.11 | 1.35 | ||
Sericite | 0.00 | 0.00 | 0.00 | 0.00 | 0.27 | 0.00 | 0.00 | 0.00 | 0.00 | ||
Unstable (%) | Chlorite | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.54 | |
Augite | 0.53 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | ||
Diopside | 1.07 | 2.41 | 0.81 | 0.52 | 0.00 | 3.51 | 0.00 | 1.32 | 0.27 | ||
Enstatite | 0.27 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | ||
Hypersthene | 0.00 | 0.00 | 0.00 | 0.26 | 0.00 | 0.00 | 0.00 | 0.26 | 0.00 | ||
Hornblende | 0.00 | 1.34 | 1.08 | 2.87 | 0.00 | 0.00 | 0.27 | 0.00 | 11.35 | ||
Olivine | 2.67 | 0.80 | 1.89 | 3.66 | 2.70 | 0.54 | 0.00 | 0.26 | 0.00 | ||
Sphalerite | 0.00 | 0.27 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | ||
Calcite | 1.07 | 0.00 | 0.27 | 7.57 | 5.14 | 0.27 | 0.00 | 0.00 | 0.00 | ||
Dolomite | 0.00 | 0.00 | 0.54 | 1.31 | 0.54 | 0.00 | 0.00 | 0.00 | 0.00 | ||
Other(%) | Siderite | 0.00 | 1.61 | 0.81 | 5.74 | 11.08 | 6.22 | 7.84 | 0.53 | 1.35 | |
Glauconite | 0.00 | 0.00 | 0.00 | 1.83 | 3.51 | 0.00 | 0.00 | 0.00 | 0.00 | ||
Opaque | 18.93 | 24.40 | 18.87 | 21.41 | 17.03 | 42.16 | 36.49 | 26.58 | 33.78 |
Unstable and highly unstable minerals include zoisite, clinozoisite, epidote, biotite and hornblende. Other heavy minerals found in smaller amounts are diopside, hypersthene, olivine and, monazite (Table 2). Opaques vary between 26-33%,
DETRITAL ZIRCON U-Pb GEOCHRONOLOGY
One hundred and fifteen grains of a sublitharenite (EJD010) from the lower member in the Ocal Section were analyzed (Figure 7). The zircons are mainly sub-rounded with minor prismatic grains. The most significant age populations are Mesoproterozoic (1000-1200 Ma) and Middle Jurassic (172 Ma) (Figure 7). Other less-abundant populations are of ages greater than 1500 Ma.
From the upper member one hundred and nineteen zircons were analyzed (quartz arenite EJD007). Most of the zircons are reddish sub-rounded to prismatic, with low sphericity. The most significant age populations are Mesoproterozoic with peaks at 1020 Ma, 1143 Ma, 1489 Ma and 1586 Ma (Figure 7).
In the El Cobre section one hundred thirty-three zircons from a sublitharenite sample (EJD014) were analyzed. The grains are predominantly colorless to oxidized, with elongated and sub-hedral shapes, and minor sub-rounded zircons. The most significant age population is from the Middle Jurassic 171 Ma, with a minor Mesoproterozoic 1000 Ma age population (Figure 7)
In a quartz arenite sample (EJD020) from the upper member In Cobre the analyzed one hundred eleven zircons are mainly sub-rounded with low sphericity and prismatic forms. Most zircons are colorless and occasionally appear oxidized with reddish colors. The most significant age populations are defined of Permian (272 Ma), and Upper Jurassic (162.9 Ma). There are less statistically representative peaks of Mesoproterozoic age with ages from 1140 Ma to 1500 Ma.
A ca. 87 Ma single zircon grain was also found. This age is statistically Limited (single grain), and is younger than the accumulation age proposed for the Caballos Formation (Aptlan - Albian) (Figure 7) and is therefore considered as related to lead loss.
DIAGENESIS
The diagenetic characteristics of the Caballos Formation are described below. Given the similarities found in the two studied sections, results are grouped by member. In the lower member, mechanical compaction includes deformed muscovite as well as muscovite schist, mudstone and volcanic lithics, that also formed a pseudo-matrix. Although samples show different types of contacts, sutured contacts in quartz grains are associated to pressure solution.
K-feldspar, volcanic and sedimentary lithics exhibit partial dissolution. In the El Ocal section the dissolution processes are advanced, with feldspar and quartz crystals characterized by a bay habit.
Primary porosity is intergranular, whereas the secondary is of ntragranular and fracture character, the later locally filled with clay and organic material.
Cements are commonly surrounding grains or occluding [62]. The most abundant cement is epitaxial siliceous (opal and chert), the former Is found as coats at the the edge of the grains, where the later occludes the porous space. Ferruginous cement Is less abundant, and It cover the grains In a thin film, or Is found as an Intra-matrix pigment, arranged as powder and partially or totally occluding the porous space.
The least common cement Is calcareous cement, which fully or partially surrounds the grains In a polkllotopic texture, that locally develops Inciplently as sparite. The calcareous cement Is also found filling pre-existing fractures In minerals such as quartz.
The upper member Includes both pure slliclclastic and mix calcareous-sillciclastlc sandstones. The mixed sillciclastlc and calcareous rocks are characterized by quartz and feldspar with corrosion bays. On the other hand, the slliclclastic rocks exhibit slightly flexed moscovites and sedimentary lithics, and pressure solution in sutured quartz.
Dissolution Is present on grains of feldspar, plagloclase, as well as in the limestone cement, in some grains of glauconite and also in quartz grains and mudstone lithics. Some fragments of mudstone and feldspars are also found as Islands. Some of the k-feldspars grains are replace by mlcrite carbonates and fine clays. In the slliclclastic rocks porosity is mainly intergranular, with additionally fracture and dissolution related porosity. In the mixed rocks, the secondary intragranular and fracture porosity predominates, with additional glauconite dissolution and less common intergranular porosity.
Calcareous cement is the most abundant in the mixed rocks, which is present fully or partially surrounding the grains in a poikilotopic texture. It is mainly sparitic, although it's also found as micrite with a syntaxial texture filling pre-existing fractures in quartz.
The most common cement in siliciclastic samples is of siliceous composition (1-4%), which is found as non-continuous syntax opal along the entire edge of the grains, whereas chert cement is found as an occluding agent. Ferruginous cement (1-3%) is present as a film in the grains and as an intra-matrix pigment, partially or totally occluding the porous space. A summary of the main diagenetic processes that affect the Caballos Formation is presented in Figure 8.
PETROPHYSICAL ANALYSIS
The lower member from the El Ocal section has porosity values between 14-24%, which are greater than those obtained in the upper member (5-6%; Table 2). In the El Cobre section, porosity increases from base to top in the lower member (3-17%), similar to the upper member that varies between 3% and 29%.
The permeability results from El Ocal Section show values of 334.25 md, while for the upper member the permeability results were near zero values. In the El Cobre section, the lower member exhibits low values of 0.43 md, while the upper member varies from 273.4 md to 3.7 md from the base to the top (Table 2).
5. RESULTS ANALYSIS
PROVENANCE
In the two analyzed sections the lower member of the Caballos Formation is characterized by the abundance of monocrystalline quartz of a straight and ondulatory extinction, as well as by the presence of sedimentary and diffuse polycrystalline quartz with more than three grains and feldspar. Lithics include micaceous schists, volcanic, plutonic, mudstone and chert. This suggests a mixed sedimentary, felsic to intermediate igneous and metamorphic lithics of low to medium grade, and short residence time and rapid burial conditions.
The high content of ultrastable phases such as zircon, rutile and tourmaline is likely associated with the recycling of sedimentary rocks, while the presence of garnet, muscovite, sillimanite, pyroxene, hornblende and olivine is likely associated with meta pelitic sources and basic rocks.
The high content of quartz is likely associated with recycled sedimentary sources, as well as a contribution of igneous and metamorphic rocks in conditions of high weathering, as suggested by the relatively high argillaceous matrix contents (6%- 12%).
The upper member from both sections is characterized by an increase in the proportion of quartz (> 93%), and an associated decrease in monocrystalline quartz of straight extinction, at the expense of an increase in quartz with ondulatory extinction, and the appearance of polycrystalline polygonal quartz with 2 to 3 grains, and feldspars (orthoclase, microcline and plagioclase). The lithics are mainly of a sedimentary nature, with minor volcanic and schist lithics.
These characteristics also suggest mixed source areas. However metamorphic sources should be of a higher grade, as suggested by the presence of polygonal quartz and feldspars.
The presence of ultra-stable minerals is also compatible with crystalline and recycled sources, while the presence of apatite, garnet, micas, epidote, hornblende, olivine and pyroxenes is also an indicator of basic to acidic crystalline sources.
The compositional characteristics of the lower member in the Ocal Creek, together with the presence of Precambrian and Jurassic detrital zircons in the lower member, suggest that the source areas have characteristics that can be related to exposed rocks, to the east, similar to those present in the Garzón massif, Serranía de La Macarena and the Putumayo region, which include Mesoproterozoic metamorphic rocks of similar ages [19,20] and may be part of the broader Vaupes Arch, as well as Lower Paleozoic siliciclastic rocks (Sedimentitary rocks of Cerro Neiva, El Higado Formation the Limestones and Sands of Batalla, and the Güejar Group [63] The volcanic lithics and probably some of the quartz with a straight extinction, together with a population of Jurassic zircons could be related to plutonic or volcanic sources that are also exposed in the eastern segment of the Garzón Massif and the Upper Magdalena Valley [21,64].
The provenance change between the lower and upper members in the Ocal Creek section, which includes the absence of zircons from Jurassic sources and the great abundance of Precambrian zircons, as well as an apparent increase in quartz, including additions of higher-grade metamorphic rocks, could be reflecting a normal denudation process, in which rocks from higher levels of the crust disappear (such as volcanic and sedimentary cover) and crystalline rocks formed in deeper levels are exposed. The proportion of quartz material and the presence of sedimentary material suggests an increase in sources of sedimentary affinity, which could be associated with a migration to the east of the source area, in which Jurassic sources located in the westernmost area -similar to those present in the Garzón Massif and in the Upper Magdalena Valley- would cease to be significant, and are replaced by more eastern sources (Figure 9), such as those that are exposed towards the easternmost segments of the Garzón Massif and the Macarena region, where Precambrian gneissic rocks and Paleozoic sedimentary rocks are exposed [20].
In the case of the Cobre Creek, the compositional characteristics and detrital populations also Indicate a significant change between the lower and upper members. For the lower member, the extensive Jurassic population and the presence of clastic material of Igneous affinity suggest that rocks similar to those exposed In the Upper Magdalena Valley or the Central Cordillera [35] would be the source areas. The sedimentary contribution could be related both to the reworking of the Paleozoic sedimentary rocks exposed on the eastern flank of the Central Cordillera towards the Magdalena Valley, as well as the Jurassic sedimentary rocks associated with the volcanic rocks from the Saldaña Formation, as suggested by the presence of sandstone lithics with feldspar In the upper member the appearance of more quartz sources, as well as the presence of Permian and Precambrian zircons, could also be associated with a change In the source area, likely more localized to the west (Figure 9), where metamorphic Permo-Trlassic rocks and Paleozoic sedimentary rocks are likely exposed [35].
This pattern of denudation change from proximal to more distal source areas, that is also associated to the exposure of deeper crustal levels (Figure 9) is characteristic of an extensional environment In which the adjacent topography Is being eroded trough time [65] Intra-basln topographic highs have been Identified In the Upper Magdalena Valley as transversal zones in the Nelva sub-basin [66] and is seismic lines [4,5], that were possibly exposed and reduced as the basin deepened and erosion exhausted the topographic highs. Such tend suggest that the generation of topography and tectonic instabllity would have already been reduced among the upper and lower members of the Caballos Formation (Figure 10), thus marking the end of the tectonic Instability of the Lower Cretaceous.
RESERVOIR QUALITY
The Caballos Formation is a proven reservoir in the Upper Magdalena Valley and Putumayo, with a wide variation in its petrophysical oroperties, with porosity values below 2% in the northeastern part of the basin, and values of 25% towards the northwestern side [45,67,6]. Some authors have related this feature to the Lateral variation of the sedimentary fácies that Limit the continuity of the reservoirs [4,5]. Other authors have considered the presence of carbonate allochemicals and their relation with the basin's temperatures as a determining factor for the variations in the reservoir's properties in the eastern sector of the basin [6].
Additional petrological controls of reservoir quality includes the primary depositional texture (grain size and selection) and the original composition of the rock and the burial history.
Although grain size and selection influenced reservoirs [68], we didn't found any relation between the petrophysical properties and such attributes.
In the Lower and upper members of the Caballos Formation quartz rich rocks are characterized by higher porosity values when compare with those that have higher feldspar and Lithics contents with Lower quartz contents (Figure 11).
Qualitative and computational diagenetic models [69] suggest that the presence of quartz and feldspar cause Less compactation for a given Lithostatic pressure than the presence of matrix rich sandstones with schistose or even volcanic lithics. Feldspars might be also affected by dissolution, and although they could initially create secondary porosity, they would Lead to the development of clay cements or other cements that would inhibit the generation of secondary porosity.
With the increase in pressure during burial, chemical compaction also tends to increase, with porosity Likely decreasing due to the appearance of concave-convex contacts. It is suggested that the Lower porosity values are likely associated with the Lower member of the Caballos Formation, where the feldspars were probably transformed into clay and a pseudo-matrix was formed due to the compactation of labile volcanic, sedimentary and schistose lithics (Figure 11).
In the upper member of the Caballos Formation, the high porosity and the better permeability values are likely related to feldspar dissolution effects (Figure 11).
Dissolution occurred after burial and development of the siliceous cement, as this cement is cut by both carbonate and ferruginous cements. It is therefore suggested that this dissolution likely occurred in a late diagenetic state, probably during exhumation, when the chemical interaction with more superficial diagenetic fluids -which would include organic acids- would promote the dissolution of unstable minerals [70,71]. The Caballos Formation was subjected to different deformational phases during the Early Cretaceous, including the orogenic phases of the Albian-Cenomanian and the Cenozoic Andean Orogeny [45,72]. Such episodes promote basin exhumation, causing the development of secondary porosity.
CONCLUSIONS
Provenance results from the Caballos Formation in the Ocal Creek in the southern segment of the Upper Magdalena basins indicates that the source areas include mix sedimentary, acidic to intermediate igneous rocks, and low to medium-grade metamorphic rocks for both the lower and upper memebers of the Formations. These sources exhibit similar characteristics, with those exposed to the east in the Garzón Massif and the Serrania de La Macarena.
In the Cobre Creek, in the northern part of the basin, the sources are mainly igneous and sedimentary rocks that have similarities with the geology of the eastern flank of the Central Cordillera.
The increase in the proportion of quartz, together with the presence of sedimentary material indicate a major change between the lower and upper members. In the Ocal Section it's suggested that source areas were located farther east whereas for the Cobre section, source areas would probably be located more to the west.
This provenance model is consistent with the extensional model proposed for the Upper Magdalena Valley [1], which suggests that the basin is likely bound by paleo-highs to the east and west and within the basin [73] that were been erode giving place to more distal and deeper crustal sources.
Reservoir quality evaluation suggest that the petrophysical properties of the lower member of the Caballos Formation are controlled by the presence of feldspars and lithics as these components are transformed into clay and produce a pseudo-matrix that reduces the porous space. Whereas in the upper member, the high porosity and the better permeability values are likely related to feldspar dissolution effects during late diagenetic exhumation.