Hydrocarbon distribution along the Soapaga thrust(Eastern Cordillera,Colombia)based on new strategic geochemistry samples

Eduardo A.Rossello ·Rolando Di Primio

Abstract The Colombian Eastern Cordillera constitutes a region with potential for finding hydrocarbon reserves that are still under-explored,despite the existence of oil and tar sand production.The intense tectonic deformation affects the petroleum systems and increases exploration challenges due to the different generation,maturation,and entrapment conditions that they produce.Diverse geochemical analysis conducted on strategic samples determines that these are likely generated by the same anoxic marine source rock type.Two source rock samples we investigated are characterized by very different biomarker signals as compared to the tar sands and oil samples.Both samples are dominated by hopanes as compared to tri-and tetracyclic terpanes,with very low(Tibasosa Formation)to no(Chipaque Formation)extended tricyclic terpanes.The Soapaga Thrust,along which the analyzed samples are located,has exercised first-rate regional structural control of the distribution of hydrocarbons in the studied area.The uplift of their Mesozoic hanging wall produces intense erosion with sediment transport towards the eastern footwall forming the present infill of the Arcabuco basin.The kitchen areas were connected with the shallower entrapment sectors where excellent reservoir levels such as the Picacho Formation before the Soapaga Thrust movement.The high degree of biodegradation observed in the Picacho Fm.tar sands support early charge before the Soapaga thrust movement,and severe biodegradation following the thrust movement when hydrocarbon supply from the kitchen was interrupted.This work aims to contribute to a better understanding of the geochemical characteristics of the petroleum systems due to the action of the Soapaga Thrust.

Keywords Hydrocarbon geochemistry·Soapaga thrust·Eastern Cordillera·Colombia

1 Introduction

The Eastern Cordillera of Colombia exhibits in the Sogamoso region the proven presence of hydrocarbons from the production of oil from the Corrales field and the exploitation of tar sands from the Mina Santa Tereza and Emilia quarry.All of these hydrocarbon occurrences around the towns of Sogamoso,Pesca,and Tibasosa in the central portion of the Eastern Cordillera are located along the N-S trending Soapaga inverse Thrust(Fig.1).

Fig.1 Geological localization of the samples taken along the Soapaga Thrust.TsT:Tilata Formation;Tp:Picacho Formation;Kt:Tibasosa Formation;Ksc:Chipaque Formation;Kv2:Une Formation;Klm:Lutitas Macanal Formation;Jg:Giron Formation;Cc:Carboniferous;D:Devonian:Pc:Precambrian(taken from Rodrı´guez-Parra and Solano-Silva 2000).1:Corrales oil field,2:Mina Santa Tereza,3:Emilia quarry,4:Tibasosa and Chipaque Fms outcrops on the hanging wall

The tar sands from Picacho Formation in the Mina Santa Tereza and Emilia are mined for paving the local roads located in the municipality of Pesca,Boyaca´from two sites of exploitation named the Emilia quarry and Mina Santa Tereza(Higuera-Sandoval et al.2012;Go´mez-Rojas et al.2018).All of these petroleum manifestations are located along the footwall of the Neogene Soapaga Thrust that superimposes their hangingwall of Mesozoic units on tertiary units(Kammer 1996;Velandia 2005;Kammer and Sa´nchez 2006;Rodrı´guez et al.2009).

To evaluate the effect of this thrusting in the distribution of hydrocarbons,strategic samples from the available hydrocarbon localities are studied from a geochemical point of view and compared to stratigraphic sequences suspected of representing their source rocks on the hanging block of the Soapaga Thrust(Fig.2).

Fig.2 Indicative pictures.A:View of the surface Corrales oil field with a high-pressure stack can be seen in the background flaring associated gas(Fig.1-1).B:outcrops of sampled rocks(Fig.1-4).C and D:Examples of il seep in the Picacho Formations around Emilia quarry(Fig.1-3).E:Underground labors of the Mina Santa Tereza showing oil seeps(Fig.1-2).F:Detail of the sandstones of the Picacho Formation saturated with oil in the Mina Santa Tereza(Fig.1-2)

The objective of this work is to carry out an analysis and interpretation of the characteristics of the existing petroleum system in this still underexplored region of the Eastern Cordillera based on the geochemical characterization of manifestations in production and the influence of the typical thrusting that affects the continuity of the oil systems.In this way,the results are expected to contribute to the better planning of current exploratory activities that seek to determine conventional and unconventional resources.

2 Geological setting

The Eastern Cordillera of Colombia,an intra-plate mountain range,has formed mainly by Cenozoic shortening due to the Andean tectonic phases(see references in Horton et al.2020).However,Mesozoic strata predominate at outcrop(Fig.3).From Coletta et al.(1990)until Horton et al.(2020)several works provided good evidence for inverted rifts in the cordillera and new subsurface data from its foreland basins.Branquet et al.(2002),consider the Eastern Cordillera was produced by repeated reactivation of older thrust thrusts,the main mechanism of mountain building in the Andes and their foreland(Cobbold et al.2007).This has led to suggestions that rifted areas became weaker than their surroundings,or that normal faults reactivated as thrusts(Cooper et al.1995).However,at some localities,Mesozoic strata are unconformable upon folded and thrusted Palaeozoic sequences.There are good examples in the Guavio massif near the center of the cordillera(Branquet et al.2002),and in the Floresta Massif,near the eastern edge of the cordillera.

Fig.3 Chrono-stratigraphic chart showing the correlation of the Eastern Cordillera with the adjacent Middle Magdalena Valley and Llanos basins(from Horton et al.2020)

Regarding the lithologies present in the Eastern Cordillera as well as the adjacent Magdalena and Llanos regions these share a similar geological context in this portion of the South American continent crust(Cooper et al.1995).Paleozoic marine sedimentary rocks are locally preserved,but generally absent from several portions of the Eastern Cordillera which is dominated by Mesozoic-Cenozoic clastic basin fill(Go´mez et al.2017).These sequences deposited into extensional sub-basins are overlain by a more regionally extensive Lower Cretaceous marine succession of clastic and subordinate carbonate facies that directly rests upon isolated Jurassic deposits of the crystalline basement(Horton et al.2020).The Upper Cretaceous to Cenozoic clastic succession exhibits the transition from marine to nonmarine deposition and rapid accumulation during Andean shortening and flexural subsidence.The Cretaceous marine records are dated by invertebrate fossils(Etayo-Serna and Laverde-Montan~o 1985)and palynomorph assemblages provide age resolution within several million years for Cenozoic basin fill across Colombia(e.g.,Jaramillo et al.2011).Strategic isotopic ages for selected volcanic horizons and syndepositional volcanogenic zircons(Bayona et al.2012;Saylor et al.2012;Go´mez et al.2015;Anderson et al.2016)provided further age control.The Upper Cretaceou-Cenozoic stratigraphic intervals along the axial Eastern Cordillera,define broad upward coarsening packages with some internal variability that makes lithostratigraphic correlations difficult(Horton et al.2020).The Floresta Basin and smaller satellite sub-basins contain a partial stratigraphic record(ca.2000 m total thickness)that spans from the Upper Cretaceous through Oligocene(Bayona et al.2008;Saylor et al.2011;Ochoa et al.2012;Silva et al.2013).

The Lower Eocene Picacho Formation is characterized by white to coarse-to-fine-grained sandstones,high porosity,and permeability,which are impregnated with hydrocarbons.The origin of this Formation is continental;it was deposited in a possibly deltaic type environment,of great lateral extension(Reyes 1984).The Picacho Formation is equivalent to the Lenguazaque sandstones of Van der Hammen(1957,taken from Reyes,1984),which represents the lower sandy level of the Bogota´Formation and correlates well with the Mirador de Norte de Santander Formation.The Picacho Formation outcrops towards the left bank of the Pesca River(Fig.1-2)and its lithological composition is made up of predominantly sandstone packages with high porosity and permeability which store natural asphalt or bitumen,these rocks are mineralogically mature and reach up to 6 m thickness,followed by a thick pack of gray to pink clays with fine-grained sandstone intercalations.

3 Samples and methodology

The following strategic samplings along both sides of the Soapaga Fault were carried out(Fig.1):

(a) Oil sample from the Corrales oil field(Fig.2A),the Bolivar-1 well(5°48′25 N-72°52′08′′W),altitude:2570 m a.s.l.,total depth:4870 feet,was drilled in 1988 by Esso Colombia up to a depth of 3.700 ft.It presents a deviation up to 29°and a vertical depth from the rotary table of 3.555 feet MD=3.690)and has been in production since July 2005.For the La Luna Formation reservoir,the crude oil API gravity produced is 190°,bubble point was determined at 1457 psig@120°F and oil viscosity varies between 9.6 cp@210°F and 114 cp@90°F.The Corrales 1D reentry,Corrales-1 deep well,production tests(2011)with final depth were 6.253 feet(MD).it produced a 25°API oil from the Tierna sandstone of Guadalupe Formation and a 19°API oil under the Soapaga Thrust in the Plaeners Member of the Guadalupe Formation,reactivated from extensional Mesozoic structures(data taken from Sarmiento 2011;Pedraza-Fracica and Marin~o-Martı´nez 2016;Rangel et al.2017).

(b) Tar sands from the Mina Santa Tereza(Fig.1-1,5°34′19.661′′N-73°3′22.904′′).This is an underground mine of sandstones of the Picacho Fm.(Lower Eocene)with high bituminous oil contents(Fig.2E,F).

(c) Tar sands from the Emilia quarry(Fig.1-3,5°35′20.255′′N-73°1′25.881′′W).This is another tar sand in sandstones of the Picacho Formation(Lower Eocene)with high bituminous contents(Fig.2C,D).The impregnated sandstones are currently exploited as asphalt material,without having an estimate of their reserves and characteristics,due to the artisanal form of its exploitation.

· Rock sample Ft taken on the northern outcrop(dipping 70°towards East)of the Floresta-Corrales route (Fig. 1-4,5°49′52.227′′N-72°51′48.241′′W):Carbonaceous pelites of Tibasosa Formation (Aptian-Albian). Assumed source rock(Fig.2B).

· Rock sample FCh taken on the northern outcrop(dipping 75°towards East)of the Floresta-Corrales route (Fig. 1-5, 5°51′37.36′′N-72°50′19.72′′W): Carbonaceous pelites in interbedded levels of the Chipaque Formation(Coniacian).Assumed source rock(Fig.2B).

The rock samples Tibasosa and Chipaque were taken from outcrops in the hanging wall of the Soapaga Thrust,in contrast,the oil and tar sands were taken from the footwall of the Soapaga Thrust(Fig.1).

Geochemical analysis of oil samples and rock extracts was performed at the Helmholtz Centre Potsdam GFZ,German Research Centre for Geosciences Potsdam,Germany.Rock samples were extracted using a dichloromethane-methanol(99:1 vol/vol)solvent mixture by Soxhlet extraction(24 h).The Soxhlet extract was concentrated by a Turbovap system and dried under a stream of N2.Extracts and oil samples were separated into compound fractions by MPLC.Before GC-MS analysis 5a-Androstane and 1-ethylpyrene were added as internal standards.The saturate fractions of extracts and oils were analyzed using a Trace GC Ultra coupled to a DSQ mass spectrometer(Thermo Electron Corp.).The GC was equipped with a Thermo PTV injection system and an SGE BPX5 fused silica capillary column(50 m×0.22 mm i.d.and 0.25 mm film thickness).Helium was used as carrier gas.The GC oven was ramped from 50 to 310°C at a rate of 3°C/min and held at the end temperature for a further 30 min.The injector temperature was programmed from 50°C to 300°C at a rate of 10°C/s.The MS was operated in single ion monitoring mode(SIM).

4 Results

The interpretation presented below on based on the GCMS analysis of the saturate fraction,focussing on the m/z 191 Hopane biomarkers(Figs.4,5,6),but also including 25 nor-hopanes(m/z 177),and steranes(m/z 217).

Three different sample types were analyzed,tar sand extracts(Figs.4a,5a,6a),source rock extracts,and an oil sample(Figs.4b,5b,6b).

The TIC(total ion current)traces shown as insets in Fig.4 clearly show a complete lack of alkanes and isoprenoids in the tar sand samples,indicating significant levels of biodegradation.The hopane distributions of the tar sand samples are characterized by high contents of tricyclic hopanes,extending up to the(tentatively identified)C31-tricyclics,as compared to the pentacyclics(Fig.4a).Ts/Tm ratios are similar and indicate an early oilwindow maturity.All tar sand samples show the presence of Oleanane,supporting a Late Cretaceous or younger source rock age(Moldowan et al.1994).

Fig.4 m/z 191 traces of tar sand samples(a),source rock extracts,and the Corrales-1 oil sample(b).Figure insets show TIC traces of the respective samples

Moretanes are relatively prominent in these samples,also in the Mina Santa Tereza 31 extract,where theβα-C31 exceeds theαβ-C31 hopanes,indicating a relative resistance of the moretanes to biodegradation.The fact that Gammacerane andβα-C31coeluted on the column make the differentiation of the two compounds difficult.Mass fragmentograms of theβα-C31peak,as marked in Fig.4,indicate that gammacerane(mass fragment 412)may elute first andβα-C31(mass fragment 426)occurs as a shoulder on the main peak.For the interpretation here it is best to assume that the differentiation of the two compounds can not be made and that the peak marked asβα-C31 or Gammacerane actually contains both compounds.

Hopane contents decrease significantly in the Mina Santa Tereza#1 extract,while Oleanane contents increase relative to the C30 Hopane.The-R epimers of the C33 and C34 homohopanes are depleted as compared to the-S epimers, indicating they are more susceptible to biodegradation,as observed by Hoffmann and Strausz(1986),Requejo et al.(1989),and Lin et al.(1989).It has been hypothesized that this is due to the size and shape of the molecules and the respective effects in the transport rates across cell membranes(Peters et al.1996).In contrast,the-S epimers of the C26 to C31-tricyclic terpanes are significantly depleted in this sample.R versus S epimer identification of the tricyclic terpanes is based on a tentative interpretation following the recommendation of APT labs in Norway.We are not advocating that the epimer naming is correct,and only use R versus S identification of the tricyclic terpanes to differentiate between the two peaks.

Comparison of the m/z 191 traces of these samples to their respective m/z 177(Fig.5a)shows the occurrence of 25-norhopanes in all tar sand samples.The Mina Santa Tereza#1 sample shows also demthylated tricylic terpanes together with the full suite of 25-norhopanes derived from biodegradation of the hopanes.The presence of demthylated tricyclic terpanes points towards a severe level of biodegradation in this sample.In addition,the observation of the selective reduction of the-S epimers in the m/z 191 trace and enrichment of the-S epimers in the m/z 177 trace indicates the preferential biodegradation of the-S epimer over-R in the samples studied here.

Fig.5 m/z 177 traces of tar sand samples(a),source rock extracts,and the Corrales#1 oil sample(b)

The Emilia#1 and#2 tar sand extracts show a similar degree of biodegradation based on the tricyclic and pentacyclic hopane fingerprints(lower as compared to the Mina Santa Tereza#1 sample).However,the steranes(Fig.6a)show a significant change going from Emilia#1 to Emilia#2 and Mina Santa Tereza#1.Emilia#1 shows a clear dominance of steranes over diasteranes,Emilia#2 is strongly depleted in the regular steranes,with an apparent stronger reduction with increasing carbon number up to the C29 steranes.C30 steranes appear unaffected,as reported earlier by Lin et al.(1989).Diasteranes are slightly higher than in the Emilia#1 samples.In the Mina Santa Tereza#1 extract essentially almost only diasteranes remain,but also in lower amounts than in the Emilia#1 indicating that also these compounds are being biodegraded.The fact that indications of diasterane biodegradation are observed in the same sample as biodegradation of tricyclic terpanes points to a severe rank of biodegradation for this sample(Conan 1984),likely above 8 in the biodegradation scale of Wenger et al.(2002).

The m/z 191 fingerprints of the source rock samples,as well as the Corrales oil sample,are shown in Fig.4b.TIC traces are displayed as insets.The Tibasosa Fm sample(M FT)shows a weak TIC signal with a clear hump in the C12-16 range.Only minor alkane contents are monitored.The hopanes signal is also poor but allows the recognition of an early oil window maturity.Hopanes dominate over tricyclic terpanes,extended tricyclics are only present in traces.The Chipaque Formation sample(M FCH)shows dominance of long-chain alkanes in the TIC,as well as a strong m/z 191 signal,where the Ts/Tm ratio indicates high maturity.Here again,hopanes dominate over tricyclics,and only trace amounts of extended tricyclic compounds are monitored.The presence of Oleanane is of special interest,as this compound was also monitored in the tar sand samples.In contrast,the Corrales-1 oil shows a strong TIC signal with alkanes peaking at C22 and some indications of biodegradation of the sample based on the predominance of i-C15 and i-C16 isoprenoids over alkanes in the early elution range.The m/z 191 traces shows abundant tricyclic compounds,as observed in the tar sand extracts,dominating over pentacyclic terpanes.Ts/Tm indicates early oil window maturity.

Figure 5b shows only very low contents of 25-norhopanes or demethylated terpanes in the m/z 177 trace,and Fig.6b shows the presence of steranes and diasteranes in all samples.The Chipaque Fm sample shows here the highest maturity,whereas the Corrales#1 oil samples,as well as the Tibabosa Fm extract,show similar,lower maturity.The sterane distributions are,however,different for these two samples,with the Corrales oil showing dominance of C27 and C28 steranes over C29,while the Tibabosa Fm.sample sterane distribution is dominated by the C29 steranes.As compared to the tar sand extracts none of the samples shown in Fig.6b show signs of sterane biodegradation.

Fig.6 m/z 217 traces of tar sand samples(a),source rock extracts,and the Corrales-1 oil sample(b)

The tri-and tetracyclic terpane distributions,prominent in all tar sand and oil samples,are probably best for the correlation of these samples.Figure 7 shows a triangular diagram with the extended tricyclic terpanes ratio(Holba et al.2001)on one axis,C26 tricyclic/C24 tetracyclic terpane ratio on the second,and C24/C23 tricyclic terpanes on the third(Peters et al.,2005).Based on the tri-and tetracyclic terpane distributions all oil sand and oil samples are very similar and distinctly different from the two source rock extracts.

Comparison of all analytical results indicates that the tar sand extracts have all been significantly biodegraded,making a correlation to the source rock extracts(very different maturities)or the Corrales#1 oil sample difficult.The presence of Oleanane in all tar sand extracts can be,however,used as a marker.Of the source rock extracts only the highly mature Chipaque Fm.sample contains quantifiable Oleanane,making it a potential source rock of the petroleum in the tar sands.The Tibabosa Fm.sample is too lean to allow any conclusive correlation.The Corrales#1 oil terpane distribution is similar to that of the tar sands,with a predominance of extended tricyclic compounds,but is also characterized by the complete lack of Oleanane,as opposed to the tar sand and source rock extracts.

In general,the oil and tar sand samples investigated show very similar characteristics,masked in part by biodegradation.The absence of Oleanane in the Corrales#1 oil can be used to argue for a different source for this sample,but may also only reflect differences in the depositional environment of the same source rock.

5 Discussion

During the Mesozoic,several sub-basins were generated limited by normal thrusts that controlled the thinning of the crust,the stretching of the lithosphere,and the beginning of an intra-arc basin as the Arcabuco rift in the studied area(Fig.8).From the end of the Cretaceous and in the Paleogene Andean compressional phases generates the uplift of the Eastern Cordillera that reaches values of more than 3 km in height,causing the erosion and transport of the sediments that cover the tectonic depressions on the flanks and interior of the Eastern Cordillera.During this tectonic event,significative inversion in some of the principal normal faulting as the Soapaga Thrust is generated(Villar et al.2017).

The Mesozoic-Cenozoic tectonosedimentary history of marine and nonmarine sedimentation recorded in the Eastern Cordillera reflects the action of different tectonic regimes.(i)Jurassic to earliest Cretaceous extension produces the development and linkage of extensional halfgraben controlled by normal faults(Toro et al.2004;Teso´n et al.2013).(ii)A post-extensional thermal subsidence phase generated a thermal sag basin across a broader region.(iii)Initial crustal shortening during the Latest Cretaceous to Paleocene,created a regional flexural basin that was successively broken by the Paleocene-Oligocene emergence of thrust/reverse-fault-related uplifts.(iv)The Neogene tectonic uplift produced a progressive shortening(commonly involved contractional reactivation of preexisting normal faulting)with the establishment of an effective starting of thick molasse deposits located in the bivergent eastern and western flanks of the Eastern Cordillera(Teso´n et al.2013).

The horizontal shortening produces the tectonic uplift and the consequent erosion of the Floresta massif above the Soapaga thrust and outcrops different metamorphic grade rocks that include Devonian and Jurassic sequences(Fig.8B).Based on regional unconformities of the Paleocene,Eocene,and Early Oligocene,compression continues during the Tertiary until the Present producing the reactivation of the Soapaga Thrust as a thrust and the inversion of the Arcabuco rift in a series of subordinate fold a thrust(Fig.8C).Kammer(1996)calculates a vertical rejection of the Soapaga Thrust on the eastern flank of the Floresta Massif greater than 5 km,considering it one of the most important discontinuities in the Eastern Cordillera.As a consequence of the tectonic displacement related to the Soapaga Thrust(Villar et al.2017),Mesozoic units as Chipaque or Une formations(the source rocks)are in contact with the Neogene Picacho Formation and equivalents(the reservoir rocks)facilitating the migration of the hydrocarbons.

Fig.8 Schematic evolutive geological transect across the Soapaga Thrust through Floresta massif(Eastern Cordillera of Colombia 06°50′N based on Cobbold 2007)showing the main sedimentary sequences related with the petroleum system.G:Granite;D:Devonian;C:Carboniferous;J1:Lower Jurassic;J2:Middle Jurassic;J3:Upper Jurassic;K1:Lower Cretaceous;K2:Middle Cretaceous;K3:Upper Cretaceous,T1:Lower Tertiary;T2:Middle Tertiary;T3:Upper Tertiary

Fig.7 Triangular diagram of selected triterpane ratios.M FT:Tibasosa Formation;M FCh Chipaque Formation

There are multiple manifestations of liquid hydrocarbons on the surface in the area,in both Paleogene,Neogene,and Cretaceous units,which indicates that the source rocks reached hydrocarbon generation conditions(ANH 2008;Rangel et al.2017).The geochemical data of the well and outcrop indicates the presence of humic-terrestrial organic matter of kerogen type III,for the entire tertiary sequence(Formations Concentracio´n,Socha Superior and Guaduas),and the presence of organic matter of the marine origin of kerogen type II,for the Cretaceous shales sequence(ANH 2008;Lozano and Zamora 2014).

Aguilera et al.(2010)interpreted the source rock geochemical information for the Eastern Cordillera Basin,including TOC and Rock-Eval Pyrolysis data from 1512 samples taken in 9 locations;additionally,369 organic petrography samples from 8 locations were studied.Crude oil and extracts information from 4 bulk analysis samples,111 liquid chromatography samples,114 gas chromatography samples,125 biomarker samples,42 isotopes,and 349 surface geochemistry samples were also interpreted.The results document that heavy oils with API gravities below 20°and sulfur content above 1%are present in the basin.There is a correlation between sulfur and API gravity,indicating that the higher the API gravity the lower the sulfur content and hence crude oil quality,likely reflecting the effect of biodegradation.The few crude oils reported in the basin suggest that API gravity should increase with depth and that hydrocarbons could be found relatively shallow in the basin.The sulfur content of the oils is above 1%,and its Ni/V ratio is below 1,suggesting that they are produced from rocks deposited in a marine suboxic to the anoxic environment.The data obtained from pyrolysis Rock-Eval of rock samples for Hydrogen Index(HI)and S2peak indicate that samples from the Cretaceous Caballos,Conejo,La Luna,Villeta,Guadalupe,Los Pinos,and Umir formations and the Cenozoic Arcillas de Socha Formation have good generation potential(HI＞200 mg HC/g TOC and S2＞5 mg HC/g rock).It is important to consider that these and other units with source rock characteristics,are or were deeply buried in the basin by thrusting,and the poor generation values obtained from many samples could reflect the depletion effect caused by the high thermal maturity reached by these rocks in subthrust sheets.

The Oxygen Index versus Hydrogen Index diagram(Van Krevelen diagram)shows that rock samples from the Cretaceous Caballos,Conejo,La Luna,Villeta,and Umir formations generally have type II oil-prone kerogen.In the case of the Cenozoic units(Guaduas,Concentracio´n and Bogota´formations)their samples are indicative of type III gas-prone kerogen to type IV kerogen.The Tmax maturity parameter vs Hydrogen Index graph shows that many samples from the Cretaceous to Cenozoic units mentioned have reached early maturity to overmature conditions in the basin.The samples from the Cretaceous(e.g.Fomeque,Chipaque,and Hilo´formations)show the highest maturities in the basin.Organic carbon content(%TOC)and S2 peak values indicate oil generation potential,indicating samples from the Cretaceous units(e.g.Caballos,Villeta,La Luna,and Umir formations)and the Cenozoic Arcillas de Socha Formation,with good to excellent oil generation potential(S2 up to 50 mg HC/g rock and%TOC up to 9).The vitrinite reflectance(%Ro)data shows that the sedimentary sequence is mature to overmature in the basin,with variable maturity trends caused probably by different burial and thermal histories controlled by the structural development of the Eastern Cordillera(Rangel et al.2017).

Aguilera et al.(2010)consider the Cretaceous rocks of the Caballos,Conejo,La Luna,Villeta,and Umir formations and the Cenozoic Arcillas de Socha Formation as the best source rocks at the basin,with good to excellent oil generation potential intervals.Tmax maturity data indicates that the Cretaceous oil-prone formations are mature and that the high thermal maturity reached by some source rocks,could produce crude oil with better characteristics than that already found.There are no good correlations reported between the few crude oils analyzed and source rock extracts in the basin.Only vague and very general comparisons have been made,for example,that the crude oil in the Picacho Formation has higher C29 steranes concentration than the rock extracts from the Guadalupe Formation,indicating more terrestrial organic matter input.In general,however,the C35/C34 Hopanes,Ts/(Ts+Tm)and diasteranes/steranes of reservoir rock extracts indicate clay-poor source rocks deposited under highly anoxic conditions.

The hydrocarbons from the Bolı´var-1 well are of low API gravity(18°)in rocks corresponding to the Lower Guadalupe Formation(Pedraza-Fracica and Marin~o-Martı´nez 2016).The maturity data and geochemical models suggest that the shales of the Chipaque and Une Medio Formations would be in the early hydrocarbon generation window(%Ro 0.55-0.7)and the shales of the Tibasosa Formation would be in the mature generation window(%Ro 0.7-1.3)in the deepest parts(Lozano and Zamora 2014).These authors took into account a poor amount of organic matter(0-2%TOC),temperatures between 430 and 450°C and%Ro of 0.8-1 of surface samples,from the Corrales well,and the Cordillera pseudo well(Sarmiento 2011)that crosses the sedimentary sequence in one of the deepest places within the basin.The modeling results show that the sequences reach maximum temperatures and depths at the end of the Miocene/earliest Pliocene(Fig.9).

Fig.9 Buried-history chart estimated for the Cordillera Oriental(performed by PetroMod Software:1D Petroleum System Modeling).Insert:Cordillera pseudo well(taken from Sarmiento 2011)

Pedraza-Fracica and Marin~o-Martı´nez(2016)calculated the following paleotemperatures in the Corrales well based on vitrinite reflectance and compared results with the present geothermal gradient.In their results,they concluded that early Cretaceous sediments reached a depth of 9000 m and a temperature of 200°C at the maximum burial.Late Cretaceous sequences reached a depth of 7000 m and 160°C.Paleocene sediments reached a depth of 4600 m and a temperature of 125°C,Eocene reached a depth of 2300 m and a temperature of 88°C,Oligocene reached a depth of 1600 m and a temperature of 52°C and finally Miocene sediments reached a depth of 700 m and a temperature of 24°C.

The interpretation of these comparisons was based on hypothetical cases of estimation of the geothermal gradient and thermal profiles associated with the flow of fluids in active geothermal systems.As a result,the paleogeotherm seems to indicate that burial was the key factor in changing the rank of the organic matter in this well.Other wells in the area show a higher than normal gradient,which could indicate enhanced heat flow at the base of the basin.In most wells,an irregular paleogeotherm could be explained through the hydrothermal process associated with intrusions and paleo regional flow.These results support the observed maturity anomalies in the central part of the Eastern Cordillera Basin and indicate a complex interaction between geological phenomena such as faulting,hydrothermal fluids,burial,and orogeny in the past affecting the petroleum system.

The rock and raw geochemical data suggest the existence of a petroleum system with source rock in the Chipaque Formation and storage rock in the Middle and Lower Guadalupe Formations,and the possibility of petroleum systems in Lower Cretaceous rocks such as Tibasosa-Une(?).In this burial model Pedraza-Fracica and Marin~o-Martı´nez(2016),observed that the sequence reaches,as in the case of the depocenter of the Corrales-1 well,with a maximum burial at 23 million years and from this moment on,an inversion process began that gave rise to the anticline structure observed today,in which the well was drilled,and the erosion of practically all the tertiary sediments.The most probable generation scenario for the depocenter immediately east of the Floresta Massif,suggests that the Lower Cretaceous rocks(Tibasosa Formation)could enter the hydrocarbon generation window in the late Cretaceous during a period of great subsidence related to the deposit of the Guaduas Formation,and from this moment until about 23 million years ago this depocenter continued to subside,at which time the sedimentary sequence would reach its maximum burial and the sequence of Cretaceous rocks its highest degree of thermal maturity,with most of this sequence in early window of generation and the Tibasosa Formation in a late stage of maturity for liquid hydrocarbons.

6 Conclusion

Based on the results of the GCMS analyses performed on tar sand extracts,source rocks,and oil the following conclusions are drawn:

(a) All tar sand samples contain heavily biodegraded oil.Emilia#1 and#2 are very similar and contain Oleanane(land plant marker).Emilia#2 appears more biodegraded than Emilia#1 based on the steranes.The Mina Santa Tereza#1 extract also contains Oleanane,but shows the highest degree of biodegradation,with obvious biodegradation of diasteranes and tricyclic terpanes.The common occurrence of Oleanane in all tar sand extracts,as well as the main tricyclic and tetracyclic terpane ratios of all extracts,indicate a strong similarity among these samples.

(b) The Corrales-1 oil sample is only slightly biodegraded and shows an early oil window maturity,similar to that of the tar sand samples.The Corrales oil sample has,however,no Oleanane.High amounts of tri-and tetracyclic terpanes as compared to hopanes as well as the main tri-and tetracyclic ratios compare favourably to the tar sand extracts.

(c) The source rock samples are characterised by very different biomarker signals as compared to the tar sands or oil samples.The two source rock samples show significantly different maturities,with the Chipaque Formation sample being highly mature,and the Tibasosa Formation sample showing an early oil window maturity.Both samples are dominated by hopanes as compared to tri-and tetracyclic terpanes,both samples show very low (Tibasosa)to no(Chipaque)extended tricyclic terpanes.Only the Chipaque Formation sample contains quantifiable amounts of Oleanane.Both samples are also characterised by a strong C24 tricyclic terpane signal,in stark contrast to the tar sand and oil samples.Due to these differences in biomarker signals,a correlation of the source rock extracts to any of the other samples is not possible.

The Corrales oil and Emilia tar sands were very likely generated by the same source rock type.The Corrales oil likely represents a slightly more marine,and less anoxic source rock facies than the samples from Emilia.The Mina Santa Tereza sample is likely the same as Emilia,but more biodegraded.The source rock samples(Tibalosa and Chipaque Fms)are most likely not representative of the source which generated the Corrales oil and the Emilia and Mina Santa Tereza tar sands.The sandstones of the Picacho Formation represent the best reservoir in the area.

The rock samples of Tibasosa and Chipaque formations were taken from outcrops in the hanging wall of the Soapaga Thrust,in contrast,the oil and tar sands were taken from the footwall of the Soapaga Thrust.The Soapaga Thrust has exercised first-rate regional structural control of the distribution of hydrocarbons in the studied area,facilitating the exhumation of previous kitchen areas into their hanging walls and severed the connection to shallower entrapment sectors with excellent reservoir levels of the Picacho Formation,leading thus to the cease of charge and ensuing strong biodegradation with no active re-charge of the Picacho reservoirs.

AknowledgementsOur thanks to Prof.H.Fonseca from the Technological and Pedagogical University of Sogamoso(Colombia),for his valuable contribution to collect the analyzed samples.Dr.Kai Mangelsdorf and Cornelia Karger from GFZ Potsdam(Germany)are kindly acknowledged for analytical support and discussion.

Declarations

Conflict of interestThe authors declares that they have no conflict of interest.

Availability of data and materialAll data generated and analyzed during this study are included in this published article and its supplementary information files.