Marine and Petroleum Geology xxx (2017) 1e14

Contents lists available at ScienceDirect

Marine and Petroleum Geology

journal homepage: www.elsevier.com/locate/marpetgeo

Research paper

Hydrocarbon seeps in Romania: Gas origin and release to the

atmosphere
Calin Baciu a, Artur Ionescu a, b, *, Giuseppe Etiope c, a
a
Faculty of Environmental Science and Engineering, Babes-Bolyai University, Cluj-Napoca, Romania
b
MTA-ELTE Volcanology Research Group, Budapest, Hungary
c
Istituto Nazionale di Geofisica e Vulcanologia, Sezione Roma 2, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Romania is one of the countries with the largest number of surface hydrocarbon seeps in the world.
Received 6 October 2016 Seeps may be an important tool for petroleum exploration as they can provide useful information
Received in revised form regarding source rock maturity, reservoir quality, and secondary gas alterations. Seeps also represent an
11 June 2017
important source of methane, ethane, and propane to the atmosphere. To date, the genetic character-
Accepted 12 June 2017
Available online xxx
ization of natural gas in Romania has only been based on molecular composition, without isotopic in-
formation. Here, we present an overview of investigations performed over the past 15 years for the main
Romanian hydrocarbon seeps, and report the molecular and isotopic compositions of gas, and the fluxes
Keywords:
Hydrocarbon seeps
of methane, ethane and propane to the atmosphere. We assessed gas origin and secondary alterations in
Mud volcanoes 17 seeps from several Romanian petroleum systems, and potential source rock types and maturity have
Gas geochemistry been evaluated. As previously inferred, gas within the Transylvanian Basin is largely microbial, but also
Methane displays indications of a minor thermogenic component that is likely related to a deep petroleum system.
Romania Carpathian Flysch and Foredeep petroleum systems contain thermogenic gas, with clear evidence of
biodegradation in some cases. Thermogenic gas generation modelling and maturity plots suggest that
most Romanian gases originate from mature type II and III kerogen (%Ro: 2e3). For cases of high flux
seeps, gas has the same hydrocarbon molecular composition as the reservoir, while in weaker seeps and
some mud volcanoes gas is altered by molecular fractionation (a loss of C2 and C3 during gas migration).
Gas seep geochemistry, in general, reflects the different geological and maturity conditions of basins
where seeps are located. A vertical sequence of petroleum systems has been suggested in some basins by
seeps displaying different maturity and secondary alterations. Measurements of methane flux to the
atmosphere from 94 seeps display a wide range of emissions (kilograms to hundreds of tons per year),
with a total, conservative estimated methane emission of approximately 3000 t y
1. Microseepage may
also release a similar quantity of methane. Consequently, seepage is a substantial contributor to natural
emissions of methane on a national level.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction
Although its available petroleum reserves are relatively small on
a global scale, Romania is still one of the most important European
oil and gas producers (BP, 2016). Romania is also among the
countries with the longest tradition of petroleum production in the
world. With more than 150 years of oil extraction and more than
100 years of gas extraction, approximately five billion barrels of oil
* Corresponding author. Faculty of Environmental Science and Engineering,
Babes-Bolyai University Cluj-Napoca, Romania.
E-mail address: artur.ionescu@ubbcluj.ro (A. Ionescu).
and 40 trillion cubic feet of gas have been cumulatively produced in
Romania. Over this period of time, a large amount of geological and
reservoir data have been acquired. A first synthesis for the
geochemistry of natural gas in Romania was published by Filipescu
and Huma (1979) who presented a geochemical database for main
hydrocarbon fields based on gas-chromatographic analyses
together with short geological descriptions of hydrocarbon-bearing
structures. No isotopic data were reported, and this type of data has
been missing from Romanian literature for three successive
decades.
The first isotopic data related to Romanian hydrocarbons were
published in 2008 for the methane seep of Gheraiesti-Bacau,
located in eastern Romania (Baciu et al., 2008). A few additional

http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
0264-8172/© 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
2 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14

papers containing isotopic data followed (Etiope et al., 2009a,
2009b; Frunzeti et al., 2012), although, to date, a comprehensive
dataset for petroleum systems in Romania is still missing from the
scientific literature. Due to intense tectonics caused by Carpathian
orogenesis and post-orogenic basin uplift (e.g., the Transylvanian
Basin), numerous hydrocarbon seepage systems (as defined by
Abrams, 2005) developed, making Romania one of the countries in
the world with the largest number of surface hydrocarbon seeps
(470 seeps have been catalogued; Etiope, 2015; Ionescu et al., 2017).
Surface manifestations include gas seeps (sometimes forming
“everlasting fires”), mud volcanoes, oil seeps, and methane-rich
water springs. A distribution of documented seeps is provided in
Fig. 1.
Seeps, generally linked to subsurface pressurized reservoirs, are
an important tool for petroleum exploration and may provide
useful information related to source rock maturity, the quality of a
reservoir, and secondary alterations (Link, 1952; Etiope et al.,
2009a, 2009b). Seeps are also important natural sources of green-
house gas (methane) and photochemical pollutants (ethane and
propane) to the atmosphere (e.g., Etiope and Milkov, 2004; Etiope,
2009; Etiope and Ciccioli, 2009). On a global scale, seepage in
sedimentary basins and gases released from geothermal manifes-
tations represent the second natural source of methane to the at-
mosphere after wetlands (Etiope, 2012, 2015). Geological gas
sources in Romania should also be assessed on a national level
because they appear to be a major contributor of natural methane
emissions in Europe.
In the context of the depletion of reservoirs, significant interest
has been given to the intensification of exploration for both
conventional and unconventional oil and gas resources. Modern
geochemical investigations, including isotopic data, are recognized
as valuable tools for more efficient petroleum exploration and for a
better understanding of the genesis and evolution of hydrocarbon
accumulations. Based on an investigation of a wide set of surface
manifestations, this research provides insight into the geochem-
istry of some of Romania's petroleum systems. In this work, and for
the first time, we present all of the molecular and isotopic data
available for 17 major Romanian seeps (Fig. 1). The data have been
gradually obtained by our research team since 2000 and have only
been partially published. In the following, we compare the mo-
lecular composition of seeping gas with that of reservoirs located
within the same petroleum system, for the main productive basins
in Romania, and shed new light on gas origins, secondary alter-
ations, and potential source rock types and maturity. The data
include measured methane fluxes from 17 seeps where we per-
formed geochemical analyses, and fluxes from 77 additional seeps
(Supplementary Material, Table S1). Using these data, we devel-
oped a wide flux dataset from which emission factors could be
derived for estimating total seepage output for the country. Based
on the molecular composition of seeping gas, total emissions were
also determined for the photochemical pollutants ethane and
propane.
2. Geological setting and description of the seepage sites
2.1. General geological setting
The Romanian territory overlaps a large portion of the

Fig. 1. Map showing the distribution of the documented seeps in Romania (blue stars). Oil fields in green, gas fields in red. Geological provinces (Pawlewicz et al., 2002): 1013 e
Ukrainian Shield; 1103 e Dobrogea Foreland; 4047 e North Carpathian Basin; 4048 e Pannonian Basin; 4057 e Transylvania; 4061 e Carpathian-Balkanian Basin; 4063 e Dobrogea
Orogen; 4064 e Western Black Sea Basin. Seeps investigated in detail in this work (violet circles): 1 e Andreiasu; 2 e Raiuti; 3 e Lopatari; 4e Lepsa; 5 e Andreiasu MV; 6 e Beciu; 7
e Paclele Mari; 8 e Paclele Mici; 9 eFierbatori; 10 e Alimpesti; 11 e Sacelu; 12 e Sarmasel; 13 e Deleni; 14 e Tauni; 15 e Praid; 16 e Homorod; 17 e Bacau. (For interpretation of
the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14 3

Carpathian Orogen and portions of the Moesian, Scythian, and
Eastern European Platforms (Sandulescu, 1984). All of these areas
contain hydrocarbon-bearing formations (Fig. 2). The basement of
these three platforms consists of Precambrian metamorphic rocks
overlain by a sedimentary cover, including Palaeozoic, Mesozoic,
and Tertiary formations. The margins of the platforms dip towards
the orogen, where foredeep basins filled with thick Tertiary sedi-
ments developed.
The Moesian Platform is a crustal block that was deformed by
Hercynian movements during the Late Devonian to Early Triassic.
The metamorphic basement is covered by sediments belonging to
the following four deposition cycles (Sandulescu, 1984): (a) the
OrdovicianeCarboniferous cycle, that is largely detrital; (b) the
PermianeTriassic cycle, with a carbonate median sequence and
detrital lower and upper portions; (c) the JurassiceCretaceous cy-
cle, with a detrital portion in the Middle Jurassic and carbonate
rocks in remaining areas; and (d) the Tertiary cycle, a detrital mo-
lasse that is Middle Miocene to Pleistocene in age. The Peceneaga-
Camena and Trotus Faults separate the north-east Moesian Plat-
form from the North Dobrogea Orogen and the Scythian Platform.
A limited portion of the Scythian Platform is located in Roma-
nian territory between the Moesian and the Eastern European
Platform. The Caledonian crystalline basement is covered by a
Paleozoic sedimentary cycle (up to the Carboniferous), and is fol-
lowed by the PermianeLower Triassic molasses and thick Jurassic
sediments. Relatively thin Cretaceous and Tertiary deposits occur
on top of the sedimentary pile.
The Moldavian Platform, representing the south-western
portion of the Eastern European Platform, is only found in
Romania. The crystalline basement is covered by three sedimen-
tation cycles: (a) the Vendian-Carboniferous cycle, largely con-
taining detrital deposits, shales, and limestones within the upper
portion; (b) the JurassiceCretaceous cycle that is largely calcareous;
and (c) the Miocene cycle, with detrital-carbonate deposits that
largely spread over the area (Mutihac, 1990). The Pericarpathian
Thrust and its equivalents delimitate the Moesian and Moldavian
Platforms from the Carpathian Foredeep, the most external unit of
the Carpathian belt.
The Romanian section of the Carpathians was generated as a
result of the subduction and continental collision of the Eastern
European/Scythian/Moesian platforms and several crustal blocks
(Tisza, Dacia, ALCAPA) during Alpine orogeny (Sandulescu, 1984;
Sandulescu and Visarion, 1988). The Carpathians largely consist of
a fold and thrust belt that includes a system of intensely deformed
nappes. The central unit of the Eastern Carpathians, corresponding
to the Median Dacides, consists of metamorphic units and marine
Mesozoic sediments, and is bordered to the west by the Neogene
volcanic belt and to the east by flysch units. Flysch deposits,
structurally corresponding to the Outer Dacides and the Molda-
vides, are involved in complex overthrusted structures. Evolution of
the flysch basin began during the early Cretaceous, continuing
during the Paleogene. Molasse-type deposits were formed during
the Miocene-Pliocene. Beginning with the Eocene, contempora-
neous with a portion of the flysch and molasse of the Eastern
Carpathians, a thick pile of sediments was deposited within the
South Carpathian Foredeep, morphologically corresponding to the
Getic Depression.
The Transylvanian Basin began to form during the Late

Fig. 2. General geological map of Romania, showing the main structural units. Modified after: Geol. Inst. 1967, Sandulescu, 1984, Mutihac 1990. EEP e Eastern European Platform;
Scy e Scythian Platform; MP e Moesian Platform; NDo e North Dobrogea Orogen; CDo e Central Dobrogea; SDo e South Dobrogea; PB e Pannonian Basin; TB e Transylvanian
Basin; AM e Apuseni Mountains; VR e volcanic belt of Eastern Carpathians; MD e Median Dacides; OD e Outer Dacides; Md e Moldavides; CF e Carpathian Foredeep; SC e
Southern Carpathians; 1 e Quaternary deposits; 2 e Neogene; 3 e Paleogene; 4 e Cretaceous; 5 e Triassic þ Jurassic; 6 e Neogene volcanics; 7 e Pre-Tertiary magmatites; 8 e
Precambrian þ Palaeozoic (metamorphites and sedimentary rocks).

Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
4 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14
Cretaceous. The most important accumulation of sediments
occurred during the Miocene. In this area, post-tectonic basin fill
may exceed 8 km in thickness (Cranganu and Deming, 1996). An
important evaporitic event occurred during the Middle Miocene,
and is responsible for the almost continuous salt layer located
throughout the basin, that is 200e400 m in thickness and that
generated diapirs on the sides.
The Pannonian Basin is only found in Romania in its easternmost
portion. This large compound Neogene Basin is separated from the
Transylvanian Basin by the Apuseni Mountains. Although similar in
many ways, the Pannonian Basin, in contrast to the Transylvanian
Basin, has a thin crust and a high heat flow important for hydro-
carbon generation.
2.2. Romanian total petroleum systems
According to Popescu (1995), approximately 30% of the Roma-
nian territory is covered by 18 petroleum systems that range from
Palaeozoic to Pliocene in age. Several Total Petroleum Systems
(TPS), spanning hydrocarbon-bearing areas in Romania, have been
defined, including the Moesian Platform Composite TPS, the Dys-
odile SchisteTertiary TPS (Pawlewicz, 2007), the Transylvanian
Composite TPS (Pawlewicz, 2005), and part of the Greater Hun-
garian Plane Neogene TPS (Dolton, 2006). In a broader sense, the
first two systems belong to the Carpathian Balkanian Basin Prov-
ince (Pawlewicz, 2007), that extends into northern Bulgaria and
southern and eastern Romania.
In the Moesian Platform, hydrocarbon deposits associated with
Devonian, Triassic, Jurassic, Cretaceous, and Upper Miocene for-
mations have been identified (Paraschiv, 1979). Unconventional
shale gas potential is inferred in the Middle Jurassic and Silurian-
Lower Devonian formations (Tari et al., 2011). The Moesian Plat-
form is bordered to the north by the Carpathian Foredeep, that
corresponds to the Tertiary TPS defined by Pawlewicz (2007). This
TPS continues along the Eastern Carpathians, beyond Romania's
northern border in Ukraine, Poland, and the Czech Republic. Oil and
gas deposits are located within the Paleogene Flysch Zone, espe-
cially within the Carpathian Foredeep that hosts the most impor-
tant oil resources of Romania. Oligocene marls, menilites, and
dysodiles are considered to be the main source rocks within the
flysch zone.
Two superposed petroleum systems have been inferred within
the Transylvanian area: (1) a deep thermogenic Mesozoic system,
and (2) a shallower system that is biogenic, and Miocene in age
(Krezsek et al., 2010). Few data regarding the Mesozoic system are
available, but its importance for hydrocarbon resources appears to
be limited. Natural gas within the main petroleum system was
generated by Miocene source rocks and is hosted in formations of
the same age (Pawlewicz, 2005). The Transylvanian Basin is the
first gas producing region in Romania and has a cumulative pro-
duction of approximately 30 TCF over more than 100 years
(Krezsek, 2011).
In the Moldavian Platform, few conventional hydrocarbon de-
posits have been identified, mainly confined to Miocene forma-
tions. In the Scythian Platform, hydrocarbon deposits are located
within Palaeozoic and Neogene formations. Relatively little infor-
mation regarding Romanian unconventional hydrocarbon plays is
currently available. According to the US-EIA (2015), the Carpathian
Foreland has significant shale gas development potential within its
Silurian black shales.
2.3. A description of seepage sites
A total of 17 hydrocarbon seeps (Fig. 1), including gas seeps, mud
volcanoes, and hydrocarbon-rich springs were investigated for
Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
their molecular and isotopic composition and gas flux, as follows:
(1) five seeps within the Transylvanian Basin (the Sarmasel “ever-
lasting fire”, Deleni, Praid, and Tauni gas seeps, and the Homorod
mud volcano), and (2) twelve seeps located in external areas of the
Carpathians within the flysch and foredeep zones of the Eastern
Carpathians (the Andreiasu, Lepsa, Lopatari, and Raiuti “everlasting
fires”, and the Andreiasu, Beciu, Fierbatori, Paclele Mari, and Paclele
Mici mud volcanoes), in the Southern Carpathian Foredeep (the
Alimpesti mud volcano and the Sacelu methane-rich spring), and in
the Moldavian Platform (the Bacau gas seep). These seeps are
briefly described below. Representative pictures are provided in the
Supplementary Material section (Figs. S1 and S2). Methane flux was
also measured in 77 additional seeps, in areas widespread within
the Transylvanian Basin, in the Carpathian Flysch and Foredeep,
and in the Moldavian Platform.
2.3.1. External Carpathian Flysch and Foredeep
The most spectacular everlasting fires occur within Paleogene-
Neogene units positioned in the southern portion of the Eastern
Carpathians, and are located relatively close to one another in
Andreiasu, Raiuti, Lopatari, and Lepsa. The famous fire from
Andreiasu (Fig. 1, no.1; Fig. 3A) consists of several vents scattered
over an area of approximately 450 m2 and generates multiple
flames that can reach heights of 0.5e1 m. A mud volcano (Fig. 1, no.
5) is located approximately 500 m NE of the everlasting fire on the
flank of a nearby hill. The mud volcano is a small bubbling muddy
pool, roughly 1 m in diameter.
The Raiuti fire (no. 2) is located near Reghiu village and sits
almost atop of the Raiuti peak. Some small vents, generating weak
flames, occur within an excavation in an area of 6 m2. The Lopatari
fire (no. 3; Fig. 3B) is a high intensity gas seepage area covering
approximately 120 m2. Several vents release gas and flames that
may reach heights of 50 cm or more. An explorative borehole was
drilled nearby in the 1970s. The Lepsa fire (no. 4) is located within
the Stramba Valley and occurs within a small excavation that is
approximately 1 m in diameter. Gas burns occasionally with flames
and does not exceed 10e12 cm in height.
The largest mud volcanoes of Romania are located on the Berca-
Arbanasi hydrocarbon bearing structure. Four distinct mud vol-
canoes are located along the crest of the anticline, from north to
south: Beciu, Paclele Mici, Paclele Mari, and Fierbatori. Beciu (no. 6;
Fig. 3C) is a smaller mud volcano located on the Berca-Arbanasi
structure and covers an area of approximately 4560 m2 (Frunzeti
et al., 2012). The main cone is surrounded by more than 40 vents
that largely appear as circular pools, and release fluid mud or
muddy water with relatively intense gas bubbling.
The Paclele Mari and Paclele Mici are located within the middle
zone of the Berca-Arbanasi structure at the intersection of the
anticline axis with strike-slip faults. Both mud volcanoes display
convex shield morphology and span areas of approximately 1.6 km2
and 0.6 km2, respectively (Etiope et al., 2004). Tens of cones that are
several meters high and circular pools that are filled with muddy
water and with active bubbling characterize the Paclele Mari
plateau (no. 7; Fig. 3D and E). The Paclele Mici mud volcano (no. 8;
Fig. 3F) is dominated by a central cone, and has smaller satellite
cones and gryphons on its flanks. Fierbatori (no. 9; Fig. 3G), located
on the southern portion of the anticline, has a flat surface, small
circular muddy pools, and a wide degassing area of dry soil without
water discharge.
One bubbling spring and one mud volcano were investigated
within the Getic Depression. The spring is located within the
Sacelu-Gorj spa (no. 11). The mud volcano from Alimpesti (no. 10)
releases mud and gas from a main cone and a large portion of its
surface (approximately 176 m2 in total) is covered by vegetation.
 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14 5

Fig. 3. Photographs from selected study sites: A e Andreiasu fire; B e Collecting gas samples from Lopatari fire; C e Beciu mud volcano; D e Paclele Mari mud volcano; E e
Collecting gas samples from a pool at Paclele Mari; F e Paclele Mici mud volcano; G e Fierbatori mud volcano; H e Deleni seep.

2.3.2. Transylvanian Basin
The Sarmasel everlasting fire (Fig. 1 e no. 12) is located within
the Sarmasel village, in central Transylvania. Several vigorous seeps
occur in an area deprived of vegetation that is roughly 1 ha. In this
area, gas may spontaneously ignite and flames can reach heights of
30e50 cm. The occurrence corresponds to the Sarmasel gas deposit,
the first gas field discovered in Transylvania, that has 15 productive
horizons, from very shallow to very deep (Paraschiv, 1979).
At Deleni, two seeping areas (no. 13), that are approximately
1500e2000 m2 wide and 200 m apart, have been identified. These
seeps release gas in 0.2e3 m wide bubbling pools that contain salty
water (Fig. 3H). The seeps are located over the Deleni gas deposit,
one of the most important in Transylvania that has 13 productive
complexes.
At Tauni (no. 14), eight craters with diameters of 2e3 m to more
than 10 m occur over an area of approximately 1 km2. The craters
are filled with water with vigorous gas bubbling. Bubble diameters
may reach 15e20 cm and the bursting frequency is ~ 2e3 bubbles
per second. The seeps are located in correspondence with the Tauni
gas deposit.
At Praid (no. 15), located on the eastern margin of the Transyl-
vanian Basin, gas bubbles in Corund Creek near the entrance of an
old salt mine. The seep is not associated with a commercial hy-
drocarbon reservoir.
The Homorod seepage site (no. 16) is also located on the eastern
margin of the Transylvanian Basin and spans an area of approxi-
mately 2000 m2. Seepage is characterized by several small mud
releasing features with flattened cones that are less than 2 m in
diameter, or bubbling pools with muddy water. A detailed
description is provided in Etiope et al. (2011a).

Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
6 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14
2.3.3. Moldavian platform
The seepage area of Bacau-Gheraiesti (Fig. 1, no. 17) is located
north of Bacau in the Gheraiesti village, and belongs to the Bacau
Petroleum System (Popescu, 1995). The gas seepage area spans
approximately 250,000 m2. Gas release was observed in 2003,
when a residential area was developed; the first gas leak was
accidentally discovered in the cellar of a newly built house (Baciu
et al., 2008). Gas leakage was subsequently observed in additional
cellars and several shallow wells were drilled in order to investigate
the gas source and to try to degas the area. The residential building
project was halted and some residents had to leave houses they had
just purchased. Isotopic analyses indicated that gas was seeping
from a deep thermogenic reservoir (Baciu et al., 2008).
3. Materials and methods
3.1. Dataset
Here, we present the molecular and isotopic composition of 17
seeps, some with analyses repeated during different years (Table 1).
Data from 8 seeps (no. 6, 7, 8, 9, 12, 15, 16, and 17) have previously
been published (Baciu et al., 2007, 2008; Etiope et al., 2009a) while
unpublished data are reported for 9 seeps (no. 1, 2, 3, 4, 5, 10, 11, 13,
and 14).
Methane flux data are from the same 17 seeps (Table 2), plus
another 77 seeps (Supplementary Material, Table S1) that were not
sampled for laboratory analyses. Out of the main 17 seeps, flux data
from five seeps have been published (Etiope et al., 2004; Baciu et al.,
2007; Spulber et al., 2010; Etiope et al., 2011a). In several seeps, flux
measurements were repeated during successive years (see Table 2)
in order to extend the investigated area using upgraded
instrumentation.
3.2. Gas sampling and analysis
Routine sampling and analytical methods were employed for all
gas seeps, as described by Etiope et al. (2011b). To reduce air
contamination, gas samples were typically collected using an
inverted funnel connected via silicone tubes to a T-valve and a
syringe. Gas samples were stored in 100 or 150 mL glass tubes
equipped with two vacuum stop-cocks.
Gas samples from 14 seeps (Table 1, seeps no. 1 to 9, 12, 13, and
15 to 17) were analysed at Isotech Labs Inc. (Illinois, USA). Molecular
composition (C1-C6 alkanes, ethylene, H2, CO2, N2, H2S, He, and Ar)
was determined via gas chromatography (Shimadzu 2010 TCD-FID,
accuracy 2%) and the isotopic composition of C1-C3 alkanes, CO2,
and N2 were determined via Isotope Ratio Mass Spectrometry
(IRMS, Finnigan Delta Plus XL, precision ±0.1‰ for 13C, ±2‰ for 2H
and ±0.3‰ for 15N).
Gas samples from three seeps (Table 1, seeps no. 10, 11, and 14)
were analysed at the Istituto Nazionale di Geofisica e Vulcanologia
in Rome. The molecular analysis was performed using a Fourier
Transform Infrared spectrometer (FTIR DX-4030, Gasmet, Finland)
with a standard spectra library for rapid, semi-quantitative (with
accuracy between 10 and 20%), and simultaneous determination of
14 gases (CH4, CO2, H2O, CO, C2H4, C2H6, C3H8, n-C4H10, i-C4H10, n-
C5H12, i-C5H12, C6H6, SO2, and COS) with typical detection limits of
1e3 ppmv. For the isotopic analysis, approximately 50e100 mL of
gas were collected in a Teflon bag and analysed using a Cavity Ring-
Down Spectroscopy (CRDS) methane carbon analyser (Picarro
G2112-I, Picarro Inc., California; precision <0.7% at 1.8 ppmv CH4
5 min, 1s, and <0.4% at 20 ppmv CH4, 5 min, 1s). A semiconductor
detector (Hydrotech Huberg, Italy; detection limit of 5 ppmv) was
used to detect H2 in the gas samples.
Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
3.3. Gas flux measurements and output estimates
Methane fluxes were measured using the closed chamber
method. The instrumental package consists of a flux-meter equip-
ped with CH4 and CO2 sensors and wireless data communication to
a palm-top computer (West Systems, Italy). Using this system,
fluxes are automatically calculated with a linear regression of gas
concentration build-up in the chamber. The methane sensor in-
cludes semiconductor (range 0e2000 ppmv, a lower detection limit
of 1 ppmv, and a resolution of 1 ppmv), catalytic (range 2000 ppmv,
3% v/v), and thermal conductivity (range 3%, 100% v/v) detectors.
The CO2 sensor is a double beam infrared detector (LI-COR).
Although measured, CO2 flux data were not relevant for this work
and are not discussed. The accumulation chamber was equipped
with a Nafion drying tube (Perma Pure, USA) for humidity removal.
To remove the excess humidity of free gas from bubbling springs, a
Drierite drying unit was connected to the inlet of the flux-meter.
Flux measurements were performed for each seep on visible
vents and in the surrounding ground where diffuse exhalation
(miniseepage) occurs. Fluxes were measured from individual bub-
ble plumes within mud volcano vents. As described in Etiope et al.
(2004), for non-accessible bubble plumes, fluxes were visually
estimated on the basis of bubble size and bursting frequency. For
miniseepage, chamber measurements were performed on points
homogeneously distributed around vents and spanned areas of a
few m2 (in small seeps) to more than 1 km2 (in large mud vol-
canoes) (see Table 2). Total emission from a given seep was esti-
mated by summing the measured fluxes from all vents within
the seepage area and miniseepage surrounding vents. Diffuse
miniseepage was estimated by applying the Natural Neighbour
interpolation method of the Surfer 10 software package (Golden-
Software Ltd). Procedures and examples of vent and miniseepage
emission estimations are described in greater detail in Etiope et al.
(2011b). For many seeps, miniseepage measurements did not cover
the entire area where gas release actually occurred. Measurements
were performed in close proximity to vents. However, diffuse
exhalation may be widespread even at large distances (hundreds
of meters) from vents (e.g., Etiope et al., 2011b). As a result, calcu-
lated total output must be considered to be a very conservative
estimate.
The fluxes of ethane and propane (Fn) were estimated by
knowing the CH4 flux and the C1/C2 and C1/C3 ratios from labora-
tory analyses, based on the following equation:
Cn
Fn ¼  FCH4  mol:corr:n
CCH4
where CCH4 is the methane concentration, Cn is the concentration of
ethane or propane, FCH4 is the flux of methane at the sampling
point, and “mol.corr.” is the molecular correction (1.875 for ethane,
and 2.75 for propane). The equation can only be used when it can
be assumed that the relative composition of gas does not signifi-
cantly change within the measured area (Etiope and Ciccioli, 2009).
4. Results and discussion
4.1. Gas origin
Molecular and isotopic composition data for the investigated
hydrocarbon seeps are provided in Table 1. All of the seeps are CH4-
dominant with the exception of Homorod, a N2-rich seep located on
the eastern margin of the Transylvanian Basin. The geochemistry of
this seep is discussed in detail in Etiope et al. (2011a). The CO2, N2,
and H2 concentrations, as well as the helium isotope ratios, are
within the typical ranges for natural gas in sedimentary basins.
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and

Table 1
Molecular composition of gas (vol. %), and isotopic composition of C1-C3 alkanes (d13C ‰ VPDB; d2H, VSMOW), CO2 (d13C ‰ VPDB), N2 (d15N, ‰ atm) and He (R/Ra) from 17 major seeps in Romania. Data are corrected for air
contamination during sampling. GS: Gas Seep; MV: Mud Volcano; WS: Gas-rich water spring.

Site no. Region Seep Seep Sampling C1 C2 C3 C4 C5 C6þ CO2 N2 Ar H2 He d13CCH4 d13CC2 d13CC3 d13CCO2 d2HCH4 d15NN2 He/4He Reference
3

type year

1 Carpathian Andreiasu EF GS 2004 95.8 2 2.2 0.001 0.03 Baciu et al.,

Flysch and 2007
Foredeep 2007 93.73 1.89 0.54 0.22 0.06 0.02 1.93 1.58 0.03
34.48
147.6 Etiope et al.,
2009a
2012 95.05 1.94 0.57 0.26 0.1 0.07 1.97 0.04 0.001
35.72
25.54
24.5 11.95
151 this work
2 Raiuti GS 2012 95.55 1.23 0.37 0.17 0.08 0.08 1.07 1.41 0.01 0.003 0.01
32.22
27.51
26.08 6.46
148.9 this work
3 Lopatari GS 2012 96.16 1.82 0.75 0.45 0.2 0.11 0.26 0.16 0.001 0.002
32.28
26.78
26.51
144.5 this work
4 Lepsa GS 2012 89.35 0.77 0.08 0.021 0.1 0.003 7.96 1.8 0.01 0.002
48.46
30.43 5.03
227.8 this work
5 Andreiasu MV MV 2012 78.94 0.05 0.04 0.082 0.001 0.002 20.94 0.02 0.003 0.001
41.67 7.58
171.6 this work

C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14

6 Beciu MV 2007 93.15 2.05 0.76 0.25 0.04 0.02 2.88 0.85 0.003 0.001
37.28
170.2 Etiope et al.,
2009a
7 Paclele Mari MV 2004 82.7 2 15.3 0.0025 0.04 Etiope et al.,
2004
2007 94.75 1.32 0.61 0.17 0.055 0.02 2.7 0.3 0.01
37.88
164.9 Etiope et al.,
2009a
8 Paclele Mici MV 2004 94.9 2.3 2.8 0.0024 0.04 Etiope et al.,
2004
2007 94.53 0.05 0.01 0.01 2.2 3.17 0.03 0.002
44.46
184.6 Etiope et al.,
2009a
9 Fierbatori MV 2004 91.3 2.5 6.2 0.0014 0.02 Etiope et al.,
2004
2007 94.95 2.24 0.83 0.44 0.12 0.03 0.93 0.41 0.02 0.002
47.96
172.5 Etiope et al.,
2009a
10 Alimpesti MV 2013 95.93 0.34 0.06 0.07 3.59 0.013
52.1 this work
11 Sacelu-Gorj WS 2013 98.48 0.02 0.03 0.008 0.01 0.03 1.41 0.013
34.3 this work
12 Transylvanian Sarmasel GS 2002 98.7 0.7 0.6 0.0027 0.02 Baciu et al.,
Basin 2007
2007 99.25 0.09 0.01 0.002 0.11 0.53 0.01 0.002
67.42
192.2 Etiope et al.,
2009a
13 Deleni 1 GS 2007 98.65 0.09 0.08 0.003 0.42 0.69 0.01 0.06
66.11
189.6
0.1 this work
Deleni 2 GS 2007 98.02 0.02 0.02 0.03 0.44 1.47 0.01 0.004
66.15
189.5
0.3 this work
14 Tauni Crater 1 GS 2013 99.12 0.01 0.01 0.86 0.007
60.3 this work
15 Praid GS 2007 84.80 0.29 0.07 0.025 4.46 7.3 0.64 0.43
28.98
193.6 0.79 Etiope et al.,
2009a
16 Homorod 2 MV 2002 0.4 5.6 92.6 1.44 0.42 Baciu et al.,
2007
Homorod 3 MV 2007 0.53 0.12 7.51 90.66 0.12 0.95
25.68
33.79 þ123.3 1.1 Etiope et al.,
2011a
Homorod 4a MV 2009 0.54 0.13 6.08 90.88 0.13 0.85
36.49
32.56
19.68 þ31 0.51 Etiope et al.,
2011a
Homorod 4b MV 2009 0.44 0.09 4.6 92.78 0.09 1
37.40
33.10
19.40 þ27 0.8 Etiope et al.,
2011a
Homorod 4c MV 2009 1.04 0.17 8.1 89.82 0.17 0.67
45.96
33.04 þ26.3 0.38 Etiope et al.,
2011a
17 Moldavian Bacau GS 2008 93.81 2.98 1.13 0.57 0.22 0.13 0.08 1.06 0.01 0.003
49.42
173.4 Baciu et al.,
Platform 2008

7
 8 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14

Higher 3He/4He values appeared at the boundary between the

Frunzeti et al., 2012

Spulber et al., 2010

Spulber et al., 2010
Spulber et al., 2010
Spulber et al., 2010
Transylvanian Basin and the Carpathian volcanic arc (the Praid and

Etiope et al., 2011

Baciu et al., 2008
Homorod seeps).
For a first evaluation of different hydrocarbon origins, stable
Reference

this work
this work
this work
this work
this work
this work

this work
this work
this work
this work
this work

this work
this work
carbon (d13C) and hydrogen (d2H) isotopic compositions of
methane were plotted using the classic genetic diagram (Schoell-
plot) (Fig. 4). Two main groups of data can be observed in Fig. 4. The
first group is represented by microbial methane from the central
Transylvanian Basin (no. 12 and 13). The second group is predom-
C3 (kg y
1)

inantly thermogenic methane from the Carpathian Paleogene

Flysch and Foredeep units.
12924
1012

4243

1325
0.03
792

576

111
889

165
The isotopic characteristics of the two Transylvanian gas seeps,
32

33

70

37
4

2
9
3
although approximately 50 km apart, are quite uniform, largely due
to similarities for hydrocarbon deposits within the entire basin.
Thermogenic gases can be divided into four smaller groups. The
C2 (kg y
1)

groups provide indications of rough maturity and similarities for

19069

the geological conditions of reservoirs. Three everlasting fires

1890
2345

7803

1637

1012

2382
0.01
950

380

126

427

246
24

72

29

28

(Andreiasu, Lopatari, and Raiuti) from the Eastern Carpathians form

2

a high maturity group (seeps no. 1 to 3). The Lepsa Fire (no. 4) has a
distinct position in the plot, displays much lower maturity, and may
C1 (t y
1)

include mixing with microbial gas (also see Fig. 8). The fires are
Output

located within the Foredeep (no. 1 to 3) and in the Paleogene Flysch

730a
383a

0.03

11.7
189

251
595
37a
1.5

1.8

2.8
4.4

0.6
50
61

27

19

17

40

(no. 4). The mud volcanoes of the Eastern Carpathian Foredeep (the
3

Andreiasu mud volcano, no. 5, and the mud volcanoes of the Berca-
Arbanasi structure, nos. 6 to 9, form a cluster between “thermo-
Area (m2)

genic with oil” and “thermogenic with condensate” fields within

1620000
630000

250000
25000

25000

the Schoell plot. The Bacau seep (no. 17) falls within the same zone
4560

1500

2200
1645

17.5
42.5
400
300

452

180
0.4
10

12

28

of the plot despite its position near the limit between the Molda-
6

vian Platform and the Carpathian Foredeep.

The isotopic composition of the Praid seep (no. 15) reflects its
2007
2012
2012
2012
2012
2012
2011
2007
2007
2007
2013
2013
2007
2008
2008
2008
2009
2013
2009
2008
Year

peculiar geographic and geological position on the eastern margin

of the Transylvanian Basin, near the volcanic belt of the Eastern
Carpathians. Here, the elevated geothermal flux may have
contributed to an increase in source rock maturity. The gas could
26.83242

26.49173
26.54812
26.83650
26.83669
26.69383
26.70634
26.70884
26.69412
23.79616
23.54002
24.18936

24.33797
24.33786
25.11407
24.08009
25.29255
26.90845

also be partially oxidized and the presence of abiotic methane as a

minor component of geothermal origin cannot be excluded. The
E

Homorod mud volcano (no. 16), also located on the same rim of the
GPS coordinates

Transylvanian Basin, is an interesting seep containing N2-enriched

gas and abiotically oxidized methane (see Etiope et al., 2011a for a
45.74907

46.00221
45.53421
45.82431
45.75191
45.37946
45.33835
45.35696
45.29474
45.12258
45.10474
46.76730

46.27455
46.27469
46.53830
46.14775
46.03078
46.61275

detailed discussion). Both the Homorod and Praid seeps have a high
helium content (1.5 vol% and 0.45 vol%, respectively; Etiope et al.,
N

2011a).
In the C1/(C2þC3) vs. d13C1 diagram (Bernard plot; Fig. 5), sam-
Total calculated output of C1-C3 hydrocarbons for the main investigated sites.

ples are grouped into six clusters. Cluster A contains microbial gas
Bacau Gheraiesti
Tauni Crater 1
Andreiasu MV

from the central Transylvanian Basin (nos. 12 to 14). Samples Deleni

Andreiasu EF

Paclele Mari
Paclele Mici

Sacelu-Gorj

1 and 2 had the same d13C but a different C1/(C2þC3) ratio. Cluster B
Fierbatori
Alimpesti

Homorod
Sarmasel

Deleni 1
Deleni 2
Lopatari

contains the Paclele Mici (no. 8) and the Andreiasu (no. 5) mud
Name

Raiuti
Lepsa

Beciu

Praid

volcanoes. As discussed in Section 4.2, gases within these areas are

likely affected by molecular fractionation and/or secondary meth-
anogenesis. Cluster C includes the Bacau (no. 17), Lepsa (no. 4),
Fierbatori (no. 9), and Alimpesti (no. 10) seeps. Seeps 4 and 10, in
Carpathian Flysch and Foredeep

particular, are located within a mixing zone between microbial and

Methane output from Etiope et al. (2004).

thermogenic gases, while seeps 9 and 17 display clearer thermo-

genic character. Cluster D contains the other everlasting fires from
Transylvanian Basin

Moldavian Platform

the Eastern Carpathian area (Andreiasu, no. 1; Raiuti, no. 2; and

Lopatari, no. 3), and mud volcanoes from Paclele Mari (no. 7) and
Beciu (no. 6) over a very narrow range of C1/(C2þC3) ratios. Cluster E
contains high maturity gases from Praid (no. 15) and Sacelu (no. 11),
Region

the latter is likely affected by molecular fractionation and/or sec-

ondary methanogenesis. Cluster F includes two samples obtained
from the Homorod mud volcano collected during different years
(no. 16, Table 1).
No. site

Overall, the two genetic diagrams, Figs. 4 and 5, indicate that gas
Table 2

seep geochemistry actually reflects the different geological and

10
11
12

13

14
15
16
17
1

2
3
4
5
6
7
8
9

maturity conditions of basins where seeps are located, and

Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14 9

Fig. 4. Isotopic composition of methane from the investigated seeps in Romania (“Schoell plot”; Schoell, 1983). Gas seeps (red squares): 1 e Andreiasu fire; 2 e Raiuti; 3 e Lopatari;
4 e Lepsa; 12 e Sarmasel; 13 e Deleni; 15 e Praid; 17 e Bacau-Gheraiesti; Mud volcanoes (black tringles): 5 e Andreiasu MV; 6 e Beciu; 7 e Paclele Mari; 8 e Paclele Mici; 9 e
Fierbatori; 16 e Homorod (estimated original isotopic composition after Etiope et al., 2011a). The theoretical relationship between d13C and vitrinite reflectance (Ro) is based on
Jenden et al. (1993). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

thermogenic components formed in situ or transferred from the

deeper Mesozoic TPS, as hypothesized in the past (Popescu, 1995).
Oil seeps located within the central portion of the Transylvanian
Basin are indeed indicated in the historical literature (Wanek,
2005).

4.2. Seep vs. reservoir and post-genetic processes

Based on the molecular composition of some reservoirs (for

Fig. 5. d13C1 vs C1/(C2þC3) diagram (“Bernard plot”) for the investigated Romanian
seeps (modified after Bernard et al., 1976; Milkov, 2011). Numbers according to Fig. 4.
Additional seeps not reported in Fig. 3: 14 e Tauni; Mud volcanoes: 10 e Alimpesti;
Bubbling springs (blue diamond): 11 e Sacelu. Letters refer to seep clusters explained
in the text. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)
highlight that within the same basin different gas sources, with
different maturity and secondary alteration processes, may coexist.
Such a result may be due to source rocks or reservoirs at various
depths, suggesting a vertical sequence for petroleum systems. Gas
in the Paleogene Flysch, in particular, may have a wide range of
maturity (seeps 3 and 4). Carpathian seeps are generally related to
more mature source rocks, especially more intensively degassing
seeps (“everlasting fires” 1, 2, and 3). As discussed in Section 4.2,
variations in C1/C2þ ratios within the same basin can then be due to
secondary alteration processes.
Transylvanian gases have measurable amounts of C2-C4 alkanes
(up to 0.1 vol %, Table 1). Their presence suggests that the gas is
not entirely microbial. Ethane and propane can only be produced
by microbes in recent seabed sediments and in very low
amounts (Formolo, 2010). Therefore, our data suggest that Tran-
sylvanian gases (Sarmasel, Deleni, and Tauni) may include minor
which isotopic data are not available) and the molecular and iso-
topic compositions of related seeps, molecular fractionation during
gas migration from the reservoir to the surface, and possible inputs
of secondary methanogenesis, can be assessed using the Bernard
diagram (C1/(C2þC3) vs. d13C1). Molecular fractionation is the sec-
ondary alteration of gas that can lead to misinterpretation of gas
origin when analysed at the surface or in shallower reservoirs. This
fractionation is a sort of segregation process or a “chromatographic
effect” that occurs during gas migration to the surface and is related
to differential solubility and molecular adsorption on solid grains of
mud such that seeping gas (at the surface or trapped in shallower,
near surface pools) is dryer (higher C1/C2þ) than the main, deeper
reservoir gas (Etiope et al., 2009a).
The seep-reservoir association in Romania was assessed at nine
sites (Lopatari, Bacau, Deleni, Praid, Sarmasel and four mud vol-
canoes on the Berca-Arbanasi structure). The results indicate that
for the case of gas seeps (especially everlasting fires) no appreciable
molecular fractionation exists (the difference between the C1/
(C2þC3) ratio between the seep and the reservoir is within ±100)
(Fig. 6). Significant fractionation (D(C1/C2þC3) > 100) occurs in the
case of mud volcanoes. In dry gas seeps, gas ascension primarily
occurs within a single-phase system and is less influenced by water
or mud than it is in mud volcanoes. In general, we observed that
molecular fractionation is inversely proportional to gas flux (e.g.
Etiope et al., 2011b). A relevant example is offered by the two seeps
at Deleni (13a and 13b) that are connected to the same reservoir.
The stronger seep, with a higher gas flux (~17 tons/year; Deleni 1,
13a), exhibits no molecular fractionation as compared to its reser-
voir, while the weaker seep (~3 tons/year Deleni 2, 13b) has a higher
C1/C2þC3 ratio. In the case of mud volcanoes, Paclele Mici (the
smaller and weaker mud volcano) displays a higher molecular
fractionation than Paclele Mari (the larger mud volcano) although

Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
10 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14

Fig. 6. Bernard plot comparing gas seeps and the corresponding reservoirs. Reservoirs are shown as horizontal lines or bars since d13C data are missing in the literature. Only low
flux, weaker seeps and mud volcanoes (5, 8, 13b) show molecular fractionation (higher Bernard ratio). Reservoir C1/(C2þC3) data from Filipescu and Huma (1979). Reservoir depth:
Sarmasel:
200 to
1440 m; Deleni (Saros):
160 to
2125 m; Praid:
2576 m; Roman-Secuieni (Bacau):
310 to
2179 m; Berca-Arbanasi:
2601 to
3331 m; Lopatari
reservoir:
3484 to
3496 m (Filipescu and Huma, 1979). Numbers according to Figs. 4 and 5.

both are located within the same hydrocarbon reservoir. The
Paclele Mici mud volcano has a Bernard ratio resembling microbial
gas ((C1/C2þC3) >1500) but the complete isotopic analysis dem-
onstrates thermogenic origin.
The seep vs. reservoir characterization is also suitable for
identifying the feeding reservoir of the seeps. A good example is the
Bacau seep (no. 17) where the reservoir is composed of a main
deeper reservoir and one or more shallower pools (Filipescu and
Huma, 1979). In this instance, the seep has the same C1/(C2þC3)
ratio of the main reservoir (Fig. 6).
The everlasting fires from Andreiasu, Raiuti, and Lopatari display
very similar isotopic ratios, suggesting that the three gases develop
under similar source rock conditions. In the case of the Andreiasu
fire, the Bernard ratio remained constant over time, between 2007
and 2013 (Table 1). Andreiasu MV (no. 5) presumably has the same
source as the Andreiasu fire (no. 1) but because migration includes
an interaction with water and sediments (mud volcanism) higher
fractionation occurs. As indicated by Etiope et al. (2009a), such a
finding is normal for mud volcanoes. If not evaluated correctly,
seeps with high C1/(C2þC3) ratios may be misinterpreted as mi-
crobial. Similar “false” microbial identifications may actually occur
in all weak flux seeps, as well as within shallow gas reservoirs that
are produced by slow accumulation of gas migrating from a main
deeper reservoir. High flux gas seeps are, instead, the best in-
dicators of reservoir molecular composition.
The relationship between the d13C-CO2 and CO2 concentration
(Fig. 7) is useful for assessing CO2 origin and biodegradation.
Samples from Andreiasu, Raiuti, and Lepsa display positive d13C-

Fig. 7. Plot for the relationship between d13C of CO2 and CO2 concentration. Curves are mixing lines of possible end-members after Jeffrey et al., 1991.

Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14
CO2 values, suggesting the biodegradation of hydrocarbons.
Biodegradation is generally followed by the generation of second-
ary microbial methane (the reason for 13C enriched CO2) that does
not have the typical d13C-CH4 values of microbial gas. As discussed
by Brown (2011), microbial methane has an isotopic character
similar to thermogenic methane, so it cannot be distinguished from
thermogenic methane.
4.3. Maturity and possible kerogen type
Published information on the source rocks of petroleum in
Romania, including vitrinite reflectance, is scarce. To approximately
evaluate kerogen type and the formation temperatures of gas from
several seeps, we combined thermogenic gas formation modelling
(Tang et al., 2000) and maturity plots (Berner and Faber, 1996).
However, since they are missing within the scientific literature, a
primary problem was the choice of kerogen d13C values. Oligo-
Miocene dysodiles and menilites are important source rocks for
petroleum accumulations along the external side of the Eastern
Carpathians, distributed in Romania, Ukraine, and Poland (Popescu,
1995). Kerogens in this area are type II and III, with a vitrinite
reflectance between 1.10 and 1.15% Ro. The stable carbon isotopic
composition of kerogen is known in Polish menilites (d13Cker
of
26.6‰ for type II kerogen;
25.2‰ for type III; Kotarba et al.,
2009) but not in Romania. By assuming that dysodiles and meni-
lites have similar d13Cker values along the Carpathians and by
combining the maturity model of Berner and Faber (1996) and the
thermogenic gas generation modelling of Tang et al. (2000), it is
possible to verify whether seep gas derives from this category of
source rocks.
Two plots (d13C1 vs. d13C2 and d13C2 vs. d13C3; Figs. 8 and 9) were
drawn based on the isotope modelling of Tang et al., (2000) (and
conducted using GeoIsochem Corp. GOR Isotope software 1.94 for
the instantaneous generation of methane, ethane, and propane
from default Type I, II, and III kerogen, with a heating rate of 5  C
per 1 million years).
The d13C1 vs. d13C2 plot suggests that Andreiasu gas was formed
from type III or II kerogen, with a relatively high maturity (Ro of
approximately 3) at roughly 200  C. Raiuti and Lopatari appear to
have a similar maturity but did not exactly fit with GOR modelling,
likely because they are a mixture of different sources, or because
the input parameters for GOR modelling (for example the heating
Fig. 8. Thermogenic gas formation modelling from marine (Type I and II) and terrestrial (Type III) kerogen, applied to seeps where both d13C1 and d 13C2 are known. The modelling is
based on GeoIsochem Corp. GOR software 1.94; heating rate of 5  C per million years (Tang et al., 2000). Horizontal arrow for the Lepsa gas suggests mixing with microbial gas (see
also Figs. 4 and 5).
Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
11
rate) are not optimal for the Carpathian Basin. As also suggested by
Figs. 4 and 5, Lepsa (no. 4) is likely mixed with microbial gas.
The d13C2 vs. d13C3 plot (Fig. 9) suggests that the gases within
Andreiasu, Raiuti, and Lopatari are all formed from type III kerogen
with a Ro between 2.5 and 3. Based on the two GOR plots, the
formation temperature of the three gases (Andreiasu, Raiuti, and
Lopatari) is inferred to be between 180 and 220  C.
The d13C1 vs. d13C2 maturity plot (Fig. 10) generated for Type II
and III kerogen with a range of d13Cker from
23 to
29‰ also
suggests that Lopatari and Raiuti gases are derived from Type III
kerogen. To fit the vitrinite values indicated by Popescu (1995;
1.1e1.2% Ro), the d13Cker should be approximately
26‰. The
interpretation for Andreiasu gas is more ambiguous. Different from
the GOR model that suggests a relative high maturity, the
Andreiasu appears closer to Type II kerogen and results in high Ro
values (Ro > 2) only for low d13Cker values, less than
28‰. Lepsa is
clearly derived from lower maturity kerogen and likely includes a
microbial component.
The Berner and Faber model (using Kotarba et al. (2009) and
Popescu (1995) data) and the GOR model are consistent in terms of
orders of magnitude of maturity. Both indicate that Lopatari and
Raiuti gases are of type III kerogen, while Andreiasu is of type II or
III, and Lepsa is a mixed gas with a low maturity thermogenic
component from type II kerogen. However, this interpretation of
source rock type and maturity should be considered to be a “best
guess” based on the poor data available in the scientific literature.
To better understand the source rocks of seeping gases, direct
measurements of d13Cker for the Romanian dysodile and menilite
formations are required.
4.4. Total hydrocarbon emissions
Measured methane output from the seeps (including fluxes
from visible vents and the invisible mini-seepage surrounding
vents, as explained in Section 3.3) has wide ranges of less than 1 kg
to hundreds of tons per year (Fig. 11; Table 2 and Table S1), obvi-
ously depending on the size and activity of the seep. Seeps are
generally small within the Transylvanian Basin, and annual output
rarely exceeds 1 ton of methane, although notable exceptions occur.
The greatest contribution is provided by the Sarmasel seep that has
a CH4 output of 595 t y
1. The eight craters from Tauni yield a total
output of approximately 440 t y
1. All of the other investigated sites
12 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14
Fig. 9. Thermogenic gas formation modelling from marine (Type I and II) and terrestrial (Type III) kerogen for ethane and propane, and calibrated for vitrinite reflectance.
Fig. 10. Maturity plot of ethane vs. methane based on Berner and Faber (1996). d13C
values used for kerogen are between
23 and
29‰, for type II and III kerogen.
from the Transylvanian Basin total approximately 60 t CH4 y
1.
In the Carpathian Foredeep area, mud volcanoes on the Berca-
Arbanasi hydrocarbon-bearing structure (Paclele Mari, Paclele
Mici, Fierbatori, and Beciu) release a total of ~1340 t CH4 y
1. Other
areas within the Carpathian Foredeep exhibit a total methane
emission of approximately 110 t y
1, with the most important seeps
occurring at Andreiasu, Lopatari, and Alimpesti. On the Moldavian
Platform, the most important seep is Bacau that has an annual CH4
output of 40 t y
1. Total emission in other sites within the Molda-
vian Platform is approximately 5 t y
1 (Pop et al., 2015; Supple-
mentary Material, Table S1).
The total CH4 emission at all of the investigated sites is
approximately 2600 t y
1. The result includes the total amount of
methane released by the 17 representative seeps plus the addi-
tional 77 seeps inventoried in the current article. The total amount
of photochemical pollutants, ethane and propane, is approximately
35.3 t y
1 and 21 t y
1, respectively (Table 2). The geo-ethane
emission is equivalent to the ethane biologically produced and
emitted in roughly 35 km2 of agricultural soil (Das et al., 2003).
Although their concentrations in the atmosphere are very low, the
two gases are important due to the ozone and photochemical
Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
pollution they produce. On a global scale, hydrocarbon seeps
represent a major natural atmospheric source of ethane and pro-
pane (Etiope and Ciccioli, 2009).
Here, the following must be stated: (a) the emission estimates
for each seep (and thus the total quantity) are quite conservative
because, as explained in Section 3.3., the measurements did not
span the entire miniseepage area; (b) the geological emissions we
document refer to macro-seepage fluxes from 94 seeps, while the
total number of inventoried seeps in Romania is much higher (470;
Ionescu et al., 2017); and (c) the total emission does not include
diffuse microseepage from soil which is generally widespread in
petroleum systems (Etiope and Klusman, 2010). The investigated
seeps are, however, the largest and most active, and, thus, represent
the main gas emitters.
By conservatively assuming that remaining seeps in Romania
individually release gas quantities on the order of 0.1e1 t CH4 y
1,
the total macro-seepage CH4 emission in Romania could reach
3000 t y
1. In the past, it has been estimated that CH4 microseepage
emissions (10e25 Mt y
1) on a global scale, are equivalent to that
from macro-seeps (13e24 Mt/y) (Etiope, 2015). By accepting that
the same macro-seepage/microseepage ratio is valid for the
Romanian territory, total seepage emissions in Romania would be
on the order of 6000 t y
1. The National Inventory Report for 2016,
issued by the Romanian Ministry of Environment and Climate
Change (UNFCCC, 2016), indicates that methane emissions from
fossil fuel combustion in Romania are 47.9 Gg y
1 (1 Gg ¼ 1000
tons) and 17 Gg y
1 from agricultural field burning. Therefore, the
total release from seepage (~6 Gg y
1 CH4) is a substantial
component of fossil fuel emission and very likely the main natural
source of methane to the atmosphere (wetlands, important on a
global scale, are relatively limited in Romania).
5. Conclusion
Seventeen hydrocarbon gas seeps from the main hydrocarbon-
prone areas of Romania were studied using molecular and iso-
topic analyses and methane flux measurements. The methane flux
was also measured in 77 minor seeps. In the external area of the
Carpathians, corresponding to the Flysch and Foredeep zones, and
on the Moldavian Platform, emitted gases have mainly a thermo-
genic origin with different degrees of maturity. Some display
 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14
Fig. 11. Dot-plot for the emission of methane from all the investigated Romanian seeps (data in Table 2 and Table S1). Values in brackets represent the number of seeps in each
category.
features of hydrocarbon biodegradation. Transylvanian gases are
largely microbial. However, thermogenic methane may be found at
the eastern edge of the basin, near the Neogene volcanic belt. In the
central portion of the Transylvanian Basin (Sarmasel, Deleni and
Tauni), gases contain measurable amounts of C2þ alkanes, which
may imply that the gas is not totally microbial. As suggested by
previous authors (e.g. Popescu, 1995), the presence of a thermo-
genic component may confirm the existence of a deeper petroleum
system in Transylvania, that may deserve future exploration.
A comparison of reservoir vs. seep geochemistry confirmed that
most seeps are actually linked to reservoirs. Therefore, the seeps
may serve as indicators of subsurface hydrocarbon origin and
quality. For high flux seeps, the gas has the same hydrocarbon
molecular composition as the reservoir (the same C1/C2þ ratio). In
weaker seeps and some mud volcanoes, gas was altered by mo-
lecular fractionation (a loss of C2 and C3 gases during gas migra-
tion). In general, seep geochemistry reflects the different geological
and maturity conditions of basins where seeps are located and
highlights the fact that different gas sources with different matu-
rities and secondary alterations exist within the same basin. This
result may be due to source rocks or reservoirs at different depths,
suggesting a vertical sequence of petroleum systems. Variations in
C1/C2þ ratios within the same basin can be due to secondary
alteration processes (e.g., molecular fractionation and secondary
methanogenesis).
Methane output to the atmosphere was estimated based on flux
measurements performed on the 94 investigated locations. The
largest gas emitters are mud volcanoes located on the Berca-
Arbanasi hydrocarbon-bearing structure and the Sarmasel seep in
Transylvania, which release hundreds of tons of methane per year.
Total estimated methane emissions from seeps in Romania are
Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
13
approximately 3000 t y
1. The value is a conservative estimate and
does not include diffuse microseepage, typically widespread over
petroleum system areas. Diffuse exhalation may double the
seepage output.
Geological seepage may represent the first natural source of
atmospheric methane in Romania. Ethane and propane emissions,
mainly related to thermogenic seeps, represent a non-negligible
source of photochemical pollutants at the country level, in addi-
tion to agricultural and oil industry sources. For the future,
methane, ethane, and propane seepage emission estimates will be
refined by extending field measurements, and should be consid-
ered amongst natural greenhouse gas sources in national inventory
reports.
Acknowledgments
This work was financially supported by the Romanian National
Research Council, Project PN-II-ID-PCE-2011-3-0537. The authors
acknowledge support from the Deep Energy Community of the
Deep Carbon Observatory. Unpublished data were acquired by C.
Pop and A. Ionescu during Ph.D. research activities at the Babes
Bolyai University.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.marpetgeo.2017.06.015.
References
Abrams, M.A., 2005. Significance of hydrocarbon seepage relative to petroleum
14 C. Baciu et al. / Marine and Petroleum Geology xxx (2017) 1e14
generation and entrapment. Mar. Pet. Geol. 22, 457e477.
Baciu, C., Caracausi, C., Etiope, G., Italiano, F., 2007. Mud volcanoes and methane
seeps in Romania: main features and gas flux. Ann. Geophys. 50, 501e512.
Baciu, C., Etiope, G., Cuna, S., Spulber, L., 2008. Methane seepage in an urban
development area (Bacau, Romania): origin, extent and hazard. Geofluids 8,
311e320.
Bernard, B.B., Brooks, J.M., Sackett, W.M., 1976. Natural gas seepage in the Gulf of
Mexico. Earth & Planet. Sci. Lett. 31, 48e54.
Berner, U., Faber, E., 1996. Empirical carbon isotope/maturity relationship for gases
from algal kerogens and terrigenous organic matter, based on dry, open-system
pyrolysis. Org. Geochem 24 (10/11), 947e955.
BP, 2016. BP Statistical Review of World Energy, 65th edition bp.com/
statisticalreview#BPstats.
Brown, A., 2011. Identification of source carbon for microbial methane in uncon-
ventional gas reservoirs. Am. Assoc. Petroleum Geol. Bull. 95, 1321e1338.
Cranganu, C., Deming, D., 1996. Heat flow and hydrocarbon generation in the
Transylvanian Basin, Romania. AAPG Bull. 80/10, 1641e1653.
Das, M., Kang, D., Aneja, V.P., Lonneman, W., Cook, D.R., Wesely, M.L., 2003. Mea-
surements of hydrocarbon air-surface exchange rates over maize. Atmos. En-
viron. 37, 2269e2277.
Dolton, G.L., 2006. Pannonian Basin Province, Central Europe (Province 4808)d
petroleum geology, total petroleum systems, and petroleum resource assess-
ment: U.S. Geol. Surv. Bull. 2204eB, 47.
Etiope, G., 2012. Methane uncovered. Nat. Geosci. 5, 373e374.
Etiope, G., 2015. Natural Gas Seepage. The Earth's Hydrocarbon Degassing. Springer,
p. 199.
Etiope, G., 2009. Natural emissions of methane from geological seepage in Europe.
Atmos. Environ. 43, 1430e1443.
Etiope, G., Ciccioli, P., 2009. Earth's Degassing: a missing ethane and propane
source. Science 323, 478.
Etiope, G., Milkov, A.V., 2004. A new estimate of global methane flux from onshore
and shallow submarine mud volcanoes to the atmosphere. Environ. Geol. 46,
997e1002.
Etiope, G., Baciu, C., Caracausi, A., Italiano, F., Cosma, C., 2004. Gas flux to the at-
mosphere from mud volcanoes in Eastern Romania. Terra Nova 16, 179e184.
Etiope, G., Baciu, C.L., Schoell, M., 2011a. Extreme methane deuterium, nitrogen and
helium enrichment in natural gas from the Homorod seep (Romania). Chem.
Geo 280, 89e96.
Etiope, G., Feyzullayev, A., Baciu, C.L., 2009a. Terrestrial methane seeps and mud
volcanoes: a global perspective of gas origin. Mar. Pet. Geol. 26, 333e344.
Etiope, G., Feyzullayev, A., Milkov, A.V., Waseda, A., Mizobe, K., Sun, C.H., 2009b.
Evidence of subsurface anaerobic biodegradation of hydrocarbons and potential
secondary methanogenesis in terrestrial mud volcanoes. Mar. Pet. Geol. 26,
1692e1703.
Etiope, G., Klusman, R.W., 2010. Microseepage in drylands: flux and implications in
the global atmospheric source/sink budget of methane. Glob. Planet. Change 72,
265e274.
Etiope, G., Nakada, R., Tanaka, K., Yoshida, N., 2011b. Gas seepage from Tokamachi
mud volcanoes, onshore Niigata Basin (Japan): origin, post-genetic alteration
and CH4-CO2 fluxes. Appl. Geochem 26, 348e359.
Filipescu, M.N., Huma, I., 1979. Geochemistry of Natural Gases. Academia Publishing
House (in Romanian).
Formolo, M., 2010. The microbial production of methane and other volatile hy-
drocarbons. In: Kenneth, N. (Ed.), Handbook of Hydrocarbon and Lipid Micro-
biology. Springer, New York, pp. 113e126.
Frunzeti, N., Baciu, C., Etiope, G., Pfanz, H., 2012. Geogenic emission of methane and
carbon dioxide at Beciu mud volcano, (Berca-Arbanasi hydrocarbon-bearing
structure, eastern Carpathians, Romania). Carpath. J. Earth. Env. 7 (3), 159e166.
Geological Institute, 1967. Geological Map of Romania 1:1,000,000 (Bucharest).
Ionescu, A., Burrato, P., Baciu, C., Etiope, G., Kis, B.-M., 2017. Inventory of onshore
hydrocarbon seeps in Romania (HYSED-RO Database). Geosciences 7 (2), 39.
Please cite this article in press as: Baciu, C., et al., Hydrocarbon seeps in Romania: Gas origin and release to the atmosphere, Marine and
Petroleum Geology (2017), http://dx.doi.org/10.1016/j.marpetgeo.2017.06.015
Jeffrey, A.W.A., Alimi, H.M., Jenden, P.D., 1991. Geochemistry of the Los Angeles basin
oil and gas systems. In: Biddle, K.T. (Ed.), Active Margin Basins, AAPG Mem, vol.
52, pp. 197e219.
Jenden, P.D., Hilton, D.R., Kaplan, I.R., Craig, H., 1993. Abiogenic hydrocarbons and
mantle helium in oil and gas fields. In: Howell, D. (Ed.), Future of Energy Gases,
vol. 1570. USGS, pp. 31e56.
Kotarba, M.J., Curtis, J.B., Lewan, M.D., 2009. Comparision of natural gases accu-
mulated in Oligocene strata with hydrous pyrolysis gases from Menilite Shales
of the Polish Outer Carpathians. Org. Geochem 40, 769e783.
Krezsek, C., 2011. Petroleum systems of Romania. AAPG-ER Newsl. 6, 5e7.
Krezsek, Cs., Filipescu, S., Silye, L., Matenco, L., Doust, H., 2010. Miocene facies as-
sociations and sedimentary evolution of the Southern Transylvanian Basin
(Romania): implication for hydrocarbon exploration. Mar. Pet. Geol. 27,
191e214.
Link, W.K., 1952. Significance of oil and gas seeps in world oil exploration. Am.
Assoc. Pet. Geol. Bull. 36, 1505e1540.
Milkov, A.V., 2011. Worldwide distribution and significance of secondary microbial
methane formed during petroleum biodegradation in conventional reservoirs.
Org. Geochem 42, 184e207.
Mutihac, V., 1990. In: Tehn (Ed.), Geological Structure of Romanian Territory, p. 419
(Bucharest).
Paraschiv, D., 1979. Romanian oil and gas fields. Tech. Ec. Stud. Inst. Geol. Geoph.
A13, 5e382.
Pawlewicz, M., 2005. Transylvanian composite total petroleum system of the
Transylvanian Basin province, Romania, Eastern Europe: U.S. Geol. Surv. Bull.
2204eE, 10.
Pawlewicz, M., 2007. Total petroleum systems of the CarpathianeBalkanian Basin
province of Romania and Bulgaria: U.S. Geol. Surv. Bull. 2204eF, 17.
Pawlewicz, M.J., Steinshouer, D.W., Gautier, D.L., 2002. Map showing geology, oil
and gas fields, and geologic provinces of Europe including Turkey. Open file
report 97e470I U.S. Geol. Surv. 14 pp.
Pop, C.I., Ionescu, A., Baciu, C., 2015. Methane seepage from geologic sources on the
moldavian platform (Eastern Romania). Carp. J. Earth Env. Sci. 10 (1), 203e210.
Popescu, B.M., 1995. Romania's petroleum systems and their remaining potential.
Petr. Geosci. 1, 337e350.
Sandulescu, M., 1984. In: Tehn (Ed.), Geotectonics of Romania (in Romanian), p. 336
(Bucharest).
Sandulescu, M., Visarion, M., 1988. La structure des plate-formes situe es dans
l’avant-pays et au-dessous des nappes du flysch des Carpathes orientales. Stud.
Teh. Econ. e Geofiz. 15, 62e67 (in French).
Schoell, M., 1983. Genetic characterization of natural gases. Am. Assoc. Petrol. Geol.
Bull. 67, 2225e2238.
Spulber, L., Etiope, G., Baciu, C., Malos, C., Vlad, S.N., 2010. Methane emissions from
natural gas seeps and mud volcanoes in Transylvania (Romania). Geofluids 10,
463e475.
Tang, Y., Perry, J.K., Jenden, P.D., Schoell, M., 2000. Mathematical modelling of stable
carbon isotope ratios in natural gases. Geochim. Cosmochim. Acta 64,
2673e2687.
Tari, G., Ciudin, D., Kostner, A., Raileanu, A., Tulucan, A., Vacarescu, G., Vangelov, D.,
2011. Play Types of the Moesian Platform of Romania and Bulgaria. Search &
Discovery Article#10311. http://www.searchanddiscovery.com/pdfz/documents/
2011/10311tari/ndx_tari.pdf.html.
UNFCCC, 2016. Romania's greenhouse gas inventory 1989-2014. Nat. Invent. Rep.
http://unfccc.int/national_reports/annex_i_ghg_inventories/national_
inventories_submissions/items/9492.php (Accessed 27 August 2016).
US Energy Information Administration (US-EIA), 2015. Technically Recoverable
Shale Oil and Shale Gas Resources: an Assessment of 137 Shale Formations in 41
Countries outside the United States. US Dept. of Energy, Washington.
Wanek, F., 2005. The true story of the discovery of natural gas in Transylvania (in
Hungarian), Muszaki Szemle 30, 27e40.