Oil is produced in the Austrian sector of the Alpine Foreland Basin from Eocene and Cenomanian reservoirs. Apart from petroleum, the basin hosts a significant geothermal potential, which is based on the regional flow of meteoric water through Malmian (Upper Jurassic) carbonate rocks. Oils are predominantly composed of n-alkanes, while some samples are progressively depleted in light aromatic components. The depletion in aromatic components relative to abundant n-alkanes is an effect of water washing. Waters coproduced with oils that are affected by water washing show a progressive reduction in salinity and depletion in 2H and 18O isotopes, indicating that the degree of water washing is mainly controlled by the inflow of meteoric water from the Malmian aquifer. In some fields with Cenomanian reservoir rocks, a hydraulic connectivity with the Malmian aquifer is evident. However, water washing is also recognized in Eocene reservoirs and in areas where the Malmian aquifer is missing. This shows that existing flow models for the regional Malmian aquifer have to be modified. Therefore, the results emphasize the importance of combining data from the petroleum and geothermal industry, which are often handled separately.

1. Introduction

The Alpine Foreland Basin (AFB) is a minor oil and moderate gas province in central Europe (Figure 1). The cumulative production of oil + condensate was about 9 mio. tons and 1.647 mio. Nm3 of associated gas (RAG production until 2015, industrial data). Main oil reservoirs are found in Cenomanian and Eocene horizons, whereas gas is mainly trapped in Oligomiocene rocks (Figure 2). Many studies have been performed to understand the petroleum systems in the basin. Within this context, organic geochemical, biomarker, and stable isotope data have been used to characterize organic matter type and source rock maturity, as well as oil migration and alteration processes (e.g., Schulz et al. (2002) [13]). However, all previous geochemical studies were based on the C15+ hydrocarbon fraction. In contrast, light hydrocarbons (C15− fraction) remained uninvestigated, although they are important proxies for facies and maturity of source rocks and migration and alteration processes (e.g., [4]).

The AFB not only is an important hydrocarbon province but also hosts a major geothermal potential (e.g., [5]), which is related to an active aquifer in Upper Jurassic carbonates (“Malmian aquifer” sensu [6]). The general characteristics of the Malmian aquifer such as charge and discharge area and residence time are reasonably well understood (e.g., [7]; Figure 3). In hydrogeological models, the Malmian aquifer is typically considered as separated from aquifers in overlying stratigraphic units [8]. However, Andrews et al. [7], Goldbrunner [9], and Gross et al. [10] suggested hydraulic connections between the Malmian aquifer and oil-bearing rocks. The interaction of water with hydrocarbons may result in the removal of relative water-soluble compounds (e.g., light aromatics: benzene, toluene, ethylbenzenes, and xylenes (BTEX)) from oil. In the petroleum industry, this process is generally called “water washing” (e.g., [11]). Simple numerical models of Lafargue and Thiez [12] showed that the removal of BTEX is limited by water velocity, if the water flow beneath the oil-water-contact is below 10 cm/year, and by rate of diffusion, if the water flow is higher. Meteoric water can also introduce microbial communities into an oil reservoir and provide oxygen or electron acceptors and inorganic nutrients necessary for microbial activity (e.g., [13, 14]). Therefore, under suitable geological conditions, both processes are concomitant.

Hence, the main aim of this study is to reveal any effect of waters from the Malmian aquifer on the composition of oil in Cretaceous and Eocene reservoirs. In addition, information on the regional and stratigraphic distribution of water washing can help to refine and constrain the hydrogeological model.

To increase the regional coverage of the sample set, selected chemistry/isotope data of water from industry and Andrews et al. [7] are included in the present study. The comparison of these old data (often determined at the beginning of oil production) and data obtained after years of oil production may also reveal any influence of hydrocarbon production on water composition.

2. Geological Background

The asymmetric Alpine Foreland Basin (AFB) stretches along the northern margin of the Alps and dips below the Alpine nappes (Figure 1(b)). In the Upper Austrian sector of the Alpine Foreland Basin, the sedimentary succession overlies crystalline basement of the Bohemian Massif and comprises the following from bottom to top: Permo-Carboniferous graben sediments, Jurassic and Upper Cretaceous mixed carbonate-siliciclastic shelf sediments, and Eocene to Upper Miocene Molasse sediments (Figure 2).

Autochthonous Jurassic and Cretaceous sediments form the basin floor. The Upper Jurassic carbonate group (Figure 2), comprising limestones and dolostones, is up to 500 m thick (Figure 3). These fractured and karstified carbonates are the most important deep aquifer for thermal water (Malmian aquifer) but are absent in the northern and eastern part of the study area due to erosion [5]. Because Malmian connate brines have been replaced by meteoric water (average total mineralization is 2.2 g/l [9]), the Malmian water differs hydrochemically and isotopically from waters in overlying horizons. High salinity water in Malmian rocks is found only in the southern part of the basin, indicating stagnant conditions in this area ([7], Figure 3). Recharge and discharge of the Malmian aquifer take place mainly at the basin edges, often through permeable Cenozoic sediments or fractured basement rocks [8, 27].

Main oil and associated gas reservoir rocks are Cenomanian and Eocene shallow marine sandstones (Figure 2). Minor hydrocarbon deposits are also found in Jurassic clastic rocks and Upper Eocene algal limestones (Lithothamnium limestone). The main source rocks for thermogenic hydrocarbons are deep marine Lower Oligocene pelitic rocks (Schöneck Fm. and Eggerding Fm. [1, 28]), which became mature beneath the Alpine nappes in Miocene time [29]. The lateral migration distance of oil in the Austrian part of the Alpine Foreland Basin varies from less than 20 to more than 50 km [2, 3]. Hydrocarbon migration commenced simultaneously with hydrocarbon generation and continued until the present day [29]. The hydrocarbon habitat is strongly influenced by Neogene uplift and erosion [30]. Neogene tilting of the basin changed migration pathways and oil-water-contacts ([30, 31]; Linzer, pers. comm.). Heavily biodegraded oils occur along the northern margin of the basin in shallow marine Oligocene sands (Figure 1).

Gratzer et al. [2] recognized two oil families. The western oil family (west of S field) contains more sulfur than the eastern family (east of S field). Dibenzothiophene/phenanthrene (DBT/Ph) ratios are higher in the western oil family, indicating enhanced availability of reduced sulfur for incorporation into organic matter [32]. These variations reflect differences in the source rock facies beneath the Alpine nappes (see [28]).

Dry gas, traditionally interpreted as microbial in origin (e.g., [33, 34]), prevails in clastic deep water sediments with an Oligocene (Lower Puchkirchen Fm.) or Miocene age (Upper Puchkirchen Fm., Hall Fm. [3537]).

3. Samples and Methods

38 oil and 15 water samples were collected from producing wells operated by Rohöl-Aufsuchungs AG (RAG, Vienna, Austria) in 2013–2015. The sample code represents the field name by a capital letter. If several wells from a single field have been sampled, these are labelled by numbers; for example, samples D1 to D4 are taken from different wells in the D field (see Figure 1(c)). Special precautions were taken during sampling and laboratory handling to avoid any possible losses of volatile hydrocarbons. Glass bottles were filled with reservoir fluids (oil and water), immediately crimped, and stored at 4°C. In the lab, water and oil were separated and stored in crimped bottles at 4°C for further investigations. In addition, oil and water samples from archives were also investigated.

Fresh and archival water samples (19 in total) were measured by ion chromatography. The samples were filtered through a 0.2 μm nylon filter prior to analysis. The filtrate was diluted and analyzed for cations and anions using two different sets of ion chromatography equipment. The anions were determined on a Dionex DX-3000 system with external suppression. For standard runs, a 25 μl sample loop was used. The cations were analyzed using a Dionex DX-120 system with electrochemical micromembrane suppression and a 25 μl sample loop. Standards were made from commercially available reference material (Merck, Certipur®). The samples were diluted with Milli-Q® 1 : 100 before analysis. For all steps in the analytical procedure, Milli-Q water was used, the analyzed components in the blank were always below detection limit. The detection limits [ppb] were established as follows: Li, 0.1; Na, 5; K, 5; Ca, 10; Mg, 10; Cl, 100; Br, 5; F, 5; J, 0.1; and SO4, 10.

The oxygen isotopic composition (δ18O) of the water samples was measured by the CO2–H2O equilibrium technique [38] with a fully automated device adapted from Horita et al. [39] coupled to a Finnigan DELTAplus mass spectrometer. Horita et al. [39] designed an automated operating procedure to analyze hydrogen and oxygen isotope ratios of the same water samples without sample change. However, in this operating procedure, the equilibrium unit with the sample vials was shaken laterally at a few Hz for 4 h to enhance the isotopic exchange reaction. In the EQ-device used in this study, the water sample is stirred individually in each vial. This leads to a higher precision of the isotope measurements. The temperature of the water bath was 24°C ± 0.1°C during water–CO2 equilibration. Measurement reproducibility of duplicates was better than ±0.05 for δ18O. Deuterium (δ2H) was measured with a Finnigan DELTAplus XP continuous flow stable isotope ratio mass spectrometer by chromium reduction using a ceramic reactor slightly modified from Morrison et al. [40]. The analytical precision of the δ2H measurements was better than 1.5. Normalization of the raw results versus the V-SMOW-SLAP scale was achieved by using a four-point calibration of in-house water standards that have been calibrated against the international reference materials V-SMOW, GISP, and VSLAP. No further corrections were applied. Stable hydrogen and oxygen isotopes of water are expressed against V-SMOW.

Oils were separated from water and treated to remove asphaltenes: 50 mg of oil was diluted in n-pentane and the insoluble fraction was separated by centrifuging. The pentane-soluble fractions were analyzed using a gas chromatograph (GC) equipped with a J&W DB-1 PONA (50 m length, ID 0.2 mm, 0.5 μm film thickness) fused silica capillary column. The sample was injected in split mode at 270°C. The GC oven temperature was programmed as follows: 32°C hold for 5 min followed by heating 2.5°C/min to 310°C and hold for 30 min. Helium was used as carrier gas with a constant flow of 1.3 ml/min. A flame ionization detector was operated at 320°C with gas flows of 350 ml/min and 35 ml/min for air and hydrogen, respectively.

4. Results

4.1. Water Samples

Total mineralization (total dissolved solids) of samples measured in the frame of this study varies between 1893.1 mg/l (sample U2; Cenomanian reservoir) and 18103.3 mg/l (sample N6; Eocene reservoir; Table 1). Notably, the average salinity of waters from Eocene reservoir (10559.7 mg/l) is higher than that of Cenomanian waters (2958.5 mg/l). This observation is also supported by data from industry (Eocene: 14164.5 mg/l; Cenomanian: 6222 mg/l) and Andrews et al. ([7]; Eocene: 11738.1 mg/l; Cenomanian: 2333.5 mg/l; Tables 1 and 2). However, unusually low salinity water from Eocene reservoir is found in oil field D (samples D1-2; Tables 1 and 2) in the northeastern part of the study area. The most important dissolved ions are Na+ and Cl. In connate brines, these constituents are mainly derived from dissolved halite.

δ2H [V-SMOW] and δ18O [V-SMOW] values of water (measured in the frame of this study) from Eocene reservoir range from −48.3 to −12.9 and from −7.1 to 0.6 for hydrogen and oxygen, respectively (Table 1). The isotopic composition of water from Cenomanian reservoir ranges from −51.4 to −41.4 and from −5.8 to −4.8 for hydrogen and oxygen, respectively.

4.2. Oil Samples

Oil samples are characterized by abundant n-alkanes up to C36. Because detailed information on biomarkers and stable isotopes based on the C15+ fractions has already been presented by Gratzer et al. [2] and Bechtel et al. [3], the present paper focuses on light hydrocarbons. The dominant light hydrocarbons are n-alkanes, although cycloalkanes and aromatics are also abundant. However, some oils are characterized by progressive depletion or almost entire removal of benzene, toluene, ethylbenzene, and xylenes (BTEX). To illustrate this phenomenon, the methylcyclohexane/toluene (Mch/Tol) and cyclohexane/benzene (Ch/B) ratios have been calculated (Table 1), which vary widely from 1.7 to 98.2 and from 2.3 to 904, respectively.

5. Discussion

5.1. Quality Control of Water Samples

To control the quality of samples with respect to possible influence of any reservoir additives used during oil production, the Na+ and Cl concentrations are cross-plotted in Figure 5. No significant deviation from the sea water dilution line is observed, suggesting no influence from reservoir additives. Nevertheless, chlorine, which is considered as a conservative constituent, is used in further interpretation.

Lécuyer et al. [41] emphasized that the isotope fractionation factor between CO2 and H2O is salinity-dependent and that increasing salt contents (KCl or NaCl) results in an increasing overestimation of oxygen isotope ratios. Because water samples investigated in the frame of this study are characterized by varying salinities, the potential effect on the study results has to be reviewed. The most saline water sample is sample N6 (18103.3 mg/l; Table 1), resulting in an overestimation of δ18O by less than 0.1, which is rather small but significant relative to analytical uncertainties. However, the quantification of the mixing between end-member waters from Malmian aquifer and Cenomanian/Eocene reservoirs performed by using 2H/1H and 18O/16O as natural traces will be negligently affected by lack of salinity-dependent corrections. In addition, it is unknown if the published isotopic compositions of samples were corrected. Therefore, the correction is not applied for the samples measured in the frame of this study.

5.2. Possible Processes Influencing the Light Hydrocarbon Fraction

The methylcyclohexane/toluene (Mch/Tol) and cyclohexane/benzene (Ch/B) ratios are cross-plotted in Figure 6 to show the significant variations in the relative contents of light hydrocarbon compounds between different oils. The observed differences may originate from natural primary and secondary processes as well as from poor storage of archival samples (e.g., changes due to evaporation losses). Therefore, it is critical to determine the consequences of each process that can generate compositional differences.

5.2.1. Influence of Source Rock Facies

Significant differences are observed in generation of hydrocarbons from marine and terrestrial organic matter. Type III kerogen yields predominantly aromatic hydrocarbons (e.g., benzene and toluene [19]), while type I/II kerogen produces more n-, iso-, and cycloalkanes (e.g., [42]). Moreover, Odden et al. [43] showed that increasing contents of terrigenous organic matter in source rocks results in higher concentration of aromatics, cyclohexane, and methylcyclohexane compared to cyclopentanes and acyclic hydrocarbons. Because the observed trend leads towards higher Mch/Tol and Ch/B ratios, increasing content of coaly facies in source rock can be ruled out (Figure 6).

5.2.2. Evaporative Fractionation

An alternative process that could influence the pattern of light hydrocarbons is called evaporative fractionation (Thompson, 1987) and describes the loss of light hydrocarbons from an oil phase (in reservoir or during migration) resulting from a later gas charge. During this process, the gas phase of a gas-saturated oil escapes from the oil, leaving behind residual oil strongly enriched in toluene and moderately enriched in cycloalkanes. In contrast, n-alkanes (e.g., heptane) will be preferentially dissolved in the escaping gas phase (Thompson, 1987). Therefore, evaporative fractionation would result in a trend opposite to the observed one (Figure 6).

5.2.3. Evaporation Losses of Light Fraction during Sample Handling

The current study is based on the quantification of relative volatile hydrocarbons. Therefore, it is critical to discuss the effect of possible losses caused by evaporation during sampling, storage, or laboratory handling. Evaporation of hydrocarbons depends on different factors like group type (linear, branched, cyclic, and aromatic), isomeric structure, molecular weight, and bulk composition of the sample [44]. However, laboratory controlled evaporation of crude oil showed that this process is primarily controlled by differences in boiling points [23]. Consequently, based on differences in vapor pressure of mixtures of two components (Mch/Tol; Ch/B), it is possible to estimate general evaporation trends. Thus, incidental loss of light hydrocarbons will result in a strong decrease of the Mch/Tol ratio and a small increase in the Ch/B ratio (see Figure 6). Hence, evaporation cannot explain the observed strong increase in Mch/Tol and Ch/B ratios, although a minor effect on samples, which have been stored for a long time (archive samples), cannot be ruled out completely.

5.3. Oil-Water Interaction

In the previous section, it could be shown that source rock facies, evaporative fractionation, and losses during sample storage and handling are not responsible for the observed BTEX depletion. In contrast, the trend in Figure 7 is interpreted to reflect the selective removal of relatively soluble aromatics during contact with water. The oil-water interaction can occur during the migration from the source rock to the reservoir and/or after oil accumulation in the reservoir. The longer the migration distance, the higher the potential interaction between oil and water. Interestingly, there is no correlation between migration distance and BTEX depletion. For example, samples A and P experienced similar migration distance (~35 km according to [3]; see Figure 1(c) for sample location), but only sample A is strongly depleted in aromatic hydrocarbons (MCH/Tol + Ch/B = 342.8). In comparison, the sum of MCH/Tol and Ch/B is only 7.7 for sample P. The same is true for the strongly altered U1–4 samples (average MCH/Tol + Ch/B = 108.7) and the less affected AB sample (MCH/Tol + Ch/B = 19.5) which show similar migration distances. Therefore, the observed water washing phenomenon is most probably not controlled migration distance. Hence, water washing probably occurs in the reservoir.

Significant removal of BTEX from bulk oil composition requires a sufficient volume of BTEX-undersaturated water. Considering hydrostatic conditions, it is unlikely that the concentration gradient between oil and the volume of associated water would be high enough to explain the observed strong BTEX depletion. This suggests that water washing is related to the Malmian aquifer, which is the only main aquifer, which is under dynamic condition and, thus, may provide sufficient undersaturated water that drives diffusion.

Because water from this aquifer is characterized by low salinity and light isotopic composition [7], water washing parameters are plotted against Cl content, δ2H [V-SMOW] and δ18O [V-SMOW] values of water coproduced with oil in Figure 7. Indeed, increasing removal of BTEX components correlates with decreasing Cl contents and isotope values (Figure 7). This shows that the original connate brines in water washed Cenomanian/Eocene reservoirs have been mixed with fresh water. The relation between water from the Malmian aquifer and water washed oil is especially obvious in the U field, where Cenomanian reservoir rocks directly overlie fresh water bearing Malmian carbonates (Figure 8). The high permeability of the Malmian carbonate, a prerequisite for significant water flow, is proven by losses of drilling mud (industry data) and allows the reinjection of thermal water in well U (Figure 8, [26]).

5.4. Impact of Water Washing on Oil Properties

Water washing can significantly change oil properties. Lafargue and Barker [11] and Kuo [45] have shown that water washing can affect biomarker ratios, like the methyl phenanthrene index (MPI), a classical maturity parameter [46], and the dibenzothiophene/phenanthrene (DBT/Ph) ratio, a parameter that is often used for oil-source and oil-oil correlations [32]. Remarkably, samples from the water washed U field are characterized by the highest MPI value of all samples in the study area and a DBT/Ph ratio, which is lower than that in comparable oils [2]. Therefore, these parameters have to be used with caution for source rock and maturity evaluations for oil samples, suspected to be water washed.

Water washing reduces API gravity because it is particularly effective for C15− hydrocarbons [11]. To test this hypothesis, the ratio of hydrocarbon fractions with less and more than 15 carbon atoms, respectively (C15−/C15+), is cross-plotted versus the sum of two ratios between cycloalkanes and aromatic hydrocarbons in Figure 9(a). It shows that decreasing concentrations of aromatic hydrocarbons correlate fairly well with decreasing contents of light hydrocarbons. This indicates that, to some extent, API gravity of oils from AFB is controlled by water washing. To exclude any effect of losses of light hydrocarbons during sampling or handling, the C15−/C15+ ratio is plotted against industrial API gravity data, obtained during decades of production (Figure 9(b), Table 1).

Water washing is often accompanied by biodegradation. In the present case, water washed samples show no signs of biodegradation in the n-C7+ range. However, Gruner et al. [47] detected metabolites of BTEX in reservoir water from water washed fields. This suggests that water washing facilitates biodegradation by making BTEX bioavailable.

5.5. Implication on the Understanding of Hydrological System

The Malmian hydrological system in the AFB is of great economic relevance, because it supplies a high number of hydrogeothermal installations and thermal spas. Assurance of sustainable use of the water is of prime importance and resulted in the establishment of a numeric, thermal hydraulic model ([8]; Figure 3).

The present study shows that heavily water washed fields occur in Cenomanian reservoirs in fields U and V and in Eocene reservoirs in fields A to D. Strong support for mixing of meteoric water and connate brine in these fields, U and D, is provided by the isotopic composition of water. For a simple quantification of mixing, two end-member waters are defined: (i) water from the Malmian aquifer and (ii) sample L3 from an Eocene reservoir (Table 1 and Figure 10). Based on this assumption, it is concluded that up to 77% of connate Cenomanian/Eocene brine has been replaced by Malmian water in field U. The percentage of meteoric water in field D is about 65%.

As discussed above (Figure 8), water washing in the U (and V) fields agrees with the current hydrogeological model. However, field D is located east of the pinch-out of Malmian rocks and the Eocene reservoir directly overlies crystalline basement (Figure 3, [48, 49]). Moreover, field D is located outside the boundaries of the regional thermal water system, proposed by the Bayerisches Landesamt für Wasserwirtschaft [8] (Figure 3). Within this context, it is noteworthy that an extensive aquifer is indicated by a strong water drive keeping the reservoir pressure constant, despite of decades of oil production. Therefore, the boundaries of the established flow model need modifications. In contrast to oils from the northern part of study area, oils from the southern part typically show no evidence of water washing (e.g., E, I, J, and T fields; Figure 6). This is in agreement with stagnant conditions in the reservoir and the Malmian aquifer in this area (Figure 3).

6. Conclusions

Light hydrocarbon geochemistry of 57 oil samples from Cenomanian and Eocene reservoirs in the Austrian sector of the Alpine Foreland Basin has been investigated in the frame of this study. Strong depletion of BTEX compounds in some oils implies water washing. Additional observed features of water washing include a decrease in API gravity related to depletion in low molecular saturated components.

Water coproduced with water washed oil shows a progressive reduction in chlorine content (min.: 888 mg/l, measured in the frame of this study) and depletion in 2H and 18O isotopes (−51.4 and −5.8, resp., measured in the frame of this study), indicating that connate brines have been partly replaced by meteoric water characteristic of the underlying Malmian carbonates, the main aquifer for geothermal water in the basin.

Most strongly affected oils are located in the shallow northern and northeastern part of the study area (fields A, D, U, and V). The U and V fields produce from Cenomanian reservoirs directly overlying the Malmian aquifer. In these fields, a hydraulic connectivity between the reservoir and the aquifer could be proven. Fields A and D are located east of the extension of the Malmian aquifer and produce from Eocene reservoirs. The Eocene reservoir rocks of field D rest directly on crystalline basement. This suggests that Malmian water is discharged (north-eastwards) through crystalline basement rocks and that previous flow models of the regional geothermal aquifer have to be reevaluated.

In contrast to the shallow northern fields, fields in the deep southern part of the basin (e.g., E-J, L-P, R, W-Z, and AD) are apparently not affected by water washing. The water in these fields shows relatively high salinity.

The results emphasize the importance of combining data from the petroleum and geothermal industry, which are often handled separately: recognition of active water flow may help to predict gravity and viscosity anomalies, biodegradation risk, and the presence of hydrodynamic traps. Additionally, identification of water washing helps to improve flow models of the underlying Malmian aquifer.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.


The authors would like to acknowledge Rohöl-Aufsuchungs AG for access to samples, geological documentation, and publication permission. Collaboration with Christoph Janka, Rohöl-Aufsuchungs AG, led to the understanding expressed in this paper. The presented data were obtained within the frame of FFG Bridge Project 836527 between Montanuniversität Leoben and Rohöl-Aufsuchungs AG. The authors would like to thank Johannes Rauball for linguistic corrections.