pT-Effects of Pleistocene Glacial Periods on Permafrost, Gas Hydrate Stability

lable at ScienceDirect

Marine and Petroleum Geology 27 (2010) 298–306

Contents lists avai

Marine and Petroleum Geology

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

pT-effects of Pleistocene glacial periods on permafrost, gas hydrate stabilityzones and reservoir of the Mittelplate oil field, northern Germany

S. Grassmann a,*, B. Cramer a, G. Delisle a, T. Hantschel b, J. Messner c, J. Winsemann d

a Federal Institute for Geosciences and Natural Resources (BGR), Stilleweg 2, D-30655 Hanover, Germanyb Schlumberger Center of Excellence for Petroleum Systems Modeling/IES GmbH, Ritterstraße 23, D-52072 Aachen, Germanyc State Authority of Mining, Energy and Geology (LBEG), Stilleweg 2, D-30655 Hanover, Germanyd Gottfried Wilhelm Leibniz University Hannover, Institute of Geology, Callinstraße 30, D-30167 Hanover, Germany

a r t i c l e i n f o

Article history:Received 25 November 2007Received in revised form31 July 2009Accepted 1 August 2009Available online 6 August 2009

Keywords:GlaciationPalaeo-gas hydratesPermafrostReservoir temperatureOverpressurePleistoceneMittelplateBasin modelling

* Corresponding author. Present address: ExxonMGmbH (EMPG), Riethorst 12, D-30659 Hanover, Germ

E-mail address: [email protected]

0264-8172/$ – see front matter � 2009 Elsevier Ltd.doi:10.1016/j.marpetgeo.2009.08.002

a b s t r a c t

During the past two million years low surface temperatures as well as episodically advancing ice sheetsfrom Scandinavia acted on the subsurface pT-regime of northern Germany. Their likely effects on thepetroleum system of Schleswig-Holstein were investigated. For the entire Quaternary mean annualground temperature (MAGT) was reconstructed at a resolution of 1000 years by calibrating oxygenisotope records from ODP-site 659 to the climate of northern Germany of the past 120 kyr. The resultingMAGT trend served as input to an ice sheet model and a permafrost model along a 2D section crossingthe petroleum bearing south-western part of Schleswig-Holstein. Here advances and retreats of theScandinavian ice sheet during Saalian and Elsterian glaciation Stages were reconstructed. Maximum icethicknesses of up to 1700 m and up to 20 periods of regional permafrost in northern Germany werereconstructed for the past 1.25 million years. Based on a basal heat flow of 50 mW/m2 permafrostthicknesses exceeded 100 m during most of these periods, temporarily extending down to depths ofmore than 300 m. Favourable surface temperatures and long durations of cold periods providedfavourable conditions for onshore gas hydrate stability zones at Mittelplate. Implementing these glacialdynamics into 2D basin modelling (PetroMod, IES, Aachen, Germany) of the Mittelplate oil field revealsfive phases of gas hydrate stability at depths down to 750 m. The latest of these events occurred duringthe Weichselian about 20 kyr ago. The effect of the ice sheets on pore pressure in the subsurface stronglydepends on the hydraulic boundary conditions at the ice base (e.g. frozen vs. temperate ice sheet base).Excess pore pressure in the reservoir of more than 10 MPa during ice overriding is possible and probable.The calculated temperature effect of the Pleistocene cooling on the Mittelplate reservoir is in the range of3–7 �C. Even today temperature in the reservoir is still lowered by about 4 �C in comparison to pre-Pleistocene times. Despite the fact that a significant influence of glacial effects on petroleum generationcan be ruled out at Mittelplate, we state that pT-effects in reservoirs related to glacial processes informerly glaciated areas have been underestimated in the past.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

1.1. The Quaternary of Schleswig-Holstein

The Quaternary is characterized by global cooling and sub-divided into cycles of colder glacial and warmer interglacial periods(e.g. Ehlers, 1994). During some glacial periods, ice sheets devel-oped on the circum-polar landmasses of Scandinavia and spreadover vast areas of northern Central Europe.

obil Production Deutschlandany. Tel.: þ49 511 641 2225.

m (S. Grassmann).

All rights reserved.

Studies on the number and the extent of ice sheets coveringSchleswig-Holstein are mainly based on the interpretation ofglacial landforms and/or the distribution of glacial deposits innorthwest and central Germany. It was found that the first advanceof the Scandinavian ice sheets across Schleswig-Holstein startedabout 350,000 years ago at the beginning of the Elsterian glaciation(e.g. Ehlers and Gibbard, 2003, 2004).

Ehlers (1990, 1994) and Stephan (1995) suggested that duringthe Elsterian glaciation two ice sheets crossed Schleswig-Holstein.The first one advanced from the north and the second one from theNE. The Elsterian glaciation of north-western Germany was asso-ciated with the formation of glacial channels (‘‘tunnel valleys’’)which were eroded by highly pressurized subglacial melt water and

Fig. 1. Study area and location of modelled 2D line intersecting the Mittelplate oil field.

S. Grassmann et al. / Marine and Petroleum Geology 27 (2010) 298–306 299

were subsequently filled with glacial deposits (Piotrowski, 1991;Piotrowski and Tulaczyk, 1999). The southern boundary of theElsterian glaciation is not known in much detail. Most reliable datafor the reconstruction of ice sheet thickness and the position offormer Elsterian ice margins come from Saxony and Thuringia(Wagenbreth, 1978; Eissmann, 1995; Unger and Kahlke, 1995;Eissmann, 1997; Junge, 1998), whereas to the west the extent of theElsterian ice sheet is poorly defined, being obscured by an overprintof the Saalian glaciation (Ehlers et al., 1984; Kaltwang, 1992;Klostermann, 1992, 1995; Caspers et al., 1995). However, all dataindicate a complete ice coverage of the Schleswig-Holstein area(Ehlers, 1990, 1994; Stephan, 1995; Ehlers and Gibbard, 2003;Ehlers and Gibbard, 2004).

During the Holsteinian interglacial, marine clays were depositedin Schleswig-Holsteins Tunnel Valleys. The beginning of the Saalianis characterized by a period of cold climate but without ice coverage(Ehlers, 1994). Three subsequent major ice advances are known forthe Saalian glaciation in northern Germany. Tills are separated byglaciolacustrine and glaciofluvial deposits, but no interglacialsediments have been found (Skupin et al., 1993; Ehlers, 1994;Caspers et al., 1995; Stephan, 1995).

Ehlers (1990) suggested that the northern part of Schleswig-Holstein was permanently covered by ice during the last two Saa-lian stages. Data from the North Sea offshore Schleswig-Holsteinindicate an ice free, periglacial area towards the west (Schwarz,1996). During the Eemian interglacial the area of the North Sea wasaffected by a transgression and marine sediments were deposited.In topographically higher areas of Schleswig-Holstein limnic sedi-ments were deposited in morphological depressions (Stephan,1995).

Most authors assume that the first advances of the Weichselianice sheet did not occur before 25 kyr BP (Stephan and Menke, 1993;Ehlers, 1994). Therefore, during most of the Weichselian timeSchleswig-Holstein was not covered by ice. The dynamics ofadvancing and retreating ice sheets in northern Europe during theWeichselian is summarized by Boulton et al. (2001). From 25 toabout 13 kyr BP at least 3 ice advances from the northeast areknown from Schleswig-Holstein, whose deposits make up most ofrecent landforms (Stephan, 1995).

1.2. Mittelplate petroleum system

The area of Schleswig-Holstein, as part of the Central EuropeanBasin System (CEBS), is an intensely explored and developedhydrocarbon province. Extensive basin modelling was carried outin this basin (e.g. Bueker et al., 1995; Neunzert et al., 1995, 1997;Erdmann, 1999; Hertle et al., 1999; Petmecky et al., 1999).

Dominating structural features are Permian (Rotliegend andZechstein) and Triassic salt diapirs, mostly striking NNE to SSW.Halokinetic movements started during the Triassic and resulted inthe formation of thick sedimentary sequences within rim synclinesbetween the salt diapirs. The most important rim synclines on theHolstein Block are the Westholstein and the Ostholstein Troughs.The importance of these Jurassic Troughs for the petroleum geologyof Schleswig-Holstein is linked to the occurrence of thick layers ofthe prominent Liassic Posidonia Shale source rock (Welte, 1979;Wehner et al., 1989). The Posidonia Shale also charged the reser-voirs at the Mittelplate oil field (Muller et al., 2004). The overlyingdeltaic Middle Jurassic Sandstones serve as prolific reservoir rocks(Berners et al., 1992). The two Jurassic Troughs contain the majorpetroleum system known so far in Schleswig-Holstein.

The Mittelplate field is the biggest oil accumulation in Germanyat present and annual production is in the order of about2.2 Mio tons of crude oil (Pasternak et al., 2006). The field is locatedin the tidal flat area at the south-western nearshore of Schleswig-

Holstein (Fig. 1). Because of its economic significance, petroleumgeology of the Mittelplate area is well known (Berners et al., 1992;Junker and Dose, 2001; Langhans et al., 2003; Muller et al., 2004).

The dynamics of petroleum generation, migration and entrap-ment were most recently investigated by an integrated petroleumsystems modelling approach using the PetroMod (IES, Aachen,Germany) software (Grassmann et al., 2005). By integrating allavailable geological data, the burial history and the temporaldevelopment of the temperature field were reconstructed. 2D basinmodelling included the organic geochemical analysis of maturityindicators, the reconstruction of sedimentation and the calibrationof the thermal history as well as the modelling of petroleumgeneration and migration. The complex history of salt diapirism isthe key in understanding the development of the Mittelplate field.Maturation and petroleum generation within the Posidonia Shalewere governed by temporal and spatial changes in temperaturecaused by extensive salt tectonics within a complex structuralframework.

1.3. Influence of cyclic glaciations on petroleum systems

The most obvious impacts of ice sheets on sedimentary basinsare changes in the structural and depositional settings. During iceadvance glaciers may erode sediments at their base, during meltingof an ice sheet glacial sediments can be deposited. In addition, theoverburden of the ice leads to isostatic subsidence of the basinwhile unburden after ice retreat results in isostatic uplift. Thespatial and temporal variations of these processes duringadvancing and retreating ice sheets can lead to changes in thepathways and the intensity of hydrocarbon migration and can evencause spilling of existing petroleum accumulations. For example,

S. Grassmann et al. / Marine and Petroleum Geology 27 (2010) 298–306300

Hermanrud et al. (1991) showed for the northern North Sea, thaterosional events during the Quaternary were related to deglacia-tion processes. These events had an effect on the palaeo-tempera-ture distribution and in consequence on hydrocarbon generationand distribution. Cloethingh et al. (1992) and Sales (1992) pointedout that for the Quaternary it is difficult to differentiate betweenuplift movements caused by glacially induced isostasy and non-glacial tectonic uplift in northwest Europe. However, the litho-sphere responds to even slight changes (<10 MPa) in local stressconditions induced by mass and fluid transfer associated withQuaternary ice sheets (Thorson, 2000). The deglaciation of the lastPleistocene ice sheet of Scandinavia even today leads to an isostaticuplift of up to several millimetres per year (Thoma and Wolf, 1999).Such uplift movements may have a strong effect on the petroleumsystems by changing the temperature and decreasing the porepressure within the sediments. The most important effect of uplifton hydrocarbon distribution is the release of formerly dissolved gasfrom the pore water leading to new accumulations of natural gas(Cramer et al., 1999, 2001).

Changes in the physical and chemical habitat of a basin due toglaciation also indirectly induce processes related to the occurrenceof petroleum. The large commercial gas accumulations within theAntrim Shale of the Michigan Basin, for example, are suggested tobe sourced from microbial methane generation within these shales(Martini et al., 1998). This microbial activity was activated bychanges in the hydro-geological setting related to the complexPleistocene glacial history.

In order to demonstrate a broader impact of cyclic glaciations onsedimentary basins, Lerche et al. (1997) conducted a modellingstudy along a synthetic 2D-line. The numerical experimentsdemonstrated that large ice sheets and their variability in cyclicglaciation frequency have a major influence on physical andchemical properties of sedimentary basins. In this way they canlead to significant changes in petroleum systems. Johansen et al.(1996), Solheim et al. (1996) and Cavanagh et al. (2006) showedthat these processes strongly affect deep temperature and pressurefields and alter hydrodynamic conditions in sedimentary basins. Inthis way they can lead to significant changes in the habitat anddistribution of the petroleum. However, a direct proof of Pleisto-cene glacial processes effecting the deeper subsurface of northernGermany is missing so far.

1.4. Temperature related processes

Surface temperature is one key parameter determining the heatflow and temperature distribution within subsurface sediments(Johansen et al., 1996; Solheim et al., 1996). The most obvious effectof reduced sediment temperatures over prolonged periods of timeis the freezing of pore water leading to the development ofpermafrost (Delisle et al., 2007). Permafrost blocks the pore spaceand reduces rock permeability. In extreme cases, this may preventupward directed migration of hydrocarbons and can lead topetroleum accumulations at the base of the permafrost. In addition,within and below the base of permafrost favourable conditionsexist for the generation of gas hydrate accumulations. A discussionabout the impact of permafrost and continental gas hydrates on thedistribution of natural gas in the sub-arctic West Siberian GasProvince was provided by Cramer et al. (1997). In this area a directinfluence of gas hydrates on the accumulation of the giant gasreservoirs is not plausible. Only in some cases gas hydrate stabilityreaches down into conventional gas reservoirs, leading to thedevelopment of a secondary gas hydrate cap within the reservoir.

Another important effect of reduced temperatures in a basin isa lowered rate of petroleum generation within source rocksreducing recharge rates of the reservoirs. Lerche et al. (1997)

showed that variable frequencies of glaciations can lead to spatialtemperature distortions in sedimentary basins. As a result hydro-carbon generation rates vary in time and space.

1.5. Pressure related processes

Large continental ice sheets also affect the pressure regime ofa sedimentary basin. Riis (1992) suggested that the overburden ofa thick ice cap will result in excess hydrostatic pressure below theice. This would cause a free gas cap to contract, while it wouldexpand when the ice melts. This so-called ‘ice pump’ effect wasbelieved to be responsible for the redistribution of oil below gasreservoirs in the North Sea area. According to Lerche et al. (1997),overpressure may also lead to fracturing of sediments, which cau-ses leakage of water, oil and gas. However, Thorson (2000) pointedout that pore pressure beneath ice sheets can only increase, if thefluid pressure fully compensates the increased vertical load of theice. This is the case in a wet-based ice sheet when liquid wateroccurs near the ice/bed interface and if there is unlimited access ofthe water to the base in an isotropic crust. Grollimund and Zoback(2000) observed a systematic variability of horizontal stress in thenorthern North Sea. In this area, the pore pressure roughly followsthis stress trend today. This implies high overpressure in areas withhigh horizontal stress and pore pressure close to hydrostaticconditions where lateral stress is decreased. Analytical andnumerical models of plate flexure (Grollimund and Zoback, 2000)suggest that the observed stress trends are the result of deglacia-tion. Overpressure appears to be caused at least partly by deglaci-ation and associated flexure.

2. Methods

2.1. Reconstruction of annual ground temperatures and ice sheetthicknesses

Our basin model requires two sets of input data to assess theimpact of the Pleistocene climate on the subsurface pT-regime: themean annual ground temperature (MAGT) and the temporal vari-ability of the ice-sheet thickness. Temperature variations in north-central Europe during the Weichselian have been reconstructed onthe basis of proxy data such as botanical macrofossils and pollencontent in Weichselian sediments by Caspers and Freund (1997).Based on their work a reconstruction of the mean annual groundtemperatures (MAGT) during the Weichselian as well as anassessment of the temporal variability of permafrost developmentin northern Germany was presented by Delisle et al. (2003). Proxydata from pre-Eemian sub-aerial deposits tend to be too sparse toenable us to reconstruct a continuous record of the regional climatefor the entire Pleistocene. Therefore, as an alternative approach, anestimate for an MAGT-curve for the Pleistocene of northernGermany based on the marine proxy record from the ODP-sites 659was developed by Delisle et al. (2007). In the basin model theMAGT-values were integrated as upper boundary condition in thecalculation of the basin’s heat budget, with a temporal resolution of1000 years. As ice sheets advance or retreat MAGT-values cannot beused as an upper thermal boundary condition for sediments, butmust be replaced by the calculated temperatures at the base of theglacier.

The timing of glacial advance and retreat was modelledaccording to a computer code by Delisle (1991), which is driven a)by a function relating air temperatures to precipitation and b) byice-rheological functions, which determine ice deformation andflow from the internal temperature field of the modelled glacier.The surface temperatures of the modelled glaciers were obtained

S. Grassmann et al. / Marine and Petroleum Geology 27 (2010) 298–306 301

by assuming a vertical lapse rate in the atmosphere of 5.5 �C/kmfrom the ground.

The deformation of the glacier was modelled based on the creeplaw for ice by Nye (1952, 1965):

_e ¼ A,sn (1)

where _e¼ creep rate (s�1); A¼ ice stiffness parameter (s�1 kPa�3);s¼ driving stress (s¼ r$g$h$sin(a)) with r¼ density of ice;g¼ gravitational acceleration; h¼ height of ice column; a¼ surfaceslope.

Horizontal (u) and vertical (v) ice flow components are relatedto _e by the relation _e ¼ 0:5$du=dz and dv=dz ¼ �du=dz.

Surface slopes are in final consequence related to differentialsnow precipitation, which in turn is related to air temperatureabove the glacier. Precipitation rates during glacial stages are notknown. We chose a ‘‘plausible’’ relation for precipitation m (iceequivalent):

m ¼ 15 cmð40þ TairÞ=40: for 0 �C < T < �40 �C (2)

This approach results in a relatively massive ice shield, whichmight be considered to represent the upper end of plausible iceshield configurations. The resulting maximum ice thicknesscalculated for Mittelplate was about 1700 m at the peak of theSaalian glaciation.

2.2. 2D basin model

The structural and thermal evolution along a 2D section crossingthe Mittelplate field (line OMS) was reconstructed utilizing 2Dbasin modelling (Grassmann et al., 2005). For modelling purposesthe software suite PetroMod by IES (Integrated ExplorationSystems, Aachen, Germany) was applied. In order to reconstruct thetemperature development of the sedimentary sequencethroughout the Mesozoic, computed vitrinite reflectance valueswere calibrated with the measured maturity of the sedimentaryorganic matter. For this purpose the EASY%Ro method (Sweeneyand Burnham, 1990) was applied. The match of modelled vitrinitereflectance (Easy%Ro) with measured vitrinite reflectance valueswas realized assuming a constant basal heat flow of 50 mW/m2

since the Early Jurassic. As a result of the maturity calibration, wewere able to compute the temperature distribution in the sedi-ments along the 2D-line (Fig. 2). Because of the ‘‘chimney effect’’ ofdiapirs (e.g. Neunzert et al., 1997; Delisle, 1998a; Cramer et al.,2005) the temperature at the base of the salt structures is signifi-cantly lowered (about 20 �C for the Mittelplate area). Obviously, theMittelplate area is part of a region with a low increase in temper-ature with depth and a moderate heat flow. The thermal evolutionof the area is characterized by rather uniform conditions persistingthroughout the Mesozoic and Early Cenozoic.

For calculating permafrost aggradation and decay throughoutthe Quaternary we used the numerical model described earlier(Delisle, 1998b). This model in particular includes the effect ofrelease and uptake of latent heat during freezing and melting ofpermafrost but has its limitations owing to a purely heat conductiveapproach.

The presented model calculates the time-dependent verticalextent of the permafrost zone during each cold period. In thecontext of the presented numerical model, the boundary definesthe level, where release (freezing) or uptake (melting) of latent heathas been completed. Various soil parameters influence permafrostgrowth and decay as well. Of major importance is the overallcontent of fluids in the soil, which at the freezing point will freeze(melt) after the latent heat content has been released (taken up).The amount of latent heat (80 kcal/kgpore water) released or being

taken up is a key component that controls the rate of permafrostgrowth or decay. The freezing/melting point of ice in permafrostalso depends on the soil type. The depression of the freezing pointin particular in fine-grained soils is a well-known effect (see e.g.Tsytovich, 1957; Washburn, 1979). However, for most clayey tosandy soils 85–97% of the pore water is frozen below a soiltemperature of �0.6 �C. To account for this effect, our calculationsinclude as approximation a mean effective freezing point of�0.6 �C, where upon freezing all latent heat has been released andupon melting is taken up. Finally, the thermal conductivity of rocksand heat flow density from the Earth’s Interior are additionalcontrolling factors.

We integrated the model of Delisle (1998b) into the existingPetroMod petroleum systems modelling technology by applying aninstant shift in lithology to intervals with a calculated temperaturebelow �0.6 �C. The associated changes of thermodynamic param-eters are summarized in Table 1.

3. Results and discussion

3.1. Permafrost and gas hydrates at Mittelplate

The MAGT-reconstruction (Fig. 3A) combined with the recon-structed ice sheet coverage (Fig. 3B) forecasts the existence ofpermafrost in northern Germany during the glacial/interglacialcycles in Central Europe during the last 1 million years (Fig. 3C).Furthermore, during the early Pleistocene prolonged periods withground temperatures below zero and phases of permafrostdevelopment existed. This is in good agreement with geologicalevidence of permafrost development in the Menapian, Eburonianand Tiglian stages, as has been reported by various workers. Asummary of the available evidence was presented by Vanden-berghe (2001), whose results appear to reflect well our perceptionof the sequence of cold stages in northern Germany for the last 1million years. The Weichselian, Saalian- and Elsterian-stages areclearly represented by permafrost as well as the Cromer- andBavel-stages.

According to our model, permafrost was a periodic and commonfeature in northern Germany during the Pleistocene, in many casesextending deeper than 100 m into the ground (Fig. 3C). The modelimplies climatic conditions were too warm for the formation ofpermafrost 2–1.2 Myr ago. However, the reconstruction for thewhole Pleistocene predicts prolonged sub-zero ground tempera-ture episodes and phases of permafrost development. Investiga-tions on the permafrost depth were based on two approaches: anestablished algorithm for the calculation of permafrost growth anddecay (Delisle, 1998a,b, 2003, 2007) was used to cross check theresults of the integrated basin model.

A comparison of permafrost for both models is shown in Fig. 3C.In general, good agreement was achieved in timing of permafrostdevelopment and decay. The increased heat flow above salt domesimpedes the downward propagation of permafrost. Our modelsuggests a reduction of the maximum vertical extent of permafrostabove salt domes by up to 100 m in comparison with the neigh-bouring rocks (Fig. 5).

Differences in thickness between both models are caused bydifferent basal heat flow scenarios. Delisle et al. (2007) applied75 mW/m2 whereas Grassmann et al. (2005) used a calibrated basalheat input of 50 mW/m2 as. However, both results appear to reflectwell our perception of the sequence of cold stages in northernGermany for the last 1 million years. The Weichselian-, Saalian- andElsterian-stages are clearly represented as well as the older stages.Also prolonged permafrost conditions in periods older than 1million years are in agreement with available geologic evidence(Vandenberghe, 2001).

Fig. 2. Modelled WE section crossing the Mittelplate oil field. Figure comprises outline of Permian evaporates, West Holstein Trough and location of VR-calibration wells. Recenttemperature distribution shows typical effect of distorted temperature field owing to contrasting thermal conductivities between salt diapirs and surrounding lithologies. Below:measured vs. calculated vitrinite reflectance calibration. Both after Grassmann et al., 2005.

S. Grassmann et al. / Marine and Petroleum Geology 27 (2010) 298–306302

Low surface temperatures lasted sufficiently long thus leading tothe development of physical conditions favourable for the forma-tion of gas hydrates below the permafrost zone (Fig. 4). Accordingto Fig. 3D, gas hydrates were stable down to depths of 750 m during

Table 1Parameters used for modelling of ice sheets and permafrost within PetroMod.

Ice lithology parameters for PetroMod

Density [kg m�3]Porosity [%]

Fluid permeability[logmD]

Thermalconductivity[W m�1 K�1]

Heat capacity[kcal kg�1 K�1]

1.000 �15 2.33 0.491

Permafrost parameters for PetroMod

Pore spacefilling

Fluid permeability[logmD]

Thermalconductivity[W m�1 K�1]

Heat capacity[kcal kg�1 K�1]

‘‘Water’’ non-permafrost Rock specific 0.6 1.0‘‘Ice’’ permafrost �16 2.33 0.49

Heat of crystallization: 80 kcal kg�1.

five different periods in the past 750,000 years. These resultssuggest that conditions for gas hydrate stability persisted several10,000 years at Mittelplate and thicknesses were up to 500 m.

The last period of gas hydrate stability ceased just recently at thebeginning of the Holocene. Earlier gas hydrate stability zonesexisted during Cromer- and Bavel-stages due to the prolongedexposure of uncovered ground to significantly lowered groundtemperatures. Favourable conditions for gas hydrate stability werelacking during Saalian- and Elsterian glacial stages due to overlyingice sheets causing smoothed temperature at its base.

Fig. 5 shows lateral variations of permafrost and depth of the gashydrate stability zone along the modelled line. The presence of saltdomes tends to reduce the maximum depth of overlying perma-frost by as much as 100 m, lowering the base of the gas hydratestability zone (GHSZ).

Despite the existence of the gas hydrate stability zones duringthe Pleistocene it is not clear if gas hydrates have in fact formed inconsiderable amounts. Thermally generated natural gas was notpresent in sufficient quantities for the formation of actual gashydrate layers (Grassmann et al., 2005). Microbial methanegeneration probably contributing to the total gas budget inthe relevant depth range for gas hydrates was not considered in the

Fig. 3. Influence of Quaternary climate effects on the subsurface at Mittelplate: A d MAGT-curve; B d ice sheet thickness; C d permafrost depth; D d gas hydrate stability zone;E d temperature effect in the reservoir; F d possible maximum pressure effect in the reservoir. Stratigraphy according to Streif (2004).

S. Grassmann et al. / Marine and Petroleum Geology 27 (2010) 298–306 303

basin model. However, even if microbial gas was periodicallytrapped in gas hydrate layers during the Pleistocene, no significanteffect of these gas hydrates on the Mittelplate oil field is expecteddue to the different depths of gas hydrate stability zones and theexisting Middle Jurassic reservoirs (Fig. 2).

3.2. Pressure and temperature effects in the Mittelplate reservoir

It is known that large continental ice sheets affect the pressureregime of an underlying sedimentary basin. Riis (1992) and Solheimet al. (1996) suggested that overburden of a thick ice cap may resultin hydrostatic pressure in excess below the ice. Thorson (2000)pointed out that pore pressure beneath ice sheets can only increase,if the fluid pressure fully compensates the increased vertical load ofthe ice. This is the case in a wet-based ice sheet with fluid water tonear surface and unlimited access of the water to the base in anisotropic crust. In the basin model along the modelled section the

ice sheets were defined as water-saturated rock layer with litho-logical parameters of the solid phase (e.g. density, heat conduc-tivity) defined as ice. Doing so, the lithostatic pressure duringperiods of glaciation increased according to changes in ice-sheetthickness. Fig. 6 shows this effect within the Mittelplate reservoirwith an increase from about 60 MPa to more than 75 MPa duringthe three main phases of glaciation.

In comparison to lithostatic pressure pore pressure effects are allthe same difficult to reconstruct. Crucial boundary conditions at thebase of the Pleistocene ice sheets are not known. To account for thelargest possible effect on pore pressure an extreme case accordingto Thorson (2000) was assumed where partly water-saturated icesheets are hydraulically connected to the groundwater. In thisscenario only the lowering of rock permeability during periods ofpermafrost formation inhibits deep penetration of the fluid pres-sure signal into the crust. With these presumptions fluid pressurein the Mittelplate reservoir increased from about 25 MPa to more

Fig. 4. Depth/temperature trends superimposed on the gas hydrates stability diagram.

S. Grassmann et al. / Marine and Petroleum Geology 27 (2010) 298–306304

than 40 MPa during ice sheet coverage (Fig. 6B). These values areconsistent with excess hydraulic pressures calculated by Cavanaghet al. (2006), who used comparable but more simplified input datato calculate the effects of ice-sheet coverage for the Barents SeaBasin.

Surface temperature is a key parameter to heat flow andtemperature distribution within sedimentary basins. Apart fromthe most obvious effect of permafrost development described

Fig. 5. Lateral variation of permafrost and gas hydrate stability

above, the temperature field in rocks below the permafrost zonecan be significantly disturbed due to the cold glacial climate. Lercheet al. (1997) showed that variable frequencies of glaciations canlead to deep spatial temperature distortions in sedimentary basins.As a result source rock maturation and hydrocarbon generationrates may vary in time and space.

In order to assess the effect on temperature in the Mittelplatereservoirs, a standard scenario with constant surface temperaturesof 10 �C provided by the software PetroMod was compared withour glacial model including the MAGT trends for the Pleistocene(Fig. 6A). According to this, temperature in the reservoir wassignificantly lower during glacial periods than predicted by stan-dard basin modelling. This effect amounts to 7 �C cooling at thebeginning of the Elsterian glaciation. Even today the reservoir isstill colder by about 5 �C. The permafrost itself also has an influenceon the temperature in the reservoir. As the thermal conductivity ofice is higher than that of (liquid) water, heat in the sedimentarybasin is more rapidly transported through permafrost. This leads toan additional cooling effect below permafrost layers. In our modelthis additional cooling related to permafrost reaches up to 1 �C(Fig. 6B).

A decrease in reservoir temperature of 7 �C due to glacial coolingcorresponds to findings by Johansen et al. (1996) who modelled thesame 7 �C temperature drop for the Norwegian Shelf area. Incontrast to the work by Lerche et al. (1997) a significant influence ofthis effect on source rock maturation can be ruled out in case of theMittelplate petroleum system. In consequence, also petroleumgeneration rates were not lowered perceptibly in our model.

In contrast to the work by Lerche et al. (1997) a significantinfluence of glacial conditions on source rock maturation in thePosidonia shale along the OMS line can be ruled out. The sourcerock is too deep and duration of glacial cooling was not sufficient toslow down maturation and petroleum generation.

zone (GHSZ) depth at Weichselian and pre-Elsterian times.

Fig. 6. Computed temperature trends in the Mittelplate reservoir for three different scenarios. Maximum cooling effect of 7 �C is observed at the beginning of the Elsterianglaciation. The difference between the two SWI-T models is due to the enhanced thermal conductivity of permafrost compared to water-saturated pore volume.

S. Grassmann et al. / Marine and Petroleum Geology 27 (2010) 298–306 305

4. Conclusions

During the past 2 million years low surface temperatures as wellas periodically advancing thick glaciers from the northeast acted onthe pT-regime in the subsurface of the Mittelplate area. To accountfor these effects, mean annual ground temperatures at a resolutionof 1000 years were reconstructed for the entire Pleistocene. Thistemperature trend was generated by applying oxygen isotoperecords from ODP-site 659 calibrated to the climate of northernGermany on the basis of previously reconstructed surface temper-atures for the past 120 kyr. Ice sheets covered the area several timesduring the Saalian and Elsterian glaciations. According to our icesheet model used to reconstruct ice sheet advances and retreatsacross the section, maximum ice thicknesses of up to 1700 m werereached.

By integrating these parameters into 2D basin modelling, up to20 periods of permafrost were reconstructed for the past 1.25million years. At a calibrated basal heat flow of 50 mW/m2

permafrost thicknesses exceeded 100 m during most of theseperiods, temporarily extending down to depths of more than300 m. Favourable surface temperatures and long durations of coldperiods also provided the stability conditions of continental gashydrates at Mittelplate during the Pleistocene. Basin modellingreveals five periods during which gas hydrates could form at depthsdown to 750 m. The last permafrost and gas hydrate event ceasedabout 20 kyr ago.

Low surface temperatures and the hydraulic system of the icesheets also significantly influenced the pT-conditions in the oilreservoir of Mittelplate. The effect of the glaciers on pore pressurein the subsurface strongly depends on the hydraulic boundaryconditions at the base of the glaciers (e.g. permafrost vs. unfrozenground) which are not well known. Nevertheless, excess porepressures in the reservoir of more than 10 MPa during glacialoverburden are possible. The temperature effects of the Pleistocene

cooling on the Mittelplate reservoirs can be constrained muchbetter and are in the order of 3–7 �C. Even today the temperaturefield in the reservoir is lowered by about 5 �C compared to the earlyPleistocene. A significant influence of these findings on the petro-leum generation potential at Mittelplate can be ruled out.

Acknowledgements

The work was funded by the German Research Foundation(CR139/1-1/-2 and WI1844/4-1/-2) as part of the DFG PriorityProgramme 1135 ‘‘Dynamics of sedimentary systems under varyingstress regimes: The example of the Central European Basin’’.

The study greatly benefited from cooperation with Dr. ThomasHantschel who implemented for the project a ‘permafrost option’into the PetroMod basin modelling software. The access to data andthe interest of the two companies operating Mittelplate, RWE-Deaand Wintershall AG, is highly appreciated.

References

Berners, H.P., Lange, H., Schneider, W.M.,1992. The Mittelplate oil field, German NorthSea; geological and sedimentological characteristics [publisher varies] [locationvaries]. In: Spencer, Anthony (Eds.), Generation, accumulation and production ofEurope’s hydrocarbons; II. Special Publication of the European Association ofPetroleum Geoscientists, vol. 2. Federal Republic of Germany, pp. 69–76.

Boulton, G., Gustafson, G., Schelkes, K., Casanova, J., Moren, L., 2001. Palae-ohydrogeology and Geoforecasting for Performance Assessment in GeosphereRepositories for Radioactive Waste Disposal (Pagepa). Report: EU 19784 –Nuclear science and technology series. Office for Official Publications of theEuropean Communities, Luxemburg. pp. 147.

Bueker, C., Littke, R., Welte, D.H., Leischner, K., 1995. Numerical simulations ofcomplex temperature, burial, and erosion histories for sedimentary basins andtheir calibration; examples from western Germany. In: Anonymous (Ed.), AAPGinternational conference and exhibition; abstracts: AAPG Bulletin, vol. 79.American Association of Petroleum Geologists, Tulsa, OK, United States, p. 1201.

Caspers, G., Freund, H., 1997. Die Vegetation und Klimaentwicklung des Weichsel-Fruh- und –Hochglazials im nordlichen Mitteleuropa. In: Freund and Caspers(Ed.), Vegetation und Palaoklima der Weichsel-Kaltzeit im nordlichen

S. Grassmann et al. / Marine and Petroleum Geology 27 (2010) 298–306306

Mitteleuropa – Ergebnisse palaobotanischer, – faunistischer und geologischerUntersuchungen, vol. 4. Schriftenreihe der Deutschen Geologischen Gesell-schaft, Heft, pp. 201–249.

Caspers, G., Jordan, H., Merkt, J., Meyer, K.-D., Muller, H., Streif, H., 1995. III. Nie-dersachsen. In: Benda, L. (Ed.), Das Quartar Deutschlands. Gebr. Borntrager,Berlin, pp. 23–58.

Cavanagh, A.J., di Primio, R., Scheck-Wenderoth, M., Horsfield, B., 2006. Severity andtiming of Cenozoic exhumation in the southwestern Barents Sea. Journal of theGeological Society, London 163, 114.

Cloethingh, S., Reemst, P., Kooi, H., Fanavoll, S., 1992. Intraplate stresses and thepost-Cretaceous uplift and subsidence in northern Atlantic basins. Norsk Geo-logisk Tidsskrift 72, 229–235.

Cramer, B., Braun, A., Poelchau, H.S., Littke, R., 1997. Gashydrate und permafrost imkontinentalen nordlichen Westsibirien. DGMK Tagungsbericht 9706, 27–36.

Cramer, B., Poelchau, H.S., Gerling, P., Lopatin, N.V., Littke, R., 1999. Methanereleased from groundwater – the source of natural gas accumulations innorthern West Siberia. Marine and Petroleum Geology 16, 225–244.

Cramer, B., Schlomer, S.,, Poelchau, H.S., 2001. Uplift related hydrocarbon accumu-lations – the release of natural gas from groundwater. Geological Society ofLondon Special Publications.

Cramer, B., Gerling, P., Bornemann, O., 2005. 3D-Beckenmodellierung zur Herkunftvon Kohlenwasserstoffen im Salzstock Gorleben. Fruhjahrstagung des DGMKFachbereichs ‘‘Aufsuchung und Gewinnung’’. pp. 165–170.

Delisle, G.,1991. Causes and consequences of the advance of a continental ice sheet fromScandinavia to Northern Germany. Geologisches Jahrbuch A127, 507–517. Hannover.

Delisle, G., 1998a. The evolution of the natural temperature field of a salt dome ingeological time. Journal of Seismic Exploration 7, 251–264.

Delisle, G., 1998b. Numerical simulation of permafrost growth and decay. Journal ofQuaternary Science 13 (4), 325–333.

Delisle, G., Caspers, G., Freund, H., 2003. Permafrost in north-central Europe duringthe Weichselian: how deep? In: Proceedings of the Eighth InternationalConference on Permafrost, Zurich, pp. 187–191.

Delisle, G., Grassmann, S., Cramer, B., Messner, J., Winsemann, J., 2007. Estimatingepisodic permafrost development in northern Germany during the Pleistocene.IAS Special Pub. 39 (Glacial Sedimentary Processes and Products), pp. 109–120.

Ehlers, J., 1990. Reconstructing the dynamics of the North-West European icesheets. Quaternary Science Reviews 9, 71–83.

Ehlers, J., 1994. Allgemeine und historische Quartargeologie, p. 358, Stuttgart (Enke).Ehlers, J., Gibbard, P.L., 2003. Extent and chronology of glaciations. Quaternary

Science Reviews 22, 1561–1568.Ehlers, J., Gibbard, P.L. (Eds.), 2004. Quaternary Glaciations: Extent and Chronology.

Part I: Europe. Developments in Quaternary Science, vol. 2a–c, p. 488.Ehlers, J., Meyer, K.D., Stephan, H.J., 1984. Pre-Weichselian glaciations of North-West

Europe. Quaternary Science Reviews 3, 1–40.Eissmann, L., 1995. VIII. Sachsen. In: Benda, L. (Ed.), Das Quartar Deutschlands. Gebr.

Borntrager, Berlin, pp. 171–198.Eissmann, L., 1997. Das quartare Eiszeitalter in Sachsen und Nordostthuringen.

Altenburger Naturwissenschaftliche Naturforschungen 8, 1–98.Erdmann, M., 1999. Gas generation from overmature upper Jurassic source rocks,

northern Viking Graben. Berichte des Forschungszentrums Juelich, ISSN: 0944-2952, p. 128.

Grassmann, S., Cramer, B., Delisle, G., Messner, J., Winsemann, J., 2005. Geologicalhistory and petroleum system of the Mittelplate oil field, Northern Germany.International Journal of Earth Sciences 94 (5–6), 979–989.

Grollimund, B., Zoback, M.D., 2000. Post glacial lithospheric flexure and inducedstresses and pore pressure changes in the northern North Sea. Tectonophysics327, 61–81.

Hermanrud, C., Eggen, S., Larsen, R.M., 1991. Investigation of the thermal regime ofthe Horda platform by basin modelling: implications for the hydrocarbonpotential of the Stord basin, northern North Sea. In: Spencer, A.M. (Ed.),Generation, Accumulation, and Production of Europe’s Hydrocarbons, vol. 1.Oxford University Press, Oxford, pp. 65–73.

Hertle, M., Littke, R., Hantschel, T., Welte, D.H., 1999. 3D modelling of burial andtemperature history in the Permo-Carboniferous Saar-Nahe Basin (SW-Germany). In: Anonymous (Ed.), American Association of Petroleum Geologists1999 annual meeting: Annual Meeting Expanded Abstracts. American Associ-ation of Petroleum Geologists, vol. 1999, pp. A58–A59.

Johansen, H., Fjeldskaar, W., Mykkeltveit, J., 1996. The Influence of Glaciation on theBasin Temperature Regime. Global Planet. Elsevier, Change, 12, Amsterdam,Netherlands. pp. 437–48.

Junge, F.W., 1998. Die Bandertone Mitteldeutschlands und angrenzender Gebiete.Altenbg. nat. Wiss. Forsch. 9, 1–210.

Junker, H., Dose, T., 2001. Charakterisierung, Modellierung und Simulation derLagerstatte Mittelplate. Fruhjahrstagung des DGMK Fachbereichs ‘‘Aufsuchungund Gewinnung’’. pp. 399–413.

Kaltwang, J., 1992. Die pleistozane Vereisungsgrenze im sudlichen Niedersachsenund im ostlichen Westfalen, vol. 33. Mitt. Geol. Inst. Univ., Hannover. pp. 1–161.

Klostermann, J., 1992. Das Quartar der Niederrheinischen Bucht – Ablagerungen derletzten Eiszeit am Niederrhein. Geol. Landesamt Nordrhein-Westfalen, Krefeld.pp. 200.

Klostermann, J., 1995. IV. Nordrhein-Westfalen. In: Benda, L. (Ed.), Das QuartarDeutschlands. Gebr. Borntrager, Berlin, pp. 59–94.

Langhans, G., Beuthan, H., Wessel, F., 2003. Erschließung und Entwicklung derLagerstatte Mittelplate – vom Pilot zum komplexen onshore/offshore Betrieb.DGMK-Fruhjahrstagung 2003 Fachbereich Aufsuchung und Gewinnung, Celle.

Lerche, I., Yu, Z., Torudbakken, B., Thomsen, R.O., 1997. Ice loading effects in sedi-mentary basins with reference to the Barents Sea. Marine and PetroleumGeology 14 (3), 277–338.

Martini, A.M., Walter, L.M., Budai, J.M., Ku, T.C.W., Kaiser, C.J., Schoell, M., 1998.Genetic and temporal relations between formation waters and biogenicmethane: upper Devonian Antrim shale, Michigan Basin, USA. Geochimica etCosmochimica Acta 62 (10), 1699–1720.

Muller, E., Brinzer, C., Szabados, A., 2004. Geochemische Zusammensetzung derErdole des Feldes Mittelplate und deren Bedeutung fur die Interpretation derOlqualitaten und Lagerstattenbildung. Fruhjahrstagung des DGMK Fachbereichs‘‘Aufsuchung und Gewinnung’’. pp. 355–362.

Neunzert, G.H., Littke, R., Idiz, E.F., 1995. An integrated approach to model gener-ation, migration, and accumulation of methane and molecular nitrogen in theNorthwest German Basin. Tulsa, OK, United States. In: Anonymous (Ed.), AAPGinternational conference and exhibition; abstracts: AAPG Bulletin. AmericanAssociation of Petroleum Geologists, vol. 79, p. 1240.

Neunzert, G.H., Littke, R., Scholten, S., 1997. Dynamic reconstruction of salt domeevolution in Northwest Germany: effects on temperature and hydrocarbongeneration. In: Anonymous (Ed.), 59th EAGE conference and technical exhibi-tion; extended abstracts. Oral and poster presentations; Petroleum Division:Conference and Technical Exhibition – European Association of Geoscientistsand Engineers, vol. 59, vol. 2. European Association of Geoscientists and Engi-neers (EAGE), International, p. D038.

Nye, J.F., 1952. A comparison between the theoretical and the measured long profileof the Unteraar glacier. Journal of Glaciology 2 (12), 103–107.

Nye, J.F., 1965. The flow of a glacier in a channel of rectangular, elliptic or paraboliccross-section. Journal of Glaciology 5 (41), 661–690.

Pasternak, M., Brinkmann, S., Messner, J., Sedlacek, R., 2006. Erdol und Erdgas in derBundesrepublik Deutschland 2005. State Authority of Mining, Energy andGeology (LBEG), Hannover. pp. 51.

Petmecky, S., Meier, L., Reiser, H., Littke, R., 1999. High thermal maturity in theLower Saxony Basin; intrusion or deep burial? Tectonophysics 304, 317–344.

Piotrowski, J.A., 1991. Quartar- und hydrogeologische Untersuchungen im Bereichder Bornhoveder Seenkette, Schleswig-Holstein Kiel.

Piotrowski, J.A., Tulaczyk, S., 1999. Subglacial conditions under the last ice sheet innorthwest Germany: ice-bed separation and enhanced basal sliding? Quater-nary Science Reviews 18, 737–751.

Riis, F., 1992. Dating and measuring of erosion, uplift and subsidence in Norway andthe Norwegian shelf in glacial periods. Norsk Geologisk Tidsskrift 72, 325–331.

Sales, J.K., 1992. Uplift and subsidence of northwestern Europe: possible causes andinfluence on hydrocarbon productivity. Norsk Geologisk Tidsskrift 72, 253–258.

Schwarz, K., 1996. Die Bohrungen 89/3, 89/4 und 89/9 auf dem deutschen Nord-seeschelf – Sedimentologische und magnetostratigraphische Befunde sowielithostratigraphische Konnektierung. Geologisches Jahrbuch A 146, 33–137.

Skupin, K., Speetzen, E., Zandstra, J.G., 1993. Die Eiszeit in Nordwestdeutschland.Geol. Landesamt Nordrhein-Westfalen, Krefeld. pp. 143.

Solheim, A., Riis, F., Elverhoi, A., Faleide, J.I., Jensen, L.N., Cloetingh, S., 1996. Impactof glaciations on basin evolution; data and models from the Norwegian marginand adjacent areas – Introduction and summary. Global Planet. Elsevier,Change, 12, Amsterdam, Netherlands. pp. 1–9.

Stephan, H.-J., 1995. I. Schleswig-Holstein. In: Benda, L. (Ed.), Das QuartarDeutschlands. Gebr. Borntrager, Berlin, pp. 1–13.

Stephan, H.-J., Menke, B., 1993. Das Pleistozan in Schleswig-Holstein. Berichteaus dem Geologischen Landesamt Schleswig-Holstein 3 (Sonderausgabe),pp. 19–62.

Streif, H., Caspers, G., Freund, H., Geyh, M.A., Kleinmann, A., Merkt, J., Meyer, K.-D.,Muller, H., Rohde, P., Schwarz, C., 2004. Das Quartar in Niedersachsen undbenachbarten Gebieten – Gliederung, geologische Prozesse, Ablagerungen undLandschaftsformen. State Authority of Mining, Energy and Geology (LBEG),Hannover.

Sweeney, J., Burnham, A.K., 1990. Evaluation of a simple model of vitrinite reflec-tance based on chemical kinetics. AAPG Bulletin 74 (10), 1559–1570.

Thoma, M., Wolf, D., 1999. Bestimmung der Mantelviskositat aus Beobachtungender Landhebung und Schwere in Fennoskandien. GeoforschungsZentrumPotsdam, Potsdam. pp. 101.

Thorson, R.M., 2000. Glacial tectonics: a deeper perspective. Quaternary ScienceReviews 19, 1391–1398.

Tsytovich, N.A., 1957. The Mechanics of Frozen Ground. McGraw-HillBook Comp.pp. 426.

Unger, K.P., Kahlke, R.-D., 1995. IX. Thuringen. In: Benda, L. (Ed.), Das QuartarDeutschlands. Gebr. Borntrager, Berlin, pp. 199–219.

Vandenberghe, J., 2001. Permafrost during the Pleistocene in northwest and centralEurope. In: Paepe, R., Melnikov, V. (Eds.), Permafrost Response on EconomicDevelopment, Environmental Security and Natural Resources. Kluver AcademicPublishers, pp. 185–194.

Wagenbreth, O., 1978. Die Feuersteinlinie in der DDR, ihre Geschichte und Popu-larisierung. Schriftenreihe fur Geologische Wissenschaften 9, 339–368.

Washburn, A.L., 1979. Geocryology – A Survey of Periglacial Processes and Envi-ronments. Edward Arnold Publ. pp. 406.

Wehner, H., Gerling, P., Hiltmann, W., Kockel, F., 1989. Erdol-Charakteristik undOl-Muttergestein-Korrelation im Niedersachsischen Becken. Nachrichten derDeutschen Geologischen Gesellschaft 41, 77–78.

Welte, D.H., 1979. Organisch-geochemische Untersuchungen zur Bildung von Erdol-Kohlenwasserstoffen an Gesteinen des mittleren Oberrhein-Grabens. FortschrGeol Rheinl Westfalen 27, 51–74.