Full text of DOI:10.1046/j.1468-8123.2001.11004.x

Geo¯uids ꢀ2001) 1, 11±36
Interpretation of observed ¯uid potential patterns in a deep
sedimentary basin under tectonic compression: Hungarian
Great Plain, Pannonian Basin
Â
Â
J. TOTH AND I. ALMASI
Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Alberta, Canada
ABSTRACT
The ꢀ 40 000 km2 Hungarian Great Plain portion of the Pannonian Basin consists of a basin ®ll of 100 m to more than
7000 m thick semi- to unconsolidated marine,deltaic,lacustrine and ¯uviatile clastic sediments of Neogene age,rest-
ing on a strongly tectonized Pre±Neogene basement of horst±and±graben topography of a relief in excess of 5000 m.
The basement is built of a great variety of brittle rocks,including ¯ysch,carbonates and metamorphics. The relatively
continuous Endrod Aquitard,with a permeability of less than 1 md /10 À15 m2) and a depth varying between 500 and
Í
5000 m,divides the basin's rock framework into upper and lower sequences of highly permeable rock units,whose
permeabilities range from a few tens to several thousands of millidarcy. Subsurface ¯uid potential and ¯ow ®elds were
inferred from 16 192 water level and pore pressure measurements using three methods of representation: pressure±
elevation pro®les; hydraulic head maps; and hydraulic cross-sections.
Pressure±elevation pro®les were constructed for eight areas. Typically,they start from the surface with a straight-
line segment of a hydrostatic gradient /gst = 9.8067 MPa kmÀ1) and extend to depths of 1400±2500 m. At high sur-
face elevations,the gradient is slightly smaller than hydrostatic,while at low elevations it is slightly greater. At greater
depths,both the pressures and their vertical gradients are uniformly superhydrostatic. The transition to the overpres-
sured depths may be gradual,with a gradient of gdÀyn1 = 10±15 MPa kmÀ1 over a vertical distance of 400±1000 m,or
abrupt,with a pressure jump of up to 10 MPa km
over less than 100 m and a gradient of gdyn > 20 MPa kmÀ1
.
According to the hydraulic head maps for 13 100±500 m thick horizontal slices of the rock framework,the ¯uid
potential in the near-surface domains declines with depth beneath positive topographic features,but it increases
beneath depressions. The approximate boundary between these hydraulically contrasting regions is the 100 m eleva-
Â
tion contour line in the Duna±Tisza inter¯uve,and the 100±110 m contours in the NyõÂrseg uplands. Below depths of
ꢀ 600 m,islets of superhydrostatic heads develop which grow in number,areal extent and height as the depth
increases; hydraulic heads may exceed 3000 m locally. A hydraulic head 'escarpment' appears gradually in the eleva-
tion range of À 1000 to À 2800 m along an arcuate line which tracks a major regional fault zone striking NE±SW:
heads drop stepwise by several hundred metres,at places 2000 m,from its north and west sides to the south and east.
The escarpment forms a `¯uid potential bank' between a `¯uid potential highland' /500±2500 m) to the north and
west,and a `¯uid potential basin' /100±500 m) to the south and east. A `potential island' rises 1000 m high above this
basin further south.
According to four vertical hydraulic sections,groundwater ¯ow is controlled by the topography in the upper
200±1700 m of the basin; the driving force is orientated downwards beneath the highlands and upwards beneath
the lowlands. However,it is directed uniformly upwards at greater depths. The transition between the two regimes
may be gradual or abrupt,as indicated by wide or dense spacing of the hydraulic head contours,respectively. Pressure
`plumes' or `ridges' may protrude to shallow depths along faults originating in the basement. The basement horsts
appear to be overpressured relative to the intervening grabens.
The principal thesis of this paper is that the two main driving forces of ¯uid ¯ow in the basin are gravitation,due to
elevation differences of the topographic relief,and tectonic compression. The ¯ow ®eld is uncon®ned in the gravita-
tional regime,whereas it is con®ned in the compressional regime. The nature and geometry of the ¯uid potential ®eld
between the two regimes are controlled by the sedimentary and structural features of the rock units in that domain,
characterized by highly permeable and localized sedimentary windows,conductive faults and fracture zones. The
transition between the two potential ®elds can be gradual or abrupt in the vertical,and island-like or ridge-like in plan
view. The depth of the boundary zone can vary between 400 and 2000 m. Recharge to the gravitational regime is
# 2001 Blackwell Science Ltd

Â
Â
12 J. TOTH & I. ALMASI
inferred to occur from in®ltrating precipitation water,whereas that to the con®ned regime is from pore volume reduc-
tion due to the basement's tectonic compression.
Key-words: gravity ¯ow systems,Hungarian Great Plain,Pannonian Basin,subsurface hydraulics,tectonic
compression
Received 20 April 2000; accepted 16 October 2000
Â
Â
Corresponding author: Jozsef Toth,Department of Earth and Atmospheric Sciences,University of Alberta,Edmonton,
Alta,T6G 2E3,Canada.
Geo¯uids ꢀ2001) 1, 11-36
Particular effort was made therefore to compile a dataset from
INTRODUCTION
the existing measurements of formation pressures and water
levels, which is as complete and as rigorously screened as pos-
sible.
The Pannonian Basin in general, and the Great Plain of
Hungary in particular, have long been the subjects of intensive
geoscience research, such as structuralgeool gy, geophysics,
petroleum geology and hydrogeology ꢀBallentine et al.
All geodetic information is expressed in the Hungarian
Â
Â
`EOV' coordinate system ꢀEOV = Egyseges Orszagos VetuÈ-
leti rendszer = Uniform National Projection system). The
system uses the horizontal YEOV- and XEOV-plane coordinate
axes to express, respectively, the Easting and Northing dis-
tances, in metres, from the system's origin. The origin is
located outside the country to the south±west so as to make
all possible EOV coordinates positive in Hungary. Elevations
are measured on the vertical ZEOV-axis and referenced to Bal-
tic sea level, with the sense being positive upwards.
Â
1991; Clayton et al. 1990; Dobos 1985; Erdelyi 1976; Hun-
Â
garian Geological Survey 1993; Royden & Horvath 1988).
As a result of these interests, large databases related to petro-
leum exploration and groundwater resources are available.
Remarkably, however, these separate databases have never
been merged or used in combination to produce a compre-
hensive hydrogeological evaluation of the basin, although
numerous studies have been conducted on a great variety
of speci®c hydrogeologically related problems.
The description of the rock framework is based on strati-
graphic, lithologic and rock physical data from 1274 bore
holes, many of which were drilled hundreds of metres into
the Pre-Neogene basement; numerous in-house reports
and published papers; and published data and maps on struc-
ture and tectonic history. The resolution of the obtained
image is adequate for the regionalportrayalof the rock frame-
work, but is insuf®cient for local analyses.
The purpose of this paper is to present our understanding
of the basic hydrogeological conditions and their causes in
the Hungarian Great Plain portion of the Pannonian Basin.
Thus the centralobjective of this study is to describe and
interpret the ¯uid potentialdistribution of the basin's forma-
tion waters, making use of the relevant components of both
databases. Such characterization allows inferences to be made
regarding the origin of ¯uid driving forces, regionalpatterns
of subsurface water ¯ow, relative amounts of water derived
from possibly different sources and the principal controls
on the ¯ow geometry, including structural or sedimentologi-
caldiscontinuities.
A grand totalof nearyl 53 000 pore pressure and water
level measurements were collected. Pore pressure data were
obtained from petroleum exploration activities in the form
of either original `Drill±Stem±Test' ꢀDST) pressure build-
up charts or already calculated initial formation pressures.
These data cover mainly the deeper zones of the basin, i.e.
below approximately 600±800 m. Nonpumping and produc-
tion water levels are largely derived from water supply develop-
ments. The sources of these data include bored and hand-dug
water table wells a few metres deep, drilled municipal and
industrial water wells reaching to depths of 1000±1500 m,
and still deeper thermal wells used for health spas and heating
of green houses and apartment buildings. Many of the
obtained records lacked essential data or were otherwise de®-
cient, such as: incomplete sets of position coordinates; miss-
ing drilling ¯uid densities; no timing, duration and/or rate of
¯uid production; and so on. Ultimately, 1294 of 3706 ꢀor
34.9%) DST pressure measurements and 14 898 of 49 265
ꢀor 30.2%) water level measurements have been retained
DATABASE
The major types of basic data used in the study are geodetical,
geological, rock physical and ¯uid dynamic. Part of this infor-
mation was obtained in the form of originalmeasurements or
processed data from the databanks of various Hungarian
industrial®rms, government departments and research orga-
nizations, while other information was obtained as in-house
reports, conference proceedings, published papers and
monographs. Owing to the large volume, great diversity, dif-
ferent forms and variable quality of the information, it was
necessary to cull, consolidate and normalize the data, and
to establish a database speci®cally for the present purposes.
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 13
and incorporated into the study's database to yield a total of
16 192 hydraulic head values out of a total of 52 971 records.
where topographic elevations vary between 90 and 95 m.
These two principal depressions are bordered by areally
extensive, but only moderately elevated, highlands. Most
notable of these hydrogeologically important morphologic
features are the WSW±ENE-striking hills along the base
of the North CentralRange, with elvations reaching
400±600 m; the broad ridge along the N±S±orientated axis
of the Duna±Tisza inter¯uve, with elevations exceeding
THE STUDY AREA
With its approximately 40 000 km2, the study area covers
most of the Hungarian Great Plain region of the Pannonian
Basin in CentralEurope ꢀFig. 1). The basin is a major,
near-circular intermontane depression of about 500 km in
diameter. It is almost completely surrounded by a belt of
high mountains comprising the Alps, the Carpathians and
the Dinarides. The highest peaks of all these ranges rise
above 2500 m, while their average elevations are ꢀ 2000 m,
ꢀ 1500 m and ꢀ 1000 m, respectively.
Í
200 m in the GoÈdoÈllo Hills at the inter¯uve's north end,
and 140 m at Illancs, in the south; and the broad sandy hills
 Â
of the Nyõrseg, in the north±east, rising over 160 m. Within
distances of 60±100 km to the east and south of the area's
eastern and southern boundaries, the Apuseni mountains
and the Dinarides complete the circle of high ground around
the entire area ꢀFig. 1). Breaking the continuity of the high
topography are the strips of lowlands where the major rivers
enter and exit from the area.
The Great Plain is bounded by the Duna River to the west;
the foothills of the Hungarian North Central Range, which is
the centralfront range of the Carpathians, to the north; and
by the internationalboundaries with Sol vakia, the Ukraine,
Romania and Serbia, respectively, to the north, north±east,
south±east and south ꢀFig. 2). The maps, cross±sections
and observation points of the study are referenced to a rectan-
gular base map prepared for the present purpose with the fol-
lowing corner points ꢀand side lengths) in the `EOV'
coordinate system: XEOV = from 60 km to 360 km ꢀ300 km)
and YEOV = from 620 km to 940 km ꢀ320 km).
Geologically, the Pannonian Basin is a complex sedimen-
tary basin which comprises a number of deep local basins,
such as the Vienna Basin, the Little Plain of Hungary and sev-
eraldeep troughs beneath the Great Plain. In the study area, a
100 m to more than 7000 m thick semi- to unconsolidated
clastic basin ®ll of marine, deltaic, lacustrine, ¯uviatile and
eolian strata of Neogene age rest on a Pre±Neogene base-
ment of brittle ¯ysch, carbonate and metamorphic rocks of
a high-relief horst-and-graben topography.
The area is well drained, primarily by the Tisza River and its
tributaries ꢀFig. 2). The Tisza enters the area in the north±
east and leaves across the centre of its south boundary. The
main tributaries, the Zagyva, the KoÈroÈs River sytem and the
Maros, ¯ow over the lowest part of the study area, namely the
extensive and ¯at centraldepression of the Great Palin. The
topographic elevation of this plane surface varies between
approximately 80 and 90 m above sea level ꢀa.s.l.). The other,
much smaller, ¯at and low±lying region is a 5±25 km wide
strip of land stretching along the left bank of the Duna River,
Neogene sedimentation started with generalrifting and
rapid subsidence in Early Miocene times ꢀ19 Ma) with the
deposition of clays, conglomerates and, locally, coals of the
Karpatian Formation ꢀFig. 3). Extension and rifting contin-
ued throughout the Middle Miocene ꢀ16.5±11.5 Ma), result-
ing in thick deposits of marine sediments, such as limestones,
marls and turbidites, forming the Badenian and Sarmatian
Formations. At the end of the rifting period, the morphology
of the basin's surface was one of high relief due to variable
amounts of extension, uplift and subsidence in different
regions. Islands of older rocks alternated with deep troughs
®lled with thousands of metres of syn-rift deposits. High-
and low-angle normal faulting, tectonic unroo®ng and trans-
fer faults gave rise to complex systems of rock framework dis-
continuities and a basin±and±range morphology by the end
Â
Â
of the Middle Miocene ꢀBerczi et al. 1988; Juhasz 1991).
In the Hungarian literature, the Late Miocene and Plio-
cene epochs of the Pannonian Basin are commonly referred
to as the Lower Pannonian ꢀ11.5±5.5 Ma) and Upper Panno-
nian ꢀ5.5±1.2 Ma), respectively. At the beginning of the
Lower Pannonian, marine transgression inundated the dee-
pest regions of the high-relief surface, with the elevated
islands coming under water later, resulting in strong lateral
facies variations. Coastalsandy congolmerates and coarse-
grained sandstones, deposited around the islands, may
change laterally into siltstones, calcareous marl, limestones
and turbidites within distances of a few kilometres, as in the
Fig. 1. Location of the Pannonian Basin and the Great Plain of Hungary in the
Â
Â
Í
Totkomlos, NagykoÈru and Szolnok Formations, listed in
regional context.
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
14 J. TOTH & I. ALMASI
Fig. 2. Topographic map of the study area showing the areas of pressure±elevation pro®les,and the traces of: seismic section Ka-27,Fig. 7; hydraulic sectio ns H1,
H2, H3, H4; and hydrostratigraphic section,Fig. 8.
ascending order in Fig. 3, and following the nomenclature
introduced by Gajdos et al. ꢀ1983). The transition from the
Lower Pannonian to the Upper Pannonian is marked by
regionally widespread deposition of delta slope and near-
pebble beds, of the Upper Pannonian ToÈrteland Zagyva For-
mations. Continued rapid subsidence during the Quaternary
ꢀ1.2±0 Ma) resulted in the deposition of up to a 1000 m thick
sequence of ®ne- to very coarse-grained, unconsolidated or
poorly consolidated clays, claystones, sands, thin-bedded
and lenticular gravels of alluvial, ¯uviatile and eolian origin.
The surface geology of the study area is thus characterized
by extensive regions of highly permeable loess, sands and
gravels.
Í
shore sediments of the Algyo Formation, consisting mainly
of mudstones, siltstones and argillaceous marl strata, inter-
Í
bedded with sandstone lenses. The Algyo Formation is, in
turn, overlain by the coarsening-upward distributory mouth
bar sequence of sandy delta front and delta plain sediments,
and the alluvial sandstones, siltstones, clays and marls, some
compacted sandstones with calcareous cement and quartz
Tectonic features, such as thrust faults, transform faults
and ¯ower structures, have been interpreted to indicate that
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 15
Fig. 4. Intraplate stresses in the Adria-Alpine-Carpathian-Dinarides region,
inferred from structural geology and geophysics /Bada et al. 1998,Fig. 12,
reproduced with permission).
Grenerczy et al. 2001, Fig. 7; Gy. Grenerczy, personal
communication).
A dense network of faults and fractures, created or rejuve-
nated by intensive Neogene tectonics, has destroyed the
regional lithologic integrity of the basement and the basin ®ll,
both in lateral and vertical directions. Depending on the site-
speci®c geological history following tectonic events of fault-
ing and fracturing, the lithologic discontinuities may become
highly conductive avenues ꢀClayton et al. 1990) or low±per-
meability barriers to pore pressure propagation and ¯uid
¯ow. The regionalcharacteristics of the patterns of Neogene
age faults can be seen in Fig. 6, as adapted from Rumpler &
Fig. 3. Lithology,stratigraphy and hydrostratigraphic units in the Great Plain
of Hungary.
Â
Horvath ꢀ1988; Fig. 11). The dominant orientation of the
the former stress regime of crustalextension started to
change to a compressionalregime at the beginning of the
Late Miocene ꢀ10.0 Ma). The compressionalregime is
thought to have been active to date with varying intensity.
Its latest major resurgence has been estimated to have started
in the Late Pliocene, ꢀ 2.5 Ma ꢀCsontos 1995; Gerner et al.
faults is NE±SW, but several faults strike also NNE±SSW or
even WNW±ESE. There are numerous deep sedimentary
troughs in the area, usually bounded by normal faults. Many
of these normalfautls, and thus their associated grabens, are
dissected by strike-slip faults, possibly effecting signi®cant
lateral displacements of the disconnected and juxtaposed
graben sections. Curving, splaying and side-stepping of
strike-slip fault segments are common.
Â
1997; Horvath & Cloetingh 1996; Peresson & Decker
1997). Based on an extensive compilation and interpretation
of earlier studies and structural geological and geophysical
data, Bada et al. ꢀ1998; Fig. 12) have envisaged the current
lateral crustal stresses in regions around the Pannonian Basin
as shown in Fig. 4. According to this interpretation, the Pan-
nonian Basin in general, and the Great Plain of Hungary in
particular, are at present subject to a compressive stress
regime of focally converging sections of the surrounding
crustalsegments. This interpretation is con®rmed quantita-
tively by satellite geodetic measurements of horizontal velo-
cities of intraplate crustal movements at various stations in
the north and north±west parts of the study area proper
and to the north and west in the entire Adria±Alpine±Car-
pathian region ꢀFig. 5; Grenerczy et al. 2000, Fig. 5;
A characteristic and important structuralstyel is shown in
Â
Fig. 7, adapted from Rumpler & Horvath ꢀ1988; Fig. 9). The
location of the seismic section is shown in Figs 2 and 6. The
Â
section Ka±27 ꢀFig. 7; Rumpler & Horvath 1988, Fig. 9)
shows normalfautls, some penetrating the basement, their
listric nature and lateral spacing, ¯ower structures, folding
and vertical displacements of strata. Many of the major faults
dissect the entire thickness of the rock framework from the
basement to the Upper Pannonian, and some even reach into
the Quaternary. Owing to the density, geometry and fre-
quent splaying of the faults, different bore holes located just
short distances apart can penetrate different faults at similar
elevations and with connections to different strata. This
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
16 J. TOTH & I. ALMASI
Fig. 5. Intraplate crustal velocities in the Adria±
Alpine±Carpathian-Dinarides region,based on
Global Positioning System satellite geodetic sur-
veys /modi®ed from Grenerczy et al. 2001;
courtesy of Gy. Grenerczy,personal communica-
tion,and with the permission of the European
Geophysical Society).
observation will be used in the interpretation of vertical pore
pressure pro®les in later sections.
till regionally representative permeability values from the
wealth of information organized and processed for other pur-
poses, one feature stands out clearly: matrix permeabilities are
generally high in the entire basin, including those of the
marlÀ and shale±dominated aquitards. The following exam-
ples of independent permeability estimates serve to illustrate
the point.
HYDROSTRATIGRAPHY
Hydrostratigraphic classi®cation was developed from original
petroleum industry records of measured core permeabilities
and DST reports, and from information found in in-house
reports and published papers. Whereas it was dif®cult to dis-
Palaeozoic metamorphics and Lower and Middle Triassic
Â
Â
carbonates of the basement: 1±10 md ꢀBerczi & Kokai
1976). Pliocene±Pleistocene: aquitards, 10±102 md; aquifers,
3
5
Â
10 ±10 md ꢀHalasz 1975). Lower Pannonian: lower, ꢀ3±5)
 10 md; middle, 5  10±4  103 md; upper, ꢀ2±5)  102
md. Upper Pannonian: lower and middle, 5 Â 102±103 md;
upper, > 103 md ꢀKorim 1966). Pliocene aquifer system,
5 Â 102±5 Â 103 md; Quaternary aquifer system, 103±104
md ꢀAlfoÈldi et al. 1978). At depths of: 1600±1800 m,
4 Â 102 md; 1800±2000 m, 2.7 Â 102 md; below 2000 m,
2.2 Â 102 md ꢀKorim & Liebe 1973). Mesozoic, 10À1 md;
Mesozoic±Miocene, 10À1À7.5 Â 10 md; Eocene±Oligocene,
10À1À102 md; Miocene, 8.5 Â 10±2.3 Â 102 md; Sarmatian
ꢀ2±6) Â 10±102 md; marls <10À1 md; Lower Pannonian,
10À1À4 Â 102 md ꢀPap 1976). Aquitards, 10À2À10 md;
aquifers, 103À105 md ꢀSzebenyi 1955). Only two docu-
Â
ments were found that gave quantitative data on anisotropy
of the matrix permeability. Korim ꢀ1966; p. 527) reports that
the vertical permeability is `30-50%' less than the horizontal
one in Pannonian sands and sandstones. In an anonymous
and undated in-house oilcompany report, anyasling
`thousands of core permeability measurements', marls show
a permeability anisotropy of slightly greater than one order
of magnitude in the depth range of ꢀ 1±2 km. Their perme-
ability is shown to be uniformly less than 10À1 md below
ꢀ 2.5 km. The degree of anisotropy reported for sandstones
does not exceed one±half of an order of magnitude anywhere
between 1 and 5 km depth. Sandstone permeability is
Fig. 6. Regional patterns of Neogene age faults in the Great Plain of
Hungary,and trace of seismic section Ka-27,Fig. 7 /modi®ed from Rumpler &
Â
Horvath 1988,®g. 11).
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 17
Â
Fig. 7. Line drawing interpretation of seismic section Ka-27,illustrating a structural style in the Great Plain of Hungary /Rumpler & Horva th 1988,®g. 9,
reproduced with the permission of the American Association of Petroleum Geologists).
shown to decrease to the range of ꢀ1±2) Â 10À1 md at a
depth of ꢀ 5 km.
runs within the body of the Pre-Pannonian rocks, but its
depth and exact shape cannot be established from available
data. Inferred from generalconsiderations concerning the
effects of subaerialerosion and tectonic events on the deve-l
opment of permeability patterns in mountainous regions, it is
assumed that the near-zero permeability surface of the Pre-
Neogene Aquiclude is a reduced amplitude replica of the
topography of the Pre-Pannonian rocks.
Based on considerations of chronostratigraphic divisions,
lithologic facies types and reported values of permeabilities
as exempli®ed above, the subsurface has been subdivided into
six regional hydrostratigraphic units, as follows ꢀFigs 2, 3 and
8).
ꢀ1) Pre-Neogene Aquiclude: the lowermost hydrostrati-
graphic unit. It is considered to form the basalboundary of
negligible permeability of the ¯ow domain. Its top surface
ꢀ2) Pre-Pannonian Aquifer: a major aquifer at the base of
the ¯ow domain, made up of the Middle Miocene Badenian
Â
Fig. 8. Schematic NW-SE hydrostratigraphic section across the Great Plain of Hungary /based on Juhasz 1991,®g. 7,with the permission of the American
Association of Petroleum Geologists).
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
18 J. TOTH & I. ALMASI
and Sarmatian Formations. Its high permeability is due to
tectonic fracturing and faulting, the abundance of coarse-
grained arkosic rocks and the remnant effects of subaerial
weathering. The hydraulic interpretation of the lithologic pro-
perties is corroborated by DST and core analyses, conducted
in the course of intensive exploration for petroleum. The
aquifer contains numerous high-yielding oil and gas reservoirs.
mers ꢀmaximum, up to 408C). The signi®cant aspect of the
climate in the present context is the fact that groundwater
recharge is suf®cient to maintain the water table near the land
surface, generally within 2±5 m in the highlands and 0±2 m in
the lowlands. Consequently, the con®guration of the water
table remains a close replica of the topographic relief at all
times, and it can thus be considered as the upper boundary
of the basin's ¯ow domain.
Í
ꢀ3) Endrod Aquitard: a regionally extensive but discontin-
uous unit primarily of low-permeability calcareous and argil-
laceous marls. At several places where it is absent, the
subjacent ꢀPre-Pannonian) and overlying ꢀSzolnok) aquifers
are in contact, thus in direct cross-formationalhydrauicl
communication.
BASINAL FLUID POTENTIAL: PATTERNS AND
INTERPRETATION
In basin±scale characterization of subsurface ¯ow ®elds, the
most critical term to evaluate correctly is the hydraulic head,
h. Its perceived regionalpattern serves as the basis for the
interpretation of the distribution of the ¯uid potentialand
its gradient, the ¯uid driving force ꢀHubbert 1940, 1953).
Flow is always in the direction of decreasing hydraulic head
because h is directly proportional to the ¯uid's potential
energy F:
ꢀ4) Szolnok Aquifer: a cyclic alternation of consolidated
sandstones, siltstones and clay±marl beds of a prodelta turbi-
dite facies. It is regionally discontinuous and occurs only in
the deep subbasins, above deep basin marls. Where it pinches
Í
out against the ¯anks of the basement highs, the Endrod
Aquitard is overlain directly by the next higher unit, the
Í
Algyo Aquitard.
Í
ꢀ5) Algyo Aquitard: an areally extensive unit of poorly to
h ˆ È=g
ꢀ1)
well-consolidated siltstone and clay±marl, containing rela-
tively highly permeable sand lenses of reworked turbidites,
rhythmic mouth bar sequences or submarine channel ®lls.
Above basement highs, the lithology is sand-dominated, giv-
ing aquifer properties to the regional aquitard locally. Owing
to its lenticularity and other sedimentological discontinuities,
where F is the mechanicalenergy per unit mass of ¯uid. If we
assume constant ¯uid density along the ¯ow path and neglect
inertialforces, then
È ˆ gz ‡ p=ꢀ
and
ꢀ2a)
Í
as well as to cross-cutting by faults and fractures, the Algyo
Aquitard is leaky in nature.
h ˆ z ‡ p=ꢀg
ꢀ2b)
ꢀ6) Great Plain *NagyalfoÈld) Aquifer: a major unit, domi-
nated by unconsolidated coarse clastics ꢀsand, gravel, loess),
and containing relatively thin clay and silt beds of varying lat-
eralextent. It inculdes the Upper Pannonian To Èrteland
Zagyva Formations, and the sur®cialQuaternary sediments
which cover the entire basin to a thickness of 2400 m. Its
chief lithologic characteristic is a predominance and good
spatial connectivity of the highly permeable bodies of silts,
and coarse sands and gravels. Large amounts of groundwater
are pumped at high rates from the upper part of this aquifer,
while thermal water is extracted from its basal ToÈrtelForma-
tion for health spas and heating.
where z is the elevation of the measurement point with
respect to a datum plane, usually sea level, and p is pore pres-
sure; thus, all components of h are measurable.
In equation 2b, rg = g is the speci®c weight of the ¯uid,
which is numerically equal to the vertical pressure gradient
in a column of static ¯uid of density r. The variables used
in the present study were the hydraulic head, h, pore ¯uid
pressure, p, and the verticalpressure gradient, g. These
parameters were derived from measured values of well
water levels and pore pressures. In calculating h from pore
pressures, the pore water density was assumed to be
uniformly 1000 kg mÀ3, as the total dissolved solids content
of the vast majority of water samples is less than 2 g LÀ1, and
the maximum values remain below 30 g LÀ1. Accordingly,
the value assumed for the vertical hydrostatic, or `nominal',
On the basis of lithologic, stratigraphic and structural con-
siderations, the entire porous and permeable framework of
the study area can be considered as a regionaland cross±for-
mational hydraulic continuum, with the base at depths below
the top of the Pre-Pannonian aquifer and with the land sur-
face being the upper boundary. Our goalis to use this under-
standing of the geological framework as a guide for the
interpretation of the mapped ¯uid potentialpatterns. The
hydraulic properties of the various units can only be repre-
sented in a generalized, order±of±magnitude sense: perme-
ability contrasts are indicated in Fig. 3.
pressure gradient is gst = 9.8067 MPa kmÀ1
.
Pressure±elevation pro®les, p2z)
The pressure±elevation pro®le, pꢀz), is a plot of the pore ¯uid
pressures versus the topographic elevation of the measure-
ment points along a vertical beneath a given location in the
®eld ꢀFigs 9±13). In such a plot, the pressure difference over
a unit vertical distance is numerically equal to the vertical
The climate of the Pannonian Basin is temperate±conti-
nentalwith midl winters ꢀminimum, À 208C) and hot sum-
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 19
Â
Fig. 10. Pressure-elevation pro®le, p/z),Szentes-Csongra
d region /for
Fig. 9. Pressure-elevation pro®le, p/z),Kiskunhalas region /for location,see
location,see Fig. 2).
Fig. 2).
component, g, of the pressure gradient at the speci®ed
elevation and is therefore indicative of the vertical sense
of the ¯uid impelling force. In the case of fresh water
ꢀr ˆ 1000 kg mÀ3), a hydrostatic value of gst = 9.8067
MPa kmÀ1 indicates stagnation of water or horizontal¯ow.
Values that are less or greater than gst, i.e. gdyn < gst or gdyn
> gst, indicate a force component driving the water down-
wards or upwards, respectively. A sharp break in the pressure
pro®le re¯ects an abrupt change in the pressures or pressure
gradient beneath the given site. Generally, such changes
result from signi®cant vertical changes in the permeability
of the rock framework. Conversely, a pressure pro®le or its
segments that are devoid of breaks in direction imply rela-
tively homogeneous rocks over the depth intervals of the
straight portions.
density. In order, however, to reduce the effects of rock het-
erogeneities and topographic elevation differences, both the
size and relief of the areas from which the wells were used for
individual pꢀz) plots were kept to a minimum.
Kiskunhalas ꢀarea boundaries: XEOV = 110±130 km ꢀ20 km),
YEOV = 673±693 km ꢀ20 km); land surface elevation: 120 m
< z0 < 150 m a.s.l.; elevation of pressure measurement
domain: z = z0 > ± 3350 m)
The area is located on the central highland of the Duna±Tisza
inter¯uve and includes the highest elevation range of its
southern regions. The pꢀz) pro®le consists of two straight
segments which have distinctly different gradients and inter-
sect at the elevation of z ꢀ À 1300 m ꢀFig. 9). The gradient
in the top segment is substantially less than hydrostatic,
Eight pressure±elevation pro®les were constructed for the
present purpose ꢀFig. 2), ®ve of which are discussed here to
exemplify the different styles of the one-dimensional vertical
pressure distribution in the study area. Ideally, a pꢀz) pro®le
should be constructed from continuous and simultaneous
measurements taken under one surface point along one single
verticalin a homogeneous ¯uid environment. In realilfe,
such idealsituations do not exist and the requirements must
be relaxed. Accordingly, data from several wells were com-
bined into single pꢀz) pro®les so as to increase control point
gz > À 1300 = 9.5099 MPa kmÀ1 < gst = 9.8067 MPa kmÀ1
while that of the one below z = À 1700 m is superhydrostatic,
gz < À 1700 = 11.7425 MPa kmÀ1 > gst = 9.8067 MPa kmÀ1
,
,
indicating downward and upward driving forces, respectively.
Descending ¯ow beneath the highest regions of the Duna±
Tisza inter¯uve can be explained by gravity drive. The super-
hydrostatic gradient of the lower segment requires an energy
source below z = À 2500 m. The existence of such an energy
source is qualitatively supported by the two points of
still higher pressure ꢀꢀ 40 MPa) at z ꢀ À 2650 m in the
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
20 J. TOTH & I. ALMASI
Â
Â
Fig. 13. Pressure-elevation pro®le, p/z),De vavanya region /for location,see
Fig. 11. Pressure-elevation pro®le, p/z),Biharkeresztes region /for location,
Fig. 2).
see Fig. 2).
Pre-Neogene basement. However, two points are considered
insuf®cient to warrant the assignment of a gradient value.
A high degree of verticalhydrauilc communication across
various lithologic units can be inferred from the straightness
of the two segments of the pro®le. This feature has a parti-
cular signi®cance in the approximate elevation range of
À 850 m > z > À 1150 m, indicating ready cross-formational
¯ow through the ꢀ 300 m thick AlgyoÄ Aquitard.
Â
Szentes±Csongrad ꢀarea boundaries: XEOV = 140±160 km
ꢀ20 km), YEOV = 730±750 km ꢀ20 km); land surface
elevation: 80 m < z0 < 90 m a.s.l.; elevation of pressure
measurement domain: z = z0 > À 2900 m)
The sample block is located along the Tisza River in one of
the lowest regions of the study area. The pꢀz) pro®le is a
straight line with dense data control down to an elevation
of z ꢀ À 2400 m ꢀFig. 10). The verticalpressure gradient
is slightly superhydrostatic, gz > À 2400 = 9.9497 MPa kmÀ1
,
indicating an upward driving force across the entire depth
range. Owing to the low topographic position of the area,
the assumption of gravitationaldischarge, at elast in the
upper strata, is warranted. The de®nition of a boundary
between ¯ow driven upward by gravity and that driven by a
deep internalenergy source, however, is not possibel. Two
measurements at z ꢀ À 2800 m show a high average over-
pressure of p ꢀ 47 MPa. Similar to the Kiskunhalas case, it
Ä
Fig. 12. Pressure-elevation pro®le, p/z),Endro d region /for location,see
Fig. 2).
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 21
is dif®cult to assign a pressure gradient value based only on
two points. However, the existence of such high pressures
is corroborated by similar values at those depths in the area
ꢀe.g. at ꢀ 120 km in Section H2, Fig. 21) and point to other
driving forces in or below the Pannonian Aquifer.
co-occurence of severalhydrostratigraphic units of signi®-
cantly different age at similar elevation levels. Attention is
drawn also to the large number of pressure measurements
Í
conducted in the Algyo Aquitard over the depth range of
z ꢀ À 1250 to À 2300 m. The obvious interest of petroleum
explorationists in this `aquitard' clearly demonstrates its
potential to contain producible reservoirs, i.e. highly perme-
able rock bodies. The hydraulically conductive nature of
Í
Good hydraulic communication between the Algyo Aqui-
tard and the overlying Great Plain ꢀNagyalfoÈld) Aquifer
ꢀFig. 3) can be inferred from the straightness of the pro®le
at the level of the boundary between these two hydrostrati-
graphic units, i.e. at z ꢀ À 2200 m ꢀFig. 10). On the other
hand, a high hydraulic resistance at the lower boundary of the
Í
the Algyo Aquitard, at least in the Biharkeresztes block, is
also corroborated by the relative straightness of the middle
segment of the pꢀz) pro®le.
Í
Algyo Aquitard, at z ꢀ À 2675 m, against ¯ow forced
Í
upwards across the Szolnok Aquifer by the postulated deep
energy source may explain the strong overpressures below.
Endrod ꢀarea boundaries: XEOV = 163±190 km ꢀ27km), Y EOV
= 764±787 km ꢀ23 km); land surface elevation: 83 m < z0 <
92 m a.s.l.; elevation of pressure measurement domain: z =
z0 > À 4050 m)
Biharkeresztes ꢀarea boundaries: XEOV = 185±214 km
ꢀ29 km), YEOV = 830±860 km ꢀ30 km); land surface elevation:
89 m < z0 < 108 m a.s.l.; elevation of pressure measurement
domain: z ˆ z0 > À 3175 m)
Í
The Endrod block is located in the central and lowest regions
of the Great Plain. The basic con®guration of its pꢀz) pro®le
resembles that of Biharkeresztes: three straight±line segments
with monotonically increasing gradients downwards ꢀFig. 12).
The top segment's gradient is slightly superhydrostatic,
gz > À 2100 ˆ 9.8785 MPa kmÀ1, indicating uniformly ascend-
ing ¯ow through a regionally homogeneous rock body above
the elevation level of z ꢀ À 2100 m. The bottom segment,
between approximately À 2400 and À 3100 m elevations,
is strongly overpressured with an apparent vertical pressure
Three well-de®ned and one questionable segment can be dis-
tinguished on the Biharkeresztes pꢀz) pro®le ꢀFig. 11). The
top segment, above z > À 1400 m, with gz
ˆ
>
À
1400
10.1258 MPa kmÀ1 is suggestive of upward-directed, grav-
ity-driven ¯ow, possibly enhanced by high pressures dissipat-
ing from the segment below. The gradient of the lowest
segment, below z ꢀ À 2150 m, is signi®cantly greater than
hydrostatic, gz <
ˆ 23.9231 MPa kmÀ1, and is consis-
gradient of g
ˆ 57.7161 MPa kmÀ1, which
À 2150
À 2400 > z > À 3100
tent with a source of high pressure below the deepest point of
measurement at z ꢀ À 3175 m. The transitionalmiddel
segment, between z ꢀ À 1300 and À 2150 m ꢀa thickness
of ꢀ 850 m), appears to re¯ect a gradualolss of pressure
energy of water ascending cross-formationally over more
is more than twice lithostatic! The middle segment's gradient
is determined at g
ˆ 22.9995 MPa kmÀ1
,
À 2100 > z > À 2400
taken over a verticaldistance of ꢀ 300 m ꢀbetween À 2100
and À 2400 m), i.e. barely more than one-third of the 850 m
thickness of the Biharkeresztes middle segment.
Í
than 800 m of the low-permeability Algyo Aquitard, impelled
The force and therefore the ¯ow conditions indicated by
the top segment are consistent with those expected in an
ascending ¯ow domain due either to gravitationaldischarge
and/or a deep-seated energy source. However, if water in this
domain is derived, indeed, from two different origins, i.e.
meteoric driven by gravity and `depth water' driven by a deep
pressure source, the region and depth zone where they
merge, or the boundary between them, cannot be deter-
mined from an analysis of the pꢀz) pro®le; the boundary, if
there is one, could be at any level above z = À 2100 m.
The simple linear interpretation of the bottom segment's
pressure distribution, as shown in Fig. 12, is considered to
be illusory. Instead, a more realistic interpretation would
probably be a series of separate, and perhaps subparallel,
gradient lines of slightly greater values than that at the
middle section, but less than lithostatic, i.e. less than
ꢀ 25.5 MPa kmÀ1. This interpretation invokes the same
conceptualimage as used to expalin the wide band of
overpressures in the bottom segment of the Biharkeresztes
pro®le. However, multiple gradient lines cannot actually be
proposed because of a lack of adequate constraints on their
possible position and orientation.
by the high overpressures below. The four pressure values
along the normal hydrostatic gradient line at depths of
z ꢀ À 2100 to À 2600 m are attributed to production
effects and are not considered to be representative of natural
conditions.
A noteworthy feature of the pressure pro®le is a wide band
of observed pressures in the bottom segment; its width may
exceed 10 MPa at some given elevation levels. This feature
can be explained by assuming that pressures were measured
in points at approximately the same elevations, but in wells
at different locations that have penetrated several faults or
sedimentary discontinuities of different permeabilities, and
of different degrees of hydraulic communication with a
deep-seated source of high pressure. From another view-
point, the close similarity of pressure values over the depth
range of 400±600 m can also be understood by assuming that
several of the above-mentioned, possibly low-angle, faults are
situated above one another ꢀas, for example, shown in Fig. 7)
and penetrated by a few closely spaced bore holes. The prob-
ability of faulting, tectonic displacement and stratigraphic
and lithologic discontinuities is apparent from the frequent
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
22 J. TOTH & I. ALMASI
Â
study, including the three reported elsewhere ꢀToth &
A fundamentaldifference between the Biharkeresztes and
Í
Â
Endrod pro®les seems to be the signi®cantly lower elevation,
Almasi 1998).
i.e. greater depth, of the top of the overpressured zone at
ꢀ1) Two vertically superposed principal pore pressure
zones can be recognized in the area: an uncon®ned upper
one, and a con®ned and overpressured lower one; a transi-
tionalzone may connect the two zones in palces.
ꢀ2) Superhydrostatic verticalpressure gradients increase
monotonically with depth below the upper boundary of the
zone of overpressures.
Í
Endrod ꢀz ꢀ À 2100 m) than at Biharkeresztes ꢀz ꢀ À
1400 m). On the other hand, at the elevation level of z ꢀ
À 3000 m pressures are in the range of 55±60 MPa in both
cases. It is suggested, therefore, that the unrealistically high
Í
gradient of the bottom segment of the Endrod pro®le is,
indeed, only `apparent', and that it is due to a combination
of the presence of a relatively thin ꢀꢀ 300 m) but low-perme-
ability aquitard situated at a depth below z ˆ À 2100 m,
effectively con®ning a number of distinct lithologic disconti-
nuities of different hydraulic conductivities, which are
connected to a deep-seated pressure source. In other words,
the low position of an effective local seal, combined with
high overpressures at a given depth, results in reduced thick-
ness of the zone of transition from overpressures to normal
pressures.
ꢀ3) The depth to the upper boundary of the zone of over-
pressures is not controlled by hydrostratigraphy or depth,
e.g.
Kiskunhalas: at À 1600 m Szolnok Aquifer, base ꢀFig. 9).
Í
Biharkeresztes: at À 1300 m Algyo Aquitard, top ꢀFig. 11).
Â
Â
Í
Devavanya: at À 2100 m Algyo Aquitard, base ꢀFig. 13).
ꢀ4) Water ¯ow is driven by gravity in the upper, uncon®ned
zone, and by a high-pressure energy source seated below the
top of the Pre-Pannonian Aquifer in the lower, con®ned
zone.
Â
Â
Devavanya ꢀarea boundaries: XEOV = 182±195 km ꢀ13 km),
ꢀ5) The regionally uncon®ned nature of the ¯ow in the
upper zone is indicated by the consistently and strictly
straight-line character of the pꢀz) pro®les above the top of
overpressures.
Y
EOV = 792±808 km ꢀ16 km); land surface elevation: 84 m <
z0 < 92 m a.s.l.; elevation of pressure measurement domain:
z = z0 > À 3000 m)
Â
Â
The Devavanya pꢀz) pro®le consists of two segments only
ꢀ6) Gravity drive is indicated by: ꢀi) verticalpressure gradi-
ents less than hydrostatic, gactual < gstatic, beneath topographi-
cally high regions: a characteristic condition of descending
groundwater ¯ow, i.e. recharge ꢀe.g. Kiskunhalas, Fig. 9);
ꢀFig. 13). The top one is a well-constrained straight line
Í
reaching to the bottom of the Algyo Aquitard at z ꢀ
À 2150 m, at a slightly superhydrostatic gradient of
gz > À 2150 ˆ 9.9920 MPa kmÀ1. It is interpreted to indicate
conditions seen in the upper segments of the other pꢀz)
ꢀii) pressure gradients greater than hydrostatic, gactual > gstatic
,
beneath topographically low regions: characteristic of ascend-
ing groundwater ¯ow, i.e. discharge ꢀe.g. Biharkeresztes, Fig.
Â
Í
pro®les ꢀSzentes±Csongrad, Biharkeresztes, Endrod), i.e.
uniformly ascending meteoric water driven by gravity with
a possible contribution of `depth water' driven by a deep
pressure source.
Â
Â
11; Devavanya, Fig. 13).
ꢀ7) A deep-seated energy source is indicated by the exis-
tence and ubiquity of superhydrostatic verticalpressure gra-
dients below the top of the Pre-Pannonian Aquifer. This type
of pressure ®eld induces uniformly upward orientated ¯ow in
the domain.
The bottom section, again, is less well constrained than the
top one, but better than those in the previous cases. The
lower scatter of the data points is probably due to the lower
number of formations tested and the smaller block size.
The signi®cant feature in this example is the absence of a
transitional segment between the overpressured and normally
pressured zones. The phenomenon can be explained by the
presence of a local seal effectively capping the few avenues
available for ¯uid ¯ow and pressure dissipation. That some
diffusion is still taking place is indicated by the strong super-
hydrostatic gradient. If the seal at the top were ideally effec-
tive, the pressures would abruptly increase directly beneath it
untilthe gradient reached hydrostatic vaul e over the entire
depth range, while the pressures at the bottom remained
unchanged. The pressures at the bottom keep re¯ecting the
strength of their source at depth or, at least, the part of that
strength reaching up to their points of measurement.
ꢀ8) The broad band of overpressures at given elevation
levels is due to the presence of multiple pathways ꢀfracture
zones, faults, sediment windows) of differing hydraulic resis-
tance to ¯uids being driven upward.
Tomographic ¯uid potential maps, hij 2x,y)
A tomographic ¯uid potentialmap, hij*x,y), is a contoured
projection of hydraulic head values in the x±y plane deter-
mined in a horizontalsubsurface domain of constant thick-
ness and speci®ed elevation. In a map, h is the hydraulic
head in, and x and y are the horizontalEOV coordinates
of, the point of measurement; i and j are, respectively, the ele-
vations of the upper and lower boundaries of the domain of
measurement. Elevation z of the measurement point is always
i ꢁ z ꢁ j. A fullportrayalof the potentialdistribution in a
given rock volume generally requires several tomographic
maps.
Summary
In summary, the following observations and inferences have
been made from all eight pressure pro®les analysed in the
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 23
Table 1 Map codes and principal data of the tomographic rock framework
slices.
Map code
/hi,j
Top zi
Thickness
Data points
)
/m)
di,j /m)
/n)
hz0,4
h4,0
Land surface, z0
z0 À 40
0
ꢀ 40
75±140
100
200
300
400
500
400
300
300
300
300
400
2304
5120
3692
2547
891
464
339
331
281
196
106
66
h0,1
h1,3
À 100
h3,6
À 300
h6,10
h10,15
h15,19
h19,22
h22,25
h25,28
h28,31
h31,35
À 600
À 1000
À 1500
À 1900
À 2200
À 2500
À 2800
À 3100
49
under regional lowlands. This type of potential distribution
is known to indicate gravity-driven ¯ow regimes in regionally
uncon®ned ¯ow domains ꢀFreeze & Witherspoon 1967;
Â
Toth 1963, 1995). Principalregions of descending ¯ow,
i.e. groundwater recharge, are: the centralridge of the
Duna±Tisza inter¯uve, in general, and its prominently ele-
vated northern and southern parts, in particular, namely the
Fig. 14. Column of tomographic ¯uid potential maps, hij*x,y),showing map
Í
GoÈdoÈllo Hills and Illancs ꢀFig. 2); the North Central Range
codes and elevation intervals of the mapped rock slices.
along the study area's northern border; and the sandy hills of
 Â
the Nyõrseg, in the north-east. The ¯uid potentials do not
The rock framework of the study area was ®rst divided into
13 horizontal slices. The position and thickness of the slices
were de®ned, so as to ensure maximum data point concentra-
tion within the boundaries of individual slices, as well as to
minimize the number and magnitude of verticalpressure
changes within the slices. Details of the nomenclature, posi-
tions, dimensions and data point controlof the rock slices are
given in Fig. 14 and Table 1. Only ®ve examples of the 13
¯uid potentialmaps prepared for the study are presented
here. Characteristics of the potential®edl identi®ed on these
maps are consistent with those seen on the other eight maps.
The full set of maps is included in an unpublished report
change perceptibly with depth along the ꢀ 100 m hydraulic
head contour line, which itself follows closely the 100 m
topographic contour on both ¯anks of the ridge. This belt
of transition, or lateral ¯ow, is at a slightly higher elevation
 Â
of ꢀ 100±110 m in the Nyõrseg. The main regionaldischarge
areas are a 5±25 km wide belt along the east bank of the Duna
River, and the coalescing broad plains along the Tisza River
and the KoÈroÈs River system, the latter two together forming
the core area of the Great Plain. The rise in head with depth
beneath the plains is de®nite but only a few metres large. The
small magnitude is probably due to the signi®cantly larger
size ꢀmore than double) of the discharge area than that of
the recharge areas, and the high permeability.
Â
Â
ꢀToth & Almasi 1998).
Tomographic ¯uid potential maps hz0,04 and h1,3
Tomographic ¯uid potential maps h3,6, h10,15 and h22,25
The relief of the hydraulic head surface in the rock slab of
i ˆ À 300 m ꢁ z ꢁ À 600 m = j ꢀh3,6 map, Fig. 17) continues
to show the gradualdownward attenuation observed in the
shallower sections, but it still retains the relatively high poten-
tials in the regions of the inter¯uve, North Central Range and
The potentialmaps hz0,04 and h1,3 ꢀFigs 15 and 16) show the
hydraulic head distribution in the 40 m thick uppermost rock
framework slab and in the slab bounded by the horizontal
planes at z = À 100 m and z = À 300 m elevations, respec-
tively ꢀTable 1).
 Â
Three major regionalproperties characterize the ¯uid
potential distribution in the elevation range between the land
surface and z = À 300 m: ꢀi) the ¯uid potentialsurfaces
resemble the con®guration of the land surface; ꢀii) hydraulic
head values decline with increasing depth in regions of high
topography; and ꢀiii) hydraulic heads increase with depth
the Nyõrseg ꢀFig. 2). However, as a new feature, islands of
`potentialmounds' appear at a few ol cations, thus reversing
the trend of the hydraulic head's downward decline. For
example, in the general area of XEOV ˆ 80±140 km, YEOV
ˆ 660±700 km, hydraulic heads decline from the near-
surface values of h ꢀ 130±140 m to h ꢀ 120 m in the
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
24 J. TOTH & I. ALMASI
Fig. 15. Tomographic ¯uid potential map, hz0,40
,
based on 2304 measurement points between the
land surface, z0,and 40 m depth.
Fig. 16. Tomographic ¯uid potential map, h1,3
,
based on 2547 measurement points in the eleva-
tion range of À 100 to À 300 m a.s.l.
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 25
Fig. 17. Tomographic ¯uid potential map, h3,6
,
based on 891 measurement points in the elevation
range of À 300 to À 600 m a.s.l.
i ˆ À 100 m ꢁ z ꢁ À 300 m ˆ j slab ꢀh1,3 map, Fig. 16),
but abruptly reverse and increase to over 200 m in the i ˆ
À 300 m ꢁ z ꢁ À 600 m ˆ j slice ꢀh3,6 map, Fig. 17).
Depending partly on the speci®c position of available data
points, the head value of h ꢀ 200 m is maintained within
the region at slightly varying locations and gradually increas-
ing arealextent, through the maps of h3,6 ꢀFig. 17), h6,10 ꢀnot
shown) and h10,15 ꢀFig. 18), to the level of z ꢀ À 1500 m.
Although no more data are available at greater depths within
the block in question, the observed values of h > 1000 m
following the coordinate positions of: XEOV/YEOV ˆ 90/
720; XEOV/YEOV ˆ 130/680; XEOV/YEOV ˆ 190/730;
XEOV/YEOV ˆ 200/820; and XEOV/YEOV ˆ 190/860.
Between this arch of high overpressures and the island at
Battonya, a potentialbasin seems to exist with vaul es gener-
ally below 100 m, in the rock slab at i = À 1000 m ꢁ z ꢁ À
1500 m = j.
The foreshadowed potential arch has developed fully in the
rock slab i = À 2200 m ꢁ z ꢁ À 2500 m = j ꢀFig. 19), dividing
the entire area into a strongly overpressured `highland' to the
west and north±west, with h values generally above 1000 m
and exceeding 2000 m in places, and a relatively mildly over-
pressured 'potentialbasin' with h values, on average, between
150 and 200 m on the east and south±east side. However,
within this generalbasin, two subbasins aslo appear with
h < 100 m, while the mound at Battonya has grown to
h > 2000 m. Also, all segments of the arcuate belt of the
`pressure escarpment' have migrated towards the centre
with increasing depth, i.e. to the east, south±east and
south, thereby shrinking the `¯uid potentialbasin's'
arealextent from the higher elvelof h10,15 ꢀFigs 18 and 19,
respectively).
0
further east along Section H1±H1 ꢀFig. 2) suggest that the
trend of downward increasing potentialcontinues.
Three strikingly strong `overpressure islands' can be
observed on map h10,15 ꢀFig. 18): ꢀi) at Szolnok, in EOV
block XEOV ˆ 180±200 km, YEOV = 720±760 km, a
mound±shaped high, with h > 200 m; ꢀii) near Biharker-
esztes, in EOV block XEOV ˆ 180±220 km, YEOV
ˆ
800±860 km, an elongated potential ridge, with h values
exceeding 500 and 800 m; and ꢀiii) near Battonya, in EOV
block XEOV ˆ 100±140 km, YEOV = 780±800 km, with
h > 300 m. On map h15,19 ꢀnot shown), the mounds
mentioned in ꢀi) and ꢀii) nearly coalesce and h values in excess
of 1000 m are common. In addition to the appearance of a
growing number of overpressure islands of increasing magni-
tudes, an arcuate spatialarrangement in their positions also is
faintly taking place. Thus the overpressure anomalies seem to
be located along a slightly curved trajectory approximately
The trend of downward rapidly increasing hydraulic
head values continues to the deepest rock slab for which
data are available ꢀFig. 14; Table 1). Observed head values
reach and regularly exceed 3000 m at the elevation range of
i ˆ À 2500 m ꢁ z ꢁ À 2800 m ˆ j, and below. However, the
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
26 J. TOTH & I. ALMASI
Fig. 18. Tomographic ¯uid potential map, h10,15
,
based on 339 measurement points in the elevation
range of À 1000 to À 1500 m a.s.l.
Fig. 19. Tomographic ¯uid potential map, h22,25
,
based on 196 measurement points in the elevation
range of À 2200 to À 2500 m a.s.l.
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 27
data point distribution below z ˆ À 2800 m is insuf®cient for
ꢀ11) The generalsense of vertical¯ow in the con®ned
regime is ubiquitously and unequivocally upwards.
ꢀ12) The principalpathways of upward ¯ow in the con®ned
regime are faults, fracture zones and lithologic discontinu-
ities.
the evaluation of pressure ®eld patterns.
Summary
In summary, the following observations and inferences have
been made from the 13 hydraulic head maps, including eight
Â
Â
not discussed here but reported elsewhere ꢀToth & Almasi
Hydraulic cross-sections, h2z,l )
1998).
ꢀ1) The ¯uid potential surfaces closely mimic the
topographic relief between the land surface and elevation
z = À 300 m, i.e. to a depth of ꢀ 380±480 m below land
surface.
A hydraulic cross-section is a contour line representation of
the subsurface hydraulic head distribution in a vertical plane
of arbitrarily chosen azimuth. In the section, h is the hydraulic
head, z is the elevation referenced to sea level and l is the sec-
tion's horizontaldirection, or azimuth.
ꢀ2) Hydraulic head values decrease uniformly with depth
beneath regions of topographic elevations above 100±110
m; they remain constant in areas of ꢀ 100±110 m and rise
in areas of lower elevation, down to a depth of at least z =
À 300 m.
Four sections of a total length of approximately 850 km
were constructed and evaluated for the present work, namely:
0
0
00
H1ÀH1 ÀH100, ꢀ 180 km; H2ÀH2 H2
,
ꢀ 220 km;
0
0
H3ÀH3 ÀH300, ꢀ 300 km; and H4ÀH4 ÀH400, ꢀ 150 km;
or, for short reference: H1, H2, H3 and H4, respectively
ꢀFig. 2). Because of space limitations, only sections H1, H2
and H4 are displayed ꢀFigs 20, 21 and 22). The contours of
the sections are based on hydraulic head values produced in
two different ways. First, point values calculated for water
wells and hydrocarbon wells found within 2.0 km from either
side of the sections' trace were projected horizontally onto
the sections. Second, contour values of the tomographic
potentialmaps were transferred to the proper elevation inter-
vals of the cross-sections at the locations where the sections
intersected the maps' contour lines. The technique intro-
duces some uncertainty to the exact verticalposition of the
contour lines within the elevation intervals of the maps, but
it provides constrained values of head in regions where other-
wise none might be available.
ꢀ3) The potentiometric surface's regionalcon®guration
resembles the topography still in the elevation range of i =
À 300 m ꢁ z ꢁ À 600 m = j, but `overpressure mounds'
appear beneath Illancs and Battonya, and in the middle of
0
00
00
segment H3 ÀH3 of the hydraulic section H3ÀH3 ꢀnot
shown).
ꢀ4) The `overpressure mounds' intensify, extend, multiply
and migrate in the range of i = À 600 m ꢁ z ꢁ À 1500 m = j; a
`¯uid potentialpal teau' and a `¯uid potentialbasin' begin to
appear on the map of h10,15
.
ꢀ5) The gradually evolving `¯uid potential plateau', to the
west and north±west, and the `¯uid potentialbasin', to the
east and south±east, are ®rmly established with depth down
to the elevation range of i = À 2200 m ꢁ z ꢁ À 2500 m =
j; they are separated by a steep `potentialescarpment'. The
`escarpment' runs through the centre of the study area strik-
ing NE±SW, with a turn to the south, west of the Tisza River,
and has severalol bate protrusions and embayments subnor-
malto its generaldirection.
The hydraulic cross-sections con®rm and render the earlier
conclusion plausible, namely that groundwater ¯ow in the
upper zones of the basin is gravity driven and regionally
uncon®ned. In these zones, hydraulic heads decline with
increasing depth under high terrains, indicating downward
¯ow of in®ltrating meteoric water. The depth of the zone
of downward drive varies strongly, depending on the relief
of the land surface, but probably also on the local hydrostra-
tigraphy and the degree of interference from the deeper,
overpressured regime. For instance, the depth of the poten-
tialinversion ꢀreversalof downward drive into upward drive
along a vertical) is ꢀ 850 m at km 16 of Section H1 ꢀFig. 20),
but it is only ꢀ 200 m along km 28±35. The effect of
high topography is felt to a depth of ꢀ 1700 m ꢀelevation
ꢀ6) The ¯uid driving forces are orientated from the
basement highs ꢀhorsts) to the sediment±®lled troughs
ꢀgrabens).
ꢀ7) Sporadic data, available for lower levels of the rock fra-
mework, indicate a continuing increase of ¯uid potentials
with increasing depth, with values exceeding h = 3000 m at
the elevation range of i = À 2900 m ꢁ z ꢁ À 3500 m = j.
ꢀ8) The strike direction of the main potentialescarpment,
and the location and direction of its embayments and lobes,
appear to correlate with the principal faults in the basement
and/or the Neogene basin ®ll ꢀFig. 6).
Í
ꢀ À 1500 m) beneath the GoÈdoÈllo Hills at km 11 of H2
ꢀ9) Two vertically superposed pore pressure regimes can be
distinguished in the area using tomographic ¯uid potential
maps: a gravity-driven, uncon®ned upper regime, and a
strongly overpressured lower regime.
ꢀFigs 2, 21), but this depth decreases to ꢀ 600 m under lower
elevations at km 63. Further south±east on the east ¯ank of
the Duna±Tisza inter¯uve and beneath the plains of the Tisza
and the KoÈroÈs River system, ¯uid drive is upward at all
depths. An apparent lack of a ¯uid potential inversion at
the north end of H4 ꢀFig. 22) can be explained by the termi-
nation of the section in the regions of lateral and ascending
ꢀ10) The source of the ¯uid driving force in the lower
regime is attributed to currently active tectonic compression
ꢀargument presented later).
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
28 J. TOTH & I. ALMASI
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 29
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
30 J. TOTH & I. ALMASI
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 31
¯ow, i.e. short of the recharge areas of the North Central
Range further north ꢀFig. 2).
aquitards and aquifers on opposite sides of the fault planes,
the angle between the fault plane and the ¯uid driving forces,
and the spatialand temporalvariations of allthese factors
within the fault zone ꢀBjùrlykke 1996; MatthaÈi & Roberts
1996; Yassir & Otto 1997). Nevertheless, fault zones appear
to be important avenues for pressure diffusion, and thus for
¯uid ¯ow, in many parts of the Great Plain Basin. This con-
clusion is based on numerous situations of observed spatial
coincidence of fault zones with `pressure plumes' and `pres-
sure ridges' along the sections as, for instance: in H1 at km
72±76, 99±104 and 138±145 ꢀFig. 20); in H2 at km
182±188 and 205±206 ꢀFig. 21); and in H4 at km 52±58
and 102±103 ꢀFig. 22). It is corroborated also by estimated
verticalmigration distances of hydrocarbon gases on the
order of 3500 m ꢀClayton et al. 1990).
The presence of strong overpressures is clearly recogniz-
able beneath the gravity-driven regime on all hydraulic sec-
tions. A noteworthy property of the hydraulic head pattern
in this zone is that the high-value contours mimic the undu-
lations of the Pre-Neogene basement surface with dampened
or magni®ed amplitudes. In general, attenuation is linked to
deep, sediment-®lled troughs and subbasins, whereas ampli-
®cation is linked to the tops or steep ¯anks of the horsts. One
feature of the potentialpattern is striking: maximum over-
pressures tend to be associated with promontories of the
Pre-Neogene basement. Fluid driving forces thus seem to
be orientated generally from basement highs towards base-
ment depressions. This force pattern is manifest in upward
convex bands of higher valued and densely spaced contour
lines running atop basement highs, while lower valued,
upward concave h contours spread out widely in the deep,
sediment-®lled depressions.
Summary
In summary, the following observations and inferences have
been made from the four hydraulic cross-sections, including
Â
one not discussed here but reported elsewhere ꢀToth &
These relations are well exempli®ed on Section H1 by the
basins located at km 40±45 and 110±130, and by the base-
ment highs at km 55 and 155 ꢀFig. 20); on Section H2 by
the basins at km 25±40 and 135±155 and the sharp basement
shoulder at km 120±125 ꢀFig. 21); and on Section H4 by the
prominent rounded ridge at km 52±64 ꢀFig. 22).
Â
Almasi 1998).
ꢀ1) Two principalpore pressure zones are recognized: ꢀi)
an upper, gravity-driven, uncon®ned zone; and ꢀii) a lower,
overpressured, con®ned zone.
The dynamic character and geometry of the transition
between the con®ned and uncon®ned hydraulic regimes
depend on the local topography and the local geological con-
ditions. The change from the high-energy lower regime to
the overlying low-energy uncon®ned one can be gradual or
stepwise. In the case of a gradualtransition, the differences
in the hydraulic heads between the lower and the upper zone
are distributed more or less evenly over vertical distances of
hundreds of metres. On the other hand, at places of stepwise
transition, the change in hydraulic heads occurs within a
short depth interval. The vertical hydraulic sections show that
the gradual-type transitions occur mainly in the deep, sedi-
ment-®lled basins as, for instance, in H1 at km 45 and 120
ꢀFig. 20); in H2 at km 30 ꢀFig. 21); and in H4 at km 40
ꢀFig. 22). On the other hand, stepwise changes are associated
primarily with basement highs or their steep ¯anks, e.g. in H1
at km 145 ꢀFig. 20); in H2 at km 110±120 and 170 ꢀFig. 21);
and in H4 at km 60 and 115±120 ꢀFig. 22). A possible inter-
pretation of the cause of the stepwise transition is that the
source of overpressure is nearby. Other possible factors, inde-
pendent from the energy source's proximity, could be the
hydrostratigraphy and tectonic structure. The stepwise char-
acter of the potential®edl transition may be enhanced by
increased hydraulic resistance of low-permeability strata.
The possible role of faults in linking the two pressure
regimes cannot be easily generalized because it depends on
too many factors, the collective effects of which can rarely
be evaluated. Such factors include the permeability of the
fault walls and of the fault gouge, the relative position of
ꢀ2) In the overpressured zone, the ¯ow's verticalcompo-
nent is uniformly upwards.
ꢀ3) The depth of the boundary between the upper and
lower zones varies between 200 and 1700 m.
ꢀ4) A lower boundary of the gravity-driven ¯ow systems
cannot be identi®ed from hydraulic observations in regional
discharge areas because there the driving forces are orientated
similarly to, and merge with, those of the ascending depth
waters.
ꢀ5) The transition between the potential®edls of the
uncon®ned and con®ned zones may be gradualor stepwise
abrupt.
ꢀ6) Gradualtransition is prevaelnt in the deep, sediment-
®lled basins and troughs; stepwise changes are associated with
basement highs and their ¯anks.
ꢀ7) The hydraulic head contours mimic the undulations of
the Pre-Neogene basement, with reduced amplitudes in the
deep, sediment-®lled basins, and with ampli®cation over the
tops and steep sides of the basement ridges.
ꢀ8) In general, the lateral components of the ¯uid driving
forces are orientated from the basement highs towards the
intervening depressions.
ꢀ9) Faults and fracture zones appear to be preferential
pathways for pressure diffusion, and thus ¯uid ¯ow, upwards.
DISCUSSION
The most striking generalfeature of the basin's ¯uid potential
®eld is the two domain-wide and vertically superimposed
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
32 J. TOTH & I. ALMASI
hydraulic regimes, namely an upper, normally pressured
regime and a lower, strongly overpressured regime. The
departures of hydraulic heads from hydrostatic values in the
upper regime are mild and consistent with topographic
effects, as can be readily observed in the pꢀz) pro®les and
hꢀz,l ) sections ꢀFigs 9±13 and 20±22, respectively). In this
regime, water is driven downwards under topographic highs,
laterally under mid-elevations, and up towards the land sur-
face beneath depressions. In the lower regime, the vertical
component of the driving force is upwards everywhere. Con-
sequently, the boundary between the two regimes is easily
distinguished where their forces are oppositely directed, i.e.
under the highlands, but it vanishes under the topographic
depressions where the water is driven upwards in both regimes.
The hijꢀx,y) and hꢀz,l ) contours show both regionaltrends
and local perturbations in both regimes. ꢀNote that `local'
may mean several kilometres laterally and hundreds of metres
vertically on the scale of presentation.) The regional trends
1976), but which are not discernible at the present scale.
These systems are generated by small local undulations of
the water table, which follows the topographic relief within
a few metres, due to the generally adequate precipitation.
The strongly variable depth of the gravity-driven regime is
the result of hydraulic interaction between the topographic
relief, lithologic heterogeneities and overpressures in the
lower hydraulic regime. In general, the gravitational regime
appears deeper above the deep, sediment-®lled basins and
grabens ꢀe.g. km 100±125, Section H1, Fig. 20; km 20±40,
Section H2, Fig. 21; and km 55±60 and 130±150, Section
H4, Fig. 22). Based on the pꢀz) pro®les and cross-sections,
this depth appears to vary between 200 and 1700 m.
The overpressured regime
The main observable characteristics of the overpressured
regime are: ubiquitously upward orientated forces; vertical
pressure gradients increasing monotonically downwards
without noticeable reversals; general congruence of high
hydraulic heads with the basement highs; and broad, band-
like scatter of pressure values on the pꢀz) pro®les. The mono-
tonic increase of the pressure gradients is suggestive of a
source depth beneath the observationaldomain, whiel the
congruence of hydraulic head maxima and basement highs
indicates driving forces orientated outwards from the crystal-
line basement and into the sedimentary basin ®lls. These
properties of the ¯uid potential®edl are interpreted here as
coherent symptoms of a currently active, high-pressure
energy source seated in the Pre-Neogene basement or
beneath it.
are manifest in the potentialmaps and cross-sections by
h
contour con®gurations that re¯ect the effect of the topogra-
phy in the shallow regime, and are generally congruent,
although with dampened amplitudes, with the basement's
relief in the overpressured regime. In the pꢀz) pro®les, the
sensitivity of the verticalpressure gradients to topographic
elevations in the shallow regime is a general feature and
re¯ects that regime's uncon®ned character. On the other
hand, the gradients, which increase monotonically with
depth in the overpressured zone, are suggestive of an energy
source in or below the Pre-Neogene basement.
The local perturbations of the hijꢀx,y) and hꢀz,l ) contours
are interpreted to re¯ect inhomogeneities of permeability due
to sedimentological or structural discontinuities. No mean-
ingfulol calperturbations can be observed in the upper seg-
ments of the pꢀz) pro®les, owing probably to scaling factors.
The broad scatter of pressure values in the lower segments is,
on the other hand, interpreted as perturbations, and indicates
lithologic heterogeneities, albeit in a lumped rather than dis-
tributed, and thus site-speci®c, manner.
Regarding the locus of the source of high pressures and the
possible role of faults and fracture zones in causing the wide
scatter of pressure values at given depths, this interpretation is
corroborated by the numericalexperiments of MatthaÈi &
Roberts ꢀ1996) ꢀFig. 23). The experiments show the pressure
and ¯ow ®elds generated through the overpressured basal
boundary of the ¯ow domain in a strati®ed and faulted rock
body to display properties similar to those observed in the
Great Plain. Opposite considerations, represented by models
of compaction-generated force ®elds, i.e. by energy sources
higher up in the ¯ow domain, tend to support the inference
concerning the depth of the energy source. Such energy
sources have been shown to induce overpressures which also
have a downward drive, as illustrated conceptually in Neuzil
ꢀ1995; Fig. 5), and by numerical models of the Gulf Coast
ꢀHarrison & Summa 1991; Fig. 15) and the Sacramento
Basin, California ꢀMcPherson & Garven 1999, Fig. 12).
The other important question arising from these observa-
tions concerns the mechanism of overpressure generation.
The clue to this question is sought in a genetic relation
between the pore pressure ®eld and the current state of
tectonic stresses and crustaldispalcements in the region. As
indicated by the geometric patterns of the ®elds of hydraulic
The gravity-driven hydraulic regime
The specialcharacteristics of the gravity-driven hydrauicl
regime are the hierarchically nested distribution of ¯ow sys-
tems and the highly variable depth of the regime.
Some examples to illustrate the ®rst point are: segments
40±95 km on hydraulic cross-section H1 ꢀFig. 20), and 40
km and 90±115 km on H2 ꢀFig. 21). In the ®rst example,
the recharge and discharge areas of a 55 km long ¯ow system
are superposed on a generally upward orientated ¯ow
domain, while in the second case several systems less than
10 km in lateral dimension are `hydraulically perched' by
upward ¯ows of larger systems. The hierarchical arrangement
continues upwards to systems less than 100 m in length as
Â
noted elsewhere ꢀAngelus 1996; Arday 1996; Erdelyi
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 33
derived from linear strain rates ꢀe.g. Fig. 5), or the character-
istic ¯ow length or effective basin size ꢀwhich are required to
calculate the strength of forcing, compression in the present
case, or the rocks' hydraulic resistance), would render the
results indefensible. Instead, another observation by Neuzil
ꢀ1994) might be invoked to resolve the apparent dichotomy
of the coexistence of high permeabilities and high overpres-
sures. Neuzilꢀ1994; p. 149) observed that, owing to their
relatively permeable subunits, argillaceous formations can
be much less permeable than the values obtained from their
large-scale mapping. This observation might be applicable to
the available permeability values for the rock framework of
the Hungarian Great Plain. High-permeability discontinu-
ities due to faults, fracture zones and sedimentary windows
are well known in the basin's rock framework ꢀe.g. Figs 7
and 8), as well as clearly manifest by high values of hydraulic
head anomalies on the potential maps ꢀFigs 17±19) and
cross-sections ꢀFigs 20±22). In addition, the major geother-
malanomailes in the area can be interpreted as indications of
localized ascent of deep-seated hot formation waters ꢀAlfoÈldi
et al. 1978; Lenkey 1993).
Fig. 23. Modelled ¯uid pressure isobars /contours in MPa) and directions of
¯uid ¯ow /arrows),generated by a pressure source at the domain's base,in the
vicinity of a conductive fault with displaced aquifers on opposite sides in a low-
Whereas the observed distributions of pore pressures,
hydraulic heads, lithologic properties and crustal movements
are consistent in suggesting tectonic compression as the prin-
cipalsource of overpressuring in the Great Plain of Hungary,
other potential mechanisms are unlikely to contribute signif-
icantly.
permeability environment. Pressure from surface to 3000
m depth is
hydrostatic; at 3400 m it is 0.8 Â lithostatic. Pressure is at steady state
/adapted from MatthaÈi & Roberts 1996,®g. 4,with the permission of the
American Association of Petroleum Geologists).
head and pore pressure in the overpressured zone, ¯ow is out
of the crystalline basement and into the adjacent or overlying
clastic sediments. If this observation is combined with the
structuralgeologicaland seismic evidence of a compressional
As mentioned before, compaction would drive ¯uids from
the centres of the grabens towards the horsts; in reality, the
opposite is true. Also, no signi®cant downward component
of the driving force seems to exist, and the region of major
overpressuring is in the most highly consolidated, i.e. least
compressible, older rocks of the basement, rather than in
the more highly deformable, younger and partially consoli-
dated basin ®ll.
Â
stress ®eld ꢀCsontos 1995; Gerner et al. 1997; Horvath &
Cloetingh 1996; Peresson & Decker 1997) as epitomized
by Fig. 4 ꢀBada et al. 1998), and with the high rates of mea-
sured convergence of intraplate segments towards the study
area ꢀbetween 2 and 4 mm yearÀ1, Fig. 5; Grenerczy et al.
2001), it is tempting to ascribe the origin of overpressuring
to tectonic compression. It is probably meaningful also to
compare these results with the numerical experiments of
McPherson & Garven ꢀ1999), who found that a crustal
shortening of 2.9±3.5 mm yearÀ1 can generate 10±15 MPa
overpressures at a depth of 3000 m in the Sacramento Basin
of California. They observed further that localized regions of
higher compressibility or lower permeability could produce
local overpressures greater than 25 MPa even at shallower
depths, results that are quite consistent with the empirical
®ndings of the present study.
Volumetric expansion of maturing hydrocarbons is an unli-
kely source of overpressures in the area. According to Szalay
& Koncz ꢀ1993; p. 305): `The deep basin sediments along
the axis of the deep troughs presently are in the main phase
of dry gas or wet gas generation.' Yet, as is clearly observable
on the hydraulic cross-sections ꢀFigs 20±22), the horizontal
components of the driving forces point in all cases from the
basement highs towards the basin centres, i.e. towards the
loci of hydrocarbon maturation.
Osmosis must also be ruled out as a potential source of
overpressures because of the inadequate salinity contrasts in
the formation waters, and the insuf®cient thickness, areal
continuity and too high permeabilities of potential shale
membranes. According to Neuzilꢀ2000; Fig. 3), the maxi-
mum osmotic pressure that could be generated in the highly
continuous and argillaceous Pierre Shale of central South
Dakota, USA ꢀk = 10À6±10À5 md) is approximately 0.4
MPa at a depth of 3 km, if a porosity of 0.2 and totaldissolved
solids content ꢀTDS) contrast of 3 g LÀ1 are assumed. This
The question may nevertheless be asked if the permeabil-
ities accepted for the Great Plain Basin's rocks are not too
high to maintain the observed disequilibrium ꢀover) pres-
sures. The answer is: the observed facts suggest apparently
not. In theory, this question can be explored quantitatively
by means of Neuzil's ꢀ1995) concept of `geological forcing'.
In practice, however, the uncertainties in estimating the
values of parameters, such as the volumetric strain rate as
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
34 J. TOTH & I. ALMASI
value is nearly two orders of magnitude less than that
observed in the Great Plain Basin.
tural, geological and geophysical observations, as well as
recent satellite geodetical measurements.
In the case of aquathermalpressuring, the loci of maximum
overpressures would be expected to coincide with the basin's
deepest and hottest subsiding centre. Also, vertical pressure
gradients should be nearly hydrostatic in the overpressured
compartments below the top seals. Neither of these situations
can be observed.
The basin hydrostratigraphy has been divided into ®ve
regionally extensive major units, namely the Pre-Pannonian
ꢀk = 100 md), Szolnok ꢀk = 10±100 md) and Great Plain ꢀk
Í
= 1000 md) Aquifers, and the Endrod ꢀk = 0.1±10 md) and
Í
Í
Algyo ꢀk = 1±10 md) Aquitards. The Algyo Aquitard contains
many high-permeability lenses, and both aquitards are dis-
sected by numerous faults and fractures.
The buoyancy of light ¯uids ꢀhydrocarbon gases, CO2, oil)
would produce overpressures orders of magnitudes less, as
well as vertical pressure gradients noticeably less, than hydro-
static. Furthermore, overpressures are ubiquitous in the basin
while the required ¯uids occur sporadically.
Characterization of the ¯uid potential®edl is based on 16
192 hydraulic head values culled from more than 52 000
water level and pore pressure measurements conducted in
shallow dug wells, drilled water wells up to 1500 m deep
and hydrocarbon wells reaching depths exceeding 4000 m.
The data were processed by three different techniques of
representation: vertical pressure-elevation pro®les, pꢀz);
tomographic hydraulic head maps, hijꢀx,y); and vertical
hydraulic head cross-sections, hꢀz,l ).
The transitional zone
Perhaps the most signi®cant characteristics of the zone of
transition between the overpressured and normally pressured
hydraulic regimes are its widely variable depth and the lack of
an identi®able single controlling factor. The depth and loca-
tion of the transition can be determined on the pꢀz) pro®les in
a given geographicalregion by the depth of the break in the
gradients away from the hydrostatic line towards higher
values; by the location and depth on the hijꢀx,y) maps where
superhydrostatic pressures appear; and by the location and
depth of bands of congested contour lines on the hꢀz,l )
cross-sections. The wide variation in depth is immediately
evident, and the lack of either a lithologic or stratigraphic
controlwas pointed out in connection with the anaylsis of
the pꢀz) pro®les. Structural discontinuities, such as faults
and fracture zones, and probably also high-permeability sedi-
mentary windows, are avenues of hydraulic pressure dissipa-
tion and ¯uid migration from the overpressured regime to
the gravity-driven zone. In addition to the structurally con-
trolled inter®ngering type of transition, diffusion also must
occur across the formations due to the high verticalgradients
and relatively high matrix permeabilities.
The observed ¯uid potentialpatterns show two superim-
posed and laterally extensive hydraulic regimes in the basin
with independent and separate sources of ¯uid drives. The
top regime is gravity driven, regionally uncon®ned, normally
pressured and is recharged from precipitation. Its depth varies
from 200 to 1700 m. The lower regime is strongly overpres-
sured ꢀhydraulic heads may exceed 2500 m locally) with ver-
tical gradients increasing monotonically with depth,
regionally con®ned, and thought to be overpressured by tec-
tonic compression of the basement. The main sites of over-
pressure are the basement's horsts, from which divergent
¯uid impelling forces are orientated upwards and towards
the interhorst basins and grabens. Communication between
the two pressure regimes occurs by diffusion across geologi-
calstrata and/or through high-permeabiilty structuraland
sedimentological discrete discontinuities. The different
methods and avenues of cross-formationalcommunication
result in vertical inter®ngering of the two regimes with a tran-
sitionalzone varying between 200 and 1700 m in depth.
Some of the high-permeability avenues for pressure dissipa-
tion and ¯uid ¯ow are manifest by perturbations of the
potential®edl . These perturbations provide diagnostic pat-
terns that may be of practical value for water resources devel-
opment and petroleum and geothermal exploration.
CONCLUSIONS
This paper presents a summary of the characteristics and
interpretation of the observed subsurface ¯uid potentialpat-
terns in the Hungarian Great Plain region of the Pannonian
Basin located in Central Europe. The study area proper cov-
ers approximately 40 000 km2 at a topographic elevation
range of ꢀ 80±200 m and is surrounded by the almost closed
ring of the Alps, Carpathians and Dinarides. In the area, thick
semi- to unconsolidated clastics of marine, deltaic, lacustrine,
¯uviatile and eolian sediments of Neogene age rest on a Pre-
Neogene basement of brittle ¯ysch, carbonate and meta-
morphic rocks of high relief with a horst-and-graben topo-
graphy. After a period of crustalextension and rifting, a
tectonic reversalto a compressionalstress regime started in
the Late Miocene. It is still ongoing, as indicated by struc-
ACKNOWLEDGEMENTS
Subsurface data for the work reported in this paper were pro-
vided by MOL Rt. ꢀHungarian Oiland Gas Co., Ltd) and,
Í
through Dr Judit Szonyi, by the Department of Applied
Â
and EnvironmentalGeoolgy, Eo ÈtvoÈs Lorand University of
Science ꢀELTE), both of Budapest, Hungary. We appreciate
Â
Â
the personalinterest and support of Dr Istva n Berczi, and are
gratefulfor the technicaland scienti®c advice, logisticalassis-
tance, continual colleagueship and personal friendship
Â
Â
 Â
received from Mr Sandor Pap and Dr Istvan Revesz, all of
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Subsurface ¯uid potentials,Hungarian Great Plain 35
Freeze RA, Witherspoon PA ꢀ1967) Theoreticalanayl sis of regional
groundwater ¯ow. 2. Effect of water table con®guration and sub-
surface permeability variation. Water Resources Research, 3, 623±34.
MOL Rt. Particular thanks are due to Mr Gy. Grenerczy of
ELTE and the Hungarian Satellite Observatory for permit-
ting the use of the intraplate velocity map ꢀFig. 5) prior to
publication in their own paper. The present version of the
work bene®tted greatly from penetrating comments and sug-
gestions by Drs Grant Garven and Mark Person. The 3 year
long project was made possible by ®nancial contributions and
logistical support from MOL Rt., and Research Grant No. A-
8504 from the NaturalSciences and Engineering Research
Councilof Canada ꢀNSERC) to the ®rst author.
Â
Gajdos I, Pap S, Somfai A, VoÈlgyi L ꢀ1983) Az alfoÈldi Pannoniai ꢀs.l.)
Â
Â
Â
Â
kepzoÄdmenyek litosztratigra®ai egysegei [Lithostratigraphic units
of the Pannonian ꢀs.l.) in the Great Hungarian Plain]. Magyar
Â
Â
Allami FoÈldtani Intezet *Hungarian Geological Survey), 70, 1±70
ꢀin Hungarian).
Â
Gerner P, Bada G, DoÈvenyi P, Mueller B, Oncescu MC, Cloetingh S,
Â
Horvath F ꢀ1999) Recent tectonic stress and crustaldeformation
in and around the Pannonian basin: data and models. In: The
Mediterranian Basins: Tertiary Extension within the Alpine Orogen
Â
Â
ꢀeds Durand B, Jolivet L, Horvath F, Seranne M), pp. 269±94.
Geological Society of London Special Publication, 126.
Grenerczy Gy, Kenyeres A, Fejes I ꢀ2000) Present crustalmovement
and strain distribution in CentralEurope inferred from GPS
measurements. Journal of Geophysical Research, 105, 21 835±21
846.
REFERENCES
Â
Â
AlfoÈldi L, Erdelyi M, Gal® J, Korim K, Liebe P ꢀ1978) A geothermal
¯ow system in the Pannonian Basin: case history of a complex
hydrogeological study at Tiszakecske. In: Hydrogeology of Great
Â
Â
Sedimentary Basins ꢀed. Ronai A), pp. 716±32. Hungarian
Grenerczy Gy, Kenyeres A, Fejes I ꢀ2001) Present crustaldeforma-
tion pattern in the PANCARDI region: constraints from space
geodesy. In: Neotectonics and Seismicity of the Pannonian Basin
and Surrounding Orogens; a Memoir on the Pannonian Basin ꢀeds
Â
Geological Institute, Memoires XI, Budapest.
Angelus B ꢀ1996) A felszõnalatti vizaramlasok tanulmanyozasa a
Duna±Tisza koÈze koÈzepsoÈ teruÈleten, Fokto es Csanytelek koÈzoÈtt
Â
Â
Â
Â
Â
Â
Â
Í Â
Â
Â
[Study of the groundwater ¯ow systems in the centralregion of the
Cloetingh S, Horvath F, Bada G, Lankreijer A). European
Í
Â
Duna±Tisza inter¯uve, between the towns of Fokto and Csany-
Geological Society Special Publication ꢀpreprint).
Â
    Â
Halasz B ꢀ1975) Retegzett hidrologiai rendszerek sajatossagai
telek]. MSc Thesis, EoÈtvoÈs Lorand University, Budapest, Hungary
Â
ꢀin Hungarian).
ꢀProperties of strati®ed hydrologic systems). Hidrologiai KoÈzloÈny
*Journal of the Hungarian Hydrological Society), 55, 505±7.
Harrison WJ, Summa LL ꢀ1991) Paleohydrology of the Gulf of
Mexico basin. American Journal of Science, 291, 109±76.
Â
Â
Â
Â
Â
Arday A ꢀ1996) A felszõnalatti vizaramlasok tanulmanyozasa a Duna±
Â
 Â
Tisza koÈze del i reszen [Study of the groundwater ¯ow systems in
the southern region of the Duna±Tisza inter¯uve]. MSc Thesis,
Â
Â
EoÈtvoÈs Lorand University, Budapest, Hungary ꢀin Hungarian).
Horvath F, Cloetingh S ꢀ1996) Stress-induced late-stage subsidence
Â
Bada G, Cloetingh S, Gerner P, Horvath F ꢀ1998) Sources of recent
anomalies in the Pannonian Basin. Tectonophysics, 266, 287±300.
Hubbert MK ꢀ1940) The theory of ground-water motion. Journal of
Geology, 48, 785±944.
Hubbert MK ꢀ1953) Entrapment of petroleum under hydrodynamic
conditions. Bulletin of the American Association of Petroleum
Geologists, 37, 1954±2026.
tectonic stress in the Pannonian region; inferences from ®nite
element modelling. Geophysical Journal International, 134, 87±
101.
Ballentine CJ, O'Nions RK, Oxburgh ER, Horvath F, Deak J ꢀ1991)
Rare gas constraints on hydrocarbon accumulation, crustal
degassing and groundwater ¯ow in the Pannonian Basin. Earth
Â
Â
Hungarian Geological Survey ꢀ1993) Field trip C: Oil and Gas,
Subsurface Water, and Geothermy in the Pannonian Basin,
and Planetary Science Letters, 105, 229±46.
Berczi I, Hamor G, Jambor A, SzentgyoÈrgyi K ꢀ1988) Neogene
Â
Â
Â
Â
 Â
Excursion Guide ꢀeds Liebe P, Revesz I). 8th Meeting of the
sedimentation in Hungary. In: The Pannonian Basin; a Study in
Association of European Geological Societies, Hungary.
Â
Â
Basin Evolution ꢀeds Royden LH, Horvath F), pp. 57±67.
Juhasz Gy. ꢀ1991) Lithostratigraphic and sedimentological frame-
American Association of Petroleum Geologists Ð Memoir 45,
work of the Pannonian ꢀs.l.) sedimentary sequence in the
Hungarian Plain ꢀAlfoÈld), eastern Hungary. Acta Geologica
Hungarica, 34, 53±72.
Â
Tulsa, OK.
Â
Â
Berczi I, Kokai I ꢀ1976) Hydrogeological features of some deep-
Â
Â
Â
Â
Â
Â
Â
 Â
 Â
basins in SE-Hungary as revealed by hydrocarbon exploration. In:
Korim K ꢀ1966) A pannoniai retegek vaztarolo- es võzadokepesseget
Â
Â
Â
Â
Hydrogeology of Great Sedimentary Basins ꢀed. Ronai A), pp. 69±
meghatarozo foÈldtani tenyezoÄk ꢀGeological factors determining
Â
93. Hungarian Geological Institute, Memoires XI, Budapest.
water storage and water supply capacity of the Pannonian strata).
Â
Bjùrlykke K ꢀ1997) Lithological control on ¯uid ¯ow in sedimentary
basins. In: Fluid Flow and Transport in Rocks Ð Mechanisms and
Effect ꢀeds Jamtveit B, Yardley BWD), pp. 1±14, Chapman & Hall,
London.
Hidrologiai KoÈzloÈny *Journal of the Hungarian Hydrological
Society), 46, 522±31.
Korim K, Liebe P ꢀ1973) A szentesi hevõztarolo rendszer ꢀThe
thermalwater storage system of Szentes). VõzuÈgyi KoÈzlemenyek
 Â
Â
Â
Â
Â
Clayton JL, Spencer CW, Koncz I, Szalay A ꢀ1990) Origin and
*Hydraulic Engineering), 55, 290±311.
 Â
Â
Í
Â
Â
migration of hydrocarbon gases and carbon dioxide, Bekes Basin,
Lenkey L ꢀ1993) A tiszakecskei hoanomalia vizsgalata termikus
Â
konvekcio numerikus modellezesevelꢀExamination of the geo-
 Â
southeastern Hungary. Organic Geochemistry, 8, 233±47.
Csontos L ꢀ1995) Tertiary tectonic evolution of the Intra-Carpathian
area: a review. Acta Vulcanologica, 7, 1±13.
Dobos I ꢀ1985) Exploration of subsurface waters in the Neogene
basins. In: Proceedings of Neogene Mineral Resources in the
Carpathian Basin; Historical Studies on Their Utilization, VIIIth
Â
thermalanomaly of Tiszake cske by numericalmodeling of thermal
convection). Magyar Geo®zika *Hungarian Geophysics), 34, 30±45.
MatthaÈi SK, Roberts SG ꢀ1996) The in¯uence of fault permeability
on single-phase ¯uid ¯ow near fault-sand intersections: results
from steady-state high-resolution models of pressure-driven ¯uid
¯ow. Bulletin of the American Association of Petroleum Geologists,
80, 1763±79.
Â
RCMNS Congress, Hungary ꢀed. Hala J), pp. 531±55. Hungarian
Geological Survey, Budapest.
Â
Erdelyi M ꢀ1976) Outlines of the hydrodynamics and hydrochem-
McPherson BJOL, Garven G ꢀ1999) Compressionaltectonics,
hydrodynamics and overpressure mechanisms in the Sacramento
Basin, California. American Journal of Science, 299, 429±66.
istry of the Pannonian Basin. Acta Geologica Academiae Scientiar-
ium Hungarica, 20, 287±309.
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36

Â
Â
36 J. TOTH & I. ALMASI
Â
Neuzil CE ꢀ1994) How permeable are clays and shales? Water
Resources Research, 30, 145±50.
NeuzilCE ꢀ1995) Abnormalpressures as hydrodynamic phenomena.
American Journal of Science, 295, 742±86.
Szalay A, Koncz I ꢀ1993) Migration and accumulation of oil and gas
generated from Neogene source rocks in the Hungarian part of the
Pannonian Basin. In: Generation, Accumulation and Production of
Europe's Hydrocarbons III ꢀed. Spencer AM), pp. 303±9. Special
Publication of the European Association of Petroleum Geoscien-
tists no. 3, Springer-Verlag, Berlin.
NeuzilCE ꢀ2000) Osmotic generation of `anomalous' ¯uid pressures
in geologic environments. Nature, 403, 182±4.
Â
Â
Â
Â
Í
Â
Â
Â
Â
Â
Í
Â
Â
  Â
Pap S ꢀ1976) AlfoÈldi es Eszaki-KoÈzephegysegi koolaj-foÈldgaztarolo
Szebenyi L ꢀ1955) Artezi vizeink fuÈggoleges iranyu mozgasarolꢀOn
Í
Â
kozetek ꢀHydrocarbon bearing reservoir rocks of the Great
the verticalmotion of our artesian waters). Hidrologiai KoÈzloÈny
Hungarian Plain and the North Hungarian Highland). FoÈldtani
KoÈzloÈny *Bulletin of the Hungarian Geological Society), 106, 555±
80.
*Journal of the Hungarian Hydrological Society), 35, 437±9.
Â
Toth J ꢀ1963) A theoretical analysis of groundwater ¯ow in small
drainage basins. Journal of Geophysical Research, 68, 4795±812.
Â
Peresson H, Decker K ꢀ1997) Far-®eld effects of Late Miocene
subduction in the Eastern Carpathians: E±W compression and
inversion of structures in the Alpine±Carpathian±Pannonian
region. Tectonics, 16, 38±56.
Toth J ꢀ1995) Hydraulic continuity in large sedimentary basins.
Hydrogeology Journal, 3, 4±16.
Â
Â
Â
Â
Â
Toth J, Almasi
I
ꢀ1998) Szenhidrogen kutatas az AlfoÈldoÈn
Â
Â
Â
 Â
Â
hidrogeologiai szempontok ®gyelembevetelevel: Zarojelentes. A
Â
Â
Â
  Â
Â
Â
 Â
Royden LH, Horvath F ꢀeds) ꢀ1988) The Pannonian Basin: A Study
Magyar Olaj- es Gazipari Rt. megbõzasabol es szamara [Hydro-
carbon exploration in the Great Plain using hydrogeological
principles: Final Report. Prepared for the Hungarian National Oil
& Gas Company, Plc.]. University of Alberta, Edmonton, Canada
ꢀin-house report, in Hungarian).
in Basin Evolution. American Association of Petroleum Geologists
Ð Memoir 45, Tulsa, OK.
Rumpler J, Horvath F ꢀ1988) Some representative seismic re¯ection
lines from the Pannonian basin and their structural interpretation.
Â
In: The Pannonian Basin; a Study in Basin Evolution ꢀeds Royden
Yassir N, Otto CJ ꢀ1997) Hydrodynamics and fault seal assessment in
the Vulcan Sub-basin, Timor Sea. Australian Petroleum Produc-
tion and Exploration Association Journal, 36, 380±9.
Â
LH, Horvath F), pp. 153±71. American Association of Petroleum
Geologists Ð Memoir 45, Tulsa, OK.
# 2001 Blackwell Science Ltd, Geo¯uids, 1, 11±36