EEE23-5-2 30/01/06 1:29 pm Page 307

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 pp. 307–332 307

Deep Structure Model and Dynamics of Mud Volcanoes,

Southwest Absheron Peninsula (Azerbaijan)

F. A. Kadirov1, I. Lerche*,1,2, I. S. Guliyev1, A. G. Kadyrov1, A. A. Feyzullayev1,

and A. Sh. Mukhtarov1
1) Geology Institute of the National Academy of Sciences of Azerbaijan, H. Javid Prospekt,
29A, 370143 Baku, Azerbaijan
2) Institut fuer Geophysik und Geologie, Universitaet Leipzig, Talstrasse 35,
04103 Leipzig, Germany

ABSTRACT
Measurements have been made of gravitational field, geodetic uplift, regional
horizontal tectonic movement, thermal patterns, and radioactivity in the general
area of the Lokbatan mud volcano after its explosion on October 25, 2001, as well
as in the crater itself. In addition, geochemical measurements of vitrinite
reflectance with depth have been done and isotopic variations of methane and
ethane are also made available. This massive compendium of information
represents the first time such a detailed investigation has been possible of the deep
structural effects of a mud volcano and also of the sources of mud and gas at
outflow time. The data are integrated into a combined picture that shows the roots
of both the mud outflow and of the gas causing the flaming eruption are at several
kilometres depth into the sedimentary pile. The overall behaviour is best served by
a model in which a relatively thin jet of liquefied mud is extruded from depth due
to action of the varying tectonic stresses in the region, as adduced from the global
positioning system (GPS) tectonic movement data. The variation of Bouguer
gravity across a profile including the Lokbatan mud volcano, and combined with
the geodetic vertical motion immediately after and long after (10 months) the
explosion, confirms this basic model. The focusing of heat flux around the volcano
prior to the explosion, and the thermal measurements made with time after the
explosion both in the crater and also in the immediate vicinity of the Lokbatan
volcano, are in accord with a thin hot jet model in which liquefied mud, with
entrained gas from deeper in the sediments, rises through a neck region and, due to
the Rayleigh–Bernard convective instability, produces a high temperature region.
The geochemical evidence, showing low vitrinite maturity (<0.6%) to a depth of
around 6 km, also indicates production of oil and gas from greater depths, as do the
isotopic carbon measurements of methane and ethane in the unburnt gases. In short,
it would seem that tectonic “squeezing” of a low-strength plastic mud layer from
depth through a narrow vent with entrained gas and mud is the primary driver

*Corresponding author. E-mail address: kadirov@excite.com (Ian Lerche)

EEE23-5-2 30/01/06 1:29 pm Page 308

308 Deep Structure Model and Dynamics of Mud Volcanoes

for mud volcano explosions. In the general regional, approximately linearly

arranged lines of mud volcanoes, with their apparent focus centred on Shemakha
from where the lines fan out, is also a strong indication of the basic tectonic origin.
The combination of a rapidly filled sedimentary region, with unconsolidated (or
deconsolidated) muds occupying a domain at several kilometre depth, and
bracketed above and below by more competent formations, together with the active
horizontal stress variations as measured by the GPS network, together form the
basis for the spectacular mud volcano effects in this part of Azerbaijan.

Key Words: Mud Volcanoes; Azerbaijan; Deep Structure; Dynamics; Thermal;

Geochemical

I. INTRODUCTION
In the South Caspian Basin, the general region of onshore and offshore Azerbaijan is
home to over 200 mud diapirs and/or mud volcanoes. These mud structures are
associated with the production of copious oil and gas, and production wells are to be
found on the flanks of many onshore mud diapirs. This association is no mere
coincidence but is related to the dynamical development of mud diapirs and the
generation, migration, and accumulation of hydrocarbons in the South Caspian Basin,
as has been detailed elsewhere (Yakubov et al., 1971).
Mud volcanism is a striking natural phenomenon. Much attention to this type of
volcanism was initiated by Gubkin and Fedorov (1938), who showed that hydrocarbon
accumulations and mud volcanoes are associated with the formation of local diapiric
structures. At present, there is a clear correlation with mud volcanism manifestations,
considered as important direct evidence for the presence of oil and gas accumulations.
However, problems that have not been adequately studied are related to the formation
mechanism of mud volcanoes, their role in the origin and/or accumulation of
hydrocarbons, the role of deep faults, and the regime of fluid motion in volcanoes.
Unfortunately, in spite of long-term complex studies of the Azerbaijan mud volcanoes,
no special geophysical investigations have been undertaken. There are available only the
results of geophysical studies obtained serendipitously while solving other oil field
problems; these investigations are not enough to compile a mud volcano model nor to
study the dynamics of its activity. Such a capability is possible only with special, very
detailed geophysical and geodetic investigations.
This article contains the results of such investigations carried out for the first time
on volcanoes in the southwest Absheron region, viz., Lokbatan, Akhtarma-Puta, and
Gushkhana, all located within one single tectonic zone (Figure 1).
The Lokbatan mud volcano was chosen because it is very active not only in
comparison to other volcanoes in Azerbaijan but in respect also of all worldwide mud
volcanoes. Lokbatan has been recorded as having erupted 21 times in the last 200 yr,
assuming no other eruptions have been missed. The last eruption occurred on October 25,
2001. The present studies were started immediately after that eruption. Together with the
gravimetric and geodetic studies, temperature, radiometric and geochemical observations
were also conducted. Profiles in different directions were chosen to carry out gravimetric
EEE23-5-2 30/01/06 1:29 pm Page 309

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 309

45˚

44˚

43˚

42˚

41˚

40˚

39˚

38˚

37˚ km
0 100 200
36˚
38˚ 40˚ 42˚ 44˚ 46˚ 48˚ 50˚ 52˚ 54˚ 56˚

Figure 1. Distribution of mud volcanoes in Azerbaijan and area of study

and geodetic measurements on areas of mud volcanoes. One profile, more extended than
the others (11 km long), begins near a key gravity-geodetic point at Lokbatan and passes
through a line joining the mud volcanoes Lokbatan–Akhtarma-Puta–Gushkhana. The
distances between measurement points are never more than 250 m. To the north of a
line connecting Shykhlar–Madrasa, GPS measurements indicate that there is a decreasing
speed of platform motion, reaching zero velocity in places. Observations of dynamic
processes in and on the volcanoes are continuing at the present time on a semi-regular
basis.

II. GLOBAL GEODYNAMICS AND THE AREA AROUND

MUD VOLCANOES
In 1998, the Azerbaijan GPS measurement network was established by the Institute of
Geology, Azerbaijan Academy of Sciences, and the Massachusetts Institute of
Technology (Reilinger) based on previously existing geodynamic sites in Azerbaijan
(Guliev et al., 2002).
The network incorporates 14 GPS stations in the Azerbaijan territory with the following
conventional designations: KATE (Katekh), SHEK (Sheki), KEBE (Gabala), SAMU
(Samur), SIYE (Siazan), MEDR (Medrasa), SHIK (Shikhlar), GOSM (Gosmalion–Lerik),
YARD (Yardymly), BILE (Bilasuvar), KURD (Kurdamir), and YEVL (Evlakh), KASP
(Agdere), and SHOU (Shusha). The first measurements were performed in September
1998 at the 10 first-listed stations. Repeat measurements were carried out in September
2000 at 12 stations. Results obtained in 1998–2000 were processed in cooperation with
the Massachusetts Institute of Technology with the use of the GAMIT/GLOBK software
(Herring, 1998; King and Bock, 1998; Oral, 1994).
EEE23-5-2 30/01/06 1:29 pm Page 310

310 Deep Structure Model and Dynamics of Mud Volcanoes

RUSSIA
42

41.5 CASPIAN SEA

41

Study
40.5 area
AZERBAIJAN
40

AZ
39.5 ER
BA
IJ
AN
39

38.5

38
45 46 47 48 49 50 51

Figure 2. Horizontal GPS velocities for the 1998–2000 period and their 95%
confidence probability ellipses for the Caucasus area in the Eurasian coordinate system

Large NE-oriented velocity vectors (9–12 mm/yr) are observed in the southeastern
segment of the Lesser Caucasus (Figure 2). Analogous velocity vectors are also
observed at the GOSM and YARD stations situated in Talysh. The same velocity
vector is typical for the BILE station.
In the Greater Caucasus, intense motions are registered at the KUDI, KATE, and
SHEK stations. At the KATE station, a large velocity (12 mm/yr) is also observed, but
is directed further to the northeast. At the SHEK station, the velocity is approximately
8 mm/yr, and the vector is orientated parallel to those at the stations in the southeastern
Lesser Caucasus. Almost zero velocity vectors are observed at stations KEBE, SAMU,
SIYE, MEDR, and SHIK, suggesting that significant deformational energy is
accumulated here along the northern boundary (Figure 2).
A decrease in the velocity and a significant accumulation of elastic energy are
observed in the southern Absheron Peninsula. This phenomenon may be responsible
for the activation of seismic events and of mud volcanoes in this region, because seven
mud volcanoes erupted during the 1998–2002 period in the Absheron and
Shamakhy–Gobustan areas. The strong earthquake (approximate magnitude 6–6.3 on
the Richter scale) in the Caspian Sea at the end of 2000, and its aftershocks, probably
represent a response to the deformational processes that are ongoing, as measured in
recent years, and the related stress accumulation at the foothills of the Greater
Caucasus, the Absheron Peninsula and the middle Caspian regions. However, the
tendency of horizontal motions in the Azerbaijan territory suggests an activation of
geologic processes (seismic activity, activation of mud volcanoes and, in adjacent
EEE23-5-2 30/01/06 1:29 pm Page 311

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 311

zones, accumulation of elastic stress). Thus, it would seem that the southeastern Lesser
Caucasus and Talysh Mountains are not separate, but rather are displaced as a unified
structure from the south to the north over the viscous asthenosphere under the action
of the northward drift of the Arabian Plate. To reinforce this preliminary conclusion,
however, the number of GPS stations should be increased to obtain a more detailed
image of the recent horizontal motion distribution.

III. GEOLOGICAL AND GEOPHYSICAL DATA

a. Morphology and classification of mud volcanoes
Mud volcanoes are widely distributed throughout the world. Their total number,
known from travelers’ notes (e.g., Ma’sudu in the 10th Century, al-Garnati in the 12th
Century) and from instrumental observations over the period from 1810 through to the
present, exceeds 800. Nearly 700 occur within oil and gas bearing basins of the Earth’s
mobile belts (Gubkin and Fedorov, 1938; Yakubov et al., 1972, 1974). Mud volcanoes,
as well as magmatic volcanoes, are mostly located at boundaries of the largest
lithospheres plates and in narrow zones of Mesozoic–Cenozoic folding, which
experienced stages of the strongest tectonic activation. Such regions include zones of
subduction (where plates collide and sink) and spreading (rifting), and areas of
development of intra-continental and intra-oceanic mud volcanoes.
More than a quarter of all known mud volcanoes are concentrated within the Caucasus.
Most of the mud volcanoes (up to 220) are located within the Abikh triangle area in
Azerbaijan (Figure 1). The following conclusions can be drawn from analysis of
Azerbaijan mud volcano eruptions based on the catalog encompassing the period
1810–1974 (Yakubov et al., 1974). Generally, two or three (occasionally more) eruptions
occur every year. Nearly 200 eruptions of more than 50 volcanoes were recorded between
1810 and 1974. The Lokbatan mud volcano, located at a distance of just 15 km from Baku,
has erupted 19 times since 1828, with intervals between eruptions ranging from a few
months to more than a few tens of years. Typical activation periods of mud volcanoes are
separated by 2–5 yr and are 2–9 yr long, a roughly 50 : 50 duty cycle. In recent years, the
mud volcano activity in Azerbaijan has intensified relative to this rough average.
Morphological features of mud volcanoes (height, shapes of slopes and craters, etc.)
depend on the activity and on the composition of erupted material. Generally,
terrestrial mud volcanoes are elevated by about 10–400 m in height relative to their
surroundings, they have conical shapes, usually with flat summits, and with bases
ranging in diameter from 100 to 400 m. The shapes of the mud volcano craters vary
from flat, convex, or shield-like, to deeply sunken (of the caldera type). The crater
plateaus of large volcanoes reach about 10 km2 in area and up to several km in
diameter, and are surrounded by concentric ramparts elevated at a height of 5–25 m
relative to the central part of the plateau. The crater diameter can often exceed 50 m
(Figure 3). All these features are also observed in marine mud volcanoes, which can be
subdivided into island and submarine types (Lerche and Bagirov, 1999).

b. Geotectonic conditions of mud volcanism manifestations

The early studies of mud volcanism manifestations in various regions of the world
(Gubkin and Fedorov, 1938) showed that large mud volcanoes are located in crustal
EEE23-5-2 30/01/06 1:29 pm Page 312

312 Deep Structure Model and Dynamics of Mud Volcanoes

Lokbatan

Akhtarma-Puta
150

Gushkhana
100
Relief (m)

50

0 4000 8000 12,000

Distance (m)

Figure 3. Change of a relief along the profile crossing the mud volcanoes
Lokbatan–Akhtarma-Puta–Gushkhana
regions that experienced the most intense subsidence and that accumulated thick
(10–20 km) sedimentary sequences in the Cenozoic. The majority of mud volcanoes
concentrate along large tectonic zones of Alpine and Central Asian foldbelts, the
Pacific mobile belt, and rift zones of the Atlantic and Indian Oceans. One of the main
prerequisites to mud volcanic activity in the mobile belt crust is the presence of faults
and sedimentary complexes crumpled into folds. Gubkin and Fedorov (1938) believed
that mud volcanoes gravitate toward “anticlinal folds of the diapiric type (with a
pierced core) or to cryptodiapirs.” These peculiar anticlinal folds were, seemingly, first
observed in Romanian oil and gas bearing regions (Guliev, private communication);
they are distinct from other anticlinal folds in that underlying geological bodies seem
to cut through the upper beds during the formation of these diapiric-type folds.
Subsequently, detailed information gathered from drilled mud volcanoes provided
additional constraints on various factors and conditions associated with the activity of
these volcanoes. The occurrence of mud volcanoes in arch and periclinal parts of local
anticlinal structures complicated by faults was substantiated by deep drilling data from
several Gobustan structures (Salaev and Kastryulin, 1977). The analysis of an immense
body of comprehensive information about the mud-volcanic processes showed that a
number of conditions are indispensable for the initiation and activity of mud volcanoes
(Peive, 1956: Shnyukov et al., 1971; Yakubov et al., 1972; Rakhmanov, 1987).
First, there is the presence of local fractured anticlinal structures well expressed in
the topography. The presence of faults is the important factor. These faults not only cut
and complicate the anticlinal structures but also cut the entire sedimentary sequence.
Second, ductile clayey rocks, which are usually widespread in intense subsidence areas
of mobile belts, should be present in the section. Third, gas accumulations, and
especially fluids in rocks with an anomalously high formation pore pressure (AHFP)
often exceeding the geostatic pressure, play an important role. Under the AHFP action,
ductile clay masses (including rock fragments) are squeezed away from the section
towards the Earth’s surface (sea or ocean floor). Mud volcano feeders are usually
EEE23-5-2 30/01/06 1:29 pm Page 313

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 313

confined to the largest faults and tectonic fractures in the crust. Peive (1956)
established that deep faults control the origination and development of crustal
deformations. These deep faults are of crucial importance for the genesis and
development of the crustal structure because they cut various folded structures and
undergo long-term development with concurrent sedimentation. Formation of folds
and folded zones unrelated to deep faults are thought not to occur. Deep faults are
antecedent structures to folds and fold zones, which are genetically interrelated to the
deep faults. Geological evidence indicates that folding in active mud volcano areas
developed concurrently with active sedimentation, the pre-existing deep faults being
the main controlling factor (Gubkin and Fedorov, 1938; Yakubov et al., 1972). Mud
volcanoes are everywhere confined to longitudinal faults or to the intersection nodes
of longitudinal and transverse faults (Gorin and Buniat-Zade, 1968) as is exemplified
by mud-volcanic manifestations in the Abikh triangle. There the mud volcanoes
concentrate mostly along deep fault lines. These faults fan out from the triangle vertex
(near Shemakha) towards the South Caspian Basin. Within the Basin, active mud
volcanoes provide surface traces of the positions of deep faults. An important feature
of the variation in the mud volcanism in the western South Caspian Basin (Gorin and
Buniat-Zade, 1968: Yakubov et al., 1974) is the southward migration of the mud
volcanic activity with time. The concept of the genetic relationship between mud
volcanoes and deep faults suggests that such a regular pattern is caused by a similar
southward migration of the fault activity.

c. Deep structural characteristics and mud-volcanic activity zones

Combined geological and geophysical studies were widely conducted in oil and gas
bearing regions and mud-volcanic zones in the 1950s and 1960s. One of the results was
the detection of the so-called buried mud volcanoes (Yakubov et al., 1972;
Rakhmanov, 1987). Further investigation revealed a non-uniform pattern of seismic
wave propagation across mud volcano zones. When travelling through such a zone, the
energy of elastic waves is attenuated, reflected waves almost completely disappear,
and low frequency components dominate the wave pattern. The quality of the related
seismic records dramatically deteriorates. A buried mud volcano and its breccia neither
reflect nor refract seismic waves at the stratigraphic depth of roots of such mud
structure. Results of these works suggest that roots of mud volcanoes are related to
sediments ranging in age from Cretaceous through Pliocene. Some relatively large mud
volcanoes can pierce the thick sedimentary cover. Therefore, their roots are usually
associated with crestal zones of deep-seated Mesozoic or Paleogene–Miocene uplifts
in areas composed of 10 km and thicker sedimentary complexes. The sources of mud
volcanoes lie in clay deposits of a sedimentary complex pierced by a feeder. The
deepest zone (root) of the mud volcanic activity, where the feeder channel originates,
is located at depths of 3–8 km (Mekhtiev and Khalilov, 1987), 5–10 km (Kropotkin and
Valyaev, 1981), or 7–10 km (Rakhmanov, 1987).
Areas around Azerbaijan mud volcanoes are reflected by minima of the Bouguer
gravity field (from –120 to –40 mGal). The GPS measurements show that decrease to
zero of the horizontal movement velocity vector takes place in the neighbourhood of
the mud volcanoes (Kadirov, 2000).
EEE23-5-2 30/01/06 1:29 pm Page 314

314 Deep Structure Model and Dynamics of Mud Volcanoes

d. Relation of deep faults and waveguide zones to mud volcanoes

The majority of mud volcanoes are located in fracture (fault) zones (Gubkin and
Fedorov, 1938; Yakubov et al., 1972, 1974; Salaev and Kastryulin, 1977; Rakhmanov,
1987; Karakin, 2001). Mud volcanoes and hydrocarbon accumulations have been
related to the largest faults considered as deep faults (Peive, 1956). Such faults are
recognised from geological, geomorphological and geophysical data.

e. Mud volcanism implications for the formation of hydrocarbon accumulations

In 1848, Abikh showed that mud volcanoes indicate the presence of deep oil and gas
accumulations. Later (1903–1904), Golubyatnikov advanced a hypothesis of a direct
relation between oil and gas accumulations and mud-volcanic processes. According to
him, volcanoes can be considered as “natural deep wells.” A major contribution based
on a large body of experimental data (Gubkin and Fedorov, 1938) consists in the
correlation of mud volcanism and oil and gas accumulation with the specific formation
conditions of local anticlinal diapiric folds. The genetic relationship between mud
volcanoes and oil and gas formation was confirmed by oil field investigations, in
particular when a well drilled on the Lokbatan mud volcano in 1933 yielded an oil
gusher with a production rate of up to 20,000 t/yr. Since that time, oil geologists have
considered that mud volcanism manifestations are a direct indication of the presence
of oil and gas at depth (Nazarov, 1964: Yakubov et al., 1972, 1974; Rakhmanov,
1987). In addition, rejection was then implied of the earlier hypothesis that mud-
volcanic processes are evidence for the last stages in the destruction of oil and gas
deposits penetrated by mud volcanoes. The 1933 well drilled on the Lokbatan mud
volcano is still operating and has yielded more than 20 MMt of oil and over 1 Bm3 of
gas. The volcano has erupted 19 times and is still active, but has never disturbed oil
production.

IV. METHODS
a. Gravity and Geodetic
Measurements of gravity differences between points were undertaken by four
gravimeters under a simple closed loop arrangement. The scale interval of the
gravimeters was determined at different temperatures using a control instrument. When
processing field measurements, corrections were made for relation of scale interval to
temperature, for non-linearity of scale of a micrometer screw, and for lunar-solar
attraction.
The longitudes and latitudes of the measurement locations were determined with the
help of GPS, and altitudes of the points with a level from the firm Carl Zeiss, Jena (the
mean quadratic error on a 1 km traverse measured in both directions was 0.2 mm).
For calculation of the Bouguer gravity anomaly, the altitude was read from the
lowest level (at the Lokbatan reference point), and the interlayer density taken to be
1.82, 2, 2.3 g/cm3.
On Figures 3 and 4 relief and observed gravity fields, respectively, are shown along
the profiles. The two curves have a similar trend as can be seen by inspection of the
Figures. Gravity modelling used a minimisation condition on a multi-parameter
functional describing the least squares difference between modeled and observed
EEE23-5-2 30/01/06 1:29 pm Page 315

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 315

40

Akhtarma-Puta

Gushkhana
Gravity field (mGal)

Lokbatan

–40
0 4000 8000 12,000
Distance (m)

Figure 4. Observed change of gravity field along the profile crossing mud volcanoes
Lokbatan–Akhtarma-Puta–Gushkhana
gravity fields and involving parameters of the initial structure model. The initial
parameters of the model are modified such that the difference between observed and
computed fields does not exceed 1mGal. Along the NE–SW profile gravity modelling
was done to investigate the depth structure and tectonic evolution of the mud
volcanoes.

b. Temperature measurements
Measurements have been made at three sites within the Lokbatan crater at depths of
0.5, 1, and 1.5 m into the mud, using a thermistor thermocouple and a balancing
Wheatstone Bridge, with resistance calibrated to true temperature in the laboratory to
a precision of ± 0.01°C using a mercury thermometer. The measurements continue at
approximately 30-day intervals at the three crater sites (with exceptions for rainy
periods when no traversal of the mud is possible or when flaming outbursts occur when
the crater sites are not approachable). Similar temperature measurements, carried out
earlier, have shown high temperature gradients in the crater area (Mukhtarov and
Adigezalov, 1997).

c. Gravity and geodetic results

Along the profile Lokbatan–Akhtarma-Puta–Gushkhana, a chart of Bouguer gravity
field anomalies is compiled for various values of intermediate stratum density (1.8, 2,
2.3 g/cm3). The results obtained show that in zones of mud volcano development
(Lokbatan, Akhtarma-Puta, Gushkhana) there are local negative anomalies of –5, –3,
and –2 mGal, respectively.
For gravity modelling, the observed gravity values and the geological/geophysical
structure section along the NE–SW profile make up the initial reference information. The
initial structure model along the profile was obtained using seismic data, well
information, geological information, and the density-depth distribution of major rock
units of the study area. The initial geologic–geophysical cross-section of the sedimentary
EEE23-5-2 30/01/06 1:29 pm Page 316

316 Deep Structure Model and Dynamics of Mud Volcanoes

10 mGal
Observed gravity anomaly
Computed gravity anomaly from initial model
Selection of a configuration and density of rocks

Puta-Akhtarma
–10

Gushkhana
Lokbatan

0 2 4 6 8 10 12

0 2
N2

1
N2

2+3
N1

4 P3 + N1

P1 + P2
Depth (km)

Mz

Initial model
Selected model

12

Figure 5. Gravity model of the Lokbatan–Puta-Akhtarma–Gushkhana profile

EEE23-5-2 30/01/06 1:29 pm Page 317

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 317

cover along the profile is provided by seven contact boundaries (International Geologic
Map of Europe legend reference in brackets):
1. Boundary along the lower Akchagil ( N 22 ) ;
2. Apparent seismic horizon in the Productive Series ( N 21 ) ;
3. Boundary along the upper Kirmaki sandy suite ( N 21 ) ;
4. Boundary along the lower part of the Kirmaki suite ( N 21 ) ;
5. Boundary separating the upper-middle Miocene and lower Miocene-Oligocene
series ( P3 − N11 ) ;
6. Boundary separating Oligocene- and Eocene-Paleocene series (P1 + P2);
7. Boundary of the Mesozoic surface (Mz).
The differential density contrasts across the seven contact boundaries (from shallow to
deep) are 0.01, 0.04, 0.08, –0.2, 0.25, 0.15, and 0.3 g/cm3, respectively. The gravity
field calculated on the basis of the described model shows that there is a difference
between observed and calculated fields in zones of mud volcanoes (Figure 5). When
computing the gravity model, 10 iterations were first done on the selection of all
boundary configurations. Then the selection on density was undertaken. The extra

800 mm

07.02–12.01

600

400

10.02–12.01

200

10.02–07.02

E W
–200
0 1000 2000 3000 4000 5000
m

Figure 6. Vertical movement along profile crossing Lokbatan mud volcano

EEE23-5-2 30/01/06 1:29 pm Page 318

318 Deep Structure Model and Dynamics of Mud Volcanoes

mass in zones of mud volcanoes in the initial model is compensated for by insertion of
zones of deconsolidation and additional contact boundaries in these parts of the profile.
The deconsolidation stretches to roughly 3 km depth. A contact boundary, representing
the volcano neck (with a differential density –0.3 g/cm3), is raised in these areas to
depths of just 5 m below the surface.
The repeated geodetic leveling on the Lokbatan volcano shows that there are
currently active geodynamic processes occurring there. During the period of November
2001 through November 2002, on the NE part of volcano (of length about 2 km), the
contact boundary was as shallow as 60 cm (Figure 6).

d. Geothermal results
Thermal information prior to the explosion of October 25, 2001 comes from production
wells on the flanks of the Lokbatan mud volcano, indicating a strong focusing of heat
flux centred on the Lokbatan volcano (Figure 7) over a regional scale of kilometers
(Sukharev et al., 1969; Yakubov and Atakishiyev, 1973; Mukhtarov et al., 2002).

Figure 7. Heat flow distribution in Lokbatan mud volcano area

EEE23-5-2 30/01/06 1:29 pm Page 319

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 319

(a) II
1a
2a

3a

4a

5a

12 B 6 7
I 13 11 26 3 8 10
6a 4 9
14 1 I'
5 2 A
46

II' 15

Lokbatan profile I-I'(W–E)

(b) 70 Crater 1,5 m
65 1m
0,5 m
Temperature (˚C)

60
55
50
45
40
35
Measurements taken on 28 November 2001
30
W 0 20 40 60 80 100 120 140 160 180 200 E
Distance (m)

Lokbatan profile II-II'

(c) 70 1,5 m
Crater
65 1m
Temperature (˚C)

60 0,5 m
55
50
45
40
35
Measurements taken on 18 December 2001
30
0 50 100 150 200 250 300
Distance (m)

Figure 8. (a) Sketch of the two thermal traversal sections in EW and NS directions
over the Lokbatan region after the explosion of October 25, 2001; (b) Temperature
profiles at 0.5, 1, and 1.5 m depth across the E-W section taken on 28 November 2001;
(c) As for Figure 8b but across the N-S section and taken on 18 December 2001
EEE23-5-2 30/01/06 1:29 pm Page 320

320 Deep Structure Model and Dynamics of Mud Volcanoes

Three measurement locations (labeled A, B, and C) within the crater and on the
crater floor were marked with iron spikes for ease of relocation. At each location, the
thermal resistance probe provides three equivalent temperature measurements at 1.5, 1,
and 0.5 m into the partially dried mud sediments.
Additionally, two thermal sections were made (one E–W, the other N–S) across the
total mud field including the crater, spanning about 800 m in both directions. Both
traverses were undertaken one month (E–W) and two months (N–S), respectively, after
the explosion, with temperature measurements at 0.5, 1, and 1.5 m into the mud (Figure 8).
A discussion of the thermal information from these two sections has been given
elsewhere (Mukhtarov et al., 2002).
The thermal observations with time after the October 25, 2001 explosion, plotted in
Figure 8a–8c, are open to several interpretations. Two of the extreme possibilities are:
(1) a systematic cool-down with time after the initial explosion with no further
complicating factors; (2) a re-heating of the surficial layers after the final explosion due
to further, more minor, flame events. The available thermal data are somewhat
ambivalent in terms of either interpretation, suggesting that more detailed thermal
measurements and other types of measurements are needed to help resolve the
uncertainties (Lerche et al., 2002). A mud volcano provides a channel for heat
advection from depth. A discussion of advection from such channels has been given
elsewhere (Mukhtarov and Adigezalov, 1997; Mukhtarov, 2002).

e. Radiometric results
In addition to the geodetic, gravity and thermal measurements, total γ-ray intensity was
also measured with a sonde on the surface across a profile of about 5 km in length, with
the Lokbatan mud volcano at the center of the profile (shown by the dashed vertical
line at 3 km on Figure 9). Two days after the explosion of October 25, 2001, the γ-ray
intensity in the crater was as high as 28 µR/hr before returning over a few days to
around 10–15 µR/hr. Apart from the isolated high intensity point at the 1.5 km marker
on Figure 9 (likely due to either radioactive waste being deposited in an old well or due
to an open fault bringing up radioactive material from depth; it is not known which),
to be noted is the broad increase in total intensity in the neighbourhood of the Lokbatan
volcano, associated with the ejected mud that was transported from depth.

f. Geochemical results
Natural gases play a key role in the energetics of mud volcanic process. The source of
these huge gas volumes is generally considered to be from the sedimentary cover
(Weber, 1935; Dadashev, 1963), but other points of view would have the origin very
much deeper (Valyaev, 1985). There are also different opinions related to the problem
of stratigraphic occurrence of the gas source. One of the first attempts to reach a
decision concerning this problem was taken by Dadashev (1963), who tied mud
volcanoes gases to Paleogene deposits on the basis of the methane to homologues ratio.
Based on hydrocarbon composition, mud volcano gases belong to the group of
“dry” gases (<2% C2H6 +). Isotopic composition data of methane carbon shows that it
belongs to the catagenic gases (average isotopic composition of CH4 lying between
–40% and –50%). Studies of inert gases (He, Ar) and their isotopic ratios show that
EEE23-5-2 30/01/06 1:29 pm Page 321

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 321

30

µR/h

25

20

10.2002

15

10

5
0 1000 2000 3000 4000 5000
m

Figure 9. Change of total γ-ray count (in units of µR/h) along a profile crossing
the Lokbatan mud volcano

mud volcano gases are similar to gases of oil and gas fields, although the stratigraphic
source of the mud volcano gases are older (Guliev et al., 1996). In the last few decades
there have been theoretical investigations that allow one to determine the dependence
between isotopic composition of methane carbon (ICC), its gaseous homologues and
vitrinite reflectance (R0). For instance, Stahl (1977) notes that:

δ13C(CH4)(‰) = 17 log R0 (%) – 42

and
δ13C(C2H6)(‰) = 22.6 log R0 (%) – 32.2

Basing on these theoretical relationships one can estimate approximate depths from
which hydrocarbon gas components migrate (methane, ethane, propane, and butane).
Knowledge of the approximate hypsometric depth of the gas source allows one to
estimate the nominal stratigraphic source of the gas (by use of geological profiles).
EEE23-5-2 30/01/06 1:29 pm Page 322

322 Deep Structure Model and Dynamics of Mud Volcanoes

Ro (%)

0.1 1 10
0

2000

4000
Depth (m)

6000 Ro = 0.6%

Early oil
Ro = 0.8%
8000
Peak oil
Ro = 1%

Late oil
10,000
Ro = 1.35%
Wet gas/condensate
12,000

Figure 10. Vitrinite measurements with depth through the Lokbatan volcano
and extrapolated trend curve based on the ICC isotopic connection given in text

Results of such an assessment method, based on ICC data of ethane and also R0 for
some mud volcanoes, have been presented in the last decade (Wavrek et al., 1996). R0
values to a depth of 6 km vary in the range 0.33–0.61%; for larger depths R0 values
were defined by extrapolation (Figure 10). Vitrinite reflectance analyses were
performed on randomly oriented particles, using a conventional microphotometric
method (Stach et al., 1982).
Use of large amounts of available methane ICC data (Valyaev 1985; Dadashev,
1985) for such estimations generally has not proven fruitful because of the mixing
of catagenic (relatively heavy ICC) and biochemical (relatively light ICC) methane
that distorts the virgin values, as can be seen on the curve for ICC methane and
ethane for mud volcanoes (Figure 11). Measurements of gas maturity have been
conducted for the mud volcanoes Lokbatan, Puta-Akhtarma and Gushkhana. The
measurements have been conducted for adjacent mud volcanoes as well with the
aim to construct maps for the change of this parameter within the area. The stable
isotopic composition of carbon in hydrocarbon gases was analysed using CJS Sigma
mass spectrometers.
EEE23-5-2 30/01/06 1:29 pm Page 323

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 323

–30
δ13C2H6(‰)

–28

–26

–40 –44 –48

δ13CH4(‰)
Figure 11. Cross-plot of the δ13C variations for methane and ethane, showing that
the mixing of catagenic and biogenic components limits use of the isotopes (see text)

Ro
Bozdak-Kobu
1.85
40.4
1.80
40.38 1.75
60
1.

1.70
40.36
1.65
40.34 1.60
Shorbulak
BibiEybat 1.55
40.32
Gushkhana
1.50
1.60

Korgoz Puta Lokbatan

40.3 1.45

Garadag 1.40
40.28
49.6 49.62 49.64 49.66 49.68 49.7 49.72 49.74 49.76 49.78 49.8 49.82 1.35

Figure 12. Two-dimensional contours of maturity across the general region

encompassing the mud volcanoes

Ethane isotopic carbon composition of these gases varies within –26.2 to –29.4%. The
change in R0 within the area, calculated according to the above equations connecting
reflectance and ICC, provides an interesting picture (Figure 12). Two neighbouring
(3–4 km apart) mud volcanoes, Lokbatan and Puta-Akhtarma, are characterised by
sharply different values of maturity; maximal on volcano Lokbatan (R0 = 1.7%) and
EEE23-5-2 30/01/06 1:29 pm Page 324

324 Deep Structure Model and Dynamics of Mud Volcanoes

minimal on volcano Puta-Akhtarama (R0 = 1.35%). Thus, if the deepest gas focus is at
volcano Lokbatan (approximately at depth 11.5 km), then the depth of gas focus is
minimal on volcano Akhtarama-Puta (approximately 10 km). On the whole, however,
there is a tendency (from NE to SW) for maturity reduction and a corresponding
reduction of depth of the gas focus.
The high activity of the Lokbatan mud volcano can be explained by a larger deep
focus for the gases. This explanation supposes, however, that a rather large interval of
sedimentary section participates in the process of gas formation (i.e., the volume of gas
generation is maximal and the energy resources of gas here are also relatively large).
Support for this explanation comes from the large volumes of gas ejected during
eruptions (more than 100 MMm3), very high frequency of eruptions (approximately
every 3–4 yr), and the long period (of order a year) of gas burning after recent
eruptions.

V. DISCUSSION
Convection of the sedimentary layers due to a Rayleigh-Bernard gravitational
instability is now reviewed based upon the combined information from all the different
lines of investigation given above.
Consider a three-layer model of the deep structure of the Alpine complex of
deposits in the region, as sketched in Figure 13. This complex (from deep to shallow)
is composed of the following layers: the lower geosynclinal layer composed of
terrigeneous and carbonaceous rocks; the middle layer, lower molassic, is composed of
plastic clayey series saturated with HC; the upper layer, upper molassic, is composed
of sandy-clayey formations.
Paleogene–Miocene deposits spread throughout the subsidence zones of the SCB
serve as an example of the interstitial plastic series. The thickness of the interstitial
layer is 4–5 km on average, but sometimes reaches to 7–8 km, and has maximum
values in zones of intensive mud volcanic activity. Exploration drilling results
demonstrate that clays and clayey rocks in the SCB compose 50–95% of the section.
The viscosity of such layers varies from 106 Pa/s to 1012 Pa/s; viscosity of the overlying
and underlying formations is five to six orders of magnitude larger. The density of
sedimentary layers depends upon composition and varies as follows:
1. For clays, 2.3–2.6 g/cm3;
2. For sandstones, 2.5–2.6 g/cm3;
3. For limestone, 2.6–2.9 g/cm3. In the zones of regional deconsolidation the clayey
series is least dense, up to 0.2 g/cm3 less differentially.
The less viscous Paleogene–Miocene series of the sedimentary complex has the
following parameters: thermal expansion coefficient α = 10–5 K–1; coefficient of
thermal diffusivity χ = 5 × 10–7 m2/s; coefficient of thermal conductivity k = 2
W/(mK). The thermal conductivity coefficients of sandstones and limestones are 5
W/(mK), respectively. For this reason, the Paleogene–Miocene deposits are heat
insulators and, no doubt, they play an important role in the formation of the thermal
field in the overlying layers.
EEE23-5-2 30/01/06 1:29 pm Page 325

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 325

1 2 3

Figure 13. Sketch of the thermal fluid jet model for liquid escape from a mud volcano

Now analyse the possibility of a Rayleigh–Bernard gravitational instability in the

intermediate, less viscous, layer of the sedimentary complex. The Rayleigh number for
the viscous interstitial layer of the complex is calculated by the formula:
R = αρgqd4/(χηk) (1)
where α, thermal expansion coefficient; g, gravitational acceleration; q, thermal flow;
d, thickness of layer; ρ, density of the intermediate weakly consolidated and highly
plastic layer; χ, coefficient of thermal diffusivity; k, heat conductivity coefficient; and
η, dynamic viscosity.
According to the linear theory of stability, if the Rayleigh number is higher than a
critical value Rc, convection movement arises. The critical Rayleigh number for
EEE23-5-2 30/01/06 1:29 pm Page 326

326 Deep Structure Model and Dynamics of Mud Volcanoes

Temperature (˚C) Temperature (˚C)

10 30 50 70 10 30 50 70
0 0
4 10
9
3
500 1 8

2
1000 2

7
1
1500 3

6
2000 4
5
To = 17˚C
Tr = 70˚C 4
2500 λm = 1.35 W/(m × K) 5
Crf = 4180 J/(kg × K)
Pf = 1000 kg/m3
3000 6
Z (m) Z (m)

Figure 14. Temperature distribution in the neck of mud volcano for the thermal jet
model with different fluid uplift velocities (in units of m/s): 1) 1 × 10–10 – 5 × 10–9;
2) 5 × 10–9 – 1 × 10–9; 3) 1 × 10–9 – 5 × 10–8; 4) 5 × 10–8 – 1 × 10–8; 5) 1 × 10–8
– 5 × 10–7; 6) 5 × 10–7 – 1 × 10–7; 7) 1 × 10–7 – 5 × 10–6; 8) 5 × 10–6 – 1 × 10–6;
9) 1 × 10–6 – 5 × 10–5; 10) 1 × 10–5 – 5 × 10–5
a horizontal layer with no-slip boundaries is Rc =1300. But the actual Rayleigh number
for the Paleocene–Miocene layer with thickness d = 5 km is
R = 3 × 10–2 × 10 × 60 × (5 × 103)4 × 2.3 × 103/(5 × 10–2 × 2 × 1012) ≈ 104
which is almost an order of magnitude higher than the critical value for the
Paleogene–Miocene layer. Accordingly, convective movements take place in this layer
having the characteristics of two-dimensional cells. The typical time for temperature
equilibrium of a liquid layer is determined by
τ = d2/Nχ (2)
where N is the Nusselt number, which characterises the efficiency of the convectional
heat transfer and shows how much more intense the convective thermal flow is
compared to the conductive flow. Figure 14 shows a schematic picture of such a thin
tube model of mudflow from depth, while Figure 15 shows the corresponding
temperature profiles with depth (on both a large scale of about 3 km and also a small
scale of a few meters from the surface) for different values of the mud flow speed. Note
the rapidity, spatially, with which the temperature reaches steady-state in both panels.
EEE23-5-2 30/01/06 1:29 pm Page 327

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 327

T
3

3
1

0
1 2 3 z

Figure 15. Sketch of the one-dimensional temperature profile along the thermal jet
of Figure 13 for three cases of advective flow. Note the “flat” region representative
of “mixing” due to dominance of convective over conductive heat transfer

For the viscous interstitial layer, the convection time is much less than the age of the
Paleogene–Miocene complex. Hence, convection in this layer is stationary. After the
establishment of convection, the temperature of the Paleogene–Miocene layer becomes
stable as well (Figure 16). The length of the disturbance wave, which characterises the
distance between two close ascending flows at fixed temperatures in the boundary of the
Paleogene–Miocene series, is calculated by the formula λ ≈ 2d. If the stresses created by
the flow exceed the strength limit of the overlying rocks, then uplift occurs above the
ascending flows (Trubitysin et al., 1998). This Rayleigh–Bernard gravitational instability
reflects one of the possible models of hydrocarbon migration in the Paleogene–Miocene
series in the SCB (Kadirov and Kadyrov, 1990; Guliev and Kadirov, 2000).
Transportation of hydrocarbons upward, together with the enclosing clayey plastic
mass of the intermediate layer, under convective processes would appear to be a
dominant mechanism of migration and accumulation in the upper parts of the series,
with further breakthrough of the overlying permeable series. As the clay is raised, the
lithostatic pressure lessens, with the result that hydrocarbon phase transitions will
occur and hydrocarbons will be in free phase.
It is quite possible that the interstitial layer is the main generator of most of the mud
volcanoes in Azerbaijan (Guliev and Kadirov, 2000). Figure 13 demonstrates an activation
scheme for mud volcanoes in a three-layer model of the sedimentary complex with the
EEE23-5-2 30/01/06 1:29 pm Page 328

328 Deep Structure Model and Dynamics of Mud Volcanoes

z
3.0

0.5 0.5

2.5
1
1

1.5

2.0

1.4
1.6
1.5

1.5

1.6 1.4

1.0 1.5

2
2

0.5

2.5 2.5

0
0.2 0.4 0.6 0.8 1.0 x

Figure 16. Two-dimensional steady-state temperature distribution for the thermal jet
model shown in Figure 13. Note the broad central region where the temperature
profile is effectively independent of depth to a much greater extent than conductive
heat flow would permit

application of convection. The underlying more competent layer is typically thought to be

of Eocene age, the convective cell motion occurs in the Paleogene–Miocene “sloppy”
sediments, while breakthrough of the thin jet material to the surface is estimated to occur
through the base of the Paleogene Tarkhan–Chokrak formation.
From December 2001 through June 2002, the following sequence of events took
place. The surface rises in and around the Lokbatan volcano. However, the amplitude of
EEE23-5-2 30/01/06 1:29 pm Page 329

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 329

rise varies strongly along the structure. To the east, measured from the neck of the
volcano, out to a distance of about 1000 m, the surface rise is 677 mm. A secondary
maximum of 218 mm is observed on the western part to a distance of about 1000 m from
the neck. The area of the crater also participates in the general rise, but lags appreciably
behind the rise of the regional sites. Geodetic observations from December 2001 through
October 2002 show that, in the period June–October 2002, to the east away from the
crater there is a relative lowering of the surface, while to the west and also in the crater
structure, the surface shows a relative rise. Comparison of the change of amplitude of
movement of the surface with the change of radioactivity along the profile, indicates
increased radioactivity in the crater and to the western side, where the differential uplift
is largest. An interpretation for this correlation is enhanced flow of liquid material as a
result of microfracture growth in and around the site of the largest uplift gradients.

VI. CONCLUSIONS
The results show that there are negative local gravity anomalies (–5, –3, and –2 mGal),
respectively, in zones of mud volcanoes. The calculated field based on the model
described shows that in zones of mud volcanoes there is a difference between observed
and calculated fields. The extra mass in mud volcano zones is compensated for by
introduction of decompaction zones and additional contact boundaries.
Clearly, gas must still be upwelling from the Lokbatan volcano to feed the flame
(currently around 0.5–1 m high) over the last 14 months, so that the thermal regime
cannot be one of simple cooling after a single flaming explosion. By mid-December
(December 18, 2001) the mud had cooled sufficiently that it was possible to complete
the N–S thermal section. The crater measurements show elevated temperatures of 47°C
(0.5 m), 53°C (1 m), and 58°C (1.5 m) with a systematic cooling within about 100 m
of the crater, indicating either thermal cooling of the ejected mud or rain enhanced
cooling by infiltration. The background temperature of 35.3°C had then been reached.
Precisely how this continued gas supply is related to geodynamic models of mud
flow from depth (6 km) in the Lokbatan mud volcano, and precisely how the dynamics
and recharging of the gas and mud flow operate in a roughly cyclic manner, need
further and more detailed observations than are presently available. Isotope gas
composition shows that gas of the Lokbatan mud volcano has very deep roots as
compared to adjacent volcanoes; such deep roots evidently account for its higher
activity. Focuses of mud and gas flow formations are different, with the mud focus
being located at a relatively shallow depth compared to that for gas.

ACKNOWLEDGEMENTS
The authors gratefully thank the US CRDF for financial support under grant NG2-2285 that
enabled this work to come to fruition. We also thank the Director of the Geological Institute
of Azerbaijan, Academician Akif Ali-zadeh for his unstinting support and encouragement.

REFERENCES
Dadashev, A. A. (1985). Peculiarities of carbon isotope composition of hydrocarbons
on the western flank of the South Caspian Basin. Abstract of Ph.D dissertation.
Moscow. VNIIYAG. 25p.
Dadashev, F. G. (1963). Hydrocarbon gases of Azerbaijan mud volcanoes. Baku.
Azerneshr. 66p. (in Russian).
EEE23-5-2 30/01/06 1:29 pm Page 330

330 Deep Structure Model and Dynamics of Mud Volcanoes

Gorin, V. A. and Z. A. Buniat-Zade. (1968). Deep faulting belts of the crust and the oil
and gas volcanism. Izv. Akad. Nauk Azerb. SSR, Ser. Nauk Zemle, no. 2, 11–17.
(in Russian).
Gubkin, I. M. and S. F. Fedorov. (1938). Gryazevye vulkany Sovetskogo Soyuza i ikh
svyaz’s genezisom neftyanykh mestorozhdenii Krymsko-Kavkazskoi geologic-
heskoi provintsii (Mud volcanoes of the Soviet Union and their implications for
the genesis of oil fields in the Crimea-Caucasus geological province). Moscow,
AN SSSR.
Guliev, I., A. R. Ismet and P. S. Djafarova. (1996). Inert components in gases of
deposits of Guneshly, Azeri, and Chirag. Azerbaijan Oil Industry. no 9, 7–11.
Guliev, I. S. and F. A. Kadirov. (2000). A mechanism of intrastratal migration of
hydrocarbons. Transactions (Doklady) of the Russian Academy of Sciences/Earth
Science Section. 373, no. 6, August-September, 941–944.
Guliev, I. S., F. A. Kadirov, R. E. Reilinger, R. I. Gasanov and A. R. Mamedov. (2002).
Active tectonics in Azerbaijan based on geodetic, gravimetric, and seismic data;
Transactions (Doklady) of the Russian Academy of Sciences/Earth Science
Section. 383, no. 2, 174–177.
Herring, T. A. (1998). GLOBK: Global Kalman Filter VLBI and GPS Analysis
Program. Version 4.1, Cambridge: Mass. Inst. Technol.
Ivanov, S. N. (1970). The ultimate depth of open cracks and the hydrodynamic
regionalization of the crust. Ezhegodnik. Osnovnye rezul’taty rabot 1969 g
(Yearbook. Main Results of Research in 1969), Sverdlovsk: Inst. Geol. Geochem.
UO AN SSSR, 212–233.
Kadirov, F. A. and A. H. Kadyrov. (1990). The possibility of the thermal convection in
the sedimentary layers of Azerbaijan. Revue Academy of Sciences of Azerbaijan,
Earth Sciences. Publishing House ELM, Baku (in Russian).
Kadirov, F. A. (2000). Gravity field and models of deep structure of Azerbaijan. Baku,
Publishers, ‘Nafta-Press’, 112p. (in Russian).
Karakin, A. V., S. A. Karakin and G. N. Kambarova. (2001). Geological-geophysical
analysis of mud volcanism regions. Izvestia, Physics of the Solid Earth 37, no 8,
684–692, MAIK “Nauka/Interperiodika”.
King, R. W. and Y. Bock. (1998). Documentation for the GAMIT analysis software.
Release 9.7, Cambridge: Mass. Inst. Technol.
Kropotkin, P. N. and B. M. Valyaev. (1981). Geodynamics of the mud volcanism activity
in relation to the oil and gas content / geological and geochemical foundations of
the prospecting for oil and gas. Kiev: Naukova Dumka, 249p. (in Russian).
Lerche, I. and E. Bagirov, (1999). Impact of natural hazards on oil and gas extraction.
Kluwer Academic/Plenum Publishers, New York, 353p.
Lerche, I., A. Sh. Mukhtarov, F. A. Kadirov and A. A. Feyzullayev. (2002). Thermal
Measurements at Lokbatan Mud Volcano: Changes after Explosion of 25 October
2001, Azerbaijan Geology (in press).
Mekhtiev, Sh. F. and E. N. Khalilov. (1987). Volcanoes and geodynamics, Priroda, no.
5, 47–49.
EEE23-5-2 30/01/06 1:29 pm Page 331

ENERGY EXPLORATION & EXPLOITATION · Volume 23 · Number 5 · 2005 331

Mukhtarov, A. Sh. (2002). Energy transfer and thermal regime in the mud volcanoes
channel. Energy Explor. Exploit. (in press).
Mukhtarov, A. Sh. and N. Z. Adigezalov. (1997). Thermal regime of mud volcanoes in
eastern Azerbaijan. Proceedings of Geology Institute 26, 221–228 (in Russian).
Mukhtarov, A. Sh., F. A. Kadirov, I. S. Guliyev and I. Lerche. (2002). Temperature
evolution in the Lokbatan mud volcano crater (Azerbaijan) after the eruption of
25 October 2001. Energy Explor. Exploit. 21, 187–208.
Oral, B. (1994). Global Positioning System (GPS) measurements in Turkey (1988–1992):
kinematics of the Africa–Arabia–Eurasia Plate collision zone, PhD. Thesis,
Cambridge: Mass. Inst. Technol.
Peive, A.V. (1956). Sedimentation, folding, magmatism, and mineral deposits in
relation to deep faults. Izv. Akad. Nauk SSSR, Ser. Geol., no. 3, 57–71.
Rakhmanov, R. R. (1987). Mud volcanoes and their implications for the oil and gas
content of deep formations, Moscow: Nedra.
Salaev, G. G. and N. S. Kastryulin. (1977). Implications of tectonic fractures for the
formation of oil and gas accumulations in Kobystan, Baku, Elm (in Russian).
Stach, E., M. T. Mackowsky, G. H. Taylor, D. Chandra, M. Teichmuller and R.
Teichmuller. (1982). Coal Petrology, 2nd Edition, Elsevier, Berlin, 428p.
Stahl, W. J. (1977). Carbon and nitrogen isotopes in hydrocarbon research and
exploration. Chem. Geol. 20, 121–149.
Sukharev, G. M., Yu. K. Taranukha and S. P. Vlasova. (1969). Heat flow from
Azerbaijan. Soviet Geology 8, 146–153. (in Russian).
Trubitsyn, V. P., F. A. Kadirov and A. H. Kadyrov. (1998). Influence of the thermal
convection in a viscous intermediate layer on distribution of temperature of a
multilayer sedimentary complex, Revue Academy of Sciences of Azerbaijan.
Earth sciences. Publishing House of Azerbaijan Academy of Sciences, Baku, no.
1. pp. 64–68 (in Russian).
Trubitsyn, V. P., O. E. Zasepina and A. G. Kadyrov. (1992). Thermal evolution of the
layered environment with an intermediate viscous layer, Izvestiya AS Russia.
Physics of the Solid Earth 8, 3–11 (in Russian).
Valyaev, B. M. (1985). Isotope image of the mud volcano gases. Moscow, Nauka,
pp. 72–78.
Vassoevich, H. B. (1986). Geochemistry of organic matter and origin of oil, Selected
Proceedings. Moscow, Nauka 336p.
Weber, V. V. (1939). Natural gases of USSR. Moscow. ONTI. 212p.
Wavrek D., J. Collister, D. Curtiss, J. Quick, I. Guliev and A. Feyzullayev. (1996).
Novel application of geochemical inversion to derive generation/expulsion
kinetic parameters for the South Caspian Petroleum Systems, Azerbaijan.
AAPG/ASPG Research Symposium “Oil and Gas Petroleum Systems in Rapidly
Subsiding Basins” October 6–9, Baku, Azerbaijan.
Yakubov, A. A., A. A. Ali-Zade and R. R. Rakhmanov. (1974). Catalog of recorded
eruptions of mud volcanoes in Azerbaijan over the period from 1810 through
1974, Baku, (in Russian).
EEE23-5-2 30/01/06 1:29 pm Page 332

332 Deep Structure Model and Dynamics of Mud Volcanoes

Yakubov, A. A., A. A. Ali-Zade and M. M. Zeynalov (eds.). (1971). Mud volcanoes of

the Azerbaijan SSR. Atlas. Academy of Sciences of the Azerbaijan SSR, Baku
(in Russian).
Yakubov, A. A. and I. S. Atakishiyev. (1973). Geothermal researches of oil-gas fields
of Apsheron, Baku: Azerneshr. 88p. (in Russian).
Yakubov, A. A., V. A. Gorin and Z. A. Buniat-Zade. (1972). Mud volcanism, geologiya
SSSR. T. 47. Azerb. SSR (Geology of the USSR. Vol. 47: Azerbaijan SSR),
Moscow, Nedra, pp. 390–403 (in Russian).