Hydro and geothermal model of Mačva

Mića Martinović, Mihailo Milivojević
Faculty of Mining and Geology, Institute of Hydrogeology, Laboratory for Geothermal Energy


Mačva is a large alluvial plain in Serbia between the Drina and Sava rivers, located about 80 km west of Belgrade. From a geotectonic point of view, Mačva is located on the southern edge of the Pannonian Basin, where it joins the Dinarides.

The Mačva hydrogeothermal system was discovered in 1982. when a large conductive geothermal anomaly was discovered in the Neogene sediments in Dublje, i.e. in the central part of Mačva (Milivojević et al., 1982). At that time, hydrogeothermal research began, but it has not yet been completed. Their results show that the low-temperature convective hydrogeothermal system “Mačva” is part of a large regional same system, which extends below Mačva, Semberija and Srem on about 2000 km2.

The results obtained so far are very interesting. Beneath the Neogene sediments is a karst reservoir in Triassic limestones from which intensive exploitation of geothermal energy is possible for heating settlements, food production and for use in industry. Above the reservoir in the central part of Mačva, the most intense conductive geothermal anomaly in the entire Pannonian basin was discovered, because thermal water with a temperature of 75oC was discovered in well BB-1 at a depth of 412 m (Milivojević et al., 1987). That is why Mačva represents the Yugoslav and Serbian “Red Spot” like the Pannonian Basin in Europe (Horvath, Bodri et al., 1979).

The results of preliminary tests performed in exploratory wells, as well as the results obtained by hydrogeothermal modeling show that the exploitation of geothermal energy with a thermal power of at least 150 MW is possible in Bogatić, but its use has not yet begun.

This paper presents the main characteristics of the Mačva hydrogeothermal system based on the results of previous research.

Geological composition of Mačva

In the paleo-relief of the Pannonian Basin, there are many shallow and deep depressions filled with Neogene and Quaternary sediments. Above one of them is the Sword. According to the results of regional geophysical investigations, the deepest part of the depression is in its southern part, where the maximum thickness of Neogene and Quaternary sediments is about 1500 m (Mladenović, 1968). The minimum thickness of Neogene and Quaternary sediments is about 200 m and was determined by drilling in the central part of Mačva. Neogene and Quaternary sediments are composed of gravels, sands and clays, and they alternate.

The paleo-relief of Neogene sediments in Mačva was only discovered in 1981. year, when the first exploratory geothermal well DB-1 was made (picture 2). At its location, it was determined that the paleo-relief consists of highly karstified Middle Triassic limestones with a thickness of more than 200 m. Karstified limestones of the Middle and Upper Triassic were found in wells BB-1 and BB-2 in Bogatić, and Triassic dolomites in well BeB-1 in Belotić. The borehole MeB-1 in Metković did not enter the paleo-relief, but ended in Neogene sediments.

The exploratory geothermal well BZ-2 is the deepest in Mačva. Its depth is 1500 m. At its location, the thickness of Quaternary and Neogene sediments is 287 m. The paleo-relief consists of thermometamorphosed sandstones and siltstones of unknown age and plagiogranites of Neogene age. The age of plagiogranite according to the K/Ar method is about 35 million years (Pezskei, 1992). The discovery of these rocks confirmed the earlier assumption about their presence in the paleo-relief of Neogene sediments in the area of Mačva (Milivojević et al., 1984). In addition to these igneous rocks, a series of ingimbrites with a thickness of about 50 m was discovered in the BB-2 well. Their K/Ar age is about 30 million years (Pezskai, 1991).

Triassic carbonate sedimentary rocks are of alpine type of development. therefore, their thickness can be up to 1000 m, as in the Dinarides.

The nearest occurrences of paleo-relief rocks on the surface of the terrain are in the area of Mount Cer and near the town of Šabac (Figure 2). In the area of Mount Cer, metamorphic rocks of Devonian-Carboniferous age, Permian limestones, Triassic and Cretaceous limestones are present. All these rocks have a periclinal slope towards the north, which is a consequence of the depression and uplift of the Cera granitoid pluton. The area of the granitoid pluton on the surface of the terrain is about 40 km2. Magmatic activity in the Cera area was carried out in several phases so that the K/Ar age of the granitoid and its vein rocks is 7-17 million years (Milivojević et al., 1986; Milivojević, 1992).

At Šabac, rocks from the pelorelief of Neogene sediments were discovered on a small area. they are made of limestones and sandstones of the Lower Triassic.

The main characteristics of the hydrogeothermal system

In Mačva, the thickness of the earth’s crust is the smallest in the territory of the former and current Yugoslavia and is 25-26 km. Similar values are also found in the area of Semberija and Srem (Dragašević, Andrić et al., 1990). In its composition, the granite layer has the greatest thickness, about 17 km (Roksandić, 1974).

The density of the terrestrial heat flow under the “sedimentary layer” in the area of Mačva is very high. It is 112 mW/m2 in well BŠ-1 in Šabac, and 120 mW/m2 in well BZ-2 in Bogatić (Milivojević, 1989).

The value of the density of the heat flow reaching the earth’s crust from the upper mantle in the area of Mačva is 55-60 mW/m2, and the temperature at the Mohorovičić discontinuity is about 1000 oC. The thickness of the lithosphere in the area of Mačva determined by the geothermal model is about 40 km (Milivojević, 1992).

Mačva is located in the area of the geothermal anomaly of Serbia (Milivojević, 1990), which is the southern part of the geothermal anomaly of the Pannonian Basin (Bodri et Bodri, 1982). Therefore, the conductive and convective geothermal anomaly of Mačva is a consequence of very high regional values of the density of the terrestrial heat flow in the earth’s crust of this part of the Pannonian Basin.

The reservoir in the Mačva hydrogeothermal system is represented by highly karstified Triassic limestones. This was determined in wells DB-1, BB-1 and BB-2. Its exact thickness is not known. According to geological data, it is a minimum of about 500 ma, according to geophysical data, it is a maximum of about 1000 m (Milivojević et al., 1987).

The backwash insulator over the reservoir consists of Neogene sediments with a thickness of 200 mu in Dublje to 620 mu in Bogatić. The temperature at the top of the tank is 35-78 oC. According to hydrogeothermometers and mixing models, the maximum temperature value in the reservoir should be around 100 oC (Milivojević, 1989; Gorgieva, 1989).

High temperatures at the top of the reservoir are the cause of very high values of conductive heat flow density and temperature in Neogene sediments. The heat flux density in them is 140-270 mW/m2. That is why the water temperature in the alluvial sediments is anomalously high and amounts to 14-16 oC. In other words, convection in the reservoir created a large conductive geothermal anomaly in the Neogene sediments of Mačva. On the other hand, anomalously high temperatures in Neogene sediments, if an inverse geothermal model is applied, are the main indicator of the presence of a limestone reservoir and high temperatures in it. Therefore, all wells that discovered the reservoir were drilled next to artesian wells with potable water from Neogene sediments in which geothermal gradients are greater than gradients at locations in Macva where no reservoir is present. In other words, geothermal gradient values greater than 0.07 oC/m are a sure indicator of the presence of limestone reservoirs with a high temperature in their substrate (Milivojevi et al., 1984). In this regard, north of Bogatić towards S.Mitrovica at the top of the reservoir, according to existing indications, temperatures should be around 90 oC.

Replenishment of the reservoir of the hydrogeothermal system “Mačva” with water is carried out by direct infiltration of precipitation along the northern rim of Mount Cer where Permian, Triassic and Cretaceous limestones have been discovered, indirectly through a thin layer of sand and gravel where its thickness is small, by direct immersion of the waters of the Drina river near Koviljače and Tavna rivers into the Triassic limestones in their river beds, and by the inflow of thermal waters from the deep parts of the system from the areas of Semberija and Srem. These hypotheses were mostly proven by isotopic tests.

Thermal waters from wells BB-1 and BeB-1 do not contain tritium, have the same deuterium content and differ only in oxygen content (18O) according to their temperatures. The thermal water from well DB-1 has a much poorer isotopic composition than the water from wells BB-1 and BeB-1 and contains 13.5 TU. -767 m. These data indicate that part of the thermal waters in the southern part of the reservoir is younger than 30 years and originates from directly infiltrated precipitation and river waters. The waters in the northern part of the reservoir originate from thermal waters over 50 years old that entered the “Mačva” system from the area of Semberija and Srem. In other words, the thermal waters in the reservoir of the “Mačva” hydrogeothermal system are a mixture of young and old waters from different recharge areas. All these data together with the data of the chemical composition indicate an active flow of water through the reservoir.

Replenishment of the reservoir with geothermal energy is carried out by conduction through its base made of shale of Paleozoic age.

The karstification of the upper part of the Triassic limestone reservoir is very large. Caverns with a diameter of 0.5-17 m were discovered in them by drilling. Therefore, preliminary hydrodynamic tests of small-diameter exploratory wells show that transmissibility values are very high for this type, from 5 x 10-3 to 1 x 10-2 m2/s. (Martinović, 1991). Yields of exploratory wells and other data on them are shown in table 1.

Table 1. The main characteristics of the conducted exploration wells

The well Location The depth of the well in reserve. The length of the well in the reserve. Diameter Generosity Temperature Pressure
Rich man
Rich man

The water-receiving parts of the exploratory wells in the reservoir are open, i.e. no filters are installed, the so-called “open hole”. That is why the inflow of thermal water from the caverns is mainly through the bottom of the wells.

Thermal waters have a low total mineralization, which is below 1 g/l. Their quality is approximately equal to the quality of drinking water. The main components of the chemical composition are shown in Table 2. Thermal waters from all wells belong to the Na-HCO3 type.

Table 2.

Chemical composition of thermal waters from Triassic limestones and dolomites
The well T pH/T On the K Approx Mg HCO3 SO4 Cl F SiO2 TDS

Hydrogeothermal model of Mačva

The assessment of the possibility of exploitation was carried out on the hydrogeothermal model, which is expressed on the basis of the geological, hydrogeological and geothermal model of the terrain. The geological model of the terrain included the area of the base of tertiary sediments in which limestone rock masses of Triassic (and Cretaceous) age have been established, or are assumed to be present, as the main reservoir of the hydrogeothermal convective system. Such a geological, i.e. the first or conceptual hydrogeothermal model was created by Milivojević & Perić (1986). In that first model, which included a large territory of about 6,000 km2, the area of Mačva is one part of it together with Semberija, Srem and Posavo-Tamnava. Such a conceptual three-dimensional model of that large hydrogeothermal system was later supplemented with new data on the hydrogeothermal characteristics of the terrain in a static and dynamic sense so that it could serve as a basis for creating a mathematical model. It was all necessary to do this in order to simulate the behavior of the hydrogeothermal site in natural conditions on the complex preliminary hydrogeothermal model obtained in this way, and then during some given future intensive exploitation.

The results of geological research obtained during the development of the OGK SFRJ, the results of oil and gas research, the results of geophysical research, as well as the results of dedicated hydrogeothermal research were used to determine the boundaries of the model. Based on their complex analysis, the boundaries of the model were defined, i.e. the limits of the possible distribution of the unique accumulation of thermal waters within the reservoir of, mainly, Triassic limestones and dolomites, and to a lesser extent from Cretaceous shale.

The southern border of the model consists of Majevica, Gučevo, Cer and Vlašić. The northern border in a narrower sense extends to the Sava River, and in a wider sense to Fruška Gora. In the west, the model extends to the SW slopes of the Majevica mountain and the Tinja river, and in the east to Obrenovac, Belgrade and the Danube. All model boundaries are defined as watertight and are far enough from the exploitation area, so their influence is negligible.

The model covers an area of about 3,000 km2, which is discretized in the mathematical model with 4,000 triangular elements. The size of the triangles, ie. the density of the network varies depending on the expected hydraulic gradients. The network is denser in areas of future exploitation, i.e. in areas of steep gradients.

In the first stage of creating the hydrogeothermal model, the initial piezometer levels within the model were reconstructed in natural conditions, i.e. a model of the natural state was expressed, which in the second phase served as an initial model for simulating its behavior during exploitation. For this purpose, the results obtained at all the exploratory hydrogeothermal wells in Mačva, Posavo-Tamnava, Srem and Semberija were used, in which the piezometer levels were measured in different temperature conditions: during extraction, i.e. pumping, immediately after closing the well or several months after closing. All these data on the measured piezometer levels should have been reduced to measurements in the same temperature conditions, i.e. eliminate the influence of thermolift, which occurs as a result of the increase in temperature in the well column, and manifests itself through an increase in artesian pressure. The greatest impact of the thermal lift was recorded at the BB-2 well in Bogatic and amounts to more than 7 m of water column. Therefore, the measured piezometer levels were reduced to the stationary state of measurement, i.e. the influence of thermolift has been eliminated. The PREDIP program (Bjornason, 1994) was used for these calculations.

The thermal logging measurements performed in the exploration wells were used to create a map of the temperature distribution in the model area, which preliminarily defined the temperature characteristics of the reservoir.

Hydrogeological parameters were determined in several ways, depending on the data available for individual parts of the model. Relatively the most representative data were obtained by the interpretation of pumping tests from exploratory wells (Bogatič, Dublje, Belotič, Debrc, Kupinovo). Hydrogeological parameters in areas for which there were no exact data were determined by interpolation and extrapolation, mapping values from areas with similar hydrogeological characteristics, as well as statistical methods.

The estimated permeability (transmissibility) value for the entire area of the model ranges from 4*10-4 to 6*10-3 m2/s, which corresponds to the values obtained by the interpretation of the pumping tests.

The specific yield is in the range of 5*10-4 to 5*10-3. For the area of Mačva, the average permeability was taken as a value of 5*10-3 m2/s, and a value of 4*10-3 was taken for the specific yield.

Modeling methodology and exploitation simulation parameters

The creation, taring and verification of the mathematical hydrogeothermal model was carried out in two phases. In the first phase, the behavior of the reservoir in its natural state was simulated on the model. The goal of this simulation is to obtain the piezometer levels of the natural state issued and their reconciliation with the measured data. The model of the natural state of the reservoir, as a result of the first phase, served as the initial model during the creation of the model of the second phase, i.e. simulations of reservoir behavior during exploitation.

In order to verify the model, it was tared based on the data of the pumping test performed on the research well BB-2, with the maximum amount of self-flow (60 l/s) for 90 days. During the testing, measurements of changes in the piezometric level were carried out at the exploratory wells BB-1 in Bogatić and DB-1 in Dublje.

Satisfactory agreement was achieved between the measured and piezometric levels obtained on the mathematical model, so we believe that this mathematical, hydrogeothermal model can be used to simulate the behavior issued in various exploitation conditions.

The second phase of the model began with the location of future geothermal sources, where the area of Bogatić was determined as one of the possible ones, because the possibility of total heating of Bogatić in the winter period and the use of thermal waters for industrial and other purposes in the summer period was analyzed as a concrete problem. In this regard, the results of the calculation of the required amount of thermal water served as the starting point for determining the exploitation quantities, so that 300 l/s of thermal water at a temperature of 80 oC was entered into the model calculations and analyses.

The microlocations of the future exploitation wells (three in total) were determined on the basis of research carried out so far and are located in the area of exploration wells BB-1 and BB-2. It is planned that the exploitation of thermal waters in the amount of 300 l/s for heating purposes and for industrial purposes is carried out continuously throughout the year. The simulation was performed for a period of 30 years.

A specific capacity of 100 l/s does not mean that it will actually be achieved per facility, i.e. exploitation well, but leaves the possibility for it to be realized in a given microlocation from two or three wells depending on their individual capabilities, i.e. quality of workmanship.

The results of the simulation based on the mathematical model show that the maximum lowering after 30 years of exploitation will be from 50 to 60 m from the initial piezometric level or 35 to 40 meters from the ground surface.

The lowering of the piezometric level during thirty years of exploitation shows that during the first two years, 80% of the lowering would be achieved, and then the piezometric level would show a tendency to decrease at a rate of 1 m per year.

The results of a thirty-year simulation show that it is possible to exploit 300 l/s of thermal water in the Mačva area with the use of well pumps, and as such can serve for the development of a techno-economic analysis of the profitability of the Bogatić heating solution using thermal water.

Complete previous research and test results on the described preliminary hydrogeothermal model show and prove that in the area of Mačva there are conditions for the exploitation of thermal waters, i.e. geothermal energy in quantities that enable its total heating and many other complex uses of it in the city. Therefore, the implementation of detailed hydrogeothermal research in the area of the city and Semberija should be started in order to create a detailed hydrogeothermal model on a micro level.

The presented results were obtained for the case of exploitation with individual independent wells in which only mass-thermal water and heat energy are taken from the reservoir. This variant was chosen as pessimistic in order to see the behavior of the reservoir in such unfavorable exploitation conditions with a thermal power of 150 MWt. This type of exploitation with irreversible pumping of fluids is less and less used in the world, because it has proven to be irrational due to the fact that with it the utilization of the reservoir is only 10%.

Since according to the given model the tank passed the test, the given exploitation capacity of 300 l/s of thermal water can be considered as minimal. Therefore, in the following analyzes that will be carried out after obtaining the first exact results of the Triassic limestone reservoir data, as should be the results of testing future exploitation wells of large diameter, exploitation with reinjection “double” systems should be simulated, because with them the utilization of the reservoir from 25-35%.

Possibilities of intensive exploitation and use of geothermal energy

The assessment of the exploitation reserves of thermal waters and geothermal energy that can be exploited from the reservoir of the hydrogeothermal system “Mačva” has not yet been completed. In our opinion, it is possible to exploit a minimum of 300 l/s and a maximum of 1500 l/s of thermal water with a temperature of 75 oC or with a thermal power of a minimum of 150 MW or 500 MW from the limestone and dolomite reservoir in the Mačva area. This forecast of ours is expressed on the basis of the expanse of the reservoir of about 800 km2, its great thickness and its extraordinary hydrodynamic characteristics. By constructing a well with a large diameter, a pressure drop in the reservoir can be achieved in the entire area of Mačva. In this way, the effect of renewing geothermal energy in the reservoir will come to the fore. In other words, the terrestrial heat flow annually brings about 2500 x 1012 W of geothermal energy into the reservoir, and with the exploitation of 150 MW during the heating period (without reinjection), about 2300 x 1012 W of geothermal energy is brought out.

Mačva is one of the most famous agricultural areas in Serbia and Yugoslavia. Therefore, geothermal resources from its karst convective hydrogeothermal system are of great importance for food production, heating of settlements and in the agricultural industry.

Future activities

Previous research has not determined the extent of the reservoir or its thickness in detail. The extent of the geothermal anomaly has also not been determined. Therefore, the main goal of future research should be the definition of these parameters, i.e. defining the regional geothermal model in three dimensions. At the same time, next to the exploratory well BB-2, an exploratory well with a large diameter up to a depth of 1000 m should be made, and then hydrodynamic tests should be carried out. Based on the results of these tests, a detailed hydrogeothermal computer model should be created and the required parameters of the reservoir and the most optimal method and mode of exploitation of thermal waters and geothermal energy should be determined. Only after that can trial exploitation and construction of a system for the use of geothermal energy in the agricultural industry begin.


A conductive geothermal anomaly in Neogene sediments in the central part of Mačva, and below it a hydrogeothermal convective anomaly in highly karstified Triassic limestones are the largest such anomalies in the Pannonian Basin. Current forecasts obtained on the basis of the hydrogeothermal model show that intensive exploitation and use of geothermal energy for the production of food and flowers and application in the agricultural industry with a thermal power of at least 150 MW is possible.


— Bodri, L. and Bodri, B., 1982. Geothermal model of the heat anomaly of the Pannonian Basin. In: V. Cermak and R. Haenel (Editors), Geothermics and geothermal energy. E. Schweizerbartsche Verlagsbuchhandlung, Stuttgart.
— Dragašević, T., Andrić, B. and Joksović, P., 1990. Structural Map of Mohorovičić discontinuity of Yugoslavia. Scale 1:500 000. Fed. Geol. Inst., Belgrade.
— Horvath, F., Bodri, L. and Ottlik, P., 1979. Geothermics of Hungary and the Tectonophysics of the Pannonian Basin “Red Spot”. In: V. Cermak and L. Rybach (Editors), Terrestrial Heat Flow in Europe. Springer Verlag, Berlin.
— Milivojević, M., Perić, J. and Pavlović, P., 1982. The newly discovered geothermal anomaly in Dublje-Mačva and its energy potential. Collection of reports 7. of the Yugoslav symposium on hydrogeology and engineering geology, Novi Sad 1982, vol. 1, Hidrogeologija, Belgrade.
— Milivojević, M. and Perić, J., 1984. The results of the investigation of the geothermal anomaly in Dublje and their significance for the assessment of the hydrogeothermal potential of Mačva and Posavo-Tamnava. Collection of reports 8. of the Yugoslav symposium on hydrogeology and engineering geology, Budva 1984, vol. 4, Hidrogeologija, Belgrade.
— Milivojević, M. and Perić, J., 1986. Geothermal potential of Mačva, Semberija and Srem. Proceedings of the XI Congress of Geologists of Yugoslavia, vol. 1, Tara, Belgrade.
— Milivojević, M. and Perić, J., 1987. Energy potential of hydrogeothermal resources of Mačva. In: Problems of geothermal energy resource research with special reference to the place and role of geophysical research methods, Committee for Geophysics SITRGMJ, Belgrade.
— Milivojević, M., Perić, J. and Simić, M., 1988. Potentiality of Karst Geothermal Resources in Mačva Region and Heat Energy Storage Feasibility. IIId Int. Colloq. an Appl. Geothermics, JIGASTOCK 88, Versailles, France Vol. 2, Paris.
— Milivojević, M., 1989. Assessment of Geothermal Resources of Serbia Excluding Autonomous Provinces. Doctoral thesis, Rud.-geol. fac. White City.
— Milivojević, M., 1991. Geothermal anomaly of the Pannonian Basin and its association with the geothermal anomaly of Serbia. In: S.Karamata (Editor), Geodynamic Evolution of The Pannonian Basin, SASA, Acad. Conf., Vol. LXII, Belgrade.
— Milivojević, M., 1992. Age of Tertiary Magmatism Rocks in the “Vardar Zone” on Serbian Territory using K/Ar Method and its Geothermal Importance. 29th IGC Kyoto.
— Papić, P., 1992. Scaling and corrosion potential of selected geothermal waters in Serbia. UNU Geothermal Training Programme, Reykjavik, Iceland, Report 9.
— Roksandić, M., 1974. Use of Geophysical Data in Explaining Tectonic Pattern and Evolution of the Dinarides. Voice CCXCII SANU, Dept. apr.-math. science, vol. 38, Belgrade.