Marko Cvetković

Application of Neural Networks in Petroleum Reservoir Lithology and Saturation Prediction

The Klostar oil fi eld is situated in the northern part of the Sava Depression within the Croatian part of the Pannonian Basin. The major petroleum reserves are confi ned to Miocene sandstones that comprise two production units: the Lower Pontian I sandstone series and the Upper Pannonian II sandstone series. We used well logs from two wells through these sandstones as input data in the neural network analysis, and used spontaneous potential and resistivity logs (R16 and R64) as the input in network training. The fi rst analysis included prediction of lithology, which was defi ned as either sandstone or marl. These two rock types were assigned categorical values of 1 or 0 which were then used in numerical analysis. The neural network was also used to predict hydrocarbon saturation in selected wells. The input dataset was extended to depth and categorical lithology. The prediction results were excellent, because the training and prediction dataset showed little disagreement between the true and predicted values. At present, this study represents the best and most useful application of neural networks in the Croatian part of the Pannonian Basin.

Reservoir Geology, Hydrocarbon Reserves and Production in the Croatian part of the Pannonian Basin System

3 of gas (52 fi elds), were recovered in the Croatian part of the Pannonian Basin System during 64 years of exploitation (1941- 2005). The production peak was attained between 1980-1989, when exploitation began in 12 new fi elds. Based on their cumulative production, the Croatian oil and gas fi elds can be divided into four groups, and the condensate fi elds into three groups. Such a division has been supported by analysis of recovery, number of reservoirs, porosity and permeability, age and lithology of reservoir rocks. The longest production period is assumed for the fi rst group of fi elds; for oil it is approximately 55 years, for condensate 46 and gas 36 years. In the favourable fi rst group the aver- age number of reservoirs is 16 for oil and 11 for gas. Lithological composition is highly favourable, because reser- voirs are represented mostly by sandstones of Pannonian and Pontian age with high porosities and permeabilities. A relatively homogeneous sandstone lithology, including good regional seals like marls, enables an increase in recov- ery through the use of secondary and tertiary recovery methods. Also, water-fl ooding will remain the dominant sec- ondary-recovery method for increased production in the future.

Sequential Indicator Simulations maps of porosity, depth and thickness of Miocene clastic sediments in the Kloštar field, Northern Croatia

Data from selected Lower Pontian sandstone reservoir in the Kloštar Field, situated in the Sava Depression (Northern Croatia), were used for mapping with Sequential Indicator Simulations rather than using a classical approach. Such approaches offer better insight in distribution of geological variables or zonal uncertainties in the cases with larger datasets (15 points or more). Obtained maps of porosity and reservoir thickness are presented here along with probability maps of certain selected cut of values of petrophysical parameters. Maps showed distinct sedimentological features that can clearly be observed on the both sets of maps.

Volume calculation of subsurface structures and traps in hydrocarbon exploration — a comparison between numerical integration and cell based models

Abstract The volume calculation of geological structures is one of the primary goals of interest when dealing with exploration or production of oil and gas in general. Most of those calculations are done using advanced software packages but still the mathematical workflow (equations) has to be used and understood for the initial volume calculation process. In this paper a comparison is given between bulk volume calculations of geological structures using trapezoidal and Simpson’s rule and the ones obtained from cell-based models. Comparison in calculation is illustrated with four models; dome ‐ 1/2 of ball/sphere, elongated anticline, stratigraphic trap due to lateral facies change and faulted anticline trap. Results show that Simpson’s and trapezoidal rules give a very accurate volume calculation even with a few inputs(isopach areas ‐ ordinates). A test of cell based model volume calculation precision against grid resolution is presented for various cases. For high accuracy, less the 1% of an error from coarsening, a cell area has to be 0.0008% of the reservoir area

Gaussian simulation of nitrate concentration distribution in the Zagreb aquifer

ABSTRACT Nitrates present one of the most common groundwater contaminants in the world and one of the five major groups of contaminants in the study area. Gaussian simulation (GS) algorithm was used for determining the spatial distribution of average nitrate concentrations from 2010 to 2015 on 95 sampling points. Results indicate two main focus areas of nitrate contamination, located on the left and right bank of the Sava River. Those areas generally extend according to groundwater flow, while areas near Sava River have much smaller concentrations. GS showed that they can be useful for this kind of mapping because they favor abrupt changes in data values which are in this case a result of heterogeneous lithological composition of the aquifer.

Petroleum system modeling and assessment of the remaining hydrocarbon potential in the eastern part of Drava Depression

AbstractThe eastern part of the Drava Depression presents a relatively small part within the Croatian part of the Pannonian Basin. A confined part within the eastern part of the Drava Depression with a working petroleum system but few hydrocarbon accumulations was selected for the evaluation of the remaining hydrocarbon potential. Four subsurface models were built with different levels of detail on which determination of source rock maturity, assessment of potential, and finally, volumes of the potentially accumulated hydrocarbons were estimated. In addition, several case scenarios, regarding source rock properties and boundary conditions were addressed to present the risk points in the model. Results proved that the amounts of hydrocarbons generated (72–129×106  m3 of oil and 1.5–3.2×1012  m3 of gas) and accumulated (14.3–40.57×106  m3 of oil and 139–182×109  m3 of gas) in the subsurface strongly support further exploration efforts even for the conservative modeling parameters (low heat flow [HF] and mod...

Low permeability Neogene lithofacies in Northern Croatia as potential unconventional hydrocarbon reservoirs

We present two examples of describing low permeability Neogene clastic lithofacies to outline unconventional hydrocarbon lithofacies. Both examples were selected from the Drava Depression, the largest macrostructure of the Pannonian Basin System located in Croatia. The first example is the Beničanci Field, the largest Croatian hydrocarbon reservoir discovered in Badenian coarse-grained clastics that consists mostly of breccia. The definition of low permeability lithofacies is related to the margins of the existing reservoir, where the reservoir lithology changed into a transitional one, which is mainly depicted by the marlitic sandstones. However, calculation of the POS (probability of success of new hydrocarbons) shows critical geological categories where probabilities are lower than those in the viable reservoir with proven reserves. Potential new hydrocarbon volumes are located in the structural margins, along the oil-water contact, with a POS of 9.375%. These potential reserves in those areas can be classified as probable. A second example was the Cremušina Structure, where a hydrocarbon reservoir was not proven, but where the entire structure has been transferred onto regional migration pathways. The Lower Pontian lithology is described from well logs as fine-grained sandstones with large sections of silty or marly clastics. As a result, the average porosity is low for conventional reservoir classification (10.57%). However, it is still an interesting case for consideration as a potentially unconventional reservoir, such as the ”tight” sandstones.

Palinspastic reconstruction of synsedimentary tectonics of Neogene and Quaternary sediments in the Kloštar Field (Sava Depression, Pannonian Basin, Croatia) [Palinspastische Rekonstruktion des Feldes Kloštar (Kroatien).]

The Klostar Oil and Gas Field lies approximately 35 km east of Zagreb in the Sava Depression, near the SW margin of the Pannonian Basin. According to detailed palinspastic reconstruction, based on well log data, the Miocene de- posits are up to 1600 m thick and transgressive over Paleozoic magmatites and metamorphites. Oil reservoirs are within Miocene sandstones and tectonised Paleozoic basement. About 7 × 106 m3 of oil have been recovered to date. The field struc- ture is anticlinal, striking NW–SE, and is cut by transverse and longitudinal faults. Different tectonic styles through the Neogene and Quaternary, palinspastically restored, resulted in 20 tectonic blocks in the field forming separate hydrody- namic units. The main longitudinal fault is part of the Northern marginal fault of the Sava Depression and crosses the central field area. The fault’s character changes along strike, from normal in the north to reverse in the south. The vertical displace- ments along formation boundaries average up to 50 m (generally larger in younger sediments) with maximum displacement in the Late “Pontian” due to long-lasting transtension gradually changing to transpression. Cumulative displacement during the Neogene and Quaternary was about 600 m.