MATERIALS and GEOENVIRONMENT MATERIALI in GEOOKOLJE RMZ - M&G, Vol. 62, No. 1 pp. 1-70 (2015) Ljubljana, April 2015 RMZ - Materials and Geoenvironment RMZ - Materiali in geookolje ISSN 1408-7073 Old title/Star naslov Mining and Metallurgy Quarterly/Rudarsko-metalurški zbornik ISSN 0035-9645, 1952-1997 Copyright © 2015 RMZ - Materials and Geoenvironment Published by/Izdajatelj Faculty of Natural Sciences and Engineering, University of Ljubljana/ Naravoslovnotehniška fakulteta, Univerza v Ljubljani Associated Publisher/Soizdajatelj Institute for Mining, Geotechnology and Environment, Ljubljana/ Inštitut za rudarstvo, geotehnologijo in okolje Velenje Coal Mine/Premogovnik Velenje Editor-in-Chief/Glavni urednik Peter Fajfar Editorial Manager/Odgovorni urednik Jakob Likar Editorial Board/Uredniški odbor Vlasta Ćosović, Sveučilište u Zagrebu Evgen Dervarič, Univerza v Ljubljani Meta Dobnikar, Ministrstvo za izobraževanje, znanost in šport Jan Falkus, Akademia Górniczno-Hutnicza im. S. Staszica w Krakowie Aleksandar Ganić, Univerzitet u Beogradu Borut Kosec, Univerza v Ljubljani Jakob Likar, Univerza v Ljubljani David John Lowe, British Geological Survey Ilija Mamuzić, Sveučilište u Zagrebu Milan Medved, Premogovnik Velenje Peter Moser, Montanuniversität Leoben Primož Mrvar, Univerza v Ljubljani Heinz Palkowski, Technische Universität Clausthal Daniele Peila, Politecnico di Torino Sebastiano Pelizza, Politecnico di Torino Jože Ratej, Inštitut za rudarstvo, geotehnologijo in okolje v Ljubljani Andrej Šmuc, Univerza v Ljubljani Milan Terčelj, Univerza v Ljubljani Milivoj Vulić, Univerza v Ljubljani Nina Zupančič, Univerza v Ljubljani Franc Zupanič, Univerza v Mariboru Editorial Office/Uredništvo Secretary/Tajnica Ines Langerholc Technical Editor/Tehnična urednica Helena Buh Editor of website/Urednik spletne strani Timotej Verbovšek Editorial Address/Naslov uredništva RMZ - Materials and Geoenvironment Aškerčeva cesta 12, p. p. 312 1001 Ljubljana, Slovenija Tel.: +386 (0)1 470 45 00 Fax.: +386 (0)1 470 45 60 E-mail: rmz-mg@ntf.uni-lj.si Linguistic Advisor/Lektor Jože Gasperič Design and DTP/Oblikovanje, prelom in priprava za tisk IDEJA.si Print/Tisk Birografika BORI, d. o. o. Printed in 200 copies./Naklada 200 izvodov. Published/Izhajanje 4 issues per year/4 številke letno Partly funded by Ministry of Education, Science and Sport of Republic of Slovenia./Pri financiranju revije sodeluje Ministrstvo za izobraževanje, znanost in šport Republike Slovenije. Articles published in Journal "RMZ M&G" are indexed in international secondary periodicals and databases:/Članki, objavljeni v periodični publikaciji „RMZ M&G", so indeksirani v mednarodnih sekundarnih virih: CA SEARCH® - Chemical Abstracts®, METADEX®, GeoRef. The authors themselves are liable for the contents of the papers./ Za mnenja in podatke v posameznih sestavkih so odgovorni avtorji. Annual subscription for individuals in Slovenia: 16.69 EUR, for institutions: 22.38 EUR. Annual subscription for the rest of the world: 20 EUR, for institutions: 40 EUR/Letna naročnina za posameznike v Sloveniji: 16,69 EUR, za inštitucije: 33,38 EUR. Letna naročnina za tujino: 20 EUR, inštitucije: 40 EUR. Current account/Tekoči račun Nova Ljubljanska banka, d. d., Ljubljana: UJP 01100-6030708186 VAT identification number/Davčna številka 24405388 Online Journal/Elektronska revija www.rmz-mg.com Table of Contents Kazalo Original scientific papers Izvirni znanstveni članki Co-Sn-Zn liquid phase thermodynamic properties investigation performed 3 by different geometric models and by CALPHAD method Raziskava termodinamičnih lastnosti tekoče faze Co-Sn-Zn z različnimi geometričnimi modeli in CALPAD-metodo Vanya Gandova Fragility curves of RC columns estimated by the CAE method 9 Ocena krivulj ranljivosti AB-stebrov s CAE-metodo Klemen Sinkovic, Iztok Perus Geochemical and isotopic characterization of coalbed gases in active excavation fields at 21 Preloge and Pesje (Velenje Basin) mining areas Geokemijska in izotopska karakterizacija premogovnih plinov iz aktivnih odkopov rudarskih območij Preloge in Pesje (Velenjski bazen) Tjaša Kanduč, Fausto Grassa, Jerneja Lazar, Simon Zavšek Indicator Kriging porosity maps of Upper Miocene sandstones, Sava Depression, 37 Northern Croatia Indikatorsko krigiranje poroznosti zgornjemiocenskih peščenjakov v Posavski kadunji v severni Hrvaški Tomislav Malvić, Karolina Novak, Kristina Novak Zelenika Professional papers Strokovni članki Design of product properties by suitable planning of a cold forging process 47 Oblikovanje lastnosti izdelka z ustreznim načrtovanjem procesa hladnega kovanja Vid Krušič, Martin Fabjan Petrophysical evaluation of reservoirs in 'y' prospect Niger delta Petrofizikalna ocena rezervoarjev na raziskovalnem območju 'y' v delti Nigra Yemisi. C. Ajisafe 55 Historical Review Zgodovinski pregled More than 90 years have passed since the University Ljubljana in Slovenia was founded in 1919. Technical fields were united in the School of Engineering that included the Geologic and Mining Division, while the Metallurgy Division was established only in 1939. Today, the Departments of Geology, Mining and Geotechnology, Materials and Metallurgy are all part of the Faculty of Natural Sciences and Engineering, University of Ljubljana. Before World War II, the members of the Mining Section together with the Association of Yugoslav Mining and Metallurgy Engineers began to publish the summaries of their research and studies in their technical periodical Rudarski zbornik (Mining Proceedings). Three volumes of Rudarski zbornik (1937, 1938 and 1939) were published. The War interrupted the publication and it was not until 1952 that the first issue of the new journal Rudarsko-metalurški zbornik - RMZ (Mining and Metallurgy Quarterly) was published by the Division of Mining and Metallurgy, University of Ljubljana. Today, the journal is regularly published quarterly. RMZ - M&G is co-issued and co-financed by the Faculty of Natural Sciences and Engineering Ljubljana, the Institute for Mining, Geotechnology and Environment Ljubljana, and the Velenje Coal Mine. In addition, it is partly funded by the Ministry of Education, Science and Sport of Slovenia. During the meeting of the Advisory and the Editorial Board on May 22, 1998, Rudarsko-metalurški zbornik was renamed into "RMZ - Materials and Geoenvironment (RMZ - Materiali in Geookolje)" or shortly RMZ - M&G. RMZ - M&G is managed by an advisory and international editorial board and is exchanged with other world-known periodicals. All the papers submitted to the RMZ - M&G undergoes the course of the peer-review process. RMZ - M&G is the only scientific and professional periodical in Slovenia which has been published in the same form for 60 years. It incorporates the scientific and professional topics on geology, mining, geotechnology, materials and metallurgy. In the year 2013, the Editorial Board decided to modernize the journal's format. A wide range of topics on geosciences are welcome to be published in the RMZ - Materials and Geoenvironment. Research results in geology, hydrogeology, mining, geotechnology, materials, metallurgy, natural and anthropogenic pollution of environment, biogeochemistry are the proposed fields of work which the journal will handle. Že več kot 90 let je minilo od ustanovitve Univerze v Ljubljani leta 1919. Tehnične stroke so se združile v tehniški visoki šoli, ki sta jo sestavljala oddelka za geologijo in rudarstvo, medtem ko je bil oddelek za metalurgijo ustanovljen leta 1939. Danes oddelki za geologijo, rudarstvo in geotehnologijo ter materiale in metalurgijo delujejo v sklopu Naravoslovnotehniške fakultete Univerze v Ljubljani. Pred 2. svetovno vojno so člani rudarske sekcije skupaj z Združenjem jugoslovanskih inženirjev rudarstva in metalurgije začeli izdajanje povzetkov njihovega raziskovalnega dela v Rudarskem zborniku. Izšli so trije letniki zbornika (1937, 1938 in 1939). Vojna je prekinila izdajanje zbornika vse do leta 1952, ko je izšel prvi letnik nove revije Rudarsko-metalurški zbornik - RMZ v izdaji odsekov za rudarstvo in metalurgijo Univerze v Ljubljani. Danes revija izhaja štirikrat letno. RMZ - M&G izdajajo in financirajo Naravoslovnotehniška fakulteta v Ljubljani, Inštitut za rudarstvo, geotehnologijo in okolje ter Premogovnik Velenje. Prav tako izdajo revije financira Ministrstvo za izobraževanje, znanost in šport. Na seji izdajateljskega sveta in uredniškega odbora je bilo 22. maja 1998 sklenjeno, da se Rudarsko-metalurški zbornik preimenuje v RMZ - Materiali in geookolje (RMZ -Materials and Geoenvironment) ali skrajšano RMZ - M&G. Revijo RMZ - M&G upravljata izdajateljski svet in mednarodni uredniški odbor. Revija je vključena v mednarodno izmenjavo svetovno znanih publikacij. Vsi članki so podvrženi recenzijskemu postopku. RMZ - M&G je edina strokovno-znanstvena revija v Sloveniji, ki izhaja v nespremenjeni obliki že 60 let. Združuje področja geologije, rudarstva, geotehnologije, materialov in metalurgije. Uredniški odbor je leta 2013 sklenil, da posodobi obliko revije. Za objavo v reviji RMZ - Materiali in geookolje so dobrodošli tudi prispevki s širokega področja geoznanosti, kot so: geologija, hidrologija, rudarstvo, geotehnologija, materiali, metalurgija, onesnaževanje okolja in biokemija. Glavni urednik Editor-in-Chief Original scientific paper Received: November 25, 2014 Accepted: January 29, 2015 Co-Sn-Zn liquid phase thermodynamic properties investigation performed by different geometric models and by CALPHAD method Raziskava termodinamičnih lastnosti tekoče faze Co-Sn-Zn z različnimi geometričnimi modeli in CALPAD-metodo Vanya Gandova University of Food Technologies, Inorganic and Physical Chemistry Department, 26 Mariza avenue, 4000 Plovdiv, Bulgaria Corresponding author. E-mail: gandova_71@abv.bg Abstract Predictions for the liquid Co-Sn-Zn alloys thermodynamic properties (molar excess Gibbs energy) were presented in this paper. The calculations were performed in the temperature range 1 000-2 000 K. Geometric models were used and the respective calculated molar excess Gibbs energies were compared to Calphad method assessments. The concentration dependences of the liquid phase thermodynamic properties along vertical sections with Sn/Zn ratios of 1 : 5, 1 : 1 and 5 : 1 were estimated. Ternary interaction parameters (L0, L1 and L2) of the liquid phase were determined using General solution (geometric) models from thermodynamic data of the binary end-systems (Co-Sn, Co-Zn, Sn-Zn). Key words: sgeneral solution model, ternary interaction parameters, ternary systems, Calphad method Izvleček V članku je predstavljena napoved termodinamičnih lastnosti (molske prebitne proste Gibbsove energije) tekočih zlitin Co-Sn-Zn. Izračuni so bili izvedeni v temperaturnem območju 1 000-2 000 K. Uporabljeni so bili geometrijski modeli. Odgovarjajoče izračunane molske prebitne proste Gibbsove energije so bile primerjane z oceno po Calpad-metodi. Termodinamične lastnosti tekoče faze v odvisnosti od koncentracije so bile določene za vertikalne prereze in razmerja Sn/Zn 1 : 5, 1 : 1 ter 5 : 1. Ternarni interakcijski koeficienti (L0, L1 and L2) so bili določeni z uporabo modela posplošene rešitve iz končnih birnarnih sistemov (Co-Sn, Co-Zn, Sn-Zn). Ključne besede: model posplošene rešitve, ternarni interakcijski parametri, ternarni sistem, Calpad-meto-da Introduction The Co-Sn-Zn system is interesting as potential materials using in industrial application like alloys applicable as lead free materials[1]. These materials are expected to be designed on the basis of systems containing low-melting elements like Sn and Zn. It is well known that the classical lead-tin based alloys represent a serious health and environmental risk. Pb-contain-ing alloys use recently[2] but the replacement of the whole variety of Sn-Pb based materials turned out to be a very difficult task[3]. The binary end-systems Co-Sn[4] and Co-Zn[5] have been intensively studied. They exhibit a large numbers of intermetallic phases. The binary system Sn-Zn represents a simple eutectic phase diagram[6]. This ternary system is included in the thermodynamic database developed by the European concerted action Solders[6] and reliable thermodynamic optimization is available. The task of the present study is to apply different ways to assess the thermochemical properties of the ternary melt Co-Sn-Zn. Theoretical fundamentals of the assessments The so-called "geometric models" give the possibility to predict the thermodynamic properties of a ternary phase (in this case - liquid] using the data for the respective binary end systems. In this work, assessments were done using the most common classic geometric models of Kohler[7], Toop[8], and Hilert[9] as well as the general solution model (GSM] developed by Chou[10' 11]. Hillert[9] classified the geometric models as symmetrical (e.g.[7]) and asymmetrical (e.g.[8, 9]). Such a universal approach was developed recently by Chou[10, 11] and was successfully applied to a variety of cases[12, 13, 14]. Nevertheless, a brief description of the techniques used is described below. The molar excess Gibbs energy (AGE, J mol-1] of the ternary liquid phase was chosen as parameter which values have to be calculated by various models and compared. This function describes the contribution of the non-ideal mixing to the thermodynamic properties of a solution phase. The molar excess Gibbs energies values of every binary end-system are necessary as starting points and calculated by means of Thermo-Calc software package[15]. The composition dependence of the binary Gibbs molar excess energies (AG. was given by Redlich-Kister formalism[16]. The Gibbs molar excess energy of a ternary phase (AGE123], consisting of the elements 1, 2 and 3, was given by the expression: AG£123 = X1X2AGE12 + X2X3AGE23 + *3*1AG*31 + (1] where AGE123 is the contribution of the ternary non-ideal mixing. In the simplest case of a regular ternary solution it may be assessed as: ^^123 - X1X2X3^123 (2] where L123 is a ternary interaction parameter that might be temperature and concentration dependent. The most essential equations, associated to the geometrical models[7, 8' 9] were used for calculations. Equation (3] can be used as an introduction of the General solution model (GSM] of Chou[10, 11]: ^123 ~~ XlX2Xlfl2 (3] Here f is the ternary interaction coefficient, related to the Redlich-Kister ternary interacts parameters L.jk (f123 = *1 X L°123 + *2 X L1123 + X3 X L2123]. are "similarity coefficients", that were defined by the term n,. called "deviation sum of squares. Equation (4] presented model of Chou. f\23 = (2^2 - 1)(L212((2^12 - + 2(*1 - xj] + L1!2> + (^23 - 1)(L223((2?23 - 1)X1 + 2(*2 - X3]] + ^3} + (2^ - 1)(L231((2?12 - 1)X2 + 2(x3 - xj] + V-J (4] Basic thermodynamic information on the binary subsystems, needed for the assessment, was taken from[6]. The optimized Redlich-Kis-ter parameters of each system are presented in 5 Table 1. They were used for the calculation of the molar excess Gibbs energies of the binary end-systems liquid phases. In this work: Co is represented as component 1, Sn - component 2 and Zn - component 3. Results and discussion Calculations of the coefficients f123 were done along three sections of the Co-Sn-Zn system with molar Sn : Zn ratios 1 : 5, 1 : 1 and 5 : 1 in the interval 1 000-2 000 K at amount fractions of cobalt equal to 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 and 1. In such a way a large amount of data was obtained and used thereafter to derive the parameters L..k. The results for the ternary parameters are shown in Table 2. Comparative reviews of the molar Gibbs excess energies (AGE, J mol-1) assessed at 1 973 K by using different geometric models (GSM[10, 11], Toop[8], Hilert[9] and Kohler[7]] and by the Cal-phad method (binary parameters only][6]] are shown in Figures 1-3. Figure 1 shows calcula- tions for the molar ratio Sn : Zn = 1 : 5. In this figure all curves exhibit positive values for the Gibbs excess energies. These positive deviations can be related to a possible miscibility gap in liquid phase at Co-Zn side in the ternary diagram. In this case, the values assessed by symmetrical models[7] were quite similar and were accompanied by the Calphad-type calculated quantities. The same conclusion is valid for the values calculated by both asymmetrical meth-ods[8, 9]. The GSM-assessed molar Gibbs excess energies deviate from all others. o- -1000 о < -2500 \ T=1973К Sn:Zn=5:1 4 1 '/ v У 1--GSM 2---TC 3* - - Toop 4----Hilert 5----Köhler > 0.2 2000- - -4 1 1000- 4'/ & 4' Ц 5 Л >. \ / » 4 . v\ \ / v w \ / Т=1973 К \ - . 4 \\\ 2 \ \\\ 4 \ . \ / Sn:Zn=1:5 / S / V 1--GSM 2---TC 3- - - Toop 4----Hilert 5----Kohler \\ \ 4\\ 0.6 Figure 1: Calculated molar Gibbs excess energies (äGE/(J mol-) of the liquid phase, along a section with constant molar Sn/Zn ratio equal to 1 : 5 at 1 973 K. Figure 2: Calculated molar Gibbs excess energies (äGE, J mol1) of the liquid phase, along a section with constant molar Sn : Zn ratio equal to 5 : 1 at 1 973 K. Most of the assessments predict (in general] negative AGE with a minimum in the composition interval of 0.5-0.55 amount fraction cobalt is shown in Figure 2. The calculations based on the GSM[10, 11] deviate from all others predicting a minimum at around 0.4-0.5 amount fractions cobalt. In Figure 3 the molar Gibbs excess energies along the section with Sn:Zn molar ratio equal to 1 : 1 all curves are with sign-changing values. Small negative values, up to around -700 J mol-1 Table 1: Optimized parameters (L0, present work; T - temperature, K , L2) for the liquid binary phases of the Co-Sn141, Co-Zn15' and Sn-Zn16' systems used in the System, i-j Ь°ЛТ)Ю mol-1) L1 a (T)/(J mol1) L*(T)/(J mol1) Co Sn -113 890 + 568.4038 * T -56 193.26 + 283.7657 * T 0 -68.169 * T * LN(T) -33.6875 * T * LN(T) Co -Zn -15 017 + 12.735 * T +51 758 - 29.752 * T 0 Sn -Zn +19 314.64 - 75.89949 * T + 8.751396 * T * LN(T) -5 696.28 + 4.20198 * T +1 037.22 + 0.98362 * T L1 6 in curves 2-5 are calculated. Sign-changing positive values are reached to 300 J mol-1. Along this section a maximum in the composition interval of 0.6-0.75 amount fractions cobalt are predicted, except by the assessment done using the GSM model (Figure 3, curve 1). The latter is sign-changing as well but deviates symmetrically from each other calculation. The reason for this discrepancy could not be found. These deviations in GSM model is observed in another ternary system - Ni-Bi-Zn[17]. 400 200 0- E -200 - Г -400 -b <1 -600 -800-1000 \ N\ ' —. _ __ ч : T=1973К Sn:Zn=1:1 / 1--GSM 2---TC 3- - - Toop 4----Hilert 5----Kohler N \ \ 2 > f 0.0 —I— 0.2 1000- 0- -1000- -2000 - -3000 - -4000 900 К 1800 К 2000 К 1000 К -1-'-1-'-1— 0.0 0.2 0.4 0.6 The values of the ternary molar Gibbs excess energies (AGE) in the temperature range 1 0002 000 K and along sections with Sn : Zn molar ratios equal to 1 : 5, 1 : 1 and 5 : 1, obtained from GSM are shown in Figures 5-7, respectively. These figures give a graphical view of the surfaces calculated at different temperature region (from 1 000 to 2 000 K) constituted by the values of the liquid phase molar Gibbs excess energies and amount fractions Co. Typically, there are maximums (positive AGE values) in the Co-rich regions and especially in Co-Zn rich solutions. From another side relatively small negative AGE values are predicted for the Co-Sn rich compositions. Figure 3: Calculated molar Gibbs excess energies (AGE, J moh1,) of the liquid phase, along a section with constant molar Sn : Zn ratio equal to 1 : 1 at 1 973 K. Figure 4 presented molar Gibbs excess energies of the liquid phase in broad temperature range 1 000-2 000 K used GSM model of Chou[10' 11]. The calculations exhibit negative Gibbs energies at low temperature. But at temperatures of 1 400 K to 2 000 K Gibbs energies are shown mix of negative and positive values. Figure 5: Ternary molar Gibbs excess energies calculated along the selected sections and at the retained temperatures. Figure 4: Calculated molar Gibbs excess energies (AGE, J mol1) Figure 6: Ternary molar Gibbs excess energies calculated of the liquid phase in all temperature range according to GSM along the selected sections and at the retained temperatures. models. 7 Figure 7: Ternary molar Gibbs excess energies calculated along the selected sections and at the retained temperatures. Figures 8-10 presented comparative revue between molar Gibbs excess energies of the liquid phase calculated at 1 973 K with binary coefficients of liquid phase[4-6] and with ternary coefficients obtained in this work. At calculations with molar ratios Sn : Zn equal to 1 : 1 and 5 : 1 observed that Gibbs energy calculated with ternary coefficients is more negative then calculated with binary parameters only. At Sn : Zn ratio equal to 1 : 5 appeared positive values of molar Gibbs excess energy with binary and ternary parameters. This is probably connected with miscibility gap in Co-Zn corner in the ternary phase diagram. \ Sn:Zn=1:1 ? О • 4 \\ -500- — — Binary coeficients • • Ternary coeficients 0.0 0.2 0.4 0.6 0.8 Figure 8: Comparative analysis between Gibbs free energy of the liquid phase, along a section with constant molar Sn : Zn ratio equal to 1 : 1. Figure 9: Comparative analysis between Gibbs free energy of the liquid phase, along a section with constant molar Sn : Zn ratio equal to 1 : 5. 1.0 Figure 10: Comparative analysis between Gibbs free energy of the liquid phase, along a section with constant molar Sn : Zn ratio equal to 5 : 1 . Conclusion Some thermodynamic properties of the Co-Sn-Zn liquid phase were predicted using the general solution model developed by Chou and have been compared with different geometrical models. The general solution model have middle place between symmetrical and asymmetrical models and give possibilities for estimating thermodynamic properties and calculating phase diagrams for ternary systems. Ternary interaction parameters (L0, L1 and L2) of the liquid phase have been determined using General solution model from thermodynamic data of the binary end-systems (Co-Sn, Co-Zn, Sn-Zn). The values of ternary parameters are: L0 = +2 384.018 - 0.7073 * T; L1 = +1 879.1670.0547 * T; L2 = +1 622.753-0.065 * T. The comparative analyses were performed between Gibbs free energy of the liquid phase with ternary parameters obtained in this work and with binary parameters of each binary system. Good agreement was found indicating that such an approach was possible in systems where no experimental data were available. References [1] Chen, S. W., Wang, C. H., Lin, S. K. (2007): Phase diagrams of Pb-free solders and their related materials systems.J. Mater. Electron., 18, pp. 19-37. [2] Directive 2002/95/EC of the European Parliament and of the Council of 27 January 2003 on the restriction of the use of certain hazardous substances in electrical and electronic equipment (2003), Official Journal of the European Union, L37, pp. 19-24. Available on: . [3] Ipser, H., Mikula, A. (2001): Memorandum of Understanding for the Implementation of a European Concerted Research Action Designated as COST 531 "Lead-free solder materials", University of Vienna. Available on: . [4] Vassilev, G. P., Lilova, K. I. (2006): Contribution to the thermodynamics of the Co-Sn system. Archives of metallurgy and materials, 51, pp. 365-375. [5] Vassilev, G. P., Jiang, M. (2004): Thermodynamic Optimization of the Co-Zn System. Journal of Phase Equilibria and Diffusion, 25, pp. 259-268. [6] Dinsdale, A., Watson, A., Kroupa, A., Vrestal, J., Zemanova, A., Vizdal, J. (2008): Atlas of Phase Diagrams for the Lead-Free Soldering, COST 531 (Lead-free Solders), Printed in the Czech Republic, Vol. 1. [7] Kohler, F. (1960): Zur Berechnung der thermodyna-mischen Daten eines ternären Systems aus den zugehörigen binären Systemen. Monatshefte fur Chemie, 91, pp. 738-740. [8] Toop, G. W. (1965): Predicting ternary activities using binary data. Trans. Metal Societe., 233, pp. 850-854. [9] Hilert, M. (1980): Empirical Methods of predicting and representing thermodynamics properties of ternary solution phases. Calphad, 4, pp. 1-12. [10] Chou, K. C. (1995): A general solution model for predicting ternary thermodynamics properties. Calphad, 19, pp. 315-325. [11] Chou, K. C., Li, W. C., Li, F., He, M. (1997): Formalism of new ternary model expressed in terms of binary regular-solution type parameters. Calphad, 20, pp. 395-406. [12] Živković, D., Katayama, I., Yamashita, H., Manasijević, D., Živković, Ž. (2006): Investigation of the thermodynamic model and ternary interaction parameter influence for Sn-Ag-Bi liquid alloys. RMZ - Materials and Geoenvironment, 53, pp. 155-161. [13] Marjanović, S., Manasijević, D., Živković, D. Gusković, D., Minić, D. (2009): Calculation of thermodynamic properties for ternary Ag-Cu-Sn system. RMZ -Materials and Geoenvironment, 56, pp. 30-37. [14] Živković, D., Manasijević, D., Mihajlović, I., Živković, Ž. (2006): Calculation of the thermodynamic properties of liquid Ag-In-Sb alloys. Serb. Chem. Soc., 71, pp. 203-211. [15] Andersson, J. O., Helander, T., Höglund, L., Shi, P., Sundman, B. (2002): Thermo-Calc & DIictra, Computational Tools For Materials Science. Calphad, 26, pp. 273-312. [16] Redlich, O., Kister, A. (1948): Algebraic Representation of Thermodynamic Properties and the Classification of Solutions. Industrial Engineering Chemistry, 40, pp. 345-348. [17] Gandova,V., Vassilev, G. (2013): Comparative analyses of thermodynamic properties assessments, performed by geometric models: application to the Ni-Bi-Zn system. J. Min. Metall. Sect. B-Metall., 49, pp. 347-352. Original scientific paper Received: June 16, 2014 Accepted: January 30, 2015 Fragility curves of RC columns estimated by the CAE method Ocena krivulj ranljivosti AB-stebrov s CAE-metodo Klemen Sinkovic, Iztok Perus* University of Ljubljana, Faculty of Civil and Geodetic Engineering, Jamova cesta 2, 1000 Ljubljana, Slovenia Corresponding author. E-mail: iperus@siol.net Abstract In the paper an alternative empirical approach for the estimation of fragility curves for RC columns is proposed. The CAE (Conditional Average Estimator) method was used. The procedure includes the LHS method, which was applied in order to take into account different uncertainties. The result of the study are estimated fragility curves for typical reinforced concrete (RC) columns (1) designed without seismic detailing, (2) designed according to first seismic codes used in the former Yugoslavia and (3) designed according to Eurocode 8. The obtained results clearly reveal the higher deformation capacity of RC columns, designed according to Eurocode 8. Additionally, the fragility curves related to various damage states, i.e. concrete crushing, longitudinal bar buckling, and longitudinal bar fracture were estimated. It is concluded that the proposed procedure offers a viable alternative to existing approaches. Key words: CAE method, drift, fragility, RC columns, performance-based earthquake engineering (PBEE) Izvleček V prispevku je predlagan alternativni empirični način za oceno krivulj ranljivosti za armiranobetonske (AB) stebre. Uporabljena je bila CAE-metoda s cenilko pogojnega povprečja. Postopek vključuje tudi LHS-me-todo, ki je bila uporabljena z namenom upoštevanja različnih negotovosti. Rezultati študije so ocene krivulj ranljivosti tipičnih AB-stebrov treh različnih obdobij, in sicer za (1) AB-stebre, projektirane brez potresnih predpisov, (2) AB-stebre, projektirane v skladu s prvimi potresnimi predpisi v nekdanji Jugoslaviji, in (3) AB--stebre, projektirane v skladu s predpisi Evrokod 8. Dobljeni rezultati jasno kažejo na večjo deformacijsko kapaciteto stebrov, projektiranih po predpisih Evrokod 8. Dodatno so bile krivulje ranljivosti ocenjene tudi za druga stanja poškodovanosti, kot je drobljenje betona, lokalni uklon vzdolžne armature in zlom vzdolžne armature. Mogoče je skleniti, da predlagani postopek ponuja dobro alternativo sedanjim načinom. Ključne besede: CAE-metoda, zamik, ranljivost, AB--stebri, potresno inženirstvo Introduction In contrast to many of the existing models for the prediction of the deformation capacity of RC columns, which provide only deterministic (point] estimations, in the performance-based design of seismic resistant buildings predictive capacity models that are unbiased and explicitly account for all the uncertainties are needed. Only in probabilistic models of capacity is variability explicitly taken into account[1]. The aim of the paper is thus to propose an alternative method for the estimation of fragility curves for RC columns, which is based on the CAE probabilistic drift capacity model. As well as in performance-based seismic design, fragility curves are very important for the estimation of the overall risk to the civil infrastructure in earthquake-prone areas. They can be used for emergency response and disaster planning by local and national authorities. Insurance companies can use them to make assessments of potential losses due to a particular scenario earthquake. Due to their numerous possible applications, a number of different research projects have been carried out worldwide. For example, Singhal and Kiremidjian[2] presented vulnerability curves and damage probability matrices for low-, mid- and high-rise RC frame structures, using Monte Carlo simulation techniques and non-linear dynamic analyses. Dumova-Jovanoska[3] presented damage probability matrices and damage indices as functions of intensity for different damage states for selected RC frame and RC wall-frame structures. Kappos et al.[4] applied a hybrid approach for the development of vulnerability curves for reinforced concrete and unreinforced masonry structures in terms of peak ground acceleration and spectral displacement. Panag-iotakos and Fardis[5] evaluated the performance of generic archetypal RC buildings according to Eurocode 8, using non-linear analyses. This study was later upgraded by the development of corresponding fragility functions[6]. Akkar et al[7] estimated vulnerability curves for low-and mid-rise infilled frame RC buildings. Pushover analyses of a number of existing buildings in Duzce were performed for buildings with a low-level of seismic design, having between two and five storeys. Good agreement of the estimated vulnerability curves with observed damage after the 1999 Duzce earthquake was observed. Rossetto & Elnashai[8] produced vulnerability curves for low-rise infilled RC frames, designed on the basis of the old Italian seismic code. The proposed analytical curves are in reasonable agreement with the empirical curves. There are several other applications. A more detailed survey of works related to fragility can be found in[6], and on the probabilistic approach to capacity models in[9]. Due to the complexity of the problem, fragility studies generally focus on generic types of structures[6]. Consequently, simplified structural models, having properties that account for the uncertainties and randomness in the structural parameters, are used to mathematically represent real buildings. The fragility curves are then presented as a function of different intensity measures, IM, (e.g. PGA, PGV, Sa, Sd) for different types of structures, different numbers of storeys, etc., taking into account different damage states (e.g. yielding, collapse]. Fragility curves can be presented for buildings as a whole or for any of their structural elements. In the latter case, demand can also be expressed in terms of deformation, not only in terms of IM. This is because the fragility of a structural member is defined as the conditional probability of failure for given values of the selected demand parameter. Hence, fragility curves can be obtained by simulations of the seismic response of structures at varying demand thresholds. Moreover, by ignoring the uncertainty in the demand, each fragility curve can display the probability that the capacity is less than or equal to a particular displacement/drift de- mand[1]. Empirical, expert opinion based, analytical and hybrid methods can be used for the assessment of fragility curves. In this paper an empirical approach to the estimation of the fragility curves of RC columns using the CAE method is presented. For example, the fragility of a RC column can be defined as the conditional probability of failure (at an ultimate drift defined at a 20 % drop in maximum strength] for given values of drift demand. Additionally, fragility can be derived in terms of other damage states, e.g. concrete crushing, longitudinal bar buckling, longitudinal bar fracture and axial loss failure. In this paper the CAE method is briefly presented. The different databases used in the study are described. The obtained results are verified by making comparisons with existing solutions from the literature. Fragility curves for typical RC columns, designed without seismic detailing, designed according to the first seismic codes used in the former Yugoslavia, and designed according to EC8, are provided as the results of the study. Procedure for the estimation of fragility curves The CAE method The fragility curves presented in this study are estimated by the CAE method. A detailed description of this method from the engineering point of view is given in Peruš et al.[10]. Here only a brief description is given. The phenomenon of the capacity of RC columns and then the estimation of the corresponding fragility curves can be described by observing N RC column specimens during the experiments. The mathematical description of the observation of one specimen during the experiment is called a model vector. As a result, the whole phenomenon can be described by a finite set of model vectors. {Xt,..., Xn,..., XN} (1) It is assumed that the observation of one particular specimen can be described by a number of variables, which are treated as components of a model vector xn = iKi.....K,' "A* C„1.....cnw -'cJ (2) The vector Xn can be further composed of two truncated vectors B and C B« = {K.....bn,.....bJanà C = (c„i' •••> cnk> ->cnM} (3a) Vector Bn is complementary to vector Cn and therefore their concatenation yields the complete data model, vector Xn. The prediction vector, too, is composed of two truncated vec- tors, i.e., the given truncated vector B and the unknown complementary vector Č B= {blt..., br ...,bD} and C- {£,,..., čp —,cM} (3b) The problem now is how an unknown complementary vector Č can be estimated from a given truncated vector B and the model vectors {X1, ..., Xn, ..., Xw}, i.e., how the drift capacity 5 can be estimated from known input parameters and the available data in the database. By using the conditional probability density function, the optimal estimator for the given problem can be expressed as 5 = V S A where A = (4) N and a = {2n)% wDn exp * съ.-bj2 fi 2w„2 (5) [6) Note that in the above equations M is assumed to be 1 and consequently с = 5 . In equations (4-6) 5 is an estimate of the displacement/drift at failure or any other damage state (e.g. flexural and shear failure, defined at a 20 % reduction in lateral strength, axial failure, bar buckling, etc.), 5n is the same output variable corresponding to the n-th model vector in the database, N is the number of model vectors in the database, bnI is the l-th input variable (e.g. axial load index, P*, shear span index, L*) of the n-th model vector in the database, and b= is the l-th input variable corresponding to the prediction vector. Note that each model vector corresponds to the results of one experiment from the database. D is the number of input variables, and defines the dimension of the sample space. A Gaussian function is used in order to achieve a smooth interpolation between the points of the model vectors. In this context the width wn is called the "smoothing" parameter that corresponds to n-th model vector from the database. In our case the same width w of the n Gaussian function is used for all the input variables. It is therefore important that the input parameters in the equation for an are normalized, i.e. generally in the range from 0 to 1. An intermediate result in the computational process is parameter p, which is defined as: P = (7) It provides a measure of how the influence of all the model vectors in the database is spread over the sample space, and strongly depends on the smoothing parameter w. It helps to detect possible less accurate predictions (indicated by low values of p) due to the manner of data distribution in the database, and due to local extrapolation outside the data range. When the expression for the displacement/ drift S at failure (Equation 4) is compared with the expression for the first order moment of the random variable X, which corresponds to the mean value m x E[X]=txiPx{x} = mx (8) i=1 the similarity between the two expressions becomes evident[11]. px(x) is the probability of the random variable X = x. and corresponds to the weights An (Equation 5), which depend on the similarity between the input variables of the prediction vector, and on the corresponding input variables pertinent to the model vectors stored in the database. Also, there is clear similarity when the central second order moment of the probability distribution of the random variable X, called its variance, given by the expression px=Var[X\=t(x-mxypx(xJ (9) is compared with the prediction of so-called "local standard deviation" in the CAE method: Èl = ì(8-S)*An (10) Л=1 The above interpretation of the CAE equations enables the estimation of the corresponding probability distribution, including the median value, as well as different percentile values. The proposed procedure is demonstrated by the simple example presented in Appendix A. The LHS method The LHS (Latin hypercube sampling) method was first described by Mackay et al.[12] and then further elaborated by Iman et al.[13]. The method, which has become increasingly popular, was used in this study in order to take into account different uncertainties (e.g. aleatory and epistemic). Aleatory uncertainties, which are inherent in the phenomenon itself, are irreducible, and cannot be influenced by the observer or the manner of observation. On the other hand, epistemic uncertainties arise from errors in measurement, from the finite size of the available observation data, and from the adopted mathematical model. Note that in the presented study no distinction was made between these two types of uncertainties. LHS allows the creation of experimental samples with as many points as needed or desired[14]. (Note, however, that the number of samples needed should not be too low[15].) It permits the use of very different statistical assumptions, and is able to treat both small and large design spaces (there are no constraints in terms of the data density and location). Additionally, LHS is flexible. If, for example, a few dimensions have to be dropped out, the resulting design is still a Latin hypercube. Moreover, the existing data can be reused without having to make any reduction in the number of sampled points. The use of LHS in its basic form does not account for any possible correlations between the variables. However, in earthquake engineering, some of the parameters of the RC structural elements are correlated (e.g. the yield strength of the reinforcement, and the concrete compressive strength), so some authors have made us of more advanced variants of LHS (e.g.[15-17]). In the presented study, however, derived input parameters were used, and since it was assumed that they are independent, a basic variant of LHS was used. This is well-justified since parametric studies (not presented here) have shown that the influence of derived input parameters has only a minor effect on the results presented in this study. The databases used in the study The application of the CAE method requires a representative database, many different databases of RC structural elements being presently available (e.g.[1, 10, 18-29]]. The deformation capacity expressed in terms of ultimate drift, representing a "near collapse" limit state, which is used in this study, is defined as a drift at a predefined drop below maximum strength. A 20 % drop in maximum strength (i.e. when the restoring force reaches 80 % of its maximum value] is commonly used, although this definition may significantly underestimate the true axial load-carrying capacity of the columns. In other cases, when lateral force resistance in columns is not reached because of premature load reversal, and also in cases, when deformation capacity is not reached because of limitation in the applied maximum displacement, the measured maximum drifts provided a lower bound of the deformation capacity. For validation and comparison purposes, the first database used in the study was the database on RC columns at axial failure[1]. In this paper it is called the DB1 database, and contains data on 28 RC column specimens. Only two input parameters are considered, namely the normalized axial load and the parameter s/d, which is related to the confinement (s is the hoop spacing, and d is the depth to the centre-line of the outermost tension reinforcement]. For the estimation of fragility curves for RC columns which failed in flexure, the PEER database was used[18]. The same input parameters as proposed by Perus et al.[10] were taken into account (i.e. an axial load index, a shear span index, the concrete compressive strength, the confinement effectiveness factor multiplied by the confinement index and the longitudinal reinforcement index]. The effective database was called DB2, and contains data on 156 RC column specimens. In order to develop fragility curves for other damage states, different from flexur-al or axial failure, an effective database, derived from the PEER database, and called the DB3 database, was used. It contains data on 80, 20 and 38 RC column specimens for which the drift at concrete crushing, longitudinal bar buckling, and longitudinal bar fracture, respectively, was measured. Fragility curves of RC columns Comparison with the existing solution In this section the results obtained by using the proposed procedure are compared with the results of the existing solution. It is assumed that all the random variables which account for different uncertainties in the input parameters (e.g. the material properties and the applied loads] are log-normally distributed. Fragility curves for axial failure were estimated, using the database DB1. The following two parameters and corresponding coefficients of variation were taken into account: - the axial load index (P*= P/Po; [0.07-0.22]]: CoV = 0.11; — the parameter s/d, related to the confinement [0.6-1.23]: CoV = 0.02 Note that the bounds which are used in the normalization process of the CAE method (see[10]) take into account the distribution of the two input parameters in the database DB1. By means of the LHS method, a database with 500 samples for each of the columns (3CLH18 and 2CLD12] from Zhu et al.[1], taking into account different uncertainties, was prepared. The CAE method was then applied. Due to the small number of input parameters, the smoothing parameter was given values of wmin = 0.05 and w = 0.1.The drift point estimates (i.e. the max r v mean and the local standard deviation] at axial failure were determined, using Eqs. 4-6 and 10, as a function of the axial load index and the parameter s/d. Considering Eqs. 4 and 8-10, the CAE empirical cumulative distribution function (CDF] was determined for each sample from the LHS generated database. The corresponding smoothed CDF for the median drift values, as well as for the 15 % and 85 % bounds, were then calculated. The estimate is based on the counted values obtained from the CAE empirical CDF. The results presented in Figure 1 indicate relatively good agreement between the CAE approach and the approach proposed by Zhou et al.[1]. The discrepancy between the results can be attributed to the functional form, which is not a priori assumed in case of the CAE method. The CAE functional form follows the data more closely, which can, however, in some cases (e.g. extrapolation) lead locally to illogical results. In the presented case, the p value (Equation 7) amounts to 0.6 and 1.2 for both of the RC sample columns, respectively. The relatively high values indicate accurate CAE predictions. The observed discrepancy between the results suggests that the uncertainties may be different (i.e. higher) to those usually estimated. Fragility curves of flexural failure for different designs of RC columns Fragility curves corresponding to flexural failure for different designs of RC columns (designed without seismic detailing - NO, designed according to first seismic codes used in the former Yugoslavia - YU, and designed according to Eurocode 8 - EC8) were estimated by CAE, using database DB2. Five input parameters were taken into account, with the following statistical parameters: — axial load index (P = P/Po): CoV = 0.11, — shear span index (L* = LV/h): CoV = 0.05, — concrete compressive strength (/c'): CoV = 0.05, — confinement index (ps* = ps /./£): CoV = 0.08 — longitudinal reinforcement index (Pi* = Pl /yl//c): CoV = 0.08. Fragility curves were estimated for L* = 3.5 and three different values of P*, which amounted to 0.05, 0.15 and 0.25, respectively. Additionally, in the case of the RC columns designed without seismic detailing, a fragility curve for P* = 0.35 was estimated. (Here it should be noted RC columns in such older buildings are, in many cases, heavily loaded due to the applied vertical loads.) Mean values of other three input parameters were assessed from the collected data on past designs. Note that only data on columns in the first storey, which is usually the critical one, were taken into account. The mean values and corresponding coefficients of variations, which account for different uncertainties, are presented in Table 1. The results which correspond to fragility curves with an 85 % confidence level are presented in Figure 2. A very small difference in the drift capacity between the RC columns designed without seismic detailing and RC columns designed according to YU codes can be observed. Table 1 Table 1: Mean values and corresponding coefficients of variation which accounts for different uncertainties NO YU EC8 p's 0.002 / 0.51 0.002 / 0.51 0.05 / 0.22 pl 0.15 / 0.41 0.15 / 0.22 0.2 / 0.22 /c 20 / 0.21 25 / 0.21 30 / 0.21 Figure 1: Fragility estimates for RC columns at axial failure for (a) column 3CLH18 and (b) column 2CLD121". Shown are discrete empirical CDF, obtained by the CAE method and by the corresponding smoothed (log-normal) CDF, and the same estimates, as obtained byZhu111. reveals, that the most notable change in the input parameters is the increase in concrete compressive strength, which only slightly improves the drift capacity and decreases dispersion. It should be noted, that despite of validation of YU codes for that particular time of being, the requirements for shear reinforcement were insufficient, which can be also observed from the data, taken from actual designs of buildings. (C) -CAE CDF for P'=0.05 -CAE CDF fa P*=0.15 -CAE CDF fa P*=0.25 —CAE CDF for P'=0.35 3% 4% 5% □rift demand S 1% 2% 3% 4% Drift demand S 1.0 0.9 - 0.8 - 0.7 - 0.6 - ss '5) 0.5 - LL 0.4 - 0.3 - 0.2 - 0.1 ■ 0.0 ■ EC8 -CAE CDF for P*=0.05 -CAE CDF for P"=0.15 -CAE CDF fa P*=0.25 2% 3% 4% 5% Drift demand S Figure 2: Fragility curves with an 85 % confidence level for a typical column, designed (a) without seismic codes, (b) according to seismic codes used in the former Yugoslavia, and (c) according to the EC8 code. In all cases L* amounts to 3.5. An important increase in drift capacity can be observed in the case of Eurocode 8, where all three parameters, especially the shear reinforcement which can significantly contribute to the drift capacity, are also increased. RC columns designed without seismic detailing exhibit a surprisingly high drift capacity. In the case of Eurocode 8 the ultimate drift capacity is reduced by a corresponding factor due to poor seismic detailing and the use of smoothed bars. The presented results suggest that a better CAE model should include additional parameters which would take into account proper seismic detailing. Fragility curves of RC columns for different damage states Besides the fragility curves corresponding to flexural and axial failure, fragility curves can also be estimated for other damage states, e.g. concrete crushing, longitudinal bar buckling and longitudinal bar fracture. The procedure is the same, only a proper database is required. In order to demonstrate the estimation of fragility curves of RC columns for other damage states, the database DB3 was used. Note that in this case uncertainties were not taken into account. The results are compared with the measured drift obtained in the case of the pseudo-dynamic testing of the 3-storey SPEAR building[30], which is representative building of old constructions in southern European Countries without specific provisions for earthquake resistance. It was designed for gravity loads alone, using the concrete design code applying in Greece between 1954 and 1995, with the construction practice and materials typical of the early 70s: for a concrete a nominal strength f' = 25 MPa was assumed while based on the scarcity of the current production, it was only possible to find reinforcement with a characteristics yield strength larger than initially requested fy x 450 MPa. The structure is regular in elevation since typical storey height is 3 m. The plan configuration is doubly non symmetric (Figure 3), with 2-bay frames spanning from 3 m to 6 m. Eight out of the nine columns have a square 25 cm x 25 cm cross-section (also the interior column C3), whereas the ninth column has a 25 cm x 75 cm cross section. Figure 3: The elevation view, the plan view and the typical reinforcement in RC columns of the SPEAR building. a) -Experiment empirical cdf- Crushing median cdf - Crushing —empirical cdf - Bar Buckling —median cdf - Bar Buckling —empirical cdf - Bar Fracture -median cdf - Bar Fracture —empirical cdf - Flexural Failure —median cdf - Fiexural Failure 2 3 4 Drift demand [%] Figure 4: Fragility curves at an 85 % confidence level for column C3 in the first storey of the SPEAR building. Considered are concrete crushing, longitudinal bar buckling, longitudinal bar fracture, and flexural failure of the column, respectively. Figure 4 reveals that concrete crushing should occur at early stages of the experiment. Figure 5 confirms the obtained result. Namely, extensive spalling and cracking was observed at the top of the column. The estimated smoothed CDF indicates that neither longitudinal bar buckling, nor longitudinal bar fracture or ultimate drift was achieved. Indeed, the ultimate drift was not achieved, and other heavy damage was not observed. At an 85 % confidence level, longitudinal bar buckling is expected at the increased drift (approx. 1.8 %). This may be followed by longitudinal bar fracture at 2.3 % drift. Flexural failure may be expected at a drift of 2.6 %. Figure 5: Damage after the PGA = 0.2 g pseudo-dynamic testing of the SPEAR building[30]: (a) extensive spalling at the top of the column, and (b) minor damage at the base of the column. Conclusions In the paper an empirical approach to the estimation of fragility curves of RC columns using the CAE method is presented. The LHS method was applied in order to take into account different uncertainties. As a result of the study fragility curves at an 85 % confidence level for typical RC columns, designed without seismic detailing, according to first seismic codes in the former Yugoslavia, and according to Eurocode 8 were proposed. The resulting increasing trend can be observed, while a small difference in the drift capacity was observed between the RC columns designed without seismic detailing and those designed according to the YU codes. Both types of columns were designed with relatively bad detailing for shear reinforcement and usually with smooth longitudinal reinforcement, which have the biggest role in case of EC8 codes and provide favourable flexural behaviour with adequate deformation capacity of RC columns. One of the possible reasons could be also the fact that the discussed existing buildings are not fully representative for buildings built before and after YU regulations. The other reason can be imperfect CAE model. Those differences were not fully discussed since the primary purpose of the article was to propose a relatively simple method for the determination of the fragility curves. On the other hand an important increase in drift capacity was observed in the case of EC8 codes, especially due to the increase and better detailing of shear reinforcement and also due to better detailing of longitudinal reinforcement with better material characteristics, which typically significantly contributes to the higher deformation capacity of RC columns. Further research is still needed in order to increase the sample size of existing buildings designed according to YU codes, and to improve the presented CAE model. Obtained conclusions may be very general, but other specific statements could be highly speculative at the moment. Fragility curves related to other damage states, i.e. concrete crushing, longitudinal bar buckling, and longitudinal bar fracture, were also estimated. It can be concluded that the proposed procedure offers a viable alternative to other procedures, and that there is no need to use closed-form equations for the prediction of point estimates. However, the presented results suggest that a better CAE model is needed to accurately describe the older designs of RC columns. It should include an additional parameter which would account for poor seismic detailing. Appendix A In order to better illustrate the use of the CAE method for the estimation the fragility curves, the calculations for the input data P" = 0.25 and L" = 3 are shown in Table A1. Note that, for illustration purposes, only two input parameters are considered. The database consists of 7 test samples of RC columns. Equations 1-7 are used, taking into account w = 0.15. It can be clearly seen that the influence of the 7 different input drift values (based on the results of measurements] on the predicted drift value depends on the similarity of the input parameters P' and L" between the measured and predicted deformation. The highest weight An is assigned to the sample #6 because its values of P' = 0.27 and L" = 2.4 are the nearest to the target values P' = 0.25 and L" = 3. In the next step the drifts S are sorted from the lowest to the highest value, together with the corresponding coefficients An (see Table A2). The results are presented in Figure A1, where Table A1: An example of the prediction of ultimate drift ö for P* = 0.5 and L* = 3 by the CAE method (w = 0.15,) Sample P* L* Norm. P" Norm. L" 8 a A SA (đ -8)2 A _£._a_a__v a t a #1_0.13 1.50 0.74 0.70 0.034 0.042 0.015 0.000 0.000 006 20 #2_0.35 3.50 0.30 0.30 0.042 0.371 0.129 0.005 0.000 020 22 #3 0.05 2.10 0.90 0.58 0.049 0.016 0.005 0.000 0.000 000 17 #4_0.00 3.30 1.00 0.34 0.078 0.004 0.001 0.000 0.000 000 77 #5_0.18 2.80 0.64 0.44 0.063 0.703 0.244 0.015 0.000 017 46 #6_0.27 2.40 0.46 0.52 0.051 0.789 0.274 0.014 0.000 003 43 #7_0.21 3.15 0.58 0.37 0.057 0.957 0.332 0.019 0.000 002 01 £ ^ 2.882 1.000 0.0545 0.000 050 26 a= V0.00005026 = 0.0071 the dotted line represents the values from the last column in Table A2 (CAE empirical CDF], The corresponding smoothed log-normal cumulative distribution function (smoothed CDF] is estimated from the mean and local standard deviation (5 = 0,045 45, a = 0,007 1], calcu- v mean ' J' lated above, Table A2: Determination of the CAE empirical and corresponding smoothed lognormal cumulative distribution function (CDF) Sample 6 A n 2A #1 0,034 0,014 70 0,014 70 #2 0,042 0,128 61 0,143 31 #3 0,049 0,005 43 0,148 74 #6 0,057 0,273 78 0,422 52 #7 0,063 0,332 18 0,754 70 #5 0,068 0,243 90 0,998 61 #4 0,081 0,001 39 1,000 00 1.0 0.6 0.4 ---CA -sm empirical □othed cdf :df / I / 1 1 / / i / i / t / / / / // // / 0.02 0.04 0.06 0.08 0.1 Drift demand ö [rad] Figure A1: Fragility function determined from the CAE empirical CDF and the corresponding smoothed lognormal CDF. Acknowledgements The results presented in this paper are based on work supported by the Slovenian Research Agency, The constructive comments provided by Prosessors P. Fajfar and M. Dolšek are greatly appreciated, References [1] Zhu, L., Elwood, K, J,, Haukaas, T, (2007): Classification and seismic safety evaluation of existing reinforced concrete columns, Journal of Structural Engineering, 133 (9), pp, 1316-1330, [2] Singhal, A, Kiremidjian, A, (1995): Method for developing motion damage relationships for reinforced concrete frames, Report #115, Blume Earthquake Engineering Centre, Stanford University, Stanford, [3] Dumova-Jovanoska, E, (2000): Fragility curves for reinforced concrete structures in Skopje (Macedonia) region, Soil Dynamics and Earthquake Engineering, 19 (6), pp, 455-466, [4] Kappos, A,, Panagopoulos, G,, Panagiotopoulos, C,, Penelis, G, (2006): A hybrid method for the vulnerability assessment of R/C and URM buildings, Bull. Earthquake Eng., 4, pp, 391-413, [5] Panagiotakos, T, B,, Fardis, M, N, (2004): Seismic performance of RC frames designed to Eurocode 8 or to the Greek Codes 2000, Bulletin of Earthquake Engineering, 2, pp, 221-259, [6] Papailia, A, (2011): Seismic fragility curves for reinforced concrete buildings, MSc Thesis (Supervisor M, N, Fardis), University of Patras, [7] Akkar, S,, Sucuoglu, H,, Yakut, A, (2005): Displacement-based fragility functions for low- and mid-rise ordinary concrete buildings, Earthquake Spectra, 21 (4), pp, 901-927, [8] Rossetto, T,, Elnashai, A, (2005): A new analytical procedure for the derivation of displacement-based vulnerability curves for populations of RC structures, Engineering Structures, 27 (3), pp, 397-409, [9] Zhu, L, (2005): Probabilistic drift capacity models for reinforced concrete columns, MSc Thesis (Supervisor K, J, Elwood, Co-supervisor T, Haukaas), The University of British Columbia, [10] Perus, I,, Poljansek, K,, Fajfar, P, (2006): Flexural deformation capacity of rectangular RC columns determined by the CAE method, Earthquake Eng. Struct. Dyn, 35 (12), pp, 1453-1470, [11] Perus, I,, Dolsek, M, (2009): The error estimation in the prediction of ultimate drift of RC columns for performance-based earthquake engineering, RMZ-Mater. Geoenviron, 56 (3), pp, 322-336, [12] McKay, M, D,, Beckman, R, J,, Conover, W, J, (1979): A Comparison of three methods for selecting values of input variables in the analysis of output from a computer code, Technometrics, 21, pp, 239-245, [13] Iman, R. L., Helton, J. C., Campbell, J. E. (1981): An approach to sensitivity analysis of computer models, Part 1. Introduction, input variable selection and preliminary variable assessment. Journal of Quality Technology, 13, pp. 174-183. [14] Viana, F. A. C. (2013): Things you wanted to know about the Latin hypercube design and were afraid to ask. 10th World Congress on Structural and Multidis-ciplinary Optimization, May 19-24, 2013, Orlando, Florida, USA. [15] Dolsek, M. (2009): Incremental dynamic analysis with consideration of modeling uncertainties. Earthquake Eng. Struct. Dyn., 38, pp. 805-825. [16] Vorechovsky, M., Novak, D. (2009): Correlation control in small-sample Monte Carlo type simulations I: A simulated annealing approach. Probabilistic Engineering Mechanics, 24, pp. 452-462. [17] Dolsek, M. (2012): Simplified method for seismic risk assessment of buildings with consideration of aleatory and epistemic uncertainty. Structure and Infrastructure Engineering: Maintenance, Management, Life-Cycle Design and Performance, 8 (10), pp. 939-953. [18] Berry, M., Parrish, M., Eberhard, M. (2004): PEER structural performance database, User's Manual, Version 1.0, 2004, Pacific Earthquake Engineering Research Center; Available on: < http://www. ce.washington.edu/~peera1/.> [19] Panagiotakos, T. B., Fardis, M. N. (2001): "Deformations of reinforced concrete members at yielding and ultimate." ACI Structural Journal, 98 (2), pp. 135-148. [20] Biskinis, D. E., Roupakias, G. K., Fardis, M. N. (2004): Degradation of Shear Strength of RC Members with Inelastic Cyclic Displacements. ACI Structural Journal, 101 (6), pp. 773-783. [21] Biskinis, D. E. (2007): Resistance and Deformation Capacity of Concrete Members with or without Retrofitting. PhD Thesis (Supervisor M. N. Fardis), Civil Engineering Department, University of Patras, Greece. (in Greek). [22] Biskinis, D. E., Fardis, M. N. (2010): Deformations at Flexural Yielding of Members with Continuous or Lap-Spliced Bars. Structural Concrete, 11 (3), pp. 127-138. [23] Biskinis, D. E., Fardis, M. N. (2010): Flexure-Controlled Ultimate Deformations of Members with Continuous or Lap-Spliced Bars, Structural Concrete, 11 (2), pp. 93-108. [24] Grammatikou, S. (2013): Strength, deformation capacity and failure mode of RC walls in seismic loading. MSc Thesis (Supervisor M. N. Fardis), Civil Engineering Department, University of Patras, Greece. (in Greek). [25] Lignos, D. G., Krawinkler, H. (2012): Sidesway collapse of deteriorating structural systems under seismic excitations. Report No. 177, 2012, The John A. Blume Earthquake Engineering Center, Stanford, CA, USA. [26] Lignos, D. G., Krawinkler, H. (2013): Development and utilization of structural component databases for performance-based earthquake engineering. Journal of Structural Engineering, Special issue 139, pp. 1382-1394. [27] Poljansek, K., Perus, I., Fajfar, P. (2009): Hyster-etic energy dissipation capacity and the cyclic to monotonic drift ratio for rectangular RC columns in flexure. Earthquake Eng. Struct. Dyn., 38 (7), pp. 907-928. [28] SERIES database (2013): RC column, beam and wall database. Available on: < http://www.dap.series. upatras.gr/>. [29] Perus, I., Biskinis, D., Fajfar, P., Fardis, M. N., Gram-matikou, S., Krawinkler, H., Lignos, D. (2014): The SERIES database of RC elements. 2nd European conference on Earthquake Engineering and Seismology, Istanbul, August 2014, Paper #145. [30] Negro, P., Mola, E., Molina, F. J., Magonette, G. E. (2004): Full-scale PSD testing of a torsionally unbalanced three-storey non-seismic RC frame. 13th WCEE, Paper # 968, August 2004, Vancouver, B. C., Canada. Original scientific paper Received: December 11, 2014 Accepted: February 2, 2015 Geochemical and isotopie characterization of coalbed gases in active excavation fields at Preloge and Pesje (Velenje Basin) mining areas Geokemijska in izotopska karakterizacija premogovnih plinov iz aktivnih odkopov rudarskih območij Preloge in Pesje (Velenjski bazen) Tjaša Kanduč1, *, Fausto Grassa2, Jerneja Lazar3, Simon Zavšek3 Jožef Stefan Institute, Jamova cesta 39, 1000 Ljubljana, Slovenia 2Istituto Nazionale di Geofisica e Vulcanologia Sezione di Palermo, Via Ugo La Malfa, 153, 90146 Palermo, Italy 3Velenje Coal mine, Partizanska 78, 3320 Velenje, Slovenia Corresponding author. E-mail: tjasa.kanduc@ijs.si Abstract Coalbed gases in excavation fields of mining areas in the Velenje coal basin have been subjected to geochemical and isotopic monitoring since the year 2000, with the aim of obtaining better insights into the origin of coalbed gases. Results from active excavation fields in the mining areas Pesje and Preloge in the year 2013 are presented in this study. Composition and isotopic composition of coalbed gases were determined with methods utilizing mass spectrometry. The chemical and isotopic composition of coalbed gases in the Velenje Basin vary and depend on the composition of the source of coalbed gas before excavation, advancement of the working face, depth of the longwall face, pre-min-ing activity and newly mined activity. The basic gas components determined in excavation fields are CO2 and methane. The isotopic composition of deuterium in methane has been determined and used to characterize the origin of methane. The isotopic compositions of carbon and hydrogen in methane in the excavation fields show its biogenic origin, while a high Carbon Dioxide Methane Index indicates the bacterial and endog-enic origin of CO2. Key words: Characterization of coalbed gases, stable isotopes, excavation fields, Velenje Basin Izvleček Geokemijski in izotopski monitoring premogovnih plinov na aktivnih odkopih v rudarskih območjih Velenjskega premogovnega bazena poteka od leta 2000 z namenom pridobiti vpogled v sestavo premogovnih plinov in njihov izvor. V tem prispevku predstavljamo geokemične in izotopske parametre iz aktivnih rudarskih območij Pesje in Preloge v letu 2013. Sestavo in izotopsko sestavo premogovnih plinov smo ugotovili z metodami masne spektrometrije. Premogovni plini v Velenjskem bazenu se spreminjajo po kemijski, kot tudi po izotopski sestavi glede na izvorno sestavo premogo-vega plina pred odkopavanjem, hitrost odkopavanja, globino odkopnega čela, odkopna čela s predhodno rudarsko aktivnostjo ali na novo odkopana območja. Glavni plinski komponenti na odkopih sta CO2 in metan. Na novo v tem prispevku predstavljamo rezultate izotopske sestave devterija v metanu, ki se uporablja za popolno karakterizacijo izvora metana. Rezultati izo-topske sestave ogljika in vodika v metanu na aktivnih odkopih kažejo na biogeni izvor, medtem ko CO2 na bakterijski in endogeni izvor, kjer se pojavlja visok indeks CO2-metan (Carbon Dioxide Methane Index). Ključne besede: karakterizacija premogovnih plinov, stabilni izotopi, aktivni odkopi, Velenjski bazen Introduction Unconventional gas resources, including coal bed methane and shale gas, are a growing part of the global energy mix, which has changed the economic and strategic picture for gas consuming and producing countries, including the USA, China and Australia that, together, are responsible for around half the currently recoverable unconventional gas resources[1]. Some other countries, including Canada, India and Indonesia, are also seeking to develop their CBM (Coalbed Methane) and shale gas resources. Exploration was initially focused on mature coals with high contents of what was thought to be thermogenic gas (e.g. Black Warrior, San Juan and Bowen basins). However, CBM production was often hindered by low permeability and mineralization in cleats and fractures, necessitating the development of cost effective horizontal drilling and completion techniques. CBM exploration then extended beyond the producing fields to lower rank coals and subsequently to less mature coal basins, in which it was thought that higher permeability and greater coal thickness would compensate for lower gas contents (e.g. Illinois, Powder River and Surat basins]. Knowledge of the stable isotope geochemistry of coal bed and shale gas and the related production water is essential to determine not only gas origins but also the dominant methanogenic pathway in the case of microbial gas[2]. Evidence for the microbial origin of methane in low rank coals, the presence of secondary microbial gas at shallow levels in higher rank coals and the presence of secondary microbial gas at shallow levels in higher rank coals have been reviewed[3-7], providing an introduction to the use of compositional and isotopic parameters in determining the origin of CBM. However, the timing of methanogenesis, factors governing coal bioavailability and the process of stepwise microbial degradation of coal to methane remain topics of considerable interest to both researchers and industry in the light of possible stimulation of enhanced microbial methane generation from coal[8-10]. In like fashion to CBM, shale gas/coalbed gas can be of thermogenic, microbial or mixed origin, the distinction being made primarily on the basis of the stable isotope compositions of the gases and co-produced waters[11-13]. These studies have shown that the relative proportions of microbial and thermogenic gas are depth related, with deeper locations containing exclusively thermogenic gas and shallower locations a mixture of microbial and thermogenic gases. The ranges in carbon isotope signatures from different sources (gas, solid and liquid phases) are presented in Figure 1. Microbial methane is enriched with 12C, while methane associated with petroleum is treated as thermogenic in origin and is enriched with 13C. Deep sourced CO2 has 613CCO2 values of -7 %o (Figure 1]. C3 plants have A13C values ranging from -30 % to -20 % and C4 plants in the range from -18 % to -10 % (Figure 1]. Vienna Pee Dee Belemnite (VPDB) is the reference material used for determining isotopic composition of carbon (613C), and has a default 5 value of 0 %[14]. Mycrobial methane Methane associated with petroleum Wet gas phases Effect of biodegradation » Petroleum h Geothermal methane Methane CO, C02 from organic matter C3 plants (^plants Deep sourced CO, PeeDee Belemnite Carbonate _i_i_i_I_I_ .»i » 613C%0 Figure 1 : Ranges in carbon isotope signatures from different sources"5. The geochemical and isotopic characterization in the period 2000-2012 of coalbed gases from different excavation fields from Velenje basin have been published[16-20]. Formation waters in Velenje Basin are not trapped in the coal seam, since large amounts of groundwater are extracted from Velenje Basin aquifers to facilitate underground mining of coal. Groundwater recharging the basin is therefore not in direct contact with the coal seam. Hydrogeochemical and isotopic characteristics of Velenje Basin groundwaters and their interactions with surface waters are described in[21]. The aim of this study is to account for the concentration, distribution and origin of coalbed gases ("free gas")[22, 23], including unconventional gas (methane), from the excavation fields K.-65/F, G3/C, K.-65/B, K.-65/E and K.-65/C in mining areas Preloge and Pesje, based on geochemical and isotopie tracers. The new tracer, isotopie composition of methane (6DCH4) was applied in this study for geochemical characterization of the origin of methane and was compared with values from other worldwide sedimentary basins. Materials and methods Coalbed gas was sampled by the operator-miner responsible for underground coalbed gas monitoring. Free gas includes both the vola-tiles filling the pores and cracks within the coal structure and some gas degassed from the coal during drilling and sampling[23]. Long boreholes (25 m] were used for monitoring coal gas concentration; the design of the borehole is described in[20]. Coalbed gases from the lignite seam were sampled at working faces Pes- je K.-65/F, Preloge G3/C, Pesje K.-65/B, Pesje K.-65/E and Pesje K.-65/C and at boreholes jpk 62 + 10°, jpk 56 + 10°, jpk 63 + 10°, jpk 70 + 10°, jpk 73 + 10° (Figure 2). Lignite was produced by the Velenje Longwall Mining Method (VLMM), a longwall top coal caving method, as described in[20]. After drilling, a capillary tube was inserted in boreholes. "Free gas"[23] emitted from the borehole was collected in a plastic syringe 50 ml and transferred to a ampoule 12 ml, which was flushed and filled with coalbed gas under pressure. After sampling the "free gas" from the boreholes, the ampoules were stored under normal atmospheric conditions until analysis. Seven ampoules were filled at each location, the third being analyzed for chemical composition (CO2, CH4, O2, N2, Ar). The other ampoules were stored for determining the isotope composition of carbon in methane (613CCH4), CO2 (613CCO2) and deuterium in methane (6DCH4). Figure 2: Map of sampling locations of coalbed gases from the lignite seam at working faces: Pesje -65/F, Preloge G3/C, Pesje -65/B, Pesje -65/E and Pesje -65/C and boreholes: jpk 62 + 10°, jpk 56 + 10°, jpk 63 + 10°, jpk 70 + 10°, jpk 73 + 10°. Analytical procedure Concentrations of methane, CO2, nitrogen, oxygen and argon were determined using a homemade mass NIER spectrometer as described in detail in[18, 20]. The isotope compositions of methane and CO2 were determined using an Europa 20-20 continuous flow isotope ratio mass spectrometer with an ANCA - TG preparation module. First, water was removed, then CO2 was analyzed directly for 13C content. For methane measurements CO2 was removed and methane then combusted over hot 10 % platinum on CuO (1 000 °C). The methane, completely converted to CO2, was then analyzed directly for the isotopic composition of carbon (513C). Working standards calibrated to IAEA (International Atomic Energy Agency] reference materials were used, with values of -4.3 %o for CO2 and -53.4 %o for methane relative to Vienna Pee Dee Belemnite (VPDB). The analytical precision for carbon isotope composition is estimated to be ±0.2 %o for CO2 and ±0.6 %o for CH4. The stable carbon isotopes are presented in 5 notation relative to VPDB standards and expressed in parts per million[14] as follows[24]: R — R 6(13C/2H]S = s X 1000 [%o] Rrm (1) Where: Rs - ratio 13C/12C in sample Rs - ratio 52H/5H in sample Rrm - ratio 13C/12C and 2H/H in reference material The 2H/H ratio of CH4 was determined using a Thermo Delta XP GC-TC/CF-IRMS coupled to a TRACEGC analyzer at the Istituto Nazionale di Geofisica e Vulcanologia Sezione di Palermo in Italy. The method of 513CCH4 determination is described in detail in[20]. Results and discussion Due to air contamination within the boreholes (used for ventilation of the coalmine) and the capillary system, samples were recalculated on an air-free basis. The percentage of oxygen in the sampled ampoules was used to calculate the amount of nitrogen, according to the ra- tio in air (N2/O2), considering Dalton's law[25]: methane migrates faster than CO2. The major gas components were CO2 and methane. Of all the samples analyzed (n = 72) in this study, seventeen were in excess in N2. Concentrations of CO2 ranged from 36.1 % to 98.2 % and methane from 1.8 % to 63.9 %. Geochemical indices, CDMI ((CO2/(CO2 + CH4) X 100 %) varied from 36.1 % to 98.2 % and stable isotope ratios varied in the following ranges: 513CCO2 from -11.0%o to 1.9 %, 513CCH4 from -71.8 %0 to -43.4 %0, 5DCH4 from -343.9 to -223.1 % and aCO2-CH4 from 1.040 to 1.071 (Table 1). The calculated gas dryness index (Ca/(C2 + C3)) ranged from 339.74 • 10-6 to 23 272.7 • 10-6 and the coal-bed gas is therefore considered as being dry, i.e. methane is the major gas component. The higher hydrocarbons determined in Velenje Basin were ethane, propane, iso-butane, n-butane, iso-pentane, n-pentane and hexane[20]. High correlation (R2 = 0.99] was obtained between methane and CO2 concentrations (Figure 3) in a lignite seam within boreholes: mining area Pesje K.-65/F, jpk 62 + 10°, mining area Preloge G3/C, jpk 56 + 10°, mining area Pesje K.-65/B 63 + 10°, mining area Pesje K.-65/E, jpk 70 + 10°, mining area Pesje K.-65/C, jpk 73 + 10°. Working faces from different mining areas exhibit different geo-chemical and isotopic compositions of coalbed gases, which also depend on the depth of the excavation field (Table 1). In addition, the total gas composition at pre-mined longwall panels and, on the other hand, at longwall panels located under fresh overburden can be different. It is known that the highest concentrations of methane gas are found at longwall panels located under fresh overburden. This occurrence is connected with the more rapid migration of CH4 up through the seam due to its smaller molecular size than that of CO2 and with the fact that CH4 is a free gas in the Velenje lignite structure while CO2 is adsorbed in the matrix of the coal. This fact applies only until the influence of active mining is not shown on the gas composition in the samples. The high advancement rate of up to 9 m per day constitutes a potential for gas outbursts[18]. The rates of advancement of working faces in this study are as follows: Pesje working face K.-65/F, jpk 62 + 10° from 0.9 m/d to 3.5 m/d, Preloge working face G3/C, jpk 56 + 10° from 3 m/d to 4 m/d, working face K.-65/B, jpk 63 + 10° from 0.8 m/d to 1.9 m/d, K.-65/E, jpk 70 + 10° from 1.7 m/d to 3.8 m/d and k.-65/c, jpk 73 + 10° 1.3 m/d (Table 1]. The rate of advancement of the working face is low (range from 0.8 m to 1.9 m] at working face K.-65/B (mining area Pesje), jpk 63 + 10° due to the high CO2 concentration and therefore higher risk of outburst generation. The areas of high and low methane concentrations (Figure 3) are in keeping with the physicochemical properties of CO2 and methane[25]. This pattern is observed for all the excavation fields investigated in this study (Figure 3]. Excavation fields K.-65/F, G3/C, K.-65/E and K.-65/C have higher ranges of gas concentrations than excavation field K.-65/B, where higher CO2 concentration is observed (Figure 3). Table 1: Geochemical ccomposition of coalbed gases and isotopie compositions of carbon in CO2 (öuCCOJ and methane (ö"CCH4) at working faces K.-65/F, G3/C, K.-65/B, K.-65/E, K.-65/C within boreholes jpk 62 + 10°, jpk 56 + 10°, jpk 63 + 10°, jpk 70 + 10°, jpk 73 + 10°. CDMI index = (COJ(CO2 + CH4) x 100 %), òDcm (isotopic composition of deuterium in methane) and aCH4-CO2 (fractionation factor between CH4 and CO 2) Working face -65F, borehole jpk 62 + 10°, mining area Pesje, z = - 74 m Date of sampling Distance of the working face (m) CH4 (vol. %) CO2 (vol. %) N2 (vol. %) CDMI index (%) 813C CO2 (%%) 813C CH4 (%) 8D CH4 (%) aCO2-CH4 January 4, 2013 190 42.5 57.5 0.0 57.5 -8.9 -52.6 -294.5 1.046 January 9, 2013 174.4 43.3 56.7 0.0 56.7 -8.8 -52.7 -296.3 1.046 January 15, 2013 156.2 43.2 56.8 0.0 56.8 -8.9 -52.3 -296.4 1.046 January 22, 2013 131.3 44.9 55.1 0.0 55.1 -9.1 -52.9 -293.6 1.046 January 28, 2013 112.1 43.3 56.7 0.0 56.7 -9.1 -54.1 -298.8 1.048 February 6, 2013 70.9 42.9 57.1 0.0 57.1 -8.9 -52.9 -294.9 1.046 February 11, 2013 59.8 53.0 47.0 0.0 47.0 -9.2 -53.2 -294.2 1.046 February 19, 2013 34.7 24.5 75.5 0.0 75.5 -9.0 -54.5 -292.6 1.048 February 20, 2013 31.2 24.0 76.0 0.0 76.0 -9.3 -55.0 -291.4 1.048 February 21, 2013 27.4 19.4 80.6 0.0 80.6 -9.1 -54.8 -292.4 1.048 February 22, 2013 25 15.9 84.1 0.0 84.1 -9.2 -55.0 -293.0 1.048 February 23, 2013 ' 21.4 19.7 80.3 0.0 80.3 -9.3 -54.7 -294.0 1.048 February 25, 2013 ' 17.9 19.1 80.9 0.0 80.9 -8.9 -54.3 -293.3 1.048 February 26, 2013 ' 14.4 19.8 80.2 0.0 80.2 -8.6 -53.5 -295.8 1.047 February 27, 2013 ' 11.1 20.4 79.6 0.0 79.6 -8.9 -54.5 -291.8 1.048 March 1, 2013 4.6 17.7 82.3 0.0 82.3 -8.8 -54.3 -294.6 1.048 March 4, 2013 1.8 13.6 86.4 0.0 86.4 -8.5 -53.2 -298.6 1.047 Working face G3/C, borehole jpk 56 + 10°, mining area Preloge, z = -46 m Date of sampling Distance of the working face (m) CH4 (vol. %) CO2 (vol. %) N2 (vol. %) CDMI index (%) S13C CO2 (%%) S13C CH4 (%) SD CH4 (%) aCO2-CH4 February 22, 2013 ' 189.9 63.6 36.4 0.0 36.4 -5.1 -43.6 1.040 February 25, 2013 ' 179.6 63.9 36.1 0.0 36.1 -2.4 -43.4 -343.5 1.043 March 8, 2013 125.4 22.2 77.8 0.0 77.8 -9.3 -54.5 -294.0 1.048 March 13, 2013 104.9 22.4 77.6 0.0 77.6 -8.8 -53.9 -292.9 1.048 March 27, 2013 60 41.5 58.5 0.0 58.5 -8.9 -52.6 -296.8 1.046 April 4, 2013 37.4 61.9 38.1 0.0 38.1 -5.4 -46.3 -343.9 1.043 April 15, 2013 19.6 36.4 63.6 0.0 63.6 -4.5 -45.6 -333.3 1.043 April 17, 2013 13.9 33.8 66.2 0.0 66.2 -4.0 -48.6 -326.4 1.047 April 19, 2013 7.3 12.0 88.0 0.0 88.0 -11.0 -71.8 1.066 April 22, 2013 3.9 40.8 59.2 0.0 59.2 Working face - 65/B, borehole jpk 63 + 10° , mining area Pesje, z = -67 m Date of sampling Distance of the working face (m) CH4 (vol. %) CO2 (vol. %) N2 (vol. %) CDMI index (%) S13C CO2 (%) S13C CH4 (%) SDCH4 (%) aCO2-CH4 March 27, 2013 200 6.0 94.0 0.0 94.0 -2.1 -55.5 -321.3 1.057 May 8, 2013 171.3 5.3 94.7 0.0 94.7 -2.9 -57.1 -310.4 1.057 May 16, 2013 155.5 4.9 95.1 0.0 95.1 -2.8 -55.9 -307.3 1.056 May 22, 2013 145.7 5.9 94.1 0.0 94.1 -2.4 -55.8 -223.1 1.057 May 30, 2013 131.2 4.0 96.0 0.0 96.0 -2.6 -57.9 -301.3 1.059 June 6, 2013 115.3 4.1 95.9 0.0 95.9 -2.5 -58.7 -299.2 1.060 June 12, 2013 103.8 4.5 95.5 0.0 95.5 -2.3 -58.8 -301.7 1.060 June 20, 2013 89.4 July 5, 2013 64.4 4.3 95.7 0.0 95.7 1.9 -64.4 -279.6 1.071 July 9, 2013 59.1 3.8 96.2 0.0 96.2 -2.1 -64.1 -269.2 1.066 July 10, 2013 56.5 4.0 96.0 0.0 96.0 -2.1 July 15, 2013 48.3 3.6 96.4 0.0 96.4 -2.0 -64.2 -273.3 1.067 July 16, 2015 45.6 3.8 96.2 0.0 96.2 -2.1 -64.2 -271.5 1.066 July 17, 2013 44 2.2 97.8 0.0 97.8 -2.0 -64.0 -271.4 1.066 July 18, 2013 41 3.7 96.0 0.0 96.3 -2.0 -63.0 -272.0 1.065 July 19, 2013 37.5 4.1 95.9 0.0 95.9 -2.1 -63.4 -277.2 1.065 July 22, 2013 33.2 4.5 95.5 0.0 95.5 -2.0 -62.9 -278.4 1.065 July 23, 2013 31.9 4.0 96.0 0.0 96.0 -1.8 -62.3 -291.7 1.064 July 24, 2013 29 2.7 97.3 0.0 97.3 -2.1 -62.2 -285.4 1.064 July 25, 2013 27 4.2 95.8 0.0 95.8 -0.3 -51.2 -263.7 1.054 July 26, 2013 24.2 7.5 92.3 0.2 92.3 -2.0 -65.6 -258.6 1.068 July 29, 2013 21.7 1.8 98.2 0.0 98.2 -2.1 -65.9 -255.3 1.068 Working face - 65/E, borehole jpk 70 + 10° , mining area Pesje, z = -78 m Date of sampling Distance of the working face (m) CH4 (vol. %) CO2 (vol. %) N2 (vol. %) CDMI index (%) S13C CO2 (%) S13C CH4 (%) SDCH4 (%) aCO2-CH4 October 3, 2013 i 152.4 57.6 42.2 0.2 42.3 -8.8 -50.3 -301.1 1.044 October 8, 2013 i 138.5 57.5 42.5 0.0 42.5 -8.4 -50.2 -300.1 1.044 October 15, 2013 123.1 6.8 93.2 0.0 93.2 -8.3 -52.0 -319.8 1.046 October 22, 2013 105.1 58.5 41.5 0.0 41.5 -8.7 -50.9 -300.9 1.044 November 5, 2013 76.7 56.8 41.7 1.5 42.3 -8.5 -50.5 -298.8 1.044 November 8, 2013 65.0 55.1 43.9 1.0 44.3 -8.8 -50.3 -299.2 1.044 November 12, 2013 55.1 55.6 44.1 0.3 44.2 -8.3 -50.9 -297.9 1.045 November 14, 2013 48.6 59.9 38.9 1.2 39.4 -8.2 -51.2 -298.7 1.045 November 18, 2013 37.2 51.8 45.1 3.1 46.5 -8.2 -50.8 -300.0 1.045 November 19, 2013 33.4 54.5 44.2 1.3 44.8 -8.2 -50.8 -300.5 1.045 November 20, 2013 30.3 51.1 46.6 2.3 47.7 -8.4 -50.8 -300.4 1.045 November 21, 2013 27.9 43.2 56.2 0.6 56.5 -8.4 -50.6 1.044 November 22, 2013 26.7 35.2 63.6 1.2 64.4 -8.4 -50.6 -299.7 1.044 November 25, 2013 20.9 21.6 77.7 0.7 78.2 -8.8 -50.4 -300.2 1.044 November 26, 2013 19.2 11.7 87.4 0.9 88.2 -8.6 -50.7 -303.7 1.044 November 27, 2013 17.4 13.5 84.9 1.6 86.3 -8.3 -50.5 -301.0 1.044 November 28, 2013 15.0 12.5 87.5 0.0 87.5 -8.6 -50.1 -305.4 1.044 November 29, 2013 12.2 10.9 89.1 0.0 89.1 -8.5 -49.9 -300.0 1.044 December 2, 2013 9.0 15.4 83.5 1.1 84.4 -8.4 -52.3 -299.4 1.046 December 3, 2013 6.5 46.2 53.8 0.0 53.8 -9.2 -47.4 -295.8 1.040 December 4, 2013 3.2 25.6 74.4 0.0 74.4 -8.5 -52.3 -286.5 1.046 Working face -65/C, borehole jpk 73 + 10°, mining area Pesje, z = -69 m Date of sampling Distance of the working face (m) CH4 (vol. %) CO2 (vol. %) N2 (vol. %) CDMI index (%) 813C CO2 (%%) 813C CH4 (%) 8D CH4 (%) aCO2-CH4 November 20, 2013 152.4 46.2 53.8 0.0 53.8 -8.4 -50.2 -298.9 1.044 December 5, 2013 133.8 7.5 91.7 0.8 92.4 -8.8 -52.4 -293.4 1.046 Figure 3: Concentration of CO2 versus concentration of methane in lignite seams in advancement of the working faces (see Figure 1) within boreholes: jpk 62, mining area Pesje, jpk 56, mining area Preloge, jpk 63, mining area Pesje, jpk 70, mining area Pesje, jpk 73 mining area Pesje. 28 Approaching of the working face also influences the stress situation at working faces. During the excavation of lignite, secondary fissures are generated due to rearrangement of primary stress conditions, enabling release of coalbed gas preserved in pores of the coalbed reservoir[18]. The results of temporal variations of CO2 and methane within boreholes at working faces 120/B, G2/C and-50/B[18] and working faces K.-5/A, G2/C, K. -130/A, K. -50/C[20] reveal similar trends. The observed maximum concentrations of methane at distances of 177 m and 122 m from the working face coincide with the minimum concentrations of CO2, but the differences found in coalbed gas composition are due to location of the longwall panels that underlie the unmined area or previously mined areas[20]. No trend is observed in CO2 and methane concentrations in the advancement of the longwall K.-65/F (jpk 62 + 10°] at a distance of longwall 120 m, as is usually observed when investigating coalbed gas concentrations in advancement of a working face. Overall, high CH4 gas content (up to 53 vol. %) was observed at the longwall panel K.-65/F, which can be explained by the longwall panel position just under fresh overburden. The highest concentrations of CO2 and minimal concentrations of methane are observed at a distance of the longwall K.-65/F of 40 m (Figure 4]. Maximum CO2 concentration is observed at a distance of 120 m from the longwall face G3/C (jpk 56], coinciding with the minimum CH4 concentration, while the maximum CH4 concentration is observed at a distance of around 40 m from the longwall and another maximum of CO2 is observed at the distance of the longwall 20 m (Figure 4]. At the excavation field K.-65/B, jpk 63 + 10° there is no change in CO2 and methane concentrations in advancement of the working face; no data concerning gas concentration is available between distances of 110 m and 70 m to allow a clearer interpretation. This excavation field lies in an area of pre-mined excavation and has a high CDMI index (Figure 4). Data relating to the coalbed gas composition in advancement of the working face -65/C (Figure 4] are not presented since only 2 values of gas concentration were obtained. The first maximum of CO2 concentration (minimum of meth- -CH4, -C02, CH4, -C02, -CH4, C02, -CH4, -C02, JPK62 JPK62 JPK 56 JPK 56 JPK 63 JPK 63 JPK 70 JPK 70 250 200 150 100 Distance of the working face ( m) 50 Figure 4: Concentration of methane and CO2 versus distance from the longwall of the working faces (-65/F, G3/C, -65/B, -65/E, -65/C) in boreholes jpk 62, jpk 56, jpk 63 and jpk 70. ane concentration] occurs at a distance of 120 m from the working face (K.-65/E, jpk 70), and, at 30 m from the working face, lignite structure was demolished (Figure 4). The longwall panel in this case was excavated under fresh overburden, which explains the high methane content compared with that of the longwall panels G3/C and K.-65/B. The methane gas content was high at 59.9 vol. %. Results from previous studies investigating coalbed gases from Velenje Basin showed that CH4 desorbs more quickly than CO2, but this is not the case from the observations of gas composition at excavation fields in this study. The reaction mechanisms of methane and CO2 generation in natural and coalbed gases are described elsewhere[3, 4' 18]. The products of these reactions support a variety of methanogens. For Velenje Basin it is known that the basin was formed in a Ca-alkaline rich environment[26], which enabled the operation of bacteria (aerobic and anaerobic), with the generation of coal-bed gas in lignite (organic rich matrix). The plots of 613CCH4 versus б 13C and 613C CCO2 and б CCH4 the origin of methane. Figures 4 and 5, in the Velenje Basin, indicate the consecutive origins of methane: microbial CO2 reduction with 613CCH4 values from -40 %o to -50 %o, microbial acetate fermentation with 613CCH4 less than -50 % and mixed origin between these two[5]. Enrichment with 13C in methane could be also due to microbial oxidation of methane, which results in enrichment of residual methane with the 13C isotope and depletion of 12C in generated CO2 (Figure 5). The method for determining the origin of methane and for distinguishing pathways of microbial methane generation (i.e. acetate fermentation from CO2 reduction) was proposed by[3] and uses the a13CCO2-CH4 fractionation factor; this was also applied in our study at five excavation fields to distinguish between the two pathways. The a13C„, „,,„ values indicate that meth- J CO2-CH4 ane was generated via both these processes at all five excavation fields investigated (K.-65/F, G3/C, K.-65/B, K.-65/E and K.-65/C) since the calculated a_n, ru„ values were below 1.06 CO2-CH4 (Table 1). Altogether, 13 samples out of a total versus 6D (Figures 5, 6) were used to explain of 72 with low values of 613C (>-60%o), and 813C, -15 -13 -11 -9 -7 C02 -5 -3 -1 microbial C02 reduction microbial C02 acetate fermentation ♦ JPK 62, mining area Pesje ■ JPK 56, mining area Preloge a JPK 63, mining area Pesje X JPK 70, mining area Pesje ж JPK 73, mining area Pesje -30 -40 -50 О VP oS •a- -60 £ iH to -70 -80 -90 Figure 5: Interpretation of the origin of methane in the Velenje basin using 6,3CC-4 versus Ö 13Cco2 in a lignite seam at boreholes (jpk 62, jpk 56, jpk 63, jpk 70 and jpk 73) ahead of the working faces (-65/F, G3/C, -65/B, -65/E, -65/C). corresponding values of 613CCO2 from -12 %o to 0 %, indicate that the methane originates from microbial CO2 reduction. Following13, 4] and using aCO2 CH4 as a parameter, most of the samples investigated in our study have aCO2-CH4 < 1.06, which is characteristic of acetate fermentation. When interpreting digenetic methanogenic pathways, it should be emphasized that the threshold for acetate fermentation/CO2 reduction was -50 %. We obtained ranges of 613CCH4 and 613CCO2 for Velenje basin similar to those obtained for Sydney and Bowen basins (Figure 1). To obtain a better geochemical characterization of methane in Velenje Basin, we used 6DCH4 tracer (Figure 6). For coalbed gases from excavation fields K.-65/F, G3/C, K.-65/B, K.-65/E and K.-65/C we conclude that parameters 613CCH4 and 6DCH4 both reveal microbial methane and mixing and none of the samples investigated in this study from excavation fields have the value for thermogenic methane (Figure 6). Plots of 613Cru„ versus 6Dru„ from the excava- CH4 CH4 tion fields K.-65/F, G3/C, K.-65/B, K.-65/E and K.-65/C (Figure 7) show that coalbed samples for Velenje Basin are similar to those for Powder River, Bowen Basin, Antrim/New Albany shales and Appalachian shales, and indicate fields of microbial methane and mixing, with values of methane enriched with 13C (values of thermogenic methane). From the study of coalbed gases[20] from years 2010 to 2012 from Velenje Basin it was found that the characteristics of methane from the Velenje Basin are similar to those from low rank coal basins such as Powder River Basin and the San Juan Basin, USA, while completely distinct from those of the Lower and Upper Silesian Basins, Poland. The origin of CO2 is interesting for its relation to outbursts in coalmines. It is known that high concentrations of carbon dioxide in seam gases in Australian coals occur in regions of igneous activity and associated faulting and this component has been described as being of presumed pneumatolytic origin[33]. We used the plot of 613CCO2 versus CDMI (Figure 8) to explain the origin of CO2 at the working faces investigated: K.-65/F, G3/C, K.-65/B, K.-65/E and K.-65/C (Figure 8). Samples from excavation field -65/B are seen to have values of 613CCO2 indicative of endogenic origin, while samples from excavation fields -65/F, G3/C, -65/E and -65/C have values characteristic of endogenic and bacterial -400 -350 -300 -250 5DCH4 (%„) -200 -150 -100 -50 Values characteristic of thermogenic methane ♦ JPK62, mining area Pesje □ JPK56, mining area Pesje JPK63, mining area Pesje X JPK 70, mining area Pesje ж JPK 73, mining area Pesje Values of microbial methane and mixing -10 -20 -30. -40 u О -50 -60 -70 -80 Figure 6: Interpretation of the origin of methane using 6'3C versus 6D in a lignite seam in boreholes (jpk 62, jpk 56, jpk 63, jpk 70 and jpk 73) ahead of the working faces (K.-65/F, G3/C, K.-65/B, K.-65/E, K.-65/C). я- COj reduction microbial methane 80 • aa DJä "SuratSasn Appalachian di ales ' о Antrim/New Albany dides □ Powder River Basin -10 -1-'-1-1-1- -400 -350 -300 -250 -200 -150 -100 6D-CH4 (%c) Figure 7: Isotopic composition of methane (6'3CCH4) versus that of deuterium in methane (òùCH4) in coalbed seam in coalmines from Bowen and Surat basins (Moranbah Gas Project[27]; Oaky Creek[2]; Avon Downs[28]. Gas samples from Powder River basins2,30], Antrim and New Albany shales"13'] and organic rich shales from North Appalachi basin[32]. In the frame, results of 6'3CCH4 and öDCH4 (see Table ' for results) from this study (excavation fields: -K.65/F, G3/C, K.-65/B, K.-65/E, K.-65/C within boreholes: jpk 62, jpk 56, jpk 63, jpk 70, jpk 73) are presented. gas formed via CO2 reduction. Typical endogenic values of 613CCO2 are about -7 %o[23] and areas of high CDMI index could be related to tectonics of the Šoštanj and Smrekovec faults systems at the time of formation of the basin[18]. The CDMI index is used as the crucial parameter for excavation fields prone to outburst generation, especially in combination with other factors such as fault zones, detrital lignite, etc. Previous studies have shown that shale gas can be of thermogenic, microbial or mixed origin, the distinction being made on the basis of the stable isotope composition of the gases and co-produced water e.g. Antrim Shale, Michigan Basin[11' 12], Barnett Shale, Fort Worth Basin, Appalachian Basin[32] and other basins studied such as New Albany Shale, Illinois Basin, Colorado Group, Western Canada Sedimentary Basin[2]. All these studies showed that the relative proportions of microbial and thermogenic gas are depth related; deeper locations contain exclusively thermogenic gas, while shallower locations contain a mixture of microbial and thermogenic methane. In the Velenje Basin, gas concentration, 613CCH4 and 613CCO2 also differ with the depth level of excavation fields (Figure 9). For Velenje Basin it is very important which excavation fields underlie the unmined or previously mined areas[20]. A high CDMI index, with values in the range of 92.5 % to 98.2 %, was observed at the excavation field K.-65/B, jpk 63 + 10° that lies under a pre-mined area. G3/C (jpk 56 + 10°) excavation field is the shallowest field to be investigated and has a larger range of all measured parameters (CO2, CH4, 613CCO2 and 613CCH4) in comparison to the other excavation fields observed in this study - K.-65/F, K.-65/E, K.-65/C (Figure 9 A-D). The 613CCO2 values show that the deepest excavation fields (K.-65/E - jpk 70, K.-65/F - jpk 62, mining area Pesje) have typical endogenic CO2 with CO2 reduction values. The excavation field G3/C, jpk 56, located at z [-46 m], had the broadest range of 613CCO2, indicating biogenic and endogenic CO2. We have not enough data to make a clear interpretation for K.-65/C, jpk 73 + 10° (mining area Pesje). A similar pattern for 613CCH4 values versus depth is observed, the widest range being for G3/C, [z = -46 m] (Figure 9 C). Figure 8: Interpretation of the origin of C02 at working faces using àuCc02 versus CDMI index in a lignite seam in boreholes (jpk 62, jpk 56, jpk 63, jpk 70 and jpk 73) ahead of the working faces (K.-65/F, G3/C, K.-65/B, K.-65/E, K.-65/C). • JPK 62, mining area Pesje C • JPK 56, mining area Preloge • JPK 63, mining area Pesje • JPK 70, mining area Pesje • JPK 73, mining area Pesje —_ i Figure 9: Concentrations of CH4 (A), C02 (B), ö'3CCH4 (C) and ö'3CC02 (D) as a function of depth (m) for excavation fields K.-65/F, jpk 56 + 10°, z [-74 m] , G3/C, jpk 56 + 10°, z [-46 m], K.-65/B, jpk 63 + 10°, z [-67 m], K.-65/E, jpk 70 + 10°, z [-78 m] K.-65/C, jpk 73 + 10° and z [-69 m]). Conclusions Geochemical (CH4, CO2, N2) and isotopic (613CCO2, 613CCH4, 6DCH4] tracers have provided a geochemical and isotopic characterization of coalbed gases from five excavation fields at Preloge and Pesje (Velenje Basin] during the year 2013. Isotopic composition of deuterium of methane (5DCH4) has been used to determine the origin of methane. Further, these characteristics of methane have been compared with those observed in other coal sedimentary basins worldwide. In shallower working faces (Preloge G3/C, jpk 56 ([z = -46 m]) and in the deepest working face (jpk 70, mining area Pesje K.-65/E, [z = -78 m]) larger ranges of methane and CO2 concentrations have been observed. The concentrations of CO2 and CH4 are relatively constant at the excavation field K.-65/B (jpk 63 + 10°]. At the deepest excavation field K.-65/E, jpk 70, depth [z] = -78 m, 613CCO2 values denote an endogenic origin of CO2 while methane, with 613CCH4 values corresponding to thermogenic methane, probably results from CO2 reduction. The wider range of 613CCH4 (biogenic and thermogenic isotopic values] and 613CCO2 (biogenic and endogenic isotopic values] are observed at the shallower excavation field G3/C, jpk 56, mining area Preloge [depth -46 m]. Each excavation field appears to be specific in terms of the geochemical characteristics of coalbed gases, making it necessary to carry out such a geochemical and isotopic investigation of coalbed gases in order to more fully understand the origin of gases, especially methane, which is considered as a greenhouse and unconventional gas. Such investigations are relevant to the development of clean coal technologies such as CBM (coalbed methane], which could have great potential when cessation of underground coal excavation is being considered. The determination of concentrations and stable isotope patterns of coalbed gases at working faces has enabled an interpretation of the origin of coalbed gases in the Velenje basin. From the results of this and previous studies, it can be concluded that coalbed CO2 from these working faces is endogenic and of bacterial ori- gin. Measurements of 613CCH4 and 6DCH4 have led to the conclusion that methane from Velenje Basin has a microbial origin, formed through microbial fermentation and/or CO2 reduction, and of mixed origin, both thermogenic and biogenic, since some 613CCH4 values are enriched with 13C isotope. Secondary processes like migration, adsorption/desorption and mixing of gases of different origin during excavation complicate the interpretation of gas origin and have been neglected in interpretations of the origin of coalbed gases. Acknowledgements This study was conducted in the framework of project L1-5451 funded by the Slovenian Research Agency (ARRS) and Velenje Coalmine d.d. The authors are grateful to Mr. Ivo Zadnik for technical support and assistance in the field sampling. References [1] International energy agency (2011): Golden rules for a golden age of gas. World Energy Outlook Special Report on Unconventional Gas, 143 p. [2] Golding, S. D., Boreham, C. J., Esterle, J. S. (2013): Stable isotope geochemistry of coal bed and shale gas and related production waters: A review. International Journal of Coal Geology, 120, pp. 24-40. [3] Whiticar, M. J. (1996): Stable isotope geochemistry of coals, humic kerogens and related natural gases. International Journal of Coal Geology, 32, pp. 191-215. [4] Whiticar, M. J. (1999): Carbon and hydrogen isotope systematics of bacterial formation and oxidation of methane. Chemical Geology, 161, pp. 291-314. [5] Clayton, J. L. (1998): Geochemistry of coalbed gas -A review. International Journal of Coal Geology, 35, pp. 159-173. [6] Faiz, M, Hendry, P. (2006): Significance of microbial activity in Australian coal bed methane reservoirs-a review. Bull. Can. Pet. Geol., 54, pp. 261-272. [7] Strapoć, D., Mastalerz, M., Dawson, K., Macala-dy, J., Callaghann, A. V., Wawrick, B., Turich, C., Ashby, M. (2011): Biogeochemistry of microbial coal-bed methane. Ann. Rev. Earth Planet. Sci., 39, pp. 617-656. [8] Scott, A. R. (1999): Improving gas recovery with microbially enhanced coalbed methane. In: Mastalerz, M., Glikson, M., Golding, S. D. (Eds.), Coalbed Methane: Scientific Environmental and Economic Evaluation. Kluwer, Dordrecht, pp. 89-110. [9] Furmann, A., Schimmelmann, A., Brassell, S. C., Mastalerz, M., Picardal, F. (2013): Chemical compound classes supporting microbial methanogenesis in coal. Chemical geology, 339, pp. 226-241. [10] Penner, C. J., Foght, J. M., Budwill, K. (2010): Micro-bila diversity of western Canadiansubsurface coal beds and methanogenic coal enrichment cultures. International Journal of Coal Geology, 82, pp. 81-93. [11] Martini, A. M., Walter, L. M., Budai, J. M., Ku, T. C. W., Kaiser, C. J., Schoell, M. (1998): Genetic and temporal relations between formation waters and biogenic methane: Upper Devonian Antrim Shale, Michigan Basin, USA. Geochimica Cosmochimica Acta, 62, pp. 1699-1720. [12] Martini, A. M., Walter, L. M., Budai, J. M., Mcintosh, J. C., Schoell, M. (2003): Microbial production and modification of gases in sedimentary basins: a geochemical case study from a Devonian shale gas play, Michigan basin. AAPG Bull., 87, pp. 1355-1375. [13] Schlegel, M. E., Mcintosh, J. C., Bates, B. I., Kirk, M. F., Martini, A. M. (2011): Comparison of fluid geochemistry and microbiology of multiple organic-rich reservoirs in the Illinois Basin, USA: evidence for controls on methanogenesis and microbial transport. Geochimica Cosmochimica Acta, 75, pp. 1903-1919. [14] Coplen, T. B. (1996): New guidenices for reporting stable hydrogen, carbon and oxygen isotopes ratio data. Geochimica Cosmochimica Acta, 60, pp. 390-3360. [15] PETROLEUM GEOCHEMISTRY GROUP, CSIRO [online]. Gas composition and isotope geochemistry. [cited 6/12/2000]. Available on: . [16] Kanduč, T. (2004): Isotopic characteristics of coalbed gases in Velenje Basin. Master Thesis, Ljubljana, University of Ljubljana, pp. 78 (in Slovene). [17] Kanduč, T., Pezdič, J. (2005): Origin and distribution of coalbed gases from the Velenje Basin, Slovenia. Geochemical Journal, 39, pp. 397-409. [18] Kanduč, T., Žula, J., Zavšek, S. (2011): Tracing coalbed gas dynamics and origin of gases in advancement of the working faces at mining areas Preloge and Pesje, Velenje basin. RMZ-Mater. Geoenvironment, 58, pp. 373-288. [19] Kanduč, T., Markič, M., Zavšek, S., McIntosh, J. (2012): Methanogenesis in the Pliocene Velenje Coal Basin, Slovenia, inferred from stable carbon isotopes. International Journal of Coal Geology, 89, 70-83. [20] Lazar, J., Kanduč, T., Jamnikar, S., Grassa, F., Zavšek, S. (2014): Distribution, composition and origin of coal-bed gases in excavation fields from the Pesje mining areas, Velenje Basin, Slovenia, International Journal of Coal Geology, 131, pp. 363-377. [21] Kanduč, T., Grassa, F., McIntosh, J., Stibilj, V., Ul-rich-Supovec, M., Supovec, I., Jamnikar, S. (2014): A geochemical and stable isotope investigation of groundwater/surface-water interactions in the Velenje Basin, Slovenia. HydrogeologyJournal, 22, pp. 971-984. [22] Kotarba, M. (1990): Isotopic geochemistry and habitat of the natural gases from the Upper Carboniferous Zacler coal - bearing formation in the Nowa Ruda coal district (Lower Silesia, Poland). Organic Geochemistry, 16, pp. 549-560 [23] Kotarba, M. J. (2001): Composition and origin of coal-bed gases in the Upper Silesian and Lublin basins, Poland. Organic Geochemistry, 32, pp. 163-180. [24] O'Neil, J. R. (1979): Stable Isotope Geochemistry of Rocks and Minerals. - V: Lectures in Isotope Geology, Jager, E., Hunzinger, J. C., (Eds). - Springer Verlag, Berlin, 235-263. [25] Atkins, P. W. (1994): Physical Chemistry, fifth ed. Oxford Univ. Press, Oxford, 1031 p. [26] Markič, M., Sachsenhofer, R. F. (1997): Petrographic composition and depositional environments of the Pliocene Velenje lignite seam (Slovenia). International Journal of Coal Geology, 33, pp. 229-254. [27] Kinnon, E. C. P., Golding, S. D., Boreham, C. J., Baublys, K. A., Esterle, J. S. (2010): Stable isotope and water quality analysis of coal bed methane roduction waters and gases from the Bowen Basin, Australia. International Journal of Coal Geology, 82, pp. 219-231. [28] Draper, J. J., Boreham, C. J. (2006): Geological controls on exploitable coal seam gas distribution in Queensland. Aust. Prod. Pet. Explor. Assoc. J., 46, pp. 343-366. [29] Bates, B. I., McIntosh, J. C., Lohse, K. A., Brooks, P.D. (2011): Influence of groundwater flowpaths, residence times, and nutrients on the extent of microbial methanogenesis in coal beds: Powder River Basin, USA. Chemical geology, 284, pp. 45-61. [30] Flores, R. M., Rice, C. A., Stricker, G. D., Warden, A., Ellis, M. S. (2008): Methanogenic pathways of coal-bed gas in Powder River Basin, United States: the geological factor. International Journal of Coal Geology, 76, pp. 52-75. [31] Mcintosh, J. C., Walter, L. M., Martini, A. M. (2002): Pleistocene recharge to mid-continent basins: effects on salinity structure and microbial gas generation. Geochimica Cosmochimica Acta, 66, pp. 1681-1700. [32] Osborn, S. G., Mcintosh, J. C. (2010): Chemical and isotopic tracers of the contribution of microbial gas in Devonian organic-rich shales and reservoir sandstones, northern Appalachian Basin. Applied Geochemistry, 25, pp. 456-471. [33] Smith, J. W., Gould, K. W. (1980): An isotopic study of the role of carbon dioxide in outbursts in coal mines. Geochemical Journal, 14, pp. 27-32. Original scientific paper Received: November 26, 2014 Accepted: February 2, 2015 Indicator Kriging porosity maps of Upper Miocene sandstones, Sava Depression, Northern Croatia Indikatorsko krigiranje poroznosti zgornjemiocenskih peščenjakov v Posavski kadunji v severni Hrvaški Tomislav Malvić1, Karolina Novak2*, Kristina Novak Zelenika1 1INA-Industry of Oil Plc., Šubićeva 29, Zagreb, Croatia 2INA-Industry of Oil Plc., Av. V. Holjevca 10, Zagreb, Croatia 3Faculty of Mining, Geology and Petroleum Engineering, Pierottijeva 6, Zagreb, Croatia Corresponding author. E-mail: karolina.novak@ina.hr Abstract There are geologically and geostatistically analysed two, presently, the most interesting Upper Miocene sandstone Croatian hydrocarbon reservoirs in the Sava Depression regarding injection of carbon-dioxide. The intention is increasing oil recovery and, simultaneously, keeps part of injected CO2 permanently in depleted reservoirs. Lithologically, both reservoirs are domi-nantly fine to medium-grained sandstones, with very restricted pelitic content, where silt, clay and marl fractions are increased only in laterally marginal parts. Lithostratigraphically (informally) those two reservoirs are named as "Gamma 3" (older, here analysed as reservoir "A") and "Gamma 4" (younger, here as reservoir "B"), both belonging to the Iva Member, i.e. the Ivanić Grad Formation. Indicator Kriging had been used to map probabilities to obtain different porosities that indicated on different lithofacies. Key words: sandstones, Indicator Kriging, porosity, Upper Miocene, Sava Depression, Croatia Izvleček V dveh na Hrvaškem sedaj najobetavnejših zgornjemiocenskih rezervoarjih ogljikovodikov so geološko in geostatistično preiskali možnost vtiskanja ogljikovega dioksida. Cilj je bil povečati izkoristek črpanja nafte, hkrati pa v izpraznjenih rezervoarjih trajno uskladiščiti del vtisnjenega CO2. Litološko sta oba rezervoarja v pretežno drobno do srednjezrnatih peščenjakih z zelo podrejenim pelitskim deležem, pri čemer so meljaste, glinene in laporne frakcije znatneje zastopane le bočno v obrobnih delih. Na osnovi litostratigrafije so rezervoarja neformalno poimenovali "Gama 3" (starejši, označen kot rezervoar "A") in "Gama 4" (mlajši, označeni kot rezervoar "B"). Oba pripadata členu Iva, tj. Ivanićgradski formaciji. Za kartiranje verjetnosti nastopanja različnih poroznosti, ki indicirajo različne lito-faciese, so uporabili metodo indikatorskega krigiranja. Ključne besede: peščenjaki, indikatorsko krigiranje, poroznost, zgornji miocen, Posavska kadunja, Hrvaška Introduction In Croatia, the projects for CO2 transport and storage can be successfully applied in order to (a) increase oil recovery from "mature" sandstone reservoirs and on a commercial basis, (b) permanent store CO2 with the purpose of greenhouse gas (abbr. GHG] emission reduction. The CO2 generated from combustion of fossil fuels or produced during production of oil or natural gas is injected into hydrocarbon reservoirs with the aim of increasing oil recovery for more then 30 years, where recovery could be averagely increased for 10-14 %[1, 2]. Currently several dozen of the Enhanced Oil Recovery (abbr. E OR] projects are operating worldwide[3-5]. The CO2 trapping in the subsurface depends on reservoir pressure, temperature and properties of cap rock. Reservoirs suitable for application of the EOR methods could be described by the following characteristics[6-8]: (a) oil viscosity < 12 mPa s, (b) oil density 825-865 kg/m3, (c) residual oil saturation, Sor > 25 %, (d) initial reservoir pressure > 100 bar. In addition CO2 mineral trapping can increase volume of trapped gas for few percentages[9]. Generally, sediments deeper than 800 m provide natural reservoir pressure sufficient to keep the CO2 in supercritical state. The quality of cap rocks is very important factor for fluid storage into subsurface reservoirs[10]. Assessment of applicability of CO2 injection into Croatian oil fields was conducted in the '70s and '80s of the last century, mostly through a pilot project on limited part of the Ivanić Field[11-13]. Its results led to a decision on implementation of the EOR project, but also initiated further geological modelling of sandstone reservoirs possibly and partially saturated with CO2. In the case of the Ivanić Field considered are medium to fine-grained sandstone reservoirs, more than 10 m thick, and still in production as oil reservoirs (but largely exhausted). The age is Upper Pannonian. There is planned to inject 400 000 m3/d of CO2 collected onto gas-processing plant Molve. The results could be twofold: (a) enhanced oil recovery and (b) reducing of CO2 emission[14, 15]. A model of injected gas behaviour is based on an analysis of several geological variables, including fluid properties and chemical reactions. However, it is important to understand distribution of porosity and permeability, as probably two important variables, within the reservoir. Geostatistical methods for reservoir mapping could be divided into deterministical and stochastical ones. Deterministical provide "the most probable" solution from available data, sometimes described like in Kriging based simulation as the median or "zero-realisation" when it is later applied in simulation. The most probable means that there are also other possibilities in representation of surface, which are "located" in uncertainty range typical for input dataset. It resulted in set of equiprobable realisations where selection of representative ones is based on several techniques, but no one of them favour any realisation as more probable. This study shows applications of Indicator Kriging (abbr. IK] probability maps, as deterministical method, in estimation of porosity and permeability. Such data are derived from Upper Miocene sandstone reservoirs into the Ivanić Field, planned for injection of CO2. Presented models are the first of such estimation made for those sandstones and consequently this is the first application of indicator transformation into analyses of CO2 target reservoirs in Croatia. Geological description of the study area within the Sava Depression During Late Pannonian and Early Pontian many depressional areas along the entire Pannonian Basin System were re-extended due to the thermal subsidence as result of 2nd transtensional phase[16-18]. Depositional spaces for the accommodation of a huge volume of sandy material were created, and the main sandstone hydrocarbon reservoirs were deposited[19]. Turbidites were dominant clastic transport mechanism in the Croatian part of the Pannonian Basin System (abbr. CPBS) during that time, originated from the Eastern Alps and transported through the Vienna Basin, the Styrian Basin and the Mura Depression, more to the east in several re-deposition phases, when it eventually entered into the Sava Depression^ 17, 20-22]. Due to long transport only the medium and finegrained sands and silts reached the depression, where in the calm periods, when turbiditic cur- rents were not active, typical calm deep-water calcite rich mad (later marls) were deposited. However, the CPBS was characterised with several prominent, even present-day, mountains that existed during entire Miocene like uplifted palaeorelief. Probably some of them were isolated islands in the Central Paratethys and, later, the Lake Pannon and, eventually, the Sava Lake, giving small part of sandy and silty detritus into dominant turbiditic and basin deposits[6 23]. Study area is located in the northwestern part of the Sava Depression (Figure 1). Present-day the Ivanić Field is asymmetrical brachianti-cline, with longer axis of northwest-southeast strike, slightly pronounced top in the southern part, and series of NW-SE faults extended along the western part (Figure 2). Lithologically, reservoir rocks are mostly medium and fine-grained sandstones, i.e. lith-oarenites, in vertical alternation with marls (Figure 3). Sandstone and marl layers are not always laterally continuous, i.e. rather they pinch out or could be eroded. As results, some sandstones deposited in separate events could be in direct contact, i.e. amalgamated. The reservoir "series" (i.e. set of connected dep-ositional events) were classified as Upper Pan-nonian (9.3-7.1 Ma[16]). Reservoirs are elongated along NW-SE direction and represented fast gravitationally dropped sediment, into "strike-slip" area as depositional minimum. Such areas were characteristic for the entire CPBS during Upper Miocene[17], and typical sedimentation model for such sediments in the western Sava Depression had been described for the Kloštar Field[24, 25]. It assumed that the coarsest SLOVENIA PANNONIAN BAS,N . HUNGARY ■„-■ Kreevd **> - v-- ' '-V,.. ZMREB % Test area г CROAZIA,« * 5 Slamski Brod Slavo ___/"K J -. J* BOSNIA and HERZEGOVINA 4___0- Osijek r SERBIA Figure 2: Structural top map of the reservoir„A" based onto archive maps[12]. Figure 1: Location map of the Ivanić Field[12]. Figure 3: Schematic geological (lithological, lithostratigraphical and chronostratigraphical) section of the Ivanić Field. material had been deposited in the central part of palaeostructure, i.e. main channel(s). Laterally, toward structure borders, depositional lithofacies gradually were replaced with basin-al marls. Basics about applied mapping methods Deterministical interpolation methods are still commonly used for reservoir characterization and modelling in Croatia12, 26-30]. However, in several cases stochastical methods are also applied114 as standard part of geomathematics as it is understood in Croatia[31]. In general, deterministic models for the same input data give always the same results if the analytical method is the same. There are many deterministic interpolation methods, but Krig-ing is commonly used in geosciences, dealing very well with clustered data and numerical outliers. The goal of the method is to determine spatial relationships between the measured data and the point where value is estimating and it is why a mapping is preceded by vario-gram (sometimes covariance or madogram) analysis. So, the point's mutual distances, not values, are crucial for interpolation. Moreover, the Kriging is marked with Kriging variance that could be calculated as regional or local. In each case the goal is minimised such variance, i.e. differences between expected and estimated values[32]. Linearity of Kriging estimation could be ex-presed by Equation 1. It means that value of regionalised variable at selected location (Zk) is estimated from all surrounding and spatially dependence values (Z), using appropriate weighting coefficient (A.). Z = £A.xZ. k i i i = l Where are: Zk - kriged value, Z. - value at location "i". i ' A - weighted coefficient. (1) Such estimation also implies that values Z . will have normal (Gaussian) distribution. It is con-seqeunce of the central limit theorem where is implied that any variable with large number of indenpendent events obtains Gaussian distribution, whatever is probability density function of events[33]. The weighting coefficients in the Kriging are calculated using matrix eqau-tions[34, 35]. One of the most frequently used Kriging method for facies identification is the Indicator Kriging, i.e. usage of porosity cutoffs. It does not require stationarity assumption of 1st or 2nd degree, or multivariate normality, and is very robust in respect to outliers. In fact, statistics is represented only with variograms, and the originally continuous distribution is discretized using cutoffs, knowing know that at a particular location the value lies in a particular interval. Indicator approach is used for mapping of two categorical (discrete) values or indicator variable, shown by 0 and 1. Intention is showing two different lithotypes or lithofacies, very often by using the cut-off values for porosity as a geological variable suitable for distinguishing clastic lithofacies[36]. The input values are transformed into indicator ones by using the cut-offs that determine presence or absence of lithofacies. Such method is described with following Equation 2: KAk = if z(x) й V if z(x) > V (2) Where are: I(x) - indicator variable, z(x) - measured value, v - cut-off. Calculation of the Indicator Kriging, as each Kriging technique, requires variogram analysis. The special in this technique is that vario-grams need to be calculating for each cut-off and standardised with sill onto 1. Moreover, recommended number of cut-offs is between 5 and 11, what requires sufficient number of measurements[36]. Indicator maps correspond to continuous assessment of iso-probability lines with values in the interval [0, 1]. Mathematically such maps show the probability of the event {z (x) < vk}. Input datasets Input datasets consisted of 16 wells with porosity and permeability values of the reservoirs "A" and "B" (Table 1 and 2). Porosities were calculated from the log analyses and cored intervals. Permeabilities were obtained by unsteady state method at a constant oil/water viscosity ratio and applying a constant differential pressure across the core samples. After correction for overburden pressure, depth correlation of log analyses and coring data was performed. Several cut-offs were defined for probability mapping: — In reservoir "A": < 15 %, 15-19 %, 19-22 %, 22-24 % and > 24 % for porosity, and < 40 x 10-3 nm2, 40-50 x 10-3 nm2, 50-60 x 10-3 nm2, 60-70 x 10-3 nm2, 70-80 x 10-3 nm2 and > 80 x 10-3 nm2 for permeability. — In reservoir "B": < 10 %, 10-15 %, 15-20 %, 20-23 % and > 23 %, and < 1 x 10-3 nm2, 1-5 x 10-3 nm2, 5-20 x 10-3 nm2, 35-75 x 10-3 nm2 and > 75 x 10-3 nm2. Histograms are shown in Figure 4. Indicator Krigingporosity subsurface maps and correlation between porosity and permeability Petrophysical parameters were mapped by the Indicator Kriging and obtained are probability maps for certain interval. Porosity intervals 19-22 % for reservoir "A" (Map 1) and 20-23 % for reservoir "B" (Map 2) are chosen as the best representing for extension of medium-grained reservoir sandstones. For reservoir "A" is also derived E-type map that shows the most probable values of porosity (Map 3). Correlation between porosity and permeability was calculated using Pearson's correlation coefficient, and coefficient of determination. The value for the reservoir "A" reservoir is R = 0.76 and R2 = 0.57 retrospectively (Figure 5] and for reservoir "B" 0.89 and 0.79 (Figure 6]. The both values are qualitatively "high" and can described stronger linear dependence between two analysed variables, especially for reservoir "B". Furthermore, both reservoirs are result of deeper lake sedimentary processes. Later, area was subdued to compaction. However, there were not existed processes that could Table 1: Input dataset for the reservoir "A" Table 2: Input dataset for the reservoir "B" reservoir Well Porosity (%) Horizontal permeability (10-3 цт2) Well Porosity (%) Horizontal permeability (10-3 цт2) Well 1 21.03 LIQUIDATED Well 1 1.90 LIQUIDATED Well 2 18.68 56.22 Well 2 0.81 1.39 Well 3 22.65 76.40 Well 3 5.10 4.07 Well 4 18.23 39.90 Well 4 19.78 32.85 Well 5 20.20 LIQUIDATED Well 5 0.00 LIQUIDATED Well 6 14.32 41.40 Well 6 0.00 0.91 Well 7 21.32 52.40 Well 7 0.00 1.39 Well 8 18.93 48.09 Well 8 0.00 1.20 Well 9 22.03 70.68 Well 9 8.62 10.39 Well 10 17.27 60.14 Well 10 23.76 77.38 Well 11 24.21 81.01 Well 11 1.99 2.54 Well 12 23.13 62.50 Well 12 23.13 34.36 Well 13 20.95 49.74 Well 13 12.07 15.83 Well 14 22.08 67.20 Well 14 19.09 29.40 Well 15 19.80 60.93 Well 15 0.00 1.63 Well 16 21.12 58.11 Well 16 21.27 36.37 Histogram of number of data in cut-off defined classes of "Gamma 3" porosity X axis: porosity (%), Y axis: number of data) Histogram of number of data in cut-off defined classes of "Gamma 4" porosity (X axis: porosity (%), Y axis: number of data) Histogram of number of data in cut-off defined classes of "Gamma 3" porosity (X axis: permeability (10-3 цт2), Y axis: number of data) Histogram of number of data in cut-off defined classes of "Gamma 4" porosity (X axis: permeability (10-3 цт2), Y axis: number of data) Figure 4: Histograms of defined classes for the reservoirs "A" and "B"porosity and permeability. Map 1: Probability map for the porosity interval 19-22 % of reservoir "A!'. Map 2: Probability map for the porosity interval 20-23 % of reservoir "B". Map 3: Indicator Kriging E-type map of reservoir "A' porosity, based on the equation of Ordinary Kriging. Porosity (%l Figure 5: Porosity and permeability regression line for reservoir "A. Porosity (%) Figure 6: Porosity and permeability regression line for reservoir "A". result in secondary porosity (tectonic caused fracturing) or filling pores with dissolved materials. So, the primary porosity played main role in fluid distribution, what is favourable for linear relation between porosity and permeability. The larger correlation between porosity and permeability in reservoir "B" is probably reflection of generally narrow range of both values in analysed lithological sequence, what is consequence of more uniform depositional environment. Results and conclusions Analysed sandstone reservoirs were created in brackish, lacustric depositional environments of the Lake Pannon. The mineral content is represented with quartz, dolomite rock fragments, micas, K-feldspars and cement. The texture is medium to fine-grained with porosity mostly between 15 % and 25 % and average permeability between 50 and 60 x 10-3 цш2. Those values can be considered as typical for the Upper Miocene sandstones in the Sava Depression. Depositional environment included periodically active regional turbidites during Late Panno-nian and Early Pontian stages. However, here are marked intervals between 19 % and 23 % as the most representative for describing depositional area of typical channel, medium-grained sandstone lithofacies. Those are the best lithology for fluid flow and injection into analysed reservoirs. For this purpose those intervals are mapped into two reservoirs planned for CO2 injection. The results (Maps 1 and 2) shows areas where such turbiditic lithofacies could be certainly followed. However, the maps are biased with elongated "bull-eyes" effect, what is not favourable feature for interpretation. It could be override using approximately omnidirectional variogram model, with larger ranges, but it was not observed in data itself. So, presented Indicator Kriging probability maps are considered as the most representative and useful for planning pilot injection. The last Map 3 is direct porosity map of the reservoir "A", but interpolated using Ordinary and Indicator Kriging algorithms simultaneously. It was possible using algorithm that indicator probability maps considered as the "weights" for the more precise application of the Ordinary Kriging. Interestingly, not each "bull-eye" shape was characterised with maximal probabilities or direct porosity estimated values. It is directed reflection of palaeo-depths of each such "micro" strike-slip on the bottom of the Ivanić Structure during Upper Pannonian, where some of them were deeper and "collected" the coarser detritus (mostly on the east and southeast] then other ones. Acknowledgments This work represents part of a multidisci-plinary geological research project entitled "Development of geomathematical methods for analysis of Neogene depositional environments in the Croatian part of Pannonian Basin System" (head T. Malvić), financially supported by the University of Zagreb in programme "Supporting of researching 2" in the 2013, the licensed version of WinGslib had been used for mapping. References [1] Awan, A. R., Teigland, R., Kleppe, J. (2008): A Survey of North Sea Enhanced-Oil-Recovery Projects Initiated During the Years 1975 to 2005. SPE Res. Eval. & Eng., 11 (3], pp. 497-512. [2] Balić, D., Velić, J. & Malvić, T. (2008): Selection of the most appropriate interpolation method for sandstone reservoirs in the Kloštar oil and gas field. Geologia Croatica, 61 (1], pp. 27-35. [3] Bennaceur, K., Monea, M., Sakurai, S., Gupta, N., Ramakrishnan, T. S., Whittaker, S., Randen, T. (2004): CO2 capture and storage a solution within. Oilfield Review, 16 (3], pp. 44-61. [4] Ma, J., Morozov, I. (2010): AVO modeling of pressure-saturation effects in Weyburn CO2 sequestration. The Leading Edge, 29 (2], pp. 178-183, doi: 10.1190/1.3304821. [5] White, D. (2009]: Monitoring CO2 storage during EOR at the Weyburn-Midale Field. The Leading Edge, 2 (7], pp. 838-842, doi: 10.1190/1.3167786. [6] Sarapa, M. (1981]: Utjecaj ugljičnog dioksida na svojstva i iscrpak nafte (Influence of carbon-dioxide on oil properties and recovery] Mag. Thesis. University of Zagreb: Faculty of Mining, Geology and Petroleum Engineering 1981; 171 p. [7] Sečen, J. (2006): Enhance oil recovery methods (Metode povećanja iscrpka nafte) (1st ed.). Zagreb: INA-Industry of Oil Plc 2007, pp. 273-364. [8] Stalkup, F.I. (1980): Carbon Dioxide Miscible Flooding: Past, Present, And Outlook for the Future. Journal of Petroleum Technology, 30 (8], 1102-1112, doi: 10.2118/7042-PA. [9] Lu, J., Wilkinson, M. & Haszeldine, R. S. (2011): Carbonate cements in Miller field of the UK North Sea: a natural analogue for mineral trapping in CO2 geological storage. Environmental Earth Sciences, 62 (3], pp. 507-517. [10] Bauer, S., Class, H., Ebert, M., Gotze, H., Holzheid, A., Kloditz, O., Rosenbaum, S., Rabbel, W. & Schafer, D. (2012]: Modelling, Parameterization, and Evaluation of Monitoring Methods for CO2 Storage in Deep Saline Formations the CO2 MoPa Project. Environmental Earth Sciences, doi: 10.1007/ s1266-012-1707-y. [11] Goričnik, B., Domitrović, D. (2003): Laboratorijska istraživanja primjenjivosti CO2 procesa na naftnim poljima u Republici Hrvatskoj (Laboratory researching of applicability of CO2 processes in oil fields in Republic of Croatia]. Naftaplin, 1, pp. 5-12. [12] Novak, K., Malvić, T., Velić, J., Simon, K. (2013): Increased hydrocarbon recovery and CO2 storage in Neogene sandstones, a Croatian example: part II. Environ Earth Sciences, 71 (8], pp. 1866-6280, doi:10.1007/s12665-013-2756-6. [13] Perić, M., Kovač, S. (2003): Simulacijska studija procesa povećanja iscrpka nafte (EOR-proce-sa] istiskivanjem ugljik-dioksidom primjenom višekomponentnog modela COMP III (Simulation Study of Enhanced Oil Recovery Process by CO2 injection Applying a Multi-component COMP III]. Naftaplin, 1, pp. 13-25. [14] Novak Zelenika, K., Velić, J., Malvić, T. (2013): Local sediment sources and palaeoflow directions in Upper Miocene turbidites of the Pannonian Basin System (Croatian part), based on mapping of reservoir properties. Geological quarterly, 57 (1], pp. 17-30. [15] Novosel, D. (2005]: Rezultati projekta pokusnog istiskivanja nafte utiskivanjem ugljičnog dioksida na Naftnom polju Ivanić (The project results of test oil replacement by carbon dioxide on the Ivanić oil field). Naftaplin, 1, pp. 51-60. [16] Haq, B. U., Eysinga, F. W. B. (1998): Geological Time Table (5th ed.). Amsterdam: Elsevier Science, (Wall Chart). [17] Malvić, T. (2012): Review of Miocene shallow marine and lacustrine depositional environments in Northern Croatia. Geological Quarterly, 56 (3), pp. 493-504. [18] Malvić, T., Velić, J. (2011). Neogene Tectonics in Croatian Part of the Pannonian Basin and Reflectance in Hydrocarbon Accumulations. In: New Frontiers in Tectonic Research: At the Midst of Plate Convergence, Schattner, U. (eds.), InTech, Rijeka 2011, pp. 215-238. [19] Velić, J. (2007): Geologija ležišta nafte i plina (Geology of oil and gas reservoirs). Faculty of Mining, Geology and Petroleum Engineering, University of Zagreb 2007; 342 p. [20] Rögl, F. (1998): Palaeographic Consideration for Mediterranean and Paratethys Seaways (Oligocene to Miocene). Ann. Naturhist. Mus. Wien 99A 1998; pp. 279-310. [21] Vrbanac, B. (1996): Palaeostructural and sedimen-tological analyses of Upper Pannonian sediments Ivanić Grad Formation in the Sava Depression. Doctoral Thesis. University of Zagreb: Faculty of Mining, Geology and Petroleum Engineering 1996; 121 p. [22] Vrbanac, B., Velić, J., Malvić, T. (2010): Sedimentation of deep-water turbidites in the SW part of the Pannonian Basin. Geologica Carpathic, 61 (1), 55—69, doi: 10.2478/v10096-010-0001-8. [23] IEA GHG Weyburn CO2 Monitoring & Storage project Summary Report 2000-2004. Petroleum Technology research Centre, Regina 2004. Available on: . [24] Novak Zelenika, K., Cvetković, M., Malvić, T., Velić, J. & Sremac, J. (2013): Sequential Indicator Simulations maps of porosity, depth and thickness of Miocene clastic sediments in the Kloštar Field, Northern Croatia. Journal of Maps, 9 (4), pp. 550-557. [25] Novak Zelenika, K. (2012): Deterministički i stohastičkigeološki modeligornjomiocenskih pješčenjačkih ležišta u naftno-plinskom polju Kloštar (Deterministic and stohastic geological models of Upper Miocene sandstones reservoirs at the Kloštar oil and gas field). Doctoral Thesis, University of Zagreb: Faculty of Mining, Geology and Petroleum Engineering 2012, 190 p. [26] Cvetković, M., Velić, J., Malvić, T. (2009): Application of neural networks in petroleum reservoir lithology and saturation prediction. Geologia Croatica, 62 (2), pp. 115-121. [27] Malvić, T. (2005): Rezultati geostatističkog kar-tiranja poroznosti polja zapadnog dijela Dravske depresije (Molve, Kalinovac, Stari Gradac). Nafta, 56 (12), pp. 472-476. [28] Malvić, T. (2006): Middle Miocene Depositional Model in the Drava Depression Described by geostatistical porosity and thickness maps (case study: Stari Gradac-Barcs Nyugat Field). Rudar-sko-geološko-naftni zbornik, 18, pp. 63-70. [29] Malvić, T., Balić, D. (2009): Linearnost i Lagrange-ov linearni multiplikator u jednadžbama običnoga kriginga (Linearity and Lagrange Linear Multipli-cator in the Equations of Ordinary Kriging). Nafta, 60 (1), pp. 31-43. [30] Malvić, T., Đureković, M. (2003): Application of methods: Inverse distance weighting, ordinary kriging and collocated cokriging in porosity evaluation, and comparison of results on the Beničan-ci and Stari Gradac fields in Croatia. Nafta, 54 (9), pp. 331-340. [31] Lapaine, M., Malvić, T. (2009): Geomathematics between Mathematics and Geosciences. Annual of the Croatian Academy of Engineering, 12, pp. 51-67. [32] Isaaks, E., Srivastava, R. (1989): An Introduction to Applied Geostatistics. New York: Oxford University Press Inc. 1989; 580 p. [33] Geiger, J. (2006): The behaviour of Sequential Gaussian Simulation in the limits. Xth Congress of Hungarian Geomathematics, (May 18-20. 2006), Morahalom, Hungary. [34] Deutsch, C. V., Journel, A. G. (1997): GSLIB: Geosta-tistical Software Library and User's Guide (2nd ed.). New York: Oxford University Press Inc. 1997; 369 p. [35] Hohn, M. E. (1988): Geostatistics and Petroleum Geology. New York: Van Nostrand Reinhold 1998; 400 p. [36] Novak Zelenika, K., Malvić, T. & Geiger, J. (2010): Mapping of the Late Miocene sandstone facies using Indicator Kriging. Nafta, 61 (5), pp. 225-233. Professional paper Received: October 25, 2014 Accepted: January 20, 2015 Design of product properties by suitable planning of a cold forging process Oblikovanje lastnosti izdelka z ustreznim načrtovanjem procesa hladnega kovanja Vid Krušič*, Martin Fabjan MAHLE Letrika, d. o. o., Polje 15, 5290 Šempeter pri Gorici, Slovenia Corresponding author. E-mail: vid.krusic@si.mahle.com Abstract One of the main advantages of the cold forming technology is improvement of mechanical properties such as tensile strength and hardness. At cold metal forging the hardness increases by plastic deformation, also known as work hardening. The degree of plastic deformation is represented by equivalent plastic strain. It is generally accepted that the mechanical property such as hardness are directly related to equivalent plastic strain. If the operating sequence to manufacture a certain product by cold forging is suitably planned, the required value of equivalent strain or appropriate hardness in the product's cross-section is achieved. The paper deals with planning of mechanical properties in manufacture of antivibration elements for the needs of the automotive industry and hubs of magnetic ignition systems for the motorcycle industry. At the production of antivibration elements the resistance to permanent plastic deformation in the case of axial dynamic loading of a product have to be assured, while in the case of hubs of magnetic ignition systems the appropriate hardness in the product's cross-section has to be achieved. A reliable design of the desired mechanical properties of products is assured by suitable planning of the forming processes in the virtual environment with the support of numerical simulations. Key words: cold forging, product property design, hardness, FE simulation Izvleček Ena izmed glavnih prednosti hladnega preoblikovanja je izboljšanje mehanskih lastnosti izdelka, kot sta natezna trdnost in trdota. Pri hladnem preoblikovanju trdota narašča s plastično deformacijo, kar je poznano kot hladno utrjevanje. Stopnja plastične deformacije je izražena z ekvivalentno plastično deformacijo. Splošno je znano, da je mehanska lastnost, kot je trdota, direktno povezana z ekvivalentno plastično deformacijo. Če je preoblikovalno zaporedje za izdelavo izdelka ustrezno načrtovano, lahko dosežemo zahtevano ekvivalentno deformacijo oziroma ustrezno trdoto po prerezu izdelka. Članek podaja načrtovanje mehanskih lastnosti pri izdelavi antivibracijskih izdelkov za potrebe avtomobilske industrije ter pest magnetnih vžigalnikov za potrebe industrije motornih koles. Pri antivibracijskih elementih moramo zagotoviti odpornost proti trajni plastični deformaciji v primeru aksialne dinamične obremenitve izdelka, medtem ko pri pestih magnetnih vžigalnikov z ustrezno trdoto po prerezu izdelka zagotavljamo obratovanje pri visokih vrtljajih. Zanesljivo oblikovanje želenih mehanskih lastnosti izdelkov lahko zagotovimo z načrtovanjem preoblikovalnih procesov v virtualnem okolju ob podpori numeričnih simulacij. Ključne besede: hladno preoblikovanje, oblikovanje lastnosti izdelka, trdota, simulacija z MKE Introduction From the perspective of production processes, cold forming (CF) is one of the most important technologies which enable production of geometrically demanding and accurate products. The CF technology aims at plastic deformation of a blank from simple initial forms by the forming process to a product of a complex form and dimensions. The main reasons for using the CF technology mostly in the automotive industry are as follows[1-3]. — Great accuracy of dimensions and good surface qualities in mass manufacture of products, — Efficient material use; generally smaller re-finishing operations, including machining, are required, — Improved mechanical properties as a result of hardening in cold, the fibres flow is uninterrupted and adjusted to the product loading, which enables the products to withstand greater dynamic loads and to be lighter accordingly, — Technology is economical and environment friendly. In cold plastic deformation, the material is hardened. This causes structural changes of material. Grains are elongated in the direction of the main deformation. Owing to the material hardening, there are changes of mechanical, physical and chemical properties of metal (Figure 1). By greater deformation, true stress, tensile strength, material hardness, and electrical resistance increase. At the same time reduction of elongation, contraction, impact toughness, resistance to corrosion, heat conductivity and elasticity module occur. 80 40060 ^ 30040 i N 200- « -Q г- Ч! о о± 0 20 40 60 80 plastic strain [%] Figure 1: Changes of mechanical properties as a consequence of cold forming process. Up to now many researches have been made towards predictions of mechanical properties of a product. If we are familiar with the ratio between the hardness and effective strain of the formed part material and the strain flow of the formed part, we can predict the course of hardness of a cold-formed part. Kim et al. performed a pressure test by the help of an experiment and FE simulation for the steel AISI 1010[4]. Hardness values regarding pressure test and effective deformations from the FE simulations have been analysed in the same point of a formed-piece. The curve and/or ratio between the hardness and the effective strain were acquired by a regression analysis. In the next study, a similar analytical model for hardness prediction of a formed piece was presented, which is based upon the known effective strain that was achieved by the FE analysis of a formed-piece[5]. The model was confirmed by an experiment, so that hardness prediction of a formed-piece is quite accurate. Tekkaya[6] carried out an analysis of final imprinting in a cold-formed part in order to find equivalent Vickers hardness and to gain the ratio between the Vickers hardness and the flow stress. Mechanical properties of a product like hardness, yield stress, elongation, residual ductility and tensile test depend directly on the equivalent deformation caused in cold forming. The article by K. Osakada and J. Yanagimoto[7] emphasizes that distribution of an equivalent deformation of the formed-piece is gained through the FE simulation and in this way we can - by a suitable ratio - define the course of mechanical properties of a product. D. Biermann et al. recommended a new way of product manufacture, which already at the planning stage considers the real material properties defined by a production process (Figure 2), while the classical way of a product planning considers only ideal material properties[8]. In this way, knowledge on local material properties are gained in tensile tests, damages, other stress, textures and micro structures used already in the process of planning, what enables product manufacture true to size. Safety factor and product weight are in this way substantially reduced, which results in reduced consumption of energy in primary production of materials as well as in the later use of a product. Hardening in cold forming can be used to one's advantage in product manufacture, where specific product properties like a defined hardness in the product cross-section or specified tensile stress are required. By suitable planning of the forming operation sequence required for a product manufacture, we can achieve the specified mechanical properties of a product. Schematic course of the process steps for a suitable planning of mechanical properties of a product with the forming process is shown in Figure 3. In this kind of planning of the specified mechanical properties, one has to pay attention to obtain a healthy product in the cross section cut, which means there are no damages of the product. The purpose of this paper is to show planning of the specified mechanical properties in the products by the cold forming technology for the needs of the automotive and motorcycle industries. First example shows planning of manufacture of the automotive antivibration bush, where in case of axial dynamic loading of a product resistance to permanent plastic deformation is required. The second example shows manufacture of a hub of magnetic ignition system for the needs of motorcycles with the requirement for a suitable tensile stress in the product cross-section. We can assure a reliable planning of the desired mechanical properties of products by creating the forming processes in the virtual environment with the support of numerical simulations. Requirements for the manufacture of an antivibration busch During driving, the car is subject to various vibrations both for external impacts, caused by the roadway, and internal impacts resulting from various drive elements. The external impacts caused by the roadway can be insulated by tyres with mufflers and dumpers, while the internal vibrations are insulated by antivibration elements. Figure 4 shows over 20 different antivibration elements, which are built into a car in order to insulate vibrations. These are elements that are used as vibration insulation as for example in vibrations from an engine, exhaust system, suspension etc. Usually these are steel, tube-like parts, whose external surface if potted by rubber. In MAHLE Letrika we make various tube-like elements for the needs of antivibration elements (Figure 5(a))[9]. These metal tube-like elements are made by the cold forming technology. The product requirements include a specific form and resistance to permanent plastic deformation. Such form and resistance can be achieved only by the cold forming technology. Figure 5 (b) shows the allowed plastic deformation in case of axial dynamic loading of a bush. Knowledge of product properties Figure 2: New approach in designing and manufacturing products18. PROCESS MODIFICATION MATERIAL : FUNCTION PRODUCT PROCESS PRODUCT REQUIRED LOADING DESIGN PLANNING PROPERTIES . OK COLD FORGING PROCESS MACHINING, HEAT TREAT. TESTING Figure 3: A flow chart for design product properties by suitable planning of the cold forging process. Figure 4: Various antivibration elements built in a car. M Figure 5: (a) metal tube-like parts of antivibration elements, (b) allowed plastic deformation in loading of an antivibration bush. Planning of antivibration busch by CF technology To make a steel Ck 25 (DIN] bush, double passing through the press is required. In the first passing, operations follow in the sequence from (a) to (d) as shown in Figure 6. First, we plan the initial blank, which has previously been annealed and surface-treated by a phosphate and molybdenum disulphide, to a suitable external diameter. Then, the subsequent operations follow: centring, backward extrusion and bottom punching. After the first passing, the formed-piece is annealed properly and again surface-treated. In the last operation (e), reduction/narrowing of the upper part of the pre-form to the appropriate final form is made. We planned the operation sequence and metal flow during the pre-forming operation by the help of the programme DEFORM 2D[10]. The final bush form and material hardening during the forming process have to ensure the requirements regarding the plastic deforma- tion in case of an axial loading (Figure 5(b)). Figure 7 (a) shows the flow of effective deformations in the product cross-section simulated by the FE method. The upper curve in the diagram (Figure 8) shows the impact of loading on the permanent plastic deformation analysed in the virtual environment. A cold formed bush achieves permanent plastic deformation of a few hundredths of a millimetre in case of loading by 200 kN, which is within the limits of the allowed deformation 0.12 mm. The results have been confirmed also by physical tests in production. We were interested also in what would the resistance be if a bush was made by the machining procedure and/or in the non-hardened condition (Figure 7 (b)). In this case, we exceed the permanent plastic deformation already at loading of 120 kN as shown in the lower curves in the diagram (Figure 8). Even at full non-hardened bush (Figure 7 (c)) the situation is essentially the same as at non-hardened bush. A critical section is the same in both bushes. Figure 6: Operational plan of the forming operation sequence for manufacture of a bush. hub Figure 7: (a) CF bush - hardening in the cross-section, (b, c) non-hardened bush. 300 200 100 plastic deformation -^on-hardened -cold forged - —-"fuN non-hardened _ -----" --- __rĆ— / ^ t / / / * / / / ! / f/ 0 0,1 0,2 0,3 0,4 Displacement [mm] Figure 8: Plastic deformation of a bush in dependence on loading. <£f> Figure 9: The flywheel, hub of magnetic ignition system and hardness in dependence on the true strain for steel 16MnCr5. Requirements for a hub of magnetic ignition system The next example of improved mechanical properties is shown in manufacture of a hub of magnetic ignition system. A hub is a component element of the magnetic ignition system, which is used in motorcycles (Figure 9). The hub's job is to transfer the torque from the shaft to the flywheel at a high rotational speed. A basic requirement of a hub made of steel 16MnCr5 is that it has a suitable hardness in the cross-section, which in this way ensures product hardness. An appropriate hardness in the product cross-section is achieved when a hub is made by the cold forming technology. The forming processes have to be planned so that they ensure a suitable deformation in the desired product points. Bigger deformation at the same time denotes greater hardness as shown in the diagram in Figure 9. Attention has to be paid so that deformation does not exceed the critical limit, which could cause material damage. Planning of an appropriate hardness in the hub cross-section As already described in the introduction, the cold forming technology enables improvement of mechanical properties due to material hardening during a forming process. Planning of an appropriate hardness in the product cross-section by the cold forming technology depends on the following factors: — It is necessary to have data on the forming properties of a material at disposal. Figure 9 shows the curve of the flow stress of, the curve of the hardness HB and the specific forming work w from the size of the true strain 4>e for steel 16MnCr5[11], — Available hardware such as presses, number of forming operations that can be carried out in one passing through the press, — Software to make simulations of material flow and hardening during the forming process. Experimental appraisal of hardness is connected to great costs in respect of tool making, test implementation on a press, measurement of hardness in the formed-piece cross-section etc. — Knowledge and experience in the field of CF metals. Definition of optimum sequence of forming operations is required, which on one hand enables achieving of the suitable hardness in the product cross-section and on the other hand as long tool life as possible, automatic operating of the formed pieces during passing through the press etc. The forming process is planned at the end, i.e. from the final form to the initial blank. First, we design a suitable pre-form for the final operation. In Figure 10 (a), (b) two pre-forms are made and flow of an appropriate effective strain of a product gained by the FE simulation. Based on the comparison we can make an analysis about the pre-form (a) causing bigger effective deformations on the final product, while the pre-form (b) ensures a more homogeneous deformation, which pretty well meets the requirements regarding the distribution of the required hardness in the product cross-section. However, to optimize a suitable pre-form more FE simulations have been made, Figure 10 show only two of them. Figure 11 shows the sequence of the forming operations gained by the FE simulation method. The machined-piece with a suitable product weight is first surface treated. The forming operations follow from operations (a)-(c) for the first passing through a vertical mechanical stress 630 t. Heat treatment follows, then the pre-form (c) process annealing and surface treatment, and the second passing through the press for operations (d)-(g). The simulated hardness flow in the product cross-section is shown in Figure 12. The highest hardness is achieved in the lower and upper side of the flange and in the inside of the hub hole. The hardness flow meets the hardness of the product requirements. Approval of suitable planning of the forming operation sequences and hardness in the hub cross-section is acquired only in the product mass production. Inspection of cold-formed hubs showed that hardness meets the planned/required hardness in the product cross-section well. An accurate analysis of the product micro-structure is required in order to determine possible material damages due to local hardening and/or forced flow of material. Ca] 00 Figure 10: (a), (b) various pre-forms and flow of the effective strain in the product final form. Figure 11: Sequence of the forming operations for the manufacture of the hub of magnetic ignition system. Hardness HV A = 0-139 B= 140-219 С = 220 - 269 D = 270 - 290 Figure 12: The hardness flow in the hub cross-section simulated by the help of FE simulation. Conclusions The paper deals with planning of suitable mechanical properties in a product by the CF technology. Two examples of product manufacture for the needs of the automotive industry and industry of motorcycles are shown, in which the functional requirement is a product resistance against the permanent plastic deformation in case of an axial loading and distribution of suitable hardness in the product cross-section. By the support of a FE simulation and suitable planning of the forming operations sequence we can ensure the specified functional requirements in the products and thus make a good use of the advantages offered by the CF technology. Acknowledgements The article was published in Conference Proceedings: New Developments in Forging Technology, Editor Prof. Dr. M. Liewald MBA, Institute for Metal Forming Technology, University of Stuttgart, Fellbach, Germany, 2013. References [1] Gologranc, F. (1991): Forming, part 1. University of Ljubljana, Faculty of Mechanical Engineering, Ljubljana, (in slovene), p. 250. [2] Altan, T., Ngaile, G., Shen, G. (2005): Cold Forging Fundamentals and Applications. ASM International, Materials Park, OH, p. 341. [3] Lange, K. (1985J: Handbook of Metal Forming, McGraw-Hill Book Company, New York, p. 1176. [4] Kim, H., Lee, S. M., Altan, T. (1996): Prediction of hardness distribution in cold backward extruded cups. Journal of Materials Processing Technology, 59, pp. 113-121. [5] Sonmez, F. O., Demir, A. (2007): Analytical relations between hardness and strain for cold formed parts. Journal of Materials Processing Technology, 186, pp.163-173. [6] Tekkaya, A. E. (2001): Improved relationship between Vickers hardness and yield stress for cold forming materials. Steel Research, 72 (8), pp. 204-310. [7] Osakada, K., Yanagimoto, J. (2010): Prediction of Property of Metal Forming Product by the Finite Element Method. Proceedings of the 1st International conference on product property prediction, Ed; D. Biermann, A. E.Tekkaya, W. Tillman, Technische Univertaet Dortmund, pp. 155-167. [8] Biermann, D., Tekkaya A. E., Tillmann, W. (2010): Tailor Made Properties - Visions for the future of Manufacturing. Proceedings of the 1st International conference on product property prediction, Ed; D. Biermann, A. E. Tekkaya, W. Tillman, Technische Univertaet Dortmund, pp. 77-91. [9] Krušič, V., Mašera, S. (2009): Improvement of mechanical characteristics of products by cold forming technology. 9th conference and exhibition Innovative Automotive Technology IAT09, Ed; M. Fajdiga, J. Klemenc, F. Trenc, Nova Gorica, (in slovene), pp. 385-392. [10] SFTC (2007): Deform 2/3D, Ver. 8.1, User Manual, Columbus, Ohio, 2007. [11] Kuzman, K. (1989): Exercises in forming techniques. University of Ljubljana, Faculty of Mechanical Engineering, Ljubljana, (in slovene), p. 71. Professional paper Received: July 12, 2014 Accepted: February 11, 2015 Petrophysical evaluation of reservoirs in 'y' prospect Niger delta Petrofizikalna ocena rezervoarjev na raziskovalnem območju 'y' v delti Nigra Yemisi. C. Ajisafe Ekiti state University, Faculty of Science, Department of Geology, Ado Ekiti, Nigeria Corresponding author. E-mail: tuaseyemi@yahoo.co.uk Abstract A suite of well logs of two wells (1 and 2) from Y Prospect Niger Delta were evaluated using GeoGraphix software, with the aim of computing the petrophysical characteristics of the reservoirs as well as identify reservoir lithology within and between wells for information on stratigraphic and lithological parameters of the wells. Three reservoirs were correlated at depth range of 1 524 m to 1800 m, with thicknesses of 10-45 m. Cross plot of neutron porosity and density porosity were used to discriminate the fluid types. Computation of petrophysical properties and reservoir evaluation were carried out to determine recoverable hydrocarbon in place in the reservoirs. Well log data shows that area was characterized by sandy shale interbeds. Porosity values for the reservoir ranged from 30-40 %, water saturation 30-45 % and hydrocarbon saturation 65-80 %. Gas zone of economic importance was detected in reservoir L300 in well 2. The reservoir properties of the wells showed that they could be fair to very good for hydrocarbon accumulation. Key words: petrophysical properties, hydrocarbon reservoir, GeoGraphix, Nigeria Izvleček Karotažne podatke iz dveh vrtin (1 in 2) v raziskovalnem območju 'Y' v delti Nigra so ovrednotili z Ge-oGraphixovimi programi z namenom izračunati pe-trofizikalne značilnosti rezervoarjev ogljikovodikov, opredeliti litološke lastnosti v vrtinah in med njima ter dobiti ustrezne podatke o stratigrafskih in litoloških parametrih. V globini med 1 524 m in 1 800 m so povezali prereze treh rezervoarjev debeline od 10 m do 45 m. Tipe fluidov v plasteh so določili iz podatkov o nevtronsko ugotovljeni poroznosti in gostoti. Količine pridobljivih ogljikovodikov v rezervoarjih so ocenili iz izračunanih petrofizikalnih lastnosti in značilnosti rezervoarjev. Karotažni podatki nakazujejo prisotnost pe-ščeno-muljastih vmesnih plasti. Vrednosti poroznosti v rezervoarjih se gibljejo med 30 % in 40 %, nasičenosti z vodo med 30 % in 45 % in nasičenosti z ogljikovodiki med 65 % in 80 %. Navzočnost ekonomsko pomembnih zalog plina so ugotovili v rezervoarju L300 v vrtini 2. Lastnosti rezervoarjev v vrtinah pričajo o dobri do zelo dobri sposobnosti za nakopičenje ogljikovodikov. Ključne besede: petrofizikalne lastnosti, rezervoar ogljikovodikov, GeoGraphix, Nigerija Introduction A well log can be defined as an indirect record showing the rock and fluid properties along borehole. Such physical properties include electrical, radioactive, and some special kinds of measurements like electrical resistivity, spontaneous potential, gamma ray intensity, density, acoustic velocity etc.[1]. Most quantitative log analyses are aimed at defining petrophysical parameters, but only few of these parameters(-Formation lithology thicknesses and depths of the reservoirs and even non-reservoirs) can be measured directly. Others have to be derived or inferred from the measurement of other physical parameters of the rocks. Three basic logs (lithology, resistivity and porosity logs) are needed for proper formation evaluation. One is required to indicate permeable zones; another is needed to measure the resistivity of the formation, while the third is important for estimating porosity values. Well logs furnish the data necessary for the quantitative evaluation of hydrocarbon in-situ. From the view point of decision making, well logging is the most important aspect of drilling and completion process[2]. The information obtained from these logs can be used to interpret geology in general and in reservoir, identify productive zones, and estimate hydrocarbon reserves. This study therefore assesses the reservoir quality of two wells: well 1 and well 2 (Figure 1) using GeoGraphix Software. The main focus is to determine some reservoir properties with a view to ascertaining if the results generated make possible to predict economic saturation and production. Geology of the study area The Niger Delta (Figure 2) is a regressive sequence of clastic sediments developed in series of offlap cycles[3]. The base of the sequence consists of massive and monotonous marine shales. These grade into interbedded shallow-marine and fluvial sands, silts, and clays, which form the typical paralic facies portion of the delta[3]. The uppermost part of the sequence is a massive non-marine sand section. The established Cainozoic sequence in the Niger delta consists, in ascending order of the marine shales (Akata Formation), paralic clastics (Agbada Formation), and continental sands (Benin Formation)[4]. Akata Formation is composed of shales, clays and silts at the base of the delta sequence. They contain a few streaks of sand, possibly of turbiditic origin, and were deposited in holomarine (delta-front to deeper marine) environments. Agbada Formation forms the hydrocarbon perspective sequence in the Niger delta. It is represented by an alternation 0 40 oo w 0 38 00 W 0 36 00 w 30,000- 0 34 00 W -3 46 00 N _ 3 44 00 N j— 3 42 00 N t N _ -240,000 -230,000 Figure 1: Base map of "Y" prospect, showing the positions of the two wells (well 1 and 2) and 3D seismic survey. Figure 2: A geological map showing the Niger delta1131. of sands, silts, and clays of various proportions and thicknesses, representing cyclic sequences of offlap units. The shallowest part of the sequence is composed almost entirely of non-marine sand. It was deposited in alluvial or upper coastal plain environments following a southward shift of deltaic depobelts (structural and stratigraphic belts)[5]. This mechanism, called the escalator regression model, postulated that the base of the Benin Formation in any of the six depobelts is coeval with the Agbada Formation in the adjacent depobelt to the south. This principle implies an abrupt shift in the age of the base of the Benin Formation across the bounding faults of depobelts and had been used to define the Northern limit of the Northern Delta depobelt[6]. Weber[7] discussed in detail the sedimentology, growth faults dynamics and hydrocarbon accumulation in the Niger Delta. Short[8] and Avbovbo[9] also, studied the hydrocarbon potentials of the Niger Delta using well data. Oomkens[10] discussed lithofacies relations in the late Quaternary period. The stratigraphy, sedimentation and structure of Niger Delta was reviewed by Schlumberger[11]. The importance of longshore drift and submarine canyons and fans in the development of the basin has been emphasized by Burke[12]. Method of study Two wells namely well 1 and well 2 exist in "Y" Prospect. Well 1 is a vertical well with a total depth of 2 332 m, and gamma-ray (GR) log, deep laterolog (LLD), compensated sonic log (BCSL), and compensated formation density log (FDC) were used in this well. Well 2 is also a vertical well with a total depth of 2 160 m. The logs used in this well include, caliper log (CALI), gamma-ray (GR) log deep, laterolog (LLD), compensated sonic log (BCSL), and compensated formation density log (FDC). Petrophysical Evaluation: The analysis of the data was done using GeoGraphix software. The data consist of logs (from two wells) namely the caliper log, the gamma ray log (GR), deep laterolog, and porosity logs (sonic, density and neutron logs). Identification and Delineation of Lithologies: The GR log was used to identify the permeable and impermeable beds. GR values greater or equal to 75 APIo were identified as shale beds while zones with GR readings below 75 APIo were identified as sandstones. Intervals where the caliper logs read values lower than 24 cm were considered as permeable zones. This is because reduction in borehole diameter is indicative of the build-up of mudcake in permeable zones. Identification of Fluids: Fluids in the permeable beds were identified, using the deep laterolog resistivity logs and a combination of the neutron and density logs. High resistivity values of deep-reading resistivity log in permeable beds are indicative of either the presence of hydrocarbon or fresh water. Determination of Volume of Shale: The presence of shale in a reservoir can adversely affect the correct evaluation of petrophysical parameters particularly resistivity, porosity and water saturation. Hilchie[13] notes that the most important effect of shale in a formation is to reduce the resistivity contrast between oil or gas and water. With sufficient shale in a reservoir, it becomes very difficult to detect a productive zone[14]. Porosity and water saturation values must be corrected for shale effect to allow for a reliable formation evaluation. The first step in making this correction is to determine the volume of shale present in the reservoir. For this study, shale volume was determined using the GRlog. The Gamma Ray Index (IGR) was calculated first from the log using the formula[15]; 58 GR, - GR . j _ log mm GR~ GR -GR . (1) Where are; GRIog = gamma ray reading of formation. GRmin = minimum gamma ray reading (clean sand) GRmax = maximum gamma ray reading (shale) Subsequently, the calculated IGR was used in the formula[10] for Cainozoic unconsolidated rocks to determine the volume of shale (Vsh). V.= 0.083 (237xIGR — 1) (2) The calculated volumes of shale are expressed in percentage. Determination of Porosity: Porosity values were obtained from sonic log, density log and a combination of neutron and density logs. Sonic porosity values were calculated using the formula proposed by Dewan[2] for undercompacted sandstones: ф, (3) The calculated sonic porosity was subsequently corrected for both shale and hydrocarbon effects. The density porosity (фГ) was computed from eqn. 4; Фо = P ma - Pfl (4) Фое = Фо-КъХФо,ъ Where are: фВе = effective density porosity V, = volume of shale = 8.26 % sh фГ1 = density porosity of adjacent shale = 0.10 Dsh The neutron log values were in API Neutron Unit and had to be converted to apparent limestone porosity. The values obtained were converted to apparent sandstone unit by the addition of 3.5 p. u (0.035). Shale effect was also corrected for to obtain the effective neutron porosity (PhiNe or фГе). Porosity values were also computed from a combination of neutron and density logs as follows. tfv = Фые + Фв Ф1 + Ф1 т Ne т p,. (for oil zones) (6) (for gas zones) (7) Where are: фте= effective neutron-density derived porosity фNe = effective neutron porosity фГе = effective density porosity Determination of Formation Water Saturation and Hydrocarbon Saturation The water saturation of the uninvaded zone (Sw) was computed from w * -- (8) S = [11] The hydrocarbon saturation (Shc) was calculated from the equation; Where are: pma = matrix (sandstone) density = 2.638 g/cm3 pb = formation bulk density pfl = fluid density The 3.5 p.u (0.035) is subtracted from the calculated density porosity to convert from apparent limestone porosity unit to apparent sandstone porosity unit. Shale effect was subsequently corrected for to give the effective density porosity (PhiDe or фГ(г). Correcting for shale effect; S =1-5 Sh W (9) The water saturation of the flushed zone (S ) xo was estimated from the Archie's formula; 5„ = FR mf R (10) The other saturation values calculated are the Moveable Hydrocarbon Saturation (MHS) and the Residual Hydrocarbon Saturation (RHS). MHS = 5—5 xO w (5) RHS =1 — 5 (11) (12) Tables 1 and 2 show the calculated parameters at sampled intervals for the calculated reservoirs. Table 1: Statistical data derived from GeoGraphix software for Well 1 DEPTH PHIN RHOB PHID DT PHIA GR V„ PHIE RT Ro SwA BVW 4 060 0.414 9 2.158 0.311 128.86 0.363 85.7 0.821 0.064 8 1.56 9.54 1 0.064 8 4 070 0.350 4 2.109 0.34 146.48 0.345 57.2 0.465 0.184 5 4.39 1.18 0.518 0.095 5 4 080 0.349 3 2.119 0.334 146.63 0.342 62.5 0.532 0.16 4.04 1.56 0.621 0.099 5 4 090 0.294 4 2.118 0.335 141.96 0.315 32.2 0.152 0.266 7 13 0.56 0.208 0.055 5 4 100 0.308 3 2.061 0.368 127 0.338 31.2 0.141 0.290 8 0.68 0.47 0.835 0.242 8 4 110 0.329 8 2.113 0.337 126.89 0.333 38.1 0.227 0.257 9 0.75 0.6 0.896 0.231 4 120 0.366 2 2.108 0.341 132.88 0.353 53.9 0.423 0.203 9 0.85 0.96 1 0.203 9 4 130 0.337 7 2.025 0.39 135.53 0.364 38.7 0.234 0.278 5 0.64 0.52 0.895 0.249 3 4 140 0.334 7 2.129 0.328 133.92 0.331 38.7 0.234 0.253 7 0.77 0.62 0.9 0.228 5 4 150 0.349 9 2.091 0.351 127.07 0.35 34 0.175 0.288 9 0.74 0.48 0.807 0.233 1 4 160 0.401 2 2.249 0.257 123.31 0.329 95.1 0.939 0.020 1 1.4 99.45 1 0.020 1 4 170 0.486 8 2.259 0.251 133.41 0.369 93.7 0.922 0.028 9 1.25 47.8 1 0.028 9 4 180 0.273 4 2.121 0.333 146.24 0.303 35.8 0.198 0.243 3 6.83 0.68 0.315 0.076 5 4 190 0.277 2.093 0.35 172.96 0.313 34.5 0.181 0.256 4 50.55 0.61 0.11 0.028 1 4 200 0.313 5 2.063 0.367 152.52 0.34 38.4 0.231 0.261 9 36.9 0.58 0.126 0.032 9 4 210 0.39 1.984 0.414 147.52 0.402 56.3 0.454 0.219 6 3.53 0.83 0.485 0.106 5 4 220 0.373 7 2.196 0.288 121.78 0.331 89.6 0.87 0.042 9 1.04 21.77 1 0.042 9 4 230 0.345 2.179 0.298 122.71 0.322 75.7 0.696 0.097 7 1.24 4.19 1 0.097 7 4 240 0.377 1 2.03 0.387 137.36 0.382 53.3 0.416 0.223 1 0.83 0.8 0.983 0.219 3 4 250 0.394 2 2.186 0.294 129.94 0.344 65.1 0.564 0.150 1 0.92 1.77 1 0.150 1 4 260 0.387 9 2.122 0.332 131.27 0.36 65.7 0.571 0.154 3 0.78 1.68 1 0.154 3 4 270 0.341 8 2.066 0.365 122.53 0.354 34.2 0.177 0.291 0.68 0.47 0.833 0.242 4 4 280 0.315 2 2.116 0.336 118.99 0.326 28.1 0.102 0.292 5 0.73 0.47 0.803 0.234 8 4 290 0.346 2 2.05 0.375 125.45 0.36 49.5 0.369 0.227 6 0.69 0.77 1 0.227 6 4 300 0.427 7 2.141 0.321 128.93 0.374 61.7 0.522 0.179 0.84 1.25 1 0.179 4 310 0.322 2 2.085 0.354 122.46 0.338 38.6 0.233 0.259 4 0.82 0.59 0.853 0.221 2 4 320 0.379 5 2.064 0.367 130.16 0.373 59.4 0.492 0.189 3 0.7 1.12 1 0.189 3 4 330 0.298 2 2.184 0.295 119.78 0.297 65 0.563 0.129 7 0.95 2.38 1 0.129 7 4 340 0.279 3 2.106 0.341 119.1 0.31 31.4 0.143 0.266 1 0.77 0.56 0.856 0.227 9 4 350 0.350 5 2.137 0.323 118.23 0.337 32.8 0.16 0.283 1 0.82 0.5 0.78 0.220 8 4 360 0.320 6 2.121 0.333 121.58 0.327 56.2 0.452 0.178 9 0.89 1.25 1 0.178 9 4 370 0.328 4 2.078 0.359 123.19 0.344 39.8 0.248 0.258 4 0.68 0.6 0.939 0.242 7 4 380 0.344 1 2.037 0.383 127.57 0.364 47.7 0.346 0.237 8 0.64 0.71 1 0.237 8 4 390 0.381 5 2.313 0.219 116.05 0.3 95 0.937 0.018 9 1.25 111.6 1 0.018 9 4 400 0.358 2.25 0.256 118.82 0.307 88.1 0.851 0.045 6 1.2 19.26 1 0.045 6 4 410 0.374 4 2.134 0.325 124.35 0.35 88.5 0.856 0.050 2 0.86 15.86 1 0.050 2 4 420 0.295 9 2.108 0.341 121.8 0.318 30.6 0.133 0.276 1 0.78 0.52 0.823 0.227 1 4 430 0.302 7 2.046 0.377 127.43 0.34 30.3 0.129 0.296 3 0.63 0.46 0.85 0.251 9 4 440 0.429 4 2.307 0.222 119.7 0.326 82.2 0.778 0.072 3 1.01 7.66 1 0.072 3 4 450 0.383 2.321 0.214 112.16 0.299 95.5 0.943 0.016 9 1.53 139.39 1 0.016 9 4 460 0.450 9 2.242 0.26 123.42 0.356 94.7 0.934 0.023 5 1.24 72.26 1 0.023 5 4 470 0.347 2 2.039 0.381 131.02 0.364 43 0.287 0.259 7 21.09 0.59 0.168 0.043 5 4 480 0.298 8 2.043 0.379 132.67 0.339 38.2 0.228 0.261 7 27.5 0.58 0.146 0.038 1 4 490 0.343 4 2.054 0.373 131.13 0.358 29.4 0.118 0.315 9 15.21 0.4 0.162 0.051 3 4 500 0.333 1 2.12 0.333 124.25 0.333 44 0.3 0.233 3 0.7 0.73 1 0.233 3 PHIN - Neutron Porosity RHOB - Bulk Density DT - Sonic log PHIND - Density Porosity PHIA - Average Porosity GR - Gamma Ray VsM - Volume of Shale PHIE - Effective Porosity RT - True Resistivity Ro - Wet Resistivity SwA - Average Water Saturation Table 2: Statistical data derived from GeoGraphix software for Well 02 DEPTH PHIN RHOB PHID DT PHIA GR V„ PHIE RT Ro SwA BVW 4 600 0.46 2.21 0.28 115.63 0.37 93.03 0.913 0.032 2 1.22 38.54 1 0.032 2 4 610 0.157 2.07 0.364 121.66 0.261 41.38 0.267 0.191 106.67 1.1 0.101 0.019 4 4 620 0.131 2.1 0.348 126.12 0.24 45.25 0.316 0.164 30.77 1.49 0.22 0.036 1 4 630 0.259 2.19 0.289 124.21 0.274 75.92 0.699 0.082 5 5.13 5.87 1 0.082 5 4 640 0.48 2.14 0.32 147.01 0.4 90.64 0.883 0.046 8 2.39 18.26 1 0.046 8 4 650 0.129 2.03 0.39 143.56 0.26 61.89 0.524 0.123 6 26.89 2.62 0.312 0.038 6 4 660 0.135 2.18 0.297 142.82 0.216 45.7 0.321 0.146 7 21.77 1.86 0.292 0.042 9 4 670 0.232 2.06 0.369 136.74 0.301 78.5 0.731 0.0808 3.14 6.13 1 0.080 8 4 680 0.195 2.07 0.362 137.64 0.279 52.89 0.411 0.164 3 10.7 1.48 0.372 0.061 1 4 690 0.058 1.95 0.434 137.59 0.246 52.09 0.401 0.147 2 13.22 1.85 0.374 0.055 4 700 0.208 2.09 0.354 133.05 0.281 39.91 0.249 0.210 9 57.14 0.9 0.125 0.026 5 4 710 0.199 2.13 0.327 131.45 0.263 42.75 0.284 0.188 1 18.82 1.13 0.245 0.046 1 4 720 0.166 2.04 0.384 135.5 0.275 62.98 0.537 0.127 2 5.83 2.47 0.651 0.082 8 4 730 0.043 1.86 0.488 139.37 0.265 45.33 0.317 0.181 2 533.33 1.22 0.048 0.008 7 4 740 0.03 1.85 0.496 151.07 0.263 47.36 0.342 0.173 1 57.14 1.34 0.153 0.026 5 4 750 0.214 2.19 0.292 142.68 0.253 38.34 0.229 0.195 2 200 1.05 0.072 0.014 1 4 760 0.018 1.9 0.465 131.11 0.241 52.89 0.411 0.142 2 24.81 1.98 0.282 0.040 2 4 770 0.197 2.02 0.393 136.45 0.295 51.5 0.394 0.179 10.67 1.25 0.342 0.061 2 4 780 0.24 2.05 0.377 152.03 0.308 42.09 0.276 0.223 3 31.37 0.8 0.16 0.035 7 4 790 0.066 1.92 0.453 147.89 0.259 53.05 0.413 0.152 2 25 1.73 0.263 0.04 4 800 0.026 1.82 0.511 140.37 0.269 45.69 0.321 0.182 4 39.51 1.2 0.174 0.031 8 4 810 0.408 2.2 0.283 131.81 0.346 96.41 0.955 0.015 5 1.08 166 1 0.015 5 4 820 0.048 1.97 0.42 125.08 0.234 56.06 0.451 0.128 7 25.4 2.42 0.308 0.039 7 4 830 0.061 1.95 0.433 137.63 0.247 50.88 0.386 0.151 5 72.73 1.74 0.155 0.023 5 4 840 0.046 1.98 0.415 128.52 0.231 46.25 0.328 0.155 1 31.37 1.66 0.23 0.035 7 4 850 0.338 2.13 0.33 130.21 0.334 76.42 0.705 0.098 3 7.05 4.14 0.766 0.075 3 4 860 0.182 2.05 0.373 133.48 0.278 75.23 0.69 0.086 4.96 5.41 1 0.086 4 870 0.442 2.08 0.355 143.16 0.399 101.48 1 0 1.95 4 880 0.117 1.97 0.424 131.52 0.27 50.59 0.382 0.166 9 19.39 1.44 0.272 0.045 4 4 890 0.021 1.94 0.442 129.29 0.232 44.78 0.31 0.16 34.41 1.56 0.213 0.034 1 4 900 0.093 1.98 0.418 135.37 0.256 42.72 0.284 0.183 1 42.1 1.19 0.168 0.030 8 4 910 0.04 1.96 0.428 136.79 0.234 50.56 0.382 0.144 5 106.67 1.92 0.134 0.019 4 4 920 0.282 2.18 0.299 134.6 0.29 83.52 0.794 0.059 8 3.11 11.18 1 0.059 8 4 930 0.013 1.82 0.512 133.72 0.263 40.2 0.253 0.196 5 228.57 1.04 0.067 0.013 2 4 940 0.125 2.01 0.396 123.34 0.261 61.94 0.524 0.124 1 13.01 2.6 0.447 0.055 5 4 950 0.04 2.03 0.386 122.22 0.213 37.39 0.217 0.166 7 118.51 1.44 0.11 0.018 4 4 960 0.034 1.9 0.467 129.7 0.251 50.23 0.378 0.155 8 61.54 1.65 0.164 0.025 5 4 970 0.246 2 0.403 137.19 0.325 81.81 0.773 0.073 8 6.28 7.35 1 0.073 8 4 980 0.08 1.96 0.426 135.98 0.253 59.89 0.499 0.126 9 20.25 2.48 0.35 0.044 4 4 990 0.348 2.21 0.282 123.25 0.315 90.98 0.887 0.035 5 2.09 31.78 1 0.035 5 5 000 0.082 2 0.404 129.35 0.243 48.86 0.361 0.155 4 10.03 1.66 0.406 0.063 1 PHIN - Neutron Porosity RHOB - Bulk Density DT- Sonic log PHIND - Density Porosity PHIA - Average Porosity GR - Gamma Ray Vshl - Volume of Shale PHIE - Effective Porosity RT - True Resistivity Ro - Wet Resistivity SwA - Average Water Saturation Reserve Estimation The volumes of hydrocarbons in place were estimated from the following formulae; OIP = 77580(1 - SJAh (13) GIP = 43560ф(1 — *Šw)Ah (14] Where are: OIP = oil in place (barrels) GIP = gas in place (cubic feet) The constants 7 758 and 43 560 are conversion factor for oil and gas barrels or cubic meter respectively. ф = porosity (decimal) Sw = formation's water saturation (decimal). Area = area of the reservoir (in acres) h = net thickness of reservoir (wet with oil or gas) (in feet) Discussion of results Gamma-ray (GR) logs were used to identify the lithology in both wells penetrated. The lithology was identified by defining shale base line (Figure 3), which is a constant line in front of the shale and in front of the sand. Thick sand at a depth of 304.8 m to 926.7 m (1 000-3 040 ft] was delineated in well 1. Well 2 contain thick sand layer at a depth of 100 m to 914.4 m (328-3 000 ft]. At a depth of 1 237.5 m to 1 371.6 m (4 060-4 500 ft], and 1 402.1 m to 1 524 m (4 600-5 800 ft] thick sand (identified reservoir sand) was also observed in well 1 and well 2 respectively. Figure 3 shows the strati-graphic cross section within the study area. The major lithologies encountered in the study area were basically shale and sand, some of which occurs as interbeds. The reservoir sandstone was evaluated quantitatively for effective porosity, water & hydrocarbon saturation and net pay (Tables 3 and 4]. Structural Cross section: Equally spaced logs Datum = 0 ft Subsea Vertical scale = 1 in per 100 ft TD = 2332 m Sandstone IH Shale Figure 3: Correlation across the wells of "Y" Prospect showing mapped sands. Table 3: Petrophysical Parameters for Well 01 SANDS ZONE TOP MD/m BASE MD/m PHIE/% Sw/% Shc/% GROSS (m) Net sand thickness (m) NTG PAY/m L300 1 1 240.54 1 524 19.95 48.5 51.5 283.46 43.65 0.154 22.9 M400 2 1 539.24 1 630.68 21.56 76.91 23.1 91.44 14.08 0.154 0.87 N500 3 1 685.54 1 699.25 12.25 98.52 1.48 13.72 3.88 0.283 - MD - Measured Depth PHIE - Effective Porosity Sw - Water Saturation Shc - Hydrocarbon Saturation NTG - Net-to-Gross Table 4: Petrophysical Parameters for Well 02 SANDS ZONE TOP MD/m BASE MD/m PHIE/% Sw/% Shc/% GROSS (m) Net sand thickness (m) NTG PAY/m L300 1 1 402.08 1524 13.46 30.82 69.18 121.92 101.07 0.829 80.75 M400 2 1 542.88 1 615.44 17.73 55.78 44 73.15 8.266 0.113 1.37 N500 3 1 676.4 1 706.88 20.26 92.38 7.62 30.48 6.858 0.225 0.14 MD - Measured Depth PHIE - Effective Porosity Sw - Water Saturation Shc - Hydrocarbon Saturation NTG - Net-to-Gross Neutron density logs were used to define hydrocarbon type (gas] present in "Y" Prospect. Petrophysical analysis of the reservoir bed was based on examination of the well logs. The combination of neutron and density logs was used for reservoir L300 in both wells to detect gas zone. At these intervals, density porosity was observed to be greater than neutron porosity and the curves cross over each other, therefore were identified as gas bearing zones (Figure 4). This is because gas in pores causes the density porosity to read very high values (gas has a lower density than oil or water] and causes the neutron porosity to be too low (there is a low concentration of hydrocarbon atoms in gas than in oil and water). Figure 5 shows the crossplot of neutron porosity with RhoB (Formation Bulk Density). Gas has a very marked effect on both density and neutron logs. If it is assumed that the formation fluid is water and the invasion zone is shallow, then gas will result in a lower bulk density (note on the cross plot, this results in a point higher on the y-axis], and a lower apparent neutron porosity (Figure 6]. 0.6 0.5 0.4 -:<'\ ■ у R2 = 0.64 R = 0.8 ♦ PHIN ■ PHID 40 60 80 100 GR Log reading of the Zome (API) Figure 5: Crossplot of Neutron Porosity (PHIN) with Density Porosity (PHID). Figure 4: Pickett crossplot of Neutron porosity(PHIN) with Bulk Density (RhoB). Figure 6: Typical log curve showing gas zone and its effect on density and neutron. Conclusion Reservoir evaluation is an attempt to find appropriate reservoir rocks and then to estimate the porosity, permeability and water saturation. In Niger Delta sands more than 15 m thick in most places represent composite bodies, and may consist of two to three stacked channels. They are poorly consolidated and have porosities as high as 40 % in oil- bearing reservoirs. Porosity reduction is gradual. All sands shallower than 3 000 m have porosities of more than 15 %, but below 4 000 m only a few sands have more than 15 % porosity. Gross, net and net-to-gross values for sandstones in well -1 are 13.72-283.46, 3.88-43.65 and 0.154-0.283, while those for well - 2 are 30.48-121.92, 6.858-101.07 and 0.113-0.829 respectively. Reservoir which contain hydrocarbon is referred to as pay zone and the porosities range 20-40 %. The average porosity (PHIA) which is the average porosity within the net is 0.23 (23 %) for well 01, it is 0.28 (28 %) for well 2. The porosity values are within the porosities of producing reservoirs in the Niger Delta. Water saturation is generally low in hydrocarbon bearing zone ranging from 1-30 % thereby implying high hydrocarbon saturation. The water saturation, values in "Y" Prospect at well 1 and well 2 are 0.85 (85 %), and 0.62 (62 %) respectively. The reservoir properties evaluated for the wells showed that they could be fair to very good for hydrocarbon accumulation. Acknowledgements The author wishes to express her sincere appreciation to Mobil Producing Nigeria Unlimited for the provision of the data used for this work. Also I am thankful to the Landmark for the licensed GeoGraphix software which was used for the well evaluation. References [1] Sheriff, R. E. (1992): Basic petrophysics and geophysics. Reservoir Geophysics, No. 7, Society of exploration Geophysicists, Tulsa, pp. 37-49. [2] Dewan, J. T. (1983): Essentials of Modern Open-Hole Log Interpretation; Tulsa, Oklahoma, U.S.A, p. 244. [3] Evamy, D. W., Haremboure, J., Kamerling, P., Knaap, W. A., Molloy, F. A., Rowlands, P. H. (1978): Hydrocarbon habitat of Tertiary Niger Delta. AAPG Bulletin, 62 (1), pp. 1-39. [4] Knox, G. J., Omatsola, E. M. (1989): Development of the Cenozoic Niger Delta in terms of the "escalator regression" model and Impact on hydrocarbon distribution; in W. J. M., van der Linden et al., eds, 1987 Proceedings KNGMG Symposium on Coastal Lowlands, geology, and Geotechnology: Dordrecht, the Netherlands, Klumer Academic Publishers, pp. 181-202. [5] Weber, K. J. (1971): Sedimentological aspects of oil fields in the Niger Delta. Geologic en Mijnbouw, 50, pp. 559-576. [6] Ekweozor, C. M., Daukoru, E. M. (1994): Northern Delta depobeltportion of the Akata-Agbada, petroleum system, Niger Delta, Nigeria; in L. B. Magoon and W. G. Dow, eds, The petroleum system - from source to trap: AAPG Memoir, pp. 599-613. [7] Weber, K. J., Daukoru, E. M. (1975): Petroleum Geology of the Niger Delta; Proceedings of the Ninth World Petroleum Congress, Tokyo, Japan, 2, pp. 209-221. [8] Short, K. C., Stauble, A. J. (1967): Outline of the Geology of Niger Delta. AAPG Bulletin, 51, pp. 761-779. [9] Avbovbo, A. A. (1978). Tertiary Lithostragraphy of the Niger Delta. AAPG, 62 (2), pp. 295-300. [10] Oomkens, E. (1974): Lithofacies relations in the late Quaternary Niger Delta complex. Sedimentology, 2, pp. 195-222. [11] Schlumberger (1991): Log interpretation principles/ applications; Houston, Schlumberger educational services, pp. 26-41. [12] Burke, K. (1972): Longshore drift, submarine canyons, and submarine fans in development of Niger Delta. AAPG Bulletin, 59 (3), pp. 414-419. [13] Hilchie, D. W. (1978): Applied Openhole Log Interpretation; Golden, Colorado, D. W. Hilchie. Inc., pp. 10-65. [14] Asquith, G., Krygowski, D. (2004): Basic Well Log Analysis; 2nd edition, AAPG, Tulsa, Oklahoma, U.S.A, p. 244. [15] Schlumberger, (1989): Log interpretation principles/ applications; Houston, Schlumberger educational services, pp.13-19. Instructions to Authors Navodila avtorjem RMZ - MATERIALS & GEOENVIRONMENT (RMZ - Materiali in geookolje) is a periodical publication with four issues per year (established in 1952 and renamed to RMZ - M&G in 1998). The main topics are Mining and Geotechnology, Metallurgy and Materials, Geology and Geoenvironment. RMZ - M&G publishes original scientific articles, review papers, preliminary notes and professional papers in English. Only professional papers will exceptionally be published in Slovene. In addition, evaluations of other publications (books, monographs, etc.), in memoriam, presentation of a scientific or a professional event, short communications, professional remarks and reviews published in RMZ - M&G can be written in English or Slovene. These contributions should be short and clear. Authors are responsible for the originality of the presented data, ideas and conclusions, as well as for the correct citation of the data adopted from other sources. The publication in RMZ - M&G obligates the authors not to publish the article anywhere else in the same form. RMZ - MATERIALS AND GEOENVIRONMENT (RMZ - Materiali in geookolje), kratica RMZ - M&G je revija (ustanovljena kot zbornik 1952 in preimenovana v revijo RMZ - M&G 1998), ki izhaja vsako leto v štirih zvezkih. V reviji objavljamo prispevke s področja rudarstva, geotehnologije, materialov, metalurgije, geologije in geookolja. RMZ - M&G objavlja izvirne znanstvene, pregledne in strokovne članke ter predhodne objave samo v angleškem jeziku. Strokovni članki so lahko izjemoma napisani v slovenskem jeziku. Kot dodatek so zaželene recenzije drugih publikacij (knjig, monografij ...), nekrologi In memoriam, predstavitve znanstvenih in strokovnih dogodkov, kratke objave in strokovne replike na članke, objavljene v RMZ - M&G v slovenskem ali angleškem jeziku. Prispevki naj bodo kratki in jasni. Avtorji so odgovorni za izvirnost podatkov, idej in sklepov v predloženem prispevku oziroma za pravilno citiranje privzetih podatkov. Z objavo v RMZ - M&G se tudi obvežejo, da ne bodo nikjer drugje objavili enakega prispevka. Specification of the Contributions Optimal number of pages is 7 to 15; longer articles should be discussed with the Editor-in-Chief prior to submission. All contributions should be written using the ISO 80000. — Original scientific papers represent unpublished results of original research. — Review papers summarize previously published scientific, research and/or expertise articles on a new scientific level and can contain other cited sources which are not mainly the result of the author(s). — Preliminary notes represent preliminary research findings, which should be published rapidly (up to 7 pages). — Professional papers are the result of technological research achievements, application research results and information on achievements in practice and industry. Vrste prispevkov Optimalno število strani je 7-15, za daljše članke je potrebno soglasje glavnega urednika. Vsi prispevki naj bodo napisani v skladu z ISO 80000. — Izvirni znanstveni članki opisujejo še neobjavljene rezultate lastnih raziskav. — Pregledni članki povzemajo že objavljene znanstvene, raziskovalne ali strokovne dosežke na novem znanstvenem nivoju in lahko vsebujejo tudi druge (citirane) vire, ki niso večinsko rezultat dela avtorjev. — Predhodna objava povzema izsledke raziskave, ki je v teku in zahteva hitro objavo obsega do sedem (7) strani. — Strokovni članki vsebujejo rezultate tehnoloških dosežkov, razvojnih projektov in druge informacije iz prakse in industrije. — Publication notes contain the author's opinion on newly published books, monographs, textbooks, etc. (up to 2 pages). A figure of the cover page is expected, as well as a short citation of basic data. — In memoriam (up to 2 pages), a photo is expected. — Discussion of papers (Comments) where only professional disagreements of the articles published in previous issues of RMZ - M&G can be discussed. Normally the source author(s) reply to the remarks in the same issue. — Event notes in which descriptions of a scientific or a professional event are given (up to 2 pages). Review Process All manuscripts will be supervised shall undergo a review process. The reviewers evaluate the manuscripts and can ask the authors to change particular segments, and propose to the Editor-in-Chief the acceptability of the submitted articles. Authors are requested to identify three reviewers and may also exclude specific individuals from reviewing their manuscript. The Editorin-Chief has the right to choose other reviewers. The name of the reviewer remains anonymous. The technical corrections will also be done and the authors can be asked to correct the missing items. The final decision on the publication of the manuscript is made by the Editor-in-Chief. Form of the Manuscript The contribution should be submitted via e-mail as well as on a USB flash drive or CD. The original file of the Template is available on RMZ - Materials and Geoenvironment Home page address: www.rmz-mg.com. The contribution should be submitted in Microsoft Word. The electronic version should be simple, without complex formatting, hyphenation, and underlining. For highlighting, only bold and italic types should be used. Composition of the Manuscript Title The title of the article should be precise, informative and not longer that 100 characters. The author should also indicate the short version of the title. The title should be written in English as well as in Slovene. — Recenzije publikacij zajemajo ocene novih knjig, monografij, učbenikov, razstav ... (do dve (2) strani; zaželena slika naslovnice in kratka navedba osnovnih podatkov). — In memoriam obsega do dve (2) strani, zaželena je slika. — Strokovne pripombe na objavljene članke ne smejo presegati ene (1) strani in opozarjajo izključno na strokovne nedoslednosti objavljenih člankov v prejšnjih številkah RMZ - M&G. Praviloma že v isti številki avtorji prvotnega članka napišejo odgovor na pripombe. — Poljudni članki, ki povzemajo znanstvene in strokovne dogodke zavzemajo do dve (2) strani. Recenzentski postopek Vsi prispevki bodo predloženi v recenzijo. Recenzent oceni primernost prispevka za objavo in lahko predlaga kot pogoj za objavo dopolnilo k prispevku. Recenzenta izbere uredništvo med strokovnjaki, ki so dejavni na sorodnih področjih, kot jih obravnava prispevek. Avtorji morajo predlagati tri recenzente. Pravico imajo predlagati ime recenzenta, za katerega ne želijo, da bi recen-ziral njihov prispevek. Uredništvo si pridržuje pravico, da izbere druge recenzente. Recenzent ostane anonimen. Prispevki bodo tudi tehnično ocenjeni in avtorji so dolžni popraviti pomanjkljivosti. Končno odločitev za objavo da glavni urednik. Oblika prispevka Prispevek lahko posredujte preko e-pošte ter na USB-mediju ali CD-ju. Predloga za pisanje članka se nahaja na spletni strani: www.rmz-mg.com. Besedilo naj bo podano v urejevalniku besedil Word. Digitalni zapis naj bo povsem enostaven, brez zapletenega oblikovanja, deljenja besed, podčrtavanja. Avtor naj označi le krepko in kurzivno poudarjanje. Zgradba prispevka Naslov Naslov članka naj bo natančen in informativen in naj ne presega 100 znakov. Avtor naj navede tudi skrajšan naslov članka. Naslov članka je podan v angleškem in slovenskem jeziku. Information on the Authors Information on the authors should include the first and last name of the authors, the address of the institution and the e-mail address of the leading author. Abstract The abstract presenting the purpose of the article and the main results and conclusions should contain no more than 180 words. It should be written in Slovene and English. Key words A list of up to 5 key words (3 to 5) that will be useful for indexing or searching. They should be written in Slovene and English. Introduction Materials and methods Results and discussion Conclusions Acknowledgements References The sources should be cited in the same order as they appear in the article. They should be numbered with numbers in square brackets. Sources should be cited according to the SIST ISO 690:1996 standards. Monograph: [1] Trček, B. (2001): Solute transport monitoring in the unsaturated zone of the karst aquifer by natural tracers. Ph. D. Thesis. Ljubljana: University of Ljubljana 2001; 125 p. Journal article: [2] Higashitani, K., Iseri, H., Okuhara, K., Hatade, S. (1995): Magnetic Effects on Zeta Potential and Diffu-sivity of Nonmagnetic Particles. Journal of Colloid and Interface Science, 172, pp. 383-388. Electronic source: CASREACT - Chemical reactions database [online]. Chemical Abstracts Service, 2000, renewed 2/15/2000 [cited 2/25/2000]. Available on: . Podatki o avtorjih Podatki o avtorjih naj vsebujejo imena in priimke avtorjev, naslov pripadajoče institucije ter elektronski naslov vodilnega avtorja. Izvleček Izvleček namena članka ter ključnih rezultatov z ugotovitvami naj obsega največ 180 besed. Izvleček je podan v angleškem in slovenskem jeziku. Ključne besede Seznam največ 5 ključnih besed (3-5) za pomoč pri indeksiranju ali iskanju. Ključne besede so podane v angleškem in slovenskem jeziku. Uvod Materiali in metode Rezultati in razprava Sklepi Zahvala Viri Uporabljane literaturne vire navajajte po vrstnem redu, kot se pojavljajo v prispevku. Označite jih s številkami v oglatem oklepaju. Literatura naj se navaja v skladu s standardom SIST ISO 690:1996. Monografija: [1] Trček, B. (2001): Solute transport monitoring in the unsaturated zone of the karst aquifer by natural tracers. doktorska disertacija. Ljubljana: Univerza v Ljubljani 2001; 125 str. Članek v reviji: [2] Higashitani, K., Iseri, H., Okuhara, K., Hatade, S. (1995): Magnetic Effects on Zeta Potential and Diffu-sivity of Nonmagnetic Particles. Journal of Colloid and Interface Science, 172, str. 383-388. Spletna stran: CASREACT - Chemical reactions database [online]. Chemical Abstracts Service, 2000, obnovljeno 15. 2. 2000 [citirano 25. 2. 2000]. Dostopno na svetovnem spletu: . Scientific articles, review papers, preliminary notes and professional papers are published in English. Only professional papers will exceptionally be published in Slovene. Annexes Annexes are images, spreadsheets, tables, and mathematical and chemical formulas. Annexes should be included in the text at the appropriate place, and they should also be submitted as a separate document, i.e. separated from the text in the article. Annexes should be originals, made in an electronic form (Microsoft Excel, Adobe Illustrator, Inkscape, AutoCad, etc.) and in .eps, .tif or .jpg format with a resolution of at least 300 dpi. The width of the annex should be at least 152 mm. They should be named the same as in the article (Figure 1, Table 1). The text in the annexes should be written in typeface Arial Regular (6 pt). The title of the image (also schemes, charts and graphs) should be indicated in the description of the image. When formatting spreadsheets and tables in text editors, tabs, and not spaces, should be used to separate columns. Each formula should have its number written in round brackets on its right side. References of the annexes in the text should be as follows: "Figure 1..." and not "as shown below:". This is due to the fact that for technical reasons the annex cannot always be placed at the exact specific place in the article. Manuscript Submission Contributions should be sent to the following e-mail address: rmz-mg@ntf.uni-lj.si. In case of submission on CD or USB flash drive, contributions can be sent by registered mail to the address of the editorial board: RMZ - Materials and Geoenvironment, Aškerčeva 12, 1000 Ljubljana, Slovenia. The contributions can also be handed in at the reception of the Faculty of Natural Sciences and Engineering (ground floor), Aškerčeva 12, 1000 Ljubljana, Slovenia with the heading "for RMZ - M&G". Znanstveni, pregledni in strokovni članki ter predhodne objave se objavijo v angleškem jeziku. Izjemoma se strokovni članek objavi v slovenskem jeziku. Priloge K prilogam prištevamo slikovno gradivo, preglednice in tabele ter matematične in kemijske formule. Priloge naj bodo vključene v besedilu, kjer se jim odredi okvirno mesto. Hkrati jih je potrebno priložiti tudi kot samostojno datoteko, ločeno od besedila v članku. Priloge morajo biti izvirne, narejene v računalniški obliki (Microsoft Excel, Adobe Illustrator, Inkscape, AutoCad...) in shranjene kot .eps, .tif ali .jpg v ločljivosti vsaj 300 dpi. Širina priloge naj bo najmanj 152 mm. Datoteke je potrebno poimenovati, tako kot so poimenovane v besedilu (Slika 1, Preglednica 1). Za besedilo v prilogi naj bo uporabljena pisava Arial navadna različica (6 pt). Naslov slikovnega gradiva, sem prištevamo tudi sheme, grafikone in diagrame, naj bo podan v opisu slike. Pri urejevanju preglednic/tabel, v urejevalniku besedila, se za ločevanje stolpcev uporabijo tabulatorji in ne presledki. Vsaka formula naj ima zaporedno številko zapisano v okroglem oklepaju na desni strani. V besedilu se je potrebno sklicevati na prilogo na način: „Slika 1 ...", in ne „... kot je spodaj prikazano:" saj zaradi tehničnih razlogov priloge ni vedno mogoče postaviti na točno določeno mesto v članku. Oddaja članka Prispevke lahko pošljete po elektronski pošti na naslov rmz-mg@ntf.uni-lj.si. V primeru oddaje prispevka na CD- ali USB-mediju le-te pošljite priporočeno na naslov uredništva: RMZ - Materials and Geoenvironment, Aškerčeva 12, 1000 Ljubljana, Slovenija ali jih oddajte na: recepciji Naravoslovnotehniške fakultete (pritličje), Aškerčeva 12, 1000 Ljubljana, Slovenija s pripisom „za RMZ - M&G". The electronic medium should clearly be marked with the name of the leading author, the beginning of the title and the date of the submission to the Editorial Office of RMZ - M&G. Information on RMZ - M&G — Editor-in-Chief Assoc. Prof. Dr. Peter Fajfar Telephone: +386 1 200 04 51 E-mail address: peter.fajfar@omm.ntf.uni-lj.si — Secretary Ines Langerholc, Bachelor in Business Administration Telephone: +386 1 470 46 08 E-mail address: ines.langerholc@omm.ntf.uni-lj.si These instructions are valid from July 2013. Elektronski mediji morajo biti jasno označeni z imenom vsaj prvega avtorja, začetkom naslova in datumom izročitve uredništvu RMZ - M&G. Informacije o RMZ - M&G — urednik izr. prof. dr. Peter Fajfar Telefon: +386 1 200 04 51 E-poštni naslov: peter.fajfar@omm.ntf.uni-lj.si — tajnica Ines Langerholc, dipl. poslov. adm. Telefon: +386 1 470 46 08 E-poštni naslov: ines.langerholc@omm.ntf.uni-lj.si Navodila veljajo od julija 2013. Inženirska geologija HidrogeoLogija Geomehanika Projektiranje ISSN 1408-7073