RESERVOIR CHARACTERIZATION AND INTEGRATED ANALYSIS OF ATTRIBUTES FOR INTERPRETATION OF 3D SEISMIC DATA OF F3 BLOCK NETHERLANDS

Abstract

Reservoir characterization is a broad process which means that the reservoir is studied in terms of the fluids in the reservoir and thus reservoir modelling and reservoir simulation are both under reservoir characterization. The search for hydrocarbon comes under the category of reservoir characterization as it includes both modelling and simulation. In this thesis I have studied the reservoir modelling part. Reservoir modelling is generating a model of a reservoir using the available geophysical and geological information about the area and thus studying the model is seismic interpretation. In the field of reservoir characterization seismic attributes are used by interpreters to better characterize a reservoir. The main objective in exploration of hydrocarbon is to correctly image the structure in time and depth and to characterize the amplitudes of the reflections. Using these amplitudes, the additional features which are derived are called attributes. Thus, a quantity which is derived from the traditional seismic data and which contain information about the better geological and geophysical interpretation of the original seismic section. The project characterizes the area F3 Block located in offshore of Netherlands and by using texture attributes along with other physical attributes and geometrical attributes on a given three-dimensional Post Stack Time Migrated seismic data (PSTM) the area is studied for potential hydrocarbon bearing zones. F3 Block is a block in offshore of Netherlands and is in the Dutch area of the North Sea. The block consists of acquired 3D seismic data which was used in the exploration and production of oil and gas in the area. The top 1200ms of the study area consist of reflections from different geological times such as Miocene and Pliocene. The area shows a geological feature such as sigmoidal bedding, downlap, onlap, and toplap with some truncative structures. The area also consists of bright spots which are studied in this dissertation thesis. The 3D Seismic data was conditioned to remove the noise and enhance the signal to noise ratio with the use of dip steered median filter. The removal of noise is important as it extracts the information about fault and fractures in the area and the faults are more visible after conditioning of the data. The attributes used to study the zone of interest are geometrical and physical attributes. The Geometrical attributes are the attributes which help the interpreter to study the geometrical v characteristics of the area for example the faults, the dip and the similarity between the traces. The physical attributes are attributes which help the interpreter to study the physical properties using the amplitude of the seismic reflections and extracting details for the interpretation of seismic data such as energy and coherency attributes. Texture based attributes were also used such as Energy, Entropy, Contrast, Homogeneity and Dissimilarity attributes were calculated on the marked horizons and the results were studied using joint analysis of all the attributes. The data consist of four Well logs in the entire 3D seismic region. However, the wells F02_1 and F03_2 only had density logs and the other two wells were not having those logs so a neural network was used to determine densities at other two wells F03_4 and F06_1. However, this is already done originally and I used the pretrained logs for interpretation. The logs also were used to calculate the porosity ⌀=2.65-rho/ (2.65-1.05) at all the wells, however my work does not require this to be done and I used the pre converted logs from the TerraNubis open data website for both the Seismic and the Well log data. The Well logging data gave the behaviour of different density and velocity logs in the area and acoustic impedance log was calculated from them that was helpful in inversion studies. Seismic inversion was done at the last to convert our volume to acoustic impedance volume. Model based deterministic inversion was used. The trend in the data were generated with the help of the low frequency information of the data and also a relative impedance result was added to the low frequency which created a model with both the low and high frequency. This is called deterministic inversion which was used in the thesis. The inversion is an important step which can define the rock property in the last to avoid any type of confusion regarding the lithology of the area. The integrated attribute and inversion study resulted in delineating a bright spot in one of the horizons. The bright spot is due to the sand deposits in the area with the presence of hydrocarbon. Also, the possibilities of presence of coal and the presence of carbonates in the area was removed using the study.