Sea Bed Logging Applications: ANOVA Analysis 2 for Synthetic Data from Electromagnetic (EM) Simulator Hanita Daud, Radzuan Razali Department of Fundamental and Applied Sciences UniversitiTeknologi PETRONAS, Tronoh, Perak, MALAYSIA hanita_daud@petronas.com.my Vijanth Asirvadam Department of Electrical and Electronics Engineering UniversitiTeknologi PETRONAS, Tronoh, Perak, MALAYSIA Abstract—The aim of this work is to determine the significant level of parameters such as sediment thickness, frequency and step size for data generated from electromagnetic (EM) simulator for sea bed logging (SBL) application that have been processed using cubic spline interpolation using Analysis of Variance (ANOVA) method. Synthetics data were generated from a simulator that is able to replicate SBL environment. SBL is a new technique that uses controlled source electromagnetic (CSEM) for finding resistive layers or hydrocarbon reservoir in deep water environment. It uses very low frequency to obtain greater wavelength for longer wave penetration. In this work frequencies of 0.5Hz, 0.25Hz, 0.125Hz and 0.0625Hz were used and sediment thicknesses were varied from 1000m to 3000m incremented every 250m. Collected data were interpolated using cubic spline with step sizes variation of 4,8,12 and 16 and mean square error (MSE) were calculated between original and interpolated data. The ANOVA method was applied to these MSE to study any significant different between frequency level and spline step size level at each sediment thickness. It was found from the ANOVA tables that frequency level and step sizes level were having significant difference to the MSE for every sediment thickness below 2250m. For sediment thicknesses of equal to 2250m and above, frequency level is not significance to the MSE but the step sizes level are still significant to the MSE. Keywords-sea bed logging, mean square error, cubic spline interpolation, analysis of variance I. INTRODUCTION Hydrocarbon exploration is the research by petroleum geologists for hydrocarbon beneath the Earth’s surface, such as oil and natural gas. Various techniques are used for this kind of exploration namely seismic method and Control Source Electromagnetic (CSEM) [1]. Seismic methods are widely accepted and considered as proven technology for hydrocarbon exploration and reservoir characterization [2]. However, although seismic method can identify structures that might be expected to contain hydrocarbons, it is practically blind to the fluids contained in the formation [3]. This explains why geophysicists prefer CSEM technique for seabed logging which is more accurate and detail in giving the result before drilling. Seabed logging is the measurement of electrical resistivity beneath the seafloor which played a crucial role in hydrocarbon exploration and reservoir assessment and development [4]. Hydrocarbon reservoirs are known to have higher resistivity than surrounding water filled sediments. Hydrocarbon reservoirs are known to have resistivity value of 30 – 500 m in contrast to sea water layer of 0.5 – 2 m and sediments of 1 -2 m. The changes in electric field around the reservoirs can be measured, and the data gained can be used to interpret the presence of hydrocarbon [5]. This SBL technique is known to be expensive and due to this it is used to complement seismic method. SBL uses horizontal electric dipole (HED) source to emit low frequency electromagnetic signal that will be received by an array of seafloor receivers. The electromagnetic energy from the source spread in all directions and rapidly attenuated in the conductive subsurface sediments due to water-filled pores. The rate of decay in the amplitude and the phase shift of the signal are controlled by geometric and skin depth effects [6]. Figure 1 from [14] shows the sea bed logging schematic diagram with electromagnetic (EM) wave transmitted are reflected, refracted and guided back from the hydrocarbon reservoir and detected by electric or magnetic field sensors placed on the sea floor [8-11]. Figure 1: Schematic Diagram of EM Transmitted and Reflected Waves with Air, Water, Sediments and Hydrocarbon Layers [14] In this paper, synthetic data are generated from EM simulator from [12]. This simulator is able to replicate the real sea bed environment and parameters are defined as required before running the simulation. In this work, parameters such as frequency are varied from 0.5Hz, 0.25Hz, 0.125Hz and 0.0625Hz and sediment thickness are varied from 1000m to 3000m incremented every 250m. Data 10 10 m 1-2 m 30-500 m 1-2 m HC Half Space Overburden Water Air Ship Source Rx4 Rx1 Rx2 Rx3 2012 IEEE Asia-Pacific Conference on Applied Electromagnetics (APACE 2012), December 11 - 13, 2012, Melaka, Malaysia 9978-1-4673-3115-9/12/$31.00 ©2012 IEEE 110