Fifteenth International Congress of the Brazilian Geophysical Society Determining the mineral composition in quaternary sediments of Amazon River Fan through regression techniques, basic well logs and gamma ray spectrometry Jefferson Falcão & Abel Carrasquilla, UENF – Macaé - RJ Copyright 2017, SBGf - Sociedade Brasileira de Geofísica This paper was prepared for presentation during the 15 th International Congress of the Brazilian Geophysical Society held in Rio de Janeiro, Brazil, 31 July to 3 August, 2017. Contents of this paper were reviewed by the Technical Committee of the 15 th Interna- tional Congress of the Brazilian Geophysical Society and do not necessarily represent any position of the SBGf, its officers or members. Electronic reproduction or storage of any part of this paper for commercial purposes without the written consent of the Brazilian Geophysical Society is prohibited. ____________________________________________________________________ Abstract The geophysical well logging records physical properties of the subsurface rocks that are useful to the exploitation of oil and mineral reserves. Through the measurements of the physical properties made from basic geophysical logs, it is possible to infer information such as: density, porosi- ty, shaliness, fluid type and saturation. More elaborate information about subsurface as: pictures, throat size of pores, and minerals can be acquired by application of advanced logs as, for example, the resistive and acousti- cal images, nuclear magnetic resonance and geochemi- cal logs. This work uses advanced geochemical log for mineralogical analysis of wells from the Amazon Fan through data collected by 155 IODP (International Ocean Discovery Program) expedition. This analysis involves the generation of basic theoretical logs, using mineral oxides derived from geochemical logging as input. These syn- thetic logs are compared with actual basic logs by the Proportional Integral Derivative (PID) controller and through this comparison, mineral oxides volumes are set. In addition to the mineralogical fit, some applications with basic and advanced loggings are made in order to know the composition of the sediments in Amazon Fan, esti- mating parameters such as: shaliness, matrix, porosity and fluid through the basic logs, and rocks classification through the geochemical log data. Introduction Geophysical well logging is one of the methods widely used in the characterization of geological formations in teh subsurface (Desbrandes, 1985.). Some relevant in- formation commonly obtained by the logs is: well diame- ter, layer thickness, porosity, density, argilosity, fluid types and lithology of the formations. The other more specific information is: size of the grains and pores, permeability and mineralogy (Serra, 1984). In relation to mineralogy, this can be obtained by the application of the geochemical log (GLT). This tool is basically composed of the components of natural gamma ray detection (NGT), aluminum activation (AACT) and gamma ray spectrometry (GST). Together, they allow the detection of up to 10 elements. The spectrum energy of these gamma rays is analyzed and the concentration of some elements in the formation is obtained. This is achieved by the emission of high energy neutrons from an artificial source installed in the logging tool, which then interacts with the drilling hole, geological formations and drilling fluid (Crain, 2015). The analysis of the mineralogical composition of the rocks, therefore, offers the possibility of a better geologi- cal characterization in the wells. Among its primary appli- cations are the well-to-well correlation, grain size calcula- tion, sandstone classification, thermal conductivity estima- tion, rock generation, among others. These logs are very useful in the evaluation of formations, for example the determination of clay volumes may be more accurate and, therefore, the estimation of the saturation of the fluids is also improved (Oord, 1991). According to Debrabant et al. (1997), the Amazonian Fan is the third largest underwater fan in the modern oceans. It extends for about 700 km from the continental shelf, towards the sea, has an area of approximately 375000 km2 and a water depth of 1000 - 4300 m. Through it, a huge area of sinuous channels approximately 1200 km from the mouth of the Amazon River spreads radially (Pirmez et al., 1997). In it can be found large channel fluvial deposits, which are usually marked by the grain descending and decreasing the thickness of the sedimen- tary packages. These deposits originate from a very large sediment discharge from the Amazon River (Figure 1). Regarding these same authors clay - rich deposits, and high sedimentation rates often characterize large fans. In addition, the variation of the distribution of these clay minerals actually expresses successive stages of tecton- ic, climatic and eustatic events, which end up controlling the size of the fan (Oliveira, 2009). In accord to Flood et al. (1995), the river carries a sedi- ment load that corresponds to 10% of the fluvial sedi- ments of the planet, which are discharged on the Amazo- nian Fan since the Andean uplift in the Miocene. This region has large sedimentary deposits that are true rec- ords of continental climate, ocean circulation and tectonic activities over time. Currently, sediments from the river do not reach beyond the continental shelf. This is because the sea level is high, but during the glacial period, when sea level was low, sediments were discharged directly into the underwater canyon (Rabinowitz et al., 1994). This work has as main objective to analyze and apply regression techniques in basic logs and geochemical data (gamma ray induction geochemical) obtained from the Amazon River Fan, collected by the International Ocean Discovery Program (IODP). These regression techniques aim to adjust the compositions of mineral oxides of the Amazonian Fan. A computer program developed by Cae-