Clasificación y determinación de litofacies aplicando metodologías de Inteligencia Artificial en el Grupo Real de la cuenca del Valle Medio del Magdalena a partir de Registros Eléctricos

dc.contributor.advisorOchoa Gutiérrez, Luis Hernánspa
dc.contributor.authorGonzález Chacón, Andrés Felipespa
dc.contributor.researchgroupMEGIAspa
dc.date.accessioned2025-07-24T19:58:07Z
dc.date.available2025-07-24T19:58:07Z
dc.date.issued2024
dc.description.abstractEl Grupo MEGIA viene realizando durante los últimos años diferentes tipos de estudios que ayudan a estudiar y entender el comportamiento de los acuíferos presentes en la Cuenca del Valle Medio del Magdalena (VMM). Uno de dichos estudios fue el petrofísico, donde litológicamente se clasificaron diferentes formaciones y facies dentro de la unidad de interés; este trabajo es una extensión de dichos estudios petrofísicos, usando metodologías de Inteligencia Artificial (IA) entre ellas diferentes algoritmos de Machine Learning (ML) que ayudan a clasificar y predecir diferentes propiedades y características de la cuenca. En la presente investigación se utilizaron dichas técnicas para replicar metodologías clásicas en petrofísica, para predecir y modelar curvas faltantes en los pozos, y para clasificar y predecir litofacies, el cual es el objetivo principal de este proyecto. Para los cálculos petrofísicos se utilizaron las ecuaciones convencionales, calculando entonces la porosidad, volumen de arcilla y saturación de agua en los pozos, a partir de curvas litológicas, como la Gamma Ray (GR) y el Factor Fotoeléctrico (PEF), además de curvas de porosidad, como la sónica (DT), densidad (RHOB) y la de Neutrón (NPHI). Para la predicción de curvas faltantes fueron necesarios diferentes mecanismos de correlación para determinar las mejores curvas que se ajusten a cada registro que se quiere predecir. Se pudo definir una buena correlación entre las curvas GR y RHOB, además de NPHI con PE. Para la predicción de la curva sónica, se utilizaron todas las curvas previamente nombradas, y se pudo hacer una buena predicción, con un ligero desfase en los intervalos más arcillosos. Se optimizaron los hiperparámetros de Random Forest, viendo un incremento no tan considerable, dado que los hiperparámetros estándar, dieron un excelente ajuste. Para la clasificación de facies se usó un método de Aprendizaje No Supervisado, KMneas, dando por lo general agrupamiento de 3 grupos de Facies, las de Lodolita, Arenita Limpia y Arenita Lodosa. Para la clasificación de las facies se usaron métodos de Aprendizaje Supervisado, utilizando el número de clústeres ya usado. Dentro de los algoritmos utilizados, el de mejores resultados de precisión, fue el de Árbol de Decisión, con valores incluso del 99% (Texto tomado de la fuente).spa
dc.description.abstractThe MEGIA Group has been conducting various types of studies in recent years to help study and understand the behavior of aquifers in the Middle Magdalena Valley Basin (VMM). One of these studies was petrophysical, where different formations and units within the unit of interest were lithologically classified. This work is an extension of these petrophysical studies, using artificial intelligence (AI) methodologies, including various machine learning (ML) algorithms to help classify and predict different properties and characteristics of the basin. In this work, these techniques were used to replicate classical petrophysical methodologies, to predict and model missing petrophysical well logs in wells, and to classify and predict lithologies, which is the main objective of this document. For petrophysical calculations, conventional equations were used to calculate porosity, shale volume, and water saturation in wells from lithological curves, such as Gamma Ray (GR) and Photoelectric Factor (PEF), and also from porosity curves, such as sonic (DT), density (RHOB), and neutron (NPHI). For the prediction of missing curves, different correlation mechanisms were necessary to determine the best well logs that fit each record to be predicted. A good correlation could be defined between the GR and RHOB well logs, as well as NPHI with PE. The Random Forest hyperparameters were optimized, seeing a not so considerable increase, since the standard hyperparameters gave an excellent fit. For the prediction of the sonic well log, all the previously named curves were used, and a good prediction could be made, with a slight lag in the most clayey intervals. For the classification of facies, a non-supervised learning method, the K-means method, was used, generally giving a grouping of 3 groups of facies, namely mudstone, clean sand, and loamy sand. Supervised learning methods were used for the classification of facies, using the number of clusters already used. Among the algorithms used, the one with the best accuracy results was the Decision Tree, with values of even 99% accuracy.eng
dc.description.degreelevelMaestríaspa
dc.description.degreenameMagister en Ciencias – Geologíaspa
dc.description.methodsPuede decirse que el presente capítulo va a presentar dos tipos de metodología: La metodología clásica del cálculo de propiedades petrofísicas, y la metodología Machine Learning (que es la base del proyecto). Esto con el fin de comparar, analizar los pasos y resultados de cada manera de hacer petrofísica.spa
dc.description.researchareaAplicación de técnicas IA/ML en las geocienciasspa
dc.format.extent166 páginasspa
dc.format.mimetypeapplication/pdfspa
dc.identifier.instnameUniversidad Nacional de Colombiaspa
dc.identifier.reponameRepositorio Institucional Universidad Nacional de Colombiaspa
dc.identifier.repourlhttps://repositorio.unal.edu.co/spa
dc.identifier.urihttps://repositorio.unal.edu.co/handle/unal/88380
dc.language.isospaspa
dc.publisherUniversidad Nacional de Colombiaspa
dc.publisher.branchUniversidad Nacional de Colombia - Sede Bogotáspa
dc.publisher.facultyFacultad de Cienciasspa
dc.publisher.placeBogotá, Colombiaspa
dc.publisher.programBogotá - Ciencias - Maestría en Ciencias - Geologíaspa
dc.relation.referencesAly, M.H., Rodgers, D.W., Thackray, G.D., and Hughes, S.S. 2009. Recent magmatotectonic activity in the Eastern Snake River Plain–Island Park region revealed by SAR interferometry. Journal of Volcanology and Geothermal Research 188: 297-304.spa
dc.relation.referencesBedrosian, P.A., and Feucht, D.W. 2014. Structure and tectonics of the northwestern United States from EarthScope USArray magnetotelluric data. Earth and Planetary Science Letters 402: 275-289.spa
dc.relation.referencesBestagini, P., Lipari, V., and Tubaro, S. 2017. A machine learning approach to facies classification using well logs.spa
dc.relation.referencesBishop, C.M. 2006. Pattern Recognition and Machine Learning. Singapore: Springer Science Business Media, LLC.spa
dc.relation.referencesCarmona, P., Climent, F., and Momparler, A. 2018. Predicting failure in the U.S. banking sector: An extreme gradient boosting approach. International Review of Economics & Finance.spa
dc.relation.referencesChen, T., and Guestrin, C. 2016. XGBoost. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '16, 785-794.spa
dc.relation.referencesDe Boissieu, F., Sevin, B., Cudahy, T., Mangeas, M., Chevrel, S., Ong, C., Rodger, A., Maurizot, P., Laukamp, C., Lau, I., Touraivane, T., Cluzel, D., and Despinoy, M. 2018. Regolith-geology mapping with support vector machine: A case study over weathered Ni-bearing peridotites, New Caledonia. International Journal of Applied Earth Observation and Geoinformation 64: 377-385.spa
dc.relation.referencesDeNosaquo, K.R., Smith, R.B., and Lowry, A.R. 2009. Density and lithospheric strength models of the Yellowstone–Snake River Plain volcanic system from gravity and heat flow data. Journal of Volcanology and Geothermal Research 188: 108-127.spa
dc.relation.referencesThongsame at al., 2018. Lithological classification from machine learning algorithm, Vol. 9, 31-43spa
dc.relation.referencesDevak, M., Dhanya, C.T., and Gosain, A.K. 2015. Dynamic coupling of support vector machine and K-nearest neighbour for downscaling daily rainfall. Journal of Hydrology 525: 286-301.spa
dc.relation.referencesDubois, M.K., Bohling, G.C., and Chakrabarti, S. 2007. Comparison of four approaches to a rock facies classification problem. Computers & Geosciences 33: 599-617.spa
dc.relation.referencesFan, J., Wang, X., Wu, L., Zhou, H., Zhang, F., Yu, X., Lu, X., and Xiang, Y. 2018. Comparison of Support Vector Machine and Extreme Gradient Boosting for predicting daily global solar radiation using temperature and precipitation in humid subtropical climates: A case study in China. Energy Conversion and Management 164: 102-111.spa
dc.relation.referencesFriedl, M.A., and Brodley, C.E. 1997. Decision tree classification of land cover from remotely sensed data. Remote Sensing of Environment 61: 399-409.spa
dc.relation.referencesGlowacz, A., and Glowacz, Z. 2016. Recognition of images of finger skin with application of histogram, image filtration and K-NN classifier. Biocybernetics and Biomedical Engineering 36: 95-101.spa
dc.relation.referencesHumphreys, E.D. 1995. Post-Laramide removal of the Farallon slab, western United States. Geology 23.spa
dc.relation.referencesKonaté, A.A., Pan, H., Khalid, M.A., Li, G., Yang, J.H., Deng, C., and Fang, S. 2015. Machine Learning Interpretation of Conventional Well Logs in Crystalline Rocks. pp. 360-370. Cham. Springer International Publishing.spa
dc.relation.referencesLonsdale, P. 2005. Creation of the Cocos and Nazca plates by fission of the Farallon plate. Tectonophysics 404: 237-264.spa
dc.relation.referencesLopes, R., and Jorge, A. 2017. Mind the Gap: A Well Log Data Analysis. The Workshop on Data Mining for Oil and Gas, Houston, Texas, USA. 1-6.spa
dc.relation.referencesRastegarzadeh, G., and Nemati, M. 2018. Primary mass discrimination of high energy cosmic rays using PNN and k-NN methods. Advances in Space Research 61: 1181-1191.spa
dc.relation.referencesShervais, J.W., Schmitt, D.R., Nielson, D., Evans, J.P., Christiansen, E.H., Morgan, L., Pat Shanks, W.C., Prokopenko, A.A., Lachmar, T., Liberty, L.M., Blackwell, D.D., Glen, J.M., Champion, L.D., Potter, K.E., and Kessler, J.A. 2013. First Results from HOTSPOT: The Snake River Plain Scientific Drilling Project, Idaho, U.S.A. Sci. Dril. 15: 36-45.spa
dc.relation.referencesSmirnoff, A., Boisvert, E., and Paradis, S.J. 2008. Support vector machine for 3D modelling from sparse geological information of various origins. Computers & Geosciences 34: 127-143.spa
dc.relation.referencesSoltani, S., Kordestani, M., and Karim Aghaee, P. 2016. New estimation methodologies for well logging problems via a combination of fuzzy Kalman filter and different smoothers. Journal of Petroleum Science and Engineering 145: 704-710.spa
dc.relation.referencesSteinbach, M., and Tan, P.-N. 2009. kNN: k-Nearest Neighbors. In Wu, X. and Kumar, V. (eds.), The Top Ten Algorithms in Data Mining, pp. 151-161. New York: Taylor & Francis Group, LLC.spa
dc.relation.referencesTester, J.W., Anderson, B.J., Batchelor, A.S., Blackwell, D.D., DiPippo, R., Drake, E.M., Garnish, J., Livesay, B., Moore, M.C., Nichols, K., Petty, S., Toksöz, M.N., Ralph W. Veatch, J., Baria, R., Augustine, C., Murphy, E., Negraru, P., and Richards, M. 2006. Geothermal ResourceBase Assessment. In The future of geothermal energy: impact of Enhanced Geothermal Systems (EGS) on the United States in the 21st Century, pp. 35-81. Geothermal Program, MS 3830 Renewable Energy and Power Department Idaho National Laboratory: U.S. Department of Energy.spa
dc.rights.accessrightsinfo:eu-repo/semantics/openAccessspa
dc.rights.licenseReconocimiento 4.0 Internacionalspa
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/spa
dc.subject.ddc550 - Ciencias de la tierra::551 - Geología, hidrología, meteorologíaspa
dc.subject.ddc550 - Ciencias de la tierra::552 - Petrologíaspa
dc.subject.lembFACIES (GEOLOGIA)spa
dc.subject.lembFacies (geology)eng
dc.subject.lembESTRATIGRAFIAspa
dc.subject.lembGeology, Stratigraphiceng
dc.subject.lembPETROLOGIAspa
dc.subject.lembPetrologyeng
dc.subject.lembAPRENDIZAJE AUTOMATICO (INTELIGENCIA ARTIFICIAL)spa
dc.subject.lembMachine learningeng
dc.subject.lembINTELIGENCIA ARTIFICIALspa
dc.subject.lembArtificial intelligenceeng
dc.subject.lembINTELIGENCIA ARTIFICIAL-PROCESAMIENTO DE DATOSspa
dc.subject.lembArtificial intelligen - data processingeng
dc.subject.lembACUIFEROSspa
dc.subject.lembAquiferseng
dc.subject.lembAGUAS SUBTERRANEASspa
dc.subject.lembWater, undergroundeng
dc.subject.proposalPetrofísicaspa
dc.subject.proposalGrupo Realspa
dc.subject.proposalInteligencia Artificialspa
dc.subject.proposalMachine Learningeng
dc.subject.proposalCLasificaciónspa
dc.subject.proposalPredicciónspa
dc.subject.proposalLitofaciesspa
dc.subject.proposalPetrophysicseng
dc.subject.proposalArtificial Intelligenceeng
dc.subject.proposalMEGIA Groupeng
dc.subject.proposalArtificial Intelligenceeng
dc.subject.proposalMachine Learningeng
dc.subject.proposalClassificationeng
dc.subject.proposalPredictioneng
dc.subject.proposalLithologieseng
dc.titleClasificación y determinación de litofacies aplicando metodologías de Inteligencia Artificial en el Grupo Real de la cuenca del Valle Medio del Magdalena a partir de Registros Eléctricosspa
dc.title.translatedLithofacies classification and determination using Artificial Intelligence methodologies in the Real Group of the Middle Magdalena Valley basin based on well log dataeng
dc.typeTrabajo de grado - Maestríaspa
dc.type.coarhttp://purl.org/coar/resource_type/c_bdccspa
dc.type.coarversionhttp://purl.org/coar/version/c_ab4af688f83e57aaspa
dc.type.contentTextspa
dc.type.driverinfo:eu-repo/semantics/masterThesisspa
dc.type.redcolhttp://purl.org/redcol/resource_type/TMspa
dc.type.versioninfo:eu-repo/semantics/acceptedVersionspa
dcterms.audience.professionaldevelopmentBibliotecariosspa
dcterms.audience.professionaldevelopmentConsejerosspa
dcterms.audience.professionaldevelopmentEstudiantesspa
dcterms.audience.professionaldevelopmentInvestigadoresspa
dcterms.audience.professionaldevelopmentMaestrosspa
dcterms.audience.professionaldevelopmentPadres y familiasspa
dcterms.audience.professionaldevelopmentPúblico generalspa
oaire.accessrightshttp://purl.org/coar/access_right/c_abf2spa

Archivos

Bloque original

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
1015438844.2025.pdf
Tamaño:
11.42 MB
Formato:
Adobe Portable Document Format
Descripción:
Tesis de Maestría en Ciencias – Geología

Bloque de licencias

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
license.txt
Tamaño:
5.74 KB
Formato:
Item-specific license agreed upon to submission
Descripción: