Estimation of Lost Circulation Zones Using Random Forest Classification and Image Logs

Lost circulation is one of the most chronic and most expensive problems in the drilling process as it leads to enormous losses in terms of finances and time. The conventional prediction methods mainly are based on the traditional well logs and correlations which have been empirically determined and which fully exploit the rich textual information available in the Formation Microresistivity Imaging (FMI) logs. The work is the innovative supervised machine learning model based on the use of the Random Forest classification algorithms to estimate the lost circulation areas based on the combination of the traditional Logging While Drilling (LWD) data with the texture features of the FMI image logs. The methodology was used on a large dataset of six wells of the Niger Delta, and this includes more than lost circulation events recorded by drilling reports. The conventional logs (gamma ray, resistivity, bulk density, caliper) were used as the input parameters complemented with the texture descriptors based on the FMI images determined by the Gray Level Co-occurrence Matrix (GLCM) methods. GLCM texture characteristics such as contrast, correlation, homogeneity and energy were calculated with the help of Open CV and at 0, 45, 90, and 135 angular orientations. Preprocessing of data included removing outliers, lithological coding, and depth matching of traditional logs with image-based characteristics. Random Forest classifier was trained on 70% of the data and was validated on the 30% of the data with 100 estimators and the maximum depth of 10 implemented in scikit-learn with Python. Accuracy, precision, recall, F1-score and ROC-AUC were used to measure model performance. Findings indicate that the combined method had an overall prediction accuracy of 86% and the image-based texture characterizations increased accuracy by 14% points as opposed to the traditional log-only models (71% to 86%). The ROC-AUC rose by 0.79 in case of conventional features and 0.91 of the combined feature set. The analysis of the feature importance showed that the most important features were texture contrast and caliper measurements. The methodology will allow the identification of lost circulation areas at an initial stage during the real-time implementation of drilling activities and allows taking proactive measures to mitigate the latter and improve the productivity of operations.

Lost Circulation, FMI Image Logs, GLCM, Random Forest, Machine Learning

1. Patel, H., Shah, S., et al. (2021). "Research on an Lost Circulation Zone Location Method Based on Transient Pressure Wave Signal." ACS Omega, 6, 25755-25769. doi:10.1021/acsomega.1c04359 [Google Scholar] [Crossref]

2. Musa, A. (2012). "Textural analysis of FMI logs to determine lithofacies in the early Paddock member." Master's Thesis, Texas Tech University. [Google Scholar] [Crossref]

3. Zhang, Y., Liu, H., Wang, L. (2023). "Application of Random Forest method in oil and water layer identification of logging data: a case study of the Liaohe depression." Earth Sciences Research Journal, 27(1), 69-75. [Google Scholar] [Crossref]

4. Kumar, S., et al. (2023). "Prediction of drilling fluid lost-circulation zone based on deep learning." Energy, 276, 127495. [Google Scholar] [Crossref]

5. Shafiabadi, M., et al. (2021). "Identification of reservoir fractures on FMI image logs using Canny edge detection algorithm." Oil & Gas Science and Technology, 76, 10. [Google Scholar] [Crossref]

6. Gonçalves, I.F.A., et al. (2022). "Predicting oil field production using the Random Forest algorithm." Conference on Graphics, Patterns and Images, 35, 134-139. [Google Scholar] [Crossref]

7. Azadivash, A. (2025). "Lost circulation intensity characterization in drilling operations: Leveraging machine learning and well log data." Heliyon, 11(1), e39758. [Google Scholar] [Crossref]

8. FMI Logs Documentation. (2023). "Formation Micro Imaging (FMI) Logging Tools." Scribd Technical Document. [Google Scholar] [Crossref]

9. Wright, C., et al. (2024). "Effects of tuning decision trees in random forest regression on predicting porosity of a hydrocarbon reservoir." Analytical Methods, 16, 5890-5903. [Google Scholar] [Crossref]

10. OnePetro. (2020). "Prediction of Lost Circulation Zones using Support Vector Machine." 20th International Petroleum Technology Conference, Paper IPTC-20054-MS. [Google Scholar] [Crossref]

11. PowerPoint Presentation. (2015). "FMI log presentation." SlideShare Educational Content. [Google Scholar] [Crossref]

12. Gonçalves, I.F.A., et al. (2022). "Predicting oil field production using the Random Forest algorithm." PDF Document, SBC Portal. [Google Scholar] [Crossref]

13. Kumar, A., Sharma, P. (2023). "Prediction of drilling fluid lost-circulation zone based on deep learning." Harvard ADS Abstract, 2023. [Google Scholar] [Crossref]

14. EPA. (2025). "Borehole Formation Microresistivity Imaging (FMI)." US Environmental Protection Agency Technical Guide. [Google Scholar] [Crossref]

15. Pratama, I., et al. (2023). "Applying random forest to oil and gas exploration in Central Sumatra Basin." Journal of Petroleum Exploration and Production, 177, 234-245. [Google Scholar] [Crossref]

16. IDEAS RePEc. (2023). "Prediction of drilling fluid lost-circulation zone based on deep learning." Energy Economics, 276, 127495. [Google Scholar] [Crossref]

17. Ahmed, K., et al. (2018). "A review on the applications of image logs in structural analysis and characterization." Marine and Petroleum Geology, 95, 139-156. [Google Scholar] [Crossref]

18. Martinez, L., et al. (2022). "Reservoir facies classification based on random forest and geostatistics modeling." Journal of Petroleum Science and Engineering, 219, 111112. [Google Scholar] [Crossref]

19. Wang, S., et al. (2023). "Identification of Shale Lithofacies from FMI Images and ECS Logs Using Machine Learning." Processes, 11(10), 2982. [Google Scholar] [Crossref]

20. Silva, R., et al. (2021). "SAR Oil Spill Detection System through Random Forest Classifiers." Remote Sensing, 13(11), 2044. [Google Scholar] [Crossref]

21. GitHub Repository. (2021). "Feature-Extraction-Gray-Level-Co-occurrence-Matrix-GLCM-with-Python." GitHub Open Source Project. [Google Scholar] [Crossref]

22. Emmanuel, O.E., et al. (2024). "Investigation of Mud Related Wellbore Problems in Drilling Well-A in Niger Delta region." International Journal of Petroleum and Gas Exploration Management, 15(1), 1-14. [Google Scholar] [Crossref]

23. Stack Overflow. (2017). "How to find GLCM of an image in OpenCV or numpy?" Programming Q&A Platform. [Google Scholar] [Crossref]

24. Andrew, N.H. (2024). "Formation Fracture Gradient Correlation for Niger Delta Basin Natural Gas Wells." ASRIC Journal on Engineering Sciences, 5(1), 73-95. [Google Scholar] [Crossref]

25. Scribd Technical Document. (2025). "Introduction To Formation Micro Imaging (FMI) Logging Tools." Industry Documentation. [Google Scholar] [Crossref]

26. Echoview Software. (2025). "GLCM Texture Feature - Technical Documentation." Software Manual. [Google Scholar] [Crossref]

27. Okafor, P., et al. (2025). "A Comprehensive Evaluation of Geologic and Radiologic Impacts of Oil Drilling in Niger Delta." QUEST Research Journal, 10(5), 49-57. [Google Scholar] [Crossref]

28. Scribd Technical Document. (2025). "Processing and Interpretation of FMI Images." Industry Manual. [Google Scholar] [Crossref]

29. scikit-image Documentation. (2025). "GLCM Texture Features — skimage 0.25.2 documentation." Python Library Documentation. [Google Scholar] [Crossref]

30. OnePetro. (2023). "Analysis of Drilling Fluid Losses, Mitigation and Recovery while Drilling in Niger Delta." 23rd North American Intelligent Energy Conference, Paper 525887. [Google Scholar] [Crossref]

31. SLB. (2008). "FMI Fullbore Formation Microimager - Well Logging Services." Schlumberger Technical Documentation. [Google Scholar] [Crossref]

32. YouTube Educational Content. (2021). "Image classification using gray-level co-occurrence matrix (GLCM)." Machine Learning Tutorial. [Google Scholar] [Crossref]

33. Ahmed, M., et al. (2020). "Assessment of pore pressure, wellbore failure and reservoir stability in Niger Delta." Journal of African Earth Sciences, 171, 103892. [Google Scholar] [Crossref]

34. OpenCV Community. (2025). "GLCM texture analysis using OpenCV C++." Programming Forum. [Google Scholar] [Crossref]

35. OnePetro. (2020). "A Case Study of an Oil Field in Niger Delta Basin, Nigeria." 20th North American Intelligent Energy Conference, Paper 449586. [Google Scholar] [Crossref]

36. GSA Conference. (2020). "Application of fullbore formation micro imager (FMI) log for characterization of Cambay Shale, Gujarat, India." Geological Society of America Annual Meeting. [Google Scholar] [Crossref]

37. MTU Educational Resource. (2025). "Image Texture Feature Extraction Using GLCM Approach." Academic Publication. [Google Scholar] [Crossref]

38. Taylor & Francis. (2023). "An assessment of the effectiveness of drilling waste treatment in Niger Delta." Environmental Technology, Online First. [Google Scholar] [Crossref]

39. GeeksforGeeks. (2025). "Random Forest Classifier using Scikit-learn." Technical Tutorial Platform. [Google Scholar] [Crossref]

40. Academic Publisher. (2025). "Filling the gaps of microresistivity imaging data by combining structure dual generative adversarial networks." Journal of Geophysics and Engineering, 22(2), 494-508. [Google Scholar] [Crossref]

41. PMC Research Article. (2025). "Lost circulation intensity characterization in drilling operations: Leveraging machine learning and well log data." PMC Research Publication, PMC11699355. [Google Scholar] [Crossref]

42. DataCamp. (2024). "Random Forest Classification with Scikit-Learn." Educational Tutorial. [Google Scholar] [Crossref]

43. ArXiv Preprint. (2022). "A Case Study on the Classification of Lost Circulation Events During Drilling using Machine Learning Techniques." arXiv:2209.01607. [Google Scholar] [Crossref]

44. Data36. (2024). "Random Forest in Python (and coding it with Scikit-learn)." Data Science Tutorial. [Google Scholar] [Crossref]

45. ScienceDirect. (2025). "Response and multi-parameter determination of microresistivity logging based on machine learning." Geophysics Journal, 149, 100764. [Google Scholar] [Crossref]

46. ScienceDirect. (2023). "Machine learning for carbonate formation drilling: Mud loss prediction." Journal of Petroleum Science and Engineering, 215, 110564. [Google Scholar] [Crossref]

47. YouTube Educational. (2023). "Random Forest Algorithm Explained with Python and scikit-learn." Ryan & Matt Data Science Channel. [Google Scholar] [Crossref]

48. OnePetro. (2024). "Automated Borehole Image Interpretation Using Computer Vision and Deep Learning." SPE Journal, 29(12), 6918-6935. [Google Scholar] [Crossref]

49. Nature Scientific Reports. (2025). "Formation permeability estimation using mud loss data by deep learning." Scientific Reports, 15, 94617. [Google Scholar] [Crossref]

50. LabEx. (2017). "How to train Random Forest in scikit-learn." Programming Tutorial. [Google Scholar] [Crossref]

51. Authorea Preprint. (2025). "Using Generative Adversarial Networks to Translate Microresistivity Images." Academic Preprint. [Google Scholar] [Crossref]

52. scikit-learn Documentation. (2025). "RandomForestClassifier — scikit-learn 1.7.1 documentation." Official Library Documentation. [Google Scholar] [Crossref]

53. ScienceDirect. (2024). "Lithology identification using electrical imaging logging image and deep learning." Journal of Applied Geophysics, 220, 105252. [Google Scholar] [Crossref]

54. SSRN Preprint. (2024). "Enhancing drilling fluid lost-circulation prediction using model stacking ensemble technique." SSRN Working Paper, 4085366. [Google Scholar] [Crossref]

55. Towards Data Science. (2025). "Understanding Random Forest using Python (scikit-learn)." Medium Publication. [Google Scholar] [Crossref]

56. De Gruyter Online. (2023). "Utilizing microresistivity image logs to recognize conglomeratic channel architectures." Open Geosciences, 15(1), 20220497. [Google Scholar] [Crossref]

57. MSSANZ Conference. (2023). "Texture feature extraction for change detection in drill core images: A comparative study." 25th International Congress on Modelling and Simulation, Darwin, Australia. [Google Scholar] [Crossref]

58. Science.gov. (2018). "Co-occurrence matrix GLCM: Topics by Science.gov." Federal Science Portal. [Google Scholar] [Crossref]

59. JSAER. (2024). "Investigating the Potential for Brittle Failure of Subsurface Formation for Field XY in Niger-Delta during Drilling Operation." Journal of Scientific and Engineering Research, 11(4), 36-44. [Google Scholar] [Crossref]

60. SCIRP. (2016). "Case Study of Onshore Niger-Delta Sedimentary Basin." Journal of Geoscience and Environment Protection, 4, 61-71. [Google Scholar] [Crossref]

61. ACS Omega. (2022). "Application of Machine Learning Methods in Modeling the Loss of Drilling Fluids." ACS Omega, 7(22), 18638-18654. [Google Scholar] [Crossref]

62. Preprints.org. (2024). "Texture Image Classification Using Wavelet Haar Approach and Grey-Level Co-occurrence Matrix." Academic Preprint, 202404.0881. [Google Scholar] [Crossref]

63. ScienceDirect. (2023). "Evaluation of formation susceptibility and sand production potential in Niger Delta oil province." Upstream Oil and Gas Technology, 11, 100092. [Google Scholar] [Crossref]

64. OnePetro. (2025). "Multidimensional Lost Circulation Risk Quantification Assessment using Ensemble Machine Learning." SPE Journal, 30(5), 2285-2301. [Google Scholar] [Crossref]

65. CSEG Recorder. (2005). "Application of texture attribute analysis to 3D seismic data." CSEG Technical Publication, September 2005. [Google Scholar] [Crossref]

66. ScienceDirect. (2023). "Pore pressure prediction in offshore Niger delta using data-driven approach." Upstream Oil and Gas Technology, 11, 100089. [Google Scholar] [Crossref]

67. The Leading Edge. (2010). "Seismic texture analysis for reservoir prediction and characterization." SEG Publication, 29(9), 1116-1121. [Google Scholar] [Crossref]

68. USGS. (1995). "Niger Delta (Akata-Agbada) Petroleum System." USGS Open-File Report, 99-50H. [Google Scholar] [Crossref]

69. OnePetro. (2023). "Case Study of New Drills in Tom Boy Field in the Swamp Niger Delta." 23rd North American Intelligent Energy Conference, Paper 525919. [Google Scholar] [Crossref]

70. SCIRP. (2023). "Challenges of Groundwater Development and Supply in the Niger Delta." Journal of Water Resource and Protection, 15, 123-141. [Google Scholar] [Crossref]

71. Labelf AI. (2022). "What is Accuracy, Precision, Recall and F1 Score?" AI Blog Tutorial. [Google Scholar] [Crossref]

72. OnePetro. (2024). "Efficient Logging-While-Drilling Image Logs Interpretation Using Deep Learning." Petrophysics, 65(3), 365-387. [Google Scholar] [Crossref]

73. Statistically Relevant. (2023). "Understanding the Confusion Matrix and ROC Curves." Data Science Blog. [Google Scholar] [Crossref]

74. YouTube Educational. (2022). "Implementing Precision, Recall, F1, & AUC in Python." Machine Learning Tutorial. [Google Scholar] [Crossref]

75. LDEO-BRG. (2019). "LWD Image Documentation." Columbia University Technical Documentation. [Google Scholar] [Crossref]

76. Stack Overflow. (2019). "Plotting ROC curve from confusion matrix - python." Programming Q&A. [Google Scholar] [Crossref]

77. Uppsala University. (2024). "A deep learning approach for drilling tool condition monitoring." Master's Thesis, UPTEC Report. [Google Scholar] [Crossref]

78. SLB. (2022). "Logging While Drilling (LWD) Formation Evaluation Services." Schlumberger Technical Documentation. [Google Scholar] [Crossref]

79. LinkedIn Professional. (2025). "Evaluating Classification Models through Confusion Matrix and ROC Curves." Professional Article. [Google Scholar] [Crossref]

80. Neptune.ai. (2025). "F1 Score vs ROC AUC vs Accuracy vs PR AUC." Machine Learning Blog. [Google Scholar] [Crossref]

81. Baker Hughes. (2024). "LWD imaging services." Technical Service Documentation. [Google Scholar] [Crossref]

82. scikit-learn Documentation. (2025). "roc_curve — scikit-learn 1.7.1 documentation." Official Library Documentation. [Google Scholar] [Crossref]

83. Neptune.ai. (2025). "20 Evaluation Metrics for Binary Classification." Machine Learning Resource. [Google Scholar] [Crossref]

84. IODP Publications. (2013). "Logging while drilling." International Ocean Discovery Program Proceedings, Volume 338. [Google Scholar] [Crossref]

85. scikit-learn Documentation. (2007). "Receiver Operating Characteristic (ROC) with cross validation." Official Examples. [Google Scholar] [Crossref]

86. Google Developers. (2025). "Classification: Accuracy, recall, precision, and related metrics." Machine Learning Crash Course. [Google Scholar] [Crossref]

87. Geoplex. (2025). "Logging While Drilling (LWD) Services." Technical Service Documentation. [Google Scholar] [Crossref]

88. scikit-learn Documentation. (2025). "confusion_matrix — scikit-learn 1.7.1 documentation." Official Library Documentation. [Google Scholar] [Crossref]

89. Encord. (2024). "F1 Score in Machine Learning Explained." AI Platform Blog. [Google Scholar] [Crossref]

90. TAQA. (2023). "Logging While Drilling (LWD)." Technical Service Documentation. [Google Scholar] [Crossref]

• The Impact of Ownership Structure on Dividend Payout Policy of Listed Plantation Companies in Sri Lanka
• Urban Sustainability in North-East India: A Study through the lens of NER-SDG index
• Performance Assessment of Predictive Forecasting Techniques for Enhancing Hospital Supply Chain Efficiency in Healthcare Logistics
• The Fractured Self in Julian Barnes' Postmodern Fiction: Identity Crisis and Deflation in Metroland and the Sense of an Ending
• Impact of Flood on the Employment, Labour Productivity and Migration of Agricultural Labour in North Bihar