1887

Abstract

Summary

Well completion reports are a vital source for subsurface and operational data, crucial for oil and gas exploration and production. These reports, containing drilling summaries, formation reports, cuttings descriptions, geochemical analysis, production data, and well plans, are generated by numerous operators over decades. However, their heterogeneity in structure, language, and format, compounded by issues like handwritten entries and poor-quality scans, presents significant challenges for extracting actionable insights.

To demonstrate the potential of digitizing such legacy data, 20 well completion reports from Equinor (formerly Statoil), dating back to the 1990s, were collected. These reports include drilling summaries and serve as valuable resources for offset well analysis in new exploration projects.

To address the challenge in accessing these vital sources, a Retrieval-Augmented Generation (RAG) framework integrated with Large Language Models (LLMs) and Large Vision Models (LVMs) was implemented. This approach digitizes, structures, and unlocks hidden insights within legacy reports, providing seamless access to valuable knowledge. The framework empowers oil and gas professionals to make informed decisions, improving efficiency and outcomes in exploration and production activities.

© EAGE Publications BV
Loading

Article metrics loading...

/content/papers/10.3997/2214-4609.202539061
2025-03-24
2026-02-06

  • From This Site

    /content/papers/10.3997/2214-4609.202539061
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Journal -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. Lewis, Patrick et al., “Retrieval-augmented generation for knowledge-intensive nlp tasks.” Advances in Neural Information Processing Systems33 (2020): 9459–9474.
     [Google Scholar]
  2. Gao, Yunfan et al., “Retrieval-augmented generation for large language models: A survey.” arXiv preprint arXiv:2312.10997 (2023)
     [Google Scholar]
  3. Abdin, Marah et al., “Phi-3 technical report: A highly capable language model locally on your phone.” arXiv preprint arXiv:2404.14219 (2024).
     [Google Scholar]
  4. Achiam, Josh et al., “Gpt-4 technical report.” arXiv preprint arXiv:2303.08774 (2023).
     [Google Scholar]
  5. Dubey, Abhimanyu et al., “The llama 3 herd of models.” arXiv preprint arXiv:2407.21783 (2024).
     [Google Scholar]
  6. Goo, J., Chovanetz, K., Chuah, C., Macleod, A., Liew, A., Hardev, P., Jong, S., Affandi, S., Garg, R., and D.Vivegananthan. “Automated Drilling Program through Digital Well Construction Planning Framework: A Case Study from Malaysia.” Paper presented at the SPE/IADC Asia Pacific Drilling Technology Conference and Exhibition, Bangkok, Thailand, August 2024. doi: https://doi.org/10.2118/219601-MS
     [Google Scholar]
  7. Singh, Ajay, Jia, Tianxia, and Varun Nalagatla. “Generative AI Enabled Conversational Chatbot for Drilling and Production Analytics.” Paper presented at the ADIPEC, Abu Dhabi, UAE, October 2023. doi: https://doi.org/10.2118/216267-MS
     [Google Scholar]
  8. Dubovik, A. and et al. “Advanced LLM Applications in Oil & Gas: Cross-Domain Implementations”, 2024, https://doi.org/10.3997/2214-4609.2024101556”. EAGE Annual 2023
     [Google Scholar]
/content/papers/10.3997/2214-4609.202539061
Loading
Unveiling Hidden Insights: A GenAI-Powered RAG Framework for Legacy Well Completion Reports
Conference Proceedings 2025, 1 (2025); https://doi.org/10.3997/2214-4609.202539061
/content/papers/10.3997/2214-4609.202539061
/content/papers/10.3997/2214-4609.202539061
Loading

Data & Media loading...

Most Cited This Month Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error