Automated Merging of Time Series and Textual Operations Data to Extract Technical Limiter Re-Design Recommendations
Bibliographic record
Abstract
Abstract During the construction planning phase of any new well, drilling engineers often look at offset well data to identify information that could be used to drill the new well more efficiently. This is generally a time-consuming process. The objective was to develop a recommender system that would automate the process of identifying potential hazards and current technical limiters. The developed methodology consisted of three parts. First, a system is developed that can parse textual information found in daily reports to identify key events that occurred in offset wells. Second, time series data from these same offset wells is processed to identify events directly from the patterns in the data, and a reconciliation is done between the time-series data and the contextual data wherever there is a conflict between the two datasets. Finally, KPIs are computed that enable the comparison of various drilling choices and their consequences across the set of offset wells and recommendation are automatically generated for improvements in the construction of a new well. The system was developed on a set of 7 recently drilled wells chosen from a specific North American land operation. The recommendations were compared to recommendation made through manual processing of the data for validation of the approach. The recommender identified invisible lost time (ILT) and potential non-productive time (NPT) scenarios, optimal depth-based drilling parameter for the new well, and recommendations on BHA and flat time improvement areas. Open-source natural language processing libraries were used in this project and were very effective in extracting events from textual data. An automated system was built to guide the drilling engineers in the planning phase of a well construction activity. Given a set of offset wells, the system combined both the time series data and textual data to arrive at these recommendations. The approach upon further refinement is expected to save 30 to 40 hours of the engineer's time per well and shorten the learning curve.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".