Modeling Rate of Penetration in a South Texas Oil Field with Aggregated Well Data in a Supercomputing Framework
Bibliographic record
Abstract
Abstract The objective of this work is to develop an accurate and practical tool for drilling engineers supporting operations to predict rate of penetration (ROP) in the tangent section of the wellbore using easily obtainable data. Historically, tacit knowledge has been used to predict both ROP and the ideal parameters. Currently, a "roadmap" of ideal parameters for each section of the wellbore is provided to site supervisors. This work was comprised of two major efforts: data acquisition and wrangling, and modeling. Data was obtained from three distinct sources: well files, drilling logs, and survey logs. This data was manually downloaded and imported into open-source software. Due to the size of the data, a university-owned high-performance computer (HPC) was required to process the data. Special care was given to optimizing for memory efficiencies that allowed the HPC to perform these operations. A test data set of 5 wells was used to pilot the data wrangling process and initial linear regression models. Four different model types were produced and evaluated: linear regression, polynomial regression, nonlinear regression, and neural networks. Neural networks provided the best prediction with a R2 of 0.85, with linear models similar in performance with much more usability. The most important variables affecting ROP in the tangent section in descending order are total pump output, rotary speed, hook load, differential pressure, and bit type. To the best of our knowledge, this is the largest dataset of ROP data found in literature, containing over 350 wells and 30 million rows of data. This workflow can be adopted to create other field-specific models or adapted to evaluate other sections of the wellbore. More immediately, this work creates a large database ready to be utilized for developing other models for other sections of the wellbore or undergirded by different computational methodologies.
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 | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| 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".