Use of Real-Time Dynamic Data to Generate Fracture Network Models
Bibliographic record
Abstract
Abstract This paper presents a new technique to generate fracture networks using static data integrated with real-time information (from smart wells) utilizing Artificial Neural Networks (ANN). Smart wells provide high-density of real-time dynamic data to be used not only for well operations but also for reservoir characterization. The latter has no application yet and we postulate that real-time data would provide significant information in updating fracture network models. We begin with a hypothetical model whose fracture network parameters and all other geological information are known. After generating the fracture networks with known characteristics using a commercial software package, the data were exported to a reservoir simulator and simulations were run. Smart wells were placed throughout the reservoir and put on production. These producers were completed in different fracture zones to create a representative pressure and production response that is affected by fracture properties. Input sensitivity analysis was performed to ensure high fracture sensitivity to production rate. We then constructed the fracture network of the reservoir employing the initially available static reservoir/well data and updated it using real-time dynamic data at different stages of production by means of artificial neural networks (ANN). Next, we compared the fracture networks obtained through this methodology and the original one to examine the proposed methodology accuracy. We also analyzed what type of real-time data would be useful in fine tuning the fracture network model. It was shown that the fracture network (or static) model could be continuously updated as more smart well data are obtained. The proposed technique contributes to managing reservoirs proactively in real-time. Introduction Naturally fractured reservoirs represent a significant percentage of oil throughout the world and have given considerable research attention. There are several fields where natural fractures are important for production and a significant proportion are in basement rocks. Our particular emphasis is on reservoirs with fracture-controlled production. This paper presents a methodology to use real-time dynamic data in mapping fracture orientation, density, length, aperture and conductivity. This allows automatic fracture network visualization at each subsequent stage of production. Intelligent well technology has enabled well operators to get timely information and response. Smart wells are equipped with surface and subsurface measurement tools and flow control devises that allow excellent well and reservoir surveillance measure. Smart well completion and control technology enables responsive actions that minimize asset risk and optimize production. New techniques in data mining capable of characterizing reservoir in real time are needed to utilize the dynamic nature of upstream oil operation and minimize the degree of uncertainty around exiting reservoir models using high density of data that are continuously generated by smart technology. These techniques can be integrated into the intelligent operation loop to update the reservoir static model. They contribute into accelerating modeling process and replacing the traditional method of viewing and interpreting production logs and then updating the model and finally managing reservoirs proactively in real-time. The application of real-time dynamic data has recently been the focus of research since the evolution of smart technology. Production data has many times been utilized as a proxy measure of fracture intensity. With the availability of extensive production data and scarcity of static data, the embedded signal in real-time dynamic data reveals much needed information to generate fracture network models integrated with the available static data. We propose a methodology that can be integrated into intelligent field concept to utilize smart production data to condition fracture network real time. This methodology can be interconnected into modern data integration to characterize fractured based reservoirs. artificial neural network (ANN) is a powerful tool to detect the underlying relationship between the historical fracture network parameters and performance data. Once the relationship has been established, the real-time dynamic can be used to update the fracture network.
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 | 0.001 | 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".