MétaCan
Menu
Back to cohort
Record W3200170290 · doi:10.2118/205837-ms

A Data-Driven Workflow for Identifying Optimum Horizontal Subsurface Targets

2021· article· en· W3200170290 on OpenAlexaff
Amir Salehi, Izzet Arslan, Lichi Deng, Hamed Darabi, Johanna Smith, Sander Suicmez, David Castiñeira, E. Gringarten

Bibliographic record

VenueSPE Annual Technical Conference and Exhibition · 2021
Typearticle
Languageen
FieldEngineering
TopicReservoir Engineering and Simulation Methods
Canadian institutionsImpact
Fundersnot available
KeywordsComputer scienceWorkflowData miningSet (abstract data type)

Abstract

fetched live from OpenAlex

Abstract Horizontal well development often increases field production and recovery due to increased reservoir contact, reduced drawdown in the reservoir, and a more efficient drainage pattern. Successful field development requires an evergreen backlog of opportunities that can be pursued, which is extremely challenging and laborious to generate using traditional workflows. Here, we present a data-driven methodology to automatically deliver a feasible, actionable inventory by combining geological knowledge, reservoir performance, production history, completion information, and multi-disciplinary expertise. This technology relies on automated geologic and engineering workflows to identify areas with high relative probability of success (RPOS) and therefore productivity potential. The workflow incorporates multiple configuration and trajectory constraints for placement of the horizontal wells, such as length/azimuth/inclination range, zone-crossing, fault-avoidance, etc. The optimization engine is initialized with an ensemble of initial guesses generated with Latin-Hypercube Sampling (LHS) to ensure all regions of POS distribution in the model are evenly considered. The advanced optimization algorithm identifies potential target locations with 3D pay tracking globally, and the segments are further optimized using an interference analysis that selects the best set of non-interfering targets to maximize production. Advanced AI-based computational algorithms are implemented under numerous physical constraints to identify the best segments that maximize the RPOS. Statistical and machine learning techniques are combined to assess neighborhood performance and geologic risks, along with physics-based analytical and upscaled parametric models to forecast phase-based production and pressure behavior. Finally, a comprehensive vetting and sorting framework is presented to ensure the final set of identified opportunities is feasible for the field development plan, given the operational constraints. This methodology has been successfully applied to a mature field in the Middle East with more than 90 vertical well producers and 50 years of production history to identify horizontal target opportunities. Rapid decline in oil production and a subpar recovery factor were the primary incentives behind switching to horizontal development. The search covered both shorter laterals accessible as a side-track from existing wells to minimize water encroachment, and longer laterals that could be drilled as new wells. After filtering based on geo-engineering attributes and rigorous vetting by domain experts, the final catalog consisted of 32 horizontal targets. After careful consideration, the top five candidates were selected for execution in the short term with an estimated total oil gain of 40,000 STB/D. The introduced AI-based methodology has many advantages over traditional simulation-centric workflows that take months to build and calibrate a model. This framework automates steps typically performed during the selection of horizontal well candidates by applying advanced algorithms and AI/ML to multi-disciplinary datasets. This enables teams to rapidly run and review different scenarios, ultimately leading to better risk management and shorter decision cycles with more than 90% speedup compared to conventional workflows.

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 imitation

Not 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.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: Simulation or modeling
GenreCandidate signal: Methods · Consensus signal: none
Teacher disagreement score0.767
Threshold uncertainty score0.632

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.000
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0000.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

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.

Opus teacher head0.066
GPT teacher head0.327
Teacher spread0.261 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one teacher head, not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designSimulation or modeling
Domainnot available
GenreMethods

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".

Quick stats

Citations6
Published2021
Admission routes1
Has abstractyes

Explore more

Same venueSPE Annual Technical Conference and ExhibitionSame topicReservoir Engineering and Simulation MethodsFrench-language works237,207