MétaCan
Menu
Back to cohort
Record W4402560686 · doi:10.1016/j.ijggc.2024.104238

Stochastic control of geological carbon storage operations using geophysical monitoring and deep reinforcement learning

2024· article· en· W4402560686 on OpenAlexafffund
Kyubo Noh, Andrei Swidinsky

Bibliographic record

VenueInternational journal of greenhouse gas control · 2024
Typearticle
Languageen
FieldEngineering
TopicReservoir Engineering and Simulation Methods
Canadian institutionsUniversity of Toronto
FundersNatural Sciences and Engineering Research Council of CanadaAlliance de recherche numérique du Canada
KeywordsReinforcement learningControl (management)ReinforcementEnvironmental scienceMining engineeringPetroleum engineeringEngineeringComputer scienceArtificial intelligenceStructural engineering

Abstract

fetched live from OpenAlex

Geological carbon storage (GCS) is the process of injecting and storing carbon dioxide (CO 2 ) in the subsurface to reduce greenhouse gas emissions. Safe and profitable GCS operations require effective decision-making in the presence of uncertain geological models, a process which can often be facilitated with geophysical monitoring. In this study, we examine how sequential decision-making algorithms can be combined with geophysical measurements for the optimal control of GCS operations. Specifically, we develop an artificial intelligence model using deep reinforcement learning (DRL) that takes geophysical time-lapse gravity and well pressure monitoring data as input and delivers an optimal CO 2 injection policy. The objective of the problem at hand is to maximize the profit of a hypothetical GCS operation while mitigating the potential for induced seismicity, by training DRL agents using combined geostatistical, reservoir and geophysical simulation. Comparisons against two benchmarks – a constant injection strategy and an injection schedule optimized using a commercial reservoir simulator toolbox – show that the stochastic control of such operations from subsurface monitoring data using deep reinforcement learning is feasible. Evaluation results show that DRL agent behavior generates profits which are on average 1 to 8 percent higher than what is possible through a constant injection approach. Furthermore, we show that DRL can generate optimal injection policies applicable to the true (yet previously unseen) subsurface given carefully managed levels of uncertainty.

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 machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.003
Version: metacan-v3-hybrid-931329e0061cValidation 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: Empirical · Consensus signal: none
Teacher disagreement score0.010
Threshold uncertainty score0.020

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0010.003
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.000
Science and technology studies0.0000.001
Scholarly communication0.0010.001
Open science0.0010.001
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0010.000

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 head0.014
GPT teacher head0.275
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 source (direct Gemma or distilled Codex), not a consensus.

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

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

Citations11
Published2024
Admission routes2
Has abstractyes

Explore more

Same venueInternational journal of greenhouse gas controlSame topicReservoir Engineering and Simulation MethodsFrench-language works237,207