Automated Monte Carlo-based quantification and updating of geological uncertainty with borehole data (AutoBEL v1.0)

Yin, Zhen Yu; Strebelle, Sebastien; Caers, Jef

doi:10.5194/gmd-13-651-2020

Cited by 48 publications

(33 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Uncertainty quantification is at the heart of decision making, especially in subsurface applications. Uncertainty about the geological structures, rocks, and fluids is because of the lack of access to the subsurface geological medium [57,58]. The uncertainty in the prediction results of all machine learning models developed in this study was directly controlled by the uncertainty on the well log data used to develop these models which were highly controlled by the depth of investigation and vertical resolution of every logging tool.…”

Section: Evaluation Of the Developed Machine Learning Modelsmentioning

confidence: 99%

Application of Machine Learning in Evaluation of the Static Young’s Modulus for Sandstone Formations

2020

View full text Add to dashboard Cite

Prediction of the mechanical characteristics of the reservoir formations, such as static Young’s modulus (Estatic), is very important for the evaluation of the wellbore stability and development of the earth geomechanical model. Estatic considerably varies with the change in the lithology. Therefore, a robust model for Estatic prediction is needed. In this study, the predictability of Estatic for sandstone formation using four machine learning models was evaluated. The design parameters of the machine learning models were optimized to improve their predictability. The machine learning models were trained to estimate Estatic based on bulk formation density, compressional transit time, and shear transit time. The machine learning models were trained and tested using 592 well log data points and their corresponding core-derived Estatic values collected from one sandstone formation in well-A and then validated on 38 data points collected from a sandstone formation in well-B. Among the machine learning models developed in this work, Mamdani fuzzy interference system was the highly accurate model to predict Estatic for the validation data with an average absolute percentage error of only 1.56% and R of 0.999. The developed static Young’s modulus prediction models could help the new generation to characterize the formation rock with less cost and safe operation.

show abstract

Section: Evaluation Of the Developed Machine Learning Modelsmentioning

confidence: 99%

Application of Machine Learning in Evaluation of the Static Young’s Modulus for Sandstone Formations

2020

View full text Add to dashboard Cite

show abstract

“…This paper also makes a key contribution in extending the DF procedure through implementing ES-MDA [26] and demonstrating that the DF with ES-MDA [27] is more robust than the standard procedure proposed in Satija et al [21]. It also provides appropriate posterior uncertainty quantification with results that can be compared to those of the methods proposed in Yin et al [22]. The paper is structured in multiple sections.…”

Section: Introductionmentioning

confidence: 96%

“…In relation to this context, Scheidt et al [20] with Satija et al [21] introduced a new protocol for making decisions under uncertainty called Bayesian evidential learning (BEL). Based on the description provided in [19,22]. BEL relies mainly on data, model, prediction, and decision under Bayesianism scientific methodologies.…”

Section: Introductionmentioning

confidence: 99%

“…This results in it being less expensive by a computational amount when compared to the standard inversion methods. Despite the applications being still lower in number, DF was successful in applying case studies related to oil reservoir management, groundwater resources, and geothermal energy problems [21][22][23][24][25]. For instance, Satija et al [23] used DF to forecast the future reservoir performance by mapping prior predictions into low-dimensional canonical space and estimating the joint distributions of historical and forecast data through linear Gaussian regression; they conclude by stating that this method displayed uncertainty estimates for production forecasts that reasonably agreed with rejection sampling.…”

Section: Introductionmentioning

confidence: 99%

“…For instance, Satija et al [23] used DF to forecast the future reservoir performance by mapping prior predictions into low-dimensional canonical space and estimating the joint distributions of historical and forecast data through linear Gaussian regression; they conclude by stating that this method displayed uncertainty estimates for production forecasts that reasonably agreed with rejection sampling. Yin et al [22] proposed an extended approach based on direct forecasting, called direct forecasting of sequential model decompositions, in which both geological model parameters and borehole data are used simultaneously. The posterior results displayed large reductions of uncertainty both spatially, through a geological model and using gas volume predictions.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Managing Uncertainty in Geological CO2 Storage Using Bayesian Evidential Learning

Tadjer

Bratvold

2021

Energies

View full text Add to dashboard Cite

Carbon capture and storage (CCS) has been increasingly looking like a promising strategy to reduce CO2 emissions and meet the Paris agreement’s climate target. To ensure that CCS is safe and successful, an efficient monitoring program that will prevent storage reservoir leakage and drinking water contamination in groundwater aquifers must be implemented. However, geologic CO2 sequestration (GCS) sites are not completely certain about the geological properties, which makes it difficult to predict the behavior of the injected gases, CO2 brine leakage rates through wellbores, and CO2 plume migration. Significant effort is required to observe how CO2 behaves in reservoirs. A key question is: Will the CO2 injection and storage behave as expected, and can we anticipate leakages? History matching of reservoir models can mitigate uncertainty towards a predictive strategy. It could prove challenging to develop a set of history matching models that preserve geological realism. A new Bayesian evidential learning (BEL) protocol for uncertainty quantification was released through literature, as an alternative to the model-space inversion in the history-matching approach. Consequently, an ensemble of previous geological models was developed using a prior distribution’s Monte Carlo simulation, followed by direct forecasting (DF) for joint uncertainty quantification. The goal of this work is to use prior models to identify a statistical relationship between data prediction, ensemble models, and data variables, without any explicit model inversion. The paper also introduces a new DF implementation using an ensemble smoother and shows that the new implementation can make the computation more robust than the standard method. The Utsira saline aquifer west of Norway is used to exemplify BEL’s ability to predict the CO2 mass and leakages and improve decision support regarding CO2 storage projects.

show abstract

Multiscale model diagnostics

Mannseth

2024

Comput Geosci

View full text Add to dashboard Cite

I consider the problem of model diagnostics, that is, the problem of criticizing a model prior to history matching by comparing data to an ensemble of simulated data based on the prior model (prior predictions). If the data are not deemed as a credible prior prediction by the model diagnostics, some settings of the model should be changed before history matching is attempted. I particularly target methodologies that are computationally feasible for large models with large amounts of data. A multiscale methodology, that can be applied to analyze differences between data and prior predictions in a scale-by-scale fashion, is proposed for this purpose. The methodology is computationally inexpensive, straightforward to apply, and can handle correlated observation errors without making approximations. The multiscale methodology is tested on a set of toy models, on two simplistic reservoir models with synthetic data, and on real data and prior predictions from the Norne field. The tests include comparisons with a previously published method (termed the Mahalanobis methodology in this paper). For the Norne case, both methodologies led to the same decisions regarding whether to accept or discard the data as a credible prior prediction. The multiscale methodology led to correct decisions for the toy models and the simplistic reservoir models. For these models, the Mahalanobis methodology either led to incorrect decisions, and/or was unstable with respect to selection of the ensemble of prior predictions.

show abstract

Automated Monte Carlo-based quantification and updating of geological uncertainty with borehole data (AutoBEL v1.0)

Cited by 48 publications

References 59 publications

Application of Machine Learning in Evaluation of the Static Young’s Modulus for Sandstone Formations

Application of Machine Learning in Evaluation of the Static Young’s Modulus for Sandstone Formations

Managing Uncertainty in Geological CO2 Storage Using Bayesian Evidential Learning

Multiscale model diagnostics

Contact Info

Product

Resources

About