Fuzzy support vector regression for permeability estimation of petroleum reservoir using well logs

Moosavi, Nastaran; Bagheri, Majid; Nabi-Bidhendi, Majid; Heidari, Reza

doi:10.1007/s11600-021-00700-8

Cited by 11 publications

(2 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Establishing a quantitative relationship between GR, shale volume, and porosity (phi) data and the calculated FZI values from core data, using Multi Regression Graph-Based Clustering (MRGC) algorithm, to make model FZIs and HFUs. The MRGC approach will likely be implemented for estimating permeability using well logs data (Moosavi et al, 2022). We compared the predicted FZI with the FZI calculated from core porosity-permeability to ensure the model worked properly.…”

Section: Rock Type Analysismentioning

confidence: 99%

Petrophysical Study and Rock Type Determination of Siliciclastic Reservoir: Case Study Sand of Bekasap Formation, AF Field, Central Sumatra Basin, Indonesia

2023

View full text Add to dashboard Cite

An integrated subsurface study has been performed for a large and mature field at the Bekasap Formation in Central Sumatra Basin. The Bekasap Formation sand represents an undeveloped reservoir because of its heterogeneity. There are five sand reservoirs (BK1, BK2, BK3, BK4, and BK5) from Bekasap Formation, which each zone or layer bounded by a flooding surface. Each sand reservoir has particular characterization based on petrophysical properties that represent geological process. The petrophysical properties consist of shale volume, porosity, and water saturation obtained by wireline log calculation. This study uses conventional core data to validate the log calculation to achieve an accurate interpretation. Bekasap reservoir is a sandstone reservoir deposited in an estuarine with tide-dominated. Formation evaluation was done to determine the interest zone by petrophysical properties. The result well-log calculation and reservoir cut-offs showed the thickest reservoir in the BK 3 with the best average values of petrophysical properties with an average shale volume 0.32; porosity of 0.245. Otherwise, in rock type determination, four lithofacies are divided in the reservoir based on flow units. The sample RT 1 and RT2 provided the best reservoir zones with HFU1 and HFU2. The RT 3 and RT 4 dominated in HFU3 and HFU4 had the lowest potential zones of reservoir. The final findings showed a good correlation between sedimentologic analysis and petrophysical properties in the rock type determination. As a result, the best reservoir quality development is controlled by the depositional environment (texture and structure) rather than the diagenetic process in this reservoir. It is proven by petrophysical properties in BK1, and BK2 is coastal barrier sand (tidal sand bar) has more clean sand rather than in BK3 and BK4 deposited in the offshore bar.

show abstract

Section: Rock Type Analysismentioning

confidence: 99%

Petrophysical Study and Rock Type Determination of Siliciclastic Reservoir: Case Study Sand of Bekasap Formation, AF Field, Central Sumatra Basin, Indonesia

2023

View full text Add to dashboard Cite

show abstract

“…Thus, by adopting a suitable kernel function, SVR is capable to produce an expected fitting, especially for the nonlinear regression [26]. Based on the super power of SVR shown in the fitting, Al-Anazi and Gates [27] and other researchers launched the SVR-based predictions for some reservoir parameters and through a comparison verified that SVR is a potential candidate in the petrophysical prediction [28,29]. Nonetheless, as the support vectors produced by SVR are unexplainable, the practical meaning of them for each test sample becomes vague, and then a deeper analysis for the relationship between learning and test samples is inaccessible, which solidly indicates the major shortcoming of SVR in the regression.…”

Section: Introductionmentioning

confidence: 97%

Petrophysical Regression regarding Porosity, Permeability, and Water Saturation Driven by Logging-Based Ensemble and Transfer Learnings: A Case Study of Sandy-Mud Reservoirs

Zhang

Gao³

et al. 2022

Geofluids

View full text Add to dashboard Cite

From a general review, most petrophysical models applied for the conventional logging interpretation imply that porosity, permeability, or water saturation mathematically have a linear or nonlinear relationship with well logs, and then arguing the prediction of these three parameters actually is accessible under a regression of logging sequences. Based on this knowledge, ensemble learning technique, partially developed for fitting problems, can be regarded as a solution. Light gradient boosting machine (LightGBM) is proved as one representative of the state-of-the-art ensemble learning, thus adopted as a potential solver to predict three target reservoir characters. To guarantee the predicting quality of LightGBM, continuous restricted Boltzmann machine (CRBM) and Bayesian optimization (Bayes) are introduced as assistants to enhance the significance of input logs and the setting of employed hyperparameters. Thereby, a new hybrid predictor, named CRBM-Bayes-LightGBM, is proposed for the prediction task. To validate the working performance of the proposed predictor, the basic data derived from the member of Chang 8, Jiyuan Oilfield, Ordos Basin, Northern China, is collected to launch the corresponding experiments. Additionally, to highlight the validating effect, three sophisticated predictors, including k-nearest neighbors (KNN), support vector regression (SVR), and random forest (RF), are introduced as competitors to implement a contrast. Since ensemble learning models universally will cause an underfitting issue when dealing with a small-volumetric dataset, transfer learning in this circumstance will be employed as an aided technique for the core predictor to achieve a satisfactory prediction. Then, three experiments are purposefully designed for four validated predictors, and given a comprehensive analysis of the gained experimented results, two critical points are concluded: (1) compared to three competitors, LightGBM-cored predictor has capability to produce more reliable predicted results, and the reliability can be further improved under a usage of more learning samples; (2) transfer learning is really functional in completing a satisfactory prediction for a small-volumetric dataset and furthermore has access to perform better when serving for the proposed predictor. Consequently, CRBM-Bayes-LightGBM combined with transfer learning is solidly demonstrated by a stronger capability and an expected robustness on the prediction of porosity, permeability, and water saturation, which then clarify that the proposed predictor can be viewed as a preferential selection when geologists, geophysicists, or petrophysicists need to finalize a characterization of sandy-mud reservoirs.

show abstract

Fuzzy-twin proximal SVM kernel-based deep learning neural network model for hyperspectral image classification

Krishna

Jeya

Deepa

2022

Neural Comput & Applic

View full text Add to dashboard Cite

Fuzzy support vector regression for permeability estimation of petroleum reservoir using well logs

Cited by 11 publications

References 28 publications

Petrophysical Study and Rock Type Determination of Siliciclastic Reservoir: Case Study Sand of Bekasap Formation, AF Field, Central Sumatra Basin, Indonesia

Petrophysical Study and Rock Type Determination of Siliciclastic Reservoir: Case Study Sand of Bekasap Formation, AF Field, Central Sumatra Basin, Indonesia

Petrophysical Regression regarding Porosity, Permeability, and Water Saturation Driven by Logging-Based Ensemble and Transfer Learnings: A Case Study of Sandy-Mud Reservoirs

Fuzzy-twin proximal SVM kernel-based deep learning neural network model for hyperspectral image classification

Contact Info

Product

Resources

About