Bayesian and Neural Network Approaches to Estimate Deep Temperature Distribution for Assessing a Supercritical Geothermal System: Evaluation Using a Numerical Model

2021 ◽  
Author(s):  
Kazuya Ishitsuka ◽  
Yosuke Kobayashi ◽  
Norihiro Watanabe ◽  
Yusuke Yamaya ◽  
Elvar Bjarkason ◽  
...  
Keyword(s):  
Neural Network ◽  
Temperature Distribution ◽  
Numerical Model ◽  
System Evaluation ◽  
Geothermal System ◽  
Deep Temperature ◽  
Network Approaches

scholarly journals Geothermal potential and genetic mechanisms of the geothermal system in the mountainous area on the northern margin of north china

2021 ◽  
pp. 014459872110506
Author(s):  
Feng Liu ◽  
Guiling Wang ◽  
Wei Zhang ◽  
Yizuo Shi ◽  
Chen Yue ◽  
...  
Keyword(s):  
Temperature Distribution ◽  
Heat Flow ◽  
Hot Springs ◽  
Geothermal System ◽  
Mountainous Area ◽  
Northern Margin ◽  
Average Value ◽  
Genetic Mechanisms ◽  
Deep Temperature ◽  
Flow Map

Geothermal resources as clean and renewable energy can be utilized for agriculture, tourism, and industry. The assessment of geothermal potential and the study of genetic mechanism of the geothermal system is an essential part of geothermal resource development. In this study, 16 steady-state temperature logs are obtained in the mountainous area on the northern margin of North China. Thermal conductivity and heat production rates are tested or collected from more than 200 rock samples of these wells and outcrops around the study area. Based on these data, for the first time, the detailed delineations of temperature distributions, genetic mechanisms of geothermal systems, and resource potential of Hot Dry Rock in the study area are achieved. The heat flow map indicates a low heat flow state with an average value of 53.1 mW/m2 in the study area, which is lower than the average value of 62.5 mW/m2 in mainland China. The distribution of hot springs in the area is mainly controlled by fault systems. Heat flow only exhibits a minor effect on the temperature of hot springs and geothermal wells. On this basis, the deep temperature distribution within 3–10 km depths of the study area is calculated using the one-dimensional steady-state heat conduction equation. With it, the reservoir depths of hot springs are estimated to be 3–5 km with temperatures ranging from 70°C to 110°C. Furthermore, a conceptual model for the geothermal system in the study area is derived. According to the results, Northeastern Chengde and northern Beijing exhibit the highest temperatures at all depths. Similar patterns are observed in the temperature distribution maps and the heat flow map, which suggest that the deep temperature distribution is mainly controlled by regional heat flow. With the depth increases, the temperature shows larger variation at each depth level, which is possibly caused by the heterogeneity of crustal composition. According to our resource assessment by volumetric method, the exploitable potential of Hot Dry Rock within the depth of 7–10 km of the study area is equivalent to about 3.1 × 1011 tons of standard coal, but the barrier is still existing for development under the current technical and economic conditions.

scholarly journals Outlier detection in scatterometer data: neural network approaches

2003 ◽  
Vol 16 (3-4) ◽  
pp. 419-426 ◽  
Author(s):  
Robert J. Bullen ◽  
Dan Cornford ◽  
Ian T. Nabney
Keyword(s):  
Neural Network ◽  
Outlier Detection ◽  
Network Approaches

Multiphase Transient Slugging Flow in Subsea Oil and Gas Production

2016 ◽  
Author(s):  
Zhenhua Zhang ◽  
Longbin Tao
Keyword(s):  
Data Analysis ◽  
Temperature Distribution ◽  
Numerical Model ◽  
Slug Flow ◽  
Field Data ◽  
Oil And Gas ◽  
Gas Production ◽  
Worst Case ◽  
Oil And Gas Production ◽  
Significant Difference

Slug flow in horizontal pipelines and riser systems in deep sea has been proved as one of the challenging flow assurance issues. Large and fluctuating gas/liquid rates can severely reduce production and, in the worst case, shut down, depressurization or damage topside equipment, such as separator, vessels and compressors. Previous studies are primarily based on experimental investigations of fluid properties with air/water as working media in considerably scaled down model pipes, and the results cannot be simply extrapolated to full scale due to the significant difference in Reynolds number and other fluid conditions. In this paper, the focus is on utilizing practical shape of pipe, working conditions and fluid data for simulation and data analysis. The study aims to investigate the transient multiphase slug flow in subsea oil and gas production based on the field data, using numerical model developed by simulator OLGA and data analysis. As the first step, cases with field data have been modelled using OLGA and validated by comparing with the results obtained using PIPESYS in steady state analysis. Then, a numerical model to predict slugging flow characteristics under transient state in pipeline and riser system was set up using multiphase flow simulator OLGA. One of the highlights of the present study is the new transient model developed by OLGA with an added capacity of newly developed thermal model programmed with MATLAB in order to represent the large variable temperature distribution of the riser in deep water condition. The slug characteristics in pipelines and temperature distribution of riser are analyzed under the different temperature gradients along the water depth. Finally, the depressurization during a shut-down and then restart procedure considering hydrate formation checking is simulated. Furthermore, slug length, pressure drop and liquid hold up in the riser are predicted under the realistic field development scenarios.

scholarly journals International agricultural trade forecasting using machine learning

Data & Policy ◽  
2021 ◽  
Vol 3 ◽  
Author(s):  
Munisamy Gopinath ◽  
Feras A. Batarseh ◽  
Jayson Beckman ◽  
Ajay Kulkarni ◽  
Sei Jeong
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Agricultural Trade ◽  
Supervised Machine Learning ◽  
Trade Patterns ◽  
Subjective Assessments ◽  
History Of ◽  
Network Approaches ◽  
Traditional Approaches

Abstract Focusing on seven major agricultural commodities with a long history of trade, this study employs data-driven analytics to decipher patterns of trade, namely using supervised machine learning (ML), as well as neural networks. The supervised ML and neural network techniques are trained on data until 2010 and 2014, respectively. Results show the high relevance of ML models to forecasting trade patterns in near- and long-term relative to traditional approaches, which are often subjective assessments or time-series projections. While supervised ML techniques quantified key economic factors underlying agricultural trade flows, neural network approaches provide better fits over the long term.

Sign in / Sign up

Export Citation Format

Share Document