Predicting the temperature, pressure, and permeability at depth is crucial for understanding natural-state geothermal systems. As direct observations of these quantities are limited to well ...locations, a reliable methodology that predicts the spatial distribution of the quantities from well observations is required. In this study, we developed a physics-informed neural network (PINN), which constrains predictions to satisfy conservation of mass and energy, for predicting spatial distributions of temperature, pressure, and permeability of natural-state hydrothermal systems. We assessed the characteristics of the proposed method by applying it to 2D synthetic models of geothermal systems. Our results showed that the PINN outperformed the conventional neural network in terms of prediction accuracy. Among the PINN-predicted quantities, the errors in the predicted temperatures in the unexplored regions were significantly reduced. Furthermore, we confirmed that the predictions decreased the loss of the conservation laws. Thus, our PINN approach guarantees physical plausibility, which has been impossible using existing machine learning approaches. As permeability investigations in geothermal wells are often limited, we also demonstrate that the resistivity model obtained using the magnetotelluric method is effective in supplementing permeability observations and improving its prediction accuracy. This study demonstrated for the first time the usefulness of the PINN to a geothermal energy problem.
•A physics-informed neural net (PINN) for geothermics is proposed for the first time.•PINN predicts temperatures, pressures, and permeabilities in hydrothermal systems.•PINN outperformed conventional neural networks in terms of prediction accuracy.•PINN enhances physical validity of the predictions by considering conservation laws.•PINN is useful for geothermal inverse modeling by combining data and physics laws.
In geothermal developments, characterizing hydrothermal flow is essential for predicting future production and designing effective development strategies. Numerical simulation models require ...determining a large number of input parameters to represent a reservoir. Most previous methods have estimated plausible parameters through a trial-and-error search with measurement data, which is time-consuming and dependent on the subjectivity of the analyst. In this study, we propose a machine-learning-based method to estimate input parameters (i.e., permeability distributions, heat source, and sink conditions) for geothermal reservoir modeling. A large amount of training data was prepared by a geothermal reservoir simulator capable of calculating pressure and temperature distributions in the natural state. Machine learning algorithms were applied to classification and regression problems, and Gradient Boosting Machine was ultimately selected. The performance evaluation scores were high for estimating parameters even for the 3D problem. The machine learning model was applied to three-dimensional data simulating a geothermal field. Although the training data had a different distribution, the estimated distribution of permeability showed the same trend as the true distribution. Recalculation using the estimated input parameters resulted in temperature distributions with good accuracy. This study successfully demonstrates the possibility of applying the model to field data.
Display omitted
•Machine learning model predicts input parameters for geothermal reservoir simulation.•Permeability and heat source and sink conditions are well estimated.•Applicability of learning model is verified by data simulating a geothermal field.•Natural-state temperature was recalculated accurately by estimated input parameters.
Os rios são importantes modeladores do relevo, pois definem novas feições geomorfológicas como as planícies e terraços fluviais. Na porção leste do Estado do Ceará, no baixo curso do rio Jaguaribe, ...precisamente na sua margem esquerda, uma área de cobertura sedimentar cenozoica sob a forma de terraços foi preferencialmente ocupada devido as condições naturais mais favoráveis. Em contrapartida, estes sistemas ambientais apresentam um quadro com severas erosões potencializadas pela pressão antrópica exercida sobre eles. Desse modo, a pesquisa se justifica pelo ineditismo, pois pouco se investigou sobre as causas e consequências da erosão acelerada na área de estudo, bem como por ela contemplar o semiárido brasileiro, região densamente ocupada. Além disso, o estudo objetiva analisar como o uso intenso e desordenado da terra nos terraços do Baixo Jaguaribe aceleraram os processos erosivos, correlacionando-os como vetores do problema. Os procedimentos metodológicos envolveram consulta bibliográfica, elaboração de material cartográfico e pesquisa de campo. Os resultados mostraram que as atividades antrópicas provocaram fortes mudanças nas paisagens em 36 anos, cujos usos desconsideraram as potencialidades e limitações dos sistemas ambientais, suprimindo a vegetação (-22,77%), aumentando a carcinicultura (334,61%), a urbanização (738,79%) e expansão de lavoura perenes (304,61%) e temporárias (1.262,94%). Assim, diante do cenário de degradação e de lacunas ainda existentes são necessárias novas pesquisas. Por fim, espera-se que este trabalho contribua para balizar estudos que visem o planejamento ambiental da área em questão.
Yerköy geothermal field which is located in the city of Yozgat in Central Anatolian Region, is one of the low temperature fields that is investigated for geothermal heating. In 2006, two ...exploration/production wells were drilled at 550 m and 750 m below the surface. The maximum bottom-hole temperatures ranged from 67 °C to 72 °C, respectively in these wells. These two exploration wells are provided the new information about the geothermal field and the reservoir of Yerköy. In this study, these well-data were evaluated for the first time together with the surface data obtained from the field. The aim of this study is to develop a conceptual model of the Yerköy geothermal system using these well data and geological, hydrogeological and hydrogeochemical studies and to simulate this model to define the three-dimensionally subsurface conditions. Numerical simulation was performed by using TOUGH2 Software. The calculated model results were compared with the measured static pressure and temperature in the wells for the calibration. According to the calibrated model, the permeability values of fractured rocks and fault are between 9.0 × 10−14 and 1.0 × 10−12 m2. The geothermal fluid with a mass flow of 18 kg/s with the enthalpy of 415 kJ/kg at the base boundary rises along the fault zones and transported into the fracture systems, which we define it as the reservoir in the immediate vicinity of the faults.
•A first field-scale 3D natural state model was developed in the low temperature, fault controlled Yerköy geothermal field.•Geology, hydrogeology and hydrohemistry data were evaluated with the well test data in the field for the conceptual model.•Initial T/P, heat and flow boundary conditions, permeability of the units and fault zone were defined for the model.•The simulation model was created by TOUGH2 code in order to characterize the geothermal reservoir.
Accurate differentiation between stereopsis assessments in the natural and dichoptic presentation states has proven challenging with commercial stereopsis measurement tools. This study proposes a ...novel method to delineate these differences more precisely.
We instituted two stereopsis test systems predicated on a pair of 4K smartphones and a modified Frisby Near Stereotest (FNS) version. Stereoacuity was evaluated both in the natural environment state (
the modified FNS) and the dichoptic state (
smartphones). Thirty subjects aged 20 to 28 years participated in the study with the best-corrected visual acuity (VA) of each eye no less than 0 logMAR and stereoauity of no worse than 40″. Varying degrees of monocular VA loss were induced using the fogging method, while this study does not explore conditions where the VA of both eyes is worse than 0 logMAR.
When the VA difference between the two eyes did not exceed 0.2 logMAR, the modified FNS produced lower stereoacuity values compared to the 4K smartphones (Wilcoxon signed-rank test: difference = 0 logMAR,
= -3.879,
< 0.001; difference = 0.1 logMAR,
= -3.478,
= 0.001; difference = 0.2 logMAR,
= -3.977,
< 0.001). Conversely, no significant differences were observed when the binocular vision difference exceeded 0.2 logMAR (difference = 0.3 logMAR,
= -1.880,
= 0.060; difference = 0.4 logMAR,
= -1.784,
= 0.074; difference = 0.5 logMAR,
= -1.812,
= 0.070).
The findings suggest that stereoacuity values measurements taken in the natural environment state surpass those derived from the dichoptic presentation. However, the observed difference diminishes as stereopsis decreases, corresponding to an increase in induced anisometropia.
Understanding the impacts of habitat modification on primate feeding ecology is essential to design effective conservation management strategies. The dietary guild (e.g., frugivore, folivore, ...insectivore, and omnivore) of primates and their degree of ecological flexibility impacts their ability to cope with human-modified habitats. The Omo River guereza (
Colobus guereza guereza
) is a subspecies of eastern black-and-white colobus monkey endemic to the western Rift Valley forests of Ethiopia, where it faces increasing anthropogenic change. While there is some understanding of how this subspecies copes with anthropogenic pressures, we compared the feeding ecology of Omo River guerezas in natural and human-modified habitats. Specifically, we collected data on two neighbouring guereza groups that inhabit adjacent plantation and natural forest habitats over 12 months in Wof-Washa Natural State Forest in the central highlands of Ethiopia. Furthermore, we conducted vegetation surveys on the botanical composition and vertical structure of both habitat types. The monthly food availability index of young leaves was higher in the natural forest than in plantation forest habitat. We observed guerezas feeding on 30 plant species in the natural forest but only 18 species in the plantation forest. Guerezas in both forest types consumed mostly young leaves, but the natural forest group relied more on mature leaves and shoots, and less on fruits and stems, than the plantation forest group.
Maesa lanceolata
leaves contributed a greater proportion of the overall diet for the plantation forest group, whereas
Vernonia leopoldi
accounted for the largest proportion of the guereza diet for the natural forest group. The top five species consumed comprised 83% of the diet in the plantation forest group and 70% in the natural forest group, indicating that relatively few plant species dominate guereza diets in these habitats. Conservation of both natural and plantation forests, especially the plant species most intensively exploited by guerezas, should be prioritized to assist in Omo River guereza conservation efforts.