Optimization of Production and Injection of Geothermal Fields: A Machine Learning Approach

Date

2021-10-01

Author

başer, ali
KÜÇÜK, SERHAT
saraçoğlu, önder
şentürk, erdinç
Akın, Serhat

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

277
views

0
downloads

Optimizing field injection and production requires a calibrated numerical model, typically consisting of thousands of grid blocks. Optimization carried out using such a model usually takes a very long time. If the numerical model is replaced by an accurate proxy model, run-time can be significantly reduced. A numerical model has been developed using TOUGH2 to optimize the production of Kızıldere geothermal field. A proxy model developed in Python is used to optimize the production and injection of the field. The results are compared with the TOUGH2 model. The proxy model results are consistent with the field model. This approach significantly reduces time and effort.

Subject Keywords

Optimization, Tough2, Proxy, Machine Learning, Kızıldere

URI

https://pangea.stanford.edu/ERE/db/WGC/mobile/mSchedule.php
https://hdl.handle.net/11511/95985
https://pangea.stanford.edu/ERE/db/WGC/papers/WGC/2020/24034.pdf

Conference Name

World Geothermal Congress 2020+1

Collections

Department of Petroleum and Natural Gas Engineering, Conference / Seminar

Suggestions

OpenMETU
Core

Optimization of Power Conversion Efficiency in Threshold Self-Compensated UHF Rectifiers With Charge Conservation Principle Gharehbaghi, Kaveh; KOÇER, FATİH; Külah, Haluk (2017-09-01) This paper presents a compact model for threshold self-compensated rectifiers that can be used to optimize circuit parameters early in the design phase instead of time-consuming transient simulations. A design procedure is presented for finding the optimum aspect ratio of transistors used in the converter and number of rectifying stages to achieve the maximum power conversion efficiency. In the presented analysis, the relation between the power conversion efficiency and the load current over the variation o...
MODELLING OF KERNEL MACHINES BY INFINITE AND SEMI-INFINITE PROGRAMMING Ozogur-Akyuz, S.; Weber, Gerhard Wilhelm (2009-06-03) In Machine Learning (ML) algorithms, one of the crucial issues is the representation of the data. As the data become heterogeneous and large-scale, single kernel methods become insufficient to classify nonlinear data. The finite combinations of kernels are limited up to a finite choice. In order to overcome this discrepancy, we propose a novel method of "infinite" kernel combinations for learning problems with the help of infinite and semi-infinite programming regarding all elements in kernel space. Looking...
Discrete sizing optimization of steel trusses under multiple displacement constraints and load cases using guided stochastic search technique Azad, S. Kazemzadeh; Hasançebi, Oğuzhan (2015-08-01) The guided stochastic search (GSS) is a computationally efficient design optimization technique, which is originally developed for discrete sizing optimization problems of steel trusses with a single displacement constraint under a single load case. The present study aims to investigate the GSS in a more general class of truss sizing optimization problems subject to multiple displacement constraints and load cases. To this end, enhancements of the GSS are proposed in the form of two alternative approaches t...
Modeling of Multi Open Phase Fault Condition of Multi-phase Permanent Magnet Synchronous Motors Fei, Marco; Zanasi, Roberto (2011-09-10) This paper deals with the modeling of multi-phase permanent magnet synchronous motors under multi open phase fault condition. The presented model is suitable for generic number of phases, generic shape of the rotor flux and generic number of open circuit faults. The motor model in fault condition can be used for faults occurring on both adjacent and not adjacent phases. The model can be very useful both for simulation and implementation of fault-tolerant control strategies.
A Bayesian Approach to Learning Scoring Systems Ertekin Bolelli, Şeyda (2015-12-01) We present a Bayesian method for building scoring systems, which are linear models with coefficients that have very few significant digits. Usually the construction of scoring systems involve manual efforthumans invent the full scoring system without using data, or they choose how logistic regression coefficients should be scaled and rounded to produce a scoring system. These kinds of heuristics lead to suboptimal solutions. Our approach is different in that humans need only specify the prior over what the ...

Citation Formats

a. başer, S. KÜÇÜK, ö. saraçoğlu, e. şentürk, and S. Akın, “Optimization of Production and Injection of Geothermal Fields: A Machine Learning Approach,” presented at the World Geothermal Congress 2020+1, Reykjavik, İzlanda, 2021, Accessed: 00, 2022. [Online]. Available: https://pangea.stanford.edu/ERE/db/WGC/mobile/mSchedule.php.