Division of Exploration Technology / Research and investigation committee on resource exploration from land to ocean to develop fundamental and big data processing techniques

[3101-06-02] Temperature estimation at the Kakkonda geothermal field based on resistivity data using artificial neural network

○Kazuya Ishitsuka¹, Toru Mogi¹, Kotaro Sugano², Toshihiro Uchida³, Tatsuya Kajiwara⁴ (1. Hokkaido University, 2. Hokkaido Electric Power Co., Inc., 3. National Institute of Advanced Industrial Science and Technology, 4. Geothermal Engineering Co., Ltd.)

Chairman: Akihisa Kizaki (Akita university)

Keywords:Temperature estimation, Neural network, Geothermal development, Resistivity, the Kakkonda geothermal field

It is well known that temperature estimation beneath a geothermal development area is important. Recently, Spichak et al. have proposed a method to estimate temperature distribution based on resistivity data estimated by MT measurements. This method constructs a neural network model that connects resistivity and temperature data at well locations, and estimates temperature distribution at other areas based on resistivity data. We examined the applicability of this method through the application to the Kakkonda geothermal field. We estimated the temperature at a deeper part of the geothermal field with the accuracy of 38 % by tuning a neural network model and training data. This results demonstrate the effectiveness of the method.