4:15 PM - 4:30 PM
[SVC29-10] Thermal property measurements by drilling cuttings obtained from a geothermal well in the Hachimantai area and thermal structure analysis
Keywords:Hachimantai area, Drilling cuttings, Thermal conductivity, Transient plane heat source method, Temperature profile
It is important to identify the factors that influence the temperature depth profile of a well in order to investigate the presence of heat convection. Especially, the thermal conductivity of rocks of a well is a vital property for quantitatively analyzing temperature logs. In general, the thermal properties of rocks can be obtained by measuring core samples. However, rock cores are often not retrieved and only drilling cuttings are collected, thus thermal conductivity measurements of drilling cuttings are required. In this study, we measured the thermal conductivities by the drilling cuttings obtained in the N19-HA-1 well in the Hachimantai area, Japan, using a transient plane heat source method. We further estimated the core-equivalent thermal conductivities from the measurements of the drilling cuttings based on a thermal conduction mixing model, and the temperature profile was quantitatively examined from the viewpoint of heat conduction.
Firstly, to validate the method of estimating the core-equivalent thermal conductivity from the cuttings, we prepared cuttings by artificially grinding andesite cores in laboratory, and compared the thermal conductivities of the cuttings and the core sample. To select the best mixing model to convert measured thermal conductivity of cuttings to core-equivalent thermal conductivity, we tested five thermal conduction mixing models (i.e., arithmetic mean, harmonic mean, geometric mean, square root mean, and effective medium theory mean). After the validation, we applied the method to 18 cuttings samples of andesite and dacitic tuff in the Hachimantai area (5 samples from depth range of 120-220 m, 4 samples from 305-385 m, and 9 samples from 925-1165 m, respectively). The estimated core-equivalent thermal conductivities were used to examine the temperature log based on Fourier's law (j=λgradT, where j is the heat flux, λ is the thermal conductivity, and gradT is the temperature gradient). This analysis of temperature profile was performed under the assumption that the heat conduction is dominant in one dimension along the well and the heat flux is constant along the analyzed depth interval. In this analysis, the average thermal conductivity of each depth zone was obtained by taking the harmonic mean of the thermal conductivities at each depth, and the heat flux was calculated from the gradient of the temperature log data.
As a result of the validation using the andesite core and cuttings, we found that difference of the thermal conductivities between the core and cuttings was minimal (1.8%) when we applied the square root mean model. In conclusion of this validation, the core-equivalent thermal conductivity could be obtained using the square root mean model. We subsequently estimated the core-equivalent thermal conductivity from the cuttings samples in the Hachimantai area by using the square root mean model and calculated the conductive temperature profile using Fourier's law. The calculated temperature profile was then compared with the observed temperature log. As the result of this analysis, we found that the heat conduction is dominant between the 120-220 m depth range. In the case of the 305-385 m depth range, the measured temperature also agrees well with the estimated conductive profile especially below the depth of 330 m, indicating that the temperature pattern is determined mainly by heat conduction. In contrast, the temperature profile of some sections between 925 m and 1165 m could not be reproduced from the present analysis. The fact that the temperature structure was not reproduced from the present analysis based on heat conduction may suggest the presence of heat convection in the depth intervals.
Firstly, to validate the method of estimating the core-equivalent thermal conductivity from the cuttings, we prepared cuttings by artificially grinding andesite cores in laboratory, and compared the thermal conductivities of the cuttings and the core sample. To select the best mixing model to convert measured thermal conductivity of cuttings to core-equivalent thermal conductivity, we tested five thermal conduction mixing models (i.e., arithmetic mean, harmonic mean, geometric mean, square root mean, and effective medium theory mean). After the validation, we applied the method to 18 cuttings samples of andesite and dacitic tuff in the Hachimantai area (5 samples from depth range of 120-220 m, 4 samples from 305-385 m, and 9 samples from 925-1165 m, respectively). The estimated core-equivalent thermal conductivities were used to examine the temperature log based on Fourier's law (j=λgradT, where j is the heat flux, λ is the thermal conductivity, and gradT is the temperature gradient). This analysis of temperature profile was performed under the assumption that the heat conduction is dominant in one dimension along the well and the heat flux is constant along the analyzed depth interval. In this analysis, the average thermal conductivity of each depth zone was obtained by taking the harmonic mean of the thermal conductivities at each depth, and the heat flux was calculated from the gradient of the temperature log data.
As a result of the validation using the andesite core and cuttings, we found that difference of the thermal conductivities between the core and cuttings was minimal (1.8%) when we applied the square root mean model. In conclusion of this validation, the core-equivalent thermal conductivity could be obtained using the square root mean model. We subsequently estimated the core-equivalent thermal conductivity from the cuttings samples in the Hachimantai area by using the square root mean model and calculated the conductive temperature profile using Fourier's law. The calculated temperature profile was then compared with the observed temperature log. As the result of this analysis, we found that the heat conduction is dominant between the 120-220 m depth range. In the case of the 305-385 m depth range, the measured temperature also agrees well with the estimated conductive profile especially below the depth of 330 m, indicating that the temperature pattern is determined mainly by heat conduction. In contrast, the temperature profile of some sections between 925 m and 1165 m could not be reproduced from the present analysis. The fact that the temperature structure was not reproduced from the present analysis based on heat conduction may suggest the presence of heat convection in the depth intervals.