California Oil, Gas, and Groundwater Program

Publication: Conference and Public Meeting Abstracts or Presentations

Authors:

Year Published:

Publication Information:

URL:

DOI:

Shimabukuro, D. H., Stephens, M. J., Gillespie, J. M., Chang, W., Finney, D. M. N. and Haugen, E.

2018

2018 Pacific Section of the American Association of Petroleum Geologists Annual Meeting, Bakersfield, California, April 22-25, 2018

http://www.psaapgconvention.org/conventions/2018/abstracts.html

Abstract

Several different methods using a combination of produced water geochemistry and borehole geophysics have been used to understand deep groundwater total dissolved solids (TDS) in and around oil and gas fields in the San Joaquin Valley.

These techniques can be extrapolated into three dimensions using geostatistical methods, such as kriging, that allow spatial heterogeneity to be better represented.Such approaches include kriging Several different methods using a combination of produced water geochemistry and borehole geophysics have been used to understand deep groundwater total dissolved solids (TDS) in and around oil and gas fields in the San Joaquin Valley.

Produced water geochemical measurements alone can yield a TDS-depth relationship. These measurements can be supplemented with borehole geophysics, such as using observed deep resistivity- TDS relationships, or using deep resistivity, porosity, and temperature to calculate TDS using the Archies-based resistivity-porosity method. These approaches are often one- dimensional in nature, yielding TDS as a function of depth for an entire well field.

Here, we discuss the advantages and disadvantages of each method, and how the availability of archived oil and gas data, along with field characteristics, such as vertical and lateral salinity gradients, the presence of diatomite, and thermal effects due to steaming, may affect the choice and accuracy of each method.produced water geochemistry, kriging Archies equation with standard or empirically- fit parameters, or machine- learning methods such as neural nets.