Utilization of a Python Algorithm to Evaluate Changes in Y-Grid Based Reservoir Temperature and Liquid Saturation at the Coso Geothermal Field

Abstract

The D'Amore and Truesdell (1985) y-grid vapor saturation (reservoir steam fraction or y-values) and gas geothermometer temperature estimates are evaluated at the Coso Geothermal Field (Coso) from the 1980's through 2020 using a Python algorithm. Production at Coso has evolved from predominantly two-phase (liquid [geothermal brine] + vapor [geothermal steam and gas]) fluids at saturated to moderate enthalpies at start-up to producing predominantly high enthalpy (> 1150 BTU/lb) steam in 2020. Evaluation of reservoir liquid saturation and reservoir temperature used gas analysis of steam samples collected and analyzed by Coso Operating Company. Changes over time in reservoir liquid saturation and temperature inform reservoir sustainability and the influence of different reservoir processes, such as boiling and injection influence. Although changes in y-values since start-up indicate the central Coso y-grid based liquid saturation has decreased over time, significant liquid reserves remain in 2020. The y-grid approach estimates the reservoir vapor (steam) fraction as it relates to reservoir liquid and reservoir temperature, as gas partitioning is temperature dependent between phases. The y-grid approach is considered more reliable than single gas-reaction geothermometers because it incorporates multiple gas-gas reactions and gas partitioning. Traditionally temperature and steam saturation are read manually from the y-grid contours by an analyst, which is time consuming and error prone. Geologica has developed a Python algorithm that automates this process and enables accurate interpretation of hundreds of geochemical samples in seconds. Python was also used to generate advanced visualizations that enable qualitative interpretation of changes in reservoir liquid saturation and gas geothermometer temperatures over time.