Tracer testing in the Svartsengi geothermal field in 2015
Author: Sigrún Brá Sverrisdóttir
Year: 2016
Supervisor: Guðni Axelsson & Ágúst Valfells
The project was conducted in cooperation with ÍSOR and HS Orka.
Abstract
The Svartsengi geothermal field has been utilized for district heating since 1976 and electricity generation since 1980. Some complications due to the utilization have arisen throughout the years. Those include a rapid pressure drawdown in the field and issues regarding waste water management. The continuous injection of spent geothermal fluid began in 1998 when SV-17 was taken into use. With the addition of SV-24 as an injection well in 2008, the injection rate was slowly increased to 60% of the extraction rate. Pressure measurements showed that the
injection provided pressure support to the field, but the addition of a new energy plant and increased production from the field reintroduced the trend of declining field pressure. An important part of geothermal field management is to monitor
injection wells and study the ability to increase reinjection. In the summer of 2015 a threefold tracer test was performed in Svartsengi for that purpose. Liquid phase tracers were injected into wells SV-17 and SV-24, 2,7-napthalene disulfonate and
2,6-napthalene disuldonate, and a steam tracer, sulfurhexafluoride, was injected into SV-24. Samples have since been taken from all production wells in the field with
the addition of the monitoring well in Eldvörp. At the time of writing the 2,6-NDS has only been detected in well SV-9, while no signs have been noted of 2,7-NDS.
However, the steam tracer has been detected in all production wells and sampling for that has been terminated, but no signs were detected in the well in Eldvörp.
The tracer returns were modelled quantitatively using a couple of programs from the ICEBOX software package from ÍSOR. TRINV was used to simulate the tracer recovery based on the theory of solute transport and one dimensional flow channel models. TRCOOL was used to predict the cooling in the production wells due to the injection. Only 0.035% of the injected steam phase tracer was recovered in the production wells, indicating a very modest recharge from the injection well to the production wells. Well SV-23 had the largest tracer recovery, but SV-11
experienced the highest tracer concentration. Cooling predictions were calculated for the current injection scenario in well SV-24 and for a scenario where the current
injection rate was doubled, for 30 years. For both scenarios the model predicted an insignificant cooling in production wells over the 30 year period.