Long-term Monitoring of a District-Scale Geothermal Exchange Field Using Fiber-Optic Distributed Temperature Sensing

Abstract

Increases in greenhouse gas (GHG) emissions and global temperature have occurred since the Industrial Revolution due to a constant increase in fossil energy consumption and emissions. Geothermal heat exchange (GHX) systems replace conventional heating and cooling methods and have higher system efficiency and decreased dependency on fossil fuels. Detailed, long-term monitoring of GHX fields is essential to improve the operation and sustainability of GHX systems. However, there is a need for more research on the long-term monitoring and performance analysis of low-temperature geothermal fields. To address this issue and evaluate a GHX system's efficiency, we collected geothermal temperature data from a large geothermal field in the Midwest. In this study, we investigate the temperature increase of a large-scale borehole field and its relationship to subsurface properties and we assessed errors and noise in the long-term temperature data. We have used fiber-optic (FO) distributed temperature sensing (DTS) for over seven years to monitor the performance of a 280-m by 360-m, 2596-borehole GHX field, to quantify the role of groundwater flow, and to evaluate the distributed thermophysical properties on subsurface heat transfer and storage. Like most GHX systems in the US, this field is subjected to cooling-dominated loads; likewise, the data show a gradual increase in subsurface temperature over the past seven years. However, high-frequency fluctuations in long-term temperature measurements are of concern. We worked to identify the sources of those fluctuations by analyzing potential errors and noise in the calibration formulation, the bath temperature measurements, and the raw data the interrogator gave. Better control over long-term noise in the data may be used to improve design approaches, develop more robust energy balance assessment formulation, and implement long-term ground temperature change models.