Utilizing deep geothermal heat has become a key climate mitigation strategy in hydrothermal regions. The temperature of the produced fluid is critical as it determines the thermal and electrical power generated. Therefore, predicting the production temperature at a planned geothermal site is crucial and typically based on field-wide static temperature models, which predict the temperature at a certain depth in undisturbed conditions. However, undisturbed and subsequent production temperatures are not well known before drilling.
Most static models rely on low-quality bottom hole temperatures (BHTs), which are thermally disturbed by drilling processes and require correction using error-prone methods. The uncertainty of the required input parameters is rarely considered. This thesis addresses this gap by proposing uncertainties for BHT correction parameters and developing a Monte Carlo-based workflow that generates correction results dependent on probabilistic principles. The workflow was applied to geothermal and hydrocarbon wells from the North Alpine Foreland Basin in Bavaria, Germany, and tested with high-quality temperature measurements. The results show that the parameter uncertainties exceed the method errors that are not taken into account in the existing models. For 5% of the boreholes, the predicted uncertainty range is > 30 K, while for 20% of the boreholes less than 5 K uncertainty is predicted.
In conductive geothermal systems, the production temperature depends on the location of hydraulically active zones and inflowing thermal water volume. These factors are hard to predict and cannot be measured in standard geothermal well design after pump installation.
To examine the difference between a geothermal well's dynamic and static states, a 3700 m MD deep production well in the study area was analyzed for its hydraulically active zones. A conventional flowmeter interpretation was conducted, and a novel fiber optic monitoring system was installed along a sucker rod into the reservoir. This system allows for distributed temperature sensing (DTS), acoustic/dynamic strain sensing (DAS), and point pressure and temperature measurements, and providing temperature values accurate within ±1.6 K standard deviation. The system monitored temperature equilibrium during and after a 16-month shutdown, characterized inflow zones through DTS profiles during and after injection tests, and verified conventional flowmeter interpretation. Under production conditions, DTS profiles quantitatively characterized inflow zones using an inverse model. Results indicate that a 25 m thick karstified zone at the reservoir top accounts for 80% of inflow, with a mixing temperature of about 97 °C at the casing inlet. This zone's inflow is significantly cooler than deeper zones with temperatures over 107 °C. The monitoring system design was tested for dynamic movement due to thermal stress, showing that future seismic monitoring with DAS requires better understanding and filtering of the sucker rod movement.
This research demonstrates the suitability of permanent fiber optic systems for characterizing hydraulically active zones. The location of flow zones significantly influences the mixing temperature profile, often making static temperature models misleading for wellhead temperature prediction, especially if model uncertainties are unclear. A BHT correction workflow based on uncertainties can enhance existing models.
«
Utilizing deep geothermal heat has become a key climate mitigation strategy in hydrothermal regions. The temperature of the produced fluid is critical as it determines the thermal and electrical power generated. Therefore, predicting the production temperature at a planned geothermal site is crucial and typically based on field-wide static temperature models, which predict the temperature at a certain depth in undisturbed conditions. However, undisturbed and subsequent production temperatures ar...
»