Large Scale Heat Storage for Solar District Heating

In the context of the global energy crisis and climate change, solar district heating systems are a crucial technology for addressing these challenges. Large-scale Pit Thermal Energy Storage (PTES) systems offer significant advantages, including high thermal capacity, low cost per cubic unit, and long lifespan. Integrating PTES with solar heating systems can significantly enhance solar heating efficiency and mitigate the limitations of renewable energy related to time and climate variability. However, there are currently only a few operating PTES systems (approximately 19), and the lack of suitable simulation and analysis methods hinders large-scale application. The research questions focus on how to generalize practical experience, improve simulation accuracy, and enhance the economic viability and performance of these systems. Therefore, this thesis aims to address these issues through data analysis, model validation, system integration, control optimization, and multienergy complementarity.

This thesis proposes a robust framework for accelerating sustainability, using a solar district heating system in Dronninglund, Denmark, as an example. The framework includes a bi-directional long short-term memory method for data correction and a balanced method for energy and exergy analyses. The solar collector efficiency is 41%, the storage efficiency is 89%, and the overall system COP is 2.9 with 77% renewable energy. The economic analysis predicts a rise in net present value from -5.5 million euros over 10 years to 15.2 million euros over 40 years, while achieving a carbon reduction of 122 kg/MWh.

An improved TRNSYS semi-analytical PTES model (Type 1535-1301) is introduced, validated with 3 years of data, showing charge/discharged energy prediction differences with measurement of less than 3% and internal average temperature differences below 3 K. The model's annual heat transfer computation time for a single grid is reduced to 57 ms, and the model is embedded into TRNSYS for system prediction. Furthermore, this thesis investigates the impact of design parameters and soil properties on the thermal performance of PTES. The results indicate that increasing the slope angle reduces heat loss and improves thermal stratification. Studies on different soil types reveal that an increase in the soil
Fourier number significantly increases heat loss. Moreover, positioning the top diffuser higher and the bottom diffuser lower leads to the highest level of thermal stratification.

The accuracy of PTES models is crucial for optimal water pit design and system efficiency. This thesis examines 6 PTES models (Type342, 343, 1531, 1534, 1535, 1536) in TRNSYS, assessing their strengths and weaknesses. Numerical models Type342, 343) offer better heat loss accuracy compared to semi-analytical models. Type343, 1535, and 1536 models provide better charged energy estimates, with Type343 performing best overall. However, Type343 becomes unstable if the mesh Fourier number exceeds 0.05. PTES models significantly influence heat pump performance but have minimal impact on the thermal performance of solar collectors. A high-precision TRNSYS model for a large-scale solar heating plant is developed and validated using data from Dronninglund, Denmark. The model integrates a solar collector field, pit heat storage, a heat pump, boilers, and an intelligent control system. The validation shows excellent agreement with the measurement, with only a difference of 0.3% in solar production and a difference of 1K in storage temperature. The study proposes a new full-time heat pump control strategy, improving heat pump efficiency by 20%, collector efficiency by 10%, reducing heat costs by up to 27% and carbon emissions by 32 kg/MWh.

Finally, a model predictive control strategy is proposed to convert excess electricity into thermal energy and store it in the large heat storage. This approach reduces electricity demand, charging frequency, and operational costs. Storing zero-cost electricity cuts heat costs by up to 80%, proving both economically feasible and operationally effective. The thesis provides valuable insights for optimizing solar heating systems and enhancing renewable energy utilization.