Research On Heat Transfer Model For Vertical Ground Heat Exchanger Uncertainty Design

Vertical ground coupled heat pump system takes the advantages of low carbon and energy saving, economical and highly efficient, and stable operation, using clean and renewable shallow geothermal resources as low-grade thermal sources or sinks. In this system, the ground heat exchanger(GHE) is one of the most important parts. The design of GHEs plays a crucial role in the efficient and stable operation of the system, and is the main reason that affects the system’s economic performance and application. However, conventional deterministic design methods neglect various inherent stochastic factors, such as uncertainties arising from heat transfer models, the calculation of building loads, the heat pump performance or soil thermal properties. Usually, the so-called worst case scenario or safety factors method is used to accommodate these uncertainties, which may cause oversize problems. Therefore, this thesis concentrates on the uncertainties of underlying heat transfer models and parameters in main design steps, aiming at establishing a theoretical framework for the GHE design, taking into account of multiple influential uncertainties, thus depicting the uncertainty propagation from parameter inputs to design outputs scientifically and reasonably.First, a few typical models for thermal response test(TRT) parameter estimation and GHE design have been selected through a detailed investigation of the related literatures. Their basic assumptions, initial and boundary conditions and solutions are thoroughly compared. This provides a theoretical basis for the developmentof a radial-vertical 2D analytical model under dynamic conditions(DCFLS). Simulation and experimental validation show that DCFLS is capable of describing the heat transfer mechanism between the GHE and its surrounding soil. Besides, based on this proposed model, a new design method for multiple borehole GHEs under variable heat flux has been developed utilizing spatial superposition and modified multiple load aggregation algorithm.After that, in order to investigate the influence of TRT on the GHE design, the problems within it, such as the selection of estimated parameters and iterative optimization algorithms, have been studied in depth. Parameter estimation has been conducted by different models using the trust region reflective method or simplex method. It is found that optimization objective function using the DCFLS is smaller than that using any other typical models.Meanwhile, the deviation of mean fluid temperature response under constant heat flux condition and that of GHE design caused by model mismatching have been analyzed quantificationally. The results show that model mismatching can cause ±5~15% calculation deviation of mean fluid temperature, and the difference between the DCFLS model and the authoritative DST model is the smallest when using self-consistent models. For GHE total length design, the relative deviation in semi-empirical design methods is small, but that in DCFLS design method is relatively large. And the design value from DCFLS method comes closest to that from TRNSYS design method when using self-consistent models.In consideration of the good accuracy, high compution speed and flexibilityof the proposed model, an uncertainty design framework based on Monte Carlo stochastic simulation using DCFLS has been proposed. The principles and implementary procedures are introduced, together with prime parameter uncertainty quantification method concerning main points in the design process.Finally, the proposed uncertainty design method has been demonstrated through a case study, where ground heat exchangers are used to providing energy for space heating and cooling in a university library building. Results indicate that uncertainty design method could provide all the possible design outputs and corresponding probabilities. The risk assessment based decision thus can be made using the visualized empirical cumulative probability distribution function, which is more reliable than conventional deterministic methods.The topic of this study is based on a basic theoretical issue refined from practical engineering application. An in-depth fundamental study on GHE design under multiple uncertainty sources is conducted through the combination of theoretical analysis, experimental validation and computer stochastic simulation. The research outputs help to extend and improve the conventional deterministic design methods, providing theoretical guidance and technical support for the development of this technology.