Thermal energy networks with the ability to integrate diverse energy sources including renewable energy sources are becoming increasingly popular for delivering thermal energy in dense urban environments. This raises new challenges for the networks to effectively deal with the large share of the fluctuating and intermittent renewable energy inputs while meeting the dynamic heat requirements of the end-users. The overall aim of this work has been to develop a computationally efficient dynamic model of heat losses and thermal responses of buried pipelines at a wide range of timescales.
Experimental facilities that represent a scaled-down district heating (DH) system have been developed to provide temperature and heat flux data for a wide range of operating conditions representative of DH systems. This has allowed the study of both the short and long timescale dynamic behaviour and collection of data to validate the novel numerical models proposed in this research. Furthermore, a three-dimensional model of a buried pipeline with turbulent fluid flow conjugate forced convection and conduction heat transfer through the surrounding soil has been developed using the Finite Volume Method. This model has been used as a reference model for the other numerical models proposed in this work.
A novel one-dimensional model has been developed that represents the dynamic thermal behaviour of pipelines. The model combines the features of plug-flow and discrete stirred tank representations that take into account the thermal capacitance of the pipe material as well as radial heat transfer. This combination enables the proposed model to simultaneously handle the simulation of momentum and energy balance as well as simulation of the longitudinal dispersion in pipelines. The proposed model has been compared to short timescale experimental measurements. The results elucidated that the proposed model is not only able to capture the outlet temperature changes due to a step change with good agreement with experimental data but also offers advantages in reduced computational expense.
The long timescale transient behaviour of buried pipe systems has been modelled based on a response factor approach using a Dynamic Thermal Network (DTN) representation. A novel experimental procedure is presented to derive the weighting functions required in the DTN approach using the measurement data obtained from step response experiments. This has validated the modelling approach and verified the applicability of some of its assumptions.
Finally, a novel approach is proposed combining the DTN model with one-dimensional discretised heat transfer fluid flow model. This combination enables the model to represent the temperature propagation through the pipeline along with transient conduction heat transfer over a wide range of timescales. The results have shown that the model is not only able to accurately simulate the dynamic behaviour of the buried pipe system with very good agreement with the experimental data, but also noticeably more computationally efficient than the finite volume method: more than five orders of magnitudes. This makes the model widely applicable for efficient thermal network design and analysis tasks.
