A comparative study for simulating heat transport in large district heating networks

Abstract

District heating networks are a convenient, economic and environmental-friendly way to supply heat to buildings connected to a central heating plant. However, the control of such a system becomes challenging if the total length of the network reaches several kilometers because the travel time of the information into the system is over hours. One solution consists in instrumenting all the parts of the network and performing a closed loop control to optimize the temperature and the mass flow rate supplied to every single consumption point. However this solution is generally expensive and difficult to implement in existing networks. What is proposed in this paper is to dynamically model the heat waves in the network to determine the temperatures and mass flow rates at key locations considering the ambient losses and the pipe thermal inertia. A study is performed to check the possibility to use the one-dimensional finite volume method to simulate heat waves propagation. First, an adiabatic pipe is considered as a reference test case to determine the limitations of this method. The results are compared to a 2D computational fluid dynamic simulation and numerical diffusion is exhibited for low spatial discretization. Therefore, an improved alternative model is developed to overcome this problem.

Modified finite volumes method for the simulation of dynamic district heating networks

“…1); ▪ heat inertia of insulation and ground are ignored. These considerations are similar to those taken in other works which report accurate results for the modeling of thermal networks [14,24,26,30,32]. Both models are further explained in the following sections.…”

“…The method, named the implicit upwind method, is compared to the characteristic line method and the authors conclude that the characteristic line is faster but the implicit upwind method provides more information on the temperature distribution within the pipe. In [30] the authors present a model based on the standard TRNSYS Type 31 component, which is based on the Lagrangian approach. The model, named the plug flow model, is compared to a 1D and a 2D Finite Volumes model.…”

Simulation Models to Size and Retrofit District Heating Systems

“…It is assumed that the fluid carrier is incompressible, which is valid if water is considered and for low pressure variations [43]. This method was previously analyzed, developed [44] and validated under Matlab for different operating conditions and pipe layouts [45], but it has now been successfully ported to Modelica [46] and the related open-source Modelica library is available in [47].…”

Optimal operation of district heating networks through demand response

“…Within this framework, full control of thermodynamic parameters assumes a key role. Indeed, the knowledge of mass flow rates flowing in the system and of the evolution of pressures and temperatures in some crucial points is essential to reduce heat losses and minimize the production cost while ensuring the comfort of end-users in buildings [3].…”