Design and numerical modelling of seasonal water-based tank-in-tank thermal energy storage for district heating systems

In district heating networks, the installation of buried water-based thermal energy storage (TES) systems has seen an extension in the last decade, as they support the decarbonization of the grid and increase its flexibility. The limited space availability in many urban areas and the high land costs require compact shapes and potentially the possibility to guarantee accessibility to the TES cover (i.e., a trafficable cover). However, the presence of a rigid self-supporting cover does not allow the natural thermal expansion of the water to be followed, which can be significant in the case of large volumes (> 50 000 m3) and large temperature excursion. A solution to this problem can be the introduction of an internal tank capable of accommodating these volumes variations. The wider outer tank will act as the main storage element, while the (smaller) inner tank will be sized according to the maximum expected water volume expansion.

Assessing thermodynamic performance of this storage solution requires the implementation of a numerical model capable of representing the behavior of both the inner and outer tanks. The energy and mass flows between the two tanks will affect the thermal losses distribution and consequently the energy available for discharge.

The main challenge in the model implementation lies in the modelling of the mass variation in the external tank and both mass and volume variation in the internal one, both following the water temperature changes. This challenge is met by dividing the model implementation into several steps, each of which introduces a specific factor: thermal conduction between the outer and inner tanks through the inner tank walls, constant predefined volumetric water flow between the outer and inner tanks, water level variation in the inner tank. The Tank-in-tank TES numerical model is implemented in COMSOL Multiphysics environment, as it allows high flexibility in terms of geometry variation and the possibility to model the surrounding soil.

In the absence of existing systems to validate the implemented model, the results (thermal losses, energy, thermal stratification) of the different simulation steps are compared with a reference model representing a standard tank TES with same capacity.