Large water tanks are used as thermal energy storage components in district heating systems to store sensible heat produced by intermittent energy sources and to decouple the production of thermal energy from its demand. Good thermal stratification is crucial for energy storage efficiency, thus flow maldistribution and mixing of water layers at different temperatures should be minimized. This paper proposes an innovative internal flow distribution configuration for a large-size thermal energy storage, and develops new simplified analytical models for the choice of its design parameters. In the novel configuration, water is injected into (and collected from) the cap volumes of the tank by flowing radially inward (outward) through several small orifices of a peripheral toroidal manifold. Two horizontal perforated plates cover the full cross sections downstream of the manifolds and rectify the vertical flow, thus reducing mixing. Uniform perforation pitch was analytically demonstrated to be the most reasonable solution both for the toroidal distributors and for the rectifying plates. A 1D model was developed to predict the time evolution of the vertical temperature distribution in the tank. The turbulence-related parameters that could not be inferred from the existing fluid-mechanics literature were initially estimated with CFD simulations. The results of CFD-calibrated model were then compared to experimental data obtained from a full-scale large water-tank facility recently built in Brescia according to the proposed design. After a re-calibration of the exponent defining the decay of homogeneous turbulence downstream of the perforated plates, good agreement was found between measured and predicted vertical temperatures. With the novel inlet design, a thermocline of about 0.5 m is established immediately downstream of the perforated plate, and remains practically constant along time. The model is important to minimize and control the thermocline thickness so as to maximize the recoverable thermal energy, not only at the tank design stage but also to identify optimal loading and unloading protocols.
Read full abstract