Optimization of the Thermal Load Profile in District Heating Networks through “Virtual Storage” at Building Level

Abstract

Thermal storage is of extreme importance in modern district heating networks in order to increase the share of waste heat and heat produced through renewable sources and cogeneration. Nevertheless, installation of large storage volumes is not always feasible, especially in dense urban areas. A possible option consists in virtual storage, which is obtained through variation of the thermal request profiles of some of the connected buildings with the goal of producing an effect similar to that obtained using storage. To perform such approach there are three crucial elements: 1) an advanced ICT solution able provide real time information about the thermal request of the buildings and the thermodynamic conditions at the thermal substations; 2) a detailed thermo fluid-dynamic model of the district heating network able to simulate the temperature evolution along the network as the function of time; 3) a compact model of the buildings in the district able to check the acceptability of the internal temperatures following the modified strategies.The model produces changes in the start-up time of the buildings connected with the network as well as possible pauses during the day. These changes in the request profiles usually involve a slightly larger heat load. Nevertheless, peak shaving is accompanied by a reduction in heat generation of boilers and an increase in the thermal production of efficient systems, such as cogeneration units. This results in a significant reduction in the primary energy consumption.An application to the Turin district heating network, which is the largest network in Italy, is presented. In particular, a subnetwork connecting the main transport network to about 100 buildings located in the central area of the town is considered. The analysis if performed in selected days where the optimization was conducted the day before on the basis of weather forecasts and then applied to the network. Despite the changes in the request profiles could be applied only to a limited number of buildings, the analysis show that the peak request can be reduced. Simulations performed considering the application of changes to a larger number of buildings show that reduction in the primary energy consumptions of the order of 5% can be obtained.