newsletter

Thermal Storages in District Heating Networks


Dear reader,

the last newsletter included a cross-analysis of many publications related to district heating and cooling networks, this time we would like to highlight one topic in more detail. For this purpose, we present a paper that deals with different possibilities for integrating thermal storage in district heating networks. The article examines both centralized and decentralized storage and compares both options with respect to the total investment for the district heating network. Based on a case study in Belgium, different combinations and their sensitivities are investigated.

The short version

This issue of our newsletter presents the article Influence of centralized and distributed thermal energy storage on district heating network design by Joseph Maria Jebamalai et. al. from the Department of Architecture and Urban Planning at the University of Ghent (Belgium).

The article deals with different storage options in a district heating network and the influence on the total investment of the network (feed-in, network, heat transfer and investment for the storage). The authors use a simulation tool for the comparison and select about 2400 buildings of the city Kortrijk in Belgium for their case study. In a scenario analysis, they vary, among other things, the storage location (central storage, distributed storage in the network and storage in buildings), the storage size (daily, multi-day and seasonal storage) and the investmens for the feed-in.

The results show that storage tanks in all three possible locations reduce the investment in a district heating network. Reductions in investment are achieved both through savings in the peak load and through pipe dimensioning in the case of decentralised storage. The savings through the installation of storage tanks in buildings are the largest; here, the optimal (cumulative) storage tank size is in the range of daily buffer storages. For distributed storage, the study shows that many, smaller storage in the grid have advantages over one larger storage. Also daily storages show to be particularly advantageous. Central storages close to the feed-in are designed larger in the study to achieve lower investments. However, they show the lowest savings potential overall, as the network dimensioning in this case does not show any savings potential.

Our full summary

The article Influence of centralized and distributed thermal energy storage on district heating network design by Joseph Maria Jebamalai et. al. from the Department of Architecture and Urban Planning at the University of Ghent (Belgium) examines different options for storages in district heating networks. The aim of integrating storage is on the one hand to avoid peak loads, and on the other hand to relieve the maximum power to be transported by decentralised storage and thus to optimise the pipe dimensioning. The authors highlight the importance of storage systems against the background of the integration of renewable energies with fluctuating feed-ins.

The article presents a case study of a theoretical district heating network in Kortrijk (Belgium). From a total of 35000 buildings in Kortrijk, about 2400 buildings were selected using a line density threshold of more than 1 MWh/year/m. The buildings are connected by means of a fictitious heat network. Based on the measured gas demand for streets and GIS data (geographic information system), the buildings are divided into residential, commercial and industrial buildings and daily load profiles are selected. Individual demand characteristics of buildings are not considered in this study. The heat source was assumed to be approx. 2 km from the district heating network. In this case study, the feed-in is constantly available throughout the year. This means that in summer there is a surplus of heat from generation that can theoretically be used in storage.

With the help of the defined case study, different scenarios are examined, with a total of 8 network parameters being varied (we only consider a selection here):

  • storage location: central storage close to the feed-in, distributed storage in the thermal grid as well as decentralised storage in the buildings.
  • storage size: daily, multi-day and seasonal storage
  • specific investment of heat generation

The different scenarios are evaluated using the total investment for the thermal network. These investments include the heat source, the network, the heat transfer and the storage. Installation costs are also added, operating costs (e.g. fuel or maintenance costs) are not taken into account.

The maximum storage capacity of the central storage corresponds to a large seasonal storage, which can theoretically reduce the peak power of the heat source from 34 MW to approx. 11 MW. Based on the investments, only storages with a significantly lower capacity are proposed. The minimum investment is for a storage capacity of about 2 days. The savings are realised by a lower peak load of the feed-in (approx. 24 MW instead of 34 MW).

In addition to a central storage facility, the case study also looks at 7, 10 and 18 distributed storage facilities in the network. In addition to the reduction of the peak load of the feed-in, pipes can be dimensioned smaller, so that the absolute savings of the investment are larger. As a result, the minimum investment is already achieved from a cumulative storage capacity of 1 day. It can be seen that the optimal number of distributed storages for the grid in Kortrijk is between 7 and 10.

The greatest savings in terms of investment can be achieved with storage in buildings. Comparable to the distributed storage systems, a cumulative storage size of 1 day is also the most favourable option for these storage systems. Due to the different variance of the assumed load profiles of the three building types, residential buildings prove to be clearly more suitable for the installation of storage. The ratio of the decrease between day and night was assumed to be larger for these buildings than for the other building types, so that larger amounts of energy can be stored in equally large storage units during the night, thus reducing the necessary peak load in the morning hours. Of course, this depends on individual demand profiles; for example, office buildings are well suited for the integration of storages.

The investment in the heat source is a major driver of the overall investment, along with the investment in the network. With the use of storages, peak power can be reduced and thus generation capacities can be saved. This is confirmed by the study of Jebamalai et. al. The minimal investment of the thermal grid shifts to larger storages if high costs for the heat source are assumed.

Jebamalai et. al. show that storages can reduce investments in thermal networks. In our projects at heatbeat, we consider both centralised and decentralised storage in dynamic operation and can observe savings not only in investments but also for operating costs. The article shows how important it is to consider storage in the planning and operation of district heating networks. In our opinion, this becomes even more important with a higher share of renewable energies in new and existing grids. The optimal storage size depends on many factors. Moreover, it is difficult to calculate storage systems with static assumptions; rather, dynamic values and control strategies must be taken into account in the early design stages. Our dynamic simulation tools can achieve this and take into account the interaction between storage, generation, distribution and buildings.

Further information

The article by Jebamalai et. al. is available at https://doi.org/10.1016/j.energy.2020.117689. In addition to the influences presented here, it also includes an investigation of temperature spread, the relationship between daily off-peak and peak load cases, and variable annual demands.

The next issue of our newsletter will be released on June 2, 2021.

Best regards,
Your heatbeat team