Control Strategies for Peak Shaving in District Heating Networks

Dear reader,

This is the fifth issue of our heatbeat Research Newsletter. This time we would like to focus on existing district heating networks and how modern simulation tools can optimize control strategies to increase energy flexibility. The selected paper presents a case study for a district heating network in Verona (Italy). The paper investigates a new control strategy for peak shaving and load shifting by adjusting flow rates in the network prior or after peak demands.

The short version

In this issue, we present our summary of the paper "Increasing the energy flexibility of existing district heating networks through flow rate variations" by Jacopo Vivian et al. from the Department of Industrial Engineering of the University of Padova in Italy.

The paper presents a simulation case study in the city of Verona (Italy). The authors apply a simulation model to investigate new control strategies in comparison to the current operation to shift loads and lower peak demands. The current control strategy is a differential pressure control at constant flow temperature, which guarantees a sufficient pressure difference at the worst point. To shift loads and avoid peak loads, for example to facilitate the integration of renewable energy, two different control strategies are modelled and simulated for the district heating network in Verona. The first strategy involves a pre-charge of the network, where the network is operated with higher flow rates before peak loads. The second strategy is a post-charge control where the network is more cooled down during peak loads and heated afterwards. The authors conclude that in terms of maximum load reduction potential, shifted energy and supply reliability, the pre-charge strategy is beneficial.

Our full summary

The paper "Increasing the energy flexibility of existing district heating networks through flow rate variations" by Jacopo Vivian et al. from the Department of Industrial Engineering of the University of Padova in Italy investigates new control strategies for existing district heating networks to decrease peak loads (peak shaving) and to shift loads for short time periods. For both peak shaving and load shifting the thermal capacitance of the water volume inside the piping system is used, thus no additional storages need to be installed. The authors state that the ability of load shifting is one key element to integrate renewable energies into existing district heating systems.

The paper presents a simulation case study of a real district heating network in Verona (Italy). The case study involves a district heating network with 25 km length and 247 connected customers. The total energy of the customers is about 70 GWh/year for space heating and domestic hot water. The peak load of the energy supply is approx. 38 MW. There are three power supply units in the network. The main supply holds about 95 % of the total installed power. Two smaller supply units involve the use of waste heat from a foundry and auxiliary gas boilers. In the current system the mass flow rate is determined by valves in the customers substations and the head of the circulating pump in the main supply unit. The head for each flow rate is set with empirically determined values, ensuring a sufficient pressure difference at the worst point in the network at every time. The flow temperature in the system is constant at 80 °C at all flow rates.

The authors present two different strategies for peak shaving and load shifting using only the capacity of the thermal network without any conventional storages. Thus, the storage capacity is not separated from the supply and return line of the network and the fluid inside the district heating network takes over two function at the same time, to transfer the heat to the customers and to store heat for later use. In the case study considered the flow temperature is constant over time, thus the only option to store energy is to increase the volume flow and thus increase the temperature of the return pipe. In the current setup, the flow rate cannot be controlled directly by the operator. Therefore the authors propose several bypasses in the network as a potential solution. Based on these preliminary considerations, two strategies are examined:

  • Pre-charge of the network: the mass flow rate increases a certain time before a peak is expected in the heat supply (e. g. Monday morning). To increase the mass flow the above-mentioned bypasses in the network are opened. The pre-charge option increases the temperature of the return pipe and stores energy in the thermal capacitance of the network.

  • Post-charge: the mass flow rate is kept below the actual necessary peak flow and remains higher after the peak load to regenerate the return flow. The reduction in mass flow is realized with a reduced head of the pumps.

The mass flow rates in both the pre-charge and post-charge have different shapes. The authors define the shapes with the duration and the power of the pre- or post-charge. To evaluate the success and suitability of the two control strategies different Key-Performance-Indicators (KPIs) are introduced. The KPIs are the power reduction and the time shift of the peak load compared to the reference control strategy, the amount of thermal energy shifted, a discomfort index based on the return temperature of the customers substations (i. e. supply safety) and the variation of supplied energy both for thermal and electric (pumps) energy at the supply stations.

The paper runs an extensive simulation study with different shapes for both control strategies and evaluates the defined KPIs and simulation results in detail. In addition to the shape of the pre- or post-charge, one main factor influencing the potential of peak shaving and load shifting in thermal networks is the flow velocity in the pipes. The higher the average velocity in the network the faster the propagation of the heat in the network and the faster the return line heats up and stores energy. This result shows the great importance of high temporal and spatial resolution simulation models for the development and evaluation of new control strategies for thermal networks.

The authors conclude that the pre-charge control strategy is more beneficial in all defined KPIs. The pre-charge strategy can reduce the peak load in the network up to 13 % for typical winter days and even 18 % for days with lower heat demand in spring. The chance of discomfort at the substations (i. e. the supply is not guaranteed) is about three times higher during times with high heat demands. The post-charge strategy performs worse in terms of thermal load shifting and peak shaving. The control strategies have minor influence on both used thermal and electric energy.

The paper presents a very interesting and up-to-date case study of new control strategies to enable flexibility of existing networks. Control strategies need to be tailored and evaluated for individual networks carefully. The results show the importance to consider, temperature, flow rates, flow velocity, pressure, and the energy demand of single customers in one combined simulation with high temporal and spatial resolution. At heatbeat, we use our advanced simulations methods to compare and evaluate different control strategies and facilitate the integration of renewable heat into existing district heating networks. As presented in our showcase we consider each customer with individual building models and calibrate them to their yearly consumption.

Further information

The original article can be found at Within the scope of our newsletter, we mainly focused on the control strategies and the presented case study. Yet, the paper includes also modelling approaches for district heating and proposes structural changes to the network to enable the flexibility of the network. Therefore, we highly recommend the paper in its entirety.

The next issue of our newsletter will be released on April 7, 2021.

Best regards,
Your heatbeat team