If you have any further questions regarding our services or see potential for cooperation, please contact us by email or phone number given below:
© heatbeat engineering GmbH & heatbeat nrw GmbH 2017-
In the 67th issue of our heatbeat Research Newsletter, we present two new studies from the district heating research community. The first challenges long-established design assumptions for domestic hot water systems, proposing reduced flow rates that better reflect real usage. The second introduces a stochastic investment optimization framework for long-term capacity planning, directly relevant to operators navigating the uncertain transition to fourth-generation district heating. In addition, this newsletter provides a brief overview of the most important developments and news regarding heatbeat and our Digital Twin.
On April 22, during our Feature Update Live Webinar, we once again provided an overview of the most important developments in the heatbeat Digital Twin over the past three months. Among the development highlights this time were several improvements to the mapping of substations and secondary networks. These can now be drawn directly into the Digital Twin’s web interface in edit mode, and during sizing and system simulation, the entire system, including multiple secondary networks, can now be evaluated holistically. You can also find a brief overview of this in this video . In addition, we were able to showcase numerous improvements in the Design Cockpit, enhanced interactivity in the building dashboards, and several new map layers for even greater information density. There have also been advancements in IT security, allowing for more granular configuration of policies regarding the use of two-factor authentication across the organization.
And since development doesn’t stop after the webinar, we’re already looking ahead to the new updates coming in April: Among other things, we’ve made our network calculations even more robust, enabling us to calculate pressure conditions more reliably, even with very large numbers of secondary networks. In addition, we have integrated a more detailed display of measurement locations, heat meters, and their data connections. This allows for an even more detailed view of the heat meters in the network, ranging from their spatial allocation to an overview of calibration deadlines.
In addition to these developments, we were delighted to participate in various exciting events in April. These included the District Heating Conference in Kassel, where we had the opportunity to engage in interesting discussions and, in our presentation “Always Better: Continuous Operational Optimization with AI in the Digital Twin,” provide insights into our combination of physically reliable models and AI methods for forecasting and fault detection. In addition, as part of the online events for the Berliner Energietage, we were able to contribute a presentation on our research project KNW-Opt II, focusing on the application of our Digital Twin to the cold local heating networks in Soest and Bad Nauheim ( link to the video ).
As the share of domestic hot water (DHW) in total building heat demand grows (reaching up to 50 % in highly energy-efficient buildings), the accuracy of design flow rates becomes increasingly critical. A preprint by Averfalk, Eklund, and Werner from Halmstad University directly challenges the current Swedish design standard (F101) and, by extension, comparable European standards, arguing that they systematically overestimate real peak hot water demand.
The study is based on 21.5 million one-minute measurements from 648 apartments across eight Swedish multi-family buildings, collected over 22–24 days in winter and spring. The key finding is clear: current F101 design flow rates substantially exceed actual demand, especially for larger buildings. For buildings with more than 30 apartments, the existing standard overestimates the 99.5th percentile by approximately 30%.
Oversized heat exchangers and control valves can cause control instability, reduce valve authority, and impair system performance. As district heating moves toward lower operating temperatures, correctly sized components become even more important. The authors note that the observed overestimation is likely linked to the widespread adoption of water-saving taps and individual metering since the F101 standard was last substantively revised in the early 2000s.
The authors therefore propose a new design formula based on the 99.5th percentile threshold. Compared to the current F101 values, this represents reductions ranging from about 4 % for two-apartment buildings to around 30 % for larger buildings.
Analysis of the exceedance events shows that approximately 70 % of all flow peaks above this level last no more than one minute, and the great majority have a separation time of more than 40 minutes. With typical supply temperatures of 50–60 °C providing a buffer of 10–20 K above the required delivery temperature, these brief and infrequent deficits are unlikely to be noticed by residents.
The authors also highlight an interesting paradox: Sweden, which applies lower design flow rates than Germany or Denmark, actually has higher measured hot water usage per floor area than both countries. This suggests that design standards and actual consumption patterns have evolved independently across countries, reinforcing the case for evidence-based revision of national and European norms alike.
The second study, published in Energy Conversion and Management, tackles one of the most consequential challenges facing district heating operators today: how to make solid infrastructure investments when future energy prices and network temperature trajectories remain deeply uncertain. The authors develop a multistage stochastic optimization framework using, applied to Stockholm's district heating network over a 2025–2050 planning horizon.
The core methodological contribution is the departure from deterministic, perfect-foresight planning (current industry standard) toward a sequential decision structure in which investments are made adaptively as uncertainties gradually resolve. The framework combines two distinct uncertainty dimensions: structural temperature pathway uncertainty (whether the network transitions to low-temperature 4GDH or remains high-temperature) and Markovian energy price dynamics governing electricity and biomass fuel costs. Vintage-based capacity tracking explicitly models the retirement of legacy assets and the commissioning of new capacity with construction lead times to capture the inertia of brownfield infrastructure that deterministic models typically simplify away.
Applying the framework to one of Europe's largest DH networks in Stockholms yields 27 %, savings (2.8 billion €) under risk-neutral evaluation relative to the deterministic baseline. Under tail-risk measures this rises to 39 %, or 5.6 billion €.
The deterministic approach under-invests in foundational, no-regret capacity at the first decision stage, producing a 330 MW capacity deficit that becomes extremely costly if the High-Temperature pathway materializes. The stochastic policy, by contrast, begins with an investment in robust technologies (Data Center Heat Pump, Wastewater Heat Pump) and then adapts aggressively in subsequent stages once the temperature trajectory is revealed. This creates a natural decision hierarchy, where no-regret options come first, while conditional ones come later.
A particularly insightful finding concerns biomass. While the median simulation under the Low-Temperature pathway shows Wood Chip CHP phasing out entirely by 2050, the model retains a small biomass capacity in a subset of scenarios not as a baseload source but as an option hedge against electricity price spikes. This nuanced role for legacy technology would be invisible in a deterministic model.
The study also identifies a smart use of the existing Bio Oil Boiler fleet (~780 MW in 2035): rather than building expensive replacement capacity prematurely, the stochastic policy deliberately exploits these legacy assets as a capacity bridge, delaying capital-intensive commitments until long-term trajectories are clearer. This 'brownfield flexibility' strategy reduces both investment cost and stranded asset risk.
For district heating planners, the 27–39 % savings quantifies what is lost by ignoring uncertainty. As networks transition to electrified heat supply becoming increasingly sensitive to electricity market volatility, the ability to make sequential, adaptive decisions becomes not just beneficial but essential.
As always, we recommend reading the full articles. The findings on domestic hot water design are highly relevant to the sizing and optimization of heat exchanger stations — an area where our heatbeat Digital Twin already supports utilities in right-sizing components and identifying substation performance anomalies. The investment optimization study resonates closely with the strategic planning challenges we address together with our clients: how to sequence investments in new heat sources and network expansions under real-world uncertainty. Our engineering team, in combination with the heatbeat Digital Twin, can support you in developing robust, evidence-based investment roadmaps for your district heating network.
In addition, we’d like to invite you to our next quarterly Feature Update Live Webinar on July 15 at 2:00 p.m . And if you’re attending the Berliner Energietage in person today, we’d be delighted if you’d also join us for our presentation on “Grid-Connected Solutions for Multi-Family Homes.” You can also meet us on May 7 at the polis Convention in Düsseldorf, where we’ll be sharing insights about our project with the City of Aachen on pilot areas for heating networks as part of the Municipal Energy Transition Practice Forum.
The next issue of our newsletter will be published on June 3, 2026.
Your heatbeat team