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 70th issue of our heatbeat Research Newsletter we present two current open-access articles from Energy Conversion and Management (Vol. 367, 2026). They describe two very different decarbonisation paths but share one operator lever – low network temperatures. The first quantifies how a new low-temperature, fifth-generation district heating and cooling network (5GDHC) can shift electricity demand through simple building-level measures. The second retrofits seasonal high-temperature aquifer thermal energy storage (HT-ATES) into an existing network and, in doing so, lowers its operating temperature.
On July 15, during our Feature Update Live Webinar, we took another look back at the past three months and presented the most important new developments in our heatbeat Digital Twin. We had already mentioned many of these changes in previous newsletters, such as our new weather dashboard, interactive management of measurement locations and heat meters, cost tables, change logs, and numerous improvements to the editability of building, pipe, and feed-in data for more efficient work within the Digital Twin.
Building on this, we were able to roll out additional new features to all users in our software updates as early as July. One of these new features relates to editing the road network. Our expanded data models and map views now allow us to import road surface conditions from OpenStreetMap for the entire road network within the study area. This provides a direct indicator of expected civil engineering costs or cost-effective route planning. Furthermore, for all road segments, users can interactively manage the expected years of expansion for district heating as well as other planned construction measures, such as fiber-optic expansion or sewer rehabilitation. These new features already demonstrate added value ranging from active municipal heat planning projects to newly launched projects for the Digital Twin as a planning and optimization tool for specific heating networks. Furthermore, in July we implemented additional improvements to our general database, including greater accuracy in elevation data for hydraulic calculations and simulations, as well as improved coverage of address data for the building stock.
5GDHC networks connect buildings through a low-temperature loop, with a water-to-water heat pump in each substation. Because every building is electrified, the network itself becomes an interface between thermal demand and the power grid. The authors build a physics-based dynamic model of a small network serving eight buildings at the ENEA research centre in Portici (Naples) and test four building-level demand-side strategies – indoor setpoint shifts of ±1 °C and ±2 °C, a thermal energy storage (TES) tank, and an electrical battery – across three ways of thermally balancing the loop: ground-coupled probes, groundwater, and a central PV-powered heat pump that conditions the loop to about 20 °C.
The potential differs sharply by lever, and the figures are relative to each building's reference electricity use over short, hourly-to-daily windows – not annual energy. Setpoint modulation alone, exploiting building thermal inertia, cut electricity demand by up to 37 % during downward events and raised it by up to 26 % during upward events – short-term flexibility at essentially no hardware cost. Battery integration widened the range to 51–60 %. Thermal storage performed best, shifting almost the entire reference demand within the event window (load modulation above 99 %) and up to 54 % of daily electricity. Over the year, flexible operation yielded primary energy savings of up to 11 % and operating-cost reductions of up to 14 %. The groundwater configuration was best overall – essentially because its stable, comparatively warm source keeps heat-pump COP high, a reminder that the loop-side source, not the control logic, sets the ceiling.
For operators the message is practical: even without new hardware, coordinated setpoint control gives a meaningful response, while storage – thermal in particular – unlocks the largest shifts. Two caveats matter. The study is simulation-based and sizes storage to the maximum daily demand, so the flexibility figures are an upper bound; realistically sized tanks and batteries would deliver clearly less. And it omits network hydraulics, pumping and investment costs. 5GDHC also presupposes low-temperature-ready buildings, so this is primarily a new-build or deep-retrofit route – but the direction is clear.
Where the first paper manages short-term electrical flexibility, the second tackles the seasonal mismatch between summer heat surplus and winter demand. The authors integrate an HT-ATES store – a warm-well / hot-well doublet – into the existing 33.9 GWh/year network of Tirano in northern Italy (864 substations). A biomass-fired ORC cogeneration plant charges the aquifer in summer, injecting water at about 81 °C into the hot well; in winter the store is discharged again. Crucially, the proposed design also lowers the network from its existing 3rd-generation level (supply above 90 °C) to a 65 °C supply with about 50 °C return – and it is this low return that lets the store discharge for longer.
Even under hydrogeology the authors call unfavourable, the store reached a thermal recovery efficiency of 25.2 % once the aquifer settled into its stable annual regime. HT-ATES covered about 8 % of annual heat (2.7 GWh), the cogeneration unit ~40 % (13.4 GWh) and the biomass boilers ~47 % (16.1 GWh); the diesel backup fell to ~5 % (1.6 GWh). In practical terms the store supplies autumn and early-winter energy but no firm capacity for the cold peak: its contribution peaked at 1032 MWh in December and fell to zero in January, once the hot well dropped below the ~70 °C cut-off needed to feed the network. It displaces fuel but replaces no generator.
Against a conventional separate-supply reference the design saved about 1.6 GWh/year of primary energy, and under the assumption that biomass combustion is climate-neutral the system approaches carbon neutrality; annual operating costs were about 1.3 M€. Two honest qualifications: a 25 % recovery efficiency is low – good northern-European aquifers reach far higher – and the great majority of the heat still comes from biomass (both the cogeneration unit and the boilers are biomass-fired), so the carbon result is a balance-sheet assumption tied to fuel supply, not a measured value. The transferable lesson is that seasonal storage can displace fossil peaking even in a difficult setting, provided the network runs cool enough to use it.
As always, we recommend reading both articles in full – they are open access. For operators the common thread is temperature: low supply and especially low return temperatures are what make both decentralised heat-pump flexibility and seasonal storage work and lowering them is a lever available in almost any network today. Both studies also stop where daily operation begins – Paper 1 leaves out network hydraulics and pumping, and both are strongly site-specific. That is the gap our heatbeat Digital Twin is built to close: we model networks together with their sources, load profiles, storage, hydraulics and operating temperatures, so that flexibility measures and seasonal storage can be assessed under real network constraints. Together with our engineering team, we support you in planning, simulating and optimising your heat networks.
The next issue of our newsletter will be published on September 2, 2026.
Your heatbeat team