Sustainable Energy, Grids and Networks Disclosing the heat density of district heating in Austria in 2050 under the remaining European CO 2 budget of the 1.5 ◦ C climate target

The core objective of this work is to downscale the cost-effective heat supply of different European decarbonization scenarios generated by the aggregate model GENeSYS-MOD from the national to the community level in Austria and thus disclose the heat density of district heating in 2050. We assume that district heating encompasses geothermal, synthetic gas, hydrogen, waste, and large-scale heat pumps as renewable heat sources. The results determine district heating in 68 Austrian communities in Austria in 2050, which corresponds to 6% of the total number of communities. We find that GENeSYS-MOD results are capable in covering local trends in district heating since high shares of the projected heat densities at the local levels achieve values indicating economic viability. Further research should follow on how locally determined district heating and heat densities could be returned into more aggregate models, such as GENeSYS-MOD, in the sense of a feedback loop. That allows refining assumptions in the large-scale upper-level models, which in turn will increase the plausibility and realism of pathways at the European level.


Introduction
To implement the pathway in line with the Paris Climate Agreement [1] as analyzed by the IPCC's Special Report on Global Warming of 1.5 • C (SR15) [2], the European Commission set deep decarbonization targets together with national governments. In particular, the EU Green Deal describes the concrete goals in Europe, namely, a climate-neutral and resource-conserving economy and society [3]. The overarching goal is to reduce CO 2 emissions to net-zero and hence achieve climate neutrality by 2050. The principles of a net-zero, decarbonized society are based on three key points: (i) reduction of the energy demand [4], (ii) deployment and generation of renewable energy technologies [5], and (iii) an increase in efficiency regarding the provision of energy services and the associated optimal utilization of sustainable energy sources.
To achieve these long-term ambitions, the European Commission recently presented Fit for 55, a roadmap with specific actions and targets until 2030. This program commits to a 55 % reduction in CO 2 emissions in 2030 compared to those in 1990 [6]. The concrete measures affect almost all sectors of the energy system and should lead to a significant efficiency improvement and a massive overall reduction in fossil fuels. It implies, among others, binding annual targets to reduce energy consumption and to extend the already established EU emissions trading system (EU ETS) to new sectors. In addition to transportation, in the future, the building sector will be part of the EU ETS. In the building sector, using the annual anchored emissions reduction means a defined roadmap to complete decarbonization of the heating and cooling demand. In this paper, we look at what deep decarbonization of building heat demands may look like in 2050 in Austria and the implications of the corresponding sustainable energy mix for district heating.

Implications of decarbonization on the heating sector
The scope of changes required by 2030/2050 in the heating sector becomes even clearer at the national level. In Europe, the share of renewable energies in the heating and cooling sector in 2018 is only just above 20 % on average [7]. In Austria, it reaches 34%. However, fossil fuels continue to dominate there as well. In 2015, the heat demand for low-temperature heat services in Austria was approximately 96 TWh. This heat volume encompasses low-temperature heat demand of residential buildings (domestic spacial heating demand, calculated on the basis of the outdoor temperature), industrial heat demand below 100 • C (e.g., food sector, machinery, and wood), and process heat demand [8]. In the residential building sector, natural gas, oil, and coal account for almost 45% of space heating and hot water demand [9]. The share of district heating reaches almost 15%, and more than one million households are connected to district heating networks. According to [10], the total heat production from district heating was around 24 TWh in 2016. Thereby, the share of renewable energy was 45%. Besides, the share of waste sources was 9%. In 2018, district heating supplied 18% of the total heat demand in the residential building and service sector with a share of 48% renewable heat sources. Thereby, the amount of district heating was 20 TWh. Table 1 provides an rough overview of the Austrian heat market, as it shows the proportion per heat source/generation technology on the total heat demand for space heating and hot water.
Nevertheless, of the nearly 4,000,000 residential dwellings in Austria, more than one million are heated with natural gas, and more than 500,000 are heated with oil [11]. If these heating systems are converted to renewable energy supply by 2040, this corresponds to a retrofitting of more than 80,000 units per year, or more than 225 per day -only in Austria. To achieve this goal, measures that go beyond the electrification of heat supply are necessary, which may require an expansion of district heating networks. This holds true even when substantial heat saving measures are implemented [12].
In Europe, good conditions for district heating exist [13], especially in the provision of heat services in densely populated or urban areas [14] because of high heat densities that are found there. In addition to heat density, the connection rate is a key factor determining the efficiency of district heating/cooling networks and thus their implementation. In Austria, a benchmark of 10 GWh/km 2 at a connection rate of 90 % is currently used when deciding whether to supply an area with district heating. 1 This reference value considers the area effectively supplied by district heating and not the total area. Thus, the exclusion of land areas that contain woodland, mountain, agricultural and other low heat density areas is crucial. The reference/benchmark value is in line 1 http://www.austrian-heatmap.gv.at/ergebnisse/. with findings regarding district heating networks also from the Scandinavian region (Denmark, Sweden, and Finland) [15]. These are rough estimates, but they do allow an initial assessment of the economic viability or feasibility of a district heating network. In a detailed consideration and evaluation of district heating networks, numerous factors play a decisive role. For example, the design and topology of district heating networks demonstrates significant impact on their cost-effectiveness [16,17]. In addition, the cost-optimized heat supply is also influenced by the location of heat generation units/sources within the networks [18]. The influence of the connection rate and linearly decreasing heat densities on the profitability of district heating networks is investigated in [19,20]. The study in [21] presents an optimization approach for district heating strategic network design. Further works also evaluate the impact of the heating system topology on energy savings [22]. When examining the economic viability of district heating networks, building renovation measures must also be taken into account [23]. Recently, the results in [24] show that a 2 − 3% building renovation rate per year results in a 19 − 28% decrease of the long-term district heating demand, which consequently also reduces the heat densities of district heating networks. However, studies show that a reduction in heat density is not necessarily a barrier to district heating networks [25]. For example, energy taxes which can certainly be expected in the future (e.g., higher taxes on fossil fuels) can improve the profitability of sparse district heating networks [26]. Following these considerations and in light of ambitious CO 2 reduction targets assumptions exists that rising CO 2 prices exhibit a similar effect. However, this is valid only in the case of deep decarbonization of the generation mix feeding into district heating networks. In general, a variety of alternatives to decarbonize the energy mix of district heating networks exists. Among others, geothermal [27], biomass [28], waste [29] and heat recovery from industrial excess heat [30] are likely to be the primary heat sources in sustainable district heating networks. Eventually, the increasing cooling demand and the co-design of district heating and cooling networks can also increase the economic viability of these and counteract the reduction of heat density from an economic point of view [31].

Implications of large-scale numerical model results at the local level
For quantifying solutions of complex planning problems, researchers use numerical models. In general, these models strike a balance between complexity and aggregation. Integrated assessment models (IAMs) are large numerical models covering complex interrelationships between climate, society, economics, policy, and technology [32]. Particularly, IAM contribute to the understanding of global energy decarbonization pathways [33]. Evaluating and discussing IAM involves, among others, the appropriate level of regional (spatial) aggregation of countries in the modeling analysis [34]. Generalizing this aspect reveals an aspect already known but essential in the context of large numerical models. Setting priorities regarding the level of detail becomes necessary for modelers, which inevitably creates trade-offs in the analysis regarding the granularity of temporal, spatial, and other dimensions [35]. Accordingly, IAMs should increasingly be supplemented with other models and analytical approaches [36]. Not least for this reason, large-scale detailed energy systems models also play a significant role in the analysis of energy systems in the context of climate change. Compared to IAMs, they more strongly emphasize the level of detail in terms of techno-economic characteristics. However, the lack of granularity remains; these global systems models consider only a highly aggregated spatial resolution. To name just two selected approaches, PRIMES [37] and Table 1 Proportion of heat sources/generation technologies on the total heat demand (space and hot water) and absolute number of households supplied for Austria in 2017. The number of households supplied by heat pumps and solarthermal is given in total. Source: [9]. GENeSYS-MOD [38] are aggregate energy system models focusing on the European energy system with a spatial resolution at the country level. Further approaches are needed to disaggregate results obtained at the country level to finer scales, such as districts, neighborhoods, and other local levels. In this context, a novel approach in the context of merging local activities/behavior in sustainable local communities into a large energy system model (bottom-up linkage) is presented in [39]. In this study, local flexibility options are integrated into the global energy system model EMPIRE, which provides, in principle, only country level resolution. This and other work confirms the emerging trend of making top-down and bottom-up linkages between different spatial-temporal levels of resolution to drive decarbonization across all sectors.

Objective and contribution of this work
Against this background, the core objective of this work is downscaling European decarbonization scenarios of the heating sector to the community levels serving end-users in 2050. In particular, downscaling considers the highly efficient and local use of sustainable heat sources in district heating (e.g., geothermal sources, co-firing synthetic gas and hydrogen in cogeneration plants, and large-scale waste utilization). In addition, the topography of district heating networks is of particular importance and plays a crucial role in applied downscaling. This allows estimates of realistic decarbonized district heating networks in 2050 to be obtained, which can be compared with existing networks. Thereby, the heat density of district heating networks serves as a comparative indicator and permits a rough estimation of the changes needed for district heating networks considering the 1.5 • C and 2.0 • C climate target. An Austrian case study is conducted, downscaling the cost-effective results of the heating sector in 2050 from the large numerical energy system model GENeSYS-MOD, from the country to the community/local administrative unit levels. In general, GENeSYS-MOD exhibits a focus on generic heat supply options based on primary energy sources, rather than specific local excess heat sources that is in general the fundamental idea of district heating. Accordingly, this study can be seen as an attempt for a stress test applying GENeSYS-MOD's heat supply in the context of district heating. The GENeSYS-MOD results, and thus the values to be downscaled implicitly, include the remaining European CO 2 budget in line with the 1.5 • C and 2.0 • C climate target.
The method applied (Section 2) consists of a simplified optimization model, which computes the amount of district heating at the local (community) level. Section 3 presents and discusses the results of this work. First, the heat supply in a representative region is presented in Sections 3.1 and 3.2. The projected heat densities at the community level in 2050 are presented in Sections 3.3 and 3.4. Finally, Section 4 concludes this work and provides an outlook for future work.

Materials and methods
This section explains the methodology developed in this work. First, Section 2.1 presents the output from the European Horizon 2020 project openENTRANCE (incl. GENeSYS-MOD results), since this is the main input for the downscaling. Therein, information about the different heat sources/generation technologies that are downscaled is provided. Section 2.2 explains the mathematical formulation of the optimization model in detail. Then, Section 2.3 shows the workflow that is used to determine the implemented shares of district heating. Finally, Section 2.4 concludes this section and presents further data and open-source tools used in this work.

Heat supply of the Austrian residential and commercial sector in 2050: Four different decarbonization scenarios
This section presents the heat generation mix covering the Austrian residential and commercial heat demand in 2050 for four different scenarios, which have been developed within the European Horizon 2020 openENTRANCE project. They are named as follows: Directed Transition, Societal Commitment, Techno-Friendly, and Gradual Development. Within each of them, specific fundamental development of the energy systems is described while aiming for a sustainable transition of the provision of energy services. The first three scenarios assume different approaches to limit global warming to around 1.5°C as laid out in the Paris Agreement. Particularly, the results of these scenarios implicitly consider the remaining European fraction of the CO 2 budget of the 1.5 • C climate target. The last scenario (Gradual Development) can be interpreted as a less ambitious scenario, limiting global warming to around 2.0°C climate target. Accordingly, the results of this scenario consider the remaining European fraction of the CO 2 budget of the 2.0 • C climate target. Below, the scenarios are described briefly, before the quantitative results at the country level are presented. For a more detailed description of the scenarios, refer to [40][41][42]. Further information is also available on the website of the project 2 and on GitHub. 3 The underlying concept of the four scenarios is a three-dimensional space consisting of the following parameters: technology, policy, and society. Each scenario describes a specific pathway to reach a decarbonized energy system taking into account a pronounced contribution of two dimensions. Regarding the third dimension, a development is assumed that leads to no significant contribution to the decarbonization of the energy system. Table 2 Heat generation by source in Austria in 2020 and the four different decarbonization scenarios in 2050 obtained from GENeSYS-MOD. Geothermal, hydrogen, synthetic gas, waste, and half of heat pump (air-sourced) generation is used in district heating. Source: [41,43,44].

Generation by source in TWh
Obtained from GENeSYS-MOD • Techno-Friendly describes a development of the energy system where a significant market-driven breakthrough of renewable energy technologies gives rise to the decarbonization of energy service supply. Additionally, society acceptance supports the penetration of clean energy technologies and the sustainable transition.
• Gradual Development differs from the other scenarios; it assumes emissions reductions that (only) stabilize the global temperature increase at 2.0°Celsius. At the same time, a combination of each possible sustainable development initiative of the energy system is realized in this scenario. Although the other three dimensions contribute to decarbonization, they do not push it sufficiently, and this results in a more conservative scenario than the others. Table 2 shows the heat generation by source/technology in Austria in 2050 for the four scenarios. These values were obtained during the course of the openENTRANCE project and are generated by the open-source aggregate model GENeSYS-MOD [8].
In this work, the naming convention of heat sources/generation technologies from GENeSYS-MOD is essentially followed to ensure consistency between aggregated (i.e., downscaling input values) and local (i.e., downscaling output values) levels. Nevertheless, we introduced the heat sources waste and geothermal that were initially not included in the list of heat sources from openENTRANCE results. We separated waste as part of biomass and geothermal from heat pump (ground-sourced) heat generation using estimates from national Austrian studies in [43,44] to complement the GENeSYS-MOD results. Note that the values obtained from GENeSYS-MOD do not explicitly include district heating, which is why its 2020's value in Table 2 cannot be specified. The total heat generation (and thus total heat demand) is significantly reduced when comparing the values of 2020 and 2050. The heat demand reduction varies between −30% and -35% and is highest in the Societal Commitment scenario. District heating (bottom row in Table 2) describes the amount of heat generation used for district heating. In this work, the assumption is made that geothermal, hydrogen, synthetic gas, waste, and half of the total heat generation by heat pumps (air-sourced) are used in district heating. Therefore, we claim that • geothermal [45] and waste [46] as renewable heat sources contribute to the decarbonization of heat supply by the integration into district heating.
• the limited amounts of synthetic gas and hydrogen are preferably used in district heating (i.e., co-firing in cogeneration plants [47]) if they supply (residential and commercial or low-temperature) heat demands [48][49][50].
• half of the cost-optimal heat supply of heat pumps (airsourced) of the aggregate model GENeSYS-MOD are used in district heating through implementation of large-scale heat pumps. Accordingly, heat pumps (air-sourced) significantly contribute to supply decarbonized district heating networks [51].

Mathematical formulation of the optimization model
Building upon the amount of district heating obtained by the aggregate model GENeSYS-MOD, this section explains the optimization model used to downscale heat supply to the LAU (Local Administrative Unit) level in detail. Before, Table 3 shows the spatial nomenclature of this work based on the NUTS nomenclature. Particularly, this includes representative examples for the LAU level. Against this background, Eq. (1) shows the objective function of the model that is used for the downscaling.

Workflow to obtain implemented shares of district heating
In order to maximize the objective function value, the described mathematical formulation of the optimization model allocates the amount of district heating to the LAU level. However, this does not necessarily ensure that obtained heat densities of district heating networks reach the benchmark of 10 GWh/km 2 being assumed in this work. Consequently, this section explains in detail how the optimal values of q dh l (i.e., district heating at the LAU level) is further processed resulting in heat densities of district heating higher than the benchmark value. The developed workflow is as follows: 1. Starting with the optimal amount of district heating q dh l at the LAU level obtained from the optimization model. 2. Identification all LAUs that do not achieve the required heat density benchmark value of 10 GWh/km 2 . 3. For each of those LAUs, the heat density of district heating within the corresponding NUTS3 region and thus network level is calculated. 4. In case that the heat density reaches values higher than the benchmark at the NUTS3 level, the supply using district heating remains since LAUs are then connected to or in the surrounding area of high heat density areas.
5. Otherwise, q dh l is set to zero as no economic viability can be expected there due to lower achieved heat densities than the benchmark.
Finally, steps 1 to 5 allow to calculate implemented district heating under the condition that either the local heat density at the LAU or the network heat density at the NUTS3 level achieves the assumed heat density benchmark value of 10 GWh/km 2 .

Further data and open-source tools used
In order to determine total heat demand at the LAU level (q total l ), we apply proportional downscaling using population as downscaling proxy. The fields of application of proportional downscaling are not limited to the modeling of energy systems but to different fields of scientific and practical studies. The reason for this is the intuitive application and that it offers possibilities for tailor-made adaptions, in particular, related to the downscaling driver and proxy. In this context, the study in [53] provides a comprehensive analysis of different proxies for the downscaling of global environmental change, including gross domestic product, emissions, and other indicators. However, downscaling aggregated values of energy system often uses proportional downscaling and population as a proxy [54]. Table 4 shows the data used to obtain heat demand at the LAU level in 2050 including population estimates for Austria until 2050. Moreover, we use STATatlas (https://www.statistik.at/atlas/) in order to set φ l for each LAU l. The four different categorizes encompass the following items: urban (I), suburban (II), and rural (III and IV). We set φ l to 0.5 for urban and suburban LAUs and equal to 1 for rural LAUs.
The developed optimization model is implemented in Python 3.8.12 using the modeling framework Pyomo version 5.7.3 [55]. It is solved with the solver Gurobi version 9.0.3. For data analysis, we use the IAMC (Integrated Assessment Modeling Consortium) common data format template with the open-source Python package pyam [56]. All materials used in this work are available in the author's GitHub webpage. We refer to the corresponding repository in [57].

Results and discussion
This section presents the results of the Austrian case study in 2050. The focus is put on the mix of heat sources/generation technologies and district heating in the four different scenarios. Section 3.1 shows the heat supply of a representative Austrian NUTS3 region in detail. Building upon, Section 3.2 compares heat supply in an urban and a rural LAU/district. Section 3.3 presents the obtained heat densities of district heating networks. Finally, Section 3.4 syntheses the results of district heating and provides indications/information that could be returned into more aggregate models, such as GENeSYS-MOD, in the sense of a feedback loop.

Heat supply in a representative NUTS3 region in 2050
This section presents the results of the NUTS3 region Salzburg and Surroundings (AT323). Fig. 1 shows the most relevant results in this region on LAU/district level for the four different scenarios. District heating supplies heat demands in 5 different LAUs/communities. In particular, the LAUs are in the surrounding area of Salzburg city (marked by the star). The remaining LAUs in the NUTS3 region are supplied decentralized/on-site. Details of the heat sources that supply heat demands in LAUs with district heating and with decentralized/on-site heat systems are presented in the following Section 3.2. The amount of district heating varies between 1045 and 1132 TWh per year (Fig. 1, top  right). The highest value is achieved in the Gradual Development scenario, since this is the scenario with the lowest heat demand reduction. The heat density of district heating in the 5 LAUs is shown in Fig. 1, bottom right. The highest heat density is achieved in Salzburg city and reaches approximately 30 GWh/km 2 in each scenario. The comparable low heat densities in two of the five LAUs (marked by rectangle and plus) is further discussed in Section 3.4.

Comparison of heat supply in urban and rural LAUs/community
Building upon the so-far presented results of the NUTS3 region Salzburg and Surroundings, this section shows the heat sources/generation technologies supplying heat demands in an urban and in a rural LAU/community. We use Salzburg city (urban community) and Abtenau (rural community) as representative LAUs. Fig. 2 shows the mix of heat sources supplying heat demands in both LAUs. The geographical location is shown on the top left in Fig. 2. In Salzburg city (marked by the orange edge), district heating supplies heat demands, which uses large-scale heat pumps (air-sourced), hydrogen, synthetic gas, and waste as heat sources/generation technologies. High shares of district heating particularly are generated by large-scale heat pumps (air) and using hydrogen. In contrary, the heat supply in the rural district Abtenau uses small-scale heat pumps (air), heat pumps (groundsourced), biomass, and direct electric heating systems. Among all four scenarios, high shares of heat demands are supplied by heat pumps (air-and ground-sourced). However, the share of each technologies varies to some extent significantly, which becomes evident when comparing exemplarily the Techno-Friendly and Gradual Development scenario. In the Techno-Friendly, smallscale heat pumps (air-sourced) are the dominant heat source, whereby heat pumps (ground-sourced) supply high shares of heat demands in the Gradual Development scenario.

Heat densities of district heating in LAUs in 2050
This section shows the heat density of district heating at the LAU/district level in 2050. Fig. 3 shows the heat density for the four different scenarios. Particularly, the values of LAU's heat densities are sorted in descending order indicating those LAUs/communities that do not reach the required heat density of economic viability, which is assumed to be 10 GWh/km 2 . Exemplarily, in the Directed Transition scenario, 107 LAUs with district heating are found. In this scenario, the highest heat density is 43.17 GWh/km 2 . 2 of the 5 LAUs in the NUTS3 region Salzburg and Surroundings are highlighted, namely, Salzburg city (marker by the star in Fig. 1) and Anif (marked by the rectangle in Fig. 1). Both LAUs are part of the same district heating network as already illustrated in the left subfigure in Fig. 1. Accordingly, the appearance of heat densities below the assumed threshold/benchmark for economic viability can be argued as those LAUs are connected to high heat density areas. The distribution of heat density values remains mostly the same between the four different scenarios. For the sake of clarity, explicit annotations are omitted in the three (smaller) scenario subfigures at the bottom.

Geographical localization of district heating in 2050
We focus in this section on those LAUs with lower heat densities than assumed to be required for economic viability for district heating and their geographical location in respect to other district heating supply areas. As indicated in Fig. 3, LAUs with low heat densities can be quite justified in case that they are located in the surrounding area of high heat density areas (e.g., Salzburg city and Anif). However, other LAUs that do not achieve the required heat density benchmark (of 10 GWh/km 2 ) and at the same time are not closely located to high heat density areas are unlikely to be implemented. Accordingly, Fig. 4 shows the heat map of district heating in Austria at the LAU level under the requirement that district heating achieves the required heat density benchmark within NUTS3 regions in the Directed Transition scenario. As previously mentioned, the model basically decides to supply heat demands in 105 LAUs by district heating. 63 of them already achieved heat densities higher than the benchmark value. The heat map in Fig. 4 still shows 68 LAUs since 5 are closely located to high heat density areas and thus achieve in total the benchmark (at the NUTS3 level).
Accordingly, district heating is unlikely to be implemented in 37 LAUs. Table 5 summarizes the results for district heating in the four different scenarios. It shows that as a result of the heat density benchmark at the NUTS3 level, the share of implemented district heating varies between 74 and 90%. In particular, this   23 26 means exemplarily that in the Techno-Friendly scenario, 74% of the assumed heat supply using district heating leads to heat density values higher than 10 GWh/km 2 . In view of the previous assumptions that 50% of heat pumps (air-sourced) are used in district heating, this results in implemented shares between 23% and 40%, whereby the highest share is achieved in the Directed Transition scenario. In view of the underlying narratives of particularly the three ambitious decarbonization scenarios from Section 2.1 (therefore excluded the Gradual Development scenario), two interesting implications can be derived from the results here: • In absolute terms, the Techno-Friendly scenario demonstrates the highest share of district heating with 20.09 TWh under the condition that district heating networks within the NUTS3 levels achieve the heat density benchmark of 10 GWh/km 2 . The main driver for this is the significant penetration of (large-scale) heat pumps (air-sourced) that characterizes this scenario.
• Nevertheless, the implemented share of district heating in GENeSYS-MOD's district heating assumptions is the highest in the Directed Transition scenario and reaches 87 %. Also, this result is reflected in the fact that the share of large-scale heat pumps (air-sourced) achieves here its maximum with 40 %.

Conclusions and recommendations
The sustainable energy transition requires methods to bridge the gap between global decarbonization pathways and the corresponding measures at local levels. This work emphasizes the development of a downscaling technique based on a simplified optimization model, which we apply to the European heating sector under several scenarios in line with the Paris Agreement and its remaining CO 2 budget. Next, we use the cost-effective European heat supply from the aggregate model GENeSYS-MOD to analyze results at the local administrative unit level in Austria. The implemented method allows to disaggregate heat supply from the country to the community level, which contributes to the linkage between the modelers and the practitioners perspective. The remaining European CO 2 budget (and related CO 2 prices) in line with the 1.5 • C climate target is considered by the GENeSYS-MOD results. The downscaling includes the claim that the renewable heat sources geothermal, green gases (synthetic gas and hydrogen), waste, and large-scale heat pumps (air-sourced) are used in district heating.
We found that the cost-effective heat supply at the European and national level in 2050 implies that district heating covers heat demand in 68 communities in Austria in 2050, which corresponds to 6% of the total number of communities. The results demonstrate that district heating continues to be picking cherries from beneficial areas (i.e., densely populated with high heat densities). However, the reduction of heat densities compared to today's values is mainly driven by a significant reduction of heat demands by building renovation measures and poses a challenge for district heating in the future. Nevertheless, the localization of district heating networks in the surrounding of urban areas indicates economic viability, too.
In view of comparing different scenarios in this work, the results indicate that the aggregate model GENeSYS-MOD is capable of handling planning approaches for decarbonization of energy systems. Particularly, the policy push in one of the scenarios (Directed Transition) is also reflected in the determined local heat densities of district heating. This can be seen in particular by the projected share of both district heating and large-scale heat pumps in GENeSYS-MOD's results.
We anticipate our work as a starting point for discussing the role of district heating as an enabler for large-scale, highly efficient, and local integration of renewable heat sources such as geothermal, synthetic gas, hydrogen, and waste in sustainable energy systems with decreasing heat demands. Further research should follow on how obtained district heating networks and their heat densities (incl. the generation of large-scale heat pump (air) units) could be returned into more aggregate models, such as GENeSYS-MOD, in the sense of a feedback loop. That allows refining assumptions in the upper-level large-scale models, which in turn will increase the plausibility and realism of pathways at the European level.