2. Review of Synergies between SGE and DHC Networks

2.1. Specific SGE Solutions Applied to Urban Environments

Figure 1. SGE-specific promising solutions in urban environments: tilted BHEs with a reduced surface footprint (a); radial groundwater horizontal wells for GWHEs (b); TAFs (c); and thermally-activated underground metro tunnels (d).

2.2. Modern DHC Installations: 5GDHC Networks

• Two pipes are enough. 5GDHC networks are bidirectional, which means that heat can flow into and out from buildings. An important distinction is made between “directed” and “non-directed” medium flow. In directed networks, a central pump station is used to always circulate water/brine in the same direction. The supply line comes from the BU at an ambient temperature, either for heating or cooling purposes, and the return line goes towards the BU. In directed networks, even 1-pipe scheme is possible (and simpler) if all buildings are topologically connected in series [94]. Under this configuration, the return temperature of a building becomes the supply temperature of the next one. However, for this configuration to be advantageous, the H&C demands need to be well balanced between adjacent buildings, and the extension of the grid is not as simple as in the previous case, where buildings would be connected in parallel. In non-directed networks, each station has its own, decentralized pump station. A warm pipe (15–25 °C) and a cool (5–15 °C) pipe act as the supply and return channels of heating demand substations, while for cooling demand it is just the opposite (see Figure 2). The direction of the heat-exchanging fluid with respect to the BU is determined by the balance between H&C demand at each moment.
• No restriction in building typology. Although highly efficient buildings are preferable nowadays under any circumstance, 5GDHC networks do not impose constraints on the type of building for its connection to them. Different degrees of thermal insulation or different emitter technologies (low or high-temperature radiators, radiant floor, fan coils, etc.) can be addressed with tailored HP solutions. Therefore, in the same network, most modern and efficient buildings can coexist along with historical buildings or tertiary buildings of any type or age, with all of them requiring water supply temperatures ranging from 40 °C to 90 °C [95].
• In 5GDHC networks, electricity is inextricably linked to heat. Unlike traditional DH networks, where steam or water is the true energy vector by itself, in 5GDHC networks, ambient-temperature water/brine can be an energy vector only if combined with electrically driven HPs. Therefore, parallel concerns about renewable and decentralized electricity generation are indispensable in designing and planning efficient 5GDHC networks.
• Increased operation complexity. For the operation of bidirectional networks, supply temperature or hydraulic pressure setpoints are not useful anymore. The temperatures of the warm and cool ring can oscillate according to the admissible range of HP performance and the BU (e.g., a BHE field). The actual temperature must be determined by an optimization function. This function considers the balance between H&C demands and a chosen optimization parameter (minimum operational cost, minimum emissions, or maximum HP efficiency, for instance). On the side of hydraulic pressure control, directed networks can be operated as traditional networks. In non-directed networks, the volumetric flow and its direction are mostly determined by the set of decentralized circulation pumps.
• Flexibility for different business models. In 5GDHC networks, old DH business models cannot apply anymore since electricity and heat are interlinked with prosumers, so new retail tariffs, investment costs, ownership, and operation schemes are necessary. As mentioned above, bidirectional metering of both electricity and heat is the most important technical prerequisite to establish any possible business model for 5GDHC networks. On one extreme, an external investor could own the network, the BU, and the set of decentralized HPs. Then, a one-time connection fee would be applied, a €/kWh price would be charged for consumption, and a different one would be charged for production. At the other extreme, a community of prosumers can associate to invest in a 5GDHC network, letting new prosumers connect to the grid by sharing its costs and ownership, and bearing the investment cost of their own HP. In between these two extremes, there are plentiful appealing choices. The following are some examples:

  ○

  When the BU is based on a UTES system, a private company that owns the whole infrastructure could charge only for the heating while offering cooling for free as a stimulus to recharge heat depletion.

  ○

  The network and the BU can be owned by a private company, while an energy service company (ESCO) would own the set of building substations. The final clients would pay only for H&C at an agreed price which, in theory, should be lower than traditional alternatives (gas boilers + air conditioning), but still high enough for the ESCO to obtain a payback for its initial investment plus an additional profit.

  ○

  The network, the BU, and the substations could be all owned by separate entities. The network would pay the BU owner, and the final prosumers would pay the network owner. Under this scheme, the owner of the BU would pursue maximizing its use with compensated H&C loads on a seasonal basis. The network owner would seek, as its main business driver, to achieve a balanced set of H&C demands on an hourly basis (reducing the need to use the BU). From the side of final prosumers, their main driver would be to compensate their own H&C demands instantly (reducing the need to use the network).

Figure 2. Illustrative scheme of a generic 2-pipe 5GDHC network, where H&C demands could take place simultaneously and the BU is a BHE field. This particular scheme corresponds to a “bidirectional energy flow with non-directional medium flow”, as defined in [16]. An additional third ring (black-colored) corresponding to electricity has been added to emphasize the coupling of heat and electricity that takes place within 5GDHC networks.

• Specific supply costs for H&C. It is defined as a yearly indicator, expressed in €/MWh_th/year.
• Exergy efficiency. It is defined as the ratio between total useful exergy and total exergy expenditures.
• System COP. This indicator is analogous to that of an individual HP, but it includes the influence of self-produced electricity by means of solar PV panels, denoting the expected close relationship between 5GDHC and renewable electricity generation:

$$CO{P}_{system}=\frac{Q_{h,dem}^{tot}+Q_{c,dem}^{tot}+W_{feed in}^{tot}}{W_{grid}^{tot}+W_{PV}^{tot}}$$

where $Q_{h,dem}^{tot}$ and $Q_{c,dem}^{tot}$ are the total H&C demand, respectively;  $W_{feed in}^{tot}$ in is the total self-produced electricity injected to the grid; $W_{grid}^{tot}$ is the total electricity consumed from the grid; and $W_{PV}^{tot}$ is the total self-produced electricity.

• Specific CO₂ emission. It is expressed as tones of CO₂ equivalent divided by MWh of thermal energy (heating + cooling). Here, the emission corresponds to the primary energy used, so it depends on the CO₂ factor of the power system mix.

Additionally, a new indicator was proposed in [98], named Demand Overlap Coefficient (DOC) in order to quantify the proportion of H&C that can be compensated between buildings within a 5GDHC network. DOC shows values from 0 to 1, being 0 for the case where the H&C demands never take place simultaneously, and 1 for the ideal case scenario defined above, when the BU is unnecessary (case 1). This indicator can be either defined for a network or a single building.

2.2.1. Overview of Concept Definition

Among the engineering community, the categorization of DH systems introduced in 2014 by Henrik Lund and co-authors [14] is nowadays widely accepted. It classifies DH systems into four different generations characterized by a set of disruptive technological and conceptual jumps in chronological order, from their conception in the nineteenth century until the present. These jumps follow a clear trend in the following aspects:

• Progressive decrease in the supply temperature of the heat-carrying fluid (steam/water) in the network.
• Progressive increase in the efficiency of the system due to a decrease in thermal losses, both in the generation and distribution of the thermal energy.
• Progressive transition from fossil fuels to renewable energy sources as the primary energy consumed by the network.

The term “5th generation DHC” (5GDHC) emerged just one year later, in 2015 [27]. This name was cleverly coined in the wake of Lund’s previous work, with the aim to attract attention to a specific concept of DHC networks. On purpose or not, by naming it with the term “5th generation”, it conveyed a sense of being the natural next step following the abovementioned trends that DHC networks had been showing throughout time (lower temperature, higher efficiency, and higher share of renewables). Furthermore, it added extra features, such as decentralized production and intrinsic coupling between electrical and thermal networks [99]. Nevertheless, the concept of 5GDHC is not newer than 4th-generation DH (4GDH) in terms of existing operative examples [16]. However, although its acceptance within Lund’s categorization is controversial [17,100,101], the fact is that 5GDHC has become the most widely and consistently used term for this kind of modern DHC projects [102]. Particularly, the research European D2Grids project has decidedly contributed to the prevalence of the term “5GDHC networks” over the rest of the nomenclatures. On its webpage, the concept is clearly and concisely stated through a list of five principles [103].

Simone Buffa and co-authors were the first to provide an exhaustive list of nomenclatures strictly fitting the concept of 5GDHC networks, but also of those corresponding to closely-related types of networks, for the sake of clarification and harmonization [16,102]. In the present work, additional comments add to the still open discussion.

The concept of “low-temperature DH” (LTDH) was before Lund’s categorization since it was a natural consequence of trying to lower the supply temperature of traditional DH networks as much as possible (T_sup = 60–70 °C) in order to integrate a larger portion of RTTs, and to lower transmission losses.

The term “ultra-low temperature DH” (ULTDH) has become unfortunately ambiguous. Initially, ULTDH was defined as an evolution from LTDH systems, where the limit of direct supply temperature (still with no booster HPs) was pushed to the lowest possible value (T_sup = 40–45 °C) [104,105]. More recently, other authors assimilate ULTDH as 5GDHC networks, where decentralized HPs are an intrinsic part of such concepts [106,107]. Although it is accurate to equate the terms LTDH and 4GDH, it is not so with ULTDH and 5GDHC. In [88], a clear distinction is made between ULTDH and 5GDHC networks, which are defined as ambient-temperature DH (ATDH) networks. In fact, LTDH and ULTDH can be considered as two subcategories of 4GDH [85], so any reference to “low“- or “ultra-low”-temperature networks should be avoided when referring to 5GDHC networks.

2.2.2. The Synergistic Effect of SGE+5GDHC Networks

Looking at the existing examples of 5GDHC networks in Europe, Simone Buffa and co-authors firstly identified 40 installations in Europe [16], mostly in Germany and Switzerland, but also in The Netherlands, Belgium, Norway, Italy, and UK. From this list, 25 out of the 40 installations used SGE in their BU. More recently, Marco Wirtz and co-authors updated the list of installations only in Germany, where 53 cases were identified [84], with some of them being a hybrid system between 4GDH and 5GDHC. In 77% of the surveyed installations, the BU was based on SGE, mainly in the form of horizontal GHEs and BHEs, but also GWHEs. From these examples, it is undeniable that SGE and its associated myriad of exploitation schemes is the natural partner of 5GDHC networks. Interestingly, the first 5GDHC network that could be identified as such by Marco Pellegrini and Augusto Bianchini [85] consists of a small DHC operative since 1991 that re-uses the groundwater inflow from the Furka Tunnel (Switzerland) as the energy source, in combination with decentralized GSHPs [108].

This is not in vain as community schemes for the exploitation of SGE are attracting attention in recent years in parallel to 5GDHC networks. In Denmark, the non-profit association Termonet [109] is named after a concept that consists of a cluster of customers sharing the same GHEs (BHEs, GWHEs, or TAFs) within a neighborhood. This kind of 5GDHC network is a simplistic version of 5GDHC networks since buildings connected to the grid are alike and usually not showing compensated thermal loads between them (Figure 3).

Figure 3. Illustrative scheme of a small 5GDHC network. Several residential units with individualized GSHPs and similar demand profiles (not necessarily simultaneous) share a vertical BHE field (BU), as a heat source (above) or as a heat sink (below).

Compared to individualized installations, a shared set of GHEs is advantageous in many ways:

• Smart and synchronized operation of decentralized HPs permits a lower aggregated peak load in the network, so the total size of the GHEs can be minimized, allowing much lower installation costs, reduced land utilization, lower impact to the ground, and minimized thermal interferences [110].
• When integrated in new urbanizing plans, its implementation is the most efficient and cost-effective. However, in existing urban areas, this option might be the only one possible to exploit the SGE resource.
• As in the general case of 5GDHC defined previously, shared GHEs permit the uptake of new business and operation models.

A representative and outstanding example of a Danish “Termonet” is in Silkeborg, which is operative since 2017. It consists of 15 individual houses sharing 6 × 120 m deep BHEs. The total installed HP power is 90 kW_th (15 units of 6 kW_th each), with no back-up equipment of any kind [111]. In Denmark, there are 11 Termonet projects (operative or under construction) [112].

In the UK, the same concept is commonly known as “Shared Ground Loop Arrays” [113,114,115,116]. In North America, the commercial term GeoGrid^TM [117] coexists with “GeoMicroDistrict” or “Networked Geothermal Energy”, which were coined by the “Home Energy Efficiency Team” (HEET) located in Massachusetts (US) [114]. This non-profit organization advocates for a progressive substitution of old and leaky natural gas grids by BHEs [118]. This is a very interesting approach where many issues converge, such as the mitigation/suppression of greenhouse gas (GHG) emissions from leaks in an extensive old gas pipe grid, the transition from fossil fuels to renewable sources, building electrification, and the shift/renewal of business and jobs. However, the implementation of such a big-scale transformation in energy generation and distribution is not evident, especially in densely packed urban areas. In locations with a markedly vertical distribution of the population (tall and packed buildings), the H&C demand density might be too large to be met by SGE systems, despite the lower demand for multi-family buildings (typically between 20% and 50% less in terms kWh_th/dwelling/year with respect to single-family detached buildings, depending on the year of construction and climate conditions [119,120]). Even if all the buildings have been previously enhanced in efficiency, the implementation of large enough GHEs would still be challenging, but not impossible.

In densely populated areas, the exploitation of the SGE resource is more efficient in the form of a DHC network rather than individual installations. Alina Walch and co-authors demonstrated that the exploitation of SGE under a DHC scheme would increase the potential of meeting H&C demands in Western Geneva by 2050, from 63% to 87% for cooling and from 55% to 85% for heating, thanks to direct heat exchange between buildings [121]. Most importantly, these numbers are independent from the Representative Concentration Pathway (RCP) scenarios defined by the International Panel on Climate Change (IPCC) [122]. Additionally, it is demonstrated that balanced seasonal ground heat exchange is highly beneficial in order to minimize thermal interference between BHEs, and to maximize the total amount of exchanged heat (comprising both H&C operation modes) along the year.