Table of Contents
Ground Tubes
Author: Berthold Kaufmann
Heat distribution via underground pipes – district heating in the neighborhood?
The power plants that will be operated with renewable energy sources in the future, as well as heating systems in residential areas that may work with combined heat and power (CHP), will be supplied with surplus energy from the summer half-year into the winter half-year. These summer surpluses of electrical energy must therefore be stored temporarily, for example by converting PV and wind power into liquid or gaseous energy sources using electrolysis etc. ('power-to-gas'). These can then be used in the winter months in the same way as the fossil fuels natural gas and crude oil. The waste heat generated during combustion in CHP plants or fuel cells can generally already be used very well today to heat buildings and provide domestic hot water. The energy content can therefore be used almost completely: electrical energy from the generator for light, power and communication and the waste heat for heating. This will become all the more important in the future because the available budget for renewable energy sources is likely to be more limited and primary energy will therefore be significantly more expensive than it is today, especially in winter.
The at least partial supply of individual neighborhoods via district or local heating networks is therefore an important component of the so called heat production transformation and is therefore an important option for heating and domestic hot water preparation.
Local heating networks are also interesting for the use of heat pumps (HPs) because the heat from larger central HP systems can be distributed to surrounding buildings via a local heating network. In this context, however, a warning should be issued about the heat distribution losses that then occur in the local heating networks: the prerequisite for the use and distribution of heat via heat-conducting pipes laid in the ground is that the heat distribution networks are optimized the best possible. This means that all heat distribution losses must be consistently balanced and minimized as far as possible.
The heat losses occurring by the distribution of waste heat from combined heat and power plants (CHP) in so-called local and district heating networks are considerable compared to the final energy made available to the user at the house connections. It is therefore instructive to look at the amount of energy lost per year compared to the amount of final energy supplied for different building types.
For passive houses and buildings renovated to the EnerPHit standard in particular, the 'energy density' achievable in the grid is relatively low because the final energy requirement for space heating and domestic hot water is only around 50 kWh/m²a and possibly even less, i.e. only around 6 000 kWh/a of final energy is delivered per residential unit (120 m²). For typical old buildings with a consumption of around 240 kWh/m²a or 29 000 kWh/a, the turnover is significantly higher [AKKP 46 FW]. These figures already include the heat distribution losses within the building.
In [AKKP 46 FW], various scenarios for the configuration of district heating and local heating distribution networks were drawn up and compared. Several different qualities of thermal insulation of the heat distribution pipes were compared with each other and the length of the distribution pipes in a network was varied. The results show that with a moderate improvement in the quality of the pipe insulation and an optimization (shortening) of the network length, acceptable heat distribution losses can already be achieved with today's pipe technology. So passive houses and districts renovated to the EnerPHit standard can be supplied with heat economically. The 'specific' network length in meters/residential unit (m/unit) is an important factor here. The first goal of network optimization must therefore be to keep the specific network length low and thus the network connection density high; values below 6 m/unit should be aimed for [AKKP 46 FW].
A corresponding spreadsheet is provided as an Excel tool for this purpose, which makes the heat distribution losses and their effect on the regenerative supply visible. In the tool all the geometry of the grid i.e. length of underground tubes and quality of thermal insulation can be defined. The characteristic values determined in this way can also be used in the PHPP to ultimately estimate the primary energy input for the heat supply.
The higher the feed-flow temperatures in the distribution networks, the more relevant are the occurring distribution losses. This is illustrated in Figure 1 using an exemplary local heating network. The buildings and the distribution pipes are shown in the site plan. Each building entrance has a branch pipe from the route and a riser pipe, the lengths of which add up to a figure total of around 12 meters per residential unit. The more residential units are connected, e.g. if the buildings are also extended during renovation, the smaller this figure can become. A re-densification in the course of refurbishment therefore also has a positive effect in this context. Figure 2 also lists the absolute totals for the individual variants in tabular form. This means that the length of the heat distribution network is at the limit of what appears to be compact and reasonable (10 m per residential unit) [AKKP 46 FW]).
The heat losses (Figure 2) are quite high at around 17 kWh/(m²a) or 13 kWh/(m²a) after the building has been extended and amount to around two thirds of the expected heating energy demand of the renovated apartments. If the feed-flow temperature is reduced from 65 °C to 55 °C, around 2 kWh/(m²a) can be saved. However, 55 °C seems questionable to provide domestic hot water with an assured temperature in heat transfer stations in every apartment.
This makes it clear that the use of centralized heat pump systems and the distribution of heat at a high temperature level greatly degrades the good energy efficiency of a heat pump. Such losses must always be taken into account in the primary energy balance. [AKKP 61 WP]
The same problem naturally also applies to heat distribution from a CHP plant: the valuable useful heat is made available at a very high temperature level and it must therefore be ensured that heat distribution losses are minimized as far as possible.
Using the Excel tool 'district_heating_losses' and the input parameters mentioned above, the characteristic values of the network configuration can be calculated as shown in Figure 2. [AKKP 61 WP]
The efficiency of the local district heating network then results from the ratio of useful heat (space heating energy demand and DHW heat demand) to the total heat demand: (useful heat plus network distribution losses).
In the PHPP, the network losses can only be calculated as a lump sum with a Y-value of the distribution routes. The 'district_heating_losses' tool allows a more detailed consideration of the district or local heating network, as e.g. storage losses in the network and other influencing factors are taken into account As can be seen from the results in the table in Figure 2, the proportion of heat distribution losses is very high compared to the useful heat for the long networks (12 m/unit) and thus the winter efficiency of the networks is quite low (only 65 %) if the losses are calculated honestly, see variant 1. If the summer conditions are considered, when only the useful heat for DHW heating is delivered in the network, then it becomes even worse because the distribution losses are at the same height in summer.
An improvement is achieved by shortening the network (variant 3), improving the thermal insulation of the underground pipes (variant 7) and, if applicable, increasing the density of apartments (variant 5) and reducing the feed flow temperatures (variants 2, 4, 6, 8). With all these measures combined, the efficiency in the example increases to around 80 % (variant 8).
Cold local district heating in underground pipes
In variants 9 and 10, the feed flow temperatures were reduced to 15 °C. As expected, this immediately drastically reduces the heat losses in the network despite standard thermal insulation. [AKKP 61 WP]
This is the situation for so-called 'cold local district heating': a heat source with very low feed flow temperatures is used for distribution via the underground pipes. In this case, the higher temperatures for the useful heat for heating and hot water in the apartments are then provided by a heat pump.
The heat pump can either be installed decentrally in each residential unit or centrally in each building. This has two advantages:
Firstly, the primary circuit of the heat pump is operated at a relatively high temperature. This can be ground temperature, i.e. a circuit fed from ground probes, for example. Or it can be waste heat from industrial processes, waste incineration or waste water, which may only be available at temperatures below 30°C. This primary temperature can then be used to operate decentralized residential and building-by-building heat pumps with very high SPFs because they have a good primary heat source available all year round.
Secondly, this concept reduces the heat distribution losses in the network to practically zero. These 'cold' heat distribution pipes only need to be slightly insulated where they run through the building in order to prevent condensation on the pipes. It would therefore be possible to completely dispense with the thermal insulation of the underground pipes. Even with Y = 1 W/mK, the network efficiency is well above 80%.
In this way, compact heat pumps have also been implemented as systems with centralized systems, i.e. each apartment has a ventilation unit with heat recovery combined with a heat pump for supply air heating and a small DHW cylinder. The heat pump is supplied via a brine circuit. [Drexel 2007] [Dimplex]
However, the investment costs will shift with this concept: instead of a central heat pump or CHP system in the entire local district heating network, a heat pump must now be installed for each building or for each residential unit. How the investment costs will compare to the reduced heat losses is beyond the scope of this report. However, an overall economic analysis could be carried out with [districtPH].
Allerdings werden sich bei diesem Konzept die Investitionskosten verschieben: anstatt eine zentrale WP oder KWK-Anlage im gesamten Nahwärmenetz muss jetzt für jedes Gebäude oder für jede Wohneinheit eine WP installiert werden.
Literature
[AKKP 61 WP] B. Kaufmann, Wärmepumpen für die Altbausanierung, in Arbeitskreis Kostengünstige Passivhäuser, Protokollband 61, Serielle energetische Sanierung nach Passivhaus Prinzipien, Darmstadt, 2024
[AKKP 46 FW] B. Kaufmann, Bewertung leitungsgebundener Energieträger: Wärmeverteilverluste von Fern-/ Nahwärmenetzen Szenarien für die optimierte Fernwärmenutzung aus KWK. In: Arbeitskreis Kostengünstige Passivhäuser, Protokollband 46, Nachhaltige Energieversorgung mit Passivhäusern, Darmstadt 2012
[Drexel 2007] C. Drexel, Tagungsband zur Internationalen Passivhaustagung 2007 (Bregenz), Energie Institut Vorarlberg (EIV) / Passivhaus Institut, Darmstadt 2007.
[Dimplex] https://www.dimplex.co.uk/professional/heat-pumps/ambient-temperature-network?
[districtPH] districtPH is used to create energy balances for neighborhoods, from smaller groups of buildings up to whole districts. A specific strength of the tool is the ability to investigate the effects that different scenarios of development and retrofitting have for the building stock. The analysis takes into account grids for electricity and heat, renewable energies, electromobility and public consumers. districtPH was developed within the EU-funded Sinfonia project. http://www.sinfonia-smartcities.eu/en/
