In the European Union, space heating (SH) and domestic hot water (DHW) account for 79% of energy use in households (192.4 Mtoe) (European Commission 2016). As around 84% of domestic heat is generated from fossil fuels, there is an urgent requirement to improve system efficiencies and support greater sustainable energy integration. Such improvements are necessary to align with the 2050 climate obligations for decarbonisation under the Kyoto Protocol (UN FCC 2014a) and more recently the UN-led Paris Agreement (UN FCC 2014b).
District heat network (DHN) research suggests that improving efficiencies and supporting sustainable energy technologies can be achieved through lowering building hot water supply and return temperatures (Lund, Werner et al. 2014) (Østergaard and Svendsen 2016a)(Olsen, Lambertsen et al. 2008). Moreover, histroically there has been a trend towards lower temperature heat medium within DHN (see Table 1), which has the following benefits:
|Generation||Period||Heat carrier||Temperature (°C)|
|1st Generation (1G)||Mid 1800s–1930s||Steam||>100|
|2nd Generation (2G) (“High-temperature”)||1930s–1970s||High temp. pressurised water||<100|
|3rd Generations (3G) (“Medium-temperature”)||1970s–2010s||High temp. pressurised water||80/40|
|4th Generation (4G) (“Low-temperature”)||2010s–present||Low-temperature water||50–55/25|
|5th Generation (5G) (“Ultra low-temperature”)||2015–present||Ultra-low temperature water||<49/<24|
The aforementioned system benefits allow for increased energy efficiency, reduced greenhouse gas emissions, fuel and infrastructure cost reduction, resource optimisation, material utilisation, ease of control, increased system flexibility and improved user comfort levels.
Currently, DHN are supplied predominantly by large-scale centralised fossil fuel (Su et al. 2015) or bioenergy (Ericsson and Nilsson, 2006) heat stations. Additionally, there has been some work on solar thermal heat sourced in DHN (Woods and Overgaard 2016, Pauschinger 2016). However, in future, a shift towards decentralised heat generation from low-temperature and sustainable heat sources should become a more feasible option. As a result, opportunities for prosumer participation, where homeowners can be both buyers and sellers of heat, will become more prevalent (Brange et al. 2016, Brand et al. 2014). For the prosumer concept to work, the discordance of low-temperature renewable energy supply and domestic demand must be balanced. One solution is thermal energy storage (Xu et al. 2018, Van Helden et al. 2016, Thomsen and Overbye 2016) or a combination of thermal and battery storage in multi-vector energy systems (MES).
At the University of Nottingham’s Creative Energy Homes test site, project SCENIC (Smart Energy Networks Integrated in Communities) is investigating how the enablement of renewable energy and the prosumer concept can be developed through the integration of community scale heat and power networks. The first phase of project SCENIC involved the design and installation of a heat network. The first of its kind in the world to combine LTDHN with distributed generation at a community level. SCENIC’s heat network configuration has been designed and installed to facilitate a low-temperature heat medium, the uptake of sustainable energy generators, the integration of heat and power networks, smart controls, community level distributed generation, and prosumer participation. A concise site overview is presented in Table 2.
|Occupants||Domestic 12; Office 35|
|Aim||To provide low-temperature heat and enable the prosumer concept|
|Heat source||Solar evacuated tube arrays, flat plate solar thermal arrays, solar PV arrays, biomass boiler, gas boilers|
|Length of network||220 m|
The design process of the network first requires heat loads for each of the connected buildings. The heat loads are modelled and act as the base for the rest of the design decisions. Once modelling and the system operating temperatures are decided – in this case 55/25°C, the pipe layout and component selection is possible. This paper presents the modelling, component sizing, system configuration, and proposed energy flows of this novel system.
To enable the sizing of heat network components, steady state peak heat demand and annual heat consumption Figures are required. Network pipe diameter, pump sizing, heat exchanger sizing, and hence the cost and capacity of the network can be projected based on the highest demand. Annual heat consumption Figures are used for heating plant and buffer tank sizing, as well as defining the heat revenues of the system.
Options for modelling peak heat demand include those set out in CIBSE TM54: Evaluating operational energy performance of buildings at the design stage guidelines. However, in order to more accurately reflect the operation of the building in practice, IES-VE software was utilised in the modelling approach. IES-VE’s advanced dynamic thermal simulation provides a higher level of heat demand modelling accuracy due to the added consideration of room occupancy, location and orientation, build material properties, energy system types and efficiencies, heat gains, and air flow (infiltration, mechanical and natural ventilation).
All seven of the CEH was modelled separately. Input data for the IES-VE simulation was taken from CIBSE and BS standards, manufacturer’s datasheets, or real-world published data on the houses.
The steady state SH load represents a simplified IES-VE building model where certain assumptions remain constant. For example, the outdoor temperature value is assigned as –3.9°C corresponding to the CIBSE guideline design winter temperature. Solar and casual heat gains are not considered so consequently the results are static (see Table 3).
|Building name||Instantaneous DHW heat load (kW)||Instantaneous DHW heat load (kW) × diversity factor (0.3)||Steady-state SH load (kW)||Total design peak heat load (kW)|
|David Wilson House||34.89||10.467||9.17||19.64|
|Mark Group House||34.89||10.467||8.22||18.69|
|Tarmac House (No. 12)||34.89||10.467||3.50||13.97|
|Tarmac House (No. 10)||34.89||10.467||2.80||13.27|
|Saint-Gobain Nottingham H.O.U.S.E||34.89||10.467||1.85||12.32|
Instantaneous DHW peak demand was calculated separately using the assumptions represented in Table 4. The instantaneous DHW peak demand is assumed to deliver hot water to 2 showers and kitchen use for 4 people in each property. The calculated 34.89 kWp/house peak seen in Table 3 is in line with the 37.5 kWp/house proposed in the Danish Standards included in CIBSE’s Heat networks: Code of Practice.
|Peak hot water demand (ṁ)||25 L/person/hour (7am–8am)||CIBSE part G – Table 2.10|
|Heated water temperature change (ΔT)||50°C (from 10°C to 60°C)||BS 8558:2015 Appendix B|
|Specific heat capacity of water (Cp)||4.187 kJ/kg.K||CIBSE part G – equation 2.2|
|Heat capacity equation||Q = ṁ Cp ΔT||CIBSE part G – equation 2.2|
|Diversity factor||0.3||CIBSE – Heat networks: Code of Practice for the UK – Figure 9|
Peak instantaneous DHW heat demand is unlikely to occur in all properties at the same time. Therefore, to ensure representative results, a diversity/coincidence factor is applied. The diversity factor is the ratio of the effective peak load used. The Danish Standard DS 439, recommended by CIBSE AM12:2013 and BS 8558/BS EB 806-3, gives the coincidence factor as a function of the number of properties served by the heat network. In this case, a site with 7 properties has a coincidence factor of 0.3.
Network configuration is determined by the placement of infrastructure, as well as the network size and heat source location. Fortunately, the Creative Energy Homes test site can accommodate a largely soft-dig trenching system with a simple branched pipe network (see Figure 1). Branched networks allow for consumers to be connected to the main loops individually by a smaller service branch. Benefits of branched networks include flexible design, easy installation, and additional branches can be added to the network loop at later stages.
As mentioned, project SCENIC is the first MES (Multi-vector Energy System) DHN with truly distributed energy generation. Consequently, a novel network configuration was required to facilitate prosumer and consumer participation at each property. A double loop system has been installed, where a network loop supplies heat to each property via a consumer HIU (heat interface unit) (see Figure 2), and a generation loop delivers heat generated at each property through a prosumer HIU to a CTS (central thermal store) (see Figure 2).
Heat networks are designed on peak load in winter, yet year round efficiency needs to be considered. Despite peak load needing to be met, oversizing needs to be balanced against technically practical and economically feasible component selection. For most of the year the system will only operate at part load with peak output is only needed for a few hours.
Initially, network operating temperature needs to be selected. Following this, when sizing a heat network, three components are considered:
Within this section, we consider the sizing of the aforementioned heat network components using the modelled results. Once component sizes have been identified it is possible to present the overall heat network layout within Section 4. Unlike standard heat network component sizing and design, project SCENIC has to consider the integration of distributed generation on all 7 networked properties.
The transition to LTDH is widely understood to be the next step in heat network efficiency. Generally, LTDH has a supply temperatures below 55°C and return temperatures of around 30°C described in numerous papers (Olsen, Lambertsen et al. 2008, Paulsen, Fan et al. 2008, Torío and Schmidt 2010, Dalla Rosa and Christensen 2011, Li and Svendsen 2012, Østergaard and Svendsen 2016b, Yang, Li et al. 2016). As a result, Project SCENIC aims to achieve a network supply temperature below 55°C for most of the year and a return temperature of 30°C.
Lund, Werner et al. (2014), the seminal work on the definition of LTDH (4th Generation) describes how future systems can achieve supply and return of 50°C/20°C for most of the year without compromising comfort levels and hygiene requirements. It is understood that operating temperatures should be kept as low as possible to avoid unnecessary heat losses. Additionally, operating temperatures will also effect:
The relationship between the aforementioned parameters is discussed more comprehensively alongside relevant system component information in the following sections.
The generation technologies have been installed before the installation of the heat network except for an evacuated tube solar thermal array on oldest property (David Wilson House). Table 5 specifies the capacity and location of each distributed generation technology. Energy flows are presented later in the paper.
|Building name||Energy generation technology||Peak installed capacity (kW)|
|David Wilson House||Evacuated tube solar thermal collector
|Mark Group House||Evacuated tube solar thermal collector||4.42|
|EON House||Gas boiler||12|
|BASF House||Flat plate solar thermal collector||3.2|
|Tarmac House (No. 12)||Biomass boiler
Flat plate solar collector
|Tarmac House (No. 10)||Flat plate solar collector||1|
|Saint-Gobain Nottingham H.O.U.S.E||Evacuated tube solar thermal collector||1.28|
|Central thermal store||Immersion heater||12|
In order to enable intermittent renewable generation, appropriate generation scheduling, and to satisfy peak loads energy storage is required. The tank is sized to match the peak flow rate of 1.99 l/s (based on formula 3) sustained over 10 minutes (600 seconds). Due to stratification, the required supply temperature is only available in one-quarter of the tank. Therefore the tank will need to accommodate 2400 seconds of flow. Hence, the required volume of the tank is:
where S is tank volume (l), F is diversity factor, and V is heat medium flow rate (l/s). With a heat network peak flow rate of 1.99 l/s, the hot water volume capacity required is 4,776 l.
The CTS selected for project SCENIC was a 10,000l accumulator tank with 100mm of polyurethane insulation with 0.23 W/mk thermal conductivity (Figure 3). The tank is capable of storing energy for days or even weeks with minimum heat losses. The tank can operate with a maximum supply temperature of 95°C, and operating pressure of 3 bar, typically 1.5 bar. The tank was oversized to account for additional energy generation installation in later projects.
The energy stored in the thermal tank is available as sensible heat in the raised water temperature. The capacity of the central thermal store is:
where E is energy content (kJ), cp is the specific heat of water (kJ/kg°C), ΔT is the temperature difference between the hot water and the surroundings, m is the mass of stored water (kg). The temperature difference between the heat medium and surrounding air is assumed to be uniform. Assuming 50% of the stored hot water is available for usage, then tank can accommodate over 202.67 kWh as sensible heat at 55°C.
In standard DHN a heat transfer fluid, usually water, is passed through a two-circuit system (flow and return). However, project SCENIC utilised a four-circuit configuration comprised of a network and generation loop (see Figure 2). Within the network loop a flow pipe transports hot water from a CTS to the consumer, a return pipe delivers cooled water back. Within the generation loop, a flow pipe delivers heat from the prosumer’s distributed energy source to a CTS, a return pipe delivers cooled water back.
For accuracy in the heat distribution design of such a system we consider the following key factors:
The pipe has been selected as the heat distribution infrastructure due to its excellent thermal insulation. The polyurethane (PU) foam insulation and corrugated outer jacket help retain heat and allow for flexibility, in turn reducing heat losses, improving ease of installation, and limiting jointing requirement (see Figure 4).
The piping system is made up of a crosslinked polyethylene (PE-Xa) water bearing carrier pipe, pentaneblown PU foam insulation and a PE-LLD outer jacket. Consequently, the product is advantageous as it durable, providing corrosion and chemical resistance, has permanently low heat and pressure loss, and very low roughness.
Each pipe section must accommodate the peak heat load of those properties down flow of the network. Moreover, when sizing pipe sections the principle, as small as possible, as big as necessary, should be followed.
Network distribution pipe diameter is inferred from the pipe flow rate and pressure loss. Flow rate can is calculated from the known heating requirement and the temperature difference between flow and return pipes. Flow rate is:
where V is the heat medium flow rate (l/s), Q is the peak heat flow (kW), Cp is the specific heat capacity of water (kJ/kg·K), ϑV is the flow temperature in the network (°C), ϑR is the return temperature in the network (°C), and ρ is the density of water (kg/l).
Using calculated flow rate and the manufactures specification the pipe should be as small as possible without exceeding a 200–250 Pa/m pressure loss, it is possible to identifiy the most appropriate branch pipe sizes.
The benefits of using a twin pipe was quantified based on the network operating temperatures and the following heat loss calculation:
where Q is heat loss during operation (W/m), U is thermal heat transfer coefficient (W/mK), ϑB is average operating temperature (55°C), and ϑE is the assumed soil temperature (10°C). Using the equation above the heat loss reduction when using twin compared to single pipes has been calculated (Table 6).
|Diameter requirement (mm)||2× single pipes||Twin pipe heat loss (W/m)||Heat loss reduction (%)|
|Thermal heat transfer coefficient (W/mK)||Heat loss (W/m)||Thermal heat transfer coefficient (W/mK)||Heat loss (W/m)|
Pumps are fundamental to heat distribution in networked hydronic systems. For DHN, pumps are not the main energy consumer in systems, but correct sizing is important to deliver minimum energy transport costs. Due to variable heat demand maximum duty pumping occurs over very few operating hours. However, pump size must be based on maximum system load, specifically maximum flow and pressure loss as well as peak efficiency. As a result, variable speed pumps with a good control strategy are ideally suited to gain lower heat, less power, and better efficiencies.
Another important factor is temperature difference within a distribution loop as it will affect pumping requirement. If ΔT is low, higher flow rates are required to satisfy heat demand, therefore causing higher pressure loss and hence pumping costs increase. ΔT should be as high as possible to minimise pumping costs.
Pump duty is sized to overcome the flow resistance in the pipework and the pressure differential at HIUs. When determining pump duty the following hydraulic requirements are essential:
Alongside these hydraulic requirements, a minimum pump energy consumption must be taken into account. Therefore, both generation and network loops use the pump which has the highest energy efficiency on the market, which is designed for pumping liquid through heating systems with variable flows. The pump duty point can be optimised to reduce energy costs. As a result, pumps were sized to work at 30–75% of the total load. It is important for correct operation that the system range lies within the duty range of the pump (see Table 7).
|Parameter||MAGNA3 (N) Single-head pumps ranges||Project SCENIC ranges|
|Flow rate, Q||21.81 l/s||1.99 l/s|
|Head, H||18 m||12 m|
|System pressure||1.6 Moa (16 bar)||Max. 3 bar|
|Liquid temperature||–10 to 110°C||30°C to 55°C|
Overall, benefits of the pump include low energy consumption, built-in differential pressure and temperature sensor, heat energy monitor, and easy system optimisation.
Heat exchangers separate the primary heat network and secondary consumer circuits. The appropriate Heat Interface Unit (HIU) containing flat plate heat exchangers are required based on peak kW SH and DHW requirements (see Table 3), and system operating temperatures.
The domestic properties in project SCENIC have preinstalled heating and hot water systems, as a result, a direct connection was not an option for SH. An indirect HIU was selected for both SH and DHW to provide a physical barrier between primary and secondary loops. In LTDH a heat exchanger barrier between loops prevents the spread of Legionella bacterium. As the temperature in the network buffer tanks is likely to fall below the required 60°C to prevent the spread of Legionella, a DHW heat exchanger is essential. DHW temperatures will be boosted on the secondary side with an immersion heater to meet the 50°C tap temperature requirement.
For the network loop, supplying SH and DHW to each property, a Danfoss FlatStation – 7 Series DS Fully Insulated was chosen (Figure 5). Furthermore, a bespoke Danfoss FlatStation – 3 Series BS Basic Fully Insulated was designed to accommodate prosumer participation within the generation loop (Figure 5). Figure 6 illustrates how the FlatStation redesign facilitated the supply of heat from distributed generation sources within buildings into the network for distribution.
Table 8 specifies the selection of HIU on each of the test site properties. Sizing was based on heating capacity and supply/return temperature. Furthermore, Danfoss HIU significantly outperform others on the market due to their specialised valve technology and energy saving features.
|Parameter||Consumer unit||Prosumer unit|
|Product||Danfoss FlatStation – 7 Series DS Fully Insulated||Danfoss FlatStation – 7 Series DS Fully Insulated||Danfoss FlatStation – 3 Series BS Basic Fully Insulated|
|Installed on||Mark Group House, David Wilson House||BASF, EON, Nottingham, Tarmac 10, Tarmac 12||All|
|Heating capacity||SH 10 kW; DHW 35 kW||SH 7 kW; DHW 35 kW||7 kW|
Typically heat networks require two energy inputs, heat energy to compensate for distribution heat losses, and electrical energy for pumping to compensate for frictional losses. Within this paper we have discussed the design of project SCENIC’s heat network, however, as a MES both heat and power networks will be linked. As a result, the energy inputs and outputs are more complicated than the typical model. Figure 7 illustrates how the energy flows and system configuration will operate.
The extent to which each energy technology will supply heat or power will depend on the scheduling scenarios to be tested within the next phase of the project.
The paper discusses the design and sizing of the first community scale MES heat and power network with distributed energy generation. Prosumer participation is facilitated through a double loop heat pipe configuration, allowing consumers to be both buyers and sellers of heat. In addition, the use of renewable energy generation and energy storage enables LTDH, which in turn reduces operational costs, heat losses, energy consumption and carbon emissions.
In order to reduce the well-documented system design margin, the SCENIC heat network will operate at low temperature. To permit LTDH, there must be adequate heat transfer through radiators inside each property. As a result, the next phase of project SCENIC will tackle the redesign of interior heating systems to allow heat to be input and output from the properties. The adequate transfer of heat is essential to maintain comfort levels and reduce the energy consumption of the network.
Heat and power network integration will be addressed in the next phase of the project through the installation of a 23 kW lithium ion battery. The battery will power an immersion heater in the CTS and distributed heat generation technologies. In turn, the battery will be charged by the numerous distributed photovoltaic arrays on site.
A formative element in achieving system efficiency is the improved operation and scheduling of the technologies. A holistic control approach for optimising decentralised heat generators, pumps, and valves will require well-functioning monitoring, communication, and decision support tools. As a result, a large part of project SCENIC will be focussed on controlling the system components designed and sized in this paper. A multi-objective, object orientated optimisation approach will be used focussing on real-time data to support decision making through a building energy management system (BEMS).
In summary, infrastructure has been installed at the University of Nottingham’s Creative Energy Homes test site to facilitate distributed generation and the prosumer concept in a combined heat and power network. For the heat network, overall configuration, component sizing, and HIU redesign have been explained in this paper. The work presented here acts as a base from which power integration, control systems, and prosumer business model design can be built upon extending the relevance of the projects outcomes.
The authors have no competing interests to declare.
Brand, L, Calvén, A, Englund, J, Landersjö, H and Lauenburg, P. 2014. Smart district heating networks – A simulation study of prosumers’ impact on technical parameters in distribution networks. Applied Energy, 129: 39–48. DOI: https://doi.org/10.1016/j.apenergy.2014.04.079
Brange, L, Englund, J and Lauenburg, P. 2016. Prosumers in district heating networks – A Swedish case study. Applied Energy, 164: 492–500. DOI: https://doi.org/10.1016/j.apenergy.2015.12.020
Dalla Rosa, A and Christensen, JE. 2011. “Low-energy district heating in energy-efficient building areas.” Energy, 36(12): 6890–6899. DOI: https://doi.org/10.1016/j.energy.2011.10.001
Dalla Rosa, A, Li, H, Svendsen, S, Werner, S, Persson, U, Ruehling, K, Felsmann, C, Crane, M, Burzynski, R and Bevilacqua, C. 2014. Annex X Final report: Toward 4th Generation District Heating: Experience and Potential of Low-Temperature District Heating.
Ericsson, K and Nilsson, LJ. 2006. Assessment of the potential biomass supply in Europe using a resource-focused approach. Biomass and Bioenergy, 30: 1–15. DOI: https://doi.org/10.1016/j.biombioe.2005.09.001
European Commission. 2016. “Heating and cooling.” Retrieved 28th February, 2017, from: https://ec.europa.eu/energy/en/topics/energy-efficiency/heating-and-cooling.
Li, H and Svendsen, S. 2012. “Energy and exergy analysis of low temperature district heating network.” Energy, 45(1): 237–246. DOI: https://doi.org/10.1016/j.energy.2012.03.056
Lund, H, Werner, S, Wiltshire, R, Svendsen, S, Thorsen, JE, Hvelplund, F and Mathiesen, BV. 2014. “4th Generation District Heating (4GDH): Integrating smart thermal grids into future sustainable energy systems.” Energy, 68: 1–11. DOI: https://doi.org/10.1016/j.energy.2014.02.089
Olsen, PK, Lambertsen, H, Bøhm, B, Svendsen, CCH, Larsen, SCT and Worm, J. 2008. A new Low-Temperature District Heating system for low-energy buildings. The 11th International Symposium on District Heating and Cooling. Reykjavik.
Østergaard, DS and Svendsen, S. 2016a. “Case study of low-temperature heating in an existing single-family house—A test of methods for simulation of heating system temperatures.” Energy and Buildings, 126: 535–544. DOI: https://doi.org/10.1016/j.enbuild.2016.05.042
Østergaard, DS and Svendsen, S. 2016b. “Replacing critical radiators to increase the potential to use low-temperature district heating – A case study of 4 Danish single-family houses from the 1930s.” Energy. DOI: https://doi.org/10.1016/j.energy.2016.03.140
Pauschinger, T. 2016. 5 – Solar thermal energy for district heating. In: Wiltshire, R (ed.), Advanced District Heating and Cooling (DHC) Systems. Oxford: Woodhead Publishing. DOI: https://doi.org/10.1016/B978-1-78242-374-4.00005-7
Su, D, Zhang, Q, Wang, G and Li, H. 2015. Market Analysis of Natural Gas for District Heating in China. Energy Procedia, 75: 2713–2717. DOI: https://doi.org/10.1016/j.egypro.2015.07.693
Thomsen, PD and Overbye, PM. 2016. 7 – Energy storage for district energy systems. In: Wiltshire, R (ed.), Advanced District Heating and Cooling (DHC) Systems. Oxford: Woodhead Publishing. DOI: https://doi.org/10.1016/B978-1-78242-374-4.00007-0
Torío, H and Schmidt, D. 2010. “Development of system concepts for improving the performance of a waste heat district heating network with exergy analysis.” Energy and Buildings, 42(10): 1601–1609. DOI: https://doi.org/10.1016/j.enbuild.2010.04.002
UN FCC see: United Nations Framework Convention on Climate Change. 2014a. “Kyoto Protocol.” Retrieved 3rd May, 2017, from: http://unfccc.int/kyoto_protocol/items/2830.php.
UN FCC see: United Nations Framework Convention on Climate Change. 2014b. “The Paris Agreement.” Retrieved 3rd May, 2017, from: http://unfccc.int/paris_agreement/items/9485.php.
Van Helden, W, Weber, R, Heinz, A, Furbo, S, Brendel, T, Vossel, M, Salg, F, Williamson, T, Gantenbein, P, Andersen, T and Jacobsen, P. 2016. Combined development of compact thermal energy storage technologies (COMTES) [Online]. Available: http://cordis.europa.eu/project/rcn/103641_en.html [Accessed 24th August 2016].
Woods, P and Overgaard, J. 2016. Historical development of district heating and characteristics of a modern district heating system. In: Wiltshire, R. (ed.), Advanced District Heating and Cooling (DHC) Systems. Elsevier. DOI: https://doi.org/10.1016/B978-1-78242-374-4.00001-X
Wu, W, Wang, B, Shang, S, Shi, W and Li, X. 2016. “Experimental investigation on NH3–H2O compression-assisted absorption heat pump (CAHP) for low temperature heating in colder conditions.” International Journal of Refrigeration, 67: 109–124. DOI: https://doi.org/10.1016/j.ijrefrig.2016.03.019
Xu, LY, Torrens, JI, Guo, F, Yandg, XD and Hensen, JLM. 2018. Application of large underground seasonal thermal energy storage in district heating system: A model-based energy performance assessment of a pilot system in Chifeng, China. Applied Thermal Engineering, 137: 319–328. DOI: https://doi.org/10.1016/j.applthermaleng.2018.03.047
Yan, J, Lee, DJ, Chou, SK, Desideri, U, Li, H, Li, H and Wang, SJ. 2014. “International Conference on Applied Energy, ICAE2014: Challenges in Smart Low-temperature District Heating Development.” Energy Procedia, 61: 1472–1475. DOI: https://doi.org/10.1016/j.egypro.2014.12.150
Yang, X, Li, H and Svendsen, S. 2016. “Evaluations of different domestic hot water preparing methods with ultra-low-temperature district heating.” Energy, 109: 248–259. DOI: https://doi.org/10.1016/j.energy.2016.04.109