Excess heat production of future net zero energy buildings within district heating areas in Denmark

Energy 48 (2012) 23e31

Contents lists available at SciVerse ScienceDirect

Energy journal homepage: www.elsevier.com/locate/energy

Excess heat production of future net zero energy buildings within district heating areas in Denmark Steffen Nielsen*, Bernd Möller Aalborg University, Vestre Havnepromenade 9, Aalborg, Denmark

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 November 2011 Received in revised form 21 March 2012 Accepted 3 April 2012 Available online 1 May 2012

Denmark’s long-term energy goal is to develop an energy system solely based on renewable energy sources by 2050. To reach this goal, energy savings in buildings is essential. Therefore, the focus on energy efficient measures in buildings and net zero energy buildings (NZEBs) has increased. Most buildings in Denmark are connected to electricity grids and around half are connected to district heating (DH) systems. Connecting buildings to larger energy systems enables them to send and receive energy from these systems. This paper’s objective is to examine how excess heat production from NZEBs influences different types of DH systems. In the analysis three different types of DH systems are analyzed, and three technology development scenarios for each are created. The examination is from a technical perspective and looks into how the overall heat production within DH areas is affected by the NZEBs excess heat production from solar thermal collectors. The main findings are that the excess heat from NZEBs can benefit DH systems by decreasing the production from production units utilizing combustible fuels. In DH areas where the heat demand in summer months is already covered by renewable energy, adding seasonal heat storage is essential to utilize the NZEB production. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Excess heat District heating Net zero energy buildings Renewable energy Solar thermal

1. Introduction In recent years, the focus on energy efficient measures in buildings and net zero energy buildings (NZEBs) has increased. This has led to many different definitions of the term [1]. There is not a general definition of what characterizes NZEBs, creating some difficulties when using the term. However, to give a common understanding of the NZEB concept as it is used throughout this article, the following definition is used: “NZEBs are buildings which have a greatly reduced energy demand that on an annual basis can be balanced out by an equivalent generation of energy from renewable energy sources. Also NZEBs are connected to other energy infrastructure like electricity grids or district heating (DH) networks” [2]. By using this definition, NZEBs are not seen as autonomous buildings that supply themselves with energy at all times during the year. Instead, they are buildings connected to the existing

* Corresponding author. Tel þ45 99408412. E-mail addresses: [email protected] (S. Nielsen), [email protected] (B. Möller). 0360-5442/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.energy.2012.04.012

energy infrastructure that have a greatly reduced energy demand and are supplemented by renewable energy sources and produce as much energy as they consume annually. Fig. 1 shows a two-step overall approach for reaching the net zero energy balance. The first step is to construct an energy efficient building with a low-energy demand. Hereafter, energy supply is added to achieve the net zero energy balance [3]. Since the net zero balance is reached annually, sometimes NZEBs produce more than they consume, and at other times, they consume more than they produce. As Lund states in [4], there is, from an overall energy system point of view, a mismatch between the hourly production and consumption in NZEBs. Even though the buildings have an annual net exchange of zero, the mismatch affects the overall energy system in different ways. The differences mainly depend on the technologies used in the NZEBs and the technologies used in the rest of the energy system. As mismatches can be both positive and negative, Lund quantifies the mismatch by proposing different factors for buildings with photovoltaics and for buildings with small-scale wind power. Lund’s article focuses on the electricity mismatch, but if NZEBs are connected to DH grids, these grids may also be affected by the excess heat production from NZEBs. In 2009 the total net heat demand in Denmark was 204 PJ, and of this 129 PJ was supplied by DH [5]. A 77.2% share of the heat within

24

S. Nielsen, B. Möller / Energy 48 (2012) 23e31

houses in the prognosis for new buildings. Since removal of buildings has a smaller percentage share of the total built area than new constructions and the new buildings have lower heat demands, these two approach each other when looking at heat demands. To reach the second objective, the method of using the heat demands found in the first objective to simulate three different sizes of typical DH systems is used. By making energy systems analyses of these systems in regard to implementing solar thermal production and storages, the effects of the excess heat from NZEBs are examined. In the present article, two main tools have been used. The first is a geographical-information system tool called a heat atlas, and the second is an energy system modeling software named energyPRO. Both are described briefly in the following sections. Fig. 1. Net Zero Balance of an NZEB [3].

2.1. Heat atlas DH areas was produced at cogeneration heat and power plants (CHPs) [5]. The resources used within DH in 2009 were: 45.8% RE (including household waste), 28.5% natural gas, 20.5% coal and 5.1% oil [5]. DH has advantages and disadvantages compared to individual alternatives. DH makes it possible to utilize waste heat from industries and waste incineration. It is also a more flexible energy source which can utilize a variety of different energy sources, e.g. geothermal heat. Moreover, the use of CHPs, which utilize waste heat from electricity production, is an important reason for having DH as a part of a flexible energy system. The significant disadvantages of DH are the grid loss of heat and the large investment costs. In general, the grid loss is around 20% for DH in Denmark [5]. Due to the disadvantages, there are limits to where DH can be used. Since such a large share of the Danish heat demand is supplied by DH, it is reasonable to analyze the effects that implementing the NZEB concept will have on these systems. As described previously, the NZEB definition makes it possible for NZEBs to connect to DH grids and thereby export excess heat production to these areas. These effects have not yet been quantified and depend very much on the location and the specific DH system that the NZEBs are connected to. An important aspect in this regard is that new, lowerenergy buildings within DH areas are not obligated to connect to the DH grids in Denmark. However, in this article, the goal is not to quantify the share of NZEBs connected to future DH areas, but to find the effect on heat production in DH grids when all new singlefamily buildings within DH areas are built as NZEBs and connected to the DH grid. By connecting all new, single-family buildings, it is possible to find the maximal effect on the DH systems. The objectives of the article are: 1. To assess in which types of DH areas that future, single-family NZEBs should be built, the heat demand of these NZEBs and the heat demand of the existing buildings. 2. To examine the effects of excess heat production of NZEBs in different typical Danish DH systems in regard to a future 2050 context. 2. Methodology and tools To reach the first objective, a prognosis of the development within the Danish building stock, with a focus on buildings within DH areas, has been made. This is done by using the Danish building register to look at historic development and assuming similar future development in building construction. The building register is also used to find out how large a share of buildings is within DH in the present system. Since it has not been possible to find the area of buildings that have been removed every year, all of the existing buildings are assumed to exist in 2050. To compensate for this assumption, the approach has been to include only single-family

An essential tool used in this article is the heat atlas [6,7]. The methodology of linking scenarios for heat demand and supply by geographical location was initially used in the Reveille project [8] and later on in relation to Heat Plan Denmark 2008 [9], where the methodology was used for the first time in combination with geographical-information system (GIS) modeling. The methodology of the heat atlas consists of four components: 1. National data from the Danish Building Register (BBR) on building level. 2. Heat consumption models based on building period, type and usage. 3. A model for calculating the costs of connecting buildings to existing or new DH-networks. 4. A geographical database for heat planning based on municipal heat plans. The outcome is a spatial database with heat demand and supply for each building. This information can be used for various analyses linked to geographic structure of the heating system [10]. In this article, the heat atlas is used to quantify heat demand and the number of current buildings located within DH areas. The heat atlas is also used to find the area of single-family houses have historically been built within different types of DH areas. This is used to prognosticate the floor area of future NZEBs within DH areas, which is later used to determine the heat demand for NZEBs within DH areas. 2.2. EnergyPRO EnergyPRO by EMD [11] is used for techno-economic analyses of different energy projects. In this article, version 4.1.1.111 of energyPRO has been used. The model is specifically suited for simulating CHP plants and DH systems with multiple energy producers. The model can be operated to calculate in ten minute steps, but for the modeling in this article, one hour steps have been used. Normally, energyPRO is used to model concrete energy projects, but for the analyses in this article, it has been used in a different manner. The reason for choosing energyPRO is its ability to model storages, CHPs, and solar thermal production as well as to connect different DH areas; the latter is very important for the interaction between NZEBs and a DH system. With energyPRO it is possible to model the heat demand from a group of NZEBs with solar thermal production and connect the group of NZEBs to a DH grid to examine the effects. 3. Geographical area and heat demand In this section the geographical boundaries of the analysis are established by finding the DH areas and dividing them into six

S. Nielsen, B. Möller / Energy 48 (2012) 23e31

overall categories. This is based on GIS analyses with data from the Danish building register and the annual energy producer count from the Danish Energy Agency. The expected floor area and heat demand for new NZEBs in 2050 is assessed by looking at the historic development within the building mass geography. Afterwards, the heat atlas methodology is used to find both the built area and the heat demand of the existing buildings within DH areas. 3.1. The DH area types In Denmark, there are approximately 415 individual DHnetworks, which are categorized by size and DH generation technology. In Denmark, the energy production method within DH areas is largely combined heat and power (CHP) plants and boiler production to a lesser extent. CHP plants can be divided into larger central and smaller decentralized plants. Currently, the registered fuel types amount to 19 [12], and within most DH areas more than one type of fuels is used. Therefore, dividing up the areas by primary fuel use has been chosen. To distinguish between central CHP and decentralized CHP, the DH areas with a total electricity capacity of 100 MW or more are chosen as central areas. Everything between 100 and 0 is considered decentralized CHP, and everything without electricity generation capacity is a boiler area. Since there is not any official measure for classifying central CHP plants, the division at 100 MW has been used to approximate the amount of units generally considered to be central in Denmark. This type of division amounts to: 12 central CHP areas, 276 decentralized CHP areas and 137 boiler areas in Denmark. In Table 1, the fuel consumption for each type of area is shown. This provides a simplified overview, where the fuels have been categorized into five fuel types. The first two categories are coal and natural gas. The third called “waste” includes both waste heat from industrial processes and waste incineration plants. The fourth category called “biomass” includes biogas, straw, biomass waste, wood chips, wood pellets, bio oil and solar thermal. The fifth category named “other” includes fuel oil, gasoil and waste oil. As shown in Table 1, coal amounts to around 70% of the total fuel consumption within central CHP areas. Within decentralized CHP areas, the most common fuel used is natural gas. However, since there are shares of waste and renewables, mainly biomass, these different fuels are used to categorize decentralized CHPs. Even though there only are a few decentralized coal CHP areas, they are included for methodological reasons. As for the boiler areas, around 90% use biomass. This amounts to the six different categories of DH areas listed below: 1. 2. 3. 4. 5. 6.

25

some areas classified as coal could get 49% of their heat from, e.g., waste incineration plants. It can also be questioned how relevant the fuel consumption in the present system is for categorizing future energy systems. The categorization of fuels is, thus, primarily given to show what is used today, and where it is reasonable to place NZEBs with solar thermal heating. The most important aspect in regard to the categorization is the different size divisions of DH areas, which also is the main difference in the scenarios. Because the DH type, by means of the heat atlas, is known for each individual building, the categorization makes it possible to examine the buildings stock within each type of DH area. 3.2. Single-family houses built within DH from 1980 to 2009 Since the objective of this paper is to look at energy use in 2050 or 39 years into the future, it is important to make a prognosis of how large an area of buildings is to be built in this period of time. This can be done by examining the historic development. As most building construction in Denmark is single-family houses, these are seen as the main building type for new NZEBs. Fig. 2 shows the development within single-family houses from 1940 to 2008. Fig. 2 shows that from the beginning of the 1960s until around 1980, the construction of single-family houses was higher than both prior and later time periods. The reasons behind this large increase in construction of single-family houses were both an increase in wealth and people moving outside the cities. Statistics from development in recent years does not point to a similar increase in construction as was seen from 1960 to 1980. Within the 1980e2008 period, there were large variations in construction levels. 1982e1988 was higher than 1989e1996 and 1997e2008 was also higher than 1989e1996. To account for these variations, the average construction rate from 1980 to 2008 is used as the prognoses for the next 39 years. The assumption is from the authors’ perspective seen as reasonable since it incorporates time periods with lower and higher than average building activity. Also it is in the line with the 2050 forecasts in both Heat Plan Denmark and the Danish Engineering Society’s climate plans. The timespan used prevents the prognosis from making too optimistic or pessimistic predictions. From 1980 to 2009 around 57,323,345 m2 single-family houses have been built in Denmark, and of these 38,448,970 m2 were inside DH areas. To determine in which type of DH area these have been built, the building area is divided into the six categories of DH, as shown in Fig. 3. Fig. 3 shows that most single-family houses have been built in relation to DH areas with central CHPs and almost as many single-

Central CHP: 12 areas Decentralized CHP e coal: 3 areas Decentralized CHP e natural gas: 221 areas Decentralized CHP e waste: 13 areas Decentralized CHP e biomass: 39 areas Boiler: 137 areas

The method used for categorizing DH is important, especially the use of primary fuel. The primary fuel is based on the annual consumption for each area, meaning that in some cases the primary fuel can be only 51% of the fuel consumption. Therefore, in theory,

Table 1 Fuel consumption in DH areas in 2008 based on [12]. TWh/year

Coal

Natural gas

Waste

Biomass

Other

Central CHP Decentralized CHP Boiler

42.8 0.8 e

9.4 9.9 0.1

6.1 5.6 0.2

4.0 3.5 2.5

2.8 0.4 0.0

Fig. 2. Single-family houses built from 1940 to 2008 (data from the Danish building register).

26

S. Nielsen, B. Möller / Energy 48 (2012) 23e31

family houses have been built within DH areas with decentralized natural gas CHPs. Assuming construction until 2050 follows the same pattern as the last 29 years, the heat demand for new NZEBs can be found in the following section. 3.3. Net zero energy single-family houses For the new NZEBs, the annual domestic hot water demand is assumed to be 18.3 kWh/m2 and space heating is assumed to be 15 kWh/m2, amounting to a total annual heat demand of 33.3 kWh/ m2 [13]. Based on the 33.3 kWh/m2, the total net heat demand for new NZEBs is found and divided into categories of DH, shown in Table 2. With the assumption that future construction follows the same pattern as the previous 29 years, most new NZEBs will be constructed within central CHP areas, and nearly the same share will be constructed within decentralized CHP areas presently using natural gas. The third largest heat demand will be within the boiler areas, and the remaining part will be built within decentralized CHP areas using coal, waste or biomass. 3.4. Existing buildings within DH areas For the existing buildings within DH areas, all building types are included to give the total heat demand. The existing buildings are assumed to be able to reduce the space heat demand by 50% compared to the present situation according to Refs. [6,14e16]. Table 3 shows the total heat demand for all areas including a 50% heat reduction from the present situation. Compared to the area of new NZEBs, existing buildings comprise the largest area and heat demand, which implies the importance of improving the existing building stock when aiming for a fossil free society. Including all building types also shows that more than half of the total heat demand is within central CHP areas. The total net heat demand is found by adding the heat demand of NZEBs, the heat demand of the existing buildings and the grid heat loss. This gives a total net heat demand of 23.76 TWh. This amount is close to the predictions made in the 2050 Climate Plan by the Danish Engineering Association which forecasts 26.55 TWh [17] and the Heat Plan Denmark 2010 estimates 25 TWh [18]. Both calculations include a 50% space heat demand reduction. 3.5. Size of the typical DH areas within each category To analyze the effects of new NZEBs in DH areas, a representative system is established for each DH type, based on average annual heat demands, shown in Table 4. These average heat demands are used as a basis for the scenarios used in the analyses. The choice to look at typical systems, based on

Table 2 Heat demand for NZEB single-family houses based on 33.3 kWh/m2. Area type

1000 m2

Heat demand (MWh/year)

Central CHP Decentralized Decentralized Decentralized Decentralized Boiler Sum

8,750,790 457,214 7,149,176 2,032,585 1,253,514 3,269,791 22,913,070

291,401 15,225 238,068 67,685 41,742 108,884 763,005

CHP CHP CHP CHP

e e e e

coal natural gas waste biomass

average values, is mainly to make it manageable to model the systems. The optimal solution would be to model each DH area separately; this approach would, however, take a large amount of resources to carry out. Therefore, a future study could be to make analyses of more specific places to see how NZEBs will affect future DH systems. 4. The scenarios Table 5 shows the scenarios modeled in energyPRO. Out of six different types of DH areas, three different sizes of areas are used in the modeling due to the fact that natural gas areas are nearly the same size as biomass areas; thus, both of these are modeled as small decentralized CHP areas. The boiler areas and the larger decentralized areas are also modeled. The decentralized waste CHP areas and the central CHP areas are excluded from the analyses because in the waste areas, solar thermal production would only reduce heat from waste, which would already be incinerated. The reason for excluding central CHP areas is that these areas are so large that they would need vast areas of land for solar thermal collectors and storage. Also these areas are usually more complex and use multiple producers and multiple types of energy sources. The use of solar heat in the large central DH areas should be addressed elsewhere since it needs more in-depth analyses of the concrete areas. In general, solar thermal production from NZEBs is not expected to have a great impact on the large central DH systems unless there are large amounts of waste heat or geothermal heat in the system during summer months. In future energy systems, solar thermal is not the only renewable resource which will be used in DH areas; other technologies like heat pumps and geothermal production will also come into play. The analyses could be improved by making more scenarios which include these technologies. This, however, should be carried out in combination with a more general analysis of the whole country since the result would depend on how much wind production is in the whole system. Geothermal production is very site-specific and can mainly be used in larger DH areas.

Fig. 3. Area of single-family houses built from 1980 to 2009 within six types of DH areas.

S. Nielsen, B. Möller / Energy 48 (2012) 23e31 Table 3 Square meters and heat demand of existing buildings with a 50% heat reduction. Area type

Area 1000 m2

Heat demand MWh/year

Central CHP Decentralized Decentralized Decentralized Decentralized Boiler

156,149 5,738 58,717 25,240 9935 22,561

13,134,027 483,625 4,686,486 2,125,879 781,781 1,786,813

278,340

22,998,610

CHP CHP CHP CHP

e e e e

Coal Natural gas Waste Biomass

27

Table 5 Scenarios used for the modeling. LTS, long-term storage. Boiler Small decentralized CHP Large decentralized CHP A1 Solar thermal A2 Solar thermal þ LTS A3

B1 B2 B3

C1 C2 C3

The next section describes the preconditions and inputs used for the energyPRO modeling.

reduced by 50%. The hourly heat distribution for NZEBs is created in a similar way by modifying the heat distribution for individual houses used in Heat Plan Denmark and by reducing space heat demand with 70%. The indexed hourly distributions are used together with the heat demands in Table 4 to create the heat demand distributions used in energyPRO.

5. Preconditions and model inputs

5.2. Solar thermal

All the inputs in this section, regarding heat demands, are based on the typical DH areas defined in Table 4. As a general choice for the 2050 DH system, the systems are modeled as energy systems with biomass fueled CHPs and/or boilers as their heat supply. In some scenarios these units are the main heat supply during the year, while in other scenarios solar thermal production has a large share of the annual production. In the last type of scenarios, CHPs and boilers can be seen as backup for solar production in the summer season and the main heat supply during winter. As mentioned earlier, other possible heat production units are not included, e.g., geothermal heat, waste heat and heat pumps, which are possible candidates for heat production in future DH areas. The heating value used for biomass is 14.50 GJ/ton, which is the same value as for straw. This is in between wood chips with a heating value of 10.05 GJ/ton and wood pellets with a heating value of 17.5 GJ/ton [19]. In a broader discussion of future energy supply and sustainability, the use of different biomass resources should be debated. This is due to the fact that biomass resources are scarce and can be used for other purposes such as transport. The biomass discussion is important, but not as relevant in the scope of this article where the aim is to minimize the use of combustible fuels like biomass by implementing solar thermal production.

There are two overall ways of utilizing solar thermal production in DH areas. The first is an individual solution, where solar panels are added to the roofs of buildings and the excess heat is exported to the DH grid. The second solution is the collective centralized model, where solar panels are placed together on a field of land. The individual solution has the benefit of using roofs of buildings instead of land, which could be used for other purposes. The collective solution has the benefit of being less investment intensive per square meter and being able to cover a larger share of the annual heat demand by using seasonal storages. From the existing collective solar thermal plants in Denmark, the average annual heat production is found to be 476 kWh/m2 [20]. This average production is used to find the needed aperture area of solar collectors for each DH area type. By dividing the heat demands in Table 4 with 0.476 MWh/m2 the needed areas for solar collectors are found. For NZEBs the area needed to produce 100% of the annual heat demand is used, while for the collective solar thermal the area needed to produce 50% of the annual heat demand is used. The areas needed are shown in Table 6. The first column shows the aperture area needed to supply the annual NZEB heat demand. The second column shows the area needed to produce enough heat to cover 50% of the heat demand in the DH areas including NZEB demand. The largest planned solar collector in Denmark is in the town of Dronninglund and will be 35,000 m2 of aperture area [18]. From Table 6, it is possible to see that the area within the central CHP used for covering 50% of the heat demand is unrealistically large. The decentralized CHPs with coal or waste are also large in size, but more realistic. The last column shows the gross area used for collective solar panels which is 2.2 times the aperture area of the collectors [9]. To model the production from solar thermal, technical data for the collector type is needed. The data used is shown in Table 7. The same type of solar collector is used in all scenarios, both for individual and collective solutions. The inclination of solar collectors is based on an average of the existing collective solar thermal collectors in Denmark, which vary between 33 and 45 degrees inclination [20]. The hourly solar radiation and the temperatures outside determine the production from solar collectors. Monthly averages of both are shown in Fig. 5. Fig. 5 shows that solar radiation and temperature follow nearly the same pattern, with high radiation and temperature in summer months. If the hourly values were shown, these would show that the solar radiation only occur in the daytime. It is also interesting to observe how the temperature decrease is delayed compared to the solar radiation. This phenomenon results in the autumn months and is characterized by low solar radiation and low heat demand.

Sum

5.1. Hourly heat demand According to Section 3.2, the model is set to operate on an hourly basis for a one-year period. To model the hourly heat demand, aggregated demands for both DH and NZEBs are used. The optimal hourly distribution would be measured data on building level for each of the different areas. However, since this kind of data is not available, the hourly distribution used is an adjusted version of the one used in Heat Plan Denmark 2010 [18]. Fig. 4 shows the hourly heat demand distribution over a year for the existing buildings, and how it is modified to a new distribution with lower space heat demand. The hourly heat demand distribution is modified in such a way that while the hot water demand is not reduced, the space heat is

Table 4 Average heat demands in DH areas for 2050 including 15% grid loss. Average heat demands

Existing MWh/year

NZEBs MWh/year

Sum MWh/year

Central CHP Decentralized Decentralized Decentralized Decentralized Boiler

1,258,678 185,389 24,387 188,059 23,053 14,999

24,283 5075 1077 5207 1070 795

1,282,961 190,465 25,464 193,265 24,123 15,794

CHP CHP CHP CHP

e e e e

coal natural gas waste biomass

28

S. Nielsen, B. Möller / Energy 48 (2012) 23e31

Fig. 4. Hourly heat distribution used for heat demands within DH areas.

5.3. Thermal storage There are different types of heat storages used in DH systems; most are used for short-term storage and used within days. However, in Heat Plan Denmark 2010, different types of long-term seasonal storages are described. These long-term storages make it possible to save heat from the summer to use in the winter months, which is very useful in combination with solar thermal heat production. The first storage type presented in Heat Plan Denmark is a traditional heat storage tank, which in this case is an insulated steel tank. The steel tank works well, but is very expensive to establish on a large scale and is also hard to fit into the landscape. An alternative to this is pond storage, which essentially is a big hole in the ground with a cover on top [18]. Other types of seasonal storages exist, e.g. in Germany where borehole and aquifer thermal energy storages are in operation [22]. According to [23], there are benefits of an economy of scale linked to the size of storages in which a larger size results in a smaller ratio between area and volume resulting in a lower relative heat loss. In the modeling, only short-term storage and seasonal-pond storage is used. To find the storage size, 0.2 m3/m2 of solar panels is used for short-term storage and for long-term storage 2.5 m3/m2 of solar collectors. Following the assumptions from [9], this is in the

lower range of long-term storages in, e.g. [24] where the ratio is between 2.94 and 4.97, but is still seen as realistic assumptions. These assumptions result in the storage sizes shown in Table 8. The short-term storages are within limits of the present sizes used for heat storages, where the largest short-term storage in Denmark according to [9] is around 75,000 m3, which is in one of the central CHP areas. The largest long-term solar thermal storage in Denmark is around 70,000 m3, and therefore, the storage size of 3,369,120 m3 in central CHP areas is seen as unrealistic. For decentralized CHP areas with heat from coal and waste, the storage size needed is more than six times as large as the present storages. Measured by today’s standards, these would be quite large storages, but they are still modeled to give an idea of their potential. The rest of the DH area types should be able to establish long-term storages in combination with solar collectors. For both types of storages, a temperature of 90  C in top and  50 C in bottom is used. In future DH areas, the forward and return temperatures could be lowered to improve the overall efficiency, but this has not been done in the present study. For short-term storage, the ambient temperature follows the temperatures shown in Fig. 5. To calculate the heat loss from the two types of storages, different factors are used for short-term and long-term storages. Short-term storages are modeled as insulated steel

Table 6 Square meters of solar thermal capacity for NZEBs and DH.

Table 7 Solar thermal collector data from [21].

Collective aperture Collective NZEB gross area, m2 area, m2 area, m2

Area type Central CHP Decentralized Decentralized Decentralized Decentralized Boiler

CHP CHP CHP CHP

e e e e

51,016 coal 10,662 natural gas 2263 waste 10,938 biomass 2249 1670

1,347,648 200,068 26,748 203,010 25,339 16,590

2,964,826 440,149 58,845 446,621 55,746 36,498

Input

Unit

Number

Latitude (UTM) Inclination of solar collector Start efficiency (h0) Loss coefficient (k0) Loss coefficient (k1) Coefficient Collector temperature (average)

Degree Degree

55 40 0.815 2.205 0.0135 4.000 65

W/(m2  C) W/(m2  C)2 

C

S. Nielsen, B. Möller / Energy 48 (2012) 23e31

29

Fig. 5. Average monthly solar radiation and temperature in Denmark.

tanks with an insulation thickness of 300 mm and thermal conductivity for mineral wool is modeled at 0.0370 W/mK [25]. Seasonal storage is difficult to model in EnergyPRO because the storage in reality uses the ground as insulation and because EnergyPRO needs both a thermal conductivity and an insulation thickness to model heat storages. Instead of modeling without heat loss for seasonal storage, assumptions are made. The thermal conductivity for the pond storage is assumed to be higher than the steel tank 2 W/mK [26] and a 5 m insulation thickness is used. Also an ambient temperature of 8  C is used since most of the storage is underground, which is more stable during the year compared to the ambient air temperature.

Scenarios A2, B2 and C2 are the only ones which do not export more than they import from the DH grid. This does not mean that they do not have as much solar thermal production as the rest of the scenarios, but shows that because of the previously implemented collective solar thermal collectors and not having longterm storage, there is not any need in the DH systems for the excess production from NZEBs. All the scenarios follow the NZEB definition, but all the number-2 scenarios produce heat which is not needed in the DH grid. A3, B3 and C3, however, show that implementing a seasonal storage would make it possible to use the excess production from NZEBs. To look at the whole system, and not only the relationship between NZEBs and DH, the heat production in all scenarios is shown in Fig. 7.

5.4. Boiler and CHPs For the boilers, the capacity is set to cover the peak hour demand plus 20%. All the boilers have a heat efficiency of 90%. For CHPs the capacity is set to cover 95% of the annual heat demand. The CHPs have an electric efficiency of 46.6% and a heat efficiency of 41% [27]. 6. Results A key issue, in regard to the results, is prioritization between solar production from NZEBs and collective solar collectors. In the EnergyPRO model, NZEBs are modeled as one area with solar thermal production, while DH areas are modeled as another area with solar thermal production. These areas have different demands and different production capacities. In NZEBs, the local solar thermal production has the highest priority, and if this production exceeds the demand of the NZEBs, it is exported to the DH area. The collective solar thermal collector mainly produces heat for the DH area, and when the production exceeds the demand, the heat is delivered to the NZEBs, if needed. In the case that the demand is met by solar production in both places, the heat will be stored. The first result is to assure that the NZEBs are within the definition of producing as much as they import on an annual basis, as shown in Fig. 6.

Fig. 6. Heat import/export for NZEBs in each scenario.

Table 8 Total storage size for each DH area in m3. Area type Central CHP Decentralized Decentralized Decentralized Decentralized Boiler

CHP CHP CHP CHP

e e e e

coal natural gas waste biomass

Short-term storage, m3

Long-term storage, m3

10,203.13 2132 453 2188 450 334

3,369,120 500,169 66,870 507,524 63,348 41,475

Fig. 7. Heat production in all scenarios shown as percentage of total.

30

S. Nielsen, B. Möller / Energy 48 (2012) 23e31

Fig. 8. Specific fuel consumption in all scenarios.

Fig. 9. Content in seasonal storage from April to December in all scenarios.

Fig. 7 shows the heat production in both DH and for NZEBs. To make all scenarios comparable, the figure shows the percentage share of heat production instead of specific values. A1 shows a case where 95% of the DH demand is covered by boilers and the remaining share is the excess solar production from NZEBs. NZEBs get a bit more than 50% of their heat from solar production. A2, where collective solar thermal collectors are added to DH, shows another situation. In this case, the collective solar collectors produce around one third of the DH demand, meaning that a smaller share of the heat production from NZEBs can be utilized in the DH area. The NZEBs also receive a part of the collective solar production, showing that their own solar production is not needed in this system. The third case, A3, where a seasonal storage is added, shows that more of the solar energy can be used in the system, saving boiler production. Half of the annual heat production in the DH area and the NZEBs comes from solar energy. It is noticeable that with long-term storage, a storage heat loss appears. Since the heat loss is relative to the surface area of the tank, the loss decreases when the storage size increases. Fig. 8 shows the specific fuel consumption in the scenarios. It is clear that the biomass consumption decreases when adding solar thermal production and decreases even further when adding a seasonal storage. From A1 to A3, a reduction of approximately 40% in fuel consumption is reached. Meanwhile from B1 to B3 and C1 to C3 fuel consumption is reduced 30%. The reason that the reduction in scenarios B and C is smaller is because these produce electricity as well. Another result is to examine when the seasonal storage is utilized most, as shown in Fig. 9. In Fig. 9, all scenarios start using the storages in the beginning of June and empty them in the beginning of November. Over this period,

or half the year, heat is supplied solely by solar thermal collectors, providing a large saving in fuels for boiler or CHP production. The seasonal storage is emptied before winter due to the way the system is simulated, where the heat demand as the first priority is covered by solar thermal production or stored thermal heat and only as a last priority the biomass units are run as backup units. In reality, the heat could be saved longer, and in some cases, the CHPs could produce the heat storage depending on fuel prices, electricity prices and related factors, which they are not allowed to be included to in this simulation. Also many DH areas combine their seasonal heat storage with heat pumps, to extend the period where the solar heat can be used. 7. Conclusion The goal of this article has been to find out how the excess heat production from NZEBs might affect the heat production in different types of DH areas in Denmark. The first step was to make a prognosis for 2050, to decide the potential area of new buildings and existing buildings. After finding these areas, the heat demands were found for both NZEBs and the existing buildings. This first step is a very important part of the analysis since it quantifies the heat demands and the geographic placement of NZEBs. An important assumption in this regard is that all new buildings within DH areas are assumed to connect to DH grids. In reality, the buildings are not obligated to be connected, but the assumption is important when examining the possible maximal effect of NZEBs in relation to future DH grids. Afterwards, all the DH areas were grouped into different types of DH areas, depending on size, existing technology and primary fuel use.

S. Nielsen, B. Möller / Energy 48 (2012) 23e31

The second step in the analysis has been to define typical DH sizes for each DH group within Denmark. Within these typical systems, the ones where it was reasonable to add solar thermal production were chosen. This amounted to three different types of systems that mainly differ in the annual heat load of the systems, which are 15,731 MWh, 25,362 MWh and 192,499 MWh, respectively. In each of these systems, biomass has been used as the fuel in boilers and CHPs to simulate how DH systems may be in 2050. Also to make scenarios of different future system, three different setups have been made. The first having only solar thermal on the roof of NZEBs, the second adding a collective solar thermal collector to the DH areas and the third adding seasonal heat storage to the DH areas. The results mainly show two things. If the DH system does not have large amounts of solar thermal production, the NZEBs excess production is useful in the DH systems because it reduce the biomass consumption in the DH areas. Even though biomass is often considered a renewable energy source, it still has the disadvantage of having a cost per unit which changes over time. Since biomass is a resource which can be expected to be used increasingly, a reduction in the use of biomass should be positive for the DH areas. Even in areas where resources other than biomass are used, the solar thermal production would be useful since it would decrease the use of that limited resource. The only exceptions to this would be if there is a significant use of industrial waste heat or heat produced on waste incineration. If the goal is to utilize these resources, heat production from NZEBs could be in conflict with this goal. The second main result is that adding NZEBs to DH systems where collective solar thermal already covers the heat demand during summer, requires seasonal storage for increasing the share of usable solar thermal production from NZEBs. These seasonal heat storages could also be a solution in areas with industrial waste heat, waste incineration or geothermal heat. There are still many aspects that need to be addressed before a general conclusion on excess heat production from buildings can be made. Concrete analyses on actual heating systems could help strengthening the conclusions. Since this article only focuses on the usefulness of excess heat from NZEBs in DH systems, analyses of the economic costs related to the implementation of solar thermal are also necessary. The storages and solar thermal technologies might not be economically feasible compared to other solutions. One could expect that there might be a lower limit to the size of a solar thermal connected to DH due to the fact that the connection to the DH would cost nearly the same, while the cost of the solar thermal panel would increase relative to the size. In addition, there might be issues linked to an economy of scale when deciding between large-scale collective solar thermal and individual small-scale solar thermal. Another topic related to economy is that while solar thermal production might benefit society as a whole, considering the benefits for homeowners and DH operators are just as important. Additional considerations could also be land-use issues since the collective solar thermal is often placed on land that could be used for other purposes. The individual solar thermal has the benefit of being placed on roofs, which results in no need for land. However, the individual solution can compete with the area for photovoltaic production. Therefore, a more general question about what the optimal use of roof areas in the future, must be answered. Also, real-life examples of individual buildings producing heat to DH grids are essential to assess the benefits and costs of the ideas put forward in this article. Such examples would also benefit from gaining knowledge of the needs for monitoring the production, extreme low grid operating temperatures and different payment schemes all of which are important aspects when implementing individual production in DH systems.

31

References [1] Marszal AJ, Heiselberg P, Bourrelle JS, Musall E, Voss K, Sartori I, et al. Zero energy building - a review of definitions and calculation methodologies. Energy and Buildings 2011;43(4):971e9. doi:10.1016/j.enbuild.2010.12.022, http://www.sciencedirect.com/science/article/B6V2V-51SPS0R-8/2/ d1f42a788ae9424ed98916ee849ccf01. [2] Heiselberg P. Nzeb definition. Tech. Rep.; 2011 [not published yet]. [3] Sartori I, Napolitano A, Marszal A, Pless S, Torcellini P, Voss K. Criteria for definition of net zero energy buildings. EuroSun, ISBN 9783901425134; 2010. [4] Lund H, Marszal A, Heiselberg P. Zero energy buildings and mismatch compensation factors. Energy and Buildings 2011;43(7):1646e54. doi:10. 1016/j.enbuild.2011.03.006, http://www.sciencedirect.com/science/article/ pii/S0378778811000843. [5] Danish Energy Authority. Energy statistics 2009. Copenhagen: Danish Energy Authority, ISBN 9788778448729; 2009. [6] Moeller B, Lund H. Conversion of individual natural gas to district heating: geographical studies of supply costs and consequences for the Danish energy system. Applied Energy 2010;87(6):1846e57. doi:10.1016/j.apenergy.2009. 12.001, http://www.sciencedirect.com/science/article/B6V1T-4Y3K186-1/2/ 5d6d11634a513526a305996c54d0e00f. [7] Lund H, Mller B, Mathiesen BV, Dyrelund A. The role of district heating in future renewable energy systems. Energy 2010;35(3):1381e90. [8] Illum K, Moeller B. Energipolitikkens teknologiske råderum. Tech. Rep.; 1998. [9] Dyrelund A, Lund H, Mller B, Mathiesen BV. Heat plan denmark, varmeplan danmark [in Danish]. Tech. Rep.; 2008. [10] Moeller B. A heat atlas for demand and supply management in denmark. Management of Environmental Quality 2008;19(4):467e79. Cited By (since 1996): 2. [11] EMD International. Energypro website; 2011. [12] The Danish Energy Agency. Energy producer count 2009. Tech. Rep.; 2010. [13] Active house project. Active house project; n.d. Available from: http://www. activehouse.info/. [14] Wittchen KB, Kragh J. Heat demand in Danish buildings in 2050. Tech. Rep.; SBI forlag; 2010. [15] Mathiesen BV, Lund H, Karlsson K. 100% Renewable energy systems, climate mitigation and economic growth. Applied Energy 2011;88(2):488e501. doi: 10.1016/j.apenergy.2010.03.001. The 5th Dubrovnik Conference on Sustainable Development of Energy, Water and Environment Systems, held in Dubrovnik September/October 2009, http://www.sciencedirect.com/science/ article/B6V1T-4YRGJJ7-1/2/9d0323c437007c60f462f3d67d76be00. [16] Sperling K, Moeller B. End-use energy savings and district heating expansion in a local renewable energy system a short-term perspective. Applied Energy 2012;92:831e42. doi:10.1016/j.apenergy.2011.08.040, http:// www.sciencedirect.com/science/article/pii/S0306261911005502. [17] Danish Engineer’s Association. The IDA climate plan 2050 e background report. Tech. Rep.; 2009. Available from: http://ida.dk/News/Dagsordener/ Klima/Klimaplan2050/Documents/Baggrundsrapport-IDAs klimaplan.pdf. [18] Ramboll Denmark and Aalborg University. Heat plan Denmark 2010: main report; 2010. [19] The Danish Energy Authority. Basic conditions for socio-economic analyses within the energy area e April 2011 (Forudsaetninger for samfundsoekonomiske analyser paa energiomraadet e April 2011). Energistyrelsen; 2011 (in Danish). [20] Danish District Heating Association, ARCON solar, BATEC solar, PlanEnergi, Raboll, Sunmark, et al., Solvarmedata; 2011. [21] ARCON solar. Data sheet ht-sa 28/10; 2010. [22] Bauer D, Marx R, Nussbicker-Lux J, Ochs F, Heidemann W, MullerSteinhagen H. German central solar heating plants with seasonal heat storage. Solar Energy 2010;84(4):612e23. doi:10.1016/j.solener.2009.05.013. International Conference CISBAT 2007, http://www.sciencedirect.com/science/ article/B6V50-4WM8WNP-1/2/b6ae9d38860e44117ff8e3c7f3fac3ec. [23] Pinel P, Cruickshank CA, Beausoleil-Morrison I, Wills A. A review of available methods for seasonal storage of solar thermal energy in residential applications. Renewable and Sustainable Energy Reviews 2011;15(7):3341e59. doi: 10.1016/j.rser.2011.04.013, http://www.sciencedirect.com/science/article/pii/ S136403211100150X. [24] Novo AV, Bayon JR, Castro-Fresno D, Rodriguez-Hernandez J. Review of seasonal heat storage in large basins: water tanks and gravel-water pits. Applied Energy 2010;87(2):390e7. doi:10.1016/j.apenergy.2009.06.033, http://www.sciencedirect.com/science/article/pii/S0306261909002694. [25] Esbensen TV, Korsgaard V. Dimensioning of the solar heating system in the zero energy house in denmark. Solar Energy 1977;19(2):195e9. doi:10.1016/ 0038-092X(77)90058-5, http://www.sciencedirect.com/science/article/pii/ 0038092X77900585. [26] Raab S, Mangold D, Mller-Steinhagen H. Validation of a computer model for solar assisted district heating systems with seasonal hot water heat store. Solar Energy 2005;79(5):531e43. doi:10.1016/j.solener.2004.10.014, http:// www.sciencedirect.com/science/article/pii/S0038092X05000289. [27] Danish Engineer’s Association. The IDA climate plan 2050. Tech. Rep.; 2009. Available from: http://ida.dk/News/Dagsordener/Klima/Documents/ IDA-Klima Hovedrapport DK.pdf.