Horizontal Building Interaction as an Element of Neighborhood Energy-Oriented Refurbishment

Abstract

The building sector has a significant impact on the environment due to its high resource and energy usage. The refurbishment of the building stock is a measure for reducing emissions. In this context, the neighborhood scale is becoming increasingly important as the level at which urban redevelopment takes place. This study contributes a new perspective and data on the scientific debate on the importance of the neighborhood as a level of action in the transformation of the building sector. It combines horizontal building interaction and a practical refurbishment approach, aiming to reduce material use and balance energy demands. Using scenario modeling, the material savings are calculated for the first time by analyzing five refurbishment scenarios of a synthetic neighborhood. The scenario, modeled with horizontal building interaction, is identified as the favorable compromise among all scenarios when considering material demand and energy efficiency. This is achieved through re-thinking energy-oriented refurbishments and optimizing the usage of locally produced renewable energy sources. The results are embedded into the scientific debate, including the works on the balance of embodied and operational energy in the construction sector.

1. Introduction

The construction sector is one of the main drivers of human-induced climate change. Emissions generated throughout the life cycle of buildings, consisting of the phases described in DIN 15978, currently account for 5–12% of total national greenhouse gas (GHG) emissions; 50% of all extracted materials are required in the building sector, and 35% of the European Union (EU)’s total waste is generated throughout this sector [1]. At the European level, GHG emissions in the construction sector are to be reduced to 55% by 2030 and 90% by 2040 compared to 1990 [2]. The levels of action to achieve these targets to date are mostly at the macro-level of (trans-)national policy making and the micro-level of practical implementation on the building scale. The already-implemented policies and measures will reduce GHG emissions from buildings to a level of 42% below 1990 levels by 2030 and to a level of 53% by 2040 [2]. The need for more far-reaching measures is therefore given, encompassing an extension of the level of action by the intermediate scale of neighborhoods.

The neighborhood has been increasingly given more consideration for several years as a level of action but is yet to be fully established [3]. To accelerate the transformation of the building sector, the level of action is shifting from the micro- to the meso-level of neighborhoods. The refurbishment of neighborhoods is a central element of this transformation. When analyzing the environmental impacts of a neighborhood energy efficiency refurbishment, buildings are addressed in a ‘building-by-building’ approach, summing the individual environmental flows of all constituting building assets [4]. This approach reflects the practice of developing and applying refurbishment measures of buildings in a neighborhood on the building scale. However, even the measures taken to reach the highest energy standards on building levels are not sufficient to achieve the climate-neutrality targets set in the building sector [5].

The concept of horizontal interaction picks up on this issue. Neighborhoods are embedded in a hierarchical system of the built environment, connected by horizontal and vertical interaction. Vertical interaction describes the integration of a unit into its hierarchical system through connection to units of a higher or lower order. Horizontal interaction refers to the connection of units of the same order in this system. In relation to the buildings of a neighborhood, vertical interaction is established, e.g., through connection to the external power grid. Horizontal interaction is still finding its way into practical implementation under various names and with different implementations. ‘Peer-to-peer’ markets, ‘local grids’ and ‘community grids’ are implementations of horizontal building interactions in practical concepts such as ‘Energiegemeinschaften’ in Austria or ‘Vor-Ort-Systeme’ in Germany [6,7,8]. Furthermore, the horizontal interaction within a neighborhood is part of the concept of Positive Energy Districts [9,10]. The EU enables horizontal building interaction throughout the policy framework of the Renewable Energy Directive (RED II) 2018/2001 and Directive 2019/944. In Germany, a transfer to national policy frameworks has yet to be performed. The overall objective of horizontal building interaction is to enable leveraging interactions among various buildings and to optimize the integration of renewable energy sources [11]. This interaction is often established but not limited to one building block. The goal is to increase the rate of self-consumption of locally generated renewable energy. Currently, studies dealing with horizontal interaction focus only on its energetic aspects [12,13,14,15]. However, carrying out energy-oriented refurbishments is connected to material demands and waste generation. Combining the energetic changes and material demands resulting from neighborhood energy-oriented refurbishment is identified as a research need [11,14].

This study meets this need by modeling five pathways of energy-oriented refurbishment of a synthetic neighborhood, applying a scenario modeling approach. The scenarios are analyzed, quantifying their respective material demands, production of waste and resulting final energy demands. The rates of self-consumption of locally generated energy from PV systems are varied by applying the concept of horizontal building interaction. Results show that the horizontal building interaction increases the degree of energy self-sufficiency in the neighborhood compared to non-interacting scenarios. This effect is intensified when combining horizontal building interaction with an adapted refurbishment approach. Further findings are that the material demand is decreased by up to 87% when a heterogeneous energy performance level prevails in the neighborhood instead of aiming for a homogeneous high-quality energy performance level.

This study is the first to identify and quantify the potential of horizontal building interaction as a balancing link between energy efficiency and material demand in energy-oriented refurbishments of neighborhoods. As such, it contributes a valuable new perspective and data on the scientific debate on the importance of the neighborhood as a level of action in the transformation of the building sector.

2. Materials and Methods

The neighborhood level is the typical operational scale used in urban planning, planning of energy provision and transport projects. The implementation of measures to achieve emission reduction targets at this scale is therefore plausible [16]. A neighborhood consists of four physical elements of the built environment: buildings, open spaces, networks and mobility [16,17]. As a meso-level between the building and (trans-)national level, the neighborhood offers the opportunity to combine several buildings within a single measure while responding to local characteristics [18]. Looking at the understanding of the term ‘neighborhood’, there is no comprehensive definition. This is due to the complexity of the neighborhood structure and the multitude of perspectives from which it is examined [19]. In this study, I use the term ‘neighborhood’ to describe the coexistence of more than one building within a local community.

I defined a synthetic neighborhood, composed of two residential building archetypes on different energy performance levels. Applying a scenario modeling approach, I developed refurbishment pathways for this synthetic neighborhood. I defined three main pathways, two with two sub-pathways each, resulting in a total of five refurbishment scenarios. I analyzed the scenarios from two perspectives: material and energy. The material perspective focuses on materials used and waste generated in the process of refurbishment. The energy perspective focuses on the resulting changes in energy demands of the neighborhood throughout the refurbishment. I identified the potential of the combination of horizontal building interaction and an adapted energy-oriented refurbishment approach aiming for a reduction of material input. I quantified material savings and balanced energy demands through maximized self-consumption rates in the neighborhood.

The Institut Wohnen und Umwelt GmbH (IWU) offers a building typology and the TABLUA WebTool as results of the framework of Intelligent Energy Europe projects TABULA and EPISCOPE. They give concrete data on residential building stocks in age classes: data on exemplary buildings, along with their commonly found construction elements and corresponding U-values. Additionally, data on exemplary heat supply systems for old buildings and energy-saving measures on two quality levels and their impact on the energy consumption are defined. From these data, two residential-building archetypes, representative of the German building stock, are selected to assemble a synthetic neighborhood. The two residential building archetypes defined can exist on three levels of energy performance. The different levels of energy performance defined by IWU are applied to reflect on the heterogeneity of energy efficiency performance levels of residential buildings in the building stock. Horizontal building interaction is, in principle, applicable to this neighborhood.

Developing the different refurbishment scenarios, the method of scenario modeling was conducted. Three main pathways of neighborhood refurbishment are discussed. The goal when working with scenarios was to understand the dynamics of changing energy and material demands in the neighborhood when applying different refurbishment approaches. Each pathway has a different set goal for the refurbishment. Following the first pathway, only minimum legal requirements in the process of refurbishment are matched to keep the short-term expenses as low as possible and use less materials. The second pathway exceeds the minimum requirements, reaching for a high energy performance level for every building, with a focus on energy savings in the use phase. The third pathway combines the two preceded ideas and is modeled as a middle course, minimizing material inputs and still saving energy in the use phase. This is reached by applying the concept of horizontal building interaction while aiming for a reduction in material input during the refurbishment. The first and second pathways are subdivided into two sub-pathways each: one with and one without an additional horizontal building interaction. The scenarios allow us to compare the effects of a stand-alone horizontal interaction of buildings and the effects of combining it with the effort of saving materials in the process of refurbishment.

To make assumptions as practical as possible in the scenario modeling, a selective review of the research literature was conducted. The calculations were made using Office 365 Excel. They were first divided into material and energy calculations for each building archetype, and then in the second step, they were combined in the scenario modeling of the neighborhood. The amount of construction material was subdivided into the six most mass-relevant materials installed into buildings during the refurbishment. For both building archetypes, the materials for the two refurbishment scenarios were calculated, matching the measures of the two performance quality levels defined by IWU. For the scenario analysis, the energy calculations were subdivided further into the pathways of implemented or not-implemented horizontal building interaction. For each pathway, the final energy demand and its coverage through withdrawal from the external grid and the use of self-generated electricity were calculated. The amount of electricity fed into the external grid was calculated by making assumptions about the self-consumption rate of the produced renewable energy in the neighborhood.

2.1. Selection of Residential Building Archetypes

The IWU building typology and the associated TABULA WebTool serve as the basis for the data used in this study. The ‘Building Typologies’ section of the WebTool provides a table of typical construction types of residential buildings by construction year classes for various countries. The data for Germany, for the construction year class ‘1958 … 1968’, were used for the calculations. These currently account for the largest share of the German residential building stock and are therefore representative. In the WebTool, within this construction year class, a distinction is made between the following building types: Single-Family House, Terraced House, Multi-Family House and Apartment Block. The dataset for Single-Family House (DE.N.SFH.05.Gen) (SFH) and for Multi-Family House (DE.N.MFH.05.Gen) (MFH) is used for the calculation. These are generic data, as indicated by the ‘.Gen’ in the designation. In the TABLUA WebTool, the German reference climate was used for any calculations of energy indicators according to DIN V 18599-10. Parameters influencing the indicators, such as the orientation of buildings and surrounding climate, including wind or solar radiation, are measured in Potsdam (Germany). The SFH has a heated living area of 110 m², a heated pitched roof, concrete ceilings, masonry made of hollow blocks, lattice bricks, wood-chip bricks or similar, and then it is plastered. The MFH has a heated living space of 2845 m² and 32 flats on 4 full stories. It has a pitched or flat roof, which is sometimes heated. As with the SFH, the masonry is made of hollow blocks, lattice bricks, wood chip bricks or similar, and the façade is plastered. Ceilings are made of reinforced concrete; there are strong thermal bridges on cantilevered balconies. The construction methods, along with the corresponding U-values and the existing heat supply system in the initial state, are summarized in

Table 1. Data for Single-Family House (SFH_E) and Multi-Family House (MFH_E), representative for constructed buildings from 1958 until 1968 in Germany, initial state of construction und heat supply system.

	Single-Family House		Multi-Family House
Construction	Description	U-value (W/(m2K))	Description	U-value (W/(m2K))
Roof/top story ceiling	Pitched roof with 5 cm insulation	0.8	Concrete ceiling with 5 cm insulation	0.6
Exterior wall	Masonry	1.2	Masonry	1.2
Windows	Wooden windows with double pane insulating glazing	2.8	PVC-U windows with double-pane insulating glazing	3.0
Flooring	Concrete ceiling with 1 cm insulation	1.6	Concrete ceiling with 1 cm insulation	1.6
System technology	Description	Energy required for 1 kWh of heat	Description	Energy required for 1 kWh of heat
Heating system	Gas boiler, low efficiency	1.38 kWh gas	Gas boiler, low efficiency	1.21 kWh gas

.

The SFH has a 168.9 m² roof surface, 141.2 m² surface of exterior walls, 115.8 m² floor surface and 27.1 m² window surface. The MFH has a 971 m² roof surface, 2039 m² surface of exterior walls, 971 m² floor surface and 507.5 m² window surface.

2.2. Definition of Building and Neighborhood Refurbishment Scenarios

Based on this initial state (SFH_E; MFH_E) of the buildings, the IWU building typology presents two alternative packages of measures for the energy modernization of the building envelope and the system technology.

Package 1 (SFH_MP1; MFH_MP1) is based on the requirements of Energieeinsparverordnung (EnEV) 2009. The standard of EnEV 2009 is used due to the data available in the IWU building typology and WebTool. In 2013, the EnEV was updated, and in 2020, it was replaced by the Gebäudeenergiegesetz (GEG). However, the U-values, which are the basic requirements from the EnEV 2009 used in this scenario definition, have remained the same in the process. As such, the information from the IWU building typology is still current. Regarding the SFH, Package 1 includes the following: removal of the old insulation in the attic and full insulation of the rafter cavity (12 cm), and insulation of the exterior walls with a 12 cm thick external thermal-insulation composite system. The windows are replaced with 2-pane thermal insulation glazing in wooden frames. Insulation boards with a thickness of 8 cm are laid under the cellar ceiling. The existing gas heating system is replaced with an electric heat pump. It is assumed that the heat pump operates with an annual coefficient of performance of 3 and that auxiliary energy of 0.3 kWh of electricity is required per kWh of heat generated. A photovoltaic system can be installed on the roof to supply energy for the operation of the heat pump. For the SFH, the resulting U-values after carried out MP1 are 0.41 W/(m²K) for the roof, 0.23 W/(m²K) for the external walls, 0.34 W/(m²K) for the floor and 1.3 W/(m²K) for the windows. For the MFH, the resulting U-values are 0.19 W/(m²K) for the roof, 0.23 W/(m²K) for the external walls, 0.31 W/(m²K) for the floor and 1.3 W/(m²K) for the windows.

Package 2 (SFH_MP2; MFH_MP2) is described as ‘forward-looking’. It provides significantly improved thermal insulation, consisting of an insulation layer of 30 cm on the roof, 24 cm on the exterior walls and 12 cm on the basement ceiling. The windows are replaced by 3-pane thermal insulation glazing in insulated frames. An electric heat pump is installed with an annual coefficient of performance of 4, requiring 0.25 kWh of auxiliary energy per kWh of heat generated. A photovoltaic system can be installed on the roof to supply energy for the operation of the heat pump. For the SFH, the resulting U-values after carrying out MP2 are 0.14 W/(m²K) for the roof, 0.13 W/(m²K) for the external walls, 0.25 W/(m²K) for the floor and 0.8 W/(m²K) for the windows. For the MFH, the resulting U-values are 0.09 W/(m²K) for the roof, 0.13 W/(m²K) for the external walls, 0.23 W/(m²K) for the floor and 0.8 W/(m²K) for the windows.

The assumptions made about the initial state of the buildings, as well as the modernization packages, are used to calculate the material demands and waste generated throughout the implementation of these measures. For the refurbishment of the roof, exterior walls and floor, the only material category included in the calculations is insulation material. Additional materials used in an actual refurbishment, such as color and plaster, are omitted because of their relatively small amounts. The mass, $m$, of insulation needed in the refurbishment is calculated as follows:

$$m=A\ast t\ast \rho$$

(1)

where $A$ is the surface of the component, $t$ is the thickness and $\rho$ is the density of the material used.

I assumed that the material used in the insulation is EPS, with ${\mathsf{\rho }}_{EPS}=15{\displaystyle \frac{kg}{m^3}}$.

For the calculation of the inventory data of the windows, I made the following assumptions based on [20]. For the initial state and MP1, I assumed that the wooden frames would use the inventory data from [20]. The data is based on the data from the software ‘GaBi 4’ [21]. For MP2, I assume that window frames made from PVC from the same literature resource are used. For a better overview, I summarize the detailed material inventory to the categories of glass, aluminum, wood, plastic materials and sheet steel in

Table 2. Material inventory of windows per state of building (_E, _MP1 and _MP2).

	Initial (_E)	_MP1	_MP2
	(kg/m2 window surface)
Glass	12.6	14.3	21.4
Plastics	0.6	0.6	8.4
Sheet steel	0	0	7.3
Aluminum	0.3	0.3	0
Wood	14.8	14.8	0

. The data are provided in mass per 1 m² window surface.

To calculate the material masses of the building archetypes’ windows, I multiplied the provided data by the respective surface of the two archetypes. The final energy consumption ($FEC$) of the building archetypes per state is calculated by multiplying the heated floor area of the respective archetype by the provided final energy consumption, both provided by the IWU building typology, summarized in

Table 3. Heated floor area and final energy consumption per building archetype (SFH and MFH) and state (_E, _MP1 and _MP2).

		SFH _E	SFH _MP1	SFH _MP2	MFH _E	MFH _MP1	MFH _MP2
Heated floor area	(m2)	2844.6			110.2
Final energy consumption $(FEC$)	(kWh/(m2a))	195	81	38	141	117	98

.

For the demand of household electricity ($ET$), it is assumed that every inhabitant of the buildings modeled demands 800 kWh electricity/year. The SFHs are modeled as 4-person households, and the MFH contains 32 flats, which leads to a total number of 96 inhabitants, assuming an average of three inhabitants per flat. The initial state electricity is only demanded to cover the household electricity. For the state MP1 and MP2, additionally, the final energy consumption is covered by electricity through the installation of a heat pump. To calculate the energy needed to operate the heat pump (${FEC}_{el}$) to cover this final energy consumption ($FEC$), the annual coefficient of performance ($COP$) is used:

$${FEC}_{el}={\displaystyle \frac{FEC}{COP}}$$

(2)

Also, the demand for auxiliary energy ($AEC$) to operate the heat pump is calculated as follows:

$$AEC={\displaystyle \frac{{FEC}_{el}}{COP}},$$

(3)

The total electric energy demand (${TEC}_{el}$) to cover the final energy demand and the energy demand to operate the heat pump is calculated as

$${TEC}_{el}=ET+{FEC}_{el}+AEC.$$

(4)

For the calculation of energy provided by the PV system (${EP}_{PV}$), it is assumed that a 5 m² roof surface is needed to provide 1 kWp and that only 30% of the roof surface, $A_{roof}$, is suitable for installing PV systems, considering shading and windows. This leads to the equation

$${EP}_{PV}=0.3\ast A_{roof}\ast 5{\displaystyle \frac{m^2}{kWp}}\ast 1000{\displaystyle \frac{kWp\ast a}{kWh}}.$$

(5)

Based on the modernization packages MP1 and MP2 described for the two building archetypes, a neighborhood is defined. The neighborhood consists of nine buildings, eight SFHs and one MFH. In the initial state of the neighborhood, four SFHs are at the initial state (SFH_E); MP1 has already been implemented on two SFHs, and MP2 on two others. The MFH has also already been upgraded to a higher energy level through MP1. This mix was determined to reflect on the heterogeneity of the energy performance of the German residential building stock. The final energy demand in the initial state totals around 445,000 kWh/a. One SFH_MP1 and both SFH_MP2s have already installed PV systems, with each producing 10,000 kWh/a electric energy. The remaining roofs of the SFH are less suitable for PV systems because of orientation and shading. Installing PV systems would result in the production of 6000 kWh/a per building. The MFH is not suitable for PV systems. Every building with an installed PV system has a battery storage that has a capacity depending on its demand. The capacity installed is 1 kWh per 1000 kWh/a electricity demand.

For the refurbishment scenarios, there are three basis scenarios defined: Business-as-Usual (BAU_I), Forward-Looking (FL_I) and Community.

In the Business-as-Usual (BAU_I) scenario, it is assumed that the minimum legal requirements are implemented. This requirement stipulates that all buildings must be brought up to the minimum standard according to EnEV 2009. Therefore, the four SFHs that are currently still in a completely unrenovated state (SFH_E) are renovated by applying MP1. After the refurbishment, all buildings will be equipped with heat pumps and meet the minimum legal standards for the condition of their building envelope. No additional PV systems are installed. The self-consumption rate of locally generated energy is assumed to be 30%.

In the Forward-Looking (FL_I) scenario, it is assumed that all buildings are to be upgraded to the forward-looking energy performance level. MP2 is carried out. This applies to all buildings on which this package has not yet been carried out: the four SFHs in the initial state, as well as the two SFHs and the MFH on which MP1 has already been carried out. For those buildings on which MP1 has already been carried out, the installed insulations are replaced by new ones. After the refurbishment, all buildings are equipped with heat pumps and exceed the minimum legal standards for the condition of their building envelope. Every SFH is equipped with a PV system, with production capacities being 10.000 kWh/year or 6000 kWh/year. The self-consumption rate of locally generated energy is assumed to be 30%.

Neither the scenario BAU_I nor the FL_I model the interaction of buildings; the premise of these scenarios is to bring the building stock to a homogeneous energy level. The remaining three scenarios, namely the Community, Business-as-Usual_EnergyCommunity (BAU_EG) and Forward-Looking_EnergyCommunity (FL_EG) scenarios, are modeled as building interaction scenarios. The buildings in these scenarios are connected to each other by horizontal interaction within the neighborhood. They form an Energy Community, optimizing the utilization of the locally generated electrical energy from the installed PV modules among themselves. The goal is to increase the self-consumption rate within the neighborhood by leveraging interactions among buildings using local storage capacities.

For the BAU_EG scenario, the resulting self-consumption rate is assumed to be 35%, and for the FL_EG scenario, 40%. The difference is caused by the assumption that the technology installed (controlled heat pump and battery storage) in the FL scenarios is more suitable for the technological complexity that comes with the interaction. This is appropriate because the main assumption of the FL scenarios’ design is the exceedance of minimum standards, which also apply for the technology of heat pumps.

The Community scenario is modeled as the fifth scenario. In this scenario, the neighborhood and its refurbishment are thought of as a holistic approach, uniting the demands of embodied energy in the process of refurbishment and the operational energy in the following use phase. The guiding principle of the scenario is the reduction of material input while applying the State-of-the-Art practical approaches of building and neighborhood refurbishments. The applied practical approaches are prioritizing refurbishments of buildings with high energy demand [22,23]; if a building undergoes a refurbishment, forward-looking refurbishment measures are applied, and the buildings of a neighborhood are connected via horizontal interaction to balance the energy demands. This transfers to the structural measure of refurbishing the four SFH_Es because their energy demands are particularly high in comparison to the other buildings in the neighborhood. This will be achieved by applying MP2, matching the practical approaches. Looking at the energy efficiency performances, this results in a remaining heterogeneous building stock limited to two energy performance levels instead of three before the refurbishment. No additional PV systems are installed because all roofs with high solar potential are already equipped with PV systems. The neighborhood has a battery storage capacity of 40 kWh. Horizontal building interaction is enabled throughout the whole neighborhood. The self-consumption rate is assumed to rise to 50%, assuming that the technical infrastructure in this scenario is optimized regarding the self-consumption rate, e.g., by integrating neighborhood mobility concepts with charging infrastructure for electromobility.

All scenarios, as well as the initial state, distinguished by their composition of energy performance levels on which the buildings exist, are summarized in Figure 1.

To calculate the amount of energy self-consumed by the buildings ($SC$), the amount of photovoltaic electricity produced (${EP}_{PV}$) is multiplied by the self-consumption rates ($SCR$) assumed:

$$SC={EP}_{PV}\ast SCR.$$

(6)

To obtain the amount of energy withdrawn from the external grid ($EGW$), the amount of energy, $SC$, is subtracted from the calculated energy demand, ${TEC}_{el}$:

$$EGW={TEC}_{el}-SC,$$

(7)

The amount of energy fed to the external grid ($EGF$) is calculated by subtracting $SC$ from the ${TEC}_{el}$:

$$EGF={TEC}_{el}-SC$$

(8)

3. Results

3.1. Amounts of Materials

The total amounts of built-in materials in the three different states of the buildings (_E, _MP1 and _MP2) are summarized in

Table 4. Selected built-in material categories per building archetype (SFH and MFH) and state (_E, _MP1 and _MP2).

	SFH _E	MFH _E	SFH _MP1	MFH _MP1	SFH _MP2	MFH _MP2
	(kg)
EPS	144	873	697	6583	1477	13,458
Glass	341	6381	387	7240	580	10,861
Plastics	17	327	17	327	227	4243
Sheet steel	0	0	0	0	197	3695
Aluminum	8	152	8	152	0	0
Wood	400	7491	400	7491	0	0

.

Both the total amount of materials used and waste generated in each of the main scenarios are summarized in

Table 5. BAU, FL and Community scenarios. Inventory of materials used and waste generated during the refurbishment.

	BAU		FL		Community
	Material Input	Waste Output	Material Input	Waste Output	Material Input	Waste Output
	(kg)
EPS	2789	576	22,319	8554	5907	576
Glass	1547	1363	14,340	9377	2320	1363
Plastics	70	33	5602	376	906	33
Sheet steel	0	1600	4878	2400	789	1600
Aluminum	0	70	0	257	0	70
Wood	0	0	0	7491	0	0
Total	4405	3642	47,139	28,454	9922	3642

. In this inventory, I omit materials used for PV systems, battery storage and heat pumps. I considered materials used and waste generated during the measures of refurbishment, not in the maintenance of the buildings during the use phase.

The demand for materials in the FL scenario exceeds that of the BAU scenario by a factor of 10.7, while that of the Community scenario is 2.3 times higher than that of the BAU scenario. Focusing on the waste generated, the FL scenario exceeds that of the BAU scenario by a factor of 7.8, while the Community scenario corresponds to the waste volume of the BAU scenario.

3.2. Energy Calculations

All five scenarios are calculated regarding their final energy demand. This demand consists of the energy needed to operate the heat pump matching the final energy demand depending on the U-value of the buildings and the demand for household electricity. The results are summarized in

Table 6. Final (electric) energy demand, household electricity demand and auxiliary electricity demand for heat pump operation.

	SFH _E	MFH _E	SFH _MP1	MFH _MP1	SFH _MP2	MFH _MP2
	(kWh/a)
FEC	21,489	401,089	8926	332,818	4188	278,771
ET	3200	76,800	3200	76,800	3.200	76,800
FECel	None *		2975	110,939	1047	69,693
AEC	None *		992	36,980	262	17,423
TECel	3200	76,800	7167	224,719	4509	163,916

* No entry is made in this cell because it refers to a value demanded for an installed heat pump. For this state of the building archetype, no heat pump is modeled.

.

The final electric energy demand is covered partially by the solar energy generated locally using the PV systems, partially by the withdrawal from the external grid.

The final energy demand of the BAU scenario is 277 MWh/year; for the FL scenario, it is 200 MWh/year; and for the Community scenario, it is 266 MWh/year, shown as dots in Figure 2. The final energy demands of the BAU and Community scenarios distinguish themselves by only 11 MWh/year. This is due to the small overall difference in refurbishment measures taken between the two scenarios. The difference results from different energy performance levels of four SFHs. The influence of this difference on the result is not significant, because the difference those four SFHs make is under 4% compared to the final energy demand of the MFH, which exists on the MP1 level in both scenarios. The coverage of final energy demand through solar energy generated locally is 3.25% or, rather, 3.8% (BAU_I and BAU_EG); 9% or, rather, 12% (FL_I and FL_EG); and 5.6% (Community scenario), shown as yellow beams in Figure 2. In every scenario, the surplus of energy generated is fed into the external grid, shown as blue beams in Figure 2. These are 21 MWh/year for the BAU-I scenario, 19,5 MWh/year for BAU_EG scenario, 42 MWh/year for the FL-I scenario, 36 MWh/year for the FL_EG scenario and 15 MWh/year for the Community scenario.

3.3. Layering Perspectives

Figure 3 shows the layering of the results of the material inventory and the energy calculations as demand of materials and share of external grid usage to cover this final energy demand. The material demand of the BAU scenario is 4405 kg. The Community scenario exceeds this by 5520 t, and the FL scenario exceeds it by 42,735 t. This surplus of materials in the FL scenario is accompanied by the installation of high capacities of PV systems: 30% (FL_I) or, rather, 40% (FL_EG) self-used in the buildings. This results in a 91% or, rather, 88% share of external grid dependency for covering the minimized final energy demand. For the BAU scenario, the share of external grid utilization is 97% (BAU_I) or, rather, 96% (BAU_EG); for the Community scenario, it is 94%. The surplus demand of 37,220 t in the FL scenario compared to the Community scenario results in a reduction of 3% less external grid usage (FL_I) and 3% more if the buildings are interacting horizontally (FL_EG).

The results of the scenario modeling and analysis show the advantages of horizontal interaction in a holistic neighborhood refurbishment approach. The base scenarios BAU and FL aim for the establishment of a building stock with homogeneous energy efficiency levels. This results in a wide range of material demand, with the two scenarios diverging from each other by over 42 kt. The Community scenario is located in between, with a demand around 5.5 kt higher than that of the BAU scenario. The differentiation of the basis scenarios BAU and FL into the sub-scenarios _I and _EG shows the effects of horizontal building interaction, precipitating in higher self-consumption rates of the self-generated energy. The share of external grid use is reduced by 1.5% (BAU) and 3% (FL) when modeling the basis scenarios with horizontally interacting buildings. The Community scenario is again located in between the total shares of the BAU_EG and BAU_FL scenarios. It is 2% lower than in BAU_EG and 6% higher than in FL_EG. In the comparison of the BAU_EG and Community scenarios, an additional demand of 13% more materials results in a lowered share of external grid usage by 22.5% in the Community scenario. Comparing the FL_EG and Community scenarios, the surplus of demand on materials of 87% results in only 77.5% less share of external grid usage.

4. Discussion

4.1. Contribution to Scientific Debates

This study contributes to the scientific debate on the nexus of operational and embodied energy. The importance of the dynamics between operational and embodied energy to achieve sustainability goals is already emphasized [24]. Several studies discuss the role of embodied and operational energy in the construction sector [25,26,27,28,29,30]. However, they do not come to a clear-cut conclusion. Both studies that highlight the predominance of operational energy and those which highlight the predominance of embodied energy exist. The former type of study highlights the importance of the operational energy mostly arguing with the duration of use of residential buildings [30]. Such studies conclude that the environmental impact of material flows in the refurbishment phase loses importance over the remaining utilization phase and accompanied need for operational energy [30]. The latter form of study argues that the embodied energy predominates the environmental impacts of buildings. Such studies insist on the vanishing impact of operational energy because of dynamic developments in the energy markets [31,32]. As the energy transition precedes, they argue that the environmental impact of operational energy is reduced because of a higher share of renewable compared to fossil energy sources [31,32]. The discussion about the predominance of operational or embodied energy in residential buildings results, therefore, mainly from the long use phase of residential buildings and different assumptions made in the assessment. The prediction of developments over years to come results in vast uncertainties. These need to be addressed by carrying out more studies on this topic with a high standard of transparency over the assumptions made. The analysis in this study takes the operational and embodied energy of the scenarios into account. As such, it delivers a new database on which future work can be conducted.

The scenario modeling and analysis show that the concept of horizontal building interaction in a neighborhood as a stand-alone element of energy-oriented refurbishments neglects the potential of possible material savings in the refurbishment. I support the potential of understanding the buildings of a neighborhood not as a sum of individual elements [4] but as a conglomerate that must be thought of as a union. It is stated that the high material demands arising from the extensive refurbishment results in not achieving the sustainability goals set for the building stock [32]. This issue translates into the design of the FL scenario in this study. Here, large amounts of material are needed to bring the buildings to a forward-looking energy efficiency standard. The required embodied energy is accompanied by environmental impacts. However, aiming only for the minimum legal requirements, as in the BAU scenario, will lead to the failure of meeting the set goals [2]. By distancing ourselves from the ideal of energetically homogeneous building stocks, seeking State-of-the-Art practice when refurbishing and applying concepts of horizontally interacting buildings, the Community scenario offers a win-win situation. It minimizes the use of embodied energy while maximizing the self-consumption rate of renewable energy sources. The Community scenario strikes a balance between the BAU and FL scenarios by treating the heterogeneity of the building stock not as an obstacle that must be overcome but as a chance for new perspectives. By not striving to achieve a homogeneous energy standard for all individual buildings in a neighborhood, significant amounts of materials are saved. At the same time, the effects on final energy demand, shares of external grid usage and self-used energy from PV-systems are balanced partially by the horizontal building interaction. This study is the first to quantify this potential.

Apart from the nexus of embodied and operational energy, several studies discuss the differentiation of self-sufficiency and decentralization of a neighborhood’s energy supply [33,34,35,36]. This study contributes to this debate by including aspects of decentralization and local systems. Through the modeling of installed PV systems and the aim for a higher self-consumption rate, the neighborhood is comprehended as a local system. The neighborhood, however, does not explicitly aim to attain self-sufficiency. This corresponds to the findings that neighborhoods with a high degree of energy self-sufficiency do not necessarily have a lower environmental impact than neighborhoods with a lower degree of self-sufficiency [36]. At the same time, the environmental opportunities of local systems at the neighborhood level are highlighted in current works [36]. The establishment of these local systems increases the local utilization rate of self-generated renewable energy, which is associated with fewer grid losses in longer infrastructures [36]. The decentralization of the energy system is therefore viewed positively in contrast to a high degree of self-sufficiency [36].

For the detailed work that I performed to cite the existing literature in the field and identify the main connecting points of my work to the ongoing scientific debates, see Appendix A.

4.2. Uncertainties

The calculations are a first push toward the direction of combining horizontal building interaction and adapted energy efficient neighborhood refurbishments. As a simplified approach, it fulfils the need to give an understanding of the potential. Further works need to be conducted according to this concept in order to develop a comprehensive and holistic framework. Uncertainties arise within different aspects of this calculation. Catching these uncertainties would result in larger case studies and calculations that did not suit the purposes of this study: a first insight into the topic. However, the uncertainties identified are addressed in the following to establish the needed transparency to interpret the results of this study.

The WebTool and IWU typology only provide synthetic data. They reflect the reality as good as possible but are still only average. This applies to the energy and material data of both building archetypes modeled in the scenario. Also, the defined measures of refurbishment MP1 and MP2 are average measures, which do not reflect the reality of actual refurbishment measures. These data are used because they possess sufficient data quality for the calculations. They suit the purpose of this study, giving a first insight into the potential of horizontal building interaction in combination with adapted refurbishment measures. However, the influence of the neighborhood’s composition and of the initial energy performance assumed for the buildings needs to be assessed. To gain more details about the influence of the buildings’ initial energy performance, I exemplarily changed the energy performance of the MFH in its initial state from MFH_MP1 to MFH_E and revised the calculation accordingly. I chose the MFH for this assessment because it has a higher influence on the calculations than the SFH because of its size. Results show a significant increase in material demands in the BAU (18,556 kg) and Community (42,178 kg) scenarios. This increase results from the additional need for refurbishment in both scenarios. The material demand in the FL scenario remains the same. The reason for that is the assumption that the materials needed for a refurbishment to the energy performance level of MP2 are independent of the building’s initial state. The results of the energy calculation do not change in this calculation, because they are only dependent on the final state of the buildings, and this state does not change. The material demands of the FL and Community scenarios therefore converge, but the Community scenario is still favorable in comparison when looking at the total amounts of material demanded. Looking at the main composition of the neighborhood, the neighborhood modeled consists only of two residential building archetypes and nine buildings altogether. It builds a random but particular example of heterogeneity in a neighborhood. It does not include non-residential buildings. Including non-residential buildings into similar calculations is interesting because of their deviating energy load profiles in comparison to residential buildings. As with the building data, the neighborhood data suit only the purpose of a first insight into the concept; thus, their representativeness is limited. Considering the uncertainties in data, I still expect similar qualitative results from working with a different set of data.

Different assumptions were made in the calculations, all described in the section Material and Methods of this study. One assumption was identified as particularly sensitive. This is the self-consumption rate assumed in the scenarios. The self-consumption rate expresses the percentage of total solar energy generation that is utilized by the household or owner of a PV system either immediately or over a longer period [37]. Translated to a neighborhood, the self-consumption rate is calculated by the utilization of the whole neighborhood. In practice, the main advantage of an increased self-consumption rate for the prosumers is a more efficient way of using the generated energy [38]. This leads to being less dependent on market prices and saving money. Environmentally, a higher self-consumption rate directly and indirectly impacts the net load demand profile of the prosumers [38]. Challenges with an increasing self-consumption rate are the increase in voltage level and the imbalance in current and voltage, leading to power factor instability [38]. In this study, one of the main results is the share of external grid usage. This share is directly dependent on the self-consumption rate assumed in the calculations. As the calculations are based on a synthetic neighborhood, it was not possible to rely on actual self-consumption rates for those assumptions. Instead, I conducted specified literature research on self-consumption rates in connected neighborhoods. As the results are directly dependent on these assumptions, this procedure is highlighted here. When working with assumptions in future work on this topic, it is recommended that researchers consider variation in consumption rates. This would increase the reliability and transferability of the calculations. Addressing the uncertainties resulting from the assumptions made for the self-consumption rate, I performed a sensitivity analysis. To test the robustness of the results, I recalculated the Community scenario with a self-consumption rate of 40%, same as in the FL_EG scenario. Results show an increased share of external grid usage (+1%), an increase of energy withdrawals from the external grid (+1.2%) and an increase in energy fed to the external grid (+20%). Additionally, the share of self-used PV energy on the final energy demand decreases (−1.13%) with the decreased self-consumption rate. The results on the material calculations remain unchanged. The sensitivity analysis on the self-consumption rate shows that the effects favoring the Community scenario over the other scenarios modeled decrease on the energy side, but they do not result in a changed order of the scenarios. Still, the Community scenario offers a compromise within the scenarios in regard to all parameters calculated. Another assumption made during the calculation is the omission of material used for the installation of PV systems, battery storage and heat pumps. This assumption and its influence on the main conclusion need to be assessed by looking at the scenario design. Additional PV systems are installed only in the FL scenario, which already has the highest demand for materials, thus strengthening the conclusion made. Battery storage, on the other hand, is modeled for all scenarios in different capacities. The storage of the BAU and FL scenarios depends on their demand for electricity, assuming 1 kWh storage capacity per 1000 kWh demand for electricity. This results in storage capacities of 16 kWh (BAU) and 36 kWh (FL). The Community scenario has an installed storage capacity of 40 kWh. When assuming a proportional progression of material demand and storage capacity, the induced material demand for the installation of battery storage is only significant for the scenario BAU. This will result in a convergence in material demand of the BAU and Community scenarios. However, analyzing the total difference between the two scenarios, it is unlikely to assume a significance in this convergence. Finally, assessing the influence of the omitted material demand for the heat pump. In all scenarios, heat pumps are installed or replaced in the building that are being refurbished. The heat pumps in the scenarios will not differ significantly in mass but gradually in technology, making the number of heat pumps installed in the respective scenario representative of the material demand. In the BAU and Community scenarios, four SFH_Es are refurbished, and in the FL scenario, an additional two SFH_MP1s and one MFH_MP1 are refurbished. This indicates that the material demand of the FL scenario will further increase in comparison to that of the Community scenario (and BAU). Consequently, the omission of the materials for PV systems, battery storage and heat pumps does not weaken the main conclusion.

It needs to be stated that the calculations made, especially for the coverage of final energy demand through locally produced energy, are simplified. Aspects such as load and PV time matching, network limits, and a control logic were not included because the focus of this article is less on the detailed work of the specific implementation of horizontal building interaction. Extensive and detailed work has been performed regarding this aspect [15]. Instead, this article aims to widen the perspective of horizontal building interaction from a stand-alone energetic dimension to the opportunities arising when considering the material dimension simultaneously.

4.3. Future Work

Future research is required to provide the definition and analysis of further neighborhood scenarios—synthetic and real—to elaborate the opportunities, possibilities and limitations of the horizontal interaction approach in more detail. This includes the adaptation of neighborhood definition and considerations regarding the resilience of such systems.

Also, the assessment of environmental impacts needs to be further worked on. Examining the scenarios of their impact on embodied and operational energy enables the comparison of the potential for material savings and the associated reduction in embodied energy during the refurbishment phase. A dynamic modeling approach is substantial for obtaining reliable results, considering, e.g., development scenarios in the energy sector [23,39,40,41]. This can be used to derive recommendations for action for the development of energy-oriented refurbishment roadmaps at the neighborhood level. Life cycle assessment is particularly suitable for modeling the environmental impacts of these refurbishment scenarios [16,42,43]. The transparency of the assumptions made also needs to be ensured. To make the results of such studies comparable, a comprehensive framework on the application of life cycle assessment on the neighborhood scale must be developed.

To enable horizontal building interaction, a regulatory framework is needed. The framework for ‘Energiegemeinschaften’ in Austria can serve as a practical orientation. This needs to be combined with an adapted neighborhood refurbishment concept. The current legislative goal of developing an energetically homogeneous building stock needs to be rethought. Similar to the automotive sector, climate targets need to be translated into fleet targets instead of individual targets, meaning that the targets of GHG reduction do not need to be achieved by every individual unit, car or building, but a manageable number of units, cars or buildings need to match this target. This leaves more room for individual solutions and new ideas, acknowledging the heterogeneity of the stock as an area of possibilities. The implementation of measures on a small scale can promote acceptance, show advantages and create the opportunity to develop the concept further. The building stock of urban housing cooperatives could suit this purpose as a living lab. These buildings are often located in close proximity to each other, and the ownership structure as an incorporated association is a favorable condition for horizontal building interaction and development of a comprehensive energy efficiency refurbishment roadmap.