The Use of Ground Source Heat Pump to Achieve a Net Zero Energy Building

: Currently, ground source heat pump (GSHP) technology is being studied, as the use of the ground as a source of renewable energy allows signiﬁcant energy savings to be obtained. Therefore, it is useful to quantify how these savings help to achieve the energy balance of a Net Zero Energy Building (NZEB) compared to an air source heat pump or a condensing boiler coupled to a chiller. This paper assesses how these savings a ﬀ ect the number of photovoltaic panels installed on the roof of a building to obtain the NZEB target. The study is conducted by dynamic simulation for a building used as a bed and breakfast, virtually placed in two Italian towns. The energy savings and reduction of CO 2 emissions, the percentage of renewable energy used, and the photovoltaic surface needed are assessed. Finally, the discounted payback period is calculated. The results show that the GSHP, unlike the systems to which it is compared, allows an NZEB to be obtained by balancing yearly energy consumption with energy production systems which only use on-site renewable energy sources (by exploiting the surface available on the roof) for both of the climatic conditions considered. GSHP also allows primary energy requests equal to or less than 57 kWh / m 2 to be obtained.


Introduction
In Europe, buildings are responsible for about 36% of total CO 2 emissions. Thus, European Directives have been issued to reduce these emissions by improving energy performance of buildings, i.e., the Energy Performance Building Directive, so-called "EPBD" [1] and "EPBD recast" [2]; successively, the European Directive 2018/844 [3] aims to completely decarbonise the building stock by 2050. This can be achieved by transforming current buildings into nearly Zero Energy Buildings (nZEBs) and employing mainly renewable energy sources [4]. However, the goal of Net Zero Energy Buildings (NZEBs) is still far from being achieved.
An NZEB is a building characterized by opaque and transparent components with optimal values of thermal parameters (for example, a low unitary thermal transmittance), in which the position of the openings helps natural ventilation and the most energy efficient systems are installed; for these reasons, this building has a very low energy requirement, completely satisfied by renewable energy sources [1,5,6]. In practice, to obtain this target, many parameters must be fixed, such as the control volume for applying the energy balance, and the time base of this balance, i.e., monthly, seasonal, or annual [7]. Berggren and Wall [8] analyse the importance of normalizing the balance referring to both internal and external boundary conditions and highlighting the importance to keep the indoor investigate the GSHP in urbanized cities where the heat from an underground tunnel (the ground surrounding the infrastructure) could be exploited to improve the heating mode operation of a GSHP. Thus, the advantage of this technology can be used to satisfy the need to use renewable energy sources for civil and non-civil use. In the literature, this system is coupled with several innovative complementary systems. In Buonomano et al. [28], geothermal energy coupled with solar energy is used and investigated in a trigeneration plant. Huang et al. [29] investigate the optimization of a large scale solar-assisted ground source heat pump for district heating, finding an optimal match between the size of the ground heat storage, the collector area, and the tank volume. In [30,31], the coupling between the GSHP and the photovoltaic/thermal system is analysed, demonstrating how the use of these systems together increases the thermal efficiency of the collectors; the electrical efficiency of the cells remains high without the risk of cell damage due to overheating. Due to its characteristics and high efficiency in terms of energy saving, CO 2 emissions reduction, and renewable energy source use, this system can easily be implemented in buildings to achieve more easily the NZEB target. However, little research has been undertaken about this system in the literature.
In Fedajev et al. [32], with reference to an nZEB, the influence of different earth heat exchangers linked to a GSHP and thermal storage is examined, for the very harsh climatic conditions of Finland; the importance of the thermal storage to achieve the goal of an nZEB is shown.
In a review article, Gao et al. [33] highlight the different potentialities of the technologies that can be coupled to the geothermal heat pump, suggesting this technology can be used in zero energy buildings.
In the technical-scientific literature to date, little examination has been made of the contribution of GSHPs to obtain a zero-energy balance in an NZEB and their influence in reducing the photovoltaic surface to be used in densely urbanized cities. Often these systems are analysed separately without any reference to the reduction of photovoltaic panels to be installed. Therefore, this paper analyses the aforementioned contribution by comparing the GSHP to other systems, such as an invertible ("reversible") air source heat pump or a condensing boiler coupled to a chiller, in order to meet the NZEB target in different climatic conditions and to evaluate the photovoltaic surface needed to achieve the status of an NZEB compared to that available on the roof. To this aim, a case study building used as a bed and breakfast (B&B) and virtually located in two Italian towns with different climatic conditions is analysed. Using the dynamic energy simulation software, DesignBuilder, a detailed energy analysis is carried out. The procedure is partially validated by comparing the building energy requirements with literature data. Moreover, other important results are obtained regarding the primary energy consumption, the reduction of CO 2 emissions, and the percentage of renewable energy used for the various solutions. Finally, a technical-economic analysis using the discounted payback is performed.

Model Description
Referring to a case study building, the seasonal energy consumption and CO 2 emissions of the building's energy systems (in particular, in the case with GSHP) are evaluated.
First, the building to be designed should be characterised by low energy needs. Therefore, optimal thermal characteristics for the building envelope must be chosen to minimize the thermal losses in winter, while the free heat gains should be maximized in winter and minimized in summer. To this end, optimal thermal insulation, compact form, and innovative technologies should be selected [16].
Subsequently, the HVAC (heating, ventilation and air-conditioning) system and other building energy systems must be chosen and designed to minimize the primary energy requirements [22] by using energy-efficient solutions.
Therefore, the use of a GSHP is considered, and it is subsequently determined if this choice makes it easier to reach the NZEB target (i.e., the energy balance equal to zero between energy required by the building-system complex and energy produced from renewable sources). The building of the case study is virtually located in two different Italian climatic zones and the energy analysis is performed through dynamic building energy performance simulation software, i.e., DesignBuilder [34], based on the EnergyPlus calculation engine. The U.S. Department of Energy has performed several validation tests referring to EnergyPlus [35,36]. The climatic data are taken from ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers) [37].
The focus of the paper is an energetic, environmental, and technical-economic comparison between various generators of thermal energy: a GSHP characterized by vertical probes, an air source heat pump, and a condensing boiler plus an air-cooled chiller.
Regarding the DesignBuilder software, the CTFM (Conduction Transfer Function Module) is selected based on the algorithm "conduction transfer function". Moreover, the algorithm DOE-2 [38] is set for the outside convection and the algorithm TARP [39] for the inside convection [40].
The heat pumps are simulated by DesignBuilder using an equation-fit based model [41,42] to predict the heat pumps' performance for cooling and heating modes. Equations (1)-(4) characterize the load curves of the heat pumps: where: • The coefficients A1-D5 are entered in the heat pump model to simulate the performances. They refer to the performance at partial loads of heat pumps present in the DesignBuilder database. After inputting the dimensions of the HVAC system and choosing the heat pump of the required size, the software chooses suitable values of these coefficients basing on information from the manufacturer. A library of templates containing pre-defined manufacturers' heat pump data is provided to allow an early stage analysis to be carried out.
In addition, the undisturbed temperature of the ground is needed in order to evaluate the seasonal efficiency of a GSHP. To this end, the equation of Kusuda [43] is applied: where: • T g (D,t): ground temperature at a depth D after t days (starting from 1 January), • C; • T av : yearly average temperature of the outdoor environment on the basis of statistical information, • C; • A: amplitude of the temperature annual oscillation, • C; • t: sequential number of the day (1 refers to 1 January); • t Tmin : sequential number of the day corresponding to the minimum ground temperature, according to statistical data (1 refers to 1 January); • D: depth of the ground, m; • α g : daily equivalent thermal diffusion of the ground (m 2 /day).
For these parameters, the values reported in Table 1 are used. The seasonal energy performance of the heat pumps and chiller (seasonal coefficient of performance (SCOP) for heating operation, seasonal energy efficiency ratio (SEER) for cooling) are calculated by means of Equations (6) and (7): where: • θ h is the absolute temperature of the hot source; • θ c is the absolute temperature of the cold source; • η II is the second law efficiency.
DesignBuilder cannot automatically derive the size of the GSHP, thus the following procedure was applied:

•
The building design thermal loads (both for heating and cooling) are evaluated; For the evaluation of the correct size of the heat pump, the thermal power necessary for handling the treated ventilation external air ("primary air") was also considered. In the examined case study, this is about 6 kW in summer; moreover, considering a ground that dissipates 50 W/m of thermal energy, the number of 4 vertical U-bend probes was obtained for both Milan and Palermo.
Regarding energy analysis, the primary energy is evaluated: the considered conversion factors are 2.2 for electricity and 1.05 for natural gas. To calculate the CO 2 emissions, the used procedure is explained in Section 4 (Results and discussion-CO 2 emissions) [44].

Procedure to Obtain an NZEB
Regarding the satisfaction of the energy balance, there are currently no specific requirements for the NZEB target. The Italian law strictly defines only nearly zero energy building (nZEB), by means of the D.M. 26 June 2015 [45]. Therefore, according to this decree, the following requirements have been satisfied for the building-system complex examined (the first three refer only to the building envelope, the fourth refers to only the building energy systems, the last to the entire building-system complex): • Building average heat transfer coefficient: The ratio between the summer equivalent solar area of the windows and the useful walking surface: A sol,equiv,summer /A walking surface ≤ reference value (-); • Energy performance indexes for the building envelope in winter and summer: EP H,nd , EP C,nd ≤ reference values (kWh/m 2 y); • Efficiencies of the building energy systems (H for heating, C for cooling, DHW for domestic hot water): Global energy performance index of the building-systems complex: EP glob,tot ≤ reference value (kWh/m 2 y).
In this way it was verified that the case study building is a nZEB; subsequently, in order to obtain an NZEB, an electricity production system from renewable sources (photovoltaic solar modules) is designed and inserted on the roof of the building. The global capturing surface of these modules is evaluated for the following different cases: GSHP, air source heat pump, condensing boiler plus chiller. Thereby, it is possible to estimate whether the GSHP would allow to achieve the NZEB target with a photovoltaic surface significantly smaller compared to the other solutions.
The derivation of the energy balance of a NZEB can be performed in different ways [6,46,47]. In this work, the control volume that coincides with the complete building is chosen, and therefore the building-systems energy demand and the energy produced by renewable sources (kWh/m 2 y) are considered in the energy balance: This energy balance must be satisfied to obtain a NZEB. To this end, an accurate design of the building-system complex is required to minimize the energy demands. Therefore, it was determined which of the analysed solutions (GSHP, air source heat pump, condensing boiler plus chiller) has the minimum energy requirement to facilitate the achievement of the NZEB objective.
The Figure 1 shows the general design and computational workflow followed in this study for obtaining the NZEB target.

Case Study
The pilot building is used as a bed and breakfast and is virtually located in two Italian towns having different climates ( Table 2): Palermo (South Italy), with mild winters and very hot summers; Milan (North Italy), with cold winters and hot summers.
The building is characterised by two floors, global area of 310 m 2 , and global volume of 900 m 3 . Accurate design of the building and passive design rules are applied in order to achieve the NZEB target: considering the sun position, the living zone is placed on the south side in order to maximize solar gains in winter, while the service rooms and sleeping zone are placed on the north side (Figures 2 and 3). Moreover, the windows are designed in order to optimise natural ventilation. A compact form of the building is chosen, so the surface-to-volume ratio is equal to only 0.20 m −1 . Low values of unitary thermal transmittance of the building envelope components are used (Table  3), particularly for the coldest town (Milan), in order to minimise the thermal losses in winter. Table 4 reports the design thermal loads of the building (for both heating and cooling conditions), calculated by the software, while Table 5 shows the values of the seasonal average energy efficiency for the three compared systems.   Table 4. Design thermal loads of the case study building for both heating and cooling.

Case Study
The pilot building is used as a bed and breakfast and is virtually located in two Italian towns having different climates ( Table 2): Palermo (South Italy), with mild winters and very hot summers; Milan (North Italy), with cold winters and hot summers. The building is characterised by two floors, global area of 310 m 2 , and global volume of 900 m 3 . Accurate design of the building and passive design rules are applied in order to achieve the NZEB target: considering the sun position, the living zone is placed on the south side in order to maximize solar gains in winter, while the service rooms and sleeping zone are placed on the north side (Figures 2 and 3). Moreover, the windows are designed in order to optimise natural ventilation. A compact form of the building is chosen, so the surface-to-volume ratio is equal to only 0.20 m −1 . Low values of unitary thermal transmittance of the building envelope components are used (Table 3), particularly for the coldest town (Milan), in order to minimise the thermal losses in winter.  Table 4 reports the design thermal loads of the building (for both heating and cooling conditions), calculated by the software, while Table 5 shows the values of the seasonal average energy efficiency for the three compared systems.    Each of the analysed thermal energy generators is connected to fan-coil units located inside the rooms and to an air handling unit (AHU) for primary air.
The design thermo-hygrometric conditions are as follows: • indoor air: temperature of 20 °C for winter and 26 °C for summer, relative humidity of 50% for both winter and summer; • supply primary air: temperature of 20 °C for winter and 12 °C for summer.  Each of the analysed thermal energy generators is connected to fan-coil units located inside the rooms and to an air handling unit (AHU) for primary air.
The design thermo-hygrometric conditions are as follows: • indoor air: temperature of 20 • C for winter and 26 • C for summer, relative humidity of 50% for both winter and summer; • supply primary air: temperature of 20 • C for winter and 12 • C for summer.
Referring to the heating mode, different working programs are selected for the two analysed towns, according to DPR 412/93 [48]: Referring to the cooling mode, the operational programs are fixed as follows: The production of domestic hot water (DHW) is obtained by means of solar thermal collectors (five for Milan, four for Palermo) and a dedicated air-to-water heat pump (COP = 2.5), while the consumption rate is considered to be 2.5 L/m 2 day.
LED devices with a linear control system are used as the lighting system (2.5 W/m 2 -100 lux). The design values of the illuminance are 300 lux in bedrooms and 500 lux in common spaces.
Few other electrical devices exist, so a range of only 1-4 W/m 2 is considered.

Energy Analysis
The primary energy demand for the different proposed solutions, divided for energy uses, is shown in Figures 4 and 5, for Palermo and Milan, respectively. The aim is to highlight the primary energy saving obtained when using the GSHP instead of the air source heat pump or the condensing boiler coupled to the chiller.
For Palermo, the following percentage savings can be obtained from the values of Figure 4: the use of GSHP involves an energy saving of 15% for heating plus cooling when compared to the air source heat pump (23% with reference to only heating, 12% for only cooling); the above mentioned saving is 38% when compared to the case with boiler plus chiller.  Regarding the entire yearly demand of primary energy for Palermo, Figure 6 shows that the GSHP results in a 5% saving compared to the air source heat pump and 16% compared to the solution with boiler plus chiller. The yearly demand of electric energy varies from 6131 kWh/y (20 kWh/m 2 y) related to the use of the air source heat pump, to 5810 kWh/y (18.7 kWh/m 2 y, i.e., −5%) related to the use of the GSHP. Figure 7 reports the pie chart relative to the energy requirements in the case of the GSHP for Palermo. The aim is to highlight the primary energy saving obtained when using the GSHP instead of the air source heat pump or the condensing boiler coupled to the chiller.
For Palermo, the following percentage savings can be obtained from the values of Figure 4: the use of GSHP involves an energy saving of 15% for heating plus cooling when compared to the air source heat pump (23% with reference to only heating, 12% for only cooling); the above mentioned saving is 38% when compared to the case with boiler plus chiller.
Regarding the entire yearly demand of primary energy for Palermo, Figure 6 shows that the GSHP results in a 5% saving compared to the air source heat pump and 16% compared to the solution with boiler plus chiller. The yearly demand of electric energy varies from 6131 kWh/y (20 kWh/m 2 y) related to the use of the air source heat pump, to 5810 kWh/y (18.7 kWh/m 2 y, i.e., −5%) related to the use of the GSHP. Figure 7 reports the pie chart relative to the energy requirements in the case of the GSHP for Palermo. Regarding the entire yearly demand of primary energy for Palermo, Figure 6 shows that the GSHP results in a 5% saving compared to the air source heat pump and 16% compared to the solution with boiler plus chiller. The yearly demand of electric energy varies from 6131 kWh/y (20 kWh/m 2 y) related to the use of the air source heat pump, to 5810 kWh/y (18.7 kWh/m 2 y, i.e., −5%) related to the use of the GSHP. Figure 7 reports the pie chart relative to the energy requirements in the case of the GSHP for Palermo.  With reference to Milan (cold winters), from the values of primary energy demand reported in Figure 5, the GSHP involves the following saving on the primary energy: 23% for heating plus cooling compared to the air source heat pump (19% for solely heating, 34% for cooling); 55% for heating plus cooling compared to boiler plus chiller.
Regarding the entire yearly demand of primary energy for Milan, Figure 8 shows that the GSHP involves a 10% saving compared to the air source heat pump and 33% compared to the solution with the condensing boiler plus chiller. The yearly demand of electric energy varies from 8171 kWh/y (26.4 kWh/m 2 y) related to the use of the air source heat pump, to 7316 kWh/y (23.6 kWh/m 2 y, i.e., −10%) related to the use of the GSHP. Although the percentage saving of total primary energy between the GSHP and the air source heat pump seems to be small, it is useful to highlight that the total primary energy consumption is equal to 57 kWh,pr/m 2 y. According to the New Building Institute of the U.S. [22], that value identifies an NZEB, and this result in a cold climate such as Milan is reached only by using a GSHP as a thermal energy generator.
Error! Reference source not found. reports the pie chart relative to the energy requirements in the case of GSHP for Milan. With reference to Milan (cold winters), from the values of primary energy demand reported in Figure 5, the GSHP involves the following saving on the primary energy: 23% for heating plus cooling compared to the air source heat pump (19% for solely heating, 34% for cooling); 55% for heating plus cooling compared to boiler plus chiller.
Regarding the entire yearly demand of primary energy for Milan, Figure 8 shows that the GSHP involves a 10% saving compared to the air source heat pump and 33% compared to the solution with the condensing boiler plus chiller. The yearly demand of electric energy varies from 8171 kWh/y (26.4 kWh/m 2 y) related to the use of the air source heat pump, to 7316 kWh/y (23.6 kWh/m 2 y, i.e., −10%) related to the use of the GSHP. Although the percentage saving of total primary energy between the GSHP and the air source heat pump seems to be small, it is useful to highlight that the total primary energy consumption is equal to 57 kWh, pr /m 2 y. According to the New Building Institute of the U.S. [22], that value identifies an NZEB, and this result in a cold climate such as Milan is reached only by using a GSHP as a thermal energy generator.
Regarding the entire yearly demand of primary energy for Milan, Figure 8 shows that the GSHP involves a 10% saving compared to the air source heat pump and 33% compared to the solution with the condensing boiler plus chiller. The yearly demand of electric energy varies from 8171 kWh/y (26.4 kWh/m 2 y) related to the use of the air source heat pump, to 7316 kWh/y (23.6 kWh/m 2 y, i.e., −10%) related to the use of the GSHP. Although the percentage saving of total primary energy between the GSHP and the air source heat pump seems to be small, it is useful to highlight that the total primary energy consumption is equal to 57 kWh,pr/m 2 y. According to the New Building Institute of the U.S. [22], that value identifies an NZEB, and this result in a cold climate such as Milan is reached only by using a GSHP as a thermal energy generator.
Error! Reference source not found. reports the pie chart relative to the energy requirements in the case of GSHP for Milan.   As shown in Figures 6 and 8, the entire primary energy requirement for both of the localities is low: 45-54 kWh/m 2 y for Palermo and 57-85 kWh/m 2 y for Milan.
Clearly, a significant primary energy saving is achieved for heating plus cooling with the use of the GSHP compared to the air source heat pump (15-23%), or to the boiler plus chiller (38-55%), mainly for the colder climate of Milan (55%). Similar results are found in García-Céspedes et al. [49]: in fact, in this paper the savings related to a GSHP are also around 20% compared with a traditional heat pump. Furthermore, according to an ENEA report [50], the hotels in Italy, including bed and breakfasts, have an electricity consumption of 30-60 kWh/m 2 for southern Italy (Palermo) and 30-90 kWh/m 2 for northern Italy (Milan); the analysed building, designed to be a NZEB, consumes between 24 and 26 kWh/m 2 of electricity.
At present, fundamental topics are sustainability and attention to the environment. In fact, climate change is accelerating due to greenhouse gas emissions. The European Union has the objective to reduce gas emissions of 20% by 2020 and to obtain total decarbonization by 2050 [44]; all of the actions tend to an energy-efficient and low-carbon economy. One of the first steps in reducing CO2 emissions is to lower energy consumption for heating and cooling by using increasingly efficient technologies.
For this reason, as an appendix of the energy analysis, a simplified analysis of the CO2 emissions was performed. This is not a Life Cycle Assessment (LCA) or a cradle-to-grave analysis, but only an investigation to compare the reduction of CO2 emissions related to the use of the building energy systems (so only during the useful life of the building), comparing the case of considering a GSHP with the other two system solutions. The reduction in CO2 emissions is assessed with reference to both heating only and global energy used during the year.
To calculate the CO2 emissions of the condensing boiler, the primary energy is multiplied by the standard emission factor of greenhouse gases referring to natural gas, equal to 0.202 kgCO2/kWh [44]. To obtain the CO2 emissions from electric heat pumps or a chiller, the final electric energy is As shown in Figures 6 and 8, the entire primary energy requirement for both of the localities is low: 45-54 kWh/m 2 y for Palermo and 57-85 kWh/m 2 y for Milan.
Clearly, a significant primary energy saving is achieved for heating plus cooling with the use of the GSHP compared to the air source heat pump (15-23%), or to the boiler plus chiller (38-55%), mainly for the colder climate of Milan (55%). Similar results are found in García-Céspedes et al. [49]: in fact, in this paper the savings related to a GSHP are also around 20% compared with a traditional heat pump. Furthermore, according to an ENEA report [50], the hotels in Italy, including bed and breakfasts, have an electricity consumption of 30-60 kWh/m 2 for southern Italy (Palermo) and 30-90 kWh/m 2 for northern Italy (Milan); the analysed building, designed to be a NZEB, consumes between 24 and 26 kWh/m 2 of electricity.
At present, fundamental topics are sustainability and attention to the environment. In fact, climate change is accelerating due to greenhouse gas emissions. The European Union has the objective to reduce gas emissions of 20% by 2020 and to obtain total decarbonization by 2050 [44]; all of the actions tend to an energy-efficient and low-carbon economy. One of the first steps in reducing CO 2 emissions is to lower energy consumption for heating and cooling by using increasingly efficient technologies.
For this reason, as an appendix of the energy analysis, a simplified analysis of the CO 2 emissions was performed. This is not a Life Cycle Assessment (LCA) or a cradle-to-grave analysis, but only an investigation to compare the reduction of CO 2 emissions related to the use of the building energy systems (so only during the useful life of the building), comparing the case of considering a GSHP with the other two system solutions. The reduction in CO 2 emissions is assessed with reference to both heating only and global energy used during the year.
To calculate the CO 2 emissions of the condensing boiler, the primary energy is multiplied by the standard emission factor of greenhouse gases referring to natural gas, equal to 0.202 kg CO 2 /kWh [44]. To obtain the CO 2 emissions from electric heat pumps or a chiller, the final electric energy is multiplied by the standard emission factor of greenhouse gases related to electricity, equal to 0.483 kg CO 2 /kWh [44].
As can be seen in the Table 6, a significant reduction of CO 2 emissions using the heat pumps instead of a condensing boiler (plus chiller) is obtained. This reduction of CO 2 emissions is 60% and 67% for only heating for Milan and Palermo, respectively, using a GSHP instead of the condensing boiler plus chiller (Figures 10 and 11). CO 2 emissions reduction relating to global energy consumption is 33% for Milan and 16% for Palermo comparing the GSHP to the condensing boiler plus chiller. The CO2 emissions reduction comparing the GSHP to the air source heat pump is also significant (19-23%) referring to only heating consumption. Conversely, when considering the global energy consumption, this reduction is only 5-11%.    Target   Tables 7 and 8 report the yearly energy balance and the number of solar thermal and photovoltaic modules necessary to obtain a zero energy balance between produced energy by renewables and consumed energy by the building, in the case of Palermo and Milan, respectively, considering the three proposed HVAC solutions and a useful surface on the roof of the building to install thermal and photovoltaic solar panels equal to 90 m 2 . The following procedure was considered to obtain a NZEB:

NZEB
• For the cases characterized by solely electric energy demand (GSHP and air source heat pump), the yearly electric energy demanded by all the systems of the building is calculated: it must be balanced by the yearly electric energy obtained by the photovoltaic modules (320 Wpeak per module); • For the case characterized by both electric energy and natural gas demand (boiler plus chiller), the yearly electric energy demanded by all the systems of the building must be balanced by the yearly electric energy obtained by the photovoltaic modules, while the yearly thermal energy related to the natural gas demanded by the boiler (3147 kWh for Palermo) is related to primary energy (3304 kWhpr), as suggested in [51] and [52], and this must be balanced by the annual electric energy obtained by the photovoltaic modules (note that, in the case of photovoltaic modules, this electric energy is equal to the corresponding primary energy).
In the case of Palermo (Table 7) when using the GSHP, 27 photovoltaic modules are necessary to obtain an NZEB, while 42 PV modules are needed when using a condensing boiler plus chiller, and 28 when using an air source heat pump. In a hot climate such as that of Palermo, both the definition of an NZEB obtained satisfying the energy balance dictated by the EPBD [2] and the definition of an NZEB suggested by the New Building Institute of U.S. reported in [22] are verified; however, using a GSHP between 5 and 92% (compared to the other two system configurations) of capturing area to be used for photovoltaics is saved. The CO 2 emissions reduction comparing the GSHP to the air source heat pump is also significant (19-23%) referring to only heating consumption. Conversely, when considering the global energy consumption, this reduction is only 5-11%. Target   Tables 7 and 8 report the yearly energy balance and the number of solar thermal and photovoltaic modules necessary to obtain a zero energy balance between produced energy by renewables and consumed energy by the building, in the case of Palermo and Milan, respectively, considering the three proposed HVAC solutions and a useful surface on the roof of the building to install thermal and photovoltaic solar panels equal to 90 m 2 . The following procedure was considered to obtain a NZEB:

NZEB
• For the cases characterized by solely electric energy demand (GSHP and air source heat pump), the yearly electric energy demanded by all the systems of the building is calculated: it must be balanced by the yearly electric energy obtained by the photovoltaic modules (320 W peak per module); • For the case characterized by both electric energy and natural gas demand (boiler plus chiller), the yearly electric energy demanded by all the systems of the building must be balanced by the yearly electric energy obtained by the photovoltaic modules, while the yearly thermal energy related to the natural gas demanded by the boiler (3147 kWh for Palermo) is related to primary energy (3304 kWh pr ), as suggested in [51,52], and this must be balanced by the annual electric energy obtained by the photovoltaic modules (note that, in the case of photovoltaic modules, this electric energy is equal to the corresponding primary energy).
In the case of Palermo (Table 7) when using the GSHP, 27 photovoltaic modules are necessary to obtain an NZEB, while 42 PV modules are needed when using a condensing boiler plus chiller, and 28 when using an air source heat pump. In a hot climate such as that of Palermo, both the definition of an NZEB obtained satisfying the energy balance dictated by the EPBD [2] and the definition of an NZEB suggested by the New Building Institute of U.S. reported in [22] are verified; however, using a GSHP between 5 and 92% (compared to the other two system configurations) of capturing area to be used for photovoltaics is saved.  With reference to Milan (Table 8), this town has a greater energy demand due to winter heating, therefore the energy saving induces a significant decrease of the photovoltaic surface: when using the GSHP, 21 PV modules are necessary to reach the NZEB target, versus 23 PV modules in the case of air source heat pump and 52 modules in the case of boiler plus chiller (the worst case), i.e., a reduction of 10 and 61% in the photovoltaic surface. In the case of the continental climate of Milan, it is not possible to satisfy the NZEB target in the case of a condensing boiler plus chiller as the photovoltaic surface needed to obtain a zero energy balance of (105 m 2 ) is greater than the surface available for installation (90 m 2 ). Furthermore, based on the definition of an NZEB reported by the New Building Institute of U.S. [22], the only case in which the primary energy consumption limit value of 57 kWh/m 2 is respected is that with the GSHP.
In Italy, there are currently specific legislative prescriptions to define an nZEB, but not an NZEB. Therefore, for all the analysed cases, the verification of the minimum requirements for an nZEB was satisfied using the TERMOLOG software [53]. Moreover, the percentage was obtained of renewable energy exploited by the various examined technologies: considering both the towns, 81% is related to the use of the GSHP, 75-76% for the air source heat pump, and 63-64% for the solution with boiler plus chiller.

Technical-Economic Analysis
For the three analysed HVAC configurations, the investment costs were evaluated [54]; then, discounted payback (DPB) periods were calculated, as reported in Table 9. To define the energy costs, the calculation was made by referring to the values recommended by Eurostat's "Statistics Explained" [55], i.e., an average value of 0.225 €/kWh for electricity consumption and 0.100 €/kWh for natural gas.
From an economic point of view, the comparison between the GSHP versus boiler plus chiller is the most convenient if it refers to a continental climate (Milan), as the energy saving related to the GSHP is relevant. In fact, the DPB is only 3 years for Milan, while for Palermo (mild climate) it is around 14 years (a not recommended value).
The GSHP is less convenient when compared to the air source heat pump: both for Milan and for Palermo, the DPB is more than the useful life, so the air source heat pump is economically more convenient than the GSHP. However, it is noteworthy that the air source heat pump may malfunction in cold climates such as Milan; thus, it may not always satisfy the user when very low external temperatures occur.

Conclusions
This paper investigates through an energy assessment the contribution that the geothermal heat pump can have on the reduction of the surface to be used for photovoltaics in order to obtain an NZEB with only on-site renewable sources. These assessments are carried out by comparing the GSHP with a common system-that is, is a condensing boiler coupled to a chiller-and with a more energy efficient system, namely, an air source heat pump.
It was shown that when using the GSHP, compared to the other systems, the NZEB target is obtained by using a smaller photovoltaic surface. Furthermore, although the comparison is also carried out with a performing technology such as the air-to-water heat pump, the GSHP is the only system that allows annual primary energy consumption equal to or lower than 57 kWh/m 2 to be obtained for both of the analysed climatic zones. This is an important result since this value in the current scientific community is taken as a reference to define an NZEB. These are the main innovative aspects obtainable from this paper, that is, the optimal coupling of a GSHP with a photovoltaic system to obtain the NZEB target, minimizing the number of photovoltaic panels.
In more detail, it was found that the use of the GSHP, compared with the two other systems, allows considerable energy savings for heating and cooling (15-38% for Palermo and 23-55% for Milan).
Moreover, a significant reduction of CO 2 emissions is obtained when using a ground source heat pump instead of a condensing boiler (plus chiller). This reduction

•
is 60 and 67% when considering only heating energy requirements for Milan and Palermo, respectively; • is instead 33% for Milan and 16% for Palermo in reference to global energy consumption.
The CO 2 emissions reduction comparing the GSHP and the air source heat pump is also significant (19-23%) regarding to heating only consumption. Conversely, when considering the global energy consumption, this reduction is only 5-11%.
Regarding the photovoltaic surface to obtain a NZEB, the use of a GSHP instead of a condensing boiler coupled to a chiller leads to reductions of 92% for Palermo and 61% for Milan. These reductions are 5% for Palermo and 10% for Milan when comparing the GSHP with a more efficient technology such as the air source heat pump. The obtained results in the context of the NZEB is very important, as in highly urbanized contexts the space to allocate the production energy systems from renewable sources is often small and even 10% of surface savings can make the difference to obtain the NZEB target. In fact, the results showed that in the case of a continental climate such as Milan, in which the winter energy demand is predominant, it is not always possible to satisfy the NZEB target due to insufficient roof surface.
Beyond satisfying the energy balance, an energy demand limit value was considered to classify a building as an NZEB: it is suggested by the New Building Institute of the U.S. and is equal to 57 kWh/m 2 . If this limit value is considered, the building in question can be considered an NZEB for both climatic conditions only by using the GSHP.
Finally, referring to the energetic-economic analysis, the discounted payback (DPB) period for the GSHP is about 3 years for Milan and 14 years for Palermo when compared to the configuration with boiler plus chiller, while it is more than the useful life when compared to an air source heat pump. Therefore, the air source heat pump seems more convenient than the GSHP from an economic point of view. However, it should be noted that the air source heat pump, especially in cold climates (Milan), can be affected by malfunctions in the case of very low external temperatures, when it may sometimes not satisfy the user.
Clearly, for the analysed cases, in order to obtain an NZEB, the ground source heat pump (compared to the common configuration with a condensing boiler coupled to a chiller and to the solution with the air source heat pump) has the highest performance from an energy point of view, but is still too expensive. Therefore, from an economic point of view (discounted payback period), it is really convenient only for continental climates (Milan) when compared to the solution with a condensing boiler coupled to a chiller.

Conflicts of Interest:
The authors declare no conflict of interest.