The results of the simulations provide insights on energy demand, which can be used as an input for economic and environmental analyses.
Section 3.1 presents the resulting energy demands and peak loads, both for heating and cooling. The results of plant sizing and of energy demand were used to estimate the installation and operational costs, and hence to assess the economic feasibility of shallow geothermal energy, with a focus on the Italian situation (
Section 3.2). Finally, the environmental benefits are assessed, estimating the reduction of fossil fuel use, of CO
2 emissions and of air pollution (
Section 3.3).
3.1. Energy Consumption
The peak heating/cooling loads and the energy demand are the key input parameters for sizing the HVAC components. To make results comparable among buildings of different size, results are shown in
Figure 4 as heating/cooling peak loads per unit area (W/m
2). Heating loads are generally higher than cooling ones in European climates, especially in poorly-insulated buildings. This is the case of the Office building (
Figure 4A,B), with values 30% to 130% higher than the other building types due to the Monday morning start-up, and of the high cooling loads (up to seven times the Hotel loads,
Figure 4C,D) due to the high internal gains. The Hotel has lower peaks since the HVAC system operates with a longer schedule (see
Figure 4) due to higher comfort standards. A better thermal insulation reduces heating loads by about 50–70% and, in warm climates, also cooling loads. On the other hand, it slightly increases cooling loads in cold climates, as the internal gains increase their relevance in the heat budget, especially in the Office case. With a good insulation, the difference between heating and cooling peaks diminishes (especially in Zones B and D) thus allowing for a better exploitation of the reversible heat pump.
The energy demand for heating and cooling is expressed per unit area (kWh/m
2/year) and is closely correlated to the heating (
Figure 5A,B) and cooling (
Figure 5C,D) degree days, thus confirming the validity of the heating/cooling signature approach [
51].
Similar to the thermal loads, the energy demand is strongly influenced by the thermal insulation, and switching from poor to good insulation can reduce it by 70–90%. The heating signatures of poorly insulated buildings (
Figure 5B) are of course characterized by steeper slopes, as well as by a higher intercept, since building heating is required even in warmer conditions. The energy signature of the Hotel has a gentler slope due to the smoothing effect of the uninterrupted schedule but, at the same time, the higher comfort requirements determine a high intercept, because heating is switched on at warmer conditions to cope with a higher setpoint compared to the other buildings. The well-insulated Office requires some cooling even in the coldest climates with low or zero CDD (
Figure 5C), due to the high internal gains. On the contrary, the House typology has relatively low cooling needs due to the low occupancy in the warmest hours of the working days.
The heating demand (
Figure 5A,B) depends more on climate compared to the peak heating load (
Figure 4A,B): the ratio between these quantities is the number of full-load equivalent operation hours (FLEH) and indicates how intensely is a heating (or cooling) plant used. Zones A and C have both a low heating demand and low FLEH, which limit the economic feasibility of installing a GCHP (
Section 3.2). A small difference between the annual heating and cooling demand is highly desirable, since it reduces the long-term thermal alteration of the ground, and it is observed mostly in well-insulated cases in the climate Zones B and D. In the other cases, the heat budget unbalance induces a ground thermal drift, reducing the system performance over time.
Domestic Hot Water (DHW) is not considered in the heat pump sizing, because it is stored and not produced instantaneously, and hence the required power is much lower than any building peak heating load. However, the energy demand for DHW may represents up to 80% of the consumption in well-insulated buildings in warm climate zones, due to the low heating needs. In the other cases, its value ranges between 24% and 1% of the total heating demand.
The time series of heating and cooling load allow to derive cumulate distributions, such as the one shown in
Figure 6, which correlates the load factor of the heat pump with the total heating/cooling demand met (TEDM), on an hourly basis. The results of all simulations are comprised among the thick black lines. For example, a heat pump sized at 60% of the peak load is able to meet 82–96% of the total yearly heating demand, consistent with a few previous studies [
31,
32,
52].
In the hybrid cases analyzed below, the HP is installed to cover the base demand, while a backup system is used to manage the peaks. The plot shown in
Figure 6 can be used for sizing the HP and the auxiliary boiler at different shares of TEDM.
The installation cost of the system with a hybrid configuration can significantly be cut due to the reduced HP capacity and BHE length. Additionally, a reduction of the frequency of on–off cycles is achieved, which could extend its operating lifetime [
43].
The efficiency of the HP strongly depends on climate and it is usually expressed by the seasonal performance factor (
), i.e., the load-weighted average COP during heating (
) or cooling (
) seasons. As shown in
Figure 7, the heat pump performs better for heating in warmer climates, while the opposite occurs for cooling. This behavior is attributable to the undisturbed ground temperature (here assumed equal to the annual mean air temperature).
Of course, as the underground temperature can be locally altered (e.g., by geothermal gradients) these conclusions may not apply in all cases.
3.2. Economic Performance
To compare a conventional heating/cooling system and a GCHP system, an economic analysis was carried out, considering either a new installation (in well-insulated buildings) or a replacement of the existing heating/cooling system (refurbishing of poorly-insulated buildings). This analysis allows distinguishing profitable investments from ineffective ones, based on three economic indicators: the net present value (NPV), the discounted payback period (DPP) and the internal rate of return (IRR) [
53]. It is straightforward that, at least, NPV should be positive at the end of the evaluation period of the investment. The net present value is calculated as:
where
(€) is the total investment cost,
(€) is the operational saving granted by the GCHP compared to the conventional system over the plant lifetime,
(€) is the difference between the total operation and maintenance cost difference of the GCHP and the conventional plant over their lifetime,
(€) are the total subsidies supplied,
is the discount rate,
is the effective discount rate [
53],
(year) is the lifetime of the GCHP system (assumed equal to 25 years [
54,
55]),
(year) is the number of years during which subsidies are provided.
The discount rate
represents the cost of capital and was assumed equal to 2% as suggested by the European Central Bank [
56]. The rate of increase of energy cost
was set to 2% same as the discount rate
but, of course, it can have different values. The installation cost,
, is the sum of the components’ costs including installation labour and taxation.
The equipment required for both conventional and new plants were not considered in the analysis; thus, AHU and fancoils were assumed to be already installed in the Office and Hotel, and their installation cost was not calculated. Conversely, fancoils were considered, along with the GCHP installation, in the House refurbishment case, where they are supposed to replace high temperature radiators.
The capacity-cost curves for each piece of the system were built with data derived from commercial catalogues [
57,
58,
59,
60,
61] and are reported in the
Supplementary Materials (Table S15), while the BHEs drilling and installation cost were assumed constant and equal to 60 €/m [
62]. After defining the components’ cost, the total investment cost of the plant was obtained by summing the single components’ costs and the labor cost for plant installation. Tax was added to the final cost considering the Italian taxation system (Value Added Tax, VAT) on the labor and equipment component of the installation costs [
63]. VAT rates on equipment (normal rate) range between 18% and 27% in the EU-28 Member States [
64], while different kind of rates could be applied to the labor costs.
The share of each factor contributing to the total cost of the system varies in the different cases. In the example of the 28 kW poorly-insulated House in Zone D, the total cost is distributed as follows: drilling and BHEs installation (44%), HP (19%), fancoils (9%), labor (6%), taxes (17%), and additional equipment (5%). In general, the cost related to boreholes represents the largest contribution to the total installation cost, with a logarithmic correlation with the BHEs total length installed as shown in the
Supplementary Materials (Figure S4).
Data obtained from the simulation were processed to calculate the annual electrical energy consumed by the heat pump and the circulation pump of the BHEs. The energy required for both the conventional and the GCHP system was not considered in the analysis (e.g., distribution pump, AHU, etc.). The price for the energy resources were extracted from Eurostat data [
65].
The annual revenue achieved with the installation of a GCHP was calculated as:
where
and
(kWh) are the building heating and cooling energy determined by the model simulation;
and
(€/kWh) are the natural gas and the electricity price, equal to 0.0719 €/kWh and 0.1818 €/kWh, respectively [
66];
is the conventional non-condensing gas boiler efficiency, equal to 76% [
41,
46];
is the seasonal coefficient of performance of the conventional air conditioning system obtained from the model simulations (for each climate zone) and based on NREL tests [
67]; and
(kWh) is the electrical energy consumed by the system calculated through simulations, considering all the components related only to the GCHP (i.e., excluding the distribution and terminal auxiliaries, since they are also included in conventional HVAC systems).
The O&M costs represent the annual expenditures for ordinary maintenance of the plant and are calculated as the difference between the GCHP and the conventional system maintenance costs, based on Refs. [
54,
55,
68].
Finally, two different scenarios were considered: absence of subsidies, which represents the case of a GCHP installation in a brand-new building, and the presence of subsidies, in the case of the refurbishment of an existing building. Subsidies were assumed equal to 65% of the total investment cost, refunded in ten equal yearly payments (
), according to the Italian incentive scheme [
63]. For this reason, while the evaluation on the unsubsidized case could be extended with a good degree of approximation to other EU countries, the subsidized case only applies for Italy.
The discounted payback period is the time required for the NPV to become equal to 0 [
53], while the internal rate of return represents the discount rate for which the net present value is null at a certain period of time [
69]. The calculated IRR corresponds to the one evaluated at the end of the plant life (i.e., after 25 years).
The GCHP system is profitable if IRR is sufficiently high, NPV is positive and DPP is not too long. According to Newnan et al. [
69], 2% is the lowest acceptable IRR for investments in energy efficiency, but companies are more focused on the short term, and hence a higher IRRs are usually requested to opt for a certain investment [
70].
Table 6, the feasibility of the different cases, with or without subsidies, is reported: in the case of negative NPV and IRR the system is not economically feasible; if the IRR is lower than the acceptable rate of return (2%), the investment should be carefully evaluated; finally, if the IRR is acceptable and the NPV positive the system installation is feasible.
In general, in the current European economic scenario, GCHP systems designed to cover the whole heating demand are unfeasible without public subsidies. Replacing an existing system in a poorly-insulated building is economically convenient due to the high energy consumption, which implies high possible revenues. This holds true in particular for hotels due to the higher number of operation hours per year. As shown in
Figure 8, the discounted payback periods (DPP) are similar throughout the heating-dominated incentivized cases with a general decrease toward high energy-consuming buildings (i.e., scarce insulation cases and cold climates) and when the full load hours are maximized and the load curve is characterized by low peaks, because of both lower installation costs and higher operational savings. DPP of 9–13 years can be achieved in most of the refurbished buildings displayed in
Figure 8. Among these, climate Zone B (in poorly insulated cases) presents balanced heating and cooling needs, and the shortest DPP (i.e., 8.6 years). Cooling-dominated buildings in Zones A and C are, instead, low energy-consuming, which results in very long payback periods. This is particularly true when subsidies are not available (generally the case for new buildings), with a minimum payback period of 16 years.
3.2.1. Hybrid System Results
The energy and economic analysis was extended to the hybrid cases. The building load curves were used to size the HP capacity to cover 90% or 70% of the annual heating demand, while the share is assumed to be covered by a gas boiler.
The estimation on load factor–TEDM curves such as in
Figure 6 proved accurate for the House (
Figure 9), with a slight underestimation of the TEDM by the backup system (3% average difference, see
Table 7). By contrast, the predicted energy covered by the boiler was highly overestimated for the Office and the Hotel (13% difference on average, see
Table 7), meaning that an ad hoc simulation process with a trial-and-error approach is required for sizing. The peak loads are reduced by the thermostat dead band and by the buffer tank, and, thus, the loads based on the building requirements are overestimated compared to the real ones.
These results suggest that an estimation of the hybrid backup system based on the building load curve is conservative for the Office and Hotel buildings, while it is correct in the House case as shown in
Table 7. In the example of the poorly-insulated Hotel in Zone D (HP covering 100% of the peak load), the payback period is equal to 18.4 years without subsidies and 9.9 years with subsidies. When sized at 90% TEDM (54% load factor), the system actually covers 97% of the heating demand, with a payback time of 14.2 years without incentives (23% reduction) and 8.6 years with incentives (13% reduction). In the 70% TEDM sized case (36% load factor) the system actually covers 86% of energy demand, with a payback equal to 12.9 years without incentives (30% reduction) and 7.3 years with incentives (−26%). Summarizing, the introduction of a backup system can reduce the payback period of the system by up to 20–40% thanks to the reduction of the HP and BHE costs. The investment becomes profitable for the Hotel cases even without subsidies. Further studies should optimize the economic output of a hybrid system and provide a reliable estimation method based on the building load curve.
3.2.2. Electricity/Fuel Price Ratio Analysis
An additional analysis was conducted on the electricity/fuel price ratio and on its effects on the economic indicators. In general, the price ratio should be lower than the heat pump SPF/boiler efficiency ratio, otherwise no saving can be achieved on the operational costs. This request is always met in the simulated scenarios, since the lowest SPF/boiler efficiency ratio equal to 4.1 (poorly-insulated Hotel in Zone F) and the electricity/fuel price ratio (
Table 8) never reaches such a high value. For this reason, the GCHP allows achieving some operational saving in all 36 cases analyzed. The ratio was evaluated in the base case of natural gas combustion (with and without the actual taxation share) in the Euro Area (EA) and in different countries (Italy, Germany, Spain, and France). A further comparison with heating oil and heating LPG [
66,
71] is made for Italy.
Figure 10 shows that the Discounted Payback Period grows exponentially with the ratio between electricity and the fossil fuel considered. In the present European price scenario, electricity is relatively expensive compared to natural gas, leaving little room for heat pump-driven revenues. This is partly due to the different taxation regimes, with a heavier taxation on electricity. If the GCHP replaced or prevented the installation of oil/LPG boilers (i.e., in areas not connected to methane pipelines), the electricity/fuel price ratio would be much lower, enabling profitable investments. The comparison between different countries shows that (geothermal) HPs installation in Germany is greatly penalized by the low gas price, while the low electricity price in France favors GCHP implementation.
3.3. Environmental Benefits
The environmental benefit of GCHP systems with respect to conventional systems was assessed, based on the simulation results, considering the non-renewable primary energy saved and the total CO
2 emission avoided. Besides the efficiency of the GCHP system, the energy and environmental benefits mainly depend, respectively, on the non-renewable energy factor and the CO
2 emission factor of the national grid. In Italy, the primary non-renewable energy factor of electrical energy is equal to 1.19 (ISPRA, 2017 [
72]), while the emission factor for electricity (424 g CO
2/kWh) and gas (240 g CO
2/kWh) according to the Joint Research Centre (JRC) (2017, [
73]).
In general, GCHPs reduce the primary energy consumption by 33–75% and CO2 emissions by 27–56% relative to conventional heating (gas boiler) and cooling (air chiller) systems. These two indicators are similar throughout the cases and mostly depend on the full load equivalent hours (FLEH) of the GCHP system operation, with lower results in cooling-dominated buildings, as the energy efficiency improvement given by GCHPs is higher in heating mode.
The effectiveness of investing in GCHPs to fight against climate changes can be expressed by the amount of avoided CO
2 per year per Euro invested (
Figure 11). With the aforementioned input data for the Italian case, this indicator ranges between 13 and 216 g CO
2/€/year for the simulated cases, and it is proportional to the full load equivalent hours (FLEH) with a reduction 7.5 g CO
2/€/year every 100 FLEH (
). This indicator strongly depends on the CO
2 emission factor of electricity, which varies depending on the source (if it is produced on site, e.g., with a PV plant) or, if it is taken from the grid, depending on the country. GCHPs global emissions are expected to reduce with time, along with the CO
2 emission factor: for example, according to JRC, the Italian grid CO
2 emission factor has fallen from 654 to 424 g CO
2/kWh from 1990 to 2013 [
73], and this reduction (−35%) is reflected on the CO
2 emission factor of heat pumps, too.
The environmental performance of hybrid systems is slightly worse than GCHP systems, due to the consumption of natural gas to manage peak loads, but effectiveness of subsidies is increased due to the lower installation cost. Considering 90% and 70% of total energy demand met by HP, the primary energy and CO2 saved decrease as the share of the backup coverage increases; however, the CO2 avoided per Euro spent is 30% higher (in the 90% TEDM case) and 64% higher (in the 70% TEDM case) on average. This means that the GCHP-hybrid configuration is especially convenient to achieve the maximum results of incentives in terms of climate change mitigation.
The analysis on CO
2 emissions performed for Italy was extended to the other 27 countries of the European Union, based on an emission assessment performed by the EU Joint Research Centre on the CO
2 emission factor of the national electrical grids, with two methods (standard IPCC and LCA). The emission factors of electricity in EU-28 countries range from 15 g CO
2/kWh of Sweden (38 g CO
2/kWh with LCA) to 1977 g CO
2/kWh of Estonia (2017 g CO
2/kWh with LCA). Italy is close to the EU-28 average (respectively, 343 and 393 g CO
2/kWh with the standard IPCC and the JRC LCA method). The full dataset is available in
Table S16 of the Supplementary Materials, along with the calculation of the CO
2 reduction achieved by a GCHP compared to a conventional HVAC system (gas boiler and air-source chiller), for which the minimum and the maximum reductions are shown. For highly carbon-intensive grids (i.e., higher than about 800 g CO
2/kWh) such as in Bulgaria, Cyprus, Czech Republic, Estonia, Greece, Malta, and Poland, the use of GCHP can even lead to higher carbon dioxide emissions compared to a gas boiler. In the detail for each case study (
Figure S6 of the Supplementary Materials), we observe that, for an average (Italy) and low (France) carbon emission factor, the highest reductions are achieved in the coldest climates (climate Zones D, E, F and, to a lesser extent, B, according to the classification by Tsikaloudaki et al. [
33], since the main benefit is achieved by heating with the electrical heat pump, compared to gas. On the other hand, if electricity has a high emission factor (e.g., Poland), the GCHP in heating mode becomes more carbon-intensive than the gas boiler; in this case, a global CO
2 emission reduction could be achieved only in the case of a cooling-dominated building, since both air-source chillers and GCHPs use electricity, and GCHPs are more efficient.
A further benefit of heat pumps (both ground and air-source ones) is related to air quality, which is a major issue especially for metropolitan areas. Heat pumps do not release pollutants in the air as they only consume electrical energy, usually generated elsewhere. The benefit compared to natural gas boilers on the air quality is rather limited, except for NOx [
74], but it is much higher when compared to oil boilers (with high SOx emissions). The need to reduce the dependence on fossil fuel, however, makes it necessary to increase the share of renewable heat production. To date, the most used renewable heat source in Europe is still wood biomass in the form of wood logs, chips or pellets [
75]. Their impact on particulate matter (PM
10 and PM
2.5), NOx and COV [
76,
77] has become hardly sustainable both in urban areas (e.g., Thessaloniki [
78]) and in larger regions (e.g., the Po Plain in Italy [
79]). Moreover, the supply chain of wood logs and pellets and hence their global environmental impact are critical, as demonstrated in various LCA studies [
80,
81]. If compared to biomass boilers, the electric energy production related to GCHP emits significantly less NOx (80–94%) and almost eliminates the particulate. The Fraunhofer Institute [
82] studied the externality cost linked to NOx and PM
10 emissions of biomasses, estimating externalities of 6 c€/kWh and 3 c€/kWh associated to wood log and pellet combustion, respectively. The internalisation of such costs would make biomasses more expensive of natural gas and, a fortiori, of heat pumps.