The paper analyzes a solar photovoltaic power plant serving an office building located in southern Italy. Space heating and cooling building demand is satisfied by a ground source heat pump partly activated by electric energy available from photovoltaic panels. Thermo-economic performance of this system varying photovoltaic peak power (4.5-7.5 kW) and electric battery capacity (3.2-9.6 kWh) is carried out by means of dynamic simulation software. The proposed system achieves primary energy saving and equivalent dioxide carbon emission reduction up to about 97 % in comparison to the reference conventional system based on a natural gas fired boiler and an electric chiller. The results show that the solar energy system is more competitive when there is no battery storage and government incentives will be provided.
An increasing interest in the reduction of primary energy requirements and related greenhouse gas emissions due to typical energy demands (space heating and cooling, domestic hot water) of residential and tertiary buildings is one of the main aim of European Union also sustained by the introduction of renewable based energy conversion systems [1]-[3]. Space heating and cooling demands in residential and tertiary reached 31.2% of final energy demand in 2012 in European Union, corresponding to about 294.8 Mtoe [4]. Final energy demand for space heating and cooling in these sectors achieved ~39.3 Mtoe in Italy in 2013, corresponding to about 33.1% of the total final energy demand [5]. To reduce primary energy demand, in the last years great efforts were focused in the introduction renewable based energy conversion systems. In particular, solar thermal heating and cooling (STHC) plants [6][7] and solar electric (photovoltaic, wind, etc.) heating and cooling (SEHC) systems interacting with an electric heat pump (EHP) or a ground source heat pump (GSHP) are nowadays under the spotlight. About 18.9 GW of photovoltaic (PV) panels were installed in 2015 in Italy, delivering 22.9 GWh. This value represents the 8.4% of global Italian electricity demand [8]. Only 18.6% of electricity available from PV was self-consumed, while the surplus was exported to the grid [9]. These data show the need to promote technologies able to increase the self-consumption of electricity. Electricity surplus from distributed PV, as well as high energy demand for space heating and cooling in buildings, have led in Italy to an increasing interest both in the introduction of GSHP and electric storage to reduce electricity exported to the grid.
In Ref. [10] there is an energy assessment by means of TRNSYS software of solar-based technologies, such as solar PV panels and solar thermal collectors, coupled with a GSHP able to satisfy domestic hot water, space heating and cooling demands for buildings located in different Italian climates (Milan, Rome and Palermo). In particular, PV plant with a peak power of 4.4 kW integrates an electric storage of 0.64 kWh, while the nominal capacity of GSHP is 12.6 kW in cooling mode and 10.5 in heating mode. The total length of the borehole heat exchanger (BHE) for Rome and Palermo is 500 m. COP (Coefficient Of Performance) and EER (Energy Efficiency Ratio), on seasonal basis for Rome, are 5.9 and 4.0, respectively.
Franco et al. reported in Ref. [11] an experimental analysis of a PV plant (3.7 kW of peak power) and a GSHP used to meet the electric and thermal demand of a single-family residential Italian building of 160 m2 and an indoor volume of about 450 m3 located in Pisa. GSHP has a nominal capacity of 14.4 kW (COP = 3.8) in heating and 11.4 kW (EER = 3) in cooling mode, while the total length of BHE is 240 m. In real operating conditions the COP on daily basis could be lower than 2, value that is very far from nominal one. Similarly, EER ranges between the nominal one and values also below 1. The lower values are obtained in the days characterized by a reduced operation of the GSHP. The analysis aims to find a possible configuration in the field of nearly zero-energy building. The experimental analysis of the systems shows a strong interaction with external electric grid due to differences between availability of electric energy from PV and electricity load. The data analysis shows that the introduction of an electric storage could reduce electricity sent to the grid.
The paper [12] analyzes the energy, exergy and environmental performance of a GSHP that is coupled with a PV plant to meet heating and cooling demands of a net zero-energy residential building. The effect of different climates of three cities of Iran (Isfahan, Yazd, and Shahrekord) on system performance is investigated. It is assumed that the building characteristics and its area equal to 200 m2 is identical for all case studies, while the cooling and heating loads, solar irradiation, and ground depth temperature are different. Exploitation of renewable sources leads to low exergy efficiency of about 10%, since solar and geothermal energy conversion facilities are considerably exergy destructive. However interesting primary energy saving in fossil fuels and reduction in pollutant emissions are observed.
In [13] there is an analysis of PV plant with a peak power of 4 kW interacting with different heat pump systems serving single-family Canadian houses located in Toronto and Vancouver. This energy model was built in TRNSYS in order to compare the performance of three distinct heat pump types, ranging from a conventional air source heat pump to a GSHP with limited thermal storage capacity. GSHP has a nominal heating capacity of 3.5 kW while the single BHE has a depth of 89 m in Toronto and 48 m in Vancouver. PV generated electricity for GSHP based system is mainly exported to the grid (55-57%) for both the locations and the self-consumption can cover only a small amount (12-14%) of the global electric demand. This is because peak PV generation (mid-day, summer months) does not meet the demand in the houses.
Thygesen and Karlsson in [14] performed an energy and economic analysis by means of TRNSYS on a PV-system with a GSHP serving a single residential building located in Västerås (Sweden) for space heating and DHW purpose. There is no battery storage in the proposed system. The nominal heating capacity of the GSHP is 5.8 kW with a depth for BHE equal to 150 m while the PV peak power is 5.19 kW. The paper reports a sensitivity analysis considering different peak powers up to 6 kW and net-metering schemes (on instantaneous, daily and monthly), also assuming the electric grid as a virtual electric storage. Solar fraction improves moving from instantaneous to monthly basis net-metering, as well as economic results. Only the introduction of daily and monthly net-metering gives positive economic feedback.
The same authors in [15] analyze on energy and economic basis a system that differs from the previous for the addition of an electric storage aiming to increase the self-consumption of renewable electricity. The total capacity of the lead acid batteries is 48 kWh with a depth of discharge limited to 50%. The introduction of the electric storage increases renewable electricity self-consumed from 56% to 89%. A sensitivity analysis shows that the most efficient battery system must be smaller than 10 kWh for the proposed system. The systems with a battery storage appears not profitable if the investment cost is higher than 50 € per kWh of battery capacity, value that is far away from actual market cost.
A further study was proposed by Thygesen and Karlsson in [16], where a TRNSYS based simulation is performed on a PV system with a peak power of 5.29 kW and a GSHP with 3 kW of heating capacity (COP = 4.29) used to meet space heating and DHW demand of a Swedish single-family building. This system is without an electric storage and has a BHE with a depth of 95 m interacting with the GSHP. The authors introduced a new GHSP controller based on weather forecast aimed to increase the PV electricity self-consumption. The self-consumption of renewable electricity increases from 56%, base case, to 63-64% in the second case in terms of self-consumption of PV electricity . The authors stated that introducing a weather forecast methods to activate a GSHP for Swedish climate is not interesting, not only because of small increase in self-consumption, but also because the controller is unprofitable, due also to a small increase in total electricity usage of the building.
Furthermore, the introduction of a GSHP could be very interesting solution in limiting urban heat island in cities. Urban heat island is very common and well documented problem and extensive studies are available for most of the major cities in the world, [17]. In particular, in cooling dominated zones, such as that one considered in the case study presented in this paper, the increase of the cooling demand due the urban heat island is much higher than the corresponding decrease of the heating demand for all types of studied buildings. In this way for space heating and cooling purpose a ground source heat pump instead of air source heat pump or natural gas boiler and chiller could be an interesting solution to this problem.
The previous literature survey showed that existing works deal with simulative or experimental analysis of solar heating and cooling plant based on ground source heat pump interacting with a PV field. The previous analysis is limited only to residential applications considering single-family houses located in different countries with PV interacting with a GSHP used to satisfy DHW, space heating and cooling demands. To the best authors’ knowledge, in scientific literature there are no data on office application of GSHP-PV based system. The analysis was performed on annual basis, in terms of energy, environmental and economic analysis. An electric storage was introduced, with the aim to reduce the electricity exported to the grid. An interesting approach to the diffusion of renewable energy is represented by the attempt of promoting the systems able to increase the self-consumption limiting the impact on the electrical grid. Furthermore, the paper reports a sensitivity analysis to evaluate the effect of different parameters, such as PV peak power, electric storage capacity, unitary electricity and natural gas costs, on the diffusion of the proposed configuration.
In this work an office building characterized by a flat roof, one floor, 200 m2, 600 m3 with 13 working persons is considered. The heating and cooling terminal units are fan-coils. The office occupancy is 8 hours per day between 9:00 and 13:00 in the morning and between 14:00 and 18:00 in the afternoon during weekdays, while in the weekend the building is empty. Seated persons with very light writing as degree of activity are considered [18]. The building envelope characteristic, reported in Table 1, respects Italian legislation [19] that imposes limitations on transmittance for building renovation. The office location is Naples, characterized by1034 heating degree days (HDD) and a latitude of 40°51.3786′ N.
According to Italian legislation the heating period for Naples is from November 15th to March 31st , with a set-point of air temperature room of 20.0 °C (+/-0.5 °C). The heating system works between 8:00 and 18:00, in a work day, while during weekend is turned off. The cooling period is from June 1st to September 30th with the same hours occupancy cooling and internal gain of heating period, while the desired temperature is 26.0 °C (+/-0.5 °C). In Figure 1 heating and cooling loads on monthly basis, including internal gains (persons, artificial lights, solar radiation, personal computers, monitors, printers, etc.) and loads due to air infiltration and ventilation, are reported. Space heating and cooling demands are 2920 kWh and 6453 kWh, respectively. The domestic hot water demand is considered negligible with respect to the heating demand.
Excluding HVAC (Heating and Ventilation Air Conditioning) requirements and according to an on-site analysis performed on electricity required by office buildings [20], electric energy is evaluated taking into account a demand of 29.64 kWh/m2 per year for small power equipment present in the buildings (PC, monitor, printers, etc.) and 11.74 kWh/m2 annum for artificial lights. The electric load profiles excluding HVAC requirements, Figure 2, are considered for three type days (heating, cooling and intermediate season) for weekdays, and only one for weekends.
In order to satisfy electric, heating and cooling demand of the office building a solar based system is here considered as proposed system (PS). The energy conversion system is based on a PV field, an inverter (INV), an electric storage (BAT) and a GSHP interacting with a borehole heat exchanger. The PV system covers both electric requirements of GSHP and end user. The system is grid connected and interacts in bidirectional way with external grid and electric battery.
The existing energy conversion system (CS, Conventional System) is based on:
• a natural gas fired boiler (B) in heating period with thermal power of 24.0 kW and thermal efficiency, , of 90.2 %;
• an electric activated chiller (CH), with a cooling power of 13.3 kW and an EER equal to 3.0;
• electric grid satisfies electric demand of the user (chiller, lighting, etc.). It was considered the average Italian efficiency () equal to 65.5 %, value that includes the thermo-electric power plant mix, renewable contribution and also transmission and distribution grid losses [21].
In Figure 3 and in Figure 4 heating and cooling plant configurations both for proposed and conventional systems are reported.
Three different PV peak powers (4.5 W, 6.0 kW, 7.5 kW) with panels facing south with a tilt angle of 31° were considered. The main PV panels characteristic at standard test conditions (STC) are shown in Table 2 [22].
On Table 3 the inverter characteristics, considered for different peak powers, are shown [23].
The main data of the Li-Ion batteries are reported in Table 4 [24]. In the following analysis a DOD (Depth Of Discharge) equal to 90% is considered.
A ground source heat pump is used to cover space heating and cooling demands, Table 5 [25]. The GSHP delivers at nominal conditions, [26], 15.9 kW of heating power with a COP of 3.90, and 13.7 kW of cooling power with an EER of 4.36.
The borefield was designed in accordance with the ASHRAE approach [27] evaluating the borehole length in heating and cooling operating modes by means of Eq. 1 and Eq. 2:
To determine the ground heat transfer rates in cooling and heating mode and net annual heat to and from the ground, the following equations were used:
where and are the nominal cooling and heating power of GSHP, respectively, HY are the hours per year (8760), EFLFc and EFLFh are the equivalent full-load cooling and heating hours, respectively. The latter have been evaluated according to literature data [27]. It was taken into account a borehole with a diameter of 0.125 m, backfilled with bentonite/cement mixture (borehole fill conductivity of 1.4 WmK-1), with shank spacing of 0.075 m and borehole separation of 6 m. It was considered a vertical U-tube with a diameter of 0.032 m with a pipe of high-density polyethylene. A ground conductivity and diffusivity of 2 WmK-1 and 9.210-7m2s-1 were accounted corresponding to the average thermal properties of normal rocky soil [27]. Table 6 reports the remaining parameters used to evaluate the bore length (LBHE). The values of effective thermal resistances of the ground (Rga, Rgst and Rgm) were evaluated according to [27]. The value of SCF has been chosen considering a ground-loop differential temperature of 3 K (see and in Table 6). The value of PLFm is corresponding to January as design month, while the value of Tpp was reported by [27]. The required bore length (LBHE) is 205 m for heating mode and 270 m for cooling mode: thus, it was assumed a borefield made up by 3 boreholes each one with a length of 90 m.
Table 6. Parameters used to evaluate bore length.
TRNSYS software [28] was used for the analysis of thermo-economic performance of proposed and conventional systems. This software is commonly used for transient simulation, to evaluate the interaction between buildings and energy conversion systems. Each component is available in TRNSYS libraries [28][29] and is modeled by subroutines (called “types”). These components can be linked between them to create very complex systems. The models considered for the main components of the above reported systems are briefly described in the following. PV modules are modeled using type 94 that is based on equations representing an empirical equivalent circuit model able to predict the current-voltage characteristics of a single module. The model is based on a “four-parameter” equivalent circuit considering PV data of manufactures [30][31]. The electric storage and its charge controller are simulated using type 47 subroutine. Any power from PV panels exceeding electric user demand is used to charge the battery or to directly feed the grid if it is completely charged. The inverter is simulated by type 48 that takes into account both the direct current available from PV modules and the electric load of the user in order to manage the interaction with the battery and the electric grid. GSHP and CH are simulated using type 927 and type 655 [32], respectively, that are based on a performance map of the devices. Natural gas boiler is modeled with a simplified subroutine, type 6, based on constant thermal efficiency.
BHE is simulated using type 557 which models a vertical heat exchanger that interacts thermally with the ground. Eventually, the building is simulated by type 56 that models the thermal behavior of a building having multiple thermal zones. The fan coils are modeled by means of type 928 that models an air to water heat exchanger delivering heating and cooling energy to the building [33].
In this paragraph an analysis of the performance of PV plant is reported. A further study is performed comparing the solar based system (proposed system, PS) with a conventional system (CS) by means of energy, environmental and economic analysis.
5.1 Proposed system analysis
An analysis of electric energy requirements on annual basis considering PS and CS is reported on Table 7. Total electric demand depends on small power office equipment, lighting, GSHP and HVAC auxiliaries (fan, circulating pumps, etc.) and the table shows that there is a small difference between proposed and conventional systems. Total electric energy, including also GSHP, required by end user for PS is 11.3 MWh corresponding to 56.5 kWh/m2 per year.
One of the key element of the performance system is the electric efficiency of solar system, that depends on PV panel, inverter efficiency and other BOS (Balance Of System) losses (dirt, reflection, cell temperature, wiring, mismatch, etc.). It reaches on annual basis about 14.6 % for 4.5 kW and 6.0 kW while is little higher (14.7 %) for 7.5 kW, without considering storage batteries. The efficiency reduces up to 14.4 % (4.5-6.0 kW) and 14.6 % (7.5 kW) including batteries with a storage capacity of 9.6 kWh. The performance of the GSHP, that includes the electric requirements of the circulators, on seasonal basis are 3.35 for COP and 4.14 for EER.
Figure 5 reports on annual basis for each PV power and electric storage capacity configuration the distribution of electricity associated to PV plant and electric grid. Electric energy available from PV is partly used by end user (blue bar) and partly exported to the grid (green bar), while the integration from the grid is always present (red bar). Increasing the size of PV plant the electricity availability increases both for self-consumption and export. The introduction of electric storage leads to increase the self-consumption of renewable electricity.
One of the main problems related to PV plants is to limit the exported electricity to the electric grid. In this way it could be important to evaluate, on the basis of electric energy required by end user, the best configuration characterized by a low percentage of exported electricity. Two indexes are introduced to highlight the self-consumption of PV electricity [34]:
• s: ratio between electric energy supplied through the inverter by PV or, if available, by batteries to end user and the total one required;
• d: ratio between electricity delivered by PV through the inverter by PV or, if available, by batteries to end user and the total one available from the inverter.
Renewable electricity covering end user demand, represented by s index, increases with PV peak power and battery capacity achieving the best result of 69.7 % for 7.5 kW and 9.6 kWh. It means that with this configuration only 30.3% of the final energy demand of the building is satisfied by external grid. For this last configuration, Figure 7 shows on monthly basis how renewable electricity is used and also the contribution from external grid. The months characterized by the highest percentage of renewable electricity covering end user load are in the intermediate period (April, May, October), due to low electric demand, and in cooling period (June to September), due to high electric energy availability from PV.
The fraction of PV electricity self-consumed, with respect to global production (index d), decreases with PV size while increases with battery storage. The solution characterized by the lower percentage of exported electricity, equal to 16 % (d = 84 %), is that one based on 4.5 kW and 9.6 kWh. In Figure 8 is reported, for the configuration characterized by the highest d value (4.5 kW, 9.6 kWh), the distribution on monthly basis of electricity available from PV, self-consumed and exported, and imported from grid. It can be noted that in heating period a small amount of electricity is exported while during intermediate and cooling period the percentage of electricity exported increases even if it appears limited.
A further possibility, aiming the reduction of electricity exported to the grid, could be the introduction of a mobile storage battery that could be represented by an electric vehicle [35].
5.2 Energy, environmental and economic analysis
The solar based system used to satisfy electric, heating and cooling demand of the office building is compared in terms of energy, environmental and economic performance indexes with a conventional system that consists of electric grid, a natural gas fired boiler and a chiller, Figure 3.
The energy performance of the proposed system has been evaluated by means of Fuel Energy Saving Ratio (FESR) index that compares the primary energy input related to fossil fuel of PS () and CS (), and is defined as:
Primary energy of CS, as stated in Eq. (7), depends on electric energy required by chiller, , pure electric load (lighting, appliances, etc.), , thermal energy satisfied by natural gas boiler, , and reference efficiency parameters (, ). A similar equation could be considered for PS, even if it depends on the electric energy drawn from () and exported to the grid (), that is considered as a credit for primary energy evaluation. FESR has a negligible dependence from electric storage size and increases with PV peak power due to the greater availability of renewable electric energy, Table 8. It ranges between about 63 % for 4.5 kW and about 97 % for 7.5 kW.
Environmental analysis is focused on a simplified approach based on the evaluation of equivalent carbon dioxide emissions (CO2). The analysis compares the avoided CO2 emissions of the proposed case () with respect to the reference one (). The CO2 emissions for each system are evaluated by means of CO2 emission factors. This factor is defined as β for natural gas and is equal to 0.205 kg for each kWh of primary energy related to the fuel. The CO2 emission factor for the electricity, defined as α, is equal to 0.360 kg for each kWh of electric energy and was evaluated considering the average Italian emissions of the thermo-electric plants mix and renewable plants including grid losses [36] [37]. Similarly to FESR, ∆CO2 is defined as:
CO2 emission related to electricity exported to the grid () is considered as a credit for evaluation. ∆CO2 has a trend and values that are similar to FESR achieving the data reported on Table 9.
The economic analysis was carried out on the basis of investment and operating costs of proposed and conventional systems. In particular a specific natural gas cost, cu,NG, equal to 0.88 €/Nm3, and an electricity price that depends on a fixed part (Cfix,el = 625.6 €/y) and a variable part, cu,el, equal to about 0.20 €/kWh, [38], that is function of the electricity drawn from the grid, both for PS and CS is here considered. Electricity exported to the grid is paid by a feed-in tariff, cu,el-exp, equal to about 0.12 €/kWh and an annual ordinary maintenance cost of 18.5 € per kW of PV peak power is considered [39]. The specific investment cost for PV decreases with plant size and is in the range 2113 €/kW (4.5 kW) to 2002 €/kW (7.5 kW). It includes all PV plant components (panels, inverter, cables, etc.), flat roof PV frame, transportation, installation, design and 10% in terms of VAT. Finally, a specific cost for battery equal to 780 € per kWh of electric capacity storage, an investment cost for GSHP, ICGSHP, of 7520 € and a cost for drilling and installation of vertical borehole heat exchanger including water-glycol mixture, ICBHE, equal to 14400 € is introduced [40]. To study the economic feasibility of the proposed system, different methods could be considered (net present value, internal rate of return, annualized life cycle cost, etc.), but in this study is introduced a simplified approach based on the simple payback period, SPB, defined as:
considering that:
• IC is the investment cost of the PS (PV plant, GSHP, BHE, batteries);
• ICPV is the total investment cost of PV plant;
• ICBAT is the investment cost of electric storage;
• Fj is the cash flow for the generic year j;
• and are the operating costs of the PS and CS;
• ΔOCj is the difference in terms of operating costs between PS and CS;
• is the maintenance cost for PV system;
• VONG is the volume of natural gas used for boiler considering a lower heating value equal to 9.52 kWh/Nm3.
Operating cost savings, ΔOC, due to introduction of PS increases with PV peak power and battery size achieving an annual saving, Figure 9, that is the range between 1300-1400 € (4.5 kW) and 2000-2100 € (7.5 kW).
The operating costs for CS related to electricity, space heating and cooling demand are about 3159 € (2857 € for electricity and 302 € for natural gas). The evaluation of SPB could be based on two different scenarios:
• grid parity: characterized by an investment cost without any economic support;
• support action: Italian government gives incentives covering 50% of IC for PV plant, battery, GSHP and BHE [41].
The economic analysis in presence of government support leads to acceptable results with SPB lower than 10 years for 7.5 kW with no or small size battery (3.2 kWh), Figure 10.
Comparing the results is clear that the SPB decreases with PV size while increases with electric battery capacity, due to high investment cost of this component. The price of electric storage needs, for this application, a further reduction when higher capacity batteries are considered.
A more detailed economic analysis has to consider:
• the reduction of PV panels efficiency, and therefore the electric energy availability during their operating life;
• the reduction of batteries performance, their actual capacity (state of health) and operating life (about 12 years) [42];
• the inverter operating life (lower than 15 years);
• the variation of natural gas and electricity unitary costs;
• the change of economic support action.
For the configuration characterized by lower electricity exported (4.5 W, 9.6 kWh) and presence of economic support action on investment cost a sensitivity analysis changing natural gas and variable part of unitary electricity costs for both proposed and conventional system is reported in Figure 11. SPB could be acceptable achieving less than 10 years only in case of high natural gas and variable part electricity prices.
A solar electric, heating and cooling plant, based on a GSHP interacting with PV field, to satisfy electric, thermal and cooling demand for an office building located in southern Italy is here considered. Dynamic simulations are carried out to evaluate the energy, economic and environmental performance of the proposed system varying PV peak power and battery capacity. Energy and environmental analysis of PS system show good and similar results, with respect to a reference energy conversion system consisting of electric grid, a natural gas fired boiler and an electric chiller. FESR and ∆CO2 are always higher than 57 %, with a maximum of about 97 % for PV peak power of 7.5 kW without batteries. Increasing the size of PV system, renewable electricity covering total demand achieves up to 69.7 %, with peak power of 7.5 kW and larger size battery (9.6 kWh), while the self-consumption of PV electricity for this configuration is 72.8 %. The percentage of self-consumed renewable electricity decreases with PV size and increases with battery capacity reaching for 4.5 kW and 9.6 kWh the best result (84 %). The economic analysis shows a reduction of operating costs of solar based system with respect to the reference one, even if investment costs, due to PV field, GSHP, BHE and batteries, highlight the need of economic support actions that lead to acceptable results only for few configurations. The introduction of batteries leads to a reduction of electricity exported to the grid even if there is an increase of SPB due to high investment cost of this component.