This paper analyses a new housing complex with high quality envelope, of which each block is equipped with a solar driven heat pump for production of space heating and domestic hot water. One of the blocks has been instrumented in order to characterize the behaviour and to assess the performance of the system. The instrumentation enabled to fully characterise the energy flows over an entire year, with less than 3% annual balance error between inputs and outputs.
In situ characterization shows that the individual subsystems (solar collectors, heat pump and heat storage) largely operate as predicted. Operation of the 10 building blocks over 2 years further demonstrated the excellent system reliability. However, despite the fact that electric backup hardly needs to be used, the system SPF3 (including backup electric heating) is only of 2.9. This value is rather low as compared to results of other systems with solar collectors and/or geothermal boreholes reported in literature (however for single-family houses). Due to a low thermal demand (68 kWh/m2/yr), the absolute electric consumption (24 kWh/m2/yr) nevertheless remains quite reasonable.
Several points which may have contributed to a relatively low SPF3 have been identified:
1. an unusually low demand for space heating along with an unusually high demand for domestic hot water (re-spectively 30% and 70% of the heat demand), so that a significant part of the heat has to be produced at a high temperature of around 60°C;
2. in absence of a load adjusted heat pump, the excess heat for domestic hot water is being stored for subse-quent floor space heating at around 30°C, resulting in a further thermodynamic inefficiency;
3. a single heat distribution circuit with decentralized domestic hot water storage, which doesn’t allow for solar preheating and thus deteriorates the potential of direct solar heat production (bypassing the heat pump);
4. no insulation of the unglazed solar collectors, which enhances their performance as absorbers on ambient air, but reduces their performance as solar collectors.
The effect of these points will be analysed in the continuation of this work, by way of a numerical simulation model, which will also enable to explore the effect of system sizing, as well as the potential performance of this concept if implemented in retrofitted buildings.
Chapter 3
Towards a SPF of 5? Numerical sensitivity analysis for various building demands
The content of this chapter is published in:
Fraga C., Hollmuller P., Mermoud F., Lachal B., 2017, Solar assisted heat pump system for multifamily buildings: to-wards a seasonal performance factor of 5? Numerical sensitivity analysis based on a monitored case study, Solar Ener-gy, 146, 543-564.
Abstract
The present work analyses the potential of a combined solar thermal and heat pump (HP) system on new and existing multifamily buildings. The study uses numerical simulation as a complement to a monitored case study. After a de-scription of the case study and a summary of the monitoring results, we present the numerical model developed for this study. Simulation results are validated with the monitored values, at component and system level, in terms of monthly profiles and yearly integrals. On this basis, we carry out an extensive sensitivity analysis concerning the prin-cipal sizing parameters of the system. Finally, we investigate the sensitivity of the system to space heating (SH) and domestic hot water (DHW) demands, in particular concerning the applicability of the analysed system in the case of building retrofit. For Geneva’s weather conditions, a sizing factor of 3 m2 solar collector per kW of HP capacity is a good compromise between system size and system performance, resulting in a system seasonal performance factor (SPFsys) between 3.1 and 4.1, depending on the SH distribution temperature. The associated electricity consumption (ranging from 12 kWh/m2 for a new low-energy building, up to 45 kWh/m2 for a non-retrofitted building) strongly depends on the heat demand. Such is also the case for the collector area (from 0.08 m2 per m2 heated area for a new low-energy building, up to 0.20 for a non-retrofitted building). Finally, a SPFsys of 5 could potentially be achieved, but only in newly constructed buildings with a high efficient envelope, a low SH distribution temperature, and with a col-lector area of at least 0.20 - 0.25 m2 per m2 heated area. However, the related investment may not be worthwhile, given the rather small associated electricity saving, not to mention that such a collector area would not fit on buildings with more than 4 storeys.
Keywords
Solar heat pump; multifamily buildings; in situ monitoring; numerical simulation; performance indicators; sizing factor.
Towards a SPF of 5? Numerical sensitivity analysis for various building demands
Nomenclature
Abbreviations
DHW domestic hot water
HP heat pump
SH space heating
Latin letters
Asol area of solar collectors [m2] Ast area of storage envelope [m2] cwat specific heat of water [J/K.kg]
COP coefficient of performance [-]
dt time step [s]
Edir electricity, direct heating [kWh/m2] Ehp electricity, HP [kWh/m2]
Etot electricity, HP + direct heating [kWh/m2] Ėdir electric load, direct heating [W]
Ėhp electric load, HP [W]
Gsol global solar irradiance, in collector plane [W/m2] Gh global solar irradiance, in horizontal plane [W/m2]
hsol,0 heat loss coefficient of solar collector, without wind [W/K.m2]
hsol,v heat loss coefficient of solar collector, proportional to wind speed [W/K.m2 per m/s]
hst heat loss coefficient of storage [W/K.m2]
Pmax.dem maximum hourly heat load, DHW + SH [W/m2]
Pmax.dhw maximum hourly heat load, DHW [W/m2]
Pmax,sh maximum hourly heat load, SH [W/m2]
Pnom,hp nominal capacity, HP (at 0°C evaporator input / 35 °C condenser output) [W/m2] Pnom,sh.0°C nominal heat load, SH (at 0°C outdoor) [W/m2]
Qdem heat demand, DHW + SH [kWh/m2] Qdhw heat demand, DHW [kWh/m2] Qevap HP heat input (evaporator) [kWh/m2] Qhp HP heat production [kWh/m2]
Qhp,dem HP heat production, to DHW + SH [kWh/m2]
Qhp,dhw HP heat production, to DHW [kWh/m2]
Qhp,sh HP heat production, to SH [kWh/m2]
Qhp,st HP heat production, to storage [kWh/m2] Qsh heat demand, SH [kWh/m2]
Qsol solar collector heat production [kWh/m2]
Qsol,dem direct solar heat production, to DHW + SH [kWh/m2] Qsol,dhw direct solar heat production, to DHW [kWh/m2]
Qsol,dir direct solar heat production, to SH + DHW + storage [kWh/m2] Qsol,hp solar collector heat production, to HP [kWh/m2]
Qsol,sh direct solar heat production, to SH [kWh/m2] Qsol,st direct solar heat production, to storage [kWh/m2]
Qst,dem storage heat discharge, to DHW + SH [kWh/m2]
Towards a SPF of 5? Numerical sensitivity analysis for various building demands
Qst,dhw storage heat discharge, to DHW [kWh/m2]
Qst,in storage heat input [kWh/m2]
Qst,loss storage heat losses [kWh/m2]
Qst,out storage heat output [kWh/m2]
Qst,sh storage heat discharge, to SH [kWh/m2]
Q̇̇evap HP input (evaporator), heat rate [W]
Q̇̇hp HP production, heat rate [W]
Q̇̇̇sol solar collector production, heat rate [W]
Q̇̇̇st,in storage input, heat rate [W]
Q̇̇̇st,loss storage losses, heat rate [W]
Q̇̇st,out storage output, heat rate [W]
SPFsys seasonal performance factor, system SPFhp seasonal performance factor, HP
Tcond temperature of condenser output [°C]
Tevap temperature of evaporator input [°C]
Text temperature of ambient [°C]
Troom temperature of technical room [°C]
Tsh.0°C temperature of SH distribution, at 0° outdoor temperature [°C]
Tsh.15°C temperature of SH distribution, at 15°C outdoor temperature [°C]
Tsh.off temperature on/off set-point for SH [°C]
Tsol temperature of solar collector [°C]
Tst temperature of storage [°C]
Tst,t-1 temperature of storage, previous time step [°C]
v wind speed [m/s]
Vst volume of storage [m3]
Greek symbols
ηsol optical efficiency, solar collector ρwat specific weight of water [kg/m3]
Towards a SPF of 5? Numerical sensitivity analysis for various building demands
3.1 Introduction
Over the past decade, heat pumps (HP) have become a key technology for the increased use of renewable energy resources. One of the major issues concerning the use of HPs is their associated electricity consumption. The perfor-mance of HP systems is therefore commonly quantified: i) at the level of the HP by the SPFhp, defined as the ratio be-tween the heat produced by the HP and the electricity consumed by the HP, in annual values; ii) at the level of the heating system by the SPFsys, defined as the ratio between total system heat production and related electricity con-sumption. Depending on the author and considered system, diverse system perimeters are taken into account for the definition of the SPFsys. In the following literature review, SPFsys is therefore used in a generic sense. For detailed in-formation on the perimeter considered in each study, the reader should refer to the specific reference.
Nowadays, the most common HP systems use air or ground as their heat source (EHPA, 2015, Observ'ER, 2015). Erb et al., 2004, monitored 199 of such systems in Switzerland, both in new and renovated buildings. They observed, in aver-age, an annual system performance factor (SPFsys) of 2.7 for the 105 air source HP, and of 3.5 for the 94 ground source HP. In another study, Miara et al., 2010, monitored 74 HP systems, the majority with underfloor heating distribution systems. The observed SPFsys were of 2.9 for the 18 air source HP systems and of 3.9 for the 56 ground source HP sys-tems.
In view of increasing the system performance, focus has lately been set on combining HPs and solar thermal collec-tors. Solar and HP systems (SHP) are composed of at least a solar collector and a HP, but they can also include other heat sources (most commonly air or ground), storages or other components. Furthermore, they can be used for space heating (SH), domestic hot water production (DHW), cooling or any combination of the latter. Consequently, their classification can become quite complex (Buker et al., 2016).
SHP systems were closely analysed by the IEA SHC Task 44 (JC. Hadorn et al., 2015). Co-authors of latter task, Frank et al., 2010, propose a system classification that focuses on the interaction between the solar collectors and the HP.
Usually, this interaction results from the following configurations: i) the collectors and the HP are not interconnected and supply heat independently (parallel systems); ii) the collectors supply heat to the HP evaporator, either exclusively or with an additional source (Series systems); iii) the collectors are used for regeneration of the HP heat source, such as the ground (Regeneration). Note that these configurations are not exclusive and can be combined. An analysis of the market availability of such SHP systems (Ruschenburg et al., 2013) shows that 61% of the available systems are parallel only, 6% series only and less than 1% regeneration only; the remaining 33% are a combination of the different configurations.
Simulation of SHP systems for residential buildings has been carried out by several authors. Concerning air + solar HP systems, Banister et al., 2015, compare a parallel/series system to a solar only system and to an electric only system;
Lerch et al., 2015, compare different concepts (parallel only, as well as diverse parallel/series combinations, involving different components); Carbonell et al., 2016, validate a model of a parallel/series system with ice storage with meas-ured data and carries out sensitivity analysis; Winteler et al., 2014, focus on a series system with ice storage; Li, H. et al., 2014, analyse the influence of major parameters on the performance of a dual HP system (an air + solar parallel HP and a seasonal energy storage HP); Haller et al., 2011, analyse the impact of a series connection in a parallel/series system; In another study, Haller et al., 2014b, study the influence of hydraulic integration and control on a parallel system with a solar combi-storage; Mojic et al., 2014, study a parallel system as well as solar + ice storage HP systems in different locations/weather as well as different building demands.
Regarding ground + solar HP systems, both Girard et al., 2015, and Bertram, 2014, simulate the impact of adding solar collectors to a ground source HP system; Haberl et al., 2014, analyse the performance of a parallel system as well as its optimization; Both Cimmino et al., , and Rad et al., 2013, evaluate the performance of regeneration systems; Reda, 2015, studies different control strategies of a parallel/regeneration system; Comodi et al., 2015, study a parallel sys-tem that includes also photovoltaic panels.
Towards a SPF of 5? Numerical sensitivity analysis for various building demands
As for studies that include both air + solar and ground + solar HP systems, Carbonell et al., 2014, simulate two differ-ent air + solar HP systems (parallel only, parallel/series with ice storage) with a parallel ground + solar HP; In another study, Carbonell et al., 2013, study the effect of weather/location on parallel air + solar and ground + solar HP sys-tems; Poppi et al., 2016, study the influence of weather/location and component sizing on parallel only air + solar HP, as well as ground + solar HP systems; Ochs et al., 2014, study the optimum share of solar thermal and solar photovol-taic in both air + solar and ground + solar HP systems.
The SPFsys documented in SHP simulation studies vary widely depending on system configurations, sizing, loads and weather conditions (Haller et al., 2014a). As an example, for Strasbourg weather conditions, a medium heat demand of a single family house (Qsh of 45 kWh/m2/yr, Qdhw of 2076 kWh/yr, defined in JC. Hadorn et al., 2015) and various system configurations and sizing, the SPFsys varies from 3.6 to 5.9 for air + solar HP systems (Carbonell et al., 2014) and from 3.6 to 6.2 for ground + solar HP system (Bertram, 2014). Apart from the SPFsys, another performance indicator that is used by most studies (Bertram, 2014, Carbonell et al., 2013, Carbonell et al., 2014, Lerch et al., 2015, Ochs et al., 2014, Poppi et al., 2016, Winteler et al., 2014) is the electricity consumption of the system. Some studies also re-port electricity savings of the SHP system when compared to a reference system without solar collectors.
Although most of the above models have their components validated with measured data, the overall system is usual-ly not validated. As an exception, Carbonell et al., 2013, Carbonell et al., 2014, perform an overall system validation by comparing the results of two different simulation software and Carbonell et al., 2016, with measured data over a 1 year.
Experimental results concerning in-situ measurements of SHP systems are also presented by several authors. With the exception of Hahne, 2000, Busato et al., 2013, and Carbonell et al., 2016, these studies concern small scale systems for single family houses. Stark et al., 2014 monitor 3 parallel only air + solar HP systems; Energie Solaire SA, 2012, studies a combined parallel/series air + solar HP system with ice storage; Balslev-Olesen, 2014, monitor a paral-lel/regeneration ground + solar HP system; Wang et al., 2010, Trillat-Berdal et al., 2007, Trillat-Berdal et al., 2006, and Loose et al., 2014, study combined parallel/series/regeneration ground + solar HP systems; Miara et al., 2010, monitor 2 parallel only ground + solar HP systems, as well as 4 parallel only air + solar HP systems; Bertram et al., 2012, analyse a series only ground + air + solar HP system with PVT collectors.
Except for Ochs et al., 2014, Li, Hong et al., 2014 and Comodi et al., 2015, all preceding simulation and monitoring studies concern single family houses. In this regard, it should be noted that, in Geneva (urban canton with high popu-lation density), 79% of the heated surface of residential buildings concern multifamily buildings, against 21% for single family buildings (Khoury, 2014). At national level the ratio is 55% for multifamily buildings against 45% for single family buildings (values given by a recently developed geo-dependent heat demand model of the Swiss building stock Schneider et al., 2016).
Within this context, a novel study concerns long term in-situ measurement of a parallel/series air + solar HP system implemented on a multifamily building in Geneva (Fraga et al., 2015). As a complement, the present study aims to understand the relatively low performance of the monitored case study, as well as to further analyse the potential of such systems. As one of the specific questions, we would like to check if parallel/series air + solar HP systems could reach a SPFsys of 5 when implemented on new or existing multifamily buildings and if so, if this aim is reasonable.
Towards a SPF of 5? Numerical sensitivity analysis for various building demands
3.2 Case Study
3.2.1 System description
The case study on which we base our study concerns a coupled solar and HP system which was implemented and commissioned in autumn 2010 in a new housing complex, called SolarCity, located in Geneva (Switzerland). The com-plex is composed of 4 buildings, each subdivided in 2 or 3 blocks of 8 flats (total of 10 blocks). The buildings present a high thermal performance envelope (Minergie, 2016) and a total living surface of 9’552 m². The monitored block has 927 m2 and a total of 32 inhabitants.
The energy concept of each block consists of a HP directly coupled to selective unglazed solar collectors as its heat source. The components of the system (Figure 3:1) are: an electrically driven HP with a thermal output of 35 kWth and COP of 4.5 (at 0°C evaporator input / 35°C condenser output, R407C working fluid); 116 m² of selective unglazed and non-insulated solar collectors (both faces subject to heat exchange with ambient air, orientation 37°E [South=0°], inclination 20°); 2 x 3’000 L of centralized water heat storage (hot and tepid tank), with an electric rod for the case of HP failure. As a particularity, the heat distribution to the flats consists of a single loop, so that SH (underfloor space heating) and DHW (domestic hot water) are supplied alternatively. Due to this particular distribution, each flat is equipped with a 300 L DHW tank. DHW distribution has priority over SH distribution.
The solar collectors can be used either for direct solar heat production, via a heat exchanger, or as absorbers for the HP (to which they have a direct connection, without intermediate storage or geothermal boreholes). Hence, when there is no solar radiation, the collectors work as a heat absorber on ambient air. Whether by direct solar heat pro-duction or via the HP, the produced heat is used for SH (underfloor heating) or DHW (heat distribution to charge the individual 300 L tanks), and the surplus is stored in the centralized heat storage for future use.
In summer, the system can also be used for night cooling, by activating the floor distribution circuit and dissipating the heat in the solar collectors. Note that this feature will be disregarded in this study.
Figure 3:1 Hydraulic diagram of the system.
Towards a SPF of 5? Numerical sensitivity analysis for various building demands
3.2.2 Monitoring results
The monitoring campaign covered 2 years of operation (November 2011-October 2013). Extensive data acquisition in 5 minute time step allowed to quantify the various energy flows of the system: solar production (direct and to evapo-rator); HP production; heat storage (charge and discharge); electricity consumption (solar circuit, HP and backup);
heat demand of the building (SH and DHW). The results over the year 2012, which are widely discussed in Fraga et al., 2015, are summarized below and in the Sankey diagram (Figure 3:2).
Figure 3:2 Sankey diagram of the monitored block with respective measurement uncertainties, year 2012 (units: kWh/m2).
Of the total heat demand (68.1 kWh/m2), 70% accounts for DHW (47.7 kWh/m2) and 28% for SH (19.1 kWh/m2). The residual 2%, which correspond to transitory regime in the single distribution loop, could not be assigned univocally to DHW or SH. Note that the observed share between DHW and SH is relatively unusual for Geneva, which points out an excellent envelope and indoor temperature control, as well as an unusually high DHW demand of 1380 kWh/pers/year (as compared to the normative value of 40 L/pers/day, equivalent to 770 kWh/pers/year (SIA385/2, 2015), or to the benchmark value of 1080 kWh/pers/year for multifamily buildings in Geneva (Khoury, 2014). Of the overall heat de-mand, 73% is supplied in winter (Oct-Apr) and 27% in summer (May-Sep).
Direct solar heat production (bypassing the HP) accounts for 19% of the total input energy (7% in winter and 49% in summer). This result is not surprising for winter but a better value was expected for DHW production in summer, due to the large solar collector area (3.6 m2 and 190 L heat storage per person, as compared to standard design values of 0.5-1.1 m2 and 60-90 L heat storage per person (SuisseEnergie, 2001). The main reason for the poor direct solar heat production is the peculiar hydraulic configuration (single distribution loop for SH and DHW, with individual DHW stor-age), which does not allow for solar preheating of DHW, but only for DHW production above 55°C (Fraga et al., 2015).
As a complement to direct solar heat production, the HP accounts for 80% of the total heat production. The annual seasonal performance factor of the HP (SPFhp) is 2.7, with only slight variations during the year (2.5 to 3), even in
Towards a SPF of 5? Numerical sensitivity analysis for various building demands
As an overall result, the renewable heat fraction corresponds to 68% of the total energy input. The complementary electricity consumption (32%) leads to an annual SPFsys of 2.9 (with a wide variation over the year, from 2.2 during cold periods to 8.6 in August, when direct solar heat production for DHW is possible). Despite a relatively low annual
As an overall result, the renewable heat fraction corresponds to 68% of the total energy input. The complementary electricity consumption (32%) leads to an annual SPFsys of 2.9 (with a wide variation over the year, from 2.2 during cold periods to 8.6 in August, when direct solar heat production for DHW is possible). Despite a relatively low annual