The role of the design and operation of individual heating systems for the energy retrofits of residential buildings

This is the accepted manuscript of the article that appeared in final form in Energy Conversion and Management 126 : 736-747 (2016) , which has been published in final form at https://doi.org/10.1016/j.enconman.2016.08.042. © 2016 Elsevier under CC BY-NCND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) © 2016. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ ENERGY CONVERSION AND MANAGEMENT https://doi.org/10.1016/j.enconman.2016.08.042 The role of the design and operation of individual heating systems for the energy retrofits of residential buildings J. Terés-Zubiaga(1); A. Campos-Celador(2); I. González-Pino(3); G. Diarce(4) (1) ENEDI Research Group, Department of Thermal Engineering, Faculty of Engineering of Bilbao, University of the Basque Country UPV/EHU, Rafael Moreno “Pitxitxi” 2, 48013 Bilbao, Spain. (2) ENEDI Research Group, Department of Thermal Engineering, Faculty of Engineering of Gipuzkoa, University of the Basque Country UPV/EHU, Avda. Otaola 29, 20600 Eibar, Spain. (3)ENEDI Research Group, Department of Thermal Engineering, Faculty of Engineering of Bilbao, University of the Basque Country UPV/EHU, Alda. Urquijo S/N, 48013 Bilbao, Spain. (4) ENEDI Research Group, Department of Mining and Metallurgical Engineering and Material Sciences, Faculty of Engineering of Bilbao, University of the Basque Country UPV/EHU, Rafael Moreno “Pitxitxi” 2, 48013 Bilbao, Spain. ABSTRACT The feasibility of individual natural gas fired boiler-based heating systems in the retrofitting of buildings constructed in the 50-60s in Bilbao (northern Spain) is evaluated in this paper. A holistic approach through dynamic simulations using TRNSYS is employed for the purpose. An existing dwelling previously monitored and used to validate the model applied is selected as a case study. 54 different scenarios are evaluated, which arise from the combination of 3 different envelope options, 2 types of heat production units, 3 heat production temperatures and 3 comfort temperature set-points. The cases are evaluated in terms of energy results, economic aspects, and the influence of user behaviour. Regarding the latter, the influence of the potential rebound effect is also evaluated. The results show energy savings nearby 10% when condensing boilers are compared with high efficiency boilers. In relation to hot water production temperature, energy savings between 5 and 10% are found when the temperature is lowered from 60 to 50ºC. The greatest impact on energy consumption is related to the occupants’ behaviour: reductions up to 89% are achieved if the indoor temperature set-point is lowered J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 2 2ºC. This is reinforced with the results related to the rebound effect, which show significant differences on energy consumption values. These evidences demonstrate that the user behaviour is an essential feature to be considered in studies regarding buildings energy performance. As a consequence, the holistic approach herein employed emerges as a key tool to be applied in further works related with the topic. Keywords: Energy supply systems; Holistic approach; Building Energy Retrofit; Energy efficiency; Rebound Effect Nomenclature BAU Business as usual BO Best option C Theoretical heating energy consumption (calculated) C op Operating cost C en Current cost of energy CB Condensing boiler DHW Domestic Hot Water E Actual energy consumption e Annual escalation rate of energy ESM Energy Saving Measures I Current cost of investment LTB Low temperature natural gas boiler; High efficiency boiler LCC Life Cycle Cost NR Non-retrofitted P eff Effective thermal power P EN442 Nominal thermal power per length (EN-442) PLR Part load ration r Discount factor RE Rebound Effect T eff Effective temperature TRV Thermostatic valve 1 Introduction Nowadays, the building sector is responsible of 40% of the overall primary energy consumption in Europe as well as one third of related global greenhouse gas emissions [1]. According to the United Nations Environment Programme, the building operational phase accounts for 80-90% of those emissions, consisting of the energy use for heating, cooling, ventilation, lighting and appliances [2]; therefore, each action directed towards increasing energy efficiency of buildings and reducing their primary energy consumption is of great importance, as it can be inferred from the numerous regulations set up during the last decade [3]. These regulations were originally focused on new construction; J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 3 however, the existing building stock is the main challenge for a substantial reduction of the energy consumption. As a result, considerable work has been done throughout the latest years to get a suitable normative framework for facing this challenge [4]. For the specific case of Spain [5], 56% of the 26 million dwellings existing by 2011 were built up before the first Spanish thermal regulation on buildings (NBE-CT 79) came into effect in 1980. Therefore, there is a doubtless requirement for retrofitting in order to meet the European objectives on 20% primary energy consumption reduction [6]. This can be achieved applying energy saving measures (ESM); reducing the energy demand through the improvement of the thermal performance of the building envelope, and/ or implementing more efficient energy systems. As far as energy systems are concerned, several works have been recently published. Defu Che et al. [7] evaluated the upgrade of a conventional gas boiler into a condensing boiler, focusing on the boiler itself, and leaving out of scope its interaction with the building and its users. Deng et al. [8] evaluated energy supply concepts for zero energy residential buildings in two different climates, by means of simulations. M. Owrak et al. [9] evaluated experimentally and by means of simulations the thermal performance of a room heated with an attached sunspace, which included water tanks with the aim of increasing the heat storage capacity. Focused on thermal installations, Obyn and van Moeseke [10] evaluated for the case of Belgium different heating systems in the renovation of an attached house. They concluded that for highly insulated dwellings, the optimum system is the most simple in terms of composition. Also for Belgium, Vrijders and Delem [11] underlined that condensing gas boilers are the cheapest heating system with low emission level. Tagliabue et al. [12] analyzed three solutions (gas condensing boiler, air source heat pump and ground source heat pump) for a residential building in Milan (Italy). It was proved that heat pumps perform better than gas condensing boilers, being the ground source heat pump the most profitable solution. Anastaselos et al. [13] carried out a comparative analysis between different technologies for a semi-detached house in Germany. Amongst the cases under evaluation, natural gas boilers showed to be the best option from an economical and environmental point of view. Nagy et al. [14] demonstrated that the implementation of a suitable low temperature heating system can be the best solution for existing buildings, even when no ESM is applied to the envelope. User behaviour is an additional factor to be considered on energy consumption. Its influence can be even larger than the building characteristics or other factors [15-17]. Many studies have pointed out noticeable differences in energy consumption for similar buildings [18, 19] due to the occupants’ J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 4 behaviour. The existing relation between behavioural patterns, user profiles and energy use was demonstrated in [20]. To illustrate this point, the energy use obtained from a field survey in 110 similar dwellings was presented in [21]. The dwelling with the maximum consumption showed an energy use 12 times higher than that dwelling with the minimum. This effect is even greater when the social building sector is analysed, as shown by Brunner et al. in [22]. The rebound effect (RE) [23] is another factor to be taken into account. It is defined as the direct increase on demand for an energy service as a result of improvements in technical efficiency in the use of energy [24, 25]. The so-called backfire occurs when the fuel use actually increases as a result of that fuel efficiency gain. Even though empirical studies suggest that backfire is not usual, many research works prove that actual energy savings in building renovations are hardly ever proportionate to the energy efficiency improvement. Whereas RE focuses on over-consumption after an energy renovation, prebound effect concept is based on the evidence of under-consumption prior to or in the absence of energy renovations [26]. The link between prebound effect and energy savings shortfalls in renovations has been studied in depth by R. Galvin [27-29], while implications of the RE in building renovations have been widely analysed in studies such as [30-32]. In some cases, the rebound effect is recognised as a co-benefit which involves social advantages like healthier conditions [33]; in others, it involves an increase of internal temperatures without occupants demanding it [34]. Despite the difficulties of quantifying these effects, Galvin and Sunikka asserted that it generally lies within the range of 10-35% [26]. Up to now, no work has been found in the literature dealing with the combined analysis of heating system and envelope retrofitting; heating system operation and user behaviour. Thus, the objective of this paper is to evaluate, under a holistic approach, the feasibility of individual natural gas fired boilerbased heating systems in the retrofitting of buildings/dwellings constructed in the 50-60s in Bilbao (northern Spain). This type of building stock has a great energy performance improvement potential, as it has been already shown in other studies [35, 36]. It should be noticed that several of these buildings in northern Spain, especially social housing, have no heating system and the dwellings are usually heated up by individual electrical radiators. Considering the absence of a central heating infrastructure and the wide availability of natural gas networks in the area, individual gas boilers appear as the most feasible option for heating installation upgrade. J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 5 The evaluation is carried out over a reference dwelling selected as a case study. This dwelling was presented in a previous paper where the authors analyzed the building envelope ESMs as a first step for energy renovation of buildings at the mentioned location [36]. For that purpose, the dwelling was modelled in TRNSYS and experimentally validated [37]. The work is herein extended, including the upgrade of heating systems and their operation as a second step for energy renovation. Different options will be studied in combination with three envelope options already analysed in [36]. An integral dynamic simulation using a validated TRNSYS model will be used for the purpose. With this aim in mind, the experimentally validated TRNSYS model used in [36] will be adapted and broadened in order to include a detailed heating installation along with the building. The energy and economic results will be evaluated, considering the interrelationship amongst the natural gas boiler technology (low temperature and condensing); its operation (hot water production temperature) and the user behaviour (indoor air set-point temperature). Regarding the latter, the influence of the rebound effect will be also addressed. The article provides two main significant contributions to the literature published so far. First, the existing lack of studies devoted to heating system upgrades in social housing buildings under mild climates is aimed to be addressed. The study is focused on retrofitting, which can be considered the actual challenge to be faced in the following years. Moreover, the simulations are performed under realistic conditions by means of an existing dwelling and using of a validated dynamic model. Second, the study will be carried out using a holistic approach, where the user behaviour and the (p)rebound effect will be evaluated along with energy and economic aspects. The article will demonstrate that these effects, rarely considered in the related literature, are an essential feature to be taken into account in further studies focused on buildings energy performance. The remainder of this paper is organized as follows: In Section 2 the methodology and main assumptions adopted and the TRNSYS energy modelling are described. In Section 3, the case study is detailed and the scenarios considered are defined. The evaluation criteria are described in Section 4. Section5 presents the results, while their discussion appears in Section 6. Finally, the main conclusions are addressed in Section 7. 2 Modelling approach The energy renovation in a building or dwelling generally consists of the energy demand reduction by improving the thermal performance of the envelope and, subsequently, the production of such demand J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 6 by more efficient energy supply systems. While the envelope retrofitting was covered in [36], in this paper the heat production systems upgrade is faced. Every heating system consists normally of the following elements: (1) heat production unit; (2) terminal units and (3) control system. The integration of these three elements makes up the heating installation. The characteristics of each of the elements for the system involved are detailed next. 2.1 Heat production unit Considering that the building stock in northern Spain is mostly comprised by individual electrical heating systems [35], only individual systems are taken into consideration in this paper. Amongst the different individual heating systems available, natural gas boilers are selected, owing to the wide natural gas network existing in the region. Natural gas boilers can be nowadays divided into two categories: high efficiency boilers (LTB) or condensing boilers (CB). Both technologies have a common operation basis, being the difference that condensing boilers recover part of the latent heat content of the exhaust by condensing their vapour water content through heat exchange with the water returning from the load side. Accordingly, the lower the temperature of the returning water, the higher the condensing level and the efficiency [38]. Besides, in boilers with modulating burners, the lower the part load operation, the higher the efficiency, since the lower flow rate of fumes implies a better heat recovery rate. This trend is maintained until a limit PLR (Part Load Ration) is reached (namely 10-15%). Below that point, the efficiency suddenly drops and thus, this condition is usually prevented by the boiler burner control. The dependency between the thermal efficiency and the return temperature and PLR can be found in Fig. 1. Fig. 1. Thermal efficiency relationship for high efficiency boilers (LTB) and condensing boilers (CB); adapted from [39] J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 7 As it can be seen, condensing boilers perform better than non-condensing ones, especially when the return temperature is lowered below the condensing temperature (around 50ºC-55ºC for natural gas) and when part load operation is boosted. This is closely related to the selection of the terminal unit and the control of the whole installation, which are subsequently discussed. 2.2 Terminal unit Radiator networks are chosen as terminal units, because they are the most common option in natural gas boiler installations. These units were originally sized for a high temperature operation; however, their design has been updated for operation at lower temperature levels with inlet temperatures around 55-60ºC, being the return temperature a function of the thermal load and the radiator thermal efficiency. Their effective thermal power can be related to the design performance by Eq. 1 [40], where n is a coefficient considered equal to 1.3 for natural convection. Eq. 1 The performance values are given by the manufacturers following the EN442 for a 50K temperature difference between the water average temperature and the room temperature (being commonly 70ºC and 20ºC, respectively) [40]. Accordingly, an average temperature of the radiator of 50ºC would give a 30K temperature difference, which results in a 50% reduction of the nominal power of the units. This makes necessary to increase the radiator length in order to lower the operation temperature when condensing boilers are used. Therefore, a relation between the radiator length increase and the operation temperatures is required. This is presented in Fig. 2, being the flow rate assumed constant. Fig. 2. Relationship between radiator average temperature and radiator length increment J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 8 The shadowed area of the plot remarks the operation points where condensing occurs (when the return temperature (Toutlet) is below 55ºC according to Fig. 1). As it can be seen, it requires longer radiator (Δlrad) units, which increase exponentially as the operation temperature drops. This means a bigger investment for the heating systems renovation which will affect the economic feasibility. Another alternative for reaching low return temperatures is to employ the radiant floor technology as terminal unit. This system works at a lower temperature, being the inlet water temperature around 40ºC. This fact ensures the condensing effect, but it also presents a higher initial investment and technical complexity. For that reason, this option is not very common on retrofitting works of social household in Spain, and thus, it has been considered out of the scope of this study. 2.3 Control system Heating systems can be controlled acting over one of the following variables: (1) On/Off control of the heating system; (2) control over the water delivery temperature; and (3) control over the water mass flow rate. The On/Off control is usually made by a thermostat, which is usually placed in the living room. The user specifies the temperature below which the heating system is activated. The control presents certain hysteresis in order to avoid too fast On/Off sequences. This hysteresis cycle is around 1ºC downward. The control over the water delivery temperature is regulated by the boiler. Modern boilers allow part load operation with the aim of meeting a given set-point temperature. This temperature can be modified by the user separately for space heating and DHW. Therefore, from a practical point of view, the hot water supply temperature can be considered as constant in individual boiler systems and then, the burner is regulated in order to meet it. Regarding the control over the water mass flow rate, this possibility presents two options: acting directly over the flow rate by the use of variable speed pumps, or acting over the pressure drop of the heating loop by the use of valves. However, none of these options are included in this paper since individual boilers integrate their own single speed pump and the pressure drop-based control could imply noise and higher pump head, and therefore, higher pumping costs. Additionally, the flow rate of the water that flows through each terminal unit can be controlled by a three way valve, bypassing part of the flow. This can act over the heat delivered by the terminal units, but is J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 9 not useful for reducing the temperature of the return water, since the bypassed flow is mixed with the water exiting from the terminal units. 2.4 TRNSYS energy modelling The dwelling, along with the heating system, is simulated with TRNSYS simulation software. The main features of the dwelling will be presented in Section 3. The model and its experimental validation were already detailed in [36]. In the current work, the heating system model is integrated into it. The analysis performed evaluates the performance of different systems made up from the combination of: high efficiency and condensing natural gas boilers as heat production units; radiators as terminal units; and different water supply temperatures and indoor air set-point temperatures. The natural gas fired boiler is simulated using the Type 700 simple boiler model developed by TESS [41]. A thermal power of 24 kW is considered as typical for this kind of boilers. The thermal efficiency for both high efficiency and condensing boilers are obtained from Cockroft et al. [39]. The On/Off operation of the boiler is controlled by Type 2b that switches it off when the air temperature of the reference room (living room) is reached. A hysteresis of ±0.5ºC is included in order to guarantee a smoother operation. The pump is integrated in the boiler and modelled within it, being controlled by the room thermostat. In the simulations, only the thermal production for space heating is considered, neglecting the operation of the DHW. The energy consumption for DHW production is added a posteriori for the economic evaluation, as detailed in Section 4. This assumption does not have significant influence on the results, since the model does not take into account the thermal mass of the boiler. Radiators are modelled by a self-tailored type, implemented as Type 211. The model consists of a lumped capacity model which is based on a first-order differential equation that accounts for the thermal inertia of the radiator. The heat delivered at any instant by the radiator at different operation conditions is obtained by applying Eq. (1). The heat released by each radiator is introduced as heat gains to each zone, a typical convective/radiative ratio of 80/20 can be considered. Piping from the boiler to the terminal units acts as heat emitters (thermal losses are released to the ambient) and adds thermal inertia to the heating installation. They are modelled by Type 31, a singlenode pipe model. Radiator networks in dwellings are better arranged by double pipe configuration; thus, water enters radiators at the same temperature and thermal unbalance is avoided. A water flow rate of 10 l/min is considered and distributed to the different rooms according to the nominal power of the J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 16 Amongst the different forms of rebound effect existing in the literature [50], only direct effects are considered in this paper. They are calculated by Eq. 4 [26]. 𝑃𝑃=𝐿𝐿−𝐸𝐸 𝐿𝐿 Eq. 4 There, C is the theoretical heating energy consumption (calculated), which is obtained considering that the indoor temperature set-point after retrofitting is the same to that it was before. E is the actual energy consumption, and is determined assuming a given increment of the indoor temperature set-point after energy efficiency upgrades. Two increment values are considered: 1 and 2 ºC. In both cases, the assumed indoor temperature set-point before the energy efficiency upgrades is 19ºC [35]; thus, the set-point temperature after the energy efficiency upgrades is considered to be 20ºC and 21ºC. 5 Results The general results of the 54 scenarios are herein presented and discussed. Prior to the evaluation, the comfort conditions were analyzed for all the cases in order to check the actual behaviour of each installation in relation to the theoretical design. The actual air temperature was qualitatively compared to the set-point, showing good agreement. A simple analysis of the comfort was made by considering the number of hours in which indoor temperature is below 18 ºC. Similar conditions were obtained for all the scenarios, where the number of hours below 18ºC ranged between 0 and a maximum value of 37 hours, depending on the indoor temperature set-point. Thus, it can be stated that all the selected designs were adequate. 5.1 Energy results Heating consumption for each scenario is depicted in Fig. 6. Data are distributed by the different envelope retrofitting options, energy supply systems and hot water production temperatures. Additionally the effect of the indoor air temperature set-point can be seen for each case. J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 17 Fig. 6. Final energy consumption (for heating) per year As expected, the envelope retrofitting option has a significant effect on the energy consumption of the dwelling. BAU scenario shows, in comparison to NR, average energy savings between 15% (6.5 - 7.5 kWh/m2; 21ºC set-point) to 27% (3.5 - 4 kWh/m2; 19ºC set-point). When NR and BO scenarios are compared, average savings from 45% (19 - 20 kWh/m2; 21ºC set-point) to 80% (11 - 13 kWh/m2; 19ºC set-point) are achieved. The final energy consumption values could be partially deduced from the demand results in [36], but here the effect of the seasonal heating system performance is also considered (which ranges for supplying the heating demand between 0.71 in the case of LTB and a setpoint temperature of 21 ºC to 0.87 in the case of CB and a set-point temperature of 19 ºC) in the evaluation. The greatest impact on the energy consumption is closely related to the occupants’ behaviour, i.e. the indoor temperature set-point. A non-linear relation between the temperature set-point and the energy consumption reduction can be clearly observed. Thus, a significantly higher reduction in the consumption is appreciated when changing from a temperature set-point of 21ºC to 20ºC. This reduction is also significant when changing from 20 to 19ºC. This influence is quantitatively presented in Table 6. It can be observed that, whereas energy savings in absolute values decrease when the envelope efficiency is higher, the impact in terms of relative values becomes the biggest in the BO scenario. This demonstrates that the user interaction plays a role more important than other design aspects, and accordingly it must be taken into account in this kind of analysis. This trend is maintained 0 10 20 30 40 50 60 NR.LTB.60 NR.LTB.55 NR.LTB.50 NR.CB.60 NR.CB.55 NR.CB.50 BAU.LTB.60 BAU.LTB.55 BAU.LTB.50 BAU.CB.60 BAU.CB.55 BAU.CB.50 BO.LTB.60 BO.LTB.55 BO.LTB.50 BO.CB.60 BO.CB.55 BO.CB.50 kWh/m2year SetPoint: 21ºC SetPoint: 20ºC SetPoint: 19ºC NR BO BAU LTB CB LTB CB LTB CB J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 18 for all the cases evaluated; thus, for the sake of clarity, a set-point of 20ºC is considered as the basis for the analysis for the rest of the paper. NR BAU BO 21 ºC 20 ºC 17.86 (40.5%) 16.95 (45.5%) 14.93 (61.4%) 21 ºC 19 ºC 29.21 (66.3%) 26.45 (70.9%) 21.6 (88.7%) Table 6. Summary of energy savings in kWh/m2.year related to set-point temperature (in brackets, percentage of the saving in relation to energy consumption with a 21 ºC set-point) Regarding hot water production temperature, a lower production temperature means a lower return temperature and, therefore, a higher efficiency. The energy consumption reduction is of 4-5% (from 60ºC to 55ºC) and of 9% (from 60ºC to 50ºC) for the NR and BAU scenarios. Energy savings are higher in the BO scenarios, 8% (from 60ºC to 55ºC) and 12% (from 60ºC to 50ºC). This is explained by the fact that the BAU case presents slightly lower demand and, therefore, for a given boiler nominal power, the part load ration is lower, meaning a higher efficiency (Fig. 1). The same trend is get regardless of the boiler type. In the case of the heating system supply, differences between LTB and CB performance are hardly found, which is owed to two main reasons: both options present the same mean return temperature of 37.6 (60ºC), 33.1 (55ºC) and 28.9ºC (50ºC), and the energy efficiency percentage variation between these return temperatures is practically the same for both boilers, around 8-9% (Fig. 1). The results of the primary energy consumption, which are obtained adding the DHW consumption to the heating consumption, are presented in Fig. 7. The values are gathered in two groups for the sake of clarity: those related to LTB and those related to CB. The results shown reinforce the aforementioned influence of the set-point temperature and, to a lesser extent, the effect of reducing the hot water production temperature. J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 19 Fig. 7. Operating Primary Energy consumption per year (indoor air temperature set-point: 20 ºC) 5.2 Economic results Results corresponding to the evaluation Approach (A) are summarized in Fig. 8. In the graph, annual costs are presented, considering them as the sum of the investment amortization and the yearly average operating cost of the system. Each envelope renovation option (BAU and BO) presents the investment (in dark gray), prorated according to the system lifespan assumed. Additional costs are related to the heating system upgrade: annual investment and fuel costs. The overinvestment needed for the condensing boiler in relation to the low-temperature one can be observed, as well as that required for lower hot water production temperature, i.e. larger radiators according to the sizing method (Table 3). Thus, under Approach (A), the joint renovation action consisting of BAU envelope and condensing boiler brings the best economic results. Amongst the different hot-water production set-points, no significant differences are observed, but the economics are better when the boiler operates at 60ºC. This means that the additional investment for operating at a lower temperature does not compensate the economic savings for a fuel usage reduction. 0 10 20 30 40 50 60 70 80 60 ºC 55 ºC 50 ºC 60 ºC 55 ºC 50 ºC Primary Energy [kWh/m2.year] Boiler setpoint temperature NR BAU BO LOW TEMPERATURE BOILER CONDENSING BOILER J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 20 Fig. 8. Economic results. Approach A. Annual cost for each scenario (amortization of investment + operating cost) considering as reference case the NR scenario with electric system. Results corresponding to Approach (B) are depicted in Fig. 9. The difference with respect to Approach (A) is that, for each case, the envelope retrofitting is assumed to be made by the user prior to the heating system upgrade. Thus, the NR scenario is the same that the one presented before in Fig. 8. For the BAU and BO scenarios, both annual costs and savings in relation to the reference case are lower. Fig. 9. Economic results. Approach B. Annual cost for each scenario (amortization of investment + operating cost) considering as reference cases the NR, BAU and BO scenarios with electrical systems The simple payback results under both approaches A and B are presented in Fig. 10. Payback periods in NR and BAU scenarios are similar regardless the renovation approach, ranging between 8 and 11 years. 0 100 200 300 400 500 600 700 NR.LTB.60 NR.LTB.55 NR.LTB.50 NR.CB.60 NR.CB.55 NR.CB.50 BAU.LTB.60 BAU.LTB.55 BAU.LTB.50 BAU.CB.60 BAU.CB.55 BAU.CB.50 BO.LTB.60 BO.LTB.55 BO.LTB.50 BO.CB.60 BO.CB.55 BO.CB.50 Cost [€/year] Operating Cost [20ºC] Investment cost (energy system) Investment Cost (envelope) Ref Cost [20 ºC] NR BAU BO 0 100 200 300 400 500 600 700 NR.LTB.60 NR.LTB.55 NR.LTB.50 NR.CB.60 NR.CB.55 NR.CB.50 BAU.LTB.60 BAU.LTB.55 BAU.LTB.50 BAU.CB.60 BAU.CB.55 BAU.CB.50 BO.LTB.60 BO.LTB.55 BO.LTB.50 BO.CB.60 BO.CB.55 BO.CB.50 Cost [€/year] Operating Cost [20ºC] Investment cost (energy system) Ref Cost [20 ºC] NR BAU BO J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 21 In BO scenarios, the renovation approach is of great importance, especially due to the relatively higher investment that it requires (i.e. windows replacement). Thus, payback ranges from 19 to 20 years for approach A and from 10 to 12 years for approach B. Fig. 10. Simple payback for each scenario for Approach A and B (indoor air set-point temperature: 20ºC) 5.3 User behaviour The occupant behaviour was partially evaluated in Section 5.1 by means of the analysis of different setpoint temperatures (Fig. 6); however, the same set-point was assumed before and after renovation. For a deeper discussion, the evaluation of the (p)rebound effect is carried out. The results obtained are presented in Table 7, where the percentages shown are the ratio between theoretical and actual savings. Set-point temperature prior to upgrades Set-point temperature after upgrades NR BAU BO 19 ºC 20 ºC 29.9% 31.5% 34.8% 19 ºC 21 ºC 60.5% 59.7% 59.5% Table 7. Rebound effect in each scenario, considering a temperature increment of 1 and 2 ºC As observed, an increment of 1ºC (from 19 to 20ºC) involves rebound effect values around 30-35%. These results agree with previous publications [26]. An increment of 2 ºC from 19 ºC to 21 ºC involves a rebound effect around 60%. Note that this study deals with social housings, where electric heaters are usually employed. These systems are expensive to operate for low-income households, and they create temperature gradients of the indoor air that make the actual room temperature significantly lower than the set-point of the system. Accordingly, increments of 2ºC are considered feasible. 0 2 4 6 8 10 12 14 16 18 20 NR.LTB.60 NR.CB.60 NR.LTB.55 NR.CB.55 NR.LTB.50 NR.CB.50 BAU.LTB.60 BAU.CB.60 BAU.LTB.55 BAU.CB.55 BAU.LTB.50 BAU.CB.50 BO.LTB.60 BO.CB.60 BO.LTB.55 BO.CB.55 BO.LTB.50 BO.CB.50 Payback period [years] Approach A Approach B NR BAU BO J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 22 The yearly savings on primary energy consumption and payback period values considering the aftermath of the rebound effect are depicted in Fig. 11 and Fig. 12. The dark grey rhombus represent the theoretical values with no rebound effect (set-point temperature of 19 ºC), while the white rectangle represents the range of depicted values when the increase of set-point temperature ranges from 1 to 2ºC. Fig. 11. Range of primary energy savings values considering the rebound effect It can be appreciated that ranges are wider in those case where a less intensive energy efficiency upgrade is carried out, and even the aforementioned backfire effect is reached in the NR.LTB.60, where negative energy savings (higher energy consumption) are found when set-point temperature increases close to 2ºC. Analogously, payback period values can be analyzed using a similar methodology. In this case, as a way of example, payback period values under approach (A) are presented considering the consequences of the (p)rebound effect. Mentioned values range between 10,5 and 12,5 years in NR and BAU scenarios when no rebound effect is considered (dark grey rhombus), increasing the payback period up to 17-18 years under some cases when the rebound effect is considered. Similar consequences are found in BO scenario; whilst payback period values are in all cases close to 24 years when no rebound effect is considered, values higher than 30 years are get in some cases, values that are higher that the considered lifespan of the system, thus compromising the feasibility of the renovation. -10 0 10 20 30 40 50 60 70 NR.LTB.60 NR.LTB.55 NR.LTB.50 NR.CB.60 NR.CB.55 NR.CB.50 BAU.LTB.60 BAU.LTB.55 BAU.LTB.50 BAU.CB.60 BAU.CB.55 BAU.CB.50 BO.LTB.60 BO.LTB.55 BO.LTB.50 BO.CB.60 BO.CB.55 BO.CB.50 kWh/m2year NR BAU BO J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 23 Fig. 12. Range of the payback period values considering the rebound effect (Approach A) 6 Discussion The obtained results have significant relevance on the planning of energy retrofitting of buildings. First, it is observed that, from the analysed issues, reducing the temperature set-point has the greatest effect on the energy consumption reduction with no additional expense, showing a saving potential of up to 80%. This makes it the measure that more effectively addresses the energy consumption reduction sought but current energy policies. Addressing this issue requires checking the concept of how the required comfort levels are reached and to emphasize the role of clothing in buildings. This need is especially remarkable in mild climates as it is the case under evaluation. The use of more efficient heating systems, such as condensing boilers, involves savings over conventional low-temperature ones, but these savings are in the order of 1-2 kWh/m2. Similar savings are got from reducing the hot water production temperature, since it allows a higher efficiency, but this could lead to problems in meeting the comfort requirements. To avoid that, in this paper the sizing of the radiators was performed according to that operating temperature. However, the sizing of radiators is usually made regardless of these issues, and problems could arise when trying to meet the set-point temperatures when reducing the hot water production temperature. Finally, the energy consumption reduction is similar in percentage terms for all the envelope retrofitting scenarios, which makes it more desirable in terms of energy intensity reduction when the envelope is less efficient. 0 5 10 15 20 25 30 35 NR.LTB.60 NR.LTB.55 NR.LTB.50 NR.CB.60 NR.CB.55 NR.CB.50 BAU.LTB.60 BAU.LTB.55 BAU.LTB.50 BAU.CB.60 BAU.CB.55 BAU.CB.50 BO.LTB.60 BO.LTB.55 BO.LTB.50 BO.CB.60 BO.CB.55 BO.CB.50 Years NR BAU BO J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 24 With the exception of the temperature set-point, the different aspects evaluated before imply different levels of investment. For example, a condensing boiler implies an overinvestment in relation to a lowtemperature one, while a lower hot water production water implies a higher radiator surface and therefore, investment. From the economic analysis, the BAU envelope with condensing boiler option and hot water production at 60ºC offers the lowest annual costs. The additional investment needed for bigger radiators when reducing the hot water temperature does not compensate the fuel consumption reduction in any case. BO envelope option needs for significantly higher investment that reduces the economic feasibility. This is mainly due to the windows replacement, which makes the investment orders of magnitudes higher than the insulation addition. However, windows replacement offers other benefits that cannot be analysed from a purely economic point of view: comfort, acoustics, etc. These results are different if the investment for the envelope retrofitting is already done when the heating system upgrade is faced (approach B). In this case, a better envelope reduces the annual costs, but also the reduction according to the reference case. In general, results have addressed the interest of integrated energy renovations, with packages that include energy savings measures with short payback periods and other measures with higher payback periods. Normally the same operating and comfort conditions are considered before and after the renovation. However, the so-called rebound effect usually occurs, meaning that energy efficiency measures lead to changes in the user that could imply a higher specific use of energy. From the results, it is observed that rebound effect can play a very important role, especially considering the low performance of the reference heating system, under which the comfort conditions were hard and expensive to meet. Thus, under a potential increase of the set-point temperature of 2ºC, some scenarios could lead to no energy savings and some retrofitting options could be directly economically unfeasible. 7 Conclusions A holistic methodology based on TRNSYS simulation has been presented for the evaluation of individual heating systems in the energy retrofitting of domestic buildings. It has been subsequently applied to a social housing building located in northern Spain, for which an already validated model is available. Different retrofitting scenarios and options have been analysed. From all of them, the temperature setpoint has resulted in the most effective to reduce the energy consumption which underlines the role played by the building user. The nature and operation of heating systems plays a significant role. Condensing boiler offer higher savings than conventional boilers but the benefits from the reduction of J. Terés-Zubiaga, A. Campos-Celador, I. González-Pino, G. Diarce. The role of the design and operation of individual heating systems for the energy retrofits of residential buildings. Energy Conversion and Management 2016, 136, 736–747 25 the operating temperature do not compensate the need for the higher investment of larger radiators. The eventual effects of the (p)rebound effect have been demonstrated very significant and it should be analysed whether this increase is a benefit or a feature not-demanded by the occupants The fact that the user interaction plays the most important role for getting significant energy savings reinforces the need of exploring new ways to achieve thermal comfort. This can be regarded as a key factor to reduce energy consumption in buildings and should be considered in further works. 8 Acknowledgements The author G. Diarce wants to thank the financial support of the Basque Government, through the Department of Education, Universities and Research's Personnel Research Training Program (2012 call). 9 References [1] Eurostat, Statistics Database - Energy statistics - Supply, transformation, consumption (2010). [2] United Nations. Sustainable Buildings & Climate Initiative, Common Carbon Metric. Protocol for Measuring Energy Use and Reporting Greenhouse Gas Emissions from Building Operations (2010). [3] European Commission, Directive 2010/31/EU on the Energy Performance of Buildings (2010). [4] C. Baek, S. Park, Changes in renovation policies in the era of sustainability, Energy Build. 47 (2012) 485-496. [5] Spanish Ministry of Public Works, Dwelling Stock Estimation (2013). [6] European Commission, Communication from the commission. Europe 2020. A strategy for smart, sustainable and inclusive growth (2010) 1-35. [7] D. Che, Y. Liu, C. Gao, Evaluation of retrofitting a conventional natural gas fired boiler into a condensing boiler, Energy Convers Manage. 45 (2004) 3251-3266. [8] S. Deng, A. Dalibard, M. Martin, Y.J. Dai, U. Eicker, R.Z. Wang, Energy supply concepts for zero energy residential buildings in humid and dry climate, Energy Convers Manage. 52 (2011) 2455-2460. [9] M. Owrak, M. Aminy, M.T. Jamal-Abad, M. Dehghan, Experiments and simulations on the thermal performance of a sunspace attached to a room including heat-storing porous bed and water tanks, Build. Environ. 92 (2015) 142-151. [10] S. Obyn, G. van Moeseke, Comparison and discussion of heating systems for single-family homes in the framework of a renovation, Energy Convers Manage. 88 (2014) 153-167. [11] J. Vrijders, L. Delem, Economical and environmental impact of low energy housing renovation, REPORT. (2008) 1-117. [12] L.C. Tagliabue, M. Maistrello, M. Fattor, Technical and Cost-optimal Evaluation of Thermal Plants for Energy Retrofitting of a Residential Building, Energy Procedia. 50 (2014) 597-602.