A comparative study on the applicability of six radiant floor, wall, and ceiling heating systems based on thermal performance analysis

A comparative study on the applicability of six radiant floor, wall, and ceiling heating systems based on thermal performance analysis ORAVEC, J.; ŠIKULA, O.; KRAJČÍK, M.; ARICI, M.; MOHAPL, M. Journal of Building Engineering Volume 36, April 2021, 102133, Pages 1-11 ISSN: 2352-7102 DOI: https://doi.org/10.1016/j.jobe.2020.102133 Accepted manuscript © 2021. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ dspace.vutbr.cz Highlights  Applicability of six floor, wall, and ceiling radiant heating systems was studied  Output, controllability, heat storage, building retrofit and costs were considered  Wall with pipes in plaster performed best, had fast response but limited storage  Classical floor heating performed consistently in all criteria  Ceiling TABS was feasible only when long-term heat storage was needed Highlights 1 A comparative study on the applicability of six radiant floor, wall, and ceiling heating systems 1 based on thermal performance analysis 2 3 Jakub Oravec1, Ondřej Šikula1, Michal Krajčík2,*, Müslüm Arıcı3, Martin Mohapl1 4 5 1Brno University of Technology, Faculty of Civil Engineering, Veveří 331/95, 60200 Brno, Czechia 6 2Slovak University of Technology, Faculty of Civil Engineering, Radlinského 11, 81005 Bratislava, 7 Slovakia 8 3Kocaeli University, Engineering Faculty, Mechanical Engineering Department, Umuttepe Campus, 9 41001 Kocaeli, Turkey 10 11 *Corresponding e-mail: [email protected] 12 13 Abstract 14 Holistic comparisons of radiant heating systems that would help make an informed decision on the 15 selection of the most convenient system for the specific application are lacking. The applicability of six 16 representative radiant floor, wall, and ceiling heating systems was therefore compared in terms of 17 thermal output and surface area required, controllability, short-term and long-term heat storage, 18 suitability for building retrofit, and investments. Temperature and heat flux distribution in the structure, 19 time constant τ63, response time τ90, and the number of operating cycles were computed by a custom20 made and verified software tool using the finite volume method. Thermal energy stored was used to 21 determine the ability of energy storage, whereas investment costs indicated affordability. Wall heating 22 with pipes attached to a thermally insulating core had the highest thermal output, was easy to control, 23 suitable for building retrofit, and most affordable while providing limited thermal storage. The 24 performance of the wall system was retained when locating the pipes in plasterboard separated from 25 the core by an air gap. Floor heating performed consistently in all the aspects evaluated. It was 26 demonstrated that inserting a metal fin between pipes and the concrete spread layer improved thermal 27 output, controllability, and storage capacity of the floor system with minor effect on investments. 28 Ceiling with pipes insulated from the core performed well when thermal storage was not required. 29 Ceiling with pipes embedded in the core was only feasible when long-term heat storage was needed. 30 31 Manuscript File [For Revision, Please upload clean version of Revised manuscript] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 2 Keywords: Radiant heating; thermal response; heat transfer; building retrofit; performance; thermally 1 active building systems (TABS) 2 3 1. Introduction 4 5 Water-based radiant systems present a potentially viable solution for space heating because they are 6 suitable for integration with renewable energy sources [1,2,3] and have the ability to create a 7 comfortable thermal environment [4,5,6,7]. The applicability of the individual system types depends on 8 their location (floor, wall, or ceiling), the configuration of material layers, and the level of thermal mass. 9 These characteristics are crucial for the selection of the most suitable system for the specific situation 10 such as the construction of new building vs. retrofitting of an existing building, thermal storage vs. fast 11 thermal response, and traditional vs. low-temperature renewable heat source. 12 Several studies have compared the radiant systems assuming various locations and 13 configurations of the material layers. The first category of studies involves calculations performed for a 14 representative fragment of the heating or cooling element. Oxizidis and Papadopoulos [8] compared 15 the energy performance and thermal comfort created by a floor cooling system, ceiling with pipes 16 embedded in plaster, wall with pipes located in the plaster, and a generic thermally active building 17 system (TABS). Although floor cooling consumed the least energy, it could not provide enough cooling 18 output to attain thermal comfort. Ning et al. [9] calculated the response times of typical radiant systems 19 as defined in ISO 11855 [10]. The systems were classified into three categories according to their 20 thermal response: fast, medium, and slow. The difference in response times of ceiling/floor 21 heating/cooling systems was explained by the different heat transfer coefficients between the radiant 22 surface and indoor air. Krajčík and Šikula [11] compared four wall cooling systems. The system’s 23 suitability depended on the requirements such as exploiting thermal storage, avoiding interventions on 24 the interior surfaces, or attaining a rapid thermal response. The cooling output was sensitive to 25 insulation thickness for the systems with pipes located in the thermal core and to pipe spacing for the 26 systems having pipes underneath the surface. 27 The second category of studies involves experiments and computer simulations on a whole-room 28 level. Mustakallio et al. [12] experimentally compared a chilled beam, chilled beam combined with a 29 radiant panel, chilled ceiling combined with mixing ventilation, and four localized cooling panels 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 3 combined with mixing ventilation. The differences in the thermal environment created by the systems 1 were small. Le Dréau and Heiselberg [13] performed computer simulations of an active chilled beam, 2 radiant floor, radiant ceiling, and radiant wall in an occupied room. Using radiant floor resulted in the 3 lowest cooling demand but created least uniform thermal conditions. The most homogeneous thermal 4 environment was attained by the radiant ceiling. Bojić et al. [14] numerically investigated the 5 performance of ceiling, wall, floor, and floor-ceiling heating systems. The floor-ceiling system 6 performed best in terms of energy and exergy saving, exergy destruction, CO2 emissions, and 7 operation costs whereas a single ceiling system was the least preferable option. Karabay et al. [15] 8 recommended using wall over floor heating because it provided higher thermal performance and more 9 favourable thermal conditions with lower water temperature, thus reducing energy consumption. On 10 the other hand, the computer simulations by Myhren and Holmberg [16] showed that the vertical 11 temperature gradient in the centre of the room was smaller for floor heating than for wall heating. 12 Thermal performance of radiant systems represented by, e.g., heating capacity, thermal 13 resistance, and thermal response, is more relevant for design, testing and control of radiant systems 14 compared to geometry and structure [9]. The thermal response is a decisive factor to determine the 15 control strategy that is appropriate for the specific application. It is especially important for the design 16 and operation of radiant systems with larger amounts of thermal mass. The related literature describes 17 several approaches to assess the dynamic thermal performance of radiant systems. The literature 18 survey has shown that response time τ63, also referred to as the time constant, is frequently used as 19 an indicator of thermal response. Time constant represents the time to reach 63% of the final value of 20 the surface temperature, thermal output, or room temperature [17,18,19]. Alternatively, other 21 percentages of the final value, e.g. 80% [20], 95% [9], may be used for the calculation. Depending on 22 the radiant system used, the response time ranges from several minutes for, e.g., suspended ceiling 23 panels up to several tens of hours for TABS [9]. A system with such a long response time requires 24 using precise control strategy to provide comfortable conditions [21,22,23]. 25 As reported by Ning et al. [9], a single number like τ63 or τ95 may not be fully representative of the 26 thermal response of radiant systems. In such a case, calculating several response times, e.g., τ25, τ50, 27 τ63, or τ80 may be needed [9,24]. Other indicators of thermal response found in the related literature 28 include the peak value of surface temperature [25], thermal admittance and transmittance [26,27], 29 visual comparison of the step-up and decay curves of surface temperature [20,28,29,30], the required 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 4 number of operation cycles to maintain thermal output between defined boundaries [11], and the heat 1 storage efficiency that takes into account the evolution of surface temperature until reaching steady2 state [31]. 3 The literature review has shown that the existing studies usually compare the systems only from 4 thermal comfort or energy efficiency point of view. Most of the research focuses on systems with 5 massive thermal layers, whereas less attention is given to lightweight systems and their comparison 6 with the massive systems. Moreover, the potential use of radiant systems for building retrofit has rarely 7 been considered. It follows that holistic comparisons of radiant heating systems that would help make 8 an informed decision on the selection of the most convenient system are lacking. The present study 9 aims to facilitate the selection of the most suitable radiant heating system for both newly constructed 10 and renovated buildings. The emphasize is on the thermal performance of the systems while also 11 considering their price and suitability for building retrofit. 12 To accomplish this, numerical investigations of six representative radiant floor, wall, and ceiling 13 heating systems (Figure 1) have been performed in terms of thermal output, area of the heating 14 surface required, controllability, short-term and long-term energy storage, suitability for building retrofit, 15 and investment costs. Thermal fields and heat flux distribution in the structure were computed to 16 evaluate the thermal output. Time constant (τ63), response time (τ90), and the number of operating 17 cycles were used to assess the thermal response and controllability. Thermal energy stored in the 18 structure and investment costs were also determined to help evaluate the systems in a broader 19 context. 20 21 2. Radiant heating systems investigated 22 23 Figure 1 shows the composition and configuration of the heating systems studied. The systems were 24 designed to cover the design heat load which consisted of the design heat loss (744 W) and the heat25 up capacity (10% of the design heat loss). The design and construction of the systems were supposed 26 to closely reflect reality. Therefore, the differences between the systems in this respect were 27 considered. The pipe spacing and diameter varied depending on the heating system used, which 28 resulted in differences in the heating area, thermal output, and surface temperature. For example, the 29 pipe spacing of the wall heating systems is denser than that of the floor heating systems, therefore the 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 5 area is smaller and the surface temperature is higher for wall heating. All the heating systems were 1 located between two conditioned rooms. The systems are described as follows: 2 (A) Floor – classical floor heating consisting of a bearing structure, system board made of a 3 thermally insulating material, pipes embedded in the system board, a concrete spread layer, 4 and wooden parquets. The outer and inner diameter of the pipes was 16 and 14 mm, 5 respectively, and the pipe spacing was 150 mm. 6 (B) Floor with metal fins – like A, but a metal fin with high conductivity was inserted between the 7 pipe and the concrete spread layer to enhance heat transfer. 8 (C) Wall (TABS) – pipes embedded in plaster underneath the surface of an internal wall. The 9 outer and inner diameter of the pipes was 12 and 10 mm, respectively, and the pipe spacing 10 was 120 mm. 11 (D) Wall (air gap) – pipes with an outer diameter of 10 mm and an inner diameter of 9 mm, spaced 12 by 80 mm and embedded in a 15 mm plasterboard. The plasterboard panels can have various 13 dimensions and be connected in series. The thermally active plasterboard is decoupled from 14 the thermal core by an air gap, which may be filled with thermal insulation. 15 (E) Ceiling (TABS) – pipes embedded in the thermal core that is insulated to prevent heat loss to 16 adjacent rooms. The outer and inner diameter of the pipes was 20 and 18 mm, respectively, 17 and the pipe spacing was 300 mm. 18 (F) Ceiling (TI) – the heat was emitted through plasterboard connected with pipes located in an air 19 gap. The thermally active layer was insulated from the main thermal mass to prevent heat 20 storage and ensure a fast thermal response. 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 6 Floor covering Screed Building structure Insulation Pipes Wall Plaster Pipes Wall Plaster Pipes Air gap Plasterboard Floor covering Screed Building structure Insulation Pipes Plaster Floor covering Screed Building structure Metal construction Pipes Plasterboard Insulation Air gap A) Floor C) Wall (TABS) D) Wall (air gap) E) Ceiling (TABS) F) Ceiling (TI) Floor covering Screed Pipes Metal fin B) Floor with metal fins Insulation 10mm 70mm 20mm 250mm 300mm 5mm 15mm 35mm 250mm 10mm 70mm 20mm 250mm 10mm 30mm 300mm 10mm 50mm 250mm 20mm 40mm Building structure 10mm 70mm 20mm 10mm 1 Figure 1. Radiant heating systems investigated: A – Floor, B – Floor with metal fins, C – Wall (TABS), 2 D – Wall (air gap), E – Ceiling (TABS), F – Ceiling (TI). 3 4 3. Methodology 5 6 The heating system was located in the living room of a residential building. The dimensions of the 7 room were 6 m x 4 m which can be considered usual in the region of Central Europe. The room design 8 heat load was determined to be 818 W following EN 12831-1 [32]. This heat loss corresponds to an 9 external temperature of -12°C and a room operative temperature of 20°C. These temperatures are 10 representative of the design conditions in the humid continental climate typical of, e.g., Central and 11 Eastern Europe or places like Maine and Michigan, USA. The mean temperature of the heating water 12 was 35°C corresponding to the thermal gradient of 38/32°C. 13 In the present study, computer simulations were used as the research method because they allow 14 to precisely formulate of the boundary conditions and retain the conditions for all the systems tested 15 while being less costly than experimental measurements. A custom-made and verified software tool was 16 employed to compute the temperature and heat flux distribution in the structure, time constant τ63, 17 response time τ90, number of operating cycles, and thermal energy stored using the finite volume 18 method. 19 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 7 3.1 Physical model and calculation principle of heat transfer 1 The heat transfer rate was calculated for a characteristic fragment of the structure. Figure 2 shows the 2 configuration and composition of the material layers for each of the fragments as defined in the 3 calculation model. The thermophysical properties of the individual materials are defined in Table 1. In 4 the analysis, the thermophysical properties of materials were assumed to be isotropic, temperature 5 independent, and constant. The heat was emitted to the environment through the surfaces facing the 6 rooms. 7 8 Figure 2. Radiant heating systems as defined in the calculation model: A – Floor, B – Floor with metal 9 fins, C – Wall (TABS), D – Wall (air gap), E – Ceiling (TABS), F – Ceiling (TI). Troom – room 10 temperature. 11 Table 1. Thermophysical properties of materials. 12 Material Colour Volumetric weight ρ (kg/m3) Thermal conductivity λ (W/(m·K)) Specific heat capacity c (J/(kg·K)) 1 – Concrete 2 – Reinforced concrete 3 – Thermal insulation 4 – Hardwood 5 – Metal fin (steel) 1 mm 6 – Air gap – system D 7 – Air gap – system F 8 – Aerated concrete 9 – PE-Xa pipe 10 – Plaster 11 – Plasterboard ███ ███ ███ ███ ███ ███ ███ ███ ███ ███ ███ 2100.0 2400.0 25.0 600.0 7850.0 1.2 1.2 400.0 1270.0 2000.0 750.0 1.23 1.58 0.04 0.22 58.00 0.20 0.01 0.15 0.45 0.99 0.22 1020 1020 1270 2010 440 1010 1010 1000 1980 790 1060 13 The heat flux and temperature distribution for stationary and dynamic analyses were obtained by 14 CalA software [33,34]. The software was developed by one of the authors of this study to calculate 15 two-dimensional heat and moisture transfer in building structures and was verified following the 16 procedure as defined in ISO 11855 [10], Part 2, Annex D. For the verification of the software, the 17 Room – Newton´s law, Troom = 20°C A B F E D C 1 3 2 4 1 2 3 4 5 8 8 6 11 3 2 1 4 4 1 2 3 11 5 7 10 10 10 Room – Newton´s law, Troom = 20°C Adiabatic boundary 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 14 1 Figure 7. Time constant τ63 and response time τ90. 2 3 4.2.2 Number of operating cycles 4 A simplified control strategy was employed to compare the controllability of the systems through the 5 number of operating cycles. This control strategy involved turning the system on and off to keep the 6 thermal output within a defined range over 24 h (Figure 8). The system was turned on at time t = 0 s 7 and continuously operated at a constant water temperature of 35°C until the thermal output reached 8 90% of its maximum value. Then the heating system was turned off until the surface cooled down to 9 the temperature corresponding to 66%, i.e. two-thirds of the maximum thermal output, after which it 10 was turned on again. The number of operating cycles was counted to evaluate the controllability. 11 Adding metal fins to the floor heating system (A, B), increased the number of operating cycles 12 thereby enhancing the system´s controllability (Figure 8). Wall TABS (C) was easy to control and had 13 the highest thermal output, whereas the wall with an air gap (D) had even better controllability at a 14 slightly lower output. The difference was most remarkable for the ceiling systems. Although their 15 maximum output was similar, ceiling with pipes insulated from the core (F) was operated dynamically 16 whereas ceiling TABS (E) gradually discharged heat providing no possibility of dynamic control. 17 18 19 20 21 22 0 10 20 30 40 50 60 70 80 90 100 Percentage of max. output Time elapsed (h) Floor Floor, metal fin Wall (TABS) Wall (air gap) Ceiling (TABS) Ceiling (TI) 63% of max. output 90% of max. output 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 15 1 2 3 4 5 6 Figure 8. Number of operation cycles (N): A – Floor, B – Floor with metal fins, C – Wall (TABS), D – 7 Wall (air gap), E – Ceiling (TABS), F – Ceiling (TI). 8 9 4.3 Thermal energy stored 10 Energy stored in the structure was computed to evaluate the systems’ suitability for discontinuous heat 11 supply. This information is useful when planning the control and operation strategy. The more energy 12 20 22 24 26 28 30 32 34 0 10 20 30 40 50 60 70 80 90 100 Temperature (°C) Thermal output (W/m2) Hour of day 20 22 24 26 28 30 32 34 0 10 20 30 40 50 60 70 80 90 100 Temperature (°C) Thermal output (W/m2) Hour of day 20 22 24 26 28 30 32 34 0 20 40 60 80 100 120 140 160 Temperature (°C) Thermal output (W/m2) Hour of day 20 22 24 26 28 30 32 34 0 20 40 60 80 100 120 140 160 Temperature (°C) Thermal output (W/m2) Hour of day 20 22 24 26 28 30 32 34 0 20 40 60 80 100 120 140 160 Temperature (°C) Thermal output (W/m2) Hour of day 20 22 24 26 28 30 32 34 0 20 40 60 80 100 120 140 160 Temperature (°C) Thermal output (W/m2) Hour of day B C D E F A Output Temperature Heating ON/OFF N = 3 N = 4.5 N = 13 N = 23.5 N = 0.5 N = 32 Heating ON Heating OFF 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 16 a heating system accumulates, the fewer operating cycles of the heat source may be needed to keep 1 the room temperature within a defined range. In the present study, two levels of thermal storage have 2 been considered: short-term and long-term. The short-term storage was defined as the energy stored 3 in the structure after three hours of operation. It shows the capability of a system to reduce the number 4 of operating cycles of the heat source or cover short outages of the source. The long-term storage was 5 defined as the energy stored in the structure after a 12-hour operation. It indicates the capability of a 6 system to accumulate and discharge heat throughout the day. Such a system can be suitable for 7 example for night heat accumulation. 8 Figure 9 shows the thermal energy stored in the structure over 24 hours. After three hours of 9 operation, floor heating with metal fins contained most thermal energy, followed by ceiling TABS and 10 floor heating. Most thermal energy was eventually stored in the ceiling TABS because of its great heat 11 storage capacity. Thus, the floor systems are suitable for short-term heat storage, whereas ceiling 12 TABS is suitable for both shortand long-term heat storage. The other systems are not suitable for 13 heat storage and require a heat source with higher heating power and operated intermittently 14 throughout the day. Alternatively, the storage capacity can be provided by a tank. 15 16 Figure 9. Thermal energy stored (Estored) in the structure over 24 hours of continuous operation. 17 18 4.4 Construction costs 19 20 Construction costs have been compared using the overall construction costs (€/m2) and price per 21 thermal output (€/W) as shown in Figure 10. All construction costs refer to EUR (€) per m2 of the floor 22 0 200 400 600 800 1,000 1,200 1,400 1,600 Thermal energy stored (Wh/m2) Time elapsed (h) Floor Floor, metal fin Wall (TABS) Wall (air gap) Ceiling (TABS) Ceiling (TI) Estored (3 h) Estored (12 h) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 17 area of the living room. The unit prices have been taken from a dedicated database which is routinely 1 used by professionals to calculate construction prices in the Czech Republic [42]. All the costs 2 reported represent standardized costs, not including VAT or any discounts. The prices of materials 3 include the heating element, distribution pipes, control elements and actuators as well as surface 4 finishing. The heat source was not included. The prices also include wages to install the equipment, 5 supporting structures, transfer of materials, and functionality tests. Figure 11 shows standard hours 6 representing a standardized amount of time needed to install the system per m2 of floor area. 7 System A is the most common one from all the radiant heating systems considered. Systems B 8 and C are from the technology point of view modifications of system A where the pipes are embedded 9 in a thermally conductive material such as concrete layer or plaster. System F requires installations on 10 the ceiling and the prices were increased correspondingly to reflect the additional effort. In system E, 11 the heating element is an inherent part of the ceiling structure. The price for this system considers the 12 space heating function of the slab, but not the statics. 13 Figures 10 and 11 show that Wall TABS (C) has the lowest price per thermal output despite taking 14 the most time to construct per m2. On the other hand, wall system D takes the least time to construct, 15 has the best output per standard hour, but has relatively high overall construction costs because it 16 contains many supporting structures and the components are expensive. The price of systems A, B, 17 D, E per m2 is similar, whereas ceiling system F is the most expensive. The reason for system F being 18 the most expensive is the high cost of the prefabricated ceiling panels. It is also shown that adding a 19 metal fin in a floor heating system is not too costly which can make this system (B) a feasible option. 20 21 0 20 40 60 80 100 120 140 160 0 0.5 1 1.5 2 2.5 3 Floor Floor, metal fin Wall (TABS) Wall (air gap) Ceiling (TABS) Ceiling (TI) Construction costs (€/m2) Price/Output (€/W) Price/Output (€/W) Construction costs (€/m2) A B C D E F 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 18 Figure 10. Construction costs of the radiant heating systems. 1 2 Figure 11. Standard hours needed to install 1 m2 of the heating system. 3 4 Operating costs can differ for various systems with respect to the differences in thermal output and 5 control strategy. The overall thermal output of the system depends on the area of the heating surface. 6 Therefore, the surface area of each system needs to be adjusted to cover the room design heat loss. 7 The differences in the costs between the systems due to the different thermal output per m2 of surface 8 area are therefore reflected in the price per output (Figure 10). 9 When each system provides the same overall thermal output to cover the room design heat loss, 10 the difference in operating costs will be only caused by differences in the control strategy. In this 11 respect, the advantage of the systems with fast thermal response is a more accurate control. The 12 favourable features of the systems with a slow thermal response are the reduced number of operating 13 cycles and lower peak power of the heat source. An accurate investigation of the effect of control 14 strategy on operation costs would require additional complex calculations and is beyond the scope of 15 this study. 16 17 5. Discussion 18 19 5.1 Sensitivity of thermal output and thermal response to system design 20 0 0.5 1 1.5 2 2.5 3 0 5 10 15 20 25 30 35 40 45 Floor Floor, metal fin Wall (TABS) Wall (air gap) Ceiling (TABS) Ceiling (TI) Standard hours Output/Standard hour (W/m2/Sh) Output/Standard hour (W/m2/Sh) Standard hours A B C D E F 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 19 The parametric simulations have confirmed the sensitivity of thermal output to pipe spacing, hi, water 1 temperature, and room temperature. The thermal output is also sensitive to the distance of pipes from 2 the surface (Table 3). The effect of concrete and insulation thickness has not been explicitly 3 considered in the present study, but certain guidance is provided in the existing studies. The studies 4 have shown that insulation thickness has a substantial effect on the thermal output for systems with 5 pipes located in the thermal core, whereas the concrete thickness does not. The insulation thickness is 6 of particular importance for the walls having a thermally conductive core [11,35]. Ning et al. [9] found 7 that pipe spacing, concrete type, and concrete thickness have a considerable effect on the response 8 time of TABS with pipes located in thermal core like system E in the present study, whereas room 9 temperature, water temperature, water flow regime, and pipe diameter do not. Mosa et al. [43] 10 emphasize that dendritic flow architecture and a compact geometry of the panels have a positive effect 11 on the heating and cooling capacity. 12 13 5.2 Thermal performance of the heating systems 14 The thermal output is greatest for the wall systems (C, D) with pipes embedded underneath the 15 surface (Table 3, Figure 5). Although the ceiling with pipes attached to plasterboard (F) also has the 16 pipes located underneath the surface, its thermal output is lower because the heat transfer is hindered 17 by locating the pipes in an air gap. Despite the thermal output of system F being lower than that of 18 systems C and D, it has better controllability as indicated by its fast thermal response (Figure 7) and 19 less operating cycles (Figure 8). 20 The floor systems (A, B) and ceiling TABS (E) tend to store heat as shown in Figure 9. Compared 21 to ceiling TABS, the floor systems have better controllability (Figure 8) while still providing the potential 22 for short-term heat storage. Using metal fin to enhance heat transfer in classical floor heating systems 23 is encouraged because it improves thermal output (Table 3) and controllability (Figure 8) while 24 increasing the heat storage capacity (Figure 11). 25 A very good performance was attained for wall TABS (C) which provided the greatest thermal 26 output (Table 3) and was easy to control (Figures 7 and 8) despite having the pipes attached to the 27 thermal core. This was caused by the low thermal conductivity of the thermal core made of aerated 28 concrete. This makes Wall TABS (C) suitable for installation on existing walls made of aerated 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 20 concrete as a part of building retrofit. Similar results can be attained in case of a conductive thermal 1 core if thermal insulation is used to separate the pipes from the thermal core. 2 3 5.3 The effect of heat load on thermal performance of the heating systems 4 The results have been elaborated for the design heating conditions. In reality, the heat load is usually 5 lower due to heat gains and milder climatic conditions. To investigate the effect of heat load reduction 6 on the thermal performance of the heating systems, the simulations were also performed at a higher 7 external temperature than the design value of -12°C. The value of 3.6°C was used, which corresponds 8 to the average external temperature in Brno, Czech Republic in the heating season. This temperature 9 can be considered usual in the region of Central Europe in winter as well as early spring and late 10 autumn. The corresponding temperature of the heating water was determined from the heating curve 11 (Figure 12). The solid line in the middle represents the mean temperature of the heating water, 12 whereas the dashed and dash-dot lines represent the inlet and return water temperatures, 13 respectively. The design temperature gradient at -12°C is 38/32°C and the range of water temperature 14 is narrowing down as the external temperature is increasing and the water temperature is decreasing. 15 As expected, a lower external temperature lead to a reduction of the thermal output and thermal 16 energy stored. The output dropped by 50±1 % for all the systems. Even though the relative drop in 17 thermal output was about the same for all systems, in absolute values the drop was more rapid for the 18 systems with higher design thermal output. For example, changing the external temperature from - 19 12°C to 3.6 °C resulted in a reduction of the thermal output of the wall system including pipes attached 20 to an insulating core (C) by 38 W/m2 whereas the output of ceiling TABS (E) dropped by only 22 W/m2. 21 Decreasing the heat load had negligible influence on the thermal response of the systems as 22 described by the time constant τ63 and response time τ90. 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 21 1 Figure 12. Determination of water temperature from the heating curve. 2 3 5.4 Suitability for building retrofit 4 The suitability of a radiant system for building retrofit depends on criteria such as the heating area 5 required, ease of installation, and the extent of changes induced in the building structures. The 6 advantage of walls over floor and ceiling systems is that they do not reduce the net story height and 7 do not require substantial changes in building structures. They have a rapid thermal response (Figures 8 7), good controllability (Figure 8), and their thermal output is higher than that of radiant floors and 9 ceilings owing to their construction and higher permissible surface temperatures (Table 3). The 10 disadvantage could be the lower angle factor between the occupant and the wall, and that 11 interventions in the wall need to be done with caution to prevent damaging the pipes. 12 Floor heating is a realistic solution for building retrofit. Besides creating a homogeneous thermal 13 environment [4,5], its heating capacity is higher than that of ceilings [3], and the angle factor between 14 the floor and the occupant is higher as compared to walls and ceilings. The limitation is that it either 15 reduces the net story height or requires destructing the existing floor. 16 Ceiling with pipes attached to plasterboard (F) is potentially suitable for building retrofit. As 17 compared to floor heating, no destruction of the floor is needed and it is less limited by obstacles such 18 as the furniture. Its thermal response is fastest from all the heating systems investigated and it 19 provides easy control of the room temperature (Figures 7 and 8). On the other hand, its heating 20 capacity is lower than that of floors and walls [3] and it reduces the story height. Ceiling TABS (E) with 21 20 22 24 26 28 30 32 34 36 38 Water temperature, Twater (°C) External air temperature, Text (°C) Inlet water temp. Mean water temp. Return water temp. Twater = -0.4828*Text + 29.207 Text = 3.6°C Twater = 27.5°C 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 22 pipes located in the core has limited applicability in building retrofit. A system with pipes attached to 1 the surface could be an alternative because it is easy to install and responds faster. 2 3 5.5 Overall evaluation 4 The performance and applicability of the heating systems are summarized in Table 4. The summary 5 given in the table is meant to facilitate the decision-making process to choose the most suitable 6 system for the specific situation. The evaluation is qualitative to a certain extent and reflects the view 7 of the authors which is based on the quantitative indicators elaborated in this study and professional 8 experience. The evaluation is based on Table 3 and Figures 5 and 6 for thermal output, Figures 7 and 9 8 for controllability, Figure 9 for energy storage, Figures 10 and 11 for construction costs, and the 10 discussion in Section 8.3 for building retrofit. 11 The summary in Table 4 shows that floor heating performs most consistently in all the aspects 12 evaluated. Adding metal fin has only a minor effect on the investment but it enhances the output and 13 energy storage of the floor heating system. Wall systems are preferable when good controllability and 14 no thermal storage are required. In such a case, system C (wall TABS) is especially suitable for new 15 buildings as well as building retrofit. 16 Table 4. Performance and applicability of the heating systems. 17 Heating system Criterion Thermal output Controllability Short-term en. storage Long-term en. storage Price/therm. output Building retrofit A Floor + + + + + + B Floor w. metal fin + + ++ + + + C Wall (TABS) ++ ++ o o ++ ++ D Wall (TI) ++ ++ - - + ++ E Ceiling (TABS) o o + ++ + - F Ceiling (TI) + ++ + - o + Key: ++ very good performance; + good performance; o mediocre performance; - not suitable 1) Applies to a thermally insulating thermal core. In case of a conductive core, the evaluation may be applicable if the pipes are thermally decoupled from the core by insulation. 2) Applies to pipes embedded in the structure. The system may be well applicable for retrofit if the pipes are attached to the core’s surface. 18 6. Conclusion 19 20 At present, holistic comparisons of various types of radiant systems that would guide the selection of 21 the most suitable heating system for the specific situation are lacking. The results elaborated in this 22 study should therefore facilitate the process of selecting the most convenient radiant heating system 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 23 for both newly constructed and renovated buildings. To accomplish this, a comparative study of six 1 representative radiant wall, floor and ceiling systems has been conducted in terms of their thermal 2 output, area of the heating surface required, controllability, short-term and long-term energy storage, 3 suitability for installation in existing buildings, and construction costs. Temperature and heat flux 4 distribution was calculated to determine the thermal output and area needed. Time constant (τ63), 5 response time (τ90), and the number of operation cycles were used to evaluate the controllability. 6 Thermal energy stored revealed the ability of shortand long-term heat storage, whereas construction 7 costs indicated affordability of the systems. The conclusions are summarized as follows: 8  The wall systems (C, D) were preferable when good controllability and no thermal storage 9 were required. Due to their versatility and fast response, they are especially suitable for 10 building retrofit. Wall TABS (C) with a thermally insulating thermal core also provided a certain 11 potential for thermal storage while being the most affordable of all the systems. 12  Floor heating (A) performed most consistently in all the aspects evaluated indicating its 13 universal use. Adding metal fins between pipes and the concrete spread layer (B) had only a 14 minor effect on the investment costs and it enhanced the thermal output and controllability 15 while increasing storage capacity. 16  Ceiling with pipes insulated from the thermal core (F) performed well when thermal storage 17 was not required and is potentially suitable for building retrofit. Ceiling TABS (E) was the 18 costliest system but might be feasible when long-term heat storage is needed. An alternative 19 with pipes attached to the core’s surface could be preferable. 20  Thermal response is a more important parameter to consider than thermal output. Maximum 21 thermal output can be adjusted by variating the pipe spacing and active surface area. The 22 thermal response is an inherent characteristic of the heating system, which has a significant 23 effect on its controllability and the room’s thermal balance. 24 Future work should include CFD simulations of a room with radiant surfaces and realistic 25 occupancy profiles. The simulations would provide additional information about temperature 26 distribution and air velocity fields in the room. Besides, the calculations should be extended to cooling 27 operation and the year-round operation of the systems should be discussed. 28 29 Acknowledgements 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65