A STUDY OF WALL SURFACE TEMPERATURE VARIATIONS FOR HOUSING IN MODERATE CLIMATES

Transcription

1 Florianópolis Brazil 212 A STUDY OF WALL SURFACE TEMPERATURE VARIATIONS FOR HOUSING IN MODERATE CLIMATES Alterman, Dariusz 1 ; Page, Adrian 2 ; Hands, Stuart 3 ; Moffiet, Trevor 4 ; Moghtaderi, Behdad 5 1 PhD, Research Fellow, The University of Newcastle, Priority Research Centre For Energy, dariusz.alterman@newcastle.edu.au 2 PhD, Emeritus Professor, The University of Newcastle, Priority Research Centre For Energy, adrian.page@newcastle.edu.au 3 Research Assistant, The University of Newcastle, Priority Research Centre For Energy, stuart.hands@newcastle.edu.au 4 PhD, Research Associate, The University of Newcastle, Priority Research Centre For Energy, trevor.moffiet@newcastle.edu.au 5 PhD, Professor, The University of Newcastle, Priority Research Centre For Energy, behdad.moghtaderi@newcastle.edu.au This paper describes an experimental investigation of internal and external surface temperature variations of heavy and lightweight walling systems under the influence of moderate weather conditions typical of the Australian climate. Four housing test modules incorporating various walling types were built on the University of Newcastle campus and the detailed thermal performance of each system was measured over a range of seasonal conditions. The temperature gradients for various locations through the thickness of each wall assemblage are examined and discussed. The analysis helps to provide an increased understanding of the dynamic behaviour of various walling systems subjected to real weather conditions, with the preliminary study demonstrating that wall thermal resistance (R-value) is only a steady state parameter which is not capable of representing the thermal performance of lightweight and heavy walling systems as a sole descriptor. The investigations indicate that energy demands for Australian weather conditions are influenced by the combination of the thermal mass and the thermal resistance of the wall components, and both need to be considered if the thermal performance is to be realistically predicted. Keywords: thermal performance, thermal mass, thermal resistance, Australian housing INTRODUCTION Material selection during the design of a building has a major influence on the energy consumption and performance over its life cycle. The increasing emphasis on energy efficiency and the resulting reduction in carbon emissions is increasing the focus on novel design principles for all buildings, supported by the use of suitable materials. Designing for energy conservation requires study of the combination of factors that affect energy consumption. This presents major challenges for builders and architects to deliver designs that are functional whilst also energy efficient. Australian housing is typically constructed from brick veneer or cavity brick walling systems, with some single skin, partially grouted and reinforced hollow masonry construction in

2 Florianópolis Brazil 212 cyclonic areas. The environmental impact of Australian housing up to the 9 s was of the little concern, with designs being governed mainly by economic and aesthetic considerations. Over the years houses have also become larger with an increasing numbers of appliances and reliance on artificial means of heating and cooling. Although building methods have improved, there is still heavy reliance on heating and cooling systems to create comfortable interior conditions with the little emphasis on energy efficiency and the environmental impact. The current regulatory climate now requires all housing to be assessed for energy performance. There is thus a need to establish the thermal characteristics of the typical walling systems in the range of Australian climates under all seasonal conditions. In this context, researchers at the University of Newcastle, in collaboration with Think Brick Australia, are involved in a major study of the thermal performance of Australian housing with a view to utilising more effectively the benefits of thermal mass which is inherent in masonry heavy walling systems. In this study four full scale housing modules have been used to provide qualitative and quantitative data on the thermal performance of various walling systems (including lightweight as well as heavy masonry walls) under real climatic conditions. The typical heat flow mechanisms were then studied to assess the actual behaviour of various walls. This paper concentrates on a study of variations in temperature profiles of conventional and innovative walling systems under various seasonal conditions. One innovative walling system being investigated is reverse brick veneer, which is made up of a lightweight external cladding supported by a timber frame with insulation and an internal non-structural brickwork skin (that is, the direct reverse of conventional brick veneer). The potential advantage of this system compared to brick veneer is that the benefits of internal thermal mass are provided by the internal brickwork skin. The thermal performance of four walling systems (insulated brick veneer; insulated cavity brick; cavity brick and insulated reverse brick veneer) is reported and discussed for typical winter and summer days. The relative performance of the four systems varied with season, but in general, compared to conventional brick veneer the thermal mass of heavy walling systems located on the interior side of the buildings was more effective in dampening the effects of the diurnal temperature swings and maintaining the internal temperatures within acceptable levels of thermal comfort. OVERVIEW OF EXPERIMENTAL MODULES The research reported here is part of an on-going eight year study which involves a detailed investigation of the performance of the various walling systems used in Australian housing, (detailed description can be found in Page (211)). All wall elements as well as walling systems were tested in a Guarded Hot Box apparatus (constructed in accordance to the ASTM Standard, (ASTM C )) to obtain their thermal resistance (R-value). Each walling system was then incorporated into a representative full scale housing module to observe its performance in a complete building under real weather conditions. The modules were constructed on the University of Newcastle Callaghan Campus in suburban Newcastle (Newcastle is located on the east coast of Australia at latitude 33 south). Over the testing period, a range of walling systems have been used (cavity brick, insulated cavity brick, brick veneer with and without insulation, lightweight construction and insulated reverse brick

3 Florianópolis Brazil 212 veneer). For this paper, only the CB (cavity brick), InsCB (insulated cavity brick), InsBV (insulated brick veneer) and InsRBV (insulated reverse brick veneer) modules with R values of.44, 1.3, 1.58 and 1.58 respectively are considered. Each module was observed with the interior space being either in a free-floating state (directly influenced by real weather conditions), or with the interior artificially heated or cooled to a preset temperature range. The typical modules are shown in Figures 1(a) and 1(b). (a) (b) Figure 1: Brick Veneer and Two Cavity Brick Masonry Modules (a) and Insulated Reverse Brick Veneer (b) The modules had a square floor plan of 6 m x 6 m and were spaced 7 m apart to avoid shading and minimise wind obstruction. With the exception of the walls and roof, the buildings were of identical construction following standard Australian practice, being built on a concrete slab-on-ground and aligned in a manner so that the north wall of each building was aligned to astronomical north. Timber trusses were used to support the roof which consisted of tiles for the CB, InsCB and InsBV modules and steel sheeting for the InsRBV module, in both cases placed over a layer of sarking. The buildings had a ceiling height of 245 mm. The ceiling consisted of 1mm thick plasterboard with glasswool insulation batts (R3.5) placed between the rafters. Since the emphasis of the investigation was on wall performance, the R3.5 insulation was selected to minimise the through-ceiling heat flow. Entry to the buildings was via tight fitting, insulated solid timber doors located on the southern face of the buildings. The roof was supported by an independent steel frame which allowed the removal and replacement of walls as required. The modules were comparable in size to other buildings used in similar studies carried out in the 8 s at Maryland, USA and Canada, (Burch at el. 82). Initial tests were performed on windowless modules. The influence of a window was then assessed by the installation of a north-facing 3-panel sliding door assembly, mm high x 284 mm wide, representing 2% of the floor area which is a typical living room window/floor area ratio, as in Figure 1a. Instrumentation recorded the external weather conditions including wind speed and direction, air temperature, relative humidity and the incident solar radiation on each wall (vertical plane) and on the roof (horizontal plane). For each module, temperature and heat flux profiles through the walls, slab and ceiling were recorded in conjunction with the internal air temperature and relative humidity. Heat flux sensors were placed on the walls, ceilings and concrete slab, adjacent to the window (in direct sunlight) and in the south-east corner. Thermocouples were placed on the surface of the slab

4 Florianópolis Brazil 212 at various locations between the window and the centre of the room. For the window, three net radiation sensors were placed at heights of 6, 12 and 18mm up the glass panel to assess the incoming/outgoing radiation. The surface temperature of the glass was recorded and additional heat flux sensors were placed on the aluminium frame to assess the influence of the frame itself. Internal air space temperatures were also monitored at heights of 6, 12 and 18mm with the relative humidity and globe temperatures being measured centrally. In total, data channels were scanned and logged every 5 minutes for each of the modules for the duration of the testing program. The results reported here for the InsRBV, InsBV, InsCB and CB modules with a northern window are from data collected from June 27 to October 29, thus covering each of the four seasons and a range of weather conditions. Each season included periods in which the interior of the modules was allowed to free float or be controlled within a temperature range of 18-24ºC and the heating/cooling energy measured. Only the free floating results for summer and winter periods are discussed in this paper. There was no artificial ventilation and the concrete floor slabs were not fitted with carpet. Since the conditions for each module were identical, direct comparison of performance can be made. The key technique used to assess the dynamic temperature profiles of the module walls was a study of histograms of internal and external surface temperatures which were simplified using statistical parameters such as means, standard deviations and range. For the comparison, the internal air temperatures for all analysed modules were also given. DYNAMIC TEMPERATURE PROFILE ANALYSIS METHOD This paper provides insight on how the various walling systems respond to real weather conditions based on a temperature profile. Such analysis was possible as temperature sensors (thermocouples) were located through the profile of the various wall systems. All thermocouples were physically fixed to solid surfaces at the time of construction and/or installed within the masonry units. Figures 2 to 5 show the measured temperature gradients across the various wall assemblies on the st May, 28 (early winter conditions). The static comparison between modules with a temperature line (red solid line) as in Figures 2 to 5 did not provide easily measured quantitative details for the full day period due to the large number of curves for each module (typically 288). The curve shown is at 1. hours when the highest solar radiation was recorded. (Note: an animation of the temperature profiles with a 5 minute sequence illustrated the difference in heat transfer through the walls for the various modules but this was not appropriate for further statistical validation as it provided qualitative variables only). The maximum and minimum temperature profiles through the walling system for the 24 hour period are shown in Figure 2 to 5 as the dotted lines. The difference in behaviour for the different forms of construction is readily apparent.

5 Florianópolis Brazil Internal Interior Brick Insulated Timber Frame Plasterboard External Figure 2: Temperature profile for the Insulated Reverse Brick Veneer Wall, ( st May, 28). The profile is shown at 1. for the western wall (when the highest solar radiation was recorded), (R-value = 1.58 m 2. K/W) Internal 21.4 Plasterboard Insulated Timber Frame Exterior Brick External Figure 3: Temperature profile for the Insulated Brick Veneer wall, ( st May, 28), (R-value = 1.58 m 2. K/W) A maximum surface temperature was measured for the InsRBV exterior lightweight skin of.6ºc, corresponding to an external air temperature of 2.3ºC (see Figure 2). The higher surface temperature was a result of the solar radiation influence. The temperature for the modules with a masonry exterior skin varied between about and 28ºC.

6 Florianópolis Brazil Internal Interior Brick Insulation Cavity Exterior Brick 24.7 External Figure 4: Temperature profile for the Insulated Cavity Brick wall, ( st May, 28), (R-value = 1.3 m 2. K/W) Internal Interior Brick Cavity Exterior Brick Figure 5: Temperature profile for the Cavity Brick wall, ( st May, 28), (R-value =.44 m 2. K/W) External Although, the exterior lightweight skin for the InsRBV module (see Figure 2) was exposed to the highest temperature, its thermal resistance properties reduced significantly the amount of energy which was transferred to the interior side of the lightweight skin for all seasons. While this on its own could be an advantage of lightweight construction, an insulated lightweight component located on the internal side of a wall can only absorb a small amount of energy throughout the day and release it back to the interior at night for moderate climates (a major disadvantage under winter conditions). This is highlighted for InsBV in Figure 2.

7 Florianópolis Brazil 212 The static temperature profiles with the profiles limited to maximum and minimum temperatures do not adequately define the dynamic performance of the walls over the diurnal cycle. However, this can be easily achieved by analysing the temperature distributions using the small time increment original data measurements from the module tests (taken every 5 minutes). The resulting histograms in Figures 6 to 9 show the response of the walls to real weather conditions for both external and internal surfaces, and are directly related to the previous temperature profiles in Figure (a) Internal surface (b) External surface Figure 6: Temperature distribution of external and internal surface for Insulated Reverse Brick Veneer wall for a diurnal cycle, (R-value = 1.58 m 2. K/W) (a) Internal surface (b) External surface Figure 7: Temperature distribution of external and internal surface for the Insulated Brick Veneer Wall for a diurnal cycle, (R-value = 1.58 m 2. K/W) For the comparison, each module with the masonry layer exposed to either the external or internal environment responded similarly, with temperature ranges varying only because of the influence of the other components involved in the heat transfer across the walls. A broad temperature distribution on the internal surface of the wall with the wide range of temperatures over a day occurred for the InsBV (high R-value but internal lightweight skin). However, narrow temperature ranges (small variations) on internal surface were observed for

8 Florianópolis Brazil 212 he InsCB, CB and InsRBV modules. The temperature range profile on the external surface of the lightweight external skin (InsRBV) experienced the widest range between minimum and maximum temperatures of 26ºC, in comparison to 16.5ºC for the InsCB module (a) Internal surface (b) External surface Figure 8: Temperature distribution of external and internal surface for the Insulated Cavity Brick wall for a diurnal cycle, (R-value = 1.3 m 2. K/W) (a) Internal surface (b) External surface Figure 9: Temperature distribution of external and internal surface for the Cavity Brick wall for a diurnal dycle, (R-value =.44 m 2. K/W) Detailed observations of the thermal behaviour of the test modules under real weather conditions in all seasons proved that the thermal resistance (R-value) of the walls is not the sole predictor of the thermal performance of a building. Similar internal conditions were achieved in modules with significantly varied R-values (.44 m 2 K/W for the CB masonry module and 1.3 m 2 K/W for the InsCB module). The following investigates the influence of other factors on the thermal performance based on a temperature profile analysis. TEMPERATURE PROFILES ANALYSIS SELECTED WINTER DAY Comparing the mean temperature and standard deviation for each temperature profile over early winter day on st May, 28 can also provide additional insight into the ability of the walling system to reduce the effect of the daily temperature variations. General trends can be 27 29

9 Florianópolis Brazil 212 inferred using simple statistics to study mean temperature, standard deviation and range between maximum and minimum temperatures. Results for the InsBV, InsCB, CB and InsRBV modules are easily comparable and presented in Table 1. Table 1: Statistical parameters of temperature distribution for st May, 28 Construction Internal ( C) Internal Surface ( C) External Surface ( C) Mean Std Dev Range Mean Std Dev Range Mean Std Dev Range InsRBV (R=1.58) InsBV (R=1.58) InsCB (R=1.3) CB (R=.44) Note: external mean temperature of 13.5ºC, standard deviation of 4.4 ºC and range of 13.3 ºC were recorded for all modules Graphical representation of the statistical parameters is also shown in Figure 1. The recognised variations between modules, shown by the previous histogram analyses, are also apparent for the InsCB and InsRBV modules as well as the InsBV and CB modules. Note: ranges of temperatures are only presented in Table 1. (a) Temperature variations InsRBV (R=1.58) InsBV (R=1.58) 1.4 InsCB (R=1.3) CB (R=.44) InsRBV (R=1.58) InsBV (R=1.58) InsCB (R=1.3) Figure 1: Modules comparison in terms of Mean and Standard Deviation Temperatures for the winter period on the internal surfaces (a) and external surfaces (b). Note: differing temperature scale CB (R=.44) The red dots in Figure 1a represent the mean internal air temperatures for each module, which almost correspond to the surface temperature for the InsBV, InsCB and CB modules. The dotted line in Figure 1(b) shows the mean external air temperature. TEMPERATURE PROFILES ANALYSIS SELECTED SUMMER MONTH The dynamic profile analysis of the modules with a north facing window is presented in Table 2 for a typical summer month (December 27). The additional insulation layer in the InsCB masonry module had the lowest surface temperatures of all the modules because of its thermal mass which allowed energy to be absorbed into both the interior and exterior skin. Although, the external surface of the InsRBV module was exposed to a very high temperature range of 45.6ºC (compared to about ºC for the masonry external surface modules). The histograms showed that the amount of the time for the surface temperature higher than 26ºC was almost 12 hours for the InsBV module, 1.5, 3.5, and 1.5 hours, respectively for the InsRBV, InsCB and CB modules. Note: the histogram plots for the summer period are not (b)

10 Florianópolis Brazil 212 presented in this paper. Table 2: Statistical parameters of temperature distribution for the summer period Construction Internal ( C) Internal Surface ( C) External Surface ( C) Mean Std Dev Range Mean Std Dev Range Mean Std Dev Range InsBV (R=1.58) InsRBV (R=1.58) InsCB (R=1.3) CB (R=.44) Note: external mean temperature of 24.7ºC, standard deviation of 6.2 ºC and range of 22.4 ºC were recorded for all modules CONCLUSIONS The comparative dynamic thermal profiles of four walling systems (including lightweight as well as heavy masonry walls) have been described and discussed for typical Australian walling systems. The study confirmed the premise that the R-value alone (a steady-state parameter) cannot represent the real dynamic performance of a building. The relative performance of the walling systems varied with season, but in general, the thermal mass of the internal skins of the CB and InsRBV modules was more effective in absorbing energy from solar ingress over winter seasons and dampening the effects of the external diurnal temperature swings over summer periods. Generally only minor differences were observed in the performance of the insulated cavity brick and reverse brick veneer walling systems. Further dynamic analyses in controlled environmental conditions (using a guarded hot box) will be performed to develop an indication of the real behaviour of walling systems. ACKNOWLEDGEMENTS This research has been supported by Think Brick Australia and Australian Research Council. Their support and the assistance of Civil Engineering laboratory staff are gratefully acknowledged. REFERENCES ASTM C , (97): Standard Test Method for the Thermal Performance of Building Assemblies by Means of a Hot Box Apparatus, American Society for Testing Materials, Philadelphia, USA, 97. BURCH, D. M., REMMERT, W. E., KRINTZ, D. F. and BARNES C. S. (82), A Field Study of the Effect of Wall Mass on the Heating and Cooling Loads of Residential Buildings, Proceedings of the Building Thermal Mass Seminar, Knoxville, Tennessee, NBS, USA, pp Page A. W., Moghtaderi B., Alterman D. and Hands S., A Study of the Thermal Performance of Australian Housing, Priority Research Centre for Energy, the University of Newcastle, 211 (available on: