Estimation of U-value of traditional North African houses

Applied Thermal Engineering 31 (2011) 1923e1928 Contents lists available at ScienceDirect Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng Estimation of U-value of traditional North African houses Bashir M. Suleiman Department of Applied Physics, College of Sciences, University of Sharjah, P.O. Box 27272, Sharjah, United Arab Emirates a r t i c l e i n f o a b s t r a c t Article history: Received 11 June 2010 Accepted 20 February 2011 Available online 9 March 2011 The heat transfer and the U-values of wall layers in traditional North African houses in Libya have been estimated. The locally produced building materials during 1970’s were used in the construction of such buildings. These materials were limited to bricks, tiles, cement plasters, mortar and ground soils without any insulation components. The masonry construction of these houses was based on concrete-backed stone masonry made of limestone bricks joined by mortar forming the outer walls of the building envelope. Based on the periodic seasonal weather conditions, within the Libyan cost region during summer and winter, measured thermal conductivity values were used to implement such seasonal variation of heat flow and the temperature variations through the walls. The estimation of the U-values is based on measurements and calculations. The experimental measured thermal conductivity values were obtained using the Hot Disk method known as Gustafsson’s probe. The numerical calculations were done using an analytical model that considers different wall and floor constructions which are characteristics of such houses. According to the obtained results, the U-values were quite high and therefore, several suggestions have been proposed to improve the thermal loading performance that will lead to a reasonable human comfort and reduce energy consumption. Ó 2011 Elsevier Ltd. All rights reserved. Keywords: U-values Building materials North African houses Thermal conductivity Heat conduction 1. Introduction There are two ways to calculate the U-value, namely the numerical and the simplified methods. The numerical methods are based on a detailed computer calculation that needs detailed input data and may include a non-uniform multi-dimensional heat flow. In general, they involve detailed variations in heat flow in all three dimensions, however, it is almost always the case that the construction is uniform in one direction, in other words, the 3-dimensional effects do not significantly affect the overall U-value. There are several numerical techniques [1e5] that can be employed, such as the finite element analysis, finite differences, and boundary integral methods, etc. Software packages based on such techniques are available for numerical analysis provided that they comply with the standards [2]. It must be noted that the numerical methods provide the most precise results, and they are an acceptable choice. They have the option to be used to calculate the U-value of the whole element or to calculate the thermal resistance of single different layers within walls, roofs and/or floors [3e8]. They are also the best choice for detailed pre-construction and design purposes. On the other hand, in the simplified method, the calculation procedures are much easier to perform. It is very suitable for E-mail address: [email protected]. 1359-4311/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.applthermaleng.2011.02.038 thermal load calculations of already constructed buildings where much of the fine details may not be readily available. According to standards, it involves one-dimension calculations of the upper and the lower limits of thermal resistance of the element. Then, the U-value of the components (the arithmetic mean of these two limits) can be used to estimate the expected U-value of the element. The expected result should lie within these limits, provided the difference between the limits is not too large. For example, for a building element which consists of plane, parallel, uniform layers where the heat flows in one dimension (straight lines) through these layers, the U-value is the inverse simple sum of the thermal resistances of each layer. However, almost all practical building components have non-uniformities due to mortar joints between masonry units that may lead to the presence of layers (thermal bridges) that are not parallel, and the heat no longer travels in straight lines. These joints affect the total heat transfer through the element and need to be included in the calculation of the U-value, when using such simplified methods. In this work, a simple U-value calculations formula (see discussion section) is used to estimate the U-values for a typical North African house with building envelope (outer walls) that are based on concrete-backed stone masonry made of limestone bricks joined by mortar, see the picture depicted in Fig. 1. These bricks are excavated from two major places (mines) in the north-eastern province of Libya. One mine produces soft limestone bricks and the other produces hard limestone type as shown in Fig. 2. These 1924 B.M. Suleiman / Applied Thermal Engineering 31 (2011) 1923e1928 Fig. 1. A group of buildings with exposed and non-exposed limestone external walls showing the mortar joints network and the plaster layers arrangement. houses were built during the seventies (1970’s) with the lack of any technical data on the construction materials such as the thermal properties of the individual components of the wall structure. Thus they were built without performing thermal transmittance (load) calculations or considering any insulation measures. This made them very unpleasant to occupants during both hot and cold seasons. Therefore, technical data such as U-values within the structure of such houses are essential for the calculations of thermal loads and/or estimating energy consumptions for the purpose of energy savings. Furthermore, such calculations could provide basis for insulation materials manufactures to produce suitable insulation materials that can be used to improve the thermal performance of such houses. 2. Location and climate Libya is located in north Africa facing the Mediterranean Sea and shares border with Egypt in the east, Tunisia and Algeria in the west Niger and Chad in the south, and Sudan in the south east. Libya covers an area of some 1,759,540 km2 and extends between latitudes 22 and 33 N and longitudes 10 and 25 E. It basically has two types of climate, desert in the south and Mediterranean in north. In the north and along the coastal strip, in extreme cases during the summer, temperatures can reach as high as 45  C and the relative humidity goes up to 65%. In winter, due to rain, the relative humidity can exceed seventy five present and temperature can go below the freezing point. Along this coastal strip lie the most populated cities namely Benghazi in the east and Tripoli in the west. These cities represent the residential sector of this relatively hot and humid region. The use of air conditioners during the summer period causes a high thermal load and high Fig. 3. A typical temperature limits and means relative humidity changes during one year within Benghazi provinces. The squares and dots temperatures variations and the triangles represent the average humidity variations. electricity consumption which adversely has indirect impact on global (CO2) environment. Therefore, the issue of energy saving in designing new residential buildings and renovation of old buildings in such locations is vital. 2.1. Climate of Benghazi Benghazi is the 2nd largest city in the northern east coast of Libya. It includes the part of the coastal province located in the north of Libya at Latitude: þ32.12 (32 07’12"N) and extends between longitudes: þ20.07 to (20 04’12"E). In Benghazi, the climate can be divided into two periods. November to March is the cold period and temperature during DeceJan can reach 10  C at night with high relative humidity that can reach 75% due to rain. June to September is the hot period, JulyeAugust are the months of relatively harsh condition of heat. Both Temperature and humidity are relatively high. Fig. 3 shows the typical daily temperature limits and mean relative humidity changes during one year within Benghazi provinces [9]. At the present time, and during the summer period, more than 80% of the buildings are using air- conditioners excessively (splitunit type) within relatively small volumes. Fig. 4, shows a typical building with relatively large number of conditioners as indicated by the arrows in the picture. This led to an increase in the consumption of electrical power to high levels that some times exceeds the available sources and cause partial blackouts within the city. Fig 2. Actual pictures of the bricks type with the mortar joints network. 1925 B.M. Suleiman / Applied Thermal Engineering 31 (2011) 1923e1928 Table 1 Characteristic properties of the building materials of the house envelope. Fig. 4. The excessive use of air conditioners (split-unit type) within small building sizes as indicated by the arrows. 3. Simplified theoretical model When considering a multi-layered construction with simplified index to represent the total heat-transfer processes, the U-value or total transmittance is given by: UT ¼ 1 1 ¼ P xi N RT i¼1 (1) li Where, N is the total number of layers, with xi referring to the thickness and li the thermal conductivity of each layer. The overall heat-transfer q through an area A and within a temperature difference ∆T, then can be estimated by: q ¼ UT ADT (2) This formula assumes steady state heat conduction and ignores the dynamic aspects of material behavior and spatial differences. In order to include the dynamic aspects of material behavior with spatial differences, the variations of thermal conductivity with temperature should be included [8]. However, as first approximation, Eq. (2) has been considered as a useful tool and is often used to estimate the heat transfer. In this work, the thermal conductivity values of the investigated building materials were measured in the laboratory at fixed temperature value of 25  C, therefore, our analysis was limited to using Eq.(1) to estimate UT values. It should be mentioned that a more detailed calculation should involve the effect of energy savings through the heating, cooling and ventilation system (HCVS). These buildings have no HCVS and the full detailed information (fine details) was not available, in fact only recently they have been equipped with the split-unit air condition type. Due to the present legislative technical difficulties, the thermal conductivity measured values were the only readily source of data that can be used. Therefore it was not possible to do such detailed calculations. In relation to the handbook of air conditioning [10], the cooling load (QCooling) of the building should be possible to estimate using such split-unit conditioners. Hatamipour et al. [11], estimated the total cooling load by considering the effect of various parameters according to the following equation: Material Thickness (m) Thermal conductivity (W/m-K) Density (kg/m3) Plaster Tiles Limestone Hard Limestone Soft Ground Soil 0.020 0.025 0.200 0.200 0.300 0.658 1.710 1.533 0.751 0.455 1708 2162 1900 1444 390 to heat generated by appliances, persons, etc. and QAirflow is the cooling load due to airflow into the building (sensible and latent heat). For each of above terms some well-known equations are developed and can be used with care [11]. For example, wall temperature could be calculated using mathematical modeling of the unsteady state heat transfer in the wall. This will depend on radiation intensity, variations of external air temperature (see Fig. 3) and accumulation of heat inside the wall. Unfortunately, in these buildings, during winter, daytime outside temperatures are warmer than inside temperatures and hence the net flux into the room continue throughout the walls until midday. During night time the temperature outside drops and as the room stays relatively warm the net flux is out of the room, such fluctuations of the heat flux complicates the problem even further. As it was mentioned, such estimations are not possible and we are planning for possible future work to incorporate some of the fine details in the thermal load calculations including the effect of intensity of the solar radiation through the windows. This can be done by incorporating a network of temperature and humidity sensors within and around the walls to collect data and build a database of information throughout the year. 4. Discussion and results Five different materials used in the construction of a typical Libyan house were used to estimate the U-values, including two types of limestone bricks, plaster, ground soft soil, and marble tiles. The Hot Disk technique [12e14] was used to measure the thermal conductivity of these materials at an average room temperature of 25  C. Depending on the time of the day and the seasonal variations, in the eastern costal province, the average outer walls surface temperature can vary from 40 to 10  C. The thermal conductivity of these samples was measured at room temperature (around 25  C), this value is a good approximation to the average temperature of the local climate average temperature. Looking at the temperature variations in the northern part of the country, in the hot summer days the temperature will not exceed 40 and in cold winter nights it will not go below 10 . Thus measuring the thermal properties of the materials at room temperature gives a good approximated QCooling ¼ QTransmission þ QSolar þ QInternal þ QAirflow where, QTransmission the cooling load due to heat transfer from exterior walls and envelopes, QSolar the cooling load due to sun radiation transmitted from windows, QInternal the cooling load due Fig. 5. A schematic drawing of the bricks type with the mortar joints network. 1926 B.M. Suleiman / Applied Thermal Engineering 31 (2011) 1923e1928 Fig. 6. Typical tile and plaster samples. average value of the thermal conductivity, and hence all measurements were preformed at an average temperature of 25  C. These measurements were done on isotropic basis with respect to microvoids distributions in the sample matrix. The average apparent densities of the samples ranging from 390 to 2162 kg/m3 were calculated using the dimensions (volume) and masses of the individual samples. The characteristic properties of these materials are shown in Table 1. These materials are the building blocks of three elements of the outer house envelope, namely the outer walls and the flooring. The outer walls are of traditional masonry type based on concretebacked stone masonry made of limestone bricks joined by mortar and covered from both sides by common (sand and cement) plaster layers of 20 mm each, see Fig. 5. The limestone bricks of outer walls are either soft or hard bricks. Fig. 2a and b, respectively, show actual Mortar fraction ¼ 1 estimation of the annual energy consumption for a split-unit airconditioning system is shown in the last column of the table, to be discussed. We can see that the calculated U-values are quite high, especially for the hard brick type. These values are without including irregular thermal bridges such as concrete beams and/or regular (frequent) thermal bridges due to the mortar joints between the bricks. We may include the mortar joints in the calculation for both inner and outer sides of walls simply by treating the bricks layer as a bridged layer. The joints, to some extent, may reduce the U-value if the difference in thermal resistance between bridging material and the bridged material is exceeding 0.1 m2 k/W. In our case, the differences are in the range of 0.1e0.25 m2 k/W. The mortar fraction between the bricks can be estimated using the ratio of the surface area of the brick with and without the mortar joints as follows: brick length  brick hight ðbrick length þ joint thicknessÞ  ðbrick height þ joint thicknessÞ  pictures of soft and hard bricks external walls without the plaster layer. The ground floors are made of a common ground soft soil enclosed with cement and sand mortar of the plaster type and covered with tiles,(see Fig. 6). The approximate dimensions of these samples were 10 by 10 cm square with various thicknesses. The choice of these five samples was based on the fact that construction of the outer walls and flooring of the ground floor envelope of such houses were made of these materials. The construction materials of the roofs were not included due to the following two reasons: First, most of these buildings are multi-stored and the effect is limited to the highest floor, which has different outer wall bricks (manufactured common cement type). Second, the large variations of roof design in terms of their thickness, methods of constructions mortar types, iron (steel) networking and types of concrete used. Table 2 shows the U-values of two types of external walls that were made of soft and hard limestone bricks covered on both sides by 2 cm thick plaster layers. Table 2 also shows the U-value of the ground floor slab. It is a combination of three layers of soil, mortar and tiles with an average thickness of 35 cm. Based on RT values an Table 2 The U-values of the multi-layered construction elements of the building envelope and the corresponding estimated annual energy consumption E. Construction element Effective thickness(m) RT (m2 k/W) UT (W/m2 k) E (kWh/m2) External wall (soft bricks) External wall (hard bricks) Slab on ground floor 0.240 0.240 0.345 0.33 0.19 0.70 3.03 5.26 1.46 40.26 69.93 18.98  Using the dimensions of the limestone bricks in meters, as 0.40, 0.20 and 0.02 for length, height and joint thickness, respectively, the calculations lead to a mortar fraction of around 13% of the total surface area. Furthermore, the possibility of including the surface resistances of naturally presented thin insulating layers of air at the inner and outer surfaces of the external walls, which are in the range of 0.03e.013 m2 K/W, may further reduce the estimated U-values. Unfortunately, including these effects, namely mortar bridging and air insulations will not help very much and the U-values are still high and far from comfortable. The annual energy consumption (E) given by the last column in Table 2 was calculated using the following equation from ref. [7]: E ¼   0:024  D kWh=m2 R  C Where, D is the degree-days of the location (D ¼ 1476); R is the thermal resistance of the construction element, and (C ¼ 2.65) is the coefficient of thermal performance (COP) of a commonly used LG wall split-unit air-conditioning system. The value of C was taken from the technical specification catalog for LG wall split-unit 2.5 ton capacity/heating & cooling mode (model #S306GH). It is worth mentioning that the efficiency of hot and cold air-conditioning systems is measured by a coefficient of performance (COP) - the amount of heat they generate compared to the amount of electricity needed to run them. A good quality air-conditioning system can have a COP of around 3 to 1.(which means you can transfer 3 kW of heat for 1 kW of electricity). The D value was estimated over a period of five years using the mean value of the heating and cooling degree-days of the location B.M. Suleiman / Applied Thermal Engineering 31 (2011) 1923e1928 HLLB (airport): Benina, Benghazi-Libya (20.27E, 32.10N) [15]. According to ref. [7],the E values can be used to calculate the total discount cost or the net present value per square meter of the construction elements. Unfortunately, this requires parameters such as the present worth factor and energy discount/inflation rates that are not available for this type of construction elements. It should be noted that UT calculations were made at dry conditions, not including the moisture effect on the thermal conductivities of the individual materials and/or the overall thermal conduction of the walls. The source of this moisture may be caused by wind-driven rains, foggy weather conditions, or relatively high outside temperature leading to high humidity. Moisture effect on thermal conductivity had been estimated in a previous work [13]. The moisture has considerable effect on the conductivity of the materials, owing to the fact that liquid water has thermal conductivity which 25 times greater than that of air. In addition to that, it is the complexity related to moisture content that is associated with several factors such as mode of heat transfer (sensible and/or latent), the moisture location, arrangement of moisture in the building blocks, etc. Such factors result in considerable effect on the measured heat conduction. Because of the complexity, there are no tables which can state that for a given percentage of moisture in the material there will be a certain percentage increase in its conductivity. The only certain statement which can be made is that moisture in the liquid phase will increase the conductivity. It is well known that the insulating value of most materials is greatly reduced by the presence of free water. Furthermore, it is generally agreed that the durability of masonry depends primarily on its resistance to the penetration of moisture into the body of masonry [16]. Therefore, we conclude that reliable insulation and/or vapor barriers are required for such buildings. In this work, it is not possible to do quantitative analysis regarding the benefits of insulation/vapor barriers, however, there are two factors that could illustrate such benefits: 1 The estimated values of the annual energy consumption depicted in Table 2 are clear indication of the effective role of insulation barriers in reducing energy consumption. 2 During hot season, vapor molecules in the form of moisture will flow into the building from the outside. Vapor barriers provide protection from external wind-driven moisture or bulk water penetration through capillary water breaks and/or micro-cracks in the wall structure. If buildings are not well ventilated during summer the indoor air quality could cause health hazard e mold conditions known as SBS, (Sick Building Syndrome). Thus, beside reduction in energy use, the benefits of improving indoor air quality are invaluable. In general, adding insulation and vapor barriers to the walls will damp temperature fluctuations and prevent vapor penetration inside the house envelope and provide better comfort [17]. As it was stated in section 3, more detailed work of thermal load calculation that include other parameters and/or factors such as the thickness of insulation as well as the surface to volume ratio [18], the location and extreme weather conditions will be presented in the near future. However, according to the calculated U-values, the external walls of these houses are in need for proper assessment and building codes to select a suitable insulation scheme to improve their performance in terms of energy conservation and human comfort. As far as my knowledge, for the time being, there are no specific energy conservation building codes to promote energy efficient buildings in Libya; however, there are some universal rules to implement energy conservation and environmental issues. This work is a very small contribution to the basis of 1927 such codes that require extensive building simulations followed by detailed measurements in model building. 5. Conclusions Five different materials used in the construction of a typical Libyan house were used to estimate the U-values for the external walls and the ground floor. The external walls are of traditional masonry type based on concrete-backed stone masonry made of limestone bricks joined by mortar and covered from both sides by common (sand and cement) plaster layers of 20 mm each. The ground floor is made of a common ground soft soil enclosed with cement and sand mortar of the plaster type and covered with tiles. Including the surface resistances of naturally presented thin insulating layers of air at the inner and outer surfaces of the external walls will not reduce the U-values more than few percentage. Furthermore, including the mortar bridging through the bricks have a little effect due to the fact that the mortar fraction surface area is less than 13% of the total surface area of the individual brick. It is anticipated that the situation is more severe if moisture reaches the walls and the insulating value is greatly reduced by the presence of free water. The calculated U-values indicate that the external walls of these houses which are built using two types of bricks are in need for proper assessment to select a suitable insulation scheme to improve their performance and to make the houses pleasant and comfortable. Thus, this work might provide the basis for insulation materials manufactures to produce suitable and reliable insulation materials that can be used to improve the thermal performance of such houses. It should be noted that measurements made by laboratory test may not be truly indicative of the properties of the materials in service and aging effects on dimensional stability (ability to retain size and shape) should be anticipated. Finally, a more detailed work of thermal load calculation which includes other factors such as the location and moisture effect and extreme weather conditions should be investigated. It is recommended to do more detailed measurements in model building and extensive building simulations in order to develop reliable building codes in Libya. Acknowledgements The financial support from the University of Sharjah for the project is gratefully acknowledged. This work is associated with research project No. 021303. References [1] R. Butler, The computation of heat flows through multi-layer slabs, Build. Environ. 19 (3) (1984) 197e206. [2] R. Yumrutas, M. Unsa, M. Kanog¢, Periodic solution of transient heat flow throw through multilayer walls and flat roofs by complex finite Fourier transform technique, Build. Environ. 40 (3) (2005) 1117e1126. [3] M.F. Zedan, A.M. Mujahid, An efficient solution for heat transfer in composite walls with periodic ambient temperature and solar radiation, Int. J. Appl. Energy 14 (2) (1993) 83e97. [4] C. Lombard, E.H. Mathews, A two-port envelope model for building heat transfer, Build. Environ. 34 (1999) 19. [5] J.J. del Coz Diaz, et al., Non-linear thermal analysis of light concrete hollow brick walls by the finite element method and experimental validation, Appl. Therm. Eng. 26 (2006) 777e786. [6] Jose-Luis Vivanocos, et al., A new model based on experimental results for thermal charactarization of bricks, Build. Environ. 44 (2009) 1047e11052. [7] L.M. Al-Hadhrami, A. Ahmad, Assessment of thermal performance of different types of masonry bricks used in Saudi Arabia, Appl. Therm. Eng. 29 (2009) 1123e1130. [8] H. Asan, Y.A. Sancaktar, Effects of wall’s thermophysical properties on time lag and decrement factor, Energ. Build. 28 (1998) 159e166. [9] BBC (2009).http://www.bbc.co.uk/weather/bbcweather/weather_forecast_media. shtml. [10] Carrier Air Conditioning Company, Handbook of Air Conditioning System Design. McGraw Hill, New York, 1965. 1928 B.M. Suleiman / Applied Thermal Engineering 31 (2011) 1923e1928 [11] M.S. Hatamipour, H. Mahiyar, M. Taheri, Evaluation of existing cooling systems for reducing cooling power consumption, Energ. Build. 39 (2007) 105e112. [12] S.E. Gustafsson, Transient plane source techniques for thermal conductivity and thermal diffusivity measurements of solid materials, Rev. Sci. Instrum. 62 (1991) 797. [13] B.M. Suleiman, Moisture effect on thermal conductivity of some major elements of a typical Libyan house envelope, J. Phys. D: Appl. Phys. 39 (2006) 547e551. [14] Suleiman Bashir M. (1993). The transient plane source technique for measurement of thermal properties of polycrystalline ceramics including high-Tc superconductors, Ph.D. Thesis, ISBN 91-7032-905-2, Physics Department, Chalmers University of Technology, Gothenburg, Sweden. [15] A free worldwide data calculation site on Heating & Cooling Degree Days.http://www.degreedays.net [accessed 03.01.11.]. [16] G.H. Galbraith, Jintang Li, G. Jiansong, David Kelly, R. McLean, Grunewald J. Craig, Evaluation of discretized transport properties for numerical modeling of heat and moisture transfer in building structures, Journal of Thermal Envelope and Building Science 24 (3) (2001) 240. [17] M. Ciampi, F. Fantozzi, F. Leccese, G. Tuoni, On the optimization of building envelope thermal performance, Civ. Eng. Environ. Syst. 4 (2003) 231e254. [18] N.K. Bansal, A. Bhattacharya, Parametric equations for energy and load estimations for buildings in India, Appl. Therm. Eng. 29 (2009) 3710e3715.