Combined use of design of experiment and dynamic building simulation in assessment of energy efficiency in tropical residential buildings

Combined use of design of experiment and dynamic building simulation in assessment of energy efficiency in tropical residential buildings

Energy and Buildings 86 (2015) 525–533 Contents lists available at ScienceDirect Energy and Buildings journal homepage:

2MB Sizes 0 Downloads 12 Views

Energy and Buildings 86 (2015) 525–533

Contents lists available at ScienceDirect

Energy and Buildings journal homepage:

Combined use of design of experiment and dynamic building simulation in assessment of energy efficiency in tropical residential buildings Aidin Nobahar Sadeghifam a,∗ , Seyed Mojib Zahraee b , Mahdi Moharrami Meynagh a , Iman Kiani a a b

Faculty of Civil Engineering, Department of Construction Management, Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia Faculty of Mechanical Engineering, Department of Industrial Engineering, Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia

a r t i c l e

i n f o

Article history: Received 6 December 2013 Received in revised form 17 August 2014 Accepted 22 October 2014 Available online 30 October 2014 Keywords: Dynamic building simulation Energy consumption Residential buildings Design of Experiment

a b s t r a c t Energy consumption has become an increasingly controversial issue in the modern world. Among the widest range of energy consumers, residential buildings consume the largest amount of energy most of which is consumed by air conditioning systems in tropical countries. This paper attempted to examine energy saving in building elements such as walls, floors, windows, roofs, and ceilings and how the integration of such optimized elements in conjunction with effective air quality factor can contribute towards an ultimate energy efficient design. A typical two-storey terraced house in Kuala Lumpur, Malaysia was chosen to model energy usage by means of dynamic building Simulation. A case study was modeled using Revit Architecture software and analyzed using energy analysis software. Current energy consumption patterns were identified and the optimal level of energy usage was determined by replacing components with new energy efficient materials. Afterward, a Design of Experiment (DOE) method was used and the best combination of factor was identified. The results indicated that in residential buildings in tropical regions, changing ceilings and ceiling materials are the most effective way to reduce energy consumption; moreover, wall materials and inside temperatures were in the next levels of significant factors respectively. These results can be used to help building designers achieve optimum cooling load savings. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Nowadays energy is a controversial issue in most countries of the world. This is mainly due to a probable future energy shortage as well as global warming. From December 1997 onward, when the Kyoto protocol was signed, most governments have attempted to reduce the release of greenhouse gases. One immediate solution to the problem of greenhouse is the efficient use of energy [1]. Malaysia as a developing country has experienced an increase in energy consumption. As indicated by statistics, the amount of energy consumed in Malaysia has roughly doubled from 2000 to 2010. Research carried out in 2006 on household energy usage by the Center for Environment, Technology and Development, Malaysia (CETDEM) (Fig. 1), pointed out that air conditioning and refrigerators consumed nearly 70% of the electricity in residential

∗ Corresponding author. E-mail address: aidin [email protected] (A.N. Sadeghifam). 0378-7788/© 2014 Elsevier B.V. All rights reserved.

buildings. Global warming is a critical issue that leads to higher building energy consumption for cooling, causing an increase in its significance [2]. Energy is considered one of the most crucial factors in economic growth and continuous development [3]. Driven by the energy crisis and demand for sustainable development, engineers attempt to provide energy saving and sustainable strategies for using in buildings. Sustainable design is the philosophy of designing built environments and physical objects that conform, to the doctrines of ecological, social, and economic sustainability. Sustainable design aims to completely remove negative environmental impacts through sensitive and skillful design. Sustainability issues are usually considered in connection with other issues such as energy consumption or choice of material [4]. Statistically, 30% of total energy of Malaysia is consumed by residential buildings. In Malaysia’s residential buildings, the thermal loss of building components is as follows: windows 50%, walls 35%, ceilings 7.5%, and floors 7.5% [5]. Sustainably designed buildings have reduced levels of energy demand resulting from heating and


A.N. Sadeghifam et al. / Energy and Buildings 86 (2015) 525–533

Fig. 1. Average home electricity consumption.

cooling needs. To develop sustainable design and efficient energy analysis and optimization, many novel methods have been used [6]. Some effort has gone into developing modern supplementary methods to efficiently provide thermal comfort in residential buildings. Reference [7] investigated the effect of colored glass, double glazed windows, light colored roofs and walls, and the use of reflective roofs on thermal comfort. Building information modeling solutions make sustainable design practices easier by enabling architects and engineers to more accurately visualize, simulate, and analyze building performance earlier in the design process [8]. Computer simulation is an useful method to analyze the different systems such as manufacturing system, construction process and energy analysis [9]. Software such as Energy Plus, Transys and Ecotect are common for facilitating energy analysis and building simulation [10,11]. Energy Plus is an extensive and complete simulation environment for the transient simulation of systems, including multi-zone buildings. Furthermore, some statistical software is commonly used to analyze variations in order to find the most appropriate combination of building elements [12]. In the 1920s, Fisher proposed a method known as the Design of Experiment (DOE) as a statistical approach for analyzing the influence of water and rain on crop production. He carried out a set of experiments using orthogonal arrays in order to restrict the number of experiments [13]. DOE can help to properly determine a combination of resources that can maximize productivity levels. In fact, DOE is known as an experiment or series of experiments that are done by changing the input process variables that may affect output responses. This technique also helps planners to find variables with the most influence on a response. Experimental design methods are considered to be practical tools that improve processes. In addition, DOE can provide an insight into the interactions between factors that can influence responses or output [14]. The designed experiments have been used to improve and comprehend a system. In this paper, a unique approach that involves both simulation and DOE was used to determine the best combination of factors to maximize process productivity for a manufacturing system process [15]. 2. Critical review of previous studies In an earlier study [16], the energy usage of a five-storey office building in hot and humid climate (Saudi Arabia) was assessed using Visual DOE (version 4.0) software. The results indicated that increasing the thickness of insulation did not influence energy efficiency. Instead, appropriate set point temperature, low-e double glazing, using VAV system, adjustable lighting options, and energy efficient lamps combined to have a significant influence on energy efficiency. The combination of these strategies led to a 36% reduction in annual energy consumption. Reference [17] determined the

climatic parameters that should be considered for designing energy efficient school buildings in a humid and hot climate. They asserted that the different strategies could interact. They demonstrated that factors such as shading, night ventilation, lighting control, infiltration, and windows size play a large role in the energy efficiency of school buildings. Reference [18] carried out a study on passive climate control in residential buildings in Singapore’s tropical climate that were ventilated naturally. Using thermal analysis software (TAS), they investigated the effects of microclimatic criteria, such as orientation, wind, and shading from surrounding buildings on ways of minimizing heat, such as using roof thermal buffers, window shades, and optimizing the use of building materials. The results showed that thermal buffers were the most efficient method for saving energy. The energy performance in buildings is influenced by many factors, such as ambient weather conditions, building structure, and characteristics, the operation of sub-level components, such as lighting and HVAC systems, and the behavior of the occupants [19]. The effect of various types of roof, wall, and flooring materials on the energy consumption of an ordinary low-density bungalow house in Malaysia was the focus of one study. Energy analysis and simulation were carried out using Energy Plus and Design Builder software. The researchers replaced heavy-weight walls with lightweight ones, which resulted in a 16% energy savings. Replacing concrete roof tiles with a white painted steel roof resulted in energy savings of 5.8%. Using appropriate floor materials decreased energy consumption by 9.4% in this study [20]. In this paper, monthly cooling load was used as the basis of comparison between types of building elements and air quality specifications. The load was used to indicate the level of cooling needed to provide thermal comfort in the building. Energy consumption levels are dependent on the efficiency of the devices used for cooling the spaces and this may mean that for the same load, different devices with different efficiencies will result in different levels of energy consumption. As a result, using the cooling load as the basis of comparison provided a more realistic view in terms of the effect of ceilings, windows, walls, and temperature as effectual measure of a building’s level of energy consumption. Moreover, this study attempted to evaluate energy use in residential buildings by evaluating the effect of different types of elements and air quality factors. The amount of saving generated by cooling loads throughout the year in the extreme tropical climate of Malaysia was examined. In addition, DOE technique was used to examine the impact of different factors on the cooling load. As a result, an optimum combination of factors was established.

3. Case study In this survey, a two-storey residential building separated into two uniform apartments was selected in Kuala Lumpur, Malaysia. The main living spaces of the case study were the living room, kitchen, four bedrooms and bathrooms, and one staircase (Figs. 2 and 3). The total area of the building was 676 m2 . The house was separated into 11 zones with separate thermal properties for each level. Structure of the building was reinforced concrete. The building had pitched roofs covered with clay tiled without any insulation in the walls and roofs. The facade of the building was covered by cement sand render and the main material of external walls was brick and plaster. Table 1 indicates the architectural properties of the building. Building orientation was East-west and it had timber framed windows with single glazing and no shading.

A.N. Sadeghifam et al. / Energy and Buildings 86 (2015) 525–533


Fig. 4. View of simulated building.

Fig. 2. Division of building in separated zones in the simulated model for level 1.

used to analyze the building. To limit the scope of this study it was assumed that the base model represented a typical residential building made with materials commonly used in Malaysia. 4.1. Simulation model development Revit Architecture software was used to simulate the building. This software is one of the most useful tools developed for dynamic building simulation software. CAD drawings were imported to Revit Architecture and specific parametric design principles were used in the simulation. Fig. 4 shows the simulated building in this software. 4.2. Model energy analysis

Fig. 3. Division of building in separated zones in the simulated model for level 2.

Regarding the location of case study building, Kuala Lumpur is situated at latitude 3.1597◦ N and longitude 101.7000◦ E, which experiences uniformly high air temperatures and relative humidity. The average mean temperature remains fairly constant. Maximum temperature hovers between 31 ◦ C and 33 ◦ C (88 ◦ F and 91 ◦ F) and never exceeds 39.3 ◦ C (102.7 ◦ F). Minimum temperature range is between 22 ◦ C and 23.5 ◦ C (72 and 74 ◦ F). The relative humidity (RH) falls between 79% and 82% and the average monthly precipitation is 202 mm [21]. 4. Methodology The methodology used in this study was based on the comparison of the physical properties of building components and various combinations of building components with air quality factors. Models simulated by building information modeling software were Table 1 Description of case study building. Parameters


Internal floor

Concrete (medium density) Cast concrete (Dense) (10 cm) + tile (1.2 cm) Cement sand render (1.3 cm) + brick (22 cm) + gypsum plastering (1.3 cm) Brick (11 cm) + inner/outer gypsum plastering (1.3 cm) Wooden batons (20 cm) + air gap (10 cm) + clay tiles (3 cm) Acoustic tile suspended (10 mm) Timber framed window, single glazing (6 mm) 0.5 ac/h Fluorescent, compact (4.6 w/m2-100 lux) 56 m2 /person

External walls Internal walls Pitched roof Ceiling Window Infiltration rate Lighting Occupancy

In order to analyze the energy consumption of the building used as a case study, a model was created in Revit Architecture. The specifications were re-assigned to Ecotect and a final sketch-up was imported to Energy Plus software. To establish a baseline for the energy consumed by a typical house, the characteristics of materials used in the initial model were defined in the software shown in Table 1. In order to achieve the most accurate analysis, variables such as the type of building, the orientation of the building and climatic data for the location of the building were essential. The simulation of weather data was based on the weather report of Kuala Lumpur, Subang weather station which is available at Energy Plus website [22]. In the model designed for this study, the building was divided into 11 different zones (for each level), each with its own particular specifications in terms of activity, HAVC systems and comfort temperature. These data are the same for both levels (1, 2), which is indicated in Table 2 for level 1. According to the comfort levels in Table 2, in a hot and humid climate, a cool set point temperature is important and can be identified in various spaces. This temperature can be defined by what the residents are habituated to as well as other adaptive standards. Adaptive temperatures (neutral temperature) have been proposed for different spaces in residential buildings by Peeters et al. [23]. Operation schedule of AC and natural ventilation of main spaces could be seen in Table 3. The output of the energy analysis software revealed the monthly cooling load for the base model. To find the most significant building elements, different floor, wall, windows, ceiling and roof materials were tested (Table 4) and the results were compared to find the optimal components. Adjustments to air quality factors uncovered the variables that affected thermal comfort. The outcomes of this step revealed the three most effective energy saving building components and these were combined to determine the most significant component.


A.N. Sadeghifam et al. / Energy and Buildings 86 (2015) 525–533

Table 2 Occupant profiles for different zones (level 1). Zone level 1

Area (m2 )

Volume (m3 )



HAVC system

Comfort band

Living room 1 Kitchen 1 Bedroom 1 Bedroom 2 Bedroom 3 Bedroom 4 WC 1 WC 2 WC 3 WC 4 Stair 1

155 55 45 25 13 10 3.5 3.5 3.5 4.5 20

465 165 135 75 39 30 10.5 10.5 10.5 13.5 60

6 1 2 2 1 1 1 1 1 1 1

Sedentary Cooking Sedentary Sedentary Sedentary Sedentary Sedentary Sedentary Sedentary Sedentary Sedentary

AC AC AC AC AC AC Natural ventilation Natural ventilation Natural ventilation Natural ventilation Natural ventilation

26 26 25 25 25 25 28 28 28 28 28

Table 3 Assumed operation schedule of AC and natural ventilation of main spaces. Abbreviation

Operation type

Operation time of weekends and holidays

Operation time of week days

Bedrooms Living Room Kitchen WC & Stair

AC AC AC Natural ventilation

8 p.m–9 a.m 9 a.m–5 p.m, 8 p.m–12 a.m 9 a.m–2 p.m, 8 p.m–10 p.m 7 p.m–7 a.m

5 p.m–7 a.m 9 a.m–12 a.m 7 a.m–2p.m, 5 p.m–8p.m 7 p.m–7 a.m

Table 4 Types of different tested material for building components. Types

Elements Wall





Type 1

Reverse Brick Veneer R20

Double Glazed Timber Frame

Clay Tiled Roof

Type 2

Reverse Brick Veneer R15

Single Glazed Alum Frame

Plaster insulation suspended Acoustic tile suspended

Type 3

Double Glazed AlumFrame

Type 4

Brick Cavity Conc Block Plaster Brick plaster

Clay Tiled Roof Ref Foil Gyproc Corrugated Metal Roof

Conc Flr Tiles Suspended Conc Flr Timber Suspended ConcSlab Tiles OnGround

DoubleGlazed LowE AlumFrame

Metal Deck

Type 5 Type 6

Brick Conc Block Plaster ConcBlock Plaster

DoubleGlazedLowE TimberFrame SingleGlazed AlumFrame Blinds

Metal Deck Insulated Plaster Foil Heat Retention Ceramic Tile

Type 7 Type 8 Type 9 Type 10

ConcBlock Render Double Brick Cavity Plaster Double Brick Cavity Render Double Brick Solid Plaster

Single Glazed Timber Frame Translucent Skylight

5. Statistical analysis

Plaster Joists Suspended

ConcSlab Timber OnGround Suspended concrete ceiling TimberFlr Suspended

building components to be used in comparing with the building components when optimised.

5.1. Design of Experiment (DOE) 6.1. Effective factors analysis A DOE technique was employed to develop a plan for an experiment that would specify the components that would lead to best process productivity. To implement a DOE technique, the following steps were followed [24]: • • • • • •

Choosing the factors and their levels, choosing a response variable, choice of experimental design, performing experiment, data analysis, promoting the best option.

6. Results and discussion A model was created by Revit Architecture using the base materials and then imported to energy analysis software (Energy Plus) in order to evaluate and determine the best option in terms of energy savings. The results of the building model assessment indicated that the annual cooling load was 214.202 kW h/m2 . This amount of energy is the result of simulating existing materials (base) in

6.1.1. Different components evaluation The five main building materials that influenced cooling load savings had high and low impact levels for different types of floors, walls, windows, ceilings and roofs. Table 5 shows the different types of materials tested for each component and the maximum and minimum effect on energy saving. The lowest cooling load indicated the highest energy savings for each of the different types of building elements. Fig. 5 shows the difference between the lowest cooling load generated by the various materials tested for each element and the base model. Result of the comparison between the base model and the model generated with different materials illustrated that in this case, the walls, windows and ceilings had major effects on energy savings. The best and worst types of elements will be discussed in the next step of the analysis. 6.1.2. Air quality factors evaluation Measuring the occupant’s thermal comfort revealed that temperature had the highest influence on the quality of indoor air compared to air movement and humidity. Having the highest

A.N. Sadeghifam et al. / Energy and Buildings 86 (2015) 525–533


Table 5 Subjected building elements to analysis. Element Floor Wall Windows Ceiling Roof

Number of material tested 5 10 8 4 6

Lowest cooling load kW h/M2

Highest cooling load kW h/M2

204.325 172.193 183.766 124.555 214.202

214.202 204.696 207.599 214.202 215.320

Fig. 5. Compare of cooling load between existing material (Base) and optimised elements of the building.

Fig. 6. Amount of cooling load for different temperatures.

impact on energy usage, temperature was chosen in the DOE analysis in the next step. Fig. 6 demonstrates the influence of air temperature on the amount of cooling load needed to maintain the occupant’s thermal comfort.

The factors chosen in this study are the important factors that play a leading role in energy usage. The variation for each factor is indicated in Table 6 where each factor has a high (+) and low (−) level.

Each level was based on discussions with building contractors and process limitations. In the above Table, −1 is the maximum cooling load and the lowest level of energy saving. On the other hand, 1 introduces the item type which results in the lower level of cooling load and higher saving levels. Table 7 sets out the properties of each element for −1 and 1 level. As discussed before, the response variable considered in the DOE for this study was the cooling load. A full factorial design was used as a design for the experiment because only a small number of factors were investigated. In factorial design, all possible combinations are considered in an experiment, which is replicated three times. As can

Table 6 Factor levels.

Table 7 Types of material for each level.

6.2. Choosing factors and response variable

Factor name

Low level (−)

A = Temperature B = Wall C = Ceiling D = Window

20 −1 −1 −1

High level (+) 26 1 1 1


Material of lowest cooling load (+1)

Material of highest cooling load (−1)

Wall Ceiling Windows

Reverse Brick Veneer R20 Plaster Insulation Suspended Double Glazed Timber Frame

Brick plaster Acoustic Tile Suspended Single Glazed Alum Frame


A.N. Sadeghifam et al. / Energy and Buildings 86 (2015) 525–533

Table 8 Result of running simulation experiment. Run order



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

20 20 26 20 20 26 26 26 26 26 20 26 20 26 26 26 26 26 26 20 20 26 20 26 20 26 20 20 20 20 26 26 26 26 20 20 20 26 20 26 26 20 20 26 20 20 20 20

−1 1 1 1 1 1 −1 −1 −1 1 1 1 1 1 1 −1 −1 −1 −1 −1 1 1 1 −1 −1 −1 −1 −1 1 −1 −1 1 −1 1 −1 −1 −1 1 −1 1 1 1 1 −1 −1 1 1 −1

Ceiling 1 1 −1 −1 1 −1 1 1 −1 1 1 1 1 −1 1 −1 −1 −1 1 1 1 1 −1 1 −1 1 −1 1 −1 −1 −1 −1 −1 −1 −1 −1 1 1 1 −1 1 −1 1 1 −1 −1 −1 1

Fig. 7. Normal probability of effects.

Window 1 −1 −1 −1 −1 1 1 −1 1 −1 1 1 1 −1 1 −1 −1 −1 1 −1 1 −1 1 1 1 −1 −1 −1 −1 1 1 1 1 1 −1 1 1 1 −1 −1 −1 1 −1 −1 −1 −1 1 1

Response 192,216 182,334 211,279 203,745 182,334 208,894 176,651 180,123 213,019 174,153 176,378 170,796 176,378 211,279 170,796 216,188 216,188 216,188 176,651 198,302 176,378 174,153 198,706 176,651 213,019 180,123 218,809 198,302 203,745 212,981 213,019 208,894 213,019 208,894 218,809 212,981 192,216 170,796 198,302 208,894 174,153 198,706 182,334 180,123 218,809 203,745 198,706 192,216

A.N. Sadeghifam et al. / Energy and Buildings 86 (2015) 525–533


Fig. 8. Pareto chart.

Table 9 ANOVA table. Source


Main effects 2-Way interactions 3-Way interactions Curvature Residual error Pure error Total

4 6 4 1 31 31 46

Seq SS 9,962,321,234 730,646,009 175,607,088 13,08,166 1,173,455,779 1,173,455,779 12,043,338,277

Adj SS

Adj MS



9,893,063,355 730,077,461 175,472,831 1,308,166 1,173,455,779 1,173,455,779

2,473,265,839 121,679,577 43,868,208 1,308,166 37,853,412 37,853,412

65.34 3.21 1.16 0.03

0.000 0.014 0.348 0.854

Table 10 Coefficient and effects of factors. Term



SE coefficient


P value

Constant Temperature (A) Wall (B) Ceiling (C) Window (D) Temperature*Wall (AB) Temperature*Ceiling (AC) Temperature*Window (AD) Wall*Ceiling (BC) Wall*Window (BD) Ceiling*Window(CD) Temperature*Wall*Ceiling (ABC) Temperature*Wall*Window (ABD) Temperature*Ceiling*Window (ACD) Wall*Ceiling*Window (BCD) Temperature*Wall*Ceiling*Window (ABCD)

−4334 −9789 −27,173 −3373 3750 −5334 2101 3090 −237 2114 −214 2786 −1339 −2243 335

196,727 −2167 −4894 −13,587 −1687 1875 −2667 1051 1545 −119 1057 −107 1393 −670 −1121 168

901.8 901.8 901.8 901.8 901.8 901.8 901.8 901.8 901.8 901.8 901.8 901.8 901.8 901.8 901.8 901.8

218.15 −2.40 −5.43 −15.07 −1.87 2.08 −2.96 1.16 1.71 −0.13 1.17 −0.12 1.54 −0.74 −1.24 0.19

0.000 0.022 0.000 0.000 0.071 0.046 0.006 0.253 0.097 0.896 0.250 0.907 0.133 0.463 0.223 0.854

be seen in Table 7, each factor had two levels and the full factorial experiment included 48 runs. 6.3. Performing simulation experiments After designing the experiment, the models and energy analysis were completed for different combinations. Table 8 sets out the results of experiments.

variance (ANOVA) performed by Minitab to identify significant factors. Decisions about the significance of a factor or its effect were made based on its P-value. If the P-value of a factor or its effect was less than 0.05, it was considered to be significant [14]. The probability of each effect is shown in Fig. 7. The effects that lie along the line are negligible, whereas the significant effects are far from the line [14]. The effects that were deemed to be significant were those caused by A (Temperature), B (Walls), C (Ceilings) and AC, and AB combinations.

6.4. Data analysis 6.4.1. Identifying significant factors In order to analyze the data shown in Table 8, a statistical computer package was required. In this study, Minitab software was used for this purpose. Table 9 shows the results of the analysis of

6.4.2. Regression model and analysis Pareto chart was applied to illustrate the sequencing of the statistical significance of both the main and interaction effects as well as to compare the relative value. Fig. 8 shows that factor A (Temperature), B (Wall), C (Ceiling) and AB, and AC in two way interactions


A.N. Sadeghifam et al. / Energy and Buildings 86 (2015) 525–533

Fig. 9. Main effects plot.

Fig. 10. Normal probability of residuals.

Fig. 11. Residuals versus fitted values.

A.N. Sadeghifam et al. / Energy and Buildings 86 (2015) 525–533

contribute the most to the cooling load. The result is the same as the one illustrated by Normal probability plot, but Pareto chart shows the ranking of the highest to lowest effects. After estimating the influence of the significant factors, the following regression model was fitted to the data. Eq. (1) calculates the regression model by considering regression coefficient and effects that was fitted to the data generated by Minitab (Table 10). Based on the regression model and the graphs generated according to amount of effect (Fig. 9), cooling load is at its optimum level when A, B and C are placed at high level. This means that to achieve the optimum cooling load, the temperature should be equal to 26 ◦ C and the walls and ceiling should use Reverse Brick Veneer R20 and Suspended Plaster Insulation respectively. Y = B0 +


BX i=1 i i



B XX i=1 ij i j

Y = 196727 + (−2167)xA (−4894)xB + (−13587)xC +(1875)xA |xB + (−2667)xA xC


Y = 196727 + (−2167)(+1) + (−4894)(+1) + (−13587)(+1) +(1875)(+1)(+1) + (−2667)(+1)(+1) Y = 175341 Residual analysis was used to validate the regression model. The residual, which indicates the difference between the predicted values and the observed values, should be located on a straight line in a normal probability plot [14]. Bases on Fig. 10, the residual followed a straight line so it confirmed the validation of model. In addition, Fig. 11 contains a graph of residual versus the predicted response for the model. According to Fig. 11, there is a less patterned structure indicating that the proposed model was adequate. 7. Conclusion Generally, building elements behave differently in terms of energy performance depending on the location, climatic conditions and physical properties of the building. In this study, a model for evaluating the performance of the floors, walls, windows, ceilings, roofs and air quality (temperature, humidity and air flow) in terms of their ability to reduce cooling loads in residential tropical buildings was introduced. The results of the simulation and energy analysis of the case study building revealed that the components of the building that have a major effect on energy saving are walls, windows and ceilings. In addition, temperature played an important role in the energy analysis process. Four factors each with two levels were selected to evaluate different alternatives to determine which would have the most influence on the energy savings. In this paper, a design of experiment technique (DOE), which is a statistical method, was applied in order to analyze the different scenarios. According to the result, ceilings, walls and temperature were the most influential factors. Interactions between temperature and walls, and temperature and ceilings had the most significant impact on energy usage. In other words, the result of this approach suggests that in these types of buildings in tropical areas, the ceiling material has the greatest effect on energy usage and saving. Wall materials also had a significant influence on energy consumption. In addition, temperature was the most influential air quality factor. Combining the effects of temperature and various wall and ceiling materials leads designers and


building providers to optimize cooling loads by using reverse brick veneer R20 for the walls, suspended plaster insulation for the ceiling, double glazed timber frame windows and keeping the temperature at 26 ◦ C. As a developing country Malaysia has faced difficulties in managing energy loss specifically in residential buildings. Providing succinct information regarding the performance of building elements and environmental considerations is intended to stimulate awareness of energy efficient factors and encourage the integration of these factors into the design and construction of new buildings. References [1] International Energy Agency, Online Source. [2] CETDAM, Working with the Community on Energy Efficiency at Household Level in Petaling Jaya, A CETDAM Study on Energy Efficiency, 2006. [3] J. Blewitt, Understanding Sustainable Development, Earthscan, London, 2008. [4] B. Hunn, Fundamental of Building Energy Dynamics in: Massachusetts, Institute of Technology, London, England, 1996. [5] S.M. Sadegh Zadeh, An energy efficiency plan for the Iranian building subsector, Energy Policy. 35 (2) (2007) 1164–1171. [6] R. Pacheco, J. Ordo˜nez, G. Martinez, Energy efficient design of building: a review, Renewable and Sustainable Energy Reviews 16 (2012) 3559–3573. [7] D.S. Parker, S.F. Barkaszi, Roof solar reflectance and cooling energy use: field research results from Florida, Energy and Buildings 25 (1997) 105–115. [8] S.E. Lee, P. Rajagopalan, Building energy efficiency labeling program in Singapore, Energy Policy 36 (10) (2008) 3982–3992. [9] S.M. Zahraee, M. Hatami, J.M. Rohani, H. Mihanzadeh, M.R. Haghighi, Comparison of different scenarios using computer simulation to improve manufacturing system productivity: a case study, Advance Materials Research 845 (2014) 770–774. [10] U.S. Department of energy, Building Energy Software Tools Directory. directory/software.cfm/ID=58/ pagename=alpha list [11] Autodesk, Sustainable design analysis and building information modeling, Autodesk White Paper, 2009. [12] U.S. Department of Energy (US-DOE), Energyplus Energy Simulation Software. 2004. [13] R. Fisher, The Design of Experiment, 2nd ed., Rinehart and Winston, New York, 1972. [14] D. Montgomery, Basic Experiment Design for Process Improvement Statistical Quality Control, John Wiley and Sons, Inc., USA, 2009. [15] S. Ng, J. Yuan, J. Yin, S. Mu, Design of experiments for simulation models with stochastic constraints, in: IEEE Conference on Industrial Engineering and Engineering Management, 2009, pp. 2094–2098. [16] M.S. IqbalI, Al-Homoud parametric analysis of alternative energy conservation measures in an office building in hot and humid climate, Building and Environment 42 (5) (2007) 2166–2177. [17] Y.V. Perez, I.G. Capeluto, Climatic considerations in school building design in the hot-humid climate for reducing energy consumption, Applied Energy 86 (3) (2009) 340–348. [18] N.H. Wong, S. Li, A study of the effectiveness of passive climate control in naturally ventilated residential buildings in Singapore, Building and Environment 42 (3) (2007) 1395–1405. [19] Z. Hai-xiang, M. Frédéric, A review on the prediction of building energy consumption, Renewable and Sustainable Energy Reviews 16 (2012) 3586–3592. [20] M.F. Saber Sabouri, M. Zain, M. Jamil, Exploring role of different floor, wall and roof details in energy efficiency of a bungalow house in Malaysia, Scientific Research and Essays 6 (30) (2011) 6331–6345. [21] M.S. Al-Homoud, A.A. Abdou, I.M. Budaiwi, Assessment of monitored energy use and thermal comfort conditions in mosques in hot-humid climates, Energy and Buildings 41 (6) (2009) 607–614. [22] U.S. Department of Energy 2010. Weather file of Kuala Lumpur, Applied for Energy PlusProgram. energyplus/cfm/weather data3.cfm/region=5 southwest pacific wmo region 5/country=MYS/cname=Malaysia [23] L. Peeters, R.D. Dear, J. Hensen, D. Haeseleer, Thermal comfort in residential buildings: comfort values and scales for building energy simulation, Applied Energy 86 (5) (2009) 772–780. [24] S.M. Zahraee, M. Hatami, N.M. Yusof, M. Rohani, F. Ziaei, Combined use of design of experiment and computer simulation for resources level determination in concrete pouring process, Jurnal Teknologi 64 (1) (2013) 43–49.