A CASE STUDY OF ENERGY EFFICIENT WINDOW SYSTEM DESIGN FOR A RENEWAL OF A HISTORICAL BUILDING

Aslihan Tavil, Assistant Professor,

Istanbul Technical University, Faculty of Architecture, Istanbul, Turkey M. Cem Altun, Assistant Professor,

Istanbul Technical University, Faculty of Architecture, Istanbul, Turkey

1. Introduction In recent years there has been a growing awareness of the impact that buildings have on environmental degradation and in response to that interest has grown in pursuing sustainable building design. For architecture sustainable design means design that delivers buildings with lower environmental impacts while enhancing health, productivity and quality of life. In this context, windows are one of the most significant elements in the design of any building impacting the building energy use and related environmental consequences [1]. In the last 20 years, the energy efficient properties of windows have been improved by innovations in glass coatings, low conductance spacers and frames that result in multiple glazing layers. Properties of window systems in the external envelope have a significant effect on the overall thermal performance of buildings and on the thermal comfort indoors [2], [3]. Because of the special constrains, requirements and objectives related to a historical building, the design of a window system of a renewal is much more important in achieving energy efficiency.

In the paper a case study comprising the methodology and implications of the design of an energy efficient window system is introduced. A renewal project of a roof and the attic space below, as an extension of a historical 19th Century building in Istanbul has been completed recently. One of the objectives in the design of the exterior envelope is to achieve an environment for education with high level of thermal comfort and low energy consumption. It becomes necessary to design all vertical exterior surfaces as transparent components for a lighter and more open appearance because of the limitations of the existing built form. The

challenge is to design a window system for the east and west façades of the extension, fulfilling the objectives of thermal comfort and energy efficiency as well as creating positive interior environment that improves occupant’s productivity. A four-step methodology for energy efficient window system design is introduced, comprising market search, computer simulations on product and building level and evaluation stages. The thermal performance of the extension is determined with different double glazed window system alternatives. The window system with optimum performance from energy efficiency and thermal comfort point of view is presented for this special case.

2. Taskisla Building and Design of the Extension Taskisla ; is a masonry building with three floors and 130 m x 100 m in width and length, with a 70 m x 40 m courtyard and towers on each of its corners. It is located on the European part of Istanbul viewing the Bosphorus and was designed by the British architect James Smith in the 19th century in a neoclassic style. Construction of the building was started in 1847 and was finished in 1852. Although the building was originally designed as a military medicine school, its function was changed into a hospital and later into military barracks. After its renovation in 1944, the building is being used as the Faculty of Architecture of Istanbul Technical University (Figure 1).

Figure 1. Taskisla Building

With the objective to create new design studio spaces, the roof and the attic space below that had been constructed in the fifties as an extension in Taskisla, is renewed. In the scope of the renewal, the structural system and the exterior envelope of the extension are redesigned and constructed. The renewal has a simple plan layout with studio spaces on both sides of a corridor between the northeast tower and the southeast tower. The open space studio with separations, faces the east is 77 m in length and 9 m in width (Figure 2). Seven “single room design studios” are placed on the other side of the corridor, facing the west and the courtyard of the building. Steel structure, aerated concrete prefabricated roof slabs, shape of the roof, external wall plane, plan layout with the studio spaces on both sides of the corridor are the constrains in the design (Figure 3).

Figure 2 Plan of the renewal

Figure 3 General view to “the open space studio” with window system facing the east.

3. Methodology for the Design of an Energy Efficient Window System

In making decision about window selection, there are many factors leading to decision-making criteria and the basic criterion relate to the original purpose of having windows in this building providing the energy efficiency in the heating season. The design process of an energy efficient window system comprises four main steps considering the objectives, requirements and constrains related to the whole attachment in general. The four main steps are; market search, quantifying energy performance data at product level with Window5 [4] simulations, quantifying energy performance data at building level with DOE 2.1E [5] simulations and evaluation / decision-making. It becomes important for the designers to integrate these criteria into a broader set of design issues and values while understanding the long-term implications of their decisions.

3.1. Market Search At the “market search” stage, appropriate glass types and their combinations are determined, with the main requirements like local availability, initial cost, system weight, aesthetics and durability [6]. The local glass company is one of the worldwide leading companies that support the research and development with its products used in this study. The appropriate glazing systems for the study are developed by using the determined glass types as a 6/12/6 mm system with air filled gap.

3.2. Energy Performance at Product Level At the second step, the energy related properties of the window system alternatives are determined by Window5 computer program. Window5 is a well known computer code, developed by Berkeley Laboratories, capable to compute glazing systems and window characteristics, considering the layers and materials (optical coefficients, conductivity, thickness and others), gaps (thickness, gas mixture, pressure, and others) and frame (U-value properties). The program is found to be quite efficient and reliable. Its results can be stored as a library file readable by other programs in particular DOE 2.1E [4]. U-value, solar heat gain coefficient (SHGC), visible transmittance (VT) and light-to-solar gain ratio (LSG) are the properties of the window systems, which can be used as the basis for quantifying energy performance and allow accurate comparison at window system level before proceeding the building level simulations. The U-value of a window system represents its overall heat transfer rate or insulating value. The ability to control the heat gain through windows by direct or indirect solar radiation regardless of outside temperature is characterized in terms of the solar heat gain coefficient (SHGC) of the window while visible transmittance (VT) is an optical property that indicates the amount of light in the visible portion of the spectrum that passes through a glazing material. The LSG ratio is defined as a ratio between visible transmittance and solar heat gain coefficient, which reflects the concept of separating solar heat gain control and light control [1].

3.3. First Level Evaluation

The window systems that cannot meet the requirements are eliminated according to their performance values of U-value, SHGC and VT obtained by Window5 at the pre evaluation stage. For example the glass systems with high U-value, low visible transmittance and low solar heat gain coefficients can be eliminated before going on the simulations at building level.

3.4. Energy Performance at Building Level At the third step, performance data of the proposed glazing systems calculated with Window5 are used as the input data for energy simulations to appraise their thermal performances at building level. The heat loss through windows, total solar heat gained from the windows and the annual heat loss of the building with glazing system alternatives are calculated by means of DOE 2.1E computer program. The DOE 2.1E program is designed to exp lore the energy behavior of buildings and their associated HVAC systems, which requires as input a geometrical description of the building and a physical description of the building construction, mechanical equipment, end-use load schedules, utility rates and hourly weather data to determine the energy consumption of the building [5]. The hourly weather data of

Istanbul that is located on the latitude of 41-degree north and longitude of 28-degree east, is prepared as ‘Typical Meteorological Year-TMY’ file and used for the simulations [7].

For the simulations one temperature zone is accepted and interior air temperature is kept constant at a temperature of 21°C throughout the year in calculating the heating energy consumption of the extension. All windows of the building are designed as double pane windows and the windows are assembled in 6cm wide white coloured aluminium frame with a thermal transmittance of 3.97 W/m2K. Aluminium spacers between the glass layers are modelled for all glazing units. The lighting type is assumed to be recessed fluorescent and maximum output is specified as 16 W/m2. The area per person is assumed to be 5.6 m2 and total 250 people are modeled to be occupying the studio at the education period. The infiltration method is defined as air-change method and the rate of air changes per hour is specified as 0.5.

3.5. Final Evaluation / Desicion Making At the final evaluation / decision making stage, the energy impacts of optical and physical properties of the developed window systems on the building’s thermal performance are analyzed in terms of annual heat gain and loss of the window system and the appropriate window system for the building is selected. The performance data determined at the product level and building system level are used together for evaluation / decision making stage. Since the building is used for education generally in the heating season, window system annual heat loss (WSHL), window system annual heat gain (WSHG), window system annual net heating load (WSNHL), building system annual heating load (BSHL) are used as the energy performance indicators at the building level. The methodology for the design of an energy efficient window system, expressed in the form of a flowchart is given in Figure 4.

4. The Effects of Glazing System Alternatives on the Thermal Performance of the Historical Building

In the case study, the appropriate glazing systems that can be used for the historical building are constituted by using 5 different glass types such as tinted, reflective, Low-E and spectrally selective Low-E coating, as well as clear glass. The coated glass types are available in the local glass market such as multifunctional (IMF-170), titanium blue (ITB 130), inox (IIN 143) and Low-E (ILE-174) [4]. The optical and physical properties of each glass type used in the developed glazing systems are given in Table 1.

Overall U value, solar heat gain coefficient (SHGC), visible transmittance (VT) and light-to solar heat gain (LSG) values of window systems are calculated with Window5 computer program for the comparison at product level by taking account of the frame properties as well. The calculated data is given in Table 2.

Figure 4. Flowchart describing the methodology for energy efficient window system design.

Table 1. The physical properties of the glass types.

Glass Type Glass ID

Description Tsol Rsol #1

Rsol #2

Tvis Rvis #1

Rvis #2

Emiss. #1

Emiss #2

Multifunctional (IMF-170) 7000

Spectrally Selective Low-E 0.486 0.213 0.320 0.792 0.082 0.049 0.840 0.060

Titaniu m blue (ITB-130) 7001

Tinted 0.208 0.152 0.328 0.300 0.121 0.265 0.840 0.550

Inox (IIN-143) 7002 Reflective

0.385 0.064 0.188 0.427 0.077 0.204 0.840 0.790

Low-E (ILE-174) 7003

Low Emissivity 0.574 0.218 0.144 0.825 0.041 0.055 0.100 0.840

Clear 7004 Clear

0.770 0.070 0.070 0.880 0.080 0.080 0.840 0.840

Tsol: Center-of-glass solar transmittance for the glazing layers at normal incidence Rsol #1 and Rsol #2: Center-of-glass solar reflectance for the glazing layers for radiation incident from the front (outside) and from the back (inside), respectively at normal incidence Tvis: Center-of-glass visible transmittance for the glazing layers, at normal incidence Rvis #1 and Rvis #2: Center-of-glass visible reflectance for the glazing layers for radiation incident from the front (outside) and from the back (inside), respectively at normal incidence Emiss#1 and Emiss#2: Thermal emmissivity of the front surface and back surface, respectively

Table 2. The appropriate window systems and their performance values.

Window ID

Glass ID (1)

Description Glass ID (2)

Description U W/m2K

SHGC VT LSG

A 7000 IMF 170 #2 7003 ILE 174 #3 1.67 0.39 0.65 1.67

B 7000 IMF 170 #2 7004 Clear 1.71 0.45 0.70 1.56

C 7001 ITB 130 #2 7003 ILE 174 #3 1.78 0.22 0.25 1.14

D 7001 ITB 130 #2 7000 IMF #3 1.71 0.20 0.24 1.20

E 7001 ITB 130 #2 7004 Clear 2.47 0.27 0.27 1.00

F 7002 IIN 143 #2 7003 ILE 174 #3 1.79 0.34 0.36 1.06

G 7002 IIN 143 #2 7000 IMF #3 1.71 0.31 0.34 1.09

H 7002 IIN 143 #2 7004 Clear 2.72 0.42 0.38 0.90

I 7004 Clear 7003 ILE 174 #3 1.79 0.59 0.73 1.24

#1 #2 #3 #4 J 7004 Clear 7000 IMF #3 1.71 0.52 0.70 1.35

K 7003 ILE 174 #2 7003 ILE 174 #3 1.70 0.47 0.68 1.45

L 7000 IMF 170 #2 7000 IMF #3 1.65 0.35 0.63 1.80

M 7004 Clear 7004 Clear 2.76 0.69 0.78 1.13

According to the results of the product level the window systems with blue tinted glass (E), reflective coated glass (H) and clear glass (M) do not thermally perform well due to their higher U-values, which vary between 2.47 W/m2K and 2.76 W/m2K when compared with the other alternatives. Moreover Windows E and H have lower visible transmittance values as 0.27 and 0.38, meaning that these systems reduce utilizing visible light in the building as well. However in case of considering heating loads for the performance evaluation of the window system, the U-value of the glazing system becomes the most significant criterion, which is affected primarily by the total number of glazing layers, their dimension, the type of gas within their cavity, and the characteristic of coatings on the various glazing surfaces. In this case the optical characteristics of the glass types are the main parameter affecting the performance of the window system.

The windows with tinted glass (C and D) and with reflective coated glass (F and G) seems to outperform related with their lower visible transmittances between 0.24 and 0.36 although their thermal performances are higher due to their low U-values varying between 1.71 W/m2K and 1.79 W/m2K. Other window system alternatives developed with spectrally selective Low-E and Low-E coated glasses (A, B, I, J, K, L) have rather low U-values between 1.65 W/m2K and 1.79 W/m2K and high visible transmittances between 0.63-0.73 while their solar heat gain coefficients vary between 0.59 and 0.35 (Figure 5).

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

A B C D E F G H I J K L M

Window Systems

SHG

C a

nd V

T

0,00

0,50

1,00

1,50

2,00

2,50

3,00

U-v

alue

and

LSG

SHGC VT U LSG

Figure 5. Comparison of the performance values of the window system alternatives

Windows A, B, I, J, K and L have LSG ratios higher than one meaning that utilizing visible light is higher while their solar heat ga in is lower. However higher solar heat gain coefficients contribute to reduce heating loads in winter. Although light-to-solar gain ratio is an indicator of performance but does not directly correlate with actual energy use. Therefore at this stage, building level simulations are required for understanding the thermal performance of window systems related with the design conditions and characteristics of the building. Although the window design and selection issues related with the window performance values as solar heat gain coefficient, visible transmittance and U-value can be overwhelming for making a design

decision if there were a simple sequence of steps, sometimes it is not easy to reach right solution based on the defined performance criteria.

In order to provide performance information for the selection of the appropriate window system, the heat loss and gains of the building at the education period are taken into consideration and annual heat loss and gain of the windows, overall heat loss of the building and the ratio of heat loss through windows to overall building heat losses are calculated by the aid of computer simulation and the results are given in Table 3.

Table 3. Annual heat loss and gain of the windows and overall heat loss of the building

Window

ID

Window Sys. Heat Loss (WSHL) MWH

Window Sys. Heat Gain (WSHG)

MWH

Window Sys. Net Heat Loss

(WSNHL) MWH

Building System Heat Loads

(BSHL) MWH

Window/Building Heat Loads

(%)

A -34.872 7.816 -27.056 -36.984 73

B -35.510 9.515 -25.995 -36.077 72

C -39.016 4.400 -34.616 -43.483 80

D -38.082 3.926 -34.156 -43.037 79

E -49.502 5.896 -43.606 -51.271 85

F -37.216 6.614 -30.602 -40.061 76

G -36.365 5.931 -30.434 -39.896 76

H -51.202 9.325 -41.877 -49.880 84

I -34.606 9.823 -24.783 -34.863 71

J -34.002 8.943 -25.059 -35.200 71

K -34.529 8.471 -26.058 -36.111 72

L -35.068 7.231 -27.837 -37.654 74

M -49.327 14.263 -35.064 -44.015 80

The window area comprises 25 % of the area of the total opaque and transparent components in the building and the percentage of the heating loads occur from the different window systems vary between 71 % and 84 %. This indicates the importance of the window system design and selection process for achieving energy efficiency since the thermal insulation levels of the opaque building elements are very high.

U-values of the window systems affect the total heat loss through windows and overall heat loss of the building in heating season. However, although the U-values of the Windows H and M are very close to each other as 2.72 W/m2K and 2.76 W/m2K, respectively, the building heat losses associated with these window systems are quite different depending on their solar heat gain coefficients as 49.88 MWH and 44.02 MWH. While heat loads decrease by reducing the U-value, it is also influenced by solar heat gain coefficient. The relationship between the U-values and the annual heat loss through windows and overall heat loss of the building with different glazing systems are given in Figure 6.

The minimum annual heat losses are attained by the Windows B, I, J and K according to the annual heat losses of the building with window system alternatives. Since the solar heat gain from windows is related with their solar heat gain coefficients, higher coefficients contribute to reduce the annual heat losses of the building, (Figure 7).

Double pane clear Window M can be considered as the base case since this type of window system is common on the new building constructions. Window E and H outperform since the rate of decrease in overall building heat loads are –16 % and –13 %, respectively while Window I and J performs best since the savings provided by these windows are 21 % and 20%, respectively when compared to the base case. From the viewpoint of sustainable design, reducing the heat loads by 20 % will contribute to the heating energy savings that will eliminate the environmental impact. Window I with Low-E coating on the second pane provides the minimum heat loss through windows and building with its highest solar heat gain coefficient although it’s U-value is not the lowest value among the window systems in consideration. Therefore for thermal performance evaluation of the window systems building level performance evaluations give more confident results than product leve l evaluations in terms of energy efficiency by taking account of the window parameter values used at the product level evaluations. Moreover Window I meet the budget when compared to the other window system alternatives for this application.

0

10

20

30

40

50

60

A B C D E F G H I J K L M

Window Systems

Hea

t Loa

ds (M

WH

)

0,00

0,50

1,00

1,50

2,00

2,50

3,00

U-v

alue

(W/m

2K)

WSNHL BSHL U

Figure 6. The relationship between the U values and the annual heat loss of windows and building

0

10

20

30

40

50

60

A B C D E F G H I J K L M

Window Systems

Hea

t Los

s (M

WH

)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

SHG

C

WSNHL BSHL SHGC

Figure 7. The relationship between the SHGC and the annual heat loss of windows and building

5. Conclusions A series of comments that summarize the most significant outcomes of this work are: • In order to achieve sensitive results in energy efficient window system design, a

methodical approach is introduced in the context of this study.

• While most windows in buildings today have a negative impact on a building’s energy load, the technological potential for windows to be designed could make a positive contribution to the building’s annual energy load. Therefore the physical properties of the glazing system such as U-value and SHGC must be taken into consideration with their effect at building level during window design process.

• Visible transmittance (VT) of glazing system should take place in window design process as another criterion to fulfill other performance attributes such as utilizing the daylight.

• The light-to-solar gain ratio does not directly correlate with actual energy use but it can be used as an indicator of performance involving both visible transmittance and solar heat gain coefficient.

• Although the information obtained at the level of the material or product can be used as the criteria for an energy efficient window design, the effort focusing on the building level impacts of these products helps to evaluate their performance in the building and to understand their energy saving potentials.

• In this case solar heat gain coefficient of the glazing system is as significant as U-value in terms of heat loss/gain in the heating season.

• According to the results the glazing system with a Low-e coating on the third surface that has higher solar heat gain coefficient, is the most efficient window system in terms of

heating loads of the building and initial cost. Moreover it provides daylight utilization with its high visible transmittance value as 0.60.

6. References

[1] Carmody, J. et al. (2004) “Window Systems for High - Performance Buildings”, W.W. Norton Company, New York. [2] Arasteh, D. (1995) “Advances in Window Technology”, Lawrence Berkeley National Laboratory Publications, LBL-36891 UC-1600. [3] Tavil, A., Özkan, E. (1998) “The Effects of Fenestration Characteristics on the Thermal Performance of Retrofitted Residential Buildings in Istanbul”, Proceedings of CIB World Building Congress, Construction and the Environment, p.1123-1131. [4] Arasteh, D. et al. (2001) “WINDOW5, User Manual for Analyzing Window Thermal Performance”, Lawrence Berkeley National Laboratory Publications, LBNL Report No-44789. [5] Winkelmann, F.C. (1981-1993) “DOE 2 Reference Manual” Version 2.1D, Part1; Part 2, Supplement Version 2.1E, Lawrence Berkeley Laboratory, California. [6] www.campazarlama.com.tr [7] Tavil,A., Sahal.N., Ozkan, E., (1997) ‘The Simulations of the Thermal Performance of Retrofitted Existing Residential Buildings in Istanbul with DOE 2.1E”, Proceedings of the 5th International IBPSA Conference, Building Simulation 97 in Prague,V.2, pp.363-371.