Energy Convers. Mgmt Vol. 39, No. 3/4, pp.

243-256, 1998
© 1997 Elsevier Science Ltd. All rights reserved
Pergamon Printed in Great Britain
P I h S0196-8904(96)00191-4 0196-8904/98 $19.00 + 0.00

ENERGY AND ENVIRONMENTAL IMPACT ANALYSIS OF

DOUBLE-GLAZED WINDOWS

G. WEIR and T. M U N E E R
Mechanical Manufacturing and Software Engineering Department, Napier University, 10 Colinton Road,
Edinburgh EHI0 5DT, U.K.

(Received 15 May 1996)

Abstract--The embodied energy of the four main materials used in the construction of an inert gas filled,
double-glazed window, of a given size, was quantified. The four main materials considered were infill gas
(argon, krypton and xenon), timber, aluminium and glass. The energy requirements of the underlying
manufacturing processes are also estimated for the processing of the finished product. The total embodied
energy for argon, krypton and xenon cavity filled windows (1.2 m by 1.2 m standard tilt and turn window),
was found to be 1031 MJ, 1539 MJ and 5531 MJ per window, respectively. For argon filled cavities, this
leads to the production of 94.7 kg of CO2, 1.2 kg of SO2 and 0.4 kg of NO~ per window. The results
presented in this paper are initial findings in the first stages of a complete Life Cycle Analysis (LCA) of
double-glazed windows for sustainable buildings. © 1997 Elsevier Science Ltd.

Life cycle assessment Embodied energy Environmental impact Multiple-glazed windows

Fenestration

1. I N T R O D U C T I O N

Since the late 1960s, Life Cycle Analysis (LCA) has become an increasingly important tool for
engineers, technologists, scientists, designers, managers and environmentalists. LCA enables the
effects that products, processes and activities have on local, regional or global environments to be
assessed. The conditions under which a system exists, the circumstances that prevail, and those
factors that influence change in a system, all define the nature of a particular environment. It is
necessary to consider the impact that raw material extraction, energy production, manufacturing
processes, transportation needs and waste disposal requirements have on both social and natural
environments. In addition to these criteria, consumer requirements and preferences must also be
built into the design phase. Impacts on the environment include effects on the atmosphere and the
world's natural resources, in addition to human health factors, animal habitats, fuel depletion,
noise pollution and the availability of raw materials and primary fuel for the future. This list is
not intended to be exhaustive.
Carbon dioxide (CO2) has many influences on the environment and has a very large part to play
in global warming issues. In the building and construction trade, CO2 is the largest environmental
concern, accounting for over 50% of the UK's total CO2 emissions. This has prompted research
into new material production, product design, manufacturing methods, recycling technology, use
patterns and waste disposal management. Table 1, extracted from Ref. [1], shows the carbon
emissions resulting from the production of some commonly used building materials.
Greenhouse gases absorb infrared radiation from the sun's rays which would normally be
reflected back into open space. This retained heat in the atmosphere is reported to be very slowly
increasing the earth's temperature. The natural greenhouse effect maintains the surface of the earth
at a temperature which enables inhabitation and is caused by important natural gases: CO2, water
vapour, methane (CH4) and ozone (03). The effects of rising concentrations of CO2, methane,
nitrous oxides (NOx) and ozone in the atmosphere, manifested in rising global temperatures, was
first noted at the beginning of the nineteenth century by Jean-Baptiste Fourier. Other gases, not
found naturally in the atmosphere, such as chlorofluorocarbons (CFCs), have also increased in
concentration since the commencement of industrialization. A worldwide rise in temperature,
243
244 WEIR and MUNEER: ANALYSISOF DOUBLE GLAZED WINDOWS

Table 1. Carbon emissionsresultingfrom the productionof commonlyused buildingmaterials.

Extracted from Ref. [1]
Carbon released Carbonstored Net carbon emitted
(kg/m3) (kg/m3) (kg/m3)
Treated timber 22 250 -228
Glue laminated timber 82 250 -168
Structural steel 8132 15 8117
Reinforced concrete 182 0 182
Aluminium 6325 0 6325

estimated at 2-6°C over the next century could result from these changes. Warming of this
magnitude would alter climates throughout the world, cause sea levels to rise and affect crop
production. Much of the CO2 in the atmosphere is a direct result of fossil fuel burning, which
accounts for about 5 billion tonnes of CO2 production annually. This alone is responsible for a
yearly CO2 concentration rise of 0.4% [2]. The other two major influences on CO2 production are
cement manufacturing and changes in tropical land use, accounting for about 1 and 2 billion
tonnes, respectively [2]. The development of products with low embodied energy contents, which
use reduced quantities of energy and which contribute reduced amounts of greenhouse gases to
the atmosphere in their use phase, has become of prime importance.
In this study, the embodied energy of raw materials and manufacturing and associated CO2, SO2
and NOx contents have been estimated for a double-glazed, timber framed window, containing an
inert gas filled cavity. The major source of data for this analysis is the result of a detailed audit
of manufacturing processes, undertaken by the authors, at the window production plant of
Nor-Dan Windows, based in Moi, Norway. This was completed during the early months of 1996.
Before details of the present work are discussed, a brief introduction to LCA is given herein.

2. LIFE CYCLE ANALYSIS

LCA takes a holistic view of the entire life cycle of a product, process or activity with a view
to identifying those stages of the life cycle which have the greatest adverse effect on the
environment. Field et al. [3] state that LCA has become one of the most actively considered
techniques for the study and analysis of strategies to meet environmental challenges and concludes
that it is one of the most promising approaches to integrating environmental knowledge and data
into a framework for action. The life cycle of a product is not simply its period of useful
employment but includes each stage from raw material extraction, manufacturing, retailing, use
and maintenance to disposal or recycling. At each stage, the transportation effects must be taken
into account, in addition to the underlying requirements of administration and work force services.
There are four main stages to an LCA:
(1) Planning,
(2) Inventory analysis,
(3) Impact assessment,
(4) Improvement analysis.
These stages are now discussed in detail.

2.1. Planning
The main purpose of the planning phase is to define the investigation boundaries. LCA may be
used for many different purposes, including the comparison of two products, identification of
harmful stages in a product life cycle, new product development, allocation of process resources,
categorizing of research and development needs or aggregation of total product environmental
burdens. During the planning of an LCA, the objective must be clearly borne in mind while the
system limitations are defined, and the methods for data collection and evaluation are prescribed.
System limitations may include such considerations as data availability, future environmental
effects and technology development requirements. Before moving to the second stage (inventory
analysis), a clear definition of investigation inclusions and exclusions and the methods of data
 WEIR and MUNEER: ANALYSISOF DOUBLE GLAZEDWINDOWS 245

retrieval must be described. The following set the boundaries for most LCAs conducted and define
the boundaries of the investigation herein.
Inclusions Exclusions
• raw material extraction • manufacture of capital equipment
• manufacturing processes • maintenance of capital equipment
• packaging requirements • manufacture of services
• transportation requirements • maintenance of services
• in-use repairs and maintenance • energy required to raise capital
• waste disposal and recycling for production
• fuel production and use
• energy generation

2.2. Inventory analysis

The inventory analysis is a methodical quantification of inputs and outputs. This is a measure
of all matter that crosses the boundary defined in the planning phase. Energy, raw materials, air
emissions, water-borne effluent and solid waste are examined and measured. This may employ
several different data acquisition methods, from measurements to database searches, surveys,
questionnaires, analysis of historical data, theoretical calculations and individual interviews.

2.3. Impact assessment

Impact assessment focuses on how the product affects the environment. This requires a
qualitative and quantitative approach to analysing how raw material use, energy generation, water
production, effluent output, air emissions and solid waste affect the environment. It is a question
of defining and characterizing the consequences which result from the inputs and outputs quantified
in the inventory analysis. The World Resource Foundation (WRF) [4] considers LCA to be a vital,
ongoing tool, in combination with the trend towards more open disclosure of environmental
information by companies, and the desire by consumers to be guided towards the least harmful
products. The WRF advises that LCA must, however, be used with care; the impact assessment
of the inventory analysis should be approached cautiously with regard to subjective judgments.

2.4. Improvement analysis

Improvement analysis involves making improvements to reduce environmental burdens. This
requires taking an objective view of the entire life cycle and assessing the impact that changes would
have on the environment. This may result in product design changes, raw material substitutions,
manufacturing process changes, improved waste management facilities or in a suggested consumer
use change. There are many benefits associated with the incorporation of LCA into an organized
environmental management system. These include the reduction of uncertainty with regard to
environmental impact and the meeting of consumer requirements due to market demands for
increased environmental awareness. A further benefit of establishing such a system, whereby waste
and energy consumption are minimized, results in improved profitability [5].
A simplified illustration of the LCA process is shown in Fig. 1.

3. WINDOW PRODUCTION
The second stage of LCA (inventory analysis) involves the quantification of all raw materials,
energy inputs, air emissions and water effluent from the entire product life cycle. This paper focuses
on the stages up to and including the manufacturing of the finished window. The manufacturing
plant used in this investigation contains 23 halls of production processes, each requiring machinery
power, lighting, heating, services, transportation and raw material distribution. There are six basic
window types produced, each having a possibility of over six different finishes. Each window
construction is available in an infinite range of dimensions, including irregular shapes for special
projects. The basic construction for each window begins with a timber flame and sash, designed
and milled to a high quality, ensuring a good fit with minimal insulative losses and sound
transmission. With so many products being manufactured in one factory, it was too complex to
246 WEIR and M U N E E R : ANALYSIS OF DOUBLE G L A Z E D WINDOWS

FUEPRODUCTION~,
L

MATERIAL MANUFACTURE USE AND

EXTRACTION ' MAINTENANCE

I tl
Fig. 1. Life Cycle Analysis (LCA).

analyse every construction possibility. As a result, one window construction was chosen, using a
popular size of 1.2 m by 1.2 m. The window type selected accounts for the largest production output
from the factory, described as a three handle tilt and turn window mechanism. This is the most
commonly ordered window, accounting for over 50% of total production.
The manufacturing process was divided into sections to simplify the procedure of data
acquisition. Timber processing, aluminium optimizing, assembly, glazing unit production, painting,
lighting, heating, administration, services and technical design needs were all analysed individually,
with a final aggregation to sum the data collected.
Administration and service needs were taken into account in the factory as a separate, but
necessary issue. For the product to be produced, marketed and sold, there must be administration
personnel, a design team, factory workers and management. To sustain a work force requires
services, heating, lighting and materials. Each of these has implications for energy consumption
and material use.

4. ENERGY CONSUMPTION
The inventory analysis covers the following:

(1) Inert infill gas,

(2) Timber sash and frame,
(3) Aluminium,
(4) Sealed glazing unit,
(5) Manufacturer.

Each of these is now discussed in detail.

4.1. Inert infill gas

There are two issues to be addressed when considering the environmental impact of cavity
infill gases. The first of these is the associated benefits, in terms of improved thermal properties
and reduced heat loss from the window. The second issue concerns the consequence which must
be addressed, regarding the energy input and environmental burdens created as a result of
production. In this analysis, three infill gases are investigated and compared with the use of air
in cavities.

• argon,
• krypton,
• xenon.
 WEIR and MUNEER: ANALYSISOF DOUBLEGLAZEDWINDOWS 247
1.8
1.6
1.4
1.2
1
0.8
~, 0.6
: 0,4
0.2
0
AIR 20ram ARGON KRYPTON XENON
CAVITY l(k~m 12ram 8ram
CAVITY CAVITY CAVITY
Fig. 2. Heat loss characteristics for air and inert gas filledcavities.

Using inert gases in place of air in sealed glazing units can substantially increase thermal efficiencies
and reduce noise transmission in multiple-glazed window systems. Decreasing the thermal losses
from a sealed window will improve building performance, reduce consumer fuel bills and place a
lesser burden on the environment from atmosphere pollution. An extensive study was made at
Napier University to define both the benefits and consequences associated with the use of heavy,
inert gases in glazing cavities. An analysis of the gas production process at BOCs Middlesborough
plant was conducted to estimate the energy content of the gases, and extensive research was
undertaken to establish the thermal improvements possible in their use. Figure 2, extracted from
Ref. [6], shows the heat loss characteristics of four similar windows, using air, argon, krypton and
xenon cavity infill gases. An optimum cavity gap exists, dependent upon the infill gas used. As the
molecular weight of the gas increases, the cavity gap required decreases, allowing designers to
manufacture very high performance windows without having to increase significantly the weight
of the frame construction. The optimum gap for air, argon, krypton and xenon was found to be
20, 16, 12 and 8 mm, respectively. The high insulative properties of inert heavy gases are due to
the hyperbolic relationship between thermal conductivity and molecular weight: as the molecular
weight rises, the thermal conductivity drops dramatically. All of the window options shown in Fig.
2 have a low-emissivity (low-E) coating on both glazings. This is standard in the window
construction adopted by Nor-Dan Windows. Applying a low-E coating to one glass surface will
reduce heat loss in an air filled window by up to 40%. Applying a low-E coating to both glazings
is not justified in air filled windows but creates significant energy savings in inert gas filled cavities
[6]. Other methods have been tried to create the same benefits, but these led to very heavy weight
glazing solutions (triple, or quadruple glazed windows) which demanded stronger frames and
increased material consumption.
Having considered the benefits to be gained from the use of heavy inert cavity infill gases,
attention is drawn to the energy content and environmental consequences of gas production.
Argon, krypton and xenon are all present in the atmosphere, and production of the pure gases
is a process of separation from the other components present. Table 2 shows the percentage content
of gases in air. Trace gases include water vapour, hydrogen, ozone, methane, carbon monoxide,
helium, and neon. The process of separating inert gases from air is shown in Fig. 3, extracted from
Ref. [7].
The equipment required for the separation of gases from air has a power rating of 12 MW. This
is the only equipment required for extracting argon gas, krypton and xenon, however, require the

Table 2. Percentagecompositionof air

Component Percentage
Nitrogen 78
Oxygen 21
CO: 0.03
Argon 0.9
Krypton 0.000114
Xenon 0.0000087
Trace gases 0.069
248 WEIR and MUNEER: ANALYSISOF DOUBLE GLAZED WINDOWS

LOSSES
8W0~)t~mN% "I"A /~ 40P%~)0N23/hr n 4600m3/hr

Airln
] 50000rn3/hr
J] 1"l °"nunil' , ~~°n
Air Sepcat' N
! 900m3/1~

CH4 Gas 02
< Kr / X¢ Purification

Kr 91.3*/,
Xe 6.9*/, (0.047m3/hr)
CH4 1.8%
Fig. 3. Gas production schematic for BOC Middlesboroughplant.

running of a crude krypton/xenon column, consisting of a reactor (4 kW), heater (10.8 kW), flash
vaporizer (4 kW), liquid oxygen pump (2.3 kW) and purification column (2.4 kW). The actual yield
of gas from the column is between 31 and 54%. The production rate data is based on an average
yield rate of 42%. Table 3 shows the yield volume per hour, energy consumption rate for extraction
and the specific energy consumption per litre of gas produced.
For a window measuring 1.2 m by 1.2 m, the area of glazing is 1.1 m 2, and the cavity gap for
argon, krypton and xenon would be 16, 12 and 8 mm, respectively, requiring a gas volume of 17.6,
13.2 and 8.8 litres, respectively. The energy requirement to produce argon, krypton and xenon gases
to fill this gap is 11.83 kJ, 508.2 MJ and 4.5 GJ, respectively. The values given for krypton and
xenon filled cavities are hypothetical, since Nor-Dan Windows is restricted, by legislation, to the
use of argon and air filled cavities alone.
Filling the glazing units with inert gas is a carefully monitored and controlled process. The panes
are filled via filling and control holes drilled in the spacer, and the amount of gas flow is regulated
using flow valves. The gas quantities and pressure regulation ensure that the mixing of gas with
the displaced air is minimized, such that there is no "flushing" process taking place. The gas is
detected using special probes, and the gas flow is stopped by the closing of a solenoid value once
the correct concentration is achieved in the pane. This ensures that gas wastage is maintained at
a minimum.

4.2. Timber sash and frame

The timber used in the window manufacturing process is 100% softwood and is obtained from
organized, well-established tree management programmes in Scandinavia (Finland 30%, Norway
30%, Sweden 40%), whereby two trees are planted to replace the felling of one; 3.5 billion tonnes
of wood are felled each year, which is equivalent to 0.7 tonnes per person per year [1]. As the world's

Table 3. Yield rate and energyconsumptionfor the production of inert gases. Extracted from Ref. [3]
Yield volumeper hour Energyconsumption rate Specificenergy consumption
Gas (hr-~) (kW) (kJ/litre)
Argon 900,000 168 0.672
Krypton 44.43 475.5 38500
Xenon 3.39 475.5 511400
 WEIR and MUNEER: ANALYSIS OF DOUBLE GLAZED WINDOWS 249

population is rising by approximately 100 million people annually, this demand for wood is ever
increasing. Softwoods, by their very nature, are less resistant to attack from rot and insects but
are faster growing and more easily processed than hardwoods. Reforestation programmes for the
felling of hardwoods are seldom as frequent or comprehensive as those for softwoods. The felling
of hardwood trees has many effects on the environment. Apart from the fact that trees absorb
carbon from the atmosphere as they grow, there are also many problems for tropical rain forest
areas, where hardwoods are predominantly felled for the commercial trade. The soil structure in
these areas has a complex ecological balance. When a tree is felled, this balance is disturbed, and
the useful topsoil is washed away by heavy rains, disabling the growth of a new tree in its place.
In fact, the land where felling has taken place can be put to very few uses, as the soil cannot support
any kind of agriculture or cultivation without major improvements. Tropical rain forests are the
home of many undiscovered plant and animal species, which could be lost forever if the natural
habitat in which they are sustained is damaged. Plantation forestry accounts for approximately
10% of timber requirements worldwide. About 2 billion tonnes of CO2 enter the atmosphere as
a direct result of forest destruction each year [2]. This is alarming and has prompted many
researchers to seek alternatives to timber in construction. Very often, however, these alternatives
have more serious implications for the environment. Alternative materials may not be renewable
or recyclable and may require larger quantities of energy to convert into useful products. This
would result in increased CO2 levels entering the atmosphere. Timber has a high strength to weight
ratio and is easily formed for many applications. Timber possesses good thermal and sound
insulating properties, does not corrode and, given the right conditions and treatment, will not rot.
One very popular treatment applied to softwoods, and adopted by Nor-Dan Windows, is
preservative impregnation. All timber components are impregnated with a white spirit based
solution, which soaks into the timber pores and slows down the process of decay in the timber,
thus lengthening the window life cycle. All timber sections are enclosed in a designated chamber,
and a vacuum is created. The preservative tank is then opened, and the chamber is filled. The
vacuum is maintained for a few minutes before a discharge pump is operated, and the vacuum is
lost. The loss of vacuum ensures that each timber pore is filled with preservative. The chamber
is then emptied of preservative, and a new vacuum at a higher pressure is created over a longer
period of time. A final discharge completes the impregnation process, and the timber remains within
a drying chamber for 1-2 days. All the stages of manufacture, including the timber impregnation,
have been taken into account when considering the energy content of the windows. This is described
in Section 4.5.
There are many different uses for timber, requiring different treatment processes and having
varying energy contents. Table 4, extracted from Ref. [1] shows the energy content for some popular
forms of timber used in construction.
This information must be used with the boundaries of analysis borne in mind. Other work has
shown energy contents that are both higher and lower than the figures shown in Table 4. The data
in Table 4 is the product of research by Buchanan and Honey [1] and allows for the energy
consumption of machinery and machinery production, including the energy required to raise
capital for production. Another analysis by West et al. [8] produced data for indigenous softwood
which includes entries for transportation, material extraction, processing and manufacture. This
work arrived at an energy content of 2.6 GJ/m 3. Further studies by Hollinger and Hunt [9], Halliday
[10] and Dinesen and Trabergy-Borup [11] arrived at figures of 0.51 GJ/m 3, 1.77GJ/m 3 and

Table 4. Energy content for some popular

forms of timber used in construction.
Extracted from Ref. [l]
Energy content
Timber type (MJ/m 3)
Rough timber 848
Glulam timber 4500
Hardboard 20600
Softwood 15470
Particleboard 12900
250 WEIR and MUNEER: ANALYSISOF DOUBLEGLAZED WINDOWS

4.03 GJ/m 3, respectively. These final values are widely dispersed for a number of possible reasons.
The working boundaries, as prescribed in the planning phase of an LCA, must be kept central to
all decision making in the analysis. Varying boundary inclusions can cause energy content
valuations to differ greatly. Manufacturing methods can also be widely variant in nature,
incorporating many differences in procedure, machinery and available technology. Work by West
et al. [8] was found to be most closely related to the aims and boundaries of this investigation,
and hence an energy content of 2.6 GJ/m 3 of softwood timber is used in this analysis.
Allowing for the waste within the window production process (discussed in Section 4.5) and given
the window dimensions and type analysed in this paper, a total mass of 37 kg of timber is used
(frame 14.94 kg, sash 10.01 kg, waste 12.05 kg). The density of softwood varies greatly, according
to the water content in the timber, but an average of 500 kg/m 3 is used. This gives an energy content
for the timber component, exclusive of machining, of 192.4 MJ.
Timber arriving at the factory site is 70% pre-cut to the correct length and is transported directly
to the storage hall. Non-standard timber accounts for approximately 30% and must be cut to size.
For the sash, timber must pass through a laminating process where two or three sections of timber
are glued and compressed together. Therefore, to complete the analysis for the embodied energy
of timber, a contribution for the glue involved in laminating must be added. Honey and Buchanan
[12], in their analysis, estimated that 0.5 kg of glue is required per square meter of laminate surface;
25% of timber laminated involves one layer of glue, and two sections of timber, while the remaining
75% requires two layers of glue and three sections of timber. Each layer covers an area of 0.07 m 2.
Therefore, taking into account that only 30% of sashes are laminated, an average of 0.04 m 2 of
glue is applied to each sash (expressing this in mass terms gives 0.02 kg of laminate glue). The glue
is assumed to have an energy intensity of 160 MJ/kg, which equates to 2.94 MJ per sash.
The total embodied energy for the timber sash and frame, inclusive of timber processes and
laminating glue, hence becomes 195.3 MJ.

4.3. Aluminium
The processes involved in cutting and assembling the aluminium components for the window
construction involve minimal quantities of energy. The main consideration for material
consumption and energy use comes at the aluminium manufacturing stage. Aluminium smelting
and forming is a highly energy intensive procedure and must be considered in the LCA of the
window unit. The process of manufacturing aluminium components from raw bauxite is a
multi-stage procedure. The most energy intensive stages in the procedure include the crushing of
raw bauxite, electrolysis of alumina to produce soluble aluminium and the casting of ingots. The
final processing is carried out in the window factory, whereby lengths of formed aluminium are
cut to the desired size. All other processing is done at the aluminium plant. Again, the energy
content of aluminium production is much debated. Buchanan and Honey [1] quote a content of
130 MJ/kg for general aluminium, while Saito and Shukuya [13] quote a content of 502.5 MJ/kg
for pressed and finished aluminium window frames. Further work by Young and Vanderburg [14]
produced data for both primary and secondary aluminium. Primary aluminium is produced from
raw bauxite, and secondary is that produced from recycled aluminium, having gross energy
requirements of 225 MJ/kg and 50 MJ/kg, respectively. This work was found to be in line with the
aims of the LCA investigation. Aluminium is used for many different purposes within the window
construction. Aluminium profiles are fitted along the bottom edge of the sash to protect the timber
from water penetration. The window opening mechanism, various sections of the ventilation grills
and the glazing unit spacer are also made of aluminium to ensure a lightweight design. The total
mass of aluminium used in a standard window of 1.2 m by 1.2 m and the total waste from each
process was calculated as shown in Table 5.
The total waste at the window factory site, 0.306 kg per window, is not discarded but is sold
to a metal merchant for recycling as secondary aluminium. The total energy input required to
produce and process the aluminium for one window, assuming that 100% primary aluminium is
used, is estimated to be 517.5 MJ. As noted previously, melting and reprocessing secondary
aluminium uses a fraction of the energy to produce 1 kg of formed aluminium (50 MJ/kg compared
to 225 MJ/kg for primary aluminium). This is because the energy intensive processes of smelting
and electrolysis are not required for secondary aluminium. UNIDO [15] estimated that 27% of the
 WEIR and MUNEER: ANALYSIS OF DOUBLE GLAZED WINDOWS 251

Table 5. Estimate of total aluminium mass incorporated into one window construction
Aluminium mass Waste Total aluminium mass
Window component (kg) (%) (kg)
Glazing unit spacer 0.241 3 0.248
Frame ventilation 0.159 17 0.186
Outer protection 1.421 17 1.663
Window mechanism 0.174 17 0.204
Total 1.995 2.301

world's total aluminium production comes from recycling aluminium. This percentage is likely to
be lower for building materials due to the design for long lifespans in construction components.
If the assumption is made that 27% of the aluminium used is recycled, then the energy intensity
for the aluminium used in the window becomes 408.8 MJ.

4.4. Sealed glazing unit

The production of large glass panes supplied to the factory is also an energy intensive process,
which must be taken into account in the LCA study. West et al. [8], Saito and Shukuya [13] and
Dinesen and Trabergy-Borup [11] report the energy content per kg of sheet glass manufactured to
be 13 MJ/kg, 16.9 MJ/kg, and 18.6 MJ/kg, respectively. There is less discrepancy between these
values than for aluminium and timber. This is perhaps because the technology used is more
standardized, and fewer differences between manufacturing methods exist. Again, the work of West
et al. [8] was found to be compatible with the aims of this analysis, and an energy content of
13 MJ/kg for sheet glass is used.
An energy and material analysis for sheet glass manufacturing was conducted. The mass of glass
required for a double-glazed unit of 1.1 m 2 was calculated to be 21.2 kg in the finished product.
A small amount of material is wasted due to breakages and small offcuts which cannot be used.
This amounts to approximately 5.5% of the total glass utilized. Taking this into account
means that the average glass consumption for one unit is 22.26 kg, having an energy content of
289 MJ.

4.5. Manufacture
Having discussed the major material inputs to the window system, the manufacturing process
itself must be considered and the energy consumption for production, administration, services,
lighting and heating estimated. For simplicity, this is split into the following:

(1) Timber sash and frame production,

(2) Sealed glass unit production,
(3) Aluminium processing,
(4) Lighting and factory services.

4.5.1. Timber sash and frame production.Energy monitoring equipment was used to measure the
average no-load machinery power, average operational power and power factor for each piece of
machinery involved in the line. Production schedules were closely monitored, and an average
percentage was calculated for the split between the time spent running on no load and the
functional time that each machine runs on operational load. Each machine is left running on no
load for a large percentage of the day (average > 50%). There are a number of reasons for this.
Machine tools have to be changed for different window profiles, differing window sizes have to
be made in batches and require different timber dimensions, and machinery can break down owing
to wood jamming or falling off the line.
The moulding lines are an energy intensive sequence of mills and saws which profile the timber.
Further milling is required to shape each sash and frame for the ironmongery and accessories added
later. Figure 4 shows the final outcome of all timber and the percentage of timber entering the
factory that is wasted.
252 WEIR and MUNEER: ANALYSISOF DOUBLE GLAZEDWINDOWS

TIMBER USE

37%
II OPTIMISING WASTE
[] DEFECTIVE LOSS
IE PROFILING WASTE
m FRAME
2% i SASH

24%
Fig. 4. Timber use and waste within the manufacturingprocess.

All the sawdust produced from these lines is extracted via a comprehensive network of ducts
and fans. Three uses are made of the wood offcuts and sawdust by-products:
• sawdust and small wood offcuts are combusted in furnaces for factory heating;
• sawdust is sold to neighbouring farmers to improve soil quality;
• larger wood offcuts are sold to factory workers for domestic heating.
The final two stages of production for the sash and frame include assembly and impregnation.
The average energy consumption for the production of one sash and one frame was estimated
to be 16.9 MJ and 16.3 MJ, respectively.
4.5. 2. Sealed glass unit production. The production process for manufacturing sealed glazing units
is shown in Fig. 5. The first task in the glass production procedure is to plan the cutting schedule.
The glass sheets are precision cut on large, programmed cutting tables. Then the edges around
low-E coated glass sheets are ground to ensure that the adhesive seals the unit adequately. Glass
panes are then washed to remove dirt and dust particles from the interior of the unit. An aluminium
spacer is fitted to separate the glazings and is kept in place by a butile compound. Before the unit
is sealed, it is fitted with inert gas, using the procedure note in Section 4.1. The unit is finally sealed
with a butile compound and is ready for assembly into the window sash and frame.
Three phase energy monitoring equipment was used to analyse the energy consumption of the
glazing unit production line. This was a fairly simple task owing to the fact that the whole line
could be examined as one load, and there was no need to measure each individual process
separately. The equipment was left for several hours in the working day, and the parameters were
set to downland the instantaneous and average apparent powers, and the power factor of the
associated machinery. The integration time for averaging the output was set to 5 min, giving 12
readings per hour and a reasonable number of data points over an eight hour day. The readings
were found to be very stable, with few fluctuations over the sample period, allowing for staff breaks
and lunch rests. The average real power load for the entire process line was found to be 132.1 kW.
Based on a 12.5 h working day, and 230 working days in the year, this gives an annual energy
consumption of 1367 GJ. The output from the glazing department averages 228,022 finished units
per year, and an allocation of 6 MJ per finished window was estimated. These are used for many
purposes; sometimes two or three are used in the same window configuration, a proportion of
supplies is sent to the nearby door factory and others are transported to other Nor-Dan factories
in Norway. To analyse the machinery energy requirements for a specific size of glazing unit would

i1 °-° 1 -°
Fig. 5. Sealedglazingunit production line processes.
 WEIR and MUNEER: ANALYSIS OF DOUBLE GLAZED WINDOWS 253

Table 6. Average power load and energy consumption for non-machining requirements
Average power Annual energy Energy consumption
load consumption per unit production
Energy use (kW) (GJ) (MJ)
Administration 69.84 462.62 2.1
Technical design 57.37 380.02 1.7
Heating 49.40 1557.88 11.3
Lighting 255.85 3389.50 24.5
Night-time use 377.86 5005.89 36.2
Weekends 337.88 3036.05 21.9
Total 97.7

have been an extremely complex exercise, and the assumption that all unit sizes consume equal
quantities of energy was made. This has slightly erroneous implications, in that some processes may
use more energy for larger or smaller units, but the differences are estimated to be so small as to
be rendered negligible.
4.5.3. Aluminium processing.Once delivered to the factory, the cutting machinery for aluminium
uses very little energy and has a load power of approximately 4 kW. This gives an annual energy
input of 26.5 G J, or 0.2 MJ per finished window. It is clear that the energy of aluminium production
from raw bauxite is of greater concern.
4.5.4. Lighting and factory services.Lighting and factory services are not directly related to the
manufacturing of the finished window but are an essential element of the overall production
process. Without these elements, production would be very difficult, if not impossible. They are,
however, not so easily subdivided into manufacturing stages and must be taken into account in
a unit overhead. Included in these are the factory services of administration, technical design,
heating, lighting and weekend and night-time loads. The average power load, annual energy
consumption and energy consumed per window is shown in Table 6.
Maintenance and repairs are kept outside the investigation boundaries for all stages of
production, including machinery, lighting, services and transportation. During the product use
phase, maintenance and repairs become important, as the length of useful life for a window is a
direct function of the treatment it receives in situ.
A summary of all the manufacturing processes and the energy requirements for each is given
in Table 7.

5. GREENHOUSE GAS IMPLICATIONS

The energy content of material requirements and manufacturing processes is discussed in detail
in Section 4. Each unit of energy consumed in developing materials and in turning them into
finished window products has its own implications for greenhouse gas production and the
environment, as outlined in the introductory paragraphs. In Norway, however, the energy structure
is virtually 100% hydroelectric power. Approximately 0.3% of the power generated is produced
by burning oil, but this is such a small proportion of the total energy demand that it is deemed
negligible.
For a chosen window of dimensions 1.2 m by 1.2 m, the total energy content, for the five elements
considered, is summarized in Table 8. A graphical representation of the percentage breakdown of
total embodied energy is the window construction, for each cavity gas option, is given in Fig. 6,
while a graph of the aggregated embodied energy is given in Fig. 7.

Table 7. Summary of energy content for manufacturing processes

Energy requirement per window
Function (MJ)
Timber sash and frame 33.2
Sealed glass unit 6.0
Aluminium processing 0.2
Lighting and factory services 97.7
Total 137.1
254 WEIR and MUNEER: ANALYSISOF DOUBLE GLAZED WINDOWS

Table 8. Summary of energy content for raw materials and manufacturing

processes
Embodied energy (MJ)
Window component/function Argon Krypton Xenon
Inert infill gas 0.01 508.2 4500
Timber 195.3 195.3 195.3
Aluminium 408.8 408.8 408.8
Glass 289.4 289.4 289.4
Manufacture 137.0 137.0 137.0
Total 1030.51 1538.7 5530.5

To begin analysing the greenhouse gases emitted owing to the burning of fossil fuels, an analysis
of the methods of energy production for each window component must be made. Figure 8
illustrates the division of total embodied energy for an argon cavity fitted window and the source
of energy generation. Since the source of some window components is changeable owing to market
forces, it is assumed that any component or material produced outside Norway is made using
energy generated from a mix of renewable energy, fossil fuels and nuclear power. The data shown
in Table 9 was produced by the Countryside Council for Wales [16] and shows the emissions
generated for each k W h of energy relative to the average U K plant mix and those emissions
generated from the production of each window type analysed. The U K plant mix present in 1994

100%
90~

70%

5O%
!!i [,=.UFACRE
[El GLASS
laALUMINIUM
4O% ]BTIMBER
| [m INERT INFILLGAS

20~
1016
O~
ARGON KRYPTON XENON
Fig. 6. Percentage of total embodied energy attributable to each window element.

6000
5000
4000
3000
2000
1000
0
ARGON KRYPTON XENON
• INERT INFILL GAS •TIMBER DALUMINIUM
[ ] GLASS [ ] MANUFACTURE

Fig. 7. Total embodied energy per window.

 WEIR and MUNEER: ANALYSIS OF DOUBLE GLAZED WINDOWS 255

Table 9. Summaryof greenhousegas emissions for argon, krypton and xenon

filled cavity windows
Emissions per window
Emissions per (kg)
kWh generated
Greenhouse gas (g) Argon Krypton Xenon
CO., 800 94.7 207.6 1094.7
SO_, 10 1.2 2.6 13.7
NO, 3.4 0.4 0.9 4.7

shows that the total consumption of primary fuel in the U K was composed of approximately 12%
coal, 54% petroleum, 26% natural gas, and 8% primary electricity [17]. These components and
materials manufactured in Norway are assumed to contribute no greenhouse gases in the
environment, since hydroelectric power gives out no substantial emissions, once the hydroelectric
power plant has been constructed. Figure 7 takes into account the location of manufacturing for
each window element and the energy structure prevailing. The emissions begin to escalate as a move
is made from argon cavities to krypton and xenon cavities. This is because a growing percentage
of the manufacturing is done outside Norway. The importance of gases in the total embodied
energy breakdown is greater for krypton and xenon than for argon. This is a hypothetical analysis
for krypton and xenon cavity windows, as no manufacturer in Norway can, by legislation, produce
these very heavy gas filled windows. Argon filled windows are manufactured as standard by
Nor-Dan Windows.

6. CONCLUSIONS
An analysis of the embodied energy of materials, manufacturing processes and greenhouse gases
generated in the production of inert gas filled timber windows was completed during the early
months of 1996. This was undertaken by the authors at the site of Nor-Dan Windows, based in
Moi, Norway. The investigation was subdivided into five main analyses: inert infill gas, timber
requirements, aluminium consumption, glass production and manufacturing processes. The total
embodied energy for argon, krypton and xenon cavity filled windows, per specified window (1.2 m
by 1.2 m tilt and turn standard timber window) was found to be 1031 MJ, 1539 MJ and 5531 MJ,
respectively. For argon filled cavity windows, this leads to the production of 94.7 kg of CO2, 1.2 kg
of SO2, and 0.4 kg of NOx per window. The results presented in this paper address the need for
qualitative and quantitative data acquisition, pertaining to the inputs to, and outputs from, the
manufacture of high-performance windows. This attempts to fill the information requirements in
providing an inventory analysis, as the first stage of a comprehensive LCA of double-glazed
windows for sustainable buildings.
Given the information available at the time of writing and the current state of technology used
in the underlying manufacturing processes, it is seen that Argon filled cavity windows involve the
least quantity of embodied energy and produce minimal amounts of greenhouse gases. Argon filled

HYDRO-POWER 70~ PLANT MIX

:o

l l INERT INFILL GAS

I
IITIMBER
DALUMINIUM
El GLASS
PLANT MIX
I MANUFACTURE

HYDRO-POWER
Fig. 8. Division of total embodied energy, and sources of energy generation.
256 WEIR and MUNEER: ANALYSISOF DOUBLE GLAZED WINDOWS

cavity windows embody marginally more energy than air filled windows (0.01 MJ), but
consideration must be given to other design issues. The air filled window has a U-value of
1.63 W/m 2 K, while the Argon filled window has an improved U-value of 1.3 W/m 2 K. Over a
window life cycle of 40 years, the air filled window will allow more heat transfer to the exterior
than the Argon filled window, resulting in increased fuel consumption, higher consumer bills and
greater quantities of CO2, SO2, and NO,. being permitted to enter the atmosphere. In addition to
this, the optimum cavity gap for air and Argon were found to be 20 mm and 16 mm, respectively.
A wider cavity gap requires the use of a wider aluminium spacer bar, involving increased levels
of embodied energy and higher levels of greenhouse gas creation.
The raw materials of gas, timber, aluminium and glass are the main constituents of the finished
window. There are, however, other steel, zinc, plastic and rubber components which are assembled
onto the finished product and which also must be taken into account in the LCA. These are the
subject of further analysis and are not reported in this paper. Further analysis is also required to
consider the energy of transportation, material recycling and the energy and materials involved in
the maintenance of a window over a 40 year lifespan. This research is ongoing and will attempt
to provide a complete Life Cycle Analysis for the above mentioned window types, taking a holistic
view of the total product life cycle and encompassing all inputs to, and outputs from, the entire
process.

Acknowledgements--The authors are indebted to Messrs. W. Davidson and C. Collins of Nor-Dan (UK) Ltd for the
provision of funding for this investigation and related ongoingwork. Thanks are also extended to Messrs. K. Rasmussen,
W. Kranenborg, O. Ausland, T. Rasmussen and A. Surdal, of Johs Rasmussen a.s., based in Moi, Norway, for their help
in data collection and technical support.

REFERENCES
1. Buchanan, A. and Honey, B., Energy and Buildings, 1994, 20, 205.
2. Callander, B., Environmental Monitoring and Assessment, 1995, 38, 129.
3. Field, F. R. III., Isaacs, J. A. and Clark, J. P., J. Minerals, Metals and Materials Soc., 1994, 46(4), 12.
4. The World Resource Foundation, Information sheet. Warmer Bulletin (46), August (1995).
5. EnvironmentalAssessments of Products: A Course on Life Cycle Assessment, ed. B. Little and B. Pedersen. UETP-EEE,
Helsinki, 1993, Ch. 2.
6. Muneer, T., Han, B. and Truslove, T., Proceedings of the International Workshop on Sustainability of Cities and
Buildings, University of Florence, 13-15 September, 1995.
7. Fernie, D. and Muneer, T., Building Services Engineering Research and Technology, (17), 1996, 43.
8. West, J., Atkinson, C. and Howard, N., Proceedings of the First International Conference of Buildings and the
Environment, CIB, 16-20 May 1994.
9. Hollinger, D. and Hunt, J., J. R. Soc., New Zealand, 1990, 20(4), 337.
10. Halliday, M., Feasibility of using timber for medium rise office structures. Research Rep. 91-3, Department of Civil
Engineering, University of Canterbury, 1991.
11. Dinesen, J. and Trabergy-Borup,S., Proceedings of the First International Conference of Buildings and the Environment,
CIB, 16-20 May 1994.
12. Honey, B. and Buchanan, A., Environmental impacts of the New Zealand building industry. Research Rep. 92-2,
Department of Civil Engineering, University of Canterbury, 1992.
13. Saito, M. and Shukuya, M., ISES '95 Congress, Harare, Zimbabwe, 1995.
14. Young, S. B. and Vanderburg, W.H., J. Minerals, Metals and Materials, 1995, 46(4), 22.
15. UNIDO, United Nations Industrial Development Organisation, Industry and Development, Global Report 1989190,
Vienna, 1989.
16. Countryside Council for Wales, Energy policy and perspectives for the Welsh countryside, 1992.
17. Department of Trade and Industry, Digest of United Kingdom Energy Statistics 1995, HMSO, London, 1995.