Multi-objective optimization of building envelope design for life cycle environmental performance

Accepted Manuscript Title: Multi-Objective Optimization of Building Envelope Design for Life Cycle Environmental Performance Author: Rahman Azari Samira Garshasbi Pegah Amini Hazem Rashed-Ali Yousef Mohammadi PII: DOI: Reference:

S0378-7788(16)30423-6 http://dx.doi.org/doi:10.1016/j.enbuild.2016.05.054 ENB 6689

To appear in:

ENB

Received date: Revised date: Accepted date:

12-10-2015 17-5-2016 18-5-2016

Please cite this article as: Rahman Azari, Samira Garshasbi, Pegah Amini, Hazem Rashed-Ali, Yousef Mohammadi, Multi-Objective Optimization of Building Envelope Design for Life Cycle Environmental Performance, Energy and Buildings http://dx.doi.org/10.1016/j.enbuild.2016.05.054 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Multi-Objective Optimization of Building Envelope Design for Life Cycle Environmental Performance Rahman Azari1,4, Samira Garshasbi2, Pegah Amini1, Hazem Rashed-Ali1, Yousef Mohammadi3 1

College of Architecture, Construction and Planning, University of Texas at San Antonio Young Researchers Club, Central Tehran Branch, Islamic Azad University, P.O. Box 13185768, Tehran, Iran 3 Petrochemical Research and Technology Company, National Petrochemical Company, P.O. Box 14358-84711, Tehran, Iran 4 Corresponding Author 2

Abstract

The building envelope incorporates significant amount of construction materials and is a key determinant of the embodied energy and environmental impacts in buildings. It is also a mediator between indoor and outdoor environmental conditions and has significant impacts on operational energy use in many types of buildings. The present article utilizes a multi-objective optimization algorithm to explore optimum building envelope design with respect to energy use and life cycle contribution to the impacts on the environment in a low-rise office building in Seattle, Washington. Design inputs of interest include insulation material, window type, window frame material, wall thermal resistance, and south and north window-to-wall ratios (WWR). The simulation tool eQuest 3.65 is used to assess the operational energy use, while Life Cycle Assessment (LCA) methodology and Athena IE are used to estimate the environmental impacts. Also, a hybrid artificial neural network and genetic algorithm approach is used as the optimization technique. The environmental impact categories of interest within the LCA include: global warming, acidification, eutrophication, smog formation, and ozone depletion. The results reveal that the optimum design scenario incorporates fiberglass-framed triple-glazed window, about 60% south WWR, 10% north WWR, and R-17 insulation.

Keywords: life cycle assessment (LCA), building envelope, optimization, genetic algorithm

1. 1. Introduction The contribution of buildings to the overall environmental impacts of human activities has been well-documented (EPA 2009, EIA 2015). According to the US Energy Information Administration (EIA 2015), 19% of the world’s primary energy is consumed in the US. The major consumers of the total energy in the US, as the EIA data reveals, include commercial buildings (18.5%), residential buildings (22%), industrial sector (32%), and transportation (27.5%). Buildings also contribute 40 percent to carbon dioxide emissions in the states (EIA 2012) and about 66% to generation of non-industrial solid-waste (EPA 2009). Life Cycle Assessment (LCA) methodology has gained increasing popularity in recent years to assess how buildings or their components contribute to the negative impacts on the environment over their entire life cycle. Considering the entire stages; i.e., cradle to grave, in LCA offers a more informed basis for decision-making, compared to other methodologies that focus on operation phase of life cycle only and rely on metrics such as operational energy use. LCA studies typically address all stages of a building’s life cycle and cover one or more of six impact categories. LCA methodology’s principles, requirements and guidelines are prescribed by ISO 14040 (2006), and ISO 14044 (2006). In 2011, the European Committee for Standardization developed EN 15978 (2011) as the standard for using LCA in assessment of the environmental performance of buildings. In addition, International Standards Organization (ISO) provides metrics and requirements for determining the carbon footprint of buildings through ISO 16745 (2015). Many studies use LCA in assessing the environmental impacts of buildings. For instance, Kosareo and Ries (2007) compare intensive and extensive green roofs versus conventional roofs

with regard to their impacts on ozone depletion, global warming, acidification and eutrophication. Pulselli et al (2009) use energy analysis and emergy evaluation methods to assess environmental costs and benefits of three different types of envelopes (an air cavity wall, a plusinsulated wall and a ventilated wall). Azari and Kim (2012) focus on curtain walls and apply a computational process-based environmental life cycle assessment (LCA) to compare the effect of change in mullion materials on curtain wall’s environmental impacts. Azari (2014) conducts a parametric LCA analysis to examine how the change of design input values in a limited number of building envelope configurations would impact the environment. Dodoo et al (2014) is an example of studies that focus on building structures. The study uses the consequential-based LCA to compare three versions of timber structures; i.e., crosslaminated timber, beam-and-column and modular structures. In an LCA study of an office building, Junnila and Horvath (2003) conduct a complete life-cycle analysis on an office building with a service life of 50 years and investigate its impacts with regard to acidification, eutrophication, climate change, etc. They later extend this study and compare the impacts of office buildings in the US and Europe (Junnila et al 2006). In Australian context, Treloar et al. (2001a, b) study the embodied energy of construction materials (2001a) in a building and, in another project, measure the embodied energy in several office buildings and study the effect of height and number of floors on changes in embodied energy. In another study, Yohanis and Norton (2002) study the variations in life-cycle embodied and operational energy as well as capital cost as a result of change in building parameters.” Finally, Tingley et al (2015) is a study at the scale of construction materials in which LCA is used to compare three different insulation materials when applied in a typical dwelling. A snapshot of a selected number of LCA studies in the field of built environment is shown in Table 1.

One limitation in many comparative LCA studies is the limited number of variables and combinations, out of all possibilities, that are studied. This in turn results in incomprehensive conclusions with regard to optimized LCA. Because of the significant resources that would be needed to analyze all possible scenarios in a comprehensive LCA study, computational optimization techniques are utilized to address the challenge. To consider all possible combinations of design inputs and values, an optimization problem is started with identification of design inputs and proceeds with determining risks and constraints, finding the objective function, setting the minimum and maximum thresholds on design inputs, choosing an optimization algorithm, and eventually obtaining the results; i.e., the optimum solution to optimization problem (Deb 2012).

The present article tries to use an intelligent optimization algorithm to explore optimum building envelope design combination in a low-rise office building from operational energy use and environmental life-cycle impacts perspectives. The design inputs of interest include wall thermal resistance (R-value), insulation material, glazing type, window-to-wall ratio (WWR) in north and south facades, and frame material. Using these variables, we attempt to find the design combination that yields the lowest operational energy use and the least environmental impact. To achieve the objective, an integrated energy and environmental life-cycle assessment methodology is used to quantify the environmental impacts associated with each design combination. The results are then used in a hybrid artificial neural network and genetic algorithm-based approach to identify the optimum design combination.

2. 2. Methods A two-phase research methodology was pursued as illustrated in Figure 1. The first phase is LCA methodology followed by optimization efforts.

Figure 1. Research Methodology

2.1. LCA Methodology LCA is a technique used to assess the environmental impacts associated with products, projects, processes, and programs (Heijungs and Suh 2002). The quantitative technique uses a four-step methodology based on requirements prescribed by ISO 14040 (2006): goal and scope definition, inventory modeling, impact assessment, and interpretation of results.

Goal and Scope Definition: The first step in LCA methodology is mainly about establishing the functional unit, reference flows, system boundary, and the impact categories to be covered. The functional unit (FU) is a measure for the function of the LCA subject and a reference to be used for assessing the reference flows. FU in the present study is one square foot of a vertical building envelope (i.e., walls and windows) that encloses a hypothetical two-storey office building in Seattle, US, with 3600.0 square feet (~335 square meters) of floor area and a service life of 60 years, which represents a typical small size office building in Seattle. The envelope’s thermal characteristic per FU was considered to provide at least an R-value of 13.5 hr.ft2.°F/Btu (2.36 m2·K/W), for the entire envelope’s cross-section including the insulation

material. While the Seattle’s energy code for commercial buildings prescribes a minimum Rvalue of 17.5, we used a lower threshold for FU because of exploratory purpose of this research. As shown in Table 2, the wall in the building envelope consists of brick veneer, air gap, insulation, vapor barrier, concrete block, and gypsum plaster board. Window system consists of clear glass panes, frame materials, and cavity gas between the panes of glass. Window system consists of low-e clear glass panes, frame materials, and 1.25-centimeter (0.5-inch) air gap between the panes of glass. The window choices used for energy modeling include low-e clear double-glazed window (glass U-factor= 1.65 W/m2·K; solar heat gain coefficient (SHGC) of 0.42) and low-e clear triple-glazed window (glass U-factor= 1.25 W/m2·K; SHGC of 0.31). Service life of materials are considered based on the common practice in industry which is used by the LCA software, Athena Impact Estimator (IE), and is reported in Azari (2014). Several design characteristics of the building envelope of interest are manipulated in this study in order to examine the changes of impacts in relation to those of a base design combination. The varying design inputs include WWR in south façade, WWR in north façade,

and external wall’s thermal resistance (R-value). Table 3 shows design inputs along with their values, as examined in this research.

Table 2. Construction details of the wall Wall Split-faced brick veneer - 10 cm (4”) Air gap - 5 cm (2”) Insulation (varying materials and characteristics) Vapor barrier (polyethylene) Concrete brick - 10 cm (4”) Gypsum plasterboard -1.25 cm (0.5”)

Table 3. Building envelope design inputs and their corresponding values Design Inputs South WWR (%) North WWR (%) Wall’s thermal resistance (R-value) Insulation material Window type Window frame material

hr.ft2.°F/Btu m2·K/W

Potential Values 10, 20, 30, 40, 50, 60 10, 20, 30, 40, 50, 60 11, 13, 15, 17, 19, 21 1.93, 2.28, 2.64, 2.99, 3.34, 3.69 batt, batt+fiberboard, batt+polystrene double-glazed, triple-glazed wood vinyl, aluminum, fiberglass/vinyl

The reference flow, another element in goal and scope definition, is about quantification of materials and energy needed per design combination. Moreover, the system boundary shows what is included in or excluded from the LCA scope. The system boundary in the present study includes all stages of building envelope life-cycle (from material extraction through manufacturing, to construction, operation and maintenance, to demolition and recovery) as well as associated transportation. The environmental impact categories of interest included global warming, acidification, eutrophication, smog formation, and fossil-fuel energy consumption.

Inventory Modeling: In the next step of LCA methodology, the environmental flows (inputs and outputs) associated with the building envelope life cycle should be identified and quantified. Given the number of envelope design inputs and their corresponding values assumed in this research, the total number of possible design combinations of building envelope was too many to be modeled considering the limited resources available to us. Optimization techniques was a reasonable methodological alternative in order to draw conclusions based on limited number of modeling runs. We considered only 91 random design combinations (called, the initial scenarios) for inventory modeling and impact assessment and then used the results in the next phase of research, to be explained in section 2.2, in a hybrid genetic algorithm (GA) and artificial neural networks (ANN) approach to develop future populations/generations of combinations and find the optimum design combination. We also made sure that diversity was present in the initial population of 91 design combinations used for inventory modeling and impact assessment. Inventory modeling and impact assessment were conducted using Athena Impact Estimator which uses the US Environmental Protection Agency’s (EPA) Tool for Reduction and Assessment of Chemicals and other environmental Impacts (TRACI) methodology for life cycle impact assessment (Athena IE 2014). While Athena IE incorporates a large useful inventory database, it is not designed to model the operational energy use associated with design combinations. To feed it with this piece of data, we relied on eQuest 3.65 (Hirsch 2009), an energy performance modeling and simulation tool which can estimate the energy consumed in the building for heating, cooling, lighting, etc. We modeled the building in eQuest 3.65 and simulated its energy use for all design combinations. For activity-, schedule- and system-related assumptions, we relied on the software’s built-in default assumptions which represent compliance with the energy code requirements. The energy simulation results were then used in the LCA tool, Athena IE, along

with the other building design and construction information. Athena IE used this information to generate the inventory results, i.e. the environmental inputs and outputs/emissions, associated with design combinations. It is important to note that the LCA’s computational approach to inventory modeling introduced by Heijungs and Suh (2002) constructs the inventory vector based on the processes in the system boundaries and the environmental inputs and outputs to those processes. The approach uses the equation below for this purpose: Equation 1: [g] = [B] * [A] -1 * [f] where g is inventory vector, B intervention matrix, A-1 inverse matrix of technology matrix, and f is final demand vector. The Athena Impact Estimator (IE) uses LCA methodology and a regionalized life-cycle inventory database in compliance with ISO 14040/14044 (Athena 2014). The software allows for specification of cities based on which the inventory data are regionalized, i.e. regional differences with regard to transportation modes and distances, electricity grid, and manufacturing technology are taken into account (Athena 2014). While the developers of the software do not provide access to its inventory database or details about the specific electricity or transportation assumptions, it is expected that hydroelectric power is considered to be the source of Seattle’s electricity generation in the software’s inventory database, as suggested by the US Energy Information Administration (2015).

Impact Assessment: Impact assessment is the LCA activity that tries to classify the environmental inputs and outputs for each design combination based on their impacts on the environment. The type of impacts we were interested in this research included global warming, acidification, eutrophication, smog formation, and ozone depletion as some of the most widely

considered impact categories in the literature. Sometimes, an environmental emission could fall into more than one impact category because of its multiple effects on the environment. Contribution of each environmental emission (methane, for instance) to each impact category (e.g., global warming) is quantified by multiplying its quantity into an impact-specific science-based factor – characterization factor – that represents relative importance of the emission of interest in that category of impacts. In other words, characterization factors convert the quantity of environmental emissions into measures that represent impact categories. The measure for global warming potential (GWP), as an example, is equivalent kilogram of CO2. This step of impact assessment is called characterization and its outcome is quantified contribution of the LCA subject, FU of building envelope, to the impact categories of interest; i.e., global warming potential (GWP), acidification potential (AP), eutrophication potential (EP), smog formation potential (SFP), and ozone depletion potential (ODP). Table 4 includes the information about design inputs and environmental outputs per FU for 91 design combinations. The top and bottom 5 percent values for environmental outputs in the Table 4 are highlighted red and green, respectively.

Interpretation of Results: The last step in LCA methodology, based on ISO 14040 (2006), is interpretation of results through which the results, limitations and recommendations are presented. We present the results following the second phase of research, optimization efforts.

Table 4. Impact assessment results.

Design Inputs

OE (MJ)

GWP (kg CO2 eq)

AP (kg SO2 eq)

ODP (kg CFC-11 eq)

EP (kg N eq)

SFP (kg O3 eq)

2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 1.93 2.28 2.99 3.34 3.69 1.93 1.93 1.93 1.93 1.93 1.93 1.93 1.93 1.93 1.93 2.28 2.28 2.28 2.28 2.28 2.28 2.28 2.28 2.28 2.28 2.99 2.99 2.99 2.99 2.99 2.99 2.99 2.99 2.99 2.99 3.34 3.34 3.34 3.34 3.34 3.34

Glazing Type

m2·K/W

15 15 15 15 15 15 15 15 15 15 15 15 15 15 11 13 17 19 21 11 11 11 11 11 11 11 11 11 11 13 13 13 13 13 13 13 13 13 13 17 17 17 17 17 17 17 17 17 17 19 19 19 19 19 19

North WWR

hr.ft2.°F/Btu

Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Vinyl Clad Wood Aluminum Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl

South WWR

Window Frame Material

Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt Batt Batt + Polystyrene Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt + Polystyrene Batt + Polystyrene Batt + Polystyrene Batt + Polystyrene Batt + Polystyrene Batt + Polystyrene Batt + Polystyrene Batt + Polystyrene Batt + Polystyrene Batt + Polystyrene Batt Batt Batt Batt Batt Batt

Wall R-Value

Insulation Material

Design Combination 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 49 50 51 52 53 54 55

Environmental Outputs per FU

40% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 10%

10% 20% 30% 40% 50% 60% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40%

DG DG DG DG DG DG DG DG DG DG DG TG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG DG

192.305 193.263 194.755 195.340 196.500 198.120 199.623 197.599 196.418 194.672 194.705 190.490 198.816 201.870 200.025 198.761 192.743 197.778 197.077 199.598 200.630 201.764 204.095 205.719 207.384 205.213 203.968 202.174 202.183 198.095 199.216 200.420 202.906 204.616 205.920 203.836 202.643 200.980 201.048 190.931 192.683 194.236 197.864 199.528 199.178 197.428 196.532 195.419 195.752 196.986 198.153 199.413 202.016 203.804 204.822

1885.933 1870.808 1866.435 1851.359 1844.278 1845.475 1888.327 1871.089 1861.763 1844.502 1846.653 1837.623 1853.104 1871.088 1869.244 1864.722 1843.507 1862.335 1589.515 1916.887 1902.228 1893.235 1877.479 1878.514 1921.585 1903.986 1894.422 1877.170 1879.423 1911.475 1897.153 1888.232 1873.054 1874.594 1915.958 1898.864 1889.492 1872.913 1875.381 1884.648 1874.651 1866.995 1859.837 1857.148 1891.624 1876.105 1868.162 1854.102 1857.518 1909.015 1894.789 1886.014 1871.125 1872.882 1913.355

6.521 6.544 6.583 6.594 6.623 6.666 6.756 6.683 6.637 6.564 6.556 6.492 6.607 6.741 6.743 6.703 6.513 6.675 4.358 6.753 6.778 6.806 6.864 6.907 7.003 6.925 6.877 6.802 6.794 6.705 6.733 6.764 6.827 6.872 6.956 6.882 6.835 6.765 6.758 6.480 6.528 6.569 6.668 6.712 6.744 6.680 6.643 6.590 6.591 6.674 6.703 6.735 6.802 6.850 6.926

8.21E-07 8.39E-07 8.57E-07 8.74E-07 8.92E-07 9.09E-07 8.21E-07 8.39E-07 8.57E-07 8.92E-07 9.09E-07 1.10E-06 8.74E-07 1.01E-06 8.69E-07 8.72E-07 8.72E-07 8.72E-07 8.71E-07 8.16E-07 8.33E-07 8.51E-07 8.87E-07 9.05E-07 8.16E-07 8.33E-07 8.51E-07 8.87E-07 9.05E-07 8.19E-07 8.36E-07 8.54E-07 8.89E-07 9.07E-07 8.19E-07 8.36E-07 8.54E-07 8.89E-07 9.07E-07 8.19E-07 8.36E-07 8.54E-07 8.89E-07 9.07E-07 8.19E-07 8.36E-07 8.54E-07 8.89E-07 9.07E-07 8.19E-07 8.36E-07 8.54E-07 8.89E-07 9.07E-07 8.19E-07

0.083 0.085 0.086 0.087 0.089 0.090 0.086 0.086 0.087 0.088 0.089 0.097 0.088 0.100 0.089 0.088 0.087 0.088 0.066 0.085 0.087 0.088 0.091 0.092 0.088 0.088 0.089 0.090 0.091 0.085 0.086 0.088 0.090 0.092 0.088 0.088 0.089 0.090 0.091 0.083 0.084 0.086 0.089 0.090 0.086 0.086 0.087 0.088 0.089 0.085 0.086 0.088 0.090 0.092 0.088

19.423 19.312 19.218 19.073 18.957 18.858 20.081 19.705 19.383 18.790 18.544 19.109 19.132 19.587 19.434 19.339 18.930 19.370 13.598 19.980 19.876 19.764 19.538 19.441 20.675 20.287 19.961 19.364 19.117 19.867 19.769 19.663 19.449 19.358 20.565 20.184 19.861 19.274 19.031 19.384 19.326 19.243 19.101 19.014 20.113 19.753 19.449 18.896 18.670 19.907 19.806 19.698 19.481 19.390 20.606

56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91

Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt + Fiberboard Batt Batt Batt+ Polystyrene Batt Batt Batt Batt Batt+ Polystyrene Batt Batt

Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Fiberglass/Vinyl Aluminum Wood Vinyl Aluminum Aluminum Aluminum Aluminum Aluminum Wood Vinyl Wood Vinyl Wood Vinyl Wood Vinyl Wood Vinyl

19 19 19 19 21 21 21 21 21 21 21 21 21 21 15 15 15 15 15 15 15 15 15 15 15 15 11 13 17 19 21 11 13 17 19 21

3.34 3.34 3.34 3.34 3.69 3.69 3.69 3.69 3.69 3.69 3.69 3.69 3.69 3.69 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 2.64 1.93 2.28 2.99 3.34 3.69 1.93 2.28 2.99 3.34 3.69

20% 30% 50% 60% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40%

40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 10% 20% 30% 50% 60% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40% 40%

DG DG DG DG DG DG DG DG DG DG DG DG DG DG TG TG TG TG TG TG TG TG TG TG TG TG DG DG DG DG DG DG DG DG DG DG

202.815 201.657 200.083 200.213 196.154 197.392 198.693 201.366 203.163 204.048 202.074 200.955 199.415 199.595 188.881 189.311 189.882 191.232 192.379 195.933 193.491 191.944 189.531 189.208 194.052 190.807 206.384 205.133 199.351 204.205 203.541 203.341 202.090 196.237 201.168 199.350

1896.479 1887.346 1870.960 1873.765 1906.575 1892.592 1883.938 1869.268 1870.858 1911.134 1894.355 1885.344 1869.030 1872.005 1875.787 1859.390 1848.255 1829.493 1829.085 1877.904 1859.448 1848.906 1830.562 1831.908 1858.428 1838.371 1903.121 1898.648 1878.508 1896.647 1890.815 1885.550 1881.077 1860.669 1879.100 1888.159

6.853 6.807 6.740 6.734 6.649 6.680 6.714 6.782 6.830 6.902 6.831 6.786 6.719 6.716 6.451 6.462 6.476 6.512 6.544 6.678 6.596 6.543 6.457 6.442 6.654 6.500 6.976 6.936 6.754 6.910 6.860 6.845 6.806 6.622 6.780 6.851

8.36E-07 8.54E-07 8.89E-07 9.07E-07 8.19E-07 8.36E-07 8.54E-07 8.89E-07 9.07E-07 8.19E-07 8.36E-07 8.54E-07 8.89E-07 9.07E-07 9.97E-07 1.03E-06 1.07E-06 1.14E-06 1.18E-06 9.97E-07 1.03E-06 1.07E-06 1.14E-06 1.18E-06 1.31E-06 1.10E-06 1.00E-06 1.01E-06 1.01E-06 1.01E-06 8.72E-07 8.69E-07 8.71E-07 8.71E-07 8.71E-07 8.71E-07

0.088 0.088 0.090 0.091 0.085 0.086 0.087 0.090 0.092 0.087 0.088 0.088 0.090 0.090 0.091 0.093 0.095 0.099 0.102 0.093 0.094 0.096 0.099 0.100 0.116 0.098 0.102 0.102 0.100 0.102 0.090 0.090 0.090 0.088 0.090 0.090

20.225 19.898 19.306 19.061 19.882 19.785 19.678 19.463 19.371 20.587 20.206 19.880 19.286 19.043 19.467 19.349 19.228 18.994 18.893 20.102 19.729 19.413 18.836 18.597 19.750 19.167 20.152 20.057 19.667 20.093 19.849 19.706 19.611 19.215 19.647 19.858

2.2. Optimization The next step in research was optimization. Several attempts have been made to classify different optimization techniques. One popular classification proposed by Cohon and Marks (1975) classifies the optimization techniques to a priori (such as global criterion method, etc.), a posteriori (such as ε-constraint, etc.) and progressive preference articulation techniques based on the way that search and decision-making problems are handled by each method. The evolutionary algorithm method (such as genetic algorithm, etc.) that is used in our research is a different kind of approach based on the theory that non-dominated solutions are chosen to remain in a population as it evolves (Chiandussi et al 2012).

The LCA results for 91 design combinations were entered into the optimization phase with the objective of identifying the optimum values for design inputs. We pursued a two-step optimization algorithm, as explained below.

Step 1: Modeling by Artificial Neural Networks (ANNs) ANNs are effective computational frameworks that imitate the complex relationships of biological networks to solve intricate nonlinear problems, especially where classical mathematical modeling processes fail to succeed (Braspenning et al 1995, Gurney 2005, Haykin 1998). In the present study, separate ANNs were used to model and optimize operational energy and environmental impacts. The process, which is the same for all outputs, is explained below using Eutrophication Potential (EP), as a random example. To model EP by an ANN, a four-stage methodology was followed. In the first stage, the normalization process was conducted on both design combination data (input data) and EP data (targets). The data normalization process minimizes the influence of magnitude and range of variations of the input variables throughout the ANN training process. Generally, a linear transformation is applied to normalize training and test data sets. To find the optimum parameters of an ANN including weights and biases, the initial data were divided into two subsets, i.e. training and test data sets. Data division, the second stage of ANN, was conducted through random selection. In this stage, 70 percent of the normalized data were utilized to train the ANN and optimize weights and biases. 30 percent of the data that was not considered in the training process was used for validation and testing process. The third stage of ANN methodology is definition of ANN architecture, including the number of layers and the number of neurons in each layer. A five-layer ANN was defined with

5-3-4-2-1 neurons in hidden and output layers. The number of neurons in the input layer is equal with the number of input variables, i.e. 6 neurons. The architecture of proposed ANN is illustrated in Figure 2.

Figure 2. Architecture of proposed ANN Logistic sigmoid, hyperbolic tangent sigmoid, and linear transfer functions are the most commonly used activation functions in the optimization of ANNs. Here, the hyperbolic tangent sigmoid was utilized as the activation function for the hidden and output layers to scale outputs of each neuron within the range of -1 to +1. The last stage in modeling the data by an ANN is to define and apply an appropriate training algorithm. In other words, ANN parameters (including weights and biases) should be adjusted and optimized to minimize the generated errors for both training and test data sets. Indeed, the performance of an ANN is evaluated via the calculation of the mean squared errors (MSEs) for both training and test data sets. These values are unique signature of the network’s

performance and defined as the average of the sum of the differences between the targets and the ANN predictions:

Equation 2:

MSE =

1 N

∑ (y

2

N

i =1

i

− y i ,t arg et )

where N, yi, and yi,target are number of data for training or test processes, the ANN output and target value for the ith experiment, respectively.

In the present study, an evolutionary optimization algorithm was used to train the defined ANNs and precisely adjust and optimize their weights and biases. A computer code was developed capable of handling a predefined ANN and stochastically computing and readjusting its weights and biases applying genetic algorithm evolutionary method. For this purpose, the network’s unknown/adjustable parameters including 67 weights and 15 biases were codified into a chromosome composed of 82 genes. Each gene within the defined chromosome can take a value between -1 and +1. So, the developed computer code was trained to scan and explore chromosomes by means of genetic algorithm and find the best chromosome results with minimum MSEs for both training and test data sets. Hence, an initial population of these chromosomes was randomly generated. For each chromosome, e.g. chromosome number j, the genes were fed into the predefined ANN structure. Then, the training data set was used to evaluate the performance of the fed ANN. The calculated MSE was considered as the fitness of chromosome number j. This training process is repeated for all chromosomes separately. Then, the chromosomes are sorted based on the calculated training MSEs. The best fitness obviously belongs to a chromosome that results in a minimum

MSE. Applying selection, mating, crossover, and mutation operators, the new generation of chromosomes is produced and sent to the evaluation unit. This optimization process is repeated till a preset training MSE is established. Then, the trained ANN (the ANN filled with the genes of the best chromosome) is evaluated via presenting test data set and calculating test MSE. Final solution occurs when a test MSE of less then or equal to a preset test MSE is achieved. Otherwise, the algorithm returns to the third stage and is repeated. The genetic algorithm parameters used in training process of ANNs are presented in Table 5. Moreover, Figure 3 illustrates the flow chart applied for training and test processes of defined ANNs. Table 5: Genetic algorithm parameters used in ANN training Optimization Parameter Initial population size Selection operator Mating operator Crossover operator Mutation rate Training and test errors* *

Value 100 Killing off the worst 50% of chromosomes Tournament selection Single-point crossover 20% of new generation 1% and 2% respectively

In case of Operational Energy (OE) the training and test errors are set to be 3% and 5%, respectively.

Figure 3: The flow chart describing the genetic algorithm based training of ANNs

Based on the five-stage computational algorithm and the flow chart presented in Figure 3, a computer code was developed in Pascal programming language (Lazarus 1.2.4 IDE) and compiled into 64-bits executable using FPC 2.6.2. A Mersenne Twister-based subroutine was used to produce the required random numbers for the modeling processes (Matsumoto and Nishimura 1998). The random number generation subroutine satisfies the tests of uniformity and serial correlation with high resolution. The cycle length of the random number generator was 219937-1.

Step 2: Multi-objective Optimization by Genetic Algorithm (GA) Genetic Algorithms (GAs) are effective stochastic optimization tools used in different fields of science and technology (Haupt and Haunt 2004, Gen and Cheng 1997). Each genetic algorithm process is initiated by randomly generating a population of predefined chromosomes. A chromosome, as a well-structured string, is made up of tightly connected genes and carries the genetic information. Each chromosome as a genotype is related to a phenotype as its corresponding physical object. Genetic algorithm, as an evolutionary artificial intelligence technique, applies genetic operators, such as mutation and crossover, to the best members from the previous generation in order to establish the new generation. These operators stochastically adjust/control the gene(s) values to help the population evolve and produce optimum chromosomes. Mutation operator, as an exploration tool, is utilized to randomly introduce new information from the search space (i.e., gene pool) and to move the solution(s) towards the global optimum. On the other hand, crossover operator, as an exploitation tool, combines information from existing chromosomes and conducts the solution(s) towards local optima.

In the present study, genetic algorithm was applied to determine the appropriate design inputs that lead to optimum operational energy and environmental impacts. To formulate the genetic algorithm, the design inputs are codified into a chromosome depicted in Figure 4.

Insulation Material

Frame Material

Wall RSouth Glazing North Type Value WWR WWR Figure 4: Schematic representation of the chromosome consists of design inputs

An initial population of the chromosomes is randomly produced. The fitness of each generated chromosome is calculated by recalling the trained ANNs of the first optimization step. Each chromosome is identified by six fitness values (objectives) including one for operational energy and five for environmental impacts. So, for each chromosome all six trained ANNs were separately recalled to calculate the corresponding fitness. To do this, the genes of each chromosome were fed into the input layer of the previously trained ANNs and the outputs of the ANNs were reported as the fitness values for that chromosome. Then, the chromosomes were sorted based on predefined target(s). If the optimization target was a single-objective optimization that minimized only one of the environmental outputs, e.g. the Eutrophication Potential (EP), then the chromosomes were totally ordered according to the EP fitness values calculated using the corresponding ANN for each chromosome. On the other hand, in case of a multi-objective optimization that simultaneously minimizes two or more targets/objectives; e.g., operational energy and/or other environmental impacts altogether, the Non-dominated Sorting Genetic Algorithm II (NSGA-II) is used (Coello et al 2002). Unlike classical multi-objective optimization methods, the evolutionary algorithms especially multi-objective versions of genetic algorithms are much more effective and faster optimization tools. The main difference between

NSGA-II and conventional GA is the sorting mechanism. Indeed, the chromosomes in NSGA-II are sorted based on the concept of domination. In general, the chromosome x dominates the chromosome y if it is not worse than the chromosome y for all objectives and strictly better than it for at least one objective:

Equation 3:

∀i , ∃j ,

f i (x ) ≤ f i ( y )

f j (x ) < f j ( y )

where, fi(x) is the fitness value of the chromosome x respect to objective i. It is clear that i and j are representatives of environmental outputs, i.e. OE, GWP, AP, ODP, EP, and SFP. Having compared all possible pairs of population, the NSGA-II inspects and evaluates the chromosomes based on the number of non-dominations and assigns a number/rank to each chromosome accordingly. So, the chromosomes are precisely classified and positioned within a set of fronts. The first front consists of completely non-dominated set of chromosomes in the current population and the second front comprise of chromosomes are dominated by the members in the first front only and the front goes on. Hence, the chromosomes in the first front are given a rank value of 1 and those in the second are assigned the rank value of 2 and so on. Then, the chromosomes within each front are sorted according to their crowding distance. The crowding distance is a measure of how close a chromosome is to its neighbors. The chromosome located in the lesser crowded region, i.e. the region with large average crowding distance, results in better diversity in the population and is more preferred. In the next stage, the selection, mating, crossover, and mutation operators were applied on the sorted population to select and reproduce new generation. The adjustable genetic

algorithm parameters applied in this optimization step are listed in Table 6. Also, the flow chart of this optimization process is shown in Figure 5.

Table 6: Main controlling parameters of the applied genetic algorithm Optimization Parameter

Value

Initial population size Selection operator Mating operator Crossover operator Mutation rate

50 Merge, Sort, and Truncate Roulette wheel selection Single-point crossover 5% of new generation

Figure 5: Genetic Algorithm flow chart

3. Results The MSE variations of the best chromosome in training the data for single-objective optimization of the data revealed a decline of errors as the number of iterations increased. The Figure 6, which illustrates the MSE variations in training the EP data, shows a decline from 5.67% to 1.00% within 946 iterations. Relatively large number of epochs (iterations) and small errors shows that the developed ANN model is able to properly predict the behavior of the data. Table 7 lists the statistical details of the ANN model training and testing for the entire data.

6

Training Error (%)

5 4 3 2 1 0

100 200 300 400 500 600 700 800 900 1000

Iteration (-) Figure 6: Iteration-dependent error variations of the ANN model for optimization of the responding variable; EP in this case.

Table 7: Statistical details of the ANN model training and testing in single-objective optimization. Response Number of epochs Training MSE Test MSE Training Error (%) Test Error (%) Max Training Error (%) Max Test Error (%) Correlation Coefficient Coefficient of Efficiency Goodness of Fit (%) Coefficient of Determination

SFP 8720 0.000399 0.000672 0.999794 1.297002 3.637830 (85) 2.714232 (15) 0.978599 0.957123 79.293394 0.957656

EP 946 0.000399 0.000243 0.999186 0.780105 2.722563 (2) 1.758720 (59) 0.981646 0.961080 80.271999 0.963630

ODP 171 0.000399 0.000249 0.999707 0.790548 2.134754 (24) 1.784440 (41) 0.992016 0.983467 87.141996 0.984097

AP 5485 0.000398 0.001154 0.998365 1.698941 3.848782 (14) 4.561337 (16) 0.969047 0.938673 75.235793 0.939053

GWP 10281 0.000399 0.001279 0.999628 1.788496 4.427776 (17) 0.968478 0.937833 75.066775 0.937949

OE 17224 0.009998 0.009718 4.999595 4.929027 14.881263 (18) 15.275572 (74) 0.931649 0.866035 63.398793 0.867970

Number of epochs in Table 7 indicates the number of times that the genetic algorithm is repeated to learn successfully how to foresee the response variable (EP, for instance) through the developed network. Reasonably low training and test MSE values are indicators of how well the models explain the data. Table 7 also reports other statistical parameters such as correlation coefficient, coefficient of efficiency, goodness of fit, and coefficient of determination for each environmental output. Coefficient of determination, for instance, shows the proportion of variations in the response variables (EP, for instance) that can be explained by the model.

The performance of the developed ANN model was further checked through comparison of predicted values of each response variable with actual experimental data. Figure 7 verifies the accuracy of ANN in prediction of all environmental impacts as response variables.

Figure 7. Relationship between the experimental data and the optimized ANN model outputs for OE, GWP, AP, SFP, ODP and EP

The single-objective optimization results shown in Table 8 reveal that the operational energy (OE) consumption is minimized in a design scenario with about 60% south WWR, about 10% north WWR, and the use of triple-glazed windows, fiberglass window frame, R-17 wall insulation, and FG batt and polystyrene insulations. Compared with the optimum scenario for OE, the GWP-optimized design scenario incorporates lesser R-value (R-13), higher north WWR (about 60%; which in turn results in lesser use of insulation in building skin), and using wood vinyl as window frame material. Among the optimized scenarios in Table 8, only GWP- and

SFP-optimized design scenarios entail relatively large WWR percentages in both south and north. This in turn means lesser use of insulation mass in building skin defined for these scenarios. Hydrofluorocarbon (HFC) blowing agents used in extruded polystyrene production make significant contribution to global warming and smog formation and the low GWP and SFP in design scenarios with lesser use of insulation can be associated with the impact of HFCs. HFCs, however, do not contain chlorine and have essentially zero ozone depletion potential. This means that the mass of insulation doesn’t have any effect on the ODP of scenarios in Table 8. On the other hand, the design scenarios with lowest WWR percentages result in minimum ODP and EP. In other words, the lowest potentials for ozone depletion and eutrophication occur in cases with the lowest mass of glass in building skin. Table 8. Results of single-objective optimization. Design inputs and environmental outputs associated with optimum (minimum) scenarios. Min OE case

Min GWP case

Min AP case

Min ODP case

Min EP case

FG Batt+ Ext. Polystyrene Fiberglass 21 3.69 10 10 Double Glazing

FG Batt+ Ext. Polystyrene Fiberglass 21 3.69 10 10 Double Glazing

Min SFP case

Design Inputs Insulation Material Window Frame Material hr.ft2.°F/Btu Wall R-value m2·K/W South WWR (%) North WWR (%) Glazing Type

FG Batt+ Ext. Polystyrene Fiberglass 17 2.99 60 10 Triple Glazing

FG Batt+ Ext. Polystyrene Wood Vinyl 13 2.28 60 60 Triple Glazing

FG Batt+ Ext. Polystyrene Aluminum 11 1.93 59 10 Triple Glazing

191.123

192.119

191.884

1804.6700 6.6764 9.0704E-07 0.0898 18.2865

1695.6097 6.6290 9.1478E-07 0.0910 17.9231

1819.9874 6.0070 8.9582E-07 0.0869 18.3499

FG Batt Wood Vinyl 11 1.93 60 50 Triple Glazing

Environmental Outputs OE (MJ) GWP (kg CO2-eq) AP (kg SO2-eq) ODP (kg CFC-11-eq) EP (kg N-eq) SFP (kg O3-eq)

197.894

1912.5770 6.71600 8.2400E-07 0.0852 20.3440

197.893

195.434

1912.5771 6.7159 8.2405E-07 0.0852 20.3440

1746.4217 6.5406 9.5233E-07 0.0920 17.6045

The result for multi-objective optimization is included in Table 9. As the table shows, the optimum design scenario that yields the optimum low environmental impacts is the one that incorporates triple-glazed windows with fiberglass frame on south elevation as about 60% WWR

and on north elevation as about 10% WWR. Interestingly, this optimum design scenario does not contain the highest R-value, but R-17. The is in line with diminishing return law based on which increasing R-value above a certain optimum point does not have a major impact on improving the building performance.

Table 9. Results of multi-objective optimization showing the characteristics of optimum scenario

Design Inputs Insulation Material Window Frame Material hr.ft2.°F/Btu Wall R-value m2·K/W South WWR (%) North WWR (%) Glazing Type

FG Batt+ Ext. Polystyrene Fiberglass 17 2.99 60 10 Triple Glazing

Environmental Outputs OE (MJ) GWP (kg CO2-eq) AP (kg SO2-eq) ODP (kg CFC-11-eq) EP (kg N-eq) SFP (kg O3-eq)

191.123 1804.6700 6.6764 9.0704E-07 0.0898 18.2865

Another observation that needs to be verified with other data is that the optimum design combination for all environmental outputs is the same as design combination that yields the least operational energy use. If verified by other data and future studies, this can imply that architects wishing to lower the life-cycle environmental impacts of the building envelopes should focus on reducing operational energy of building more than any other impact category. In real practice and in the context of global environmental challenges, however, some environmental impacts such as global warming pose more immediate threats to the environment compared with other impacts. In this sense, a minimum-GWP design combination currently can be more important than an optimum design combination in achieving an eco-friendly building. We emphasize that the results of this research should be interpreted with caution because there are many other

design-related and location-related factors not captured in this study that can affect the environmental impacts of building envelopes.

Conclusion While significant research has been conducted on the impact of building envelopes on energy performance of buildings, the knowledge is still relatively vague about their life-cycle environmental impacts. In the present research, we considered 6 basic design characteristics of building envelopes and used quantitative and simulation-based techniques to explore combinations of 6 design characteristics that yield optimum environmental impacts in 6 categories of operational energy, global warming potential, acidification potential, ozone depletion potential, eutrophication potential, and smog formation potential. The results of this research once validated by further study in this area by other researchers have the potential to address the architecture and construction research community’s need to design guidelines that reduce the operational energy consumption and environmental impacts simultaneously. For this purpose, the methodology that we used in this research can be followed to develop design guidelines. More than its results, the significance of our research is in the methodology that we applied. A future direction of the research in this area could be to develop a software tool that can help conduct the entire methodology within one tool. Such tools need to include all design and construction inputs beyond what we studied in our research. Also, it is important to note that we treated all design inputs as well as environmental outputs equally with regard to their significance in design decision-making. In real-life situations, however, architects may assign more weight to some design inputs or environmental impacts because of design priorities, or in order to overcome the conflicting effects of inputs or impacts, so future

methodologies and tools should provide capability to assign weight to factors in their decision support mechanism. We also encourage future research to address some of the limitations we faced in doing this study. This study focused on Seattle as the location of project, mainly because the city is among the few cities that the version of Athena Impact Estimator that was used in this research supported with regard to inventory modeling. Future research could use other LCA software, such as SimaPro, to cross-examine the results of the present research in the context of other geographical locations. Indeed, different locations enjoy different climates potentially, which would affect energy performance of buildings. Various geographical locations also incorporate different manufacturing practices, electricity generation and distribution, transportation and logistic systems, and consuming patterns, which all would affect the environmental life-cycle impacts of buildings and their components. Future research can also expand the scope of this study and examine more design characteristics or other building components and systems. Finally, future research should also focus on development of architect-friendly software and tools that can work as plugin to architectural modeling tools to optimize the environmental impacts of design.

References Asdrubali, A., Baldassarri, C., Fthenakis, V. 2013. Life cycle analysis in the construction sector: Guiding the optimization of conventional Italian buildings. Energy and Buildings, 64:73-89. Athena IE (2014). Athena Impact Estimator (IE) User Guide. Athena Sustainable Materials Institute. Available from: http://calculatelca.com/wpcontent/uploads/2014/10/IE4B_v5_User_Guide_September_2014.pdf Azari, R. 2014. Integrated energy and environmental life-cycle assessment of office building envelopes. Energy and Buildings, 82: 156-162. Azari, R. and Kim, Y.W. 2012. Comparative assessment of life cycle impacts of curtain wall mullions. Building and Environment, 48: 135-145. Braspenning, P.J., Thuijsman, F., Weijters, A.J.M.M. 1995. Artificial Neural Networks: An Introduction to ANN Theory and Practice”, Springer, Berlin, Germany. Chiandussi, G., Codegone, M., Ferrero, S., Varesio, F.E. (2012). Comparison of multi-objective optimization methodologies for engineering applications. Computers and Mathematics with Applications 63. 912-942.

Coello, C.A., Van Veldhuizen, D.A., Lamont, G.B. 2002. Evolutionary Algorithms for Solving MultiObjective Problems, Kluwer Academic Publishers, New York. Cohon, D.H. Marks. 1975. A review and evaluation of multiobjective programming techniques, Water Resour. Res. 11, 208–220. Deb, K. 2012. Optimization for Engineering Design – Algorithms and Examples. PHI Publishing. Dodoo, A., Gustavsson, L., Sathre, R. 2014. Lifecycle carbon implications of conventional and lowenergy multi-storey timber building systems. Energy and Buildings, 82: 194-210. EIA 2015. Monthly energy review. Energy Information Administration. http://www.eia.gov/totalenergy/data/monthly/pdf/sec2_3.pdf EN 15978. 2011. Sustainability of Construction Works. Assessment of Environmental Performance of Buildings – Calculation Method. European Committee for Standardization. Engelbrecht, A.P. 2007. Computational Intelligence: An Introduction. 2nd Edition”, John Wiley and Sons Inc. EPA 2009. Buildings and the Environment: a Statistical Summary. http://ww2.harford.edu/faculty/eaugusti/Environ%20Health/green%20building%20stats.pdf Gen, M., Cheng, R. 1997. Genetic Algorithms and Engineering Design. John Wiley and Sons Inc., 1997. Gurney, K. 2005. An Introduction to Neural Networks. Taylor & Francis. London, UK. Haupt, R.L., Haupt, S.E. 2004. Practical Genetic Algorithms. 2nd Edition, John Wiley & Sons Inc. Haykin, S. 1998. Neural Networks: A Comprehensive Foundation. Prentice Hall PTR Upper Saddle River, NJ, USA. Heijungs, R. and Suh, S. 2002. Computational Structure of Life Cycle Assessment. Kluwer Academic Publishers, The Netherlands. Hirsch, J. (2009). The Quick Energy Simulation Tool (eQuest) [software tool]. Available from: http://www.doe2.com/equest/ International Organization of Standards, ISO 14040. 2006. Environmental Management– Life Cycle Assessment – Principle and Framework, National Standard Authority of Ireland, Ireland. International Organization of Standards, ISO 14044. 2006. Environmental Management– Life Cycle Assessment – Requirements and Guidelines, National Standard Authority of Ireland, Ireland. International Organization of Standards, ISO 16745. 2015. Environmental Performance of Buildings – Carbon Metric of a Building, Use Stage, National Standard Authority of Ireland, Ireland. Junnila, S., and Horvath, A. 2003. Life-cycle environmental effects of an office Building. J. Infrastruct. Syst., 9(4), 157-166. Junnila, S., Horvath, A., Guggemos, A.A. 2006. Life-Cycle Assessment of Office Buildings in Europe and the United States. J. Infrastruct. Syst., 12(1), 10-17. Kosareo, L., Ries, R. 2007. Comparative environmental life cycle assessment of green roofs. Building and Environment, 42(7):2606-2613. Matsumoto, M., Nishimura, T. 1998. Mersenne twister: a623-dimensionally equidistributed uniform pseudo-random number generator”, ACM Transitions on Modeling and Computer Simulation. 8(1), 3-30. Ottele, M., Perini, K., Fraaij, A.L.A., Haas, E.M., Raiteri, R. 2011. Comparative life cycle analysis for green façades and living wall systems. Energy and Buildings, 43(12): 3419-3429. Pulselli, R.M., Simoncini, E., Marchettini, N. 2009. Energy and emergy based cost–benefit evaluation of building envelopes relative to geographical location and climate. Building and Environment, 44(5): 920-928. Svozil, D., Kvasnicka, V., Jiri, P. 1997. Introduction to multi-layer feed-forward neural networks”, Chemometrics and Intelligent Laboratory Systems. 39(1), 43-62. Stazi, F., Mastrucci, A., Munafo, P. (2012). Life cycle assessment approach for the optimization of sustainable building envelopes: An application on solar wall systems. Building and Environment, 58: 278-288. Tingley, D.D., Hathway, A., Davison, B. 2015. An environmental impact comparison of external wall insulation types. Building and Environment, 85: 182-189.

Treloar, G., Fay, R., Ilozor, B., and Love, P. 2001a. Building materials selection: Greenhouse strategies for built facilities. Facilities, 19(3/4), 139-149.
 Treloar, G., Fay, R., Ilozor, B., and Love, P. 2001b. "An analysis of the embodied energy of office buildings by height." Facilities, 19(5/6), 204-214. Yohanis, Y. G., and Norton, B. 2002. Life-cycle operational and embodied energy for a generic singlestorey office building in the U.K. Energy, 27, 77-92.

Table 1. An overview of some LCA literature Subject of LCA

Impact Categories

            

Azari & Kim (2012)





   

Dodoo et al (2014)





    

Junnila and Horvath (2003)



      

Junnila et al (2006)



      

     



    



     



Ottele et al (2011)



      



Pulselli et al (2009)



     

Stazi et al (2012)



   



Treloar et al (2001a)



 



Kosareo & Ries (2007)

Tingley et al (2015)

Acidification

     

Eutrophication



Energy Use

Azari (2014)

Global Warming



Smog Formation

Asdrubali et al (2013)

Human Toxicity

Other

Demolition & Recovery

Operation & Maintenance

Construction

Manufacturing

Material Extraction

Buildings

Building Envelope, Components, and Systems

Construction Materials

Reference

LCA stages

  

 

                  



Treloar et al (2001b) Yohanis and Norton (2002)