Preliminary assessment of greenhouse gas emissions ...

Building and Environment 78 (2014) 95e102

Contents lists available at ScienceDirect

Building and Environment journal homepage: www.elsevier.com/locate/buildenv

Preliminary assessment of greenhouse gas emissions for atactic polypropylene (APP) modified asphalt membrane roofs Warren Vaz*, John Sheffield Missouri University of Science and Technology, Mechanical and Aerospace Engineering, 194 Toomey Hall, 400 W. 13th St., Rolla, MO 65409, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 February 2014 Received in revised form 7 April 2014 Accepted 23 April 2014 Available online 5 May 2014

In recent years, there has been an increased awareness about the environmental impact of buildings and construction materials with increasing regulatory pressure to reduce this impact. Properly assessing this impact is the first step towards achieving a green construction sector. Asphalt roofs are an important component of the construction sector in the United States. GaBiÒ, a life cycle assessment (LCA) software package, was used to evaluate the greenhouse gas (GHG) emissions of an atactic polypropylene (APP) asphalt roofing product. Due to the complexity of the manufacturing process, an average approach where the raw materials and energy use to manufacture the product over an entire year were summed up and the results were normalized per functional unit of the product. Global warming potential (GWP) was used to evaluate the overall GHG emissions. The GHG emissions were estimated to be 75.2 kg CO2-eq per roll. The manufacturing stage was found to be responsible for a majority of the GHG emissions. Certain shortcomings in the LCA study were identified and suggestions to reduce GHG emissions were made. This study differs from previous studies on roofs in that it focuses on the environmental impact of roofing product itself, unlike previous studies that have investigated energy efficiencies or reduction of environmental impact as part of a larger study. Ó 2014 Elsevier Ltd. All rights reserved.

Keywords: Asphalt roof Atactic polypropylene (APP) Building GaBiÒ Greenhouse gas emissions Life cycle assessment

1. Introduction The governments of several countries have promulgated regulations aimed at reducing the environmental impact of buildings. In the United Kingdom, the Climate Change Act [1] was enacted to reduce carbon dioxide emissions by 80% of the 1990 level by 2050. To achieve this target, all new homes from 2016 onwards are required to be “zero carbon” homes [2]. In the United States, all federal buildings must achieve “zero-net-energy” by 2030 [3] and in the state of California, all major renovations and new state buildings must be zero-net-energy starting in 2025 [4]. Efforts to reduce the environmental impact of buildings through the reduction of energy consumption must be complemented by efforts to reduce the environmental impact associated with the production of materials used in the construction of buildings. Roofs have a significant effect on the overall environmental impact of buildings [5e12]. There are a few studies [13e16] that focus on the environmental impact caused by the manufacturing, utilization, and disposal of the roofing materials themselves. While

* Corresponding author. Tel.: þ1 573 341 4661; fax: þ1 573 341 4607. E-mail address: [email protected] (W. Vaz). http://dx.doi.org/10.1016/j.buildenv.2014.04.019 0360-1323/Ó 2014 Elsevier Ltd. All rights reserved.

there are studies on the environmental impact associated with certain types of building materials and products [14,17,18] including different kinds of asphalt roofing products [19e21], studies on atactic polypropylene (APP) modified asphalt membrane roofing products are lacking. Blom et al. [22] concluded that the greatest impact categories for climate systems are energy consumption (during operation) and material use (during manufacturing). Pajchrowskia et al. [23] concluded that the environmental impact of the production of building materials will be more significant as the energy demand during the lifetime of buildings, usually the life cycle stage with the greatest environmental impact, is increasingly met by renewable sources. With asphalt membrane roofs comprising about 15.1% of the United States roofing market [24] or about 34 million squares (1 square ¼ 100 ft2), assessing the overall environmental impact becomes critical. This study focuses on the environmental impact, quantifying greenhouse gas (GHG) emissions in particular, associated with the life cycle of an APP asphalt membrane roofing product. Properly assessing this impact is the first step towards reducing the environmental impact of buildings and achieving a green construction sector. To evaluate the GHG emissions of the asphalt roofing product, a life cycle assessment (LCA) was conducted. LCA is a powerful tool to assess the overall impact of a product throughout its life cycle. The

96

W. Vaz, J. Sheffield / Building and Environment 78 (2014) 95e102

LCA approach takes into account the various energies, processes, and raw materials throughout the entire life cycle of a product from design, manufacturing, and utilization to disposal and recycling. Once the impact of each stage is known, the stage with the greatest impact is identified. The impact on the environment is most effectively reduced by making changes to that stage before other stages. LCA has been adopted by several groups in recent years as a tool to evaluate the overall environmental impact of a product and to come up with ways to minimize the impact [5,13e15,25e27]. LCA has been used in several recent studies in evaluating the net carbon footprint or overall GHG emissions of a process or a product [28e32]. LCA aims to help organizations minimize their impact on the environment, whether it is by reducing energy consumption, land use, water consumption, or GHG emissions. This study is a first step in assessing the overall environmental impact of an APP asphalt roofing product and focuses on GHG emissions. The rest of the paper is organized as follows: Section 2 describes the LCA study including data collection, modeling, and assumptions; Section 3 presents the results including certain shortcomings in the LCA study, and Section 4 concludes the results and details certain suggestions to reduce the overall GHG emissions. 2. LCA approach Membrane roofs are a roofing solution for commercial flat roofs. The composition and structure of these roofs varies by location and manufacturer. In order to mitigate wear and tear, manufacturers use several modifiers and solvents that reflect ultraviolet light and resist contraction and expansion. A high thermal emissivity and reflectivity help reduce cooling costs and energy consumption. A product that is durable and free of leaks helps reduce maintenance costs. Modified asphalt roofs are a type of membrane roof and are of two kinds: atactic polypropylene (APP) or styrene butadiene styrene (SBS). APP roofs typically have the following structure: an insulation layer, a base sheet, and a cap sheet. The different pieces of roofing materials are connected by either a cold process, like cold adhesive application, or a hot process, like melting the edges together to form a seal. In today’s market, APP roofs like the one studied in this paper are rated for as much as 30 years by the manufacturer. A long life of 30 years or more definitely reduces the overall impact on the environment since the product has to be replaced less frequently by a customer, reducing the overall consumption of energy, labor, and raw materials. The life cycle data for the roofing material in this study were obtained from a manufacturing facility located in Kansas City, Missouri, United States. In order to manufacture such a product, the manufacturing facility uses a complex feedstock that changes daily. In addition, some of the product is recycled internally and used as raw material. The manufacturing facility uses electricity from the local grid, but also generates solar energy on site. Fifteen different kinds of products to suit general and specific applications are manufactured on site. In this study, only one of the fifteen products was studied. The asphalt roofing product studied is one of the flagship cap sheets produced at the facility, designed to suit a variety of applications. It had the largest share in the total production at 14.1% per year. The manufacturer has raw materials supplied from different parts of the United States and a couple of locations in Europe. Its customers are located all over the United States. Unlike a lot of products which tend to have a fixed formula, this particular manufacturer varies the mix of raw materials used for its roofs based on what is available from suppliers, the status of the production line, and what the application calls for keeping in view its goals of longevity in product life and reduction of environmental impact. The overall structure of the product remains the same, but the mix differs from batch to batch, which would imply that the

GHG emissions would also differ from batch to batch. To account for this, a strategy similar to the one use in a study by Mila i Canals et al. [33] was employed. In a study on the environmental impact of the production of various food products produced by Knorr, an average product mix was used rather than a using a specific mix given the thousands of different products produced by the company. A metaproduct approach was used where the different product types studied were representative of the company’s product range. In a similar manner, all the raw materials and energy resources that were consumed at the facility over an entire year, in this case 2010, were summed up. Based on the specific formula for the cap sheet studied and based on the total amount of the cap sheet produced, the appropriate fractions of raw materials and energy resources were included in LCA. The results were normalized per functional unit of the product, in this case one roll of cap sheet (100 ft2). An example of such normalization per unit of the product is a study on the global warming potential of milk on a New Zealand farm [34]. 2.1. Modeling GaBiÒ 4.4 Education, an LCA software package, was used for modeling [35]. GaBiÒ was chosen because it is not restricted by product type or region unlike certain other LCA software packages. This would allow for easy comparison and flexibility with future studies. Additionally, GaBiÒ uses Sankey diagrams and makes it easy to visualize the modeling process. A plan for each of the stages of the life cycle was created and linked to create the overall life cycle. Within each plan, representative processes from the Life Cycle Inventory (US LCI) database were linked using flows. Flows represented raw materials (asphalt, limestone, etc.), finished products (asphalt roofing product, etc.), or energy (diesel, electricity, etc.). Processes represented physical processes during any of the life cycle stages (packaging, transportation by truck, etc.). All other data used in the LCA were obtained from the manufacturer. 2.2. Data collection All relevant manufacturing data from the manufacturing facility were compiled into a series of spreadsheets. Raw materials were tracked from suppliers in order to account for resources consumed from raw material production and transportation. The energy and raw material consumption for each manufacturing process were recorded. An inventory of all the motors used in the manufacturing of the product was obtained from the manufacturing facility along with load factors and rated capacity. An inventory of customers all over the country was obtained from the manufacturing facility in order to account for resources consumed during installation and transportation. Finally, an inventory of landfills where the product was disposed of at the end of its life cycle was obtained as well. This helped to account for GHG emissions resulting from disposal. 2.3. Life cycle description The life cycle of the product was divided into three stages: manufacturing, utilization, and disposal. Fig. 1 shows the flow of raw materials and energy throughout the life cycle of the product. 2.3.1. Manufacturing Raw materials were shipped from suppliers located in different parts of Europe and the United States. The GaBiÒ software package automatically accounts for emissions and other environmental effects resulting from extracting, producing, and using the raw materials based on preexisting mixes in the US LCI database. The only input is the quantity. Table 1 shows the quantities that were input into the software package as flows to the appropriate processes


97

Fig. 1. Life cycle of an asphalt roofing product.

normalized per roll. To account for the environmental impact resulting from transportation by ship or truck, several transportation processes were defined. The total mass of the raw material being transported and the distance from the supplier to the manufacturing facility were the inputs. Based on the location of the supplier, the appropriate fuel blend was chosen from a list of fuel blends in the US LCI database. Table 1 also shows the distance between the supplier and manufacturer for each type of raw material along with the modes of transportation. The manufacturing of the product within the facility was broken up into major groups (preparation, cutting, etc.) with each major group comprising several smaller manufacturing processes. The different manufacturing processes were defined as unique processes in the life cycle model with each process having the appropriate flows going into it and coming out of it. To display the entire life cycle conveniently, only the major groups are shown in Fig. 1. The only environmental impact resulting from the different manufacturing processes themselves stemmed from use of energy for various heaters, motor, pumps, etc. Electricity and natural gas were input as flows to the manufacturing processes that required them. Electricity consumption was calculated by multiplying the load factor, the numbers of hours of operation, and the rated power. While most of the load factors were available, some had to be estimated based on the experience of the manufacturer. The appropriate electrical grid mix was selected based on the location of the manufacturing facility. It is important to note that the

Table 1 Raw materials and transportation for asphalt roofing manufacture (per roll). Raw material

Type

Amount (kg)

Mode

Distance (km)

Aluminum trihydrate Asphalt

Fire retardant Blend Blend Blend

8.89 15.58 0.61 1.43 0.79 1.75

Truck Truck Truck Truck Truck Truck Ship Truck Truck Truck e Truck Truck Truck

361 390 319 85 1267 225 6058 1917 1196 1267 e 507 1505 726

Polypropylene

Rework Limestone Mat (glass fibers) Sand

Blend Granules Base Packaging

0.95 1.11 0.64 14.97 1.59 1.59 49.90

environmental impact of construction of the manufacturing facility itself including all the equipment and the machinery and the financial costs due to labor and maintenance were not a part of this study. Some of the raw materials used are the result of recycling at other facilities. Since the recycling history from the suppliers was unavailable, all raw materials were treated as primary raw materials, i.e. produced from natural substances without any recycling as reported in the US LCI database. Some of the finished asphalt roofing product that may have been defective or lacking in some way was reprocessed and used as feedstock instead of being discarded or sold at a discount. The average amount of rework that comprises each roll is listed in Table 1. It may be noted that the manufacturing facility generated a significant amount of solar energy on site. This factored into the LCA study as a reduction in the overall electricity consumption. Table 2 shows all the energy resources involved in the life cycle normalized to represent one roll.

2.3.2. Utilization The utilization stage was the longest stage of the life cycle but it is also the simplest. The only two processes in this stage are packaging the finished product and installing it. The emissions resulting from transporting the finished product to different customers all over the United Stated was treated in a manner similar to the emissions resulting from transporting the raw materials to the manufacturing facility from the suppliers. Using customer data from the manufacturer, the appropriate masses transported and distances traveled were entered into the software package. Table 3 shows the product sales by location, including the number of rolls

Table 2 Energy resources used in asphalt roofing life cycle (per roll). Energy Manufacturing On-site solar energy Electricity Natural gas Diesel Fuel oil Propane Utilization Diesel Disposal Diesel

Type

Amount (kg)

Amount (kWh)

Power generation Production Production Transportation Transportation Transportation

e e 1.42 0.40 0.04 0.04

0.07 7.25 e e e e

Transportation

0.76

e

Transportation

0.03

e

98


delivered to the location and the distance between the location and the manufacturer. Certain cities had multiple customers. The distance to each customer was calculated and entered into the software. However, only average distance values for each city are shown in Table 3. This was done for the sake of convenience considering that there were 775 customers in 2010. The average distance between a customer and the manufacturer was calculated to be 1246.0 km. Negative values in Table 3 indicate the number of

Table 3 Product sales by location. Location

Number of rolls

AL Birmingham AR Siloam Springs AZ Phoenix Prescott CA Los Angeles Napa Van Nuys Walnut CO Denver Englewood FL Orlando Pompano Beach Tampa GA Athens Atlanta Lagrange IL Chicago Glenview Joliet Mapleton IN Carlisle Fort Wayne Frankfort Indianapolis Marion Mishawaka Sullivan West Lafayette KS El Dorado Emporia Kansas City Lawrence Wakarusa KY Louisville LA Hammond Harahan New Orleans MD Baltimore MI Madison Heights Utica MO Belton Independence Kansas City Saint Louis Springfield St. Louis

360 360 400 400 1640 1520 120 252 8 125 147 28 600 580 20 4080 920 500 2660 2640 340 1840 460 3430 620 120 690 2000 5400 440 1020 400 1760 120 40 1500 120 3951 160 300 2671 400 420 20 20 1729 560 20 1149 160 160 1426 1286 140 4901 172 1683 1486 1260 180 120

Distance (km) 1092.8 360.0 1931.2 1928.0 2598.4 2856.0 2616.0 2566.4 972.8 982.4 1979.2 2284.8 2006.4 1371.2 1276.8 1300.8 824.0 836.8 761.6 552.0 702.4 974.4 809.6 768.0 904.0 945.6 689.6 785.6 275.2 179.2 16.0 73.6 124.8 808.0 1259.2 1336.0 1345.6 1715.2 1249.6 1272.0 28.8 17.6 16.0 392.0 259.2 392.0

Location

Number of rolls

MS Pearl NC Raleigh NJ Avenel Moorestown Morristown Plainfield NM Hueferno Tohatchi Tohatchi NV Las Vegas NY Brooklyn OH Akron Cincinnati Cleveland Elyria Glouster Twinsburg Wickliffe OK Broken Arrow Tulsa PA Carnegie Monaca Pittsburgh Saltsburg State College University Park TN Chattanooga Knoxville Oak Ridge Red Bank TX Cedar Creek Fort Worth Houston Midlothian Round Rock Saginaw San Antonio Taylor VA Richmond WA Seattle Spokane Valley Sumner Vancouver WI Marinette Milwaukee

200 200 400 400 302 20 2 260 20 580 160 280 140 80 80 20 20 1540 280 60 440 140 120 400 100 520 40 480 1460 120 380 352 8 340 260 1680 200 260 1080 140 2180 440 440 300 140 20 200 900 140 1 1 2571 900 971 460 240 1040 180 860 43,563

Distance (km)

rolls that were returned. Individual variations in customer installation practices, if any, were neglected. For example, the distance between the point of delivery of the product by the manufacturer and the installation site was uniformly assumed to be so small when compared with the distance between the manufacturer and the point of delivery that it could be neglected. No energy was used during the installation process. Even though the product has a long life, maintenance is to be expected. Specific maintenance data for 2010 were unavailable and would have to be collected by the supplier from customers in the future. Minor on-site maintenance activities are not thought to significantly increase the GHG emissions. Major maintenance would entail replacing existing membranes with new membranes. Such data could easily be incorporated into the LCA study by adjust the total number of rolls consumed.

1054.4 1715.2 1899.2 1819.2 1872.0 1872.0 1459.2 1521.6 1521.6 2174.4 1934.4 1246.4 937.6 1273.6 1235.2 1164.8 1262.4 1292.8 424.0 427.2 1328.0 1345.6 1342.4 1398.4 1553.6 1553.6 1092.8 1168.0 1140.8 1102.4 1209.6 881.6 1214.4 873.6 1152.0 875.2 1307.2 1136.0 1705.6 2998.4 2521.6 3014.4 2886.4 1126.4 913.6

2.3.3. Disposal The disposal stage was tackled by breaking it up into two processes: uninstallation and disposal to landfills. No energy was used during the uninstallation process. Due to the long life of the product and the fact that most customers take care of this particular stage themselves, the landfill data had to be approximated for several customers where data were not available. This was done by finding the location of the landfill closest to the customer and assuming that the entire asphalt roofing product was disposed of at that site. It was also not known whether customers reuse any portion of the old roof or whether all of it was discarded. In this study, it was assumed that the entire old roof is discarded, so the values estimated are a “worst case” scenario. The masses transported and the distances traveled were entered into the software in order to account for GHG emissions resulting from transporting the product during the disposal stage. The average distance for disposal was calculated to be 49.8 km.

3. Results and discussion In order to generate results, GaBiÒ was used to perform a balance of flows, which is a function that takes into account all the relevant flows into the different processes in the life cycle and all the relevant flows out of those processes before computing the net environmental impact. The functional unit for LCA was one roll of asphalt roofing product (1 roll ¼ 1 square ¼ 100 ft2), which weighs 49.9 kg. The results have been normalized by taking the overall mass in each category and dividing by the total number of rolls produced in a year, which was 60,000 rolls in 2010. It may be noted that, although 60,000 rolls were produced, only 43,563 rolls were delivered as shown in Table 3.

3.1. Consumption of energy resources Fig. 2 shows a plot of the non-renewable energy resources that are consumed. Crude oil was the biggest consumed resource followed by natural gas. Crude oil consumption was estimated to be 27.2 kg per roll. This is to be expected as crude oil was a raw material in the production of several of the raw materials used to manufacture the asphalt roofing product. Additionally, crude oil is a precursor to diesel and fuel oil, which were both used in the transportation stage of the life cycle. Natural gas was the next highest after crude oil; it too was used in the production of raw materials and transportation. The use of coal was from the consumption of electricity. Electricity generation is heavily dependent on coal in most parts of the United States [36].


99

Fig. 4. Raw emissions to air by stage. Fig. 2. Consumption of non-renewable energy resources.

3.2. Overall emissions to air Fig. 3 shows the overall emissions to air. Carbon dioxide was the largest contributor to the overall emissions to air by almost two orders of magnitude. It comprised over 96% of the overall emissions. Carbon dioxide emissions were estimated to be 47.7 kg per roll. Given the raw materials and the energy-intensive production process, the fact that carbon dioxide was the leader in the emissions to air was a foregone conclusion. Methane was the next highest contributor to the overall emissions to air and comprised a little over 2% of the overall emissions. Over 83% of the methane emitted came from landfills. Nitrogen oxides and sulfur dioxide were released in significant quantities mostly due to the use of fossil fuels to provide energy. 3.3. Emissions to air by stage Fig. 4 shows the emissions to air by the stage of the life cycle without normalization. A trend can be observed: the manufacturing stage was responsible for a majority of the emissions with the disposal stage being next and the utilization stage being last in almost every category. During the course of the life cycle, the manufacturing stage was where all of the raw materials and most of the energy were consumed so its majority share in the emissions was not surprising. This is typical of this type of product unlike buildings as a whole, where most of the emissions stem from the utilization stage. Of the manufacturing processes modeled, the melting and cooling/saturation processes consumed the most electricity at about 42% of the total electricity consumption. These observations are expected to be common to other modified asphalt roofing products as well.

Fig. 3. Overall emissions to air.

Table 4 100-year global warming potential (GWP) for significant emissions in asphalt roof LCA [37]. Designation

Chemical formula

GWP

Carbon dioxide Methane Nitrous oxide

CO2 CH4 N2O

1 25 298

The utilization stage was chronologically the longest but had the lowest impact in terms of emissions. This is convenient because it is also the stage that is the most difficult to control given that the product is scattered across dozens of locations. It is also easier for the manufacturer to effect changes to the life cycle on site than for it to expect all its customers to do so. Methane emissions were higher in the disposal stage than in the manufacturing stage. This was to be expected considering the asphalt roofing product was disposed of in landfills. Another break in the trend was the emission of nitrogen oxides. The reason the utilization stage had higher emissions than the disposal stage was due to the transportation of the product over vast distances to customers located across the United States. While the product has to be transported to landfills at the end of its life as well, the distance between a landfill and a customer was, on average, a lot lower than the distance between a customer and the manufacturing facility. 3.4. Evaluation of greenhouse (GHG) emissions When evaluating the GHG emissions, the global warming potential (GWP) must be considered. It helps in comparing different categories of emissions by converting different types of emissions into a carbon dioxide equivalent (CO2-eq). The 100-year GWP values of the GHGs emitted in significant quantities (greater than 106 kg per roll) are listed in Table 4.

Fig. 5. Emissions to air by global warming potential.

100


Table 5 Comparison of GHG emissions for different types of roofing membranes (per 49.9 kg of roofing membrane). Product type

Atactic polypropylene (APP) modified asphalta

Styrene butadiene styrene (SBS) modified asphalt

Ethylene propylene diene monomer (EDPM) rubber

Polyvinyl chloride

Polyisocyanurate insulation

CO2 (kg) CH4 (kg) N2O (g) GHG Emissions (kg CO2-eq)

32.07 0.17 0.13 36.4

38.93 0.11 0.13 41.7

264.52 0.39 0.57 274.4

116.79 0.47 0.47 128.7

100.65 0.29 0.40 108.0

a

Emissions data were obtained from Ref. [21] except for APP roofing membranes, GWP was calculated using Table 4.

Fig. 5 shows a comparison of the GHG emissions by the stage of the life cycle after adjusting for GWP. The manufacturing stage was found to responsible for a majority of the GHG emissions at almost 69%. The overall GHG emissions were estimated to be 75.2 kg CO2eq per roll with carbon dioxide comprising 63.4% of GHG emissions and methane comprising almost 36%. It is clear that methane emissions are as much of a concern as carbon dioxide emissions. About 84% of the methane emissions were during the disposal stage primarily due to landfill disposal. Internal recycling of the product during the manufacturing process was accounted for. It was estimated that by recycling defective rolls rather than discarding them, the manufacturer reduces overall GHG emissions by up to 2%. The GHG emissions for different customers would be slightly different since the distance that the product is transported is different for each customer. Thus value 75.2 kg CO2-eq per roll is an average value for the overall GHG emissions for 2010. All of the remaining fourteen products have a composition and manufacturing process very similar to the one studied here, therefore significant differences in GHG emissions are not expected. A feature of LCA is that similar processes or products can be directly compared with one another, as demonstrated by Nilsson et al. in their comparison of the environmental impact of butter and margarine in different European markets [38]. The overall GWP of one roll of asphalt roofing product weighing 49.9 kg was estimated to be 75.2 kg per roll over a 30-year lifespan. As comparison, Table 5 shows a summary of the GHG emissions for four types of commercial roofing products studied in Ref. [21] and the roofing product studied in this paper. The results in Ref. [21] were for 1 kg of each roofing membrane; these were been scaled to represent 49.9 kg of roofing membrane for comparison. Additionally, since [21] presented only production LCA data, the GHG emissions appearing in Table 5 under APP modified asphalt roofing membranes are only for the manufacturing stage of the life cycle. Modified asphalt roofs have similar GHG emissions in the manufacturing/production stage, with SBS being a little higher. This is most probably because crude oil comprises a larger fraction of the raw materials, almost 75% as reported in Ref. [21], compared to under 45% for APP. Looking at other types of roofing products [14], quantifies the GWP of 1 kg of ceramic roof tiles, concrete roof titles, and fiber cement roof slate as 0.406, 0.270, and 1.392 kg CO2-eq respectively (20.3, 13.5, and 69.5 kg CO2eq when considering 49.9 kg). It must be noted that these figures are subject to the assumptions and conditions of the LCA study in Ref. [14]. Furthermore, the carbon dioxide emissions per roll of the same cap sheet produced by the parent company in Europe, which uses a similar production process, were found to be within 10% of the product in this study. 3.5. Uncertainty The results presented in this paper are subject to certain assumptions and limitations that have been outlined. This section examines these assumptions and limitations in order to estimate their impact on the results. Due to the nature of the manufacturing

process, an average approach to determine GHG emissions had to be adopted. This approach does not take individual differences from batch to batch into account. These differences may be difficult to ascertain and quantify leading to some uncertainty in the results presented in this study. Some of them are as follows. There are several grid mixes defined in the US LCI database to account for geographical variations in electricity generation sources. While the most suitable regional grid mix was chosen for this study, certain differences between the local and the regional grid mixes are to be expected. Transportation fuels such as diesel and fuel oil were input based on origin: trucks transporting raw materials were assumed to have fueled up in the supplier’s city and trucks transporting the asphalt roofing product were assumed to have fueled up in the manufacturing facility’s city rather than in the customer’s city. This may not reflect the actual fueling habits of truck drivers for 2010. The GHG emissions from the production of the raw materials used in this study were included solely based on the US LCI database, which may not reflect individual supplier practices in terms of manufacturing and recycling throughout the year. The load factors used to calculate electricity consumption were not available for each and every process and so estimates based on experience were used in some cases. There was a difference of about 812,942 kWh of electricity consumption between billed consumption and calculated consumption as provided by the manufacturer. The most probable reason for this discrepancy is that the calculations for electricity consumption done by the manufacturer were very conservative. Since the billed consumption was a summed value and was lower than the calculated consumption (for which a breakdown was available for each process), the calculated consumption was used. If the billed consumption is used, the estimate of overall GHG emissions would be reduced by 1.3 kg CO2-eq per roll or approximately 1.8%. The US LCI database estimates the amount of electricity produced from landfill disposal. This electricity was not factored into the LCA study since the manufacturing facility did not directly benefit from it. If the electricity generated from landfills is factored in, the overall electricity consumption would be reduced by 6.32 kWh per roll and the estimate of overall GHG emissions would be reduced by 5.6 kg CO2-eq per roll or approximately 7.5%. Landfill emissions data in LCA databases tend to be generic. That is, while modeling disposal to a landfill, only the amount of waste is specified. The type of waste is not considered. It is treated as generic landfill waste for which a preexisting mix of emissions has been calculated. Some software packages such as GaBiÒ classify waste into categories such as construction material, paper, etc., which is going one step further. In general, GHG emissions estimated by LC software packages due to landfill disposal may not reflect GHG emissions of the product being studied. Better data on landfills are needed for more accurate results. 3.6. Suggestions to reduce emissions Every stage of the life cycle studied was heavily dependent on fossil fuels especially considering the raw materials used in


manufacturing. Diesel and natural gas are cleaner fuels than coal, but the indirect use of coal was unavoidable due to the fact that electrical generation in the United States is heavily dependent on coal. The following suggestions were made to the manufacturer to help to reduce its GHG emissions. 1. The manufacturing stage accounted for 70e80% of carbon dioxide emissions, mostly due to fossil fuel-based raw materials. Finding substitutes for hydrocarbon-based raw materials such as asphalt and polypropylene must be a long-term goal. 2. Electricity consumption accounted for 20e25% of carbon dioxide emissions in the manufacturing stage. Monitoring consumption and revising load factors so that billed consumption and calculated consumption match must be done in order to ensure the accuracy of LCA. 3. Diesel consumption for transportation accounted for 7e8% of carbon dioxide emissions in the manufacturing stage. Finding raw materials closer to the manufacturing facility will reduce the number of miles raw materials are transported using fossil fuels thereby reducing emissions. 4. About 0.07 kWh per roll of solar energy was produced on site (in 2010). Increasing the production of on-site solar energy would reduce net overall electricity consumption. For every 1.0 kWh of solar energy produced on site, about 0.7 kg of carbon dioxide generation is avoided [39]. 5. Even though the utilization stage only accounted for 5e6% of carbon dioxide emissions, most of these emissions come from transportation. Using more efficient methods of transportation such as hybrid trucks or railroad instead of conventional trucks would reduce the consumption of fossil fuels. 6. The disposal stage accounted for 20e25% of carbon dioxide emissions and 80e85% of methane emissions, most of which was produced when the product was disposed of in landfills. It has been shown that methane is as much of a concern as carbon dioxide. Recycling the used product to obtain raw materials would result in a further reduction of carbon dioxide and methane emissions. If not raw materials, then roofing scrap can be recycled in other ways like repairing pot holes [40]. 4. Conclusions The LCA study presented in this paper was done to assess the overall GHG emissions of an asphalt roofing product. It is seen as the first step towards assessing the overall environmental impact of asphalt roofing material. The produce studied was the flagship cap sheet produced by a manufacturing facility located in the United States. The overall GHG emissions were estimated to be 75.2 kg CO2-eq per roll. Carbon dioxide was estimated to constitute 64.3% of the GHG emissions with methane being next at almost 36%. The manufacturing stage was found to be responsible for a majority of the GHG emissions. However, almost 84% of the methane emissions were during the disposal stage. The influence of LCA assumptions and the uncertainty in the results were presented and certain observations and shortcomings of the LCA study were discussed. The length of the study was one year. The manufacturer’s goals are longevity in product life and minimization of environmental impact. It aims to certify its products according to the ISO 14000 series [41] and to back its claims that its products are greener than competing products. The use of a complex feedstock that changed from batch to batch necessitated the use of an average approach to LCA where all the raw materials and energy resources involved in the life cycle over an entire year were accounted for and the calculated emissions were normalized per functional unit, in this case one roll of asphalt roofing product. Normalization allows easy interpretation of results and direct comparison across different

101

brands and different products. This would facilitate an effort into green labeling or creating an environmental declarations sheet for products [42]. When labeling a product as green, it is necessary to ensure the LCA data used to evaluate the product are not lacking in anyway, as pointed out in a study on green labeling of products [42]. More research is needed to solve the issues discussed in Sections 2 and 3 including LCA databases that are lacking in certain cases and a lack of experience by the manufacturer in tracking its own manufacturing, utilization, and disposal data. In the future, this study would have to include all fifteen products produced at the manufacturing facility. Using an average approach with normalization of the results would allow direct comparison between the different products. It would allow the manufacturer to learn how exactly differences in raw materials and manufacturing processes affect the overall GHG emissions of different products. Acknowledgment The authors are grateful to PE International GmBH for providing GaBiÒ 4.4 Education, which was used for LCA. References [1] HM Government. Climate change act 2008: Elizabeth II. Chapter 27. London: The Stationery Office; 2008. [2] CLG (Department for Communities and Local Government). Building a greener future: policy statement. London: CLG; 2007. [3] Executive Order No. 13,514. Federal leadership in environmental, energy, and economic performance. Fed. Reg October 5, 2009;74(194). [4] Executive Order No. B-18e12. State of California; April 25, 2012. [5] Bianchini F, Hewage K. Probabilistic social cost-benefit analysis for green roofs: a lifecycle approach. Build Environ 2012;58:152e62. [6] Xu T, Sathaye J, Akbari H, Garg V, Tetali S. Quantifying the direct benefits of cool roofs in an urban setting: reduced cooling energy use and lowered greenhouse gas emissions. Build Environ 2012;48:1e6. [7] Niachou A, Papakonstantinou K, Santamouris M, Tsangrassoulis A, Mihalakakou G. Analysis of the green roof thermal properties and investigation of its energy performance. Energy Build 2007;33(7):719e29. [8] Getter KL, Rowe DB, Robertson GP, Cregg BM, Andresen JA. Carbon sequestration potential of extensive green roofs. Environ Sci Technol 2009;43(19): 7564e70. [9] Levinson R, Akbari H, Reilly JC. Cooler tile-roofed buildings with nearinfrared-reflective non-white coatings. Build Environ 2007;42(7):2591e605. [10] Saber HH, Swinton MC, Kalinger P, Paroli RM. Long-term hygrothermal performance of white and black roofs in North American climates. Build Environ 2012;50:141e54. [11] Fioretti R, Palla A, Lanza LG, Principi P. Green roof energy and water related performance in the Mediterranean climate. Build Environ 2010;45(8): 1890e904. [12] Al-Sanea SA. Thermal performance of building roof elements. Build Environ 2002;37(7):665e75. [13] Bianchini F, Hewage K. How “green” are the green roofs? Lifecycle analysis of green roof materials. Build Environ 2012;48:57e65. [14] Bribián IZ, Capilla AV, Usón AA. Life cycle assessment of building materials: comparative analysis of energy and environmental impacts and evaluation of the eco-efficiency improvement potential. Build Environ 2011;46:1133e40. [15] Saiz S, Kennedy C, Bass B, Pressnail K. Comparative life cycle assessment of standard and green roofs. Environ Sci Technol 2006;40(13):4312e6. [16] Kosareo L, Robert R. Comparative environmental life cycle assessment of green roofs. Build Environ 2007;42(7):2606e13. [17] Gustavsson L, Sathre R. Variability in energy and carbon dioxide balances of wood and concrete building materials. Build Environ 2006;41(7):940e51. [18] Traverso M, Rizzo G, Finkbeiner M. Environmental performance of building materials: life cycle assessment of a typical Sicilian marble. Int J Life Cycle Assess 2010;15:104e14. [19] Franklin Associates. A life cycle inventory for road and roofing asphalt. AthenaÔ Sustainable Materials Institute; March 2001. retrieved on 31st March 2014 from: http://www.athenasmi.org/resources/publications/. [20] Venta GJ, Nisbet M. Life cycle analysis of residential roofing products. AthenaÔ Sustainable Materials Institute; March 2000. retrieved on 31st March 2014 from: http://www.athenasmi.org/resources/publications/. [21] Franklin Associates. A life cycle inventory of selected commercial roofing products. AthenaÔ Sustainable Materials Institute; April 2001. retrieved on 31st March 2014 from: http://www.athenasmi.org/resources/publications/. [22] Blom I, Itard L, Meijer A. LCA-based environmental assessment of the use and maintenance of heating and ventilation systems in Dutch dwellings. Build Environ 2010;45:2363e72.

102


[23] Pajchrowskia G, Noskowiaka A, Lewandowskab A, Strykowski W. Materials composition or energy characteristic e what is more important in environmental life cycle of buildings? Build Environ 2014;72:15e27. [24] Dodson M. Low-slope market. Western Roofing Magazine; 2007. [25] Ferrari AM, Barbieri L, Folloni B, Neri P. Life cycle assessment of advertising folders. Int J Life Cycle Assess 2012;17:625e34. [26] Khoo HH, Tan RBH. Environmental impacts of conventional plastic and biobased carrier bags (Part 2). Int J Life Cycle Assess 2010;15:338e45. [27] Wernet G, Conradt S, Isenring HP, Jimenez-Gonzalez C, Hungerbuhler K. Life cycle assessment of fine chemical production: a case study of pharmaceutical synthesis. Int J Life Cycle Assess 2010;15:294e303. [28] Gonzalez-Garcia S, Silva FJ, Moreira MT, Pascual RC, Lozano RG, Gabarrell X, et al. Combined application of LCA and eco-design for the sustainable production of wood boxes for wine bottles storage. Int J Life Cycle Assess 2011;16:224e37. [29] Dominguez-Ramos A, Held M, Aldaco R, Fischer M, Irabien A. Preliminary CO2 emissions from energy supplying systems: photovoltaic systems and conventional grid within Spanish frame conditions. Int J Life Cycle Assess 2010;15:557e66. [30] Nakatani J, Fujii M, Moriguchi Y, Hirao M. Life cycle assessment of domestic and transboundary recycling of post-consumer PET bottles. Int J Life Cycle Assess 2010;15:590e7. [31] Suebsiri J, Wilson M. A model of carbon capture and storage with demonstration of global warming potential and fossil fuel resource use efficiency. Energy Procedia 2011;4:2465e9. [32] Thomas NJ, Chang NB, Qi C. Preliminary assessment in global warming potential of leading contributory gases from a 40-in LCD flat-screen television. Int J Life Cycle Assess 2011;17:96e104. [33] Mila i Canals L, Sim S, Garcia-Suarez T, Neuer G, Herstein K, Kerr C, et al. Estimating the greenhouse gas footprint of Knorr. Int J Life Cycle Assess 2011;16:50e8.

[34] Basset-Mens C, Kelliher F, Ledgard S, Cox N. Uncertainty of global warming potential for milk production on a New Zealand farm and implications for decision making. Int J Life Cycle Assess 2009;14:630e8. [35] PE International. GaBiÒ 4.4 Education. Software and database for life cycle engineering. PE International GmBH and LBP University of Stuttgart; 2011. [36] Itten R, Frischknecht R, Stucki M. Life cycle inventories of electricity mixes and grid. Uster, Switzerland: ESU Services Ltd; 2013. [37] Forster P, Ramaswamy V, Artaxo P, Berntsen T, Betts R, Fahey DW, et al. Changes in atmospheric constituents and in radiative forcing. Climate change 2007: The physical science basis. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL, editors. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UK and New York, NY, USA: Cambridge University Press; 2007. [38] Nilsson K, Flysjo A, Davis J, Sim S, Unger N, Bell S. Comparative life cycle assessment of margarine and butter consumed in UK, Germany, and France. Int J Life Cycle Assess 2010;15:916e28. [39] Environmental Protection Agency. Clean energy: calculations and references. Climate change 2012; retrieved from: http://www.epa.gov/cleanenergy/ energy-resources/refs.html. [40] Rabasca L. Recycling facility of the month: reclaim finds success in recycling roofs. Waste Age 1994;25(5):65e8. [41] International Organization for Standardization. ISO 14000 e Environmental management. Management system standards 2006; retrieved on 12th January 2014 from: http://www.iso.org/iso/home/standards/management-standards/ iso14000.htm. [42] Rajagopalan N, Bilec MM, Landis AE. Life cycle assessment evaluation of green product labeling systems for residential construction. Int J Life Cycle Assess 2012;17:753e63.