Goal and scope
The goal of this study is to assess and compare the life cycle environmental impacts associated with polymer (+PLUSPAK™) and glass bottles used for delivering contrast media for X-ray procedures. A functional unit is defined to ensure that any comparisons are made on a functionally equivalent basis; in this study, the functional unit is defined as the packaging of contrast media required to deliver one dose of 96 mL to a patient for an X-ray procedure. The study examines both bottle types, from cradle-to-grave, by assessing the impacts associated with raw material extraction, production, distribution, use, and disposal. Several bottle sizes are assessed including 50, 100, 200, and 500 mL. The 100-mL bottle size with a dose of 96 mL is selected as the primary basis of comparison.
Life cycle inventory analysis
The +PLUSPAK™ polymer bottle consists of a pharmaceutical-grade polypropylene vial, rubber stopper, and polypropylene cap (all virgin materials); the glass bottle consists of a glass vial (containing 10–30 % glass cullet), rubber stopper, and a crimp seal made of aluminum and polypropylene.
Primary data and product information (including material types and quantities, manufacturing processes and associated parameters, supply chain, and distribution transport logistics including breakage and product loss) are provided by GE Healthcare and its suppliers. Secondary data (including cradle-to-gate materials extraction and refining and end-of-life treatment) are obtained from the ecoinvent life cycle inventory database (ecoinvent Centre 2010) and from the literature. Process data for autoclaving of the glass bottles at end-of-life are purchased (Environmental Clarity 2012). All material, energy, and environmental flows are quantified for each step in the product life cycle in accordance with the functional unit definition. Life cycle inventories for supply chain, assembly and filling, transport, and use are described in the Electronic Supplementary Material along with discussion of cutoff criteria, excluded data, and assumptions.
The use and disposal of GE Healthcare’s contrast media packaging takes place in hospitals across the globe. This study examines a wide range of end-of-life scenarios because of the various treatment methods used in different countries as well as within countries.
The default end-of-life treatments in this study are based on the most likely disposal methods for plastic and glass contrast media bottles in the USA. The polymer bottles are likely to be treated as municipal waste (Ed Krisiunas, President, WNWN International, personal communication, 2012), and the study assumes an average split of 80 % landfill and 20 % incineration (US EPA 2012). The glass bottles are likely to be disposed of in a sharps container (Morrison and Odle 2007; Blackburn and Hawley 2006) and would therefore be treated by autoclaving and landfilling.
Additional treatment scenarios are considered (Table 1). In the European Union, most types of waste are stratified by the European Waste Catalog (European Waste Catalogue and Hazardous Waste List 2002). Some of the likely treatments for the sharps waste and other potentially infectious waste include (1) autoclaving and incineration; (2) incineration; and (3) pre-shred, autoclave, and incineration.
The study also considers recycling of materials at end-of-life. Two approaches are used to allocate the environmental impacts associated with recycling. The cutoff approach (Frischknecht 2010; Frischknecht et al. 2007) assigns environmental burdens of material production (raw material extraction, processing, etc.) to the first life of a material and assigns material refurbishment impacts (collection and scrap processing) to its second life. The sensitivity of the results to choice of allocation method is tested by using the market-based approach for system expansion (Weidema 2003), in which the level of material utilization in the recycled materials market determines how the environmental burdens and benefits of recycling are allocated.
The cutoff allocation approach is also applied to landfill (no credit for landfill gas, if any) and incineration (no credit for heat recovery or electricity generation, if any). The market-based allocation approach is applied to incineration of the bottles by applying a credit for the heat and electricity generated.
Life cycle impact assessment
Since this study is directed toward health-care professionals, human health impacts are of primary importance. The health-care industry is also keenly aware of energy and solid waste issues. To address these perspectives, this study uses human health midpoint impact categories and ecosystem quality and resources end point damage categories from the internationally accepted method ReCiPe (H) (Goedkoop et al. 2009). The climate change impact category within the ReCiPe Midpoint (H) method includes all greenhouse gases specified in the Kyoto Protocol using global warming potentials from the IPCC Fourth Assessment Report with a 100-year time horizon (IPCC 2007). The cumulative energy demand (CED) method (Jungbluth and Frischknecht 2010) is used to evaluate energy demand associated with a product’s life cycle. The IMPACT 2002+ (Jolliet et al. 2003) and USEtox (Rosenbaum et al. 2008) methods are used to assess the sensitivity of the results to choice of impact assessment methodology.
Sensitivity and uncertainty analyses
Sensitivity analyses are performed to assess the effect of different bottle sizes (50, 100, 200, 500 mL), electricity grids (Europe, China, US), glass bottle recycled content (10–60 % cullet), polymer scrap rate (5–50 %), mode of distribution transport (air vs. ocean freight), different contrast media solutions (X-ray vs. MRI diagnostic agents), secondary packaging and shipping container (single-pack vs. multi-pack), and choice of impact assessment method (ReCiPe H vs. IMPACT 2002+ vs. USEtox).
Uncertainty analysis is performed to understand how data quality affects the reliability and robustness of the study results for the comparison of the 100-mL glass and +PLUSPAK™ polymer bottles using the default end-of-life treatments. Flows and parameters within the model are input as probabilistic values, typically using lognormal distributions based on the pedigree matrix approach (Weidema and Wesnæs 1996). Monte Carlo simulations are performed (1,000 runs), and a distribution is plotted indicating the percentage of runs in which one bottle type exhibits lower impact compared to the other bottle type in each impact category.