Emissions of chemicals from the economy-wide stock of plastic material - a first model iteration for Sweden
Paper in proceeding, 2011
Estimating the size of the problem with release, fate, exposure and effects from the human use of chemical substances of materials and consumer products is daunting
More than 100,000 chemical substances are in commercial use and a reasonable description of their existence in, and release from, plastic polymers, glues, paints, fibres, lubricants etc. comprise a big challenge.
In this study, a generic emission model has been developed and applied to estimate emissions of a set of organic chemicals from products. The scope of the estimate is emissions from products containing plastic materials during their average lifetime within the geographical boundaries of Sweden. The customs Combined Nomenclature has been used to divide the products into categories for which the chemical composition, surface area, thickness and accumulated stock in society has been described and estimated using several approaches for approximation in cases where data have been lacking. For this, information from national trade statistics as well as lifecycle assessments and building product declarations was used
Anti-oxidants, flame retardants and plasticisers are among the most interesting use categories of additives that are emitted in significant quantities. Thus for anti-oxidants in plastic material in the economy-wide product stock in Sweden as an example, the emissions in the service-life were estimated to be almost 500 tons/year divided on 32 individual chemical organic substances, stemming from a stock of anti-oxidants of 84000 tons, contained in a stock of plastic material of 24 million tons.
Until now, the method has only been used to estimate emissions of additives from plastic materials, but it is believed to also be applicable to other materials. However, the uncertainties in the estimations are currently large, which is partly due to model uncertainties but to a large extent also due to uncertainties in the input data.